- Название: New approaches for detecting alcohol abuse
- Автор: EPSHP
Novel Use of Biomarkers and Their Combinations
for Detecting Excessive Drinking
To be presented, with the permission of
the Faculty of Medicine of the University of Tampere,
for public discussion in the auditorium of
Mediwest Health Technology Center, Koskenalantie 16, Seinäjoki,
on February 23rd, 2007, at 12 o’clock.
(Simultaneous video conference connection
in the small auditorium of Building K,
Medical School of the University of Tampere,
Teiskontie 35, Tampere)
U N I V E R S I T Y O F TA M P E R E
University of Tampere, Medical School
Seinäjoki Central Hospital, Department of Laboratory Medicine and Medical Research Unit
Professor Onni Niemelä
University of Tampere
Docent Kari Pulkki
University of Turku
Professor Kaija Seppä
University of Tampere
P.O. Box 617
33014 University of Tampere
Tel. +358 3 3551 6055
Fax +358 3 3551 7685
Cover design by
Acta Universitatis Tamperensis 1210
ISBN 978-951-44-6856-8 (print)
Tampereen Yliopistopaino Oy – Juvenes Print
Acta Electronica Universitatis Tamperensis 594
ISBN 978-951-44-6857-5 (pdf )
To my family
Excessive alcohol consumption and consequent medical disorders create a major burden for modern
health care. In addition to medical and social problems, excessive drinking causes considerable
strains on the national economy. In order to improve the diagnosis and treatment of patients suffering from ethanol-related health problems, reliable and accurate methods for recognizing excessive
alcohol consumption in its early phase need to be developed. Currently, excessive drinkers tend to
escape detection, which may lead to delays in intervention.
The present study was set out to develop new approaches for detecting excessive drinking based
on conventional and new laboratory tests and their combinations and to address the relationships
between such markers and alcoholic liver disease. Conventional laboratory markers of excessive
drinking (GGT, CDT, MCV, AST, ALT), a mathematically formulated combination of GGT and
CDT, and autoimmune responses to proteins modified with acetaldehyde, the first metabolite of
ethanol, were measured in alcoholics with or without liver disease, moderate drinkers and abstainers. Cytokine profiles of subjects were also studied in order to clarify the associations between alcohol intake and pathogenesis of alcoholic liver disease.
The results show that even moderate drinking may increase levels of gamma-glutamyl transferase (GGT). When GGT was combined logarithmically with carbohydrate-deficient transferrin
(CDT), the diagnostic performance of the combination GGT-CDT was found to markedly exceed
that of the traditional markers, reaching a sensitivity of 90 % whereas the sensitivities of its parent
components remained at 63 % (CDT) and 58 % (GGT).
Alcohol consumption was also found to induce alterations in the immune system. An association
between cytokine levels and alcohol use was observed, with most evident alterations in alcoholics
with liver disease. Pro-inflammatory cytokines (IL-2, IL-6, IL-8, TNF-α) were found to increase in
alcoholic liver disease whereas the levels of anti-inflammatory cytokines, particularly TGF-β1
showed a slight decrease.
Acetaldehyde adducts are formed when acetaldehyde reacts with proteins and cellular constituents. Antibodies directed against acetaldehyde-modified proteins were found in the circulation of
alcoholic patients. The highest anti-adduct IgA and IgG titres occurred in patients with alcoholic
liver disease, while specific class IgM antibodies were most abundant in alcoholics without liver
disease. The possible usefulness of anti-adduct IgA as a marker of excessive drinking was subsequently studied in alcoholics without liver disease, and the mean anti-adduct IgA levels were significantly higher than those in the moderate drinkers or abstainers (p < 0.001), while the difference
between the moderate drinkers and abstainers was also significant (p < 0.05). Mean daily ethanol
consumption during the previous month was found to correlate significantly with anti-adduct IgA
levels. Based on these findings, anti-adduct IgA antibodies could serve as markers of alcohol con-
sumption. A combination of the anti-adduct IgA and CDT results showed improved diagnostic performance for this marker (IgA-CDT) as compared to traditional ones.
In the light of these results, it may be suggested that GGT, CDT and IgAs against acetaldehydederived epitopes either alone or combined by means of mathematical equations can serve as useful
markers of alcohol consumption. The elevation of GGT in moderate drinkers as compared with abstainers may affect its reference ranges, since they are usually calculated from general population in
which the proportion of abstainers is decreasing. The results also indicate that the antibody responses and alterations in cytokine balance may contribute to the pathogenesis and progression of
alcoholic liver disease in humans.
LIST OF ORIGINAL PUBLICATIONS ...................................................................13
2. REVIEW OF THE LITERATURE........................................................................16
2.1. Ethanol: past and present................................................................................................. 16
2.2. Main features of ethanol metabolism .............................................................................. 16
2.3. Effects of ethanol on health ............................................................................................. 17
2.3.1. Ethanol and the liver ........................................................................................... 18
2.3.2. Ethanol and extrahepatic tissues ......................................................................... 19
2.3.3. Ethanol and the immune system ......................................................................... 20
2.3.4. Ethanol, psychiatric disorders and accidents ...................................................... 20
2.3.5. Suggested positive effects of ethanol.................................................................. 21
2.4. Assessing alcohol consumption....................................................................................... 22
2.4.1. Amount and pattern of drinking.......................................................................... 22
2.4.2. Self-reported alcohol consumption ..................................................................... 22
2.4.3. Laboratory markers of alcohol consumption ...................................................... 23
2.4.4. Laboratory markers of ethanol-induced tissue injury ......................................... 25
2.4.5. Gender issues ...................................................................................................... 25
2.5. Routinely used biomarkers of ethanol consumption ....................................................... 25
2.5.1. Ethanol concentration in blood, breath or urine.................................................. 25
2.5.2. Gamma-glutamyl transferase .............................................................................. 25
2.5.3. Carbohydrate-deficient transferrin ...................................................................... 26
2.5.4. Mean corpuscular volume ................................................................................... 27
2.5.5. Aminotransferases............................................................................................... 27
2.6. New biomarkers of ethanol consumption........................................................................ 27
2.6.1. Acetaldehyde adducts and anti-adduct antibodies .............................................. 27
2.6.2. 5-Hydroxytryptophol .......................................................................................... 28
2.6.3. Ethyl glucuronide ................................................................................................ 28
2.6.4. Fatty acid ethyl esters.......................................................................................... 28
2.6.5. Phosphatidylethanol ............................................................................................ 29
2.6.6. Sialic acid ............................................................................................................ 29
2.6.7. Other postulated markers .................................................................................... 29
2.7. Marker combinations....................................................................................................... 30
2.8. Laboratory markers for the follow-up of alcoholics........................................................ 31
2.9. Immune responses related to alcohol consumption......................................................... 31
2.9.1. Immune responses to ethanol metabolites........................................................... 31
2.9.2. Alcohol-induced cytokine responses................................................................... 33
3. AIMS OF THE PRESENT RESEARCH...............................................................34
4. MATERIALS AND METHODS ...........................................................................35
4.1. Subjects............................................................................................................................ 35
4.2. Measurements of laboratory markers (I-II, IV) ............................................................... 36
4.3. Determination of marker combinations (II, IV) .............................................................. 37
4.3.1. GGT-CDT (II)..................................................................................................... 37
4.3.2. IgA-CDT ............................................................................................................. 37
4.4. Measurement of antibodies against acetaldehyde adducts (III-IV) ................................. 37
4.5. Serum cytokines .............................................................................................................. 38
4.6. Statistical methods........................................................................................................... 38
5.1. Influence of drinking levels on markers (I, II, IV) .......................................................... 39
5.2. Clinical characteristics of marker combinations (II, IV)................................................. 39
5.2.1. GGT-CDT (II)..................................................................................................... 39
5.2.2. IgA-CDT (IV) ..................................................................................................... 40
5.3. Antibody and cytokine responses in alcoholics (III–IV)................................................. 41
5.3.1. Alcohol-related changes in cytokine and antibody production (III-IV).............. 41
5.3.2. Anti-adduct IgAs as a marker of alcohol consumption (IV)............................... 41
5.4. Markers in the follow-up of alcoholics (II-IV)................................................................ 42
6.1. Influence of moderate drinking on marker levels (I-IV) ................................................. 43
6.2. Marker combinations (II, IV) .......................................................................................... 44
6.3. Antibody and cytokine responses in alcohol abusers (III-IV)......................................... 45
6.3.1. Changes in antibody and cytokine production attributable to alcohol and liver
disease (III-IV) .............................................................................................................. 45
6.3.2. Anti-adduct IgAs as a marker of alcohol consumption....................................... 47
6.4. Follow-up studies ............................................................................................................ 47
6.5. Possible limitations of this study..................................................................................... 48
6.6. Future considerations....................................................................................................... 48
7. CONCLUSIONS ....................................................................................................49
IgA, IgG, IgM
Area under the curve
Alcohol Use Disorders Identification Test
Body mass index
Cut down, Annoyed, Guilty, Eye-opener (acronym)
Combined clinical and laboratory index
CDT as a percentage of total transferrin
Coronary heart disease
Combined morphological index
Cytochrome P450 IIE1
Diagnostic and Statistical Manual of Mental Disorders. Fourth Edition.
Early onset alcoholics
Fetal alcohol effects
Fatty acid ethyl esters
Fetal alcohol syndrome
High density lipoprotein -cholesterol
Immunoglobulin A, G, M
Late onset alcoholics
Michigan Alcoholism Screening Test
Mean corpuscular volume
Microsomal ethanol oxidizing system
Negative predictive value
Aminoterminal propeptide of type I collagen
Aminoterminal propeptide of type III collagen
Positive predictive value
Sialic acid index of apolipoprotein J
Transforming growth factor-β1
Timeline follow-back -method
Tumor necrosis factor-α
List of original publications
Hietala J, Puukka K, Koivisto H, Anttila P, Niemelä O (2005) Serum gamma-glutamyl
transferase in alcoholics, moderate drinkers and abstainers: effect on GT reference intervals at population level. Alcohol Alcohol 40: 511–514.
Hietala J, Koivisto H, Anttila P, Niemelä O (2006) Comparison of the combined
marker GGT-CDT and the conventional laboratory markers of alcohol abuse in heavy
drinkers, moderate drinkers and abstainers. Alcohol Alcohol 41: 528–533.
Latvala J, Hietala J, Koivisto H, Järvi K, Anttila P, Niemelä O (2005) Immune responses to ethanol metabolites and cytokine profiles differentiate alcoholics with or
without liver disease. Am J Gastroenterol 100: 1303–1310.
Hietala J, Koivisto H, Latvala J, Anttila P, Niemelä O (2006) IgAs against acetaldehyde-modified red cell protein as a marker of ethanol consumption in male alcoholic
subjects, moderate drinkers, and abstainers. Alcohol Clin Exp Res 30: 1693–1698.
The original articles are referred to in the text with the above Roman numerals.
Excessive alcohol consumption and consequent medical disorders are considerable problems in
many Western countries (Österberg 2006, Blocker, Jr. 2006). In addition to medical and social
problems at the individual level, excessive drinking causes significant problems for the national
economy. Resources should therefore be focused on reducing the prevalence of alcoholism through
more effective diagnosis and early intervention (Anderson 1993, Sharpe 2001, Fleming et al. 2002,
Latt and Saunders 2002, Niemelä 2002, Rehm et al. 2006). It has been estimated that only 20–50 %
of patients with alcoholism are actually identified in health care and thus more reliable and accurate
methods of recognizing excessive alcohol consumption in clinical work are urgently needed (Reid
et al. 1986, Moore et al. 1989, Sharpe 2001).
The diagnosis of excessive alcohol consumption is often based on patients' own reports and answers to questionnaires. This approach suffers from a lack of reliability, because patients are usually
unwilling to admit excessive drinking (Rosman and Lieber 1994, Sharpe 2001, Niemelä 2002).
Laboratory markers have been integrated into the diagnostics and may lead to a considerable improvement in detecting excessive drinking (Lieber 1995, Niemelä 2002). The main advantage of
laboratory markers is their objective nature. The results of laboratory tests revealing excessive alcohol consumption and possible tissue injury could motivate patients more than a verbal report by a
clinician (Allen et al. 1992, Rosman and Lieber 1994). It has also been suggested that a better understanding of the physiological effects of ethanol and the mechanisms by which alcohol exerts its
effects on different tissues would lead to improved diagnostics and treatment of alcohol use disorders.
This work was aimed at investigating the usefulness of various laboratory markers and their
combinations for detecting excessive drinking. The effect of abstaining and moderate drinking on
levels of these markers was studied. The aims were also to clarify the effects of alcohol on specific
antibody and cytokine responses and to investigate the possibility of using such responses as markers of alcohol consumption and tissue injury.
2. Review of the literature
2.1. Ethanol: past and present
Alcohol has been used in human societies from the beginning of recorded history. The earliest alcoholic beverages were fermented drinks, and distilled alcohol products emerged about 3000 years
ago. Alcohol has been used over the centuries for both medicinal and refreshment purposes (Millikan 1999), and the production of alcoholic beverages has gradually become industrialized and has
turned into a worldwide business (Walsh 1997, Room 1997). After the temperance movements of
late 19th and early 20th centuries, alcohol consumption has become more accepted and is no longer
viewed as a threat to all individuals (Blocker, Jr. 2006). Consequently, consumption volumes have
grown continuously in many industrialized countries since World War II, with the exception of certain Mediterranean countries, which have shown a slow downward trend during recent decades
(Pyörälä 1990, Medical Research Council 1998, Stakes 2005, Blocker, Jr. 2006, IAS 2006). Alcohol consumption by women and young people in particular has increased considerably since the
war, and attention has also been focused more recently on the increasing use of alcohol among the
It has been known for centuries that alcohol consumption is somehow associated with increased
illness and death, and the scientific study of alcohol-related mortality that started in the 1920s has
continued with the establishment of a connection between alcohol and liver disease (Mann et al.
2003). Over the past few decades, alcohol and alcoholism have been studied intensively in order to
achieve more effective diagnosis, treatment and prevention.
Alcoholism and alcohol-related disorders place a considerable burden on society nowadays, and
alcohol strategies should therefore be focused on the prevention and early detection of alcoholism
(Wilson 1995, Room et al. 2005, Rehm et al. 2006). The simultaneous use of alcohol and drugs has
also become more common, posing a new challenge to health care. The price and availability of
alcoholic beverages are among the major factors that have contributed to alcohol consumption, and
these should be controlled in order to reduce consumption at the population level (Chaloupka et al.
1998, Room et al. 2005).
2.2. Main features of ethanol metabolism
Ethanol is a simple molecule, which can affect cell membranes and signaling pathways throughout
the body. Ethanol in itself is not very toxic, but its first metabolite, acetaldehyde, is highly toxic and
may cause various adverse effects in cells and tissues (Niemelä 2001, Eriksson 2001). Acetaldehyde
is carcinogenic, mutagenic and can interfere with DNA synthesis and repair. It can also react with
proteins and cellular constituents, forming stable or unstable acetaldehyde adducts. The process of
ethanol metabolism also generates free oxygen radicals, which are extremely reactive and may
thereby cause major damage to cells (Molina et al. 2003, Caro and Cederbaum 2004).
After ingestion, ethanol is absorbed in the gastrointestinal tract, mainly in the small intestine.
Approximately 2–10 % of the ethanol absorbed will be removed unaltered in the respiration and
urine (Lieber 2005), while the rest will be metabolized. The oxidation of ethanol to acetaldehyde
begins in gastric wall through the action of the alcohol dehydrogenase enzyme (ADH), in a process
referred to as gastric first-pass metabolism (Julkunen et al. 1985, Frezza et al. 1990, Seitz and
Pöschl 1997). The microbes in the alimentary tract are also capable of metabolizing ethanol. The
resulting acetaldehyde may accumulate in the colon because of insufficient aldehyde dehydrogenase
(ALDH) activity and may cause adverse effects there (Salaspuro 1996, Koivisto and Salaspuro
Ethanol is rapidly distributed among the various tissues of the body. Its metabolism takes place
primarily in the liver, where it is oxidized to acetaldehyde, mainly via the ADH pathway. The minor
metabolic pathways are the microsomal ethanol oxidizing system (MEOS) and the catalase pathway. The MEOS pathway, in which hepatic P450 cytochromes (mainly CYP2E1) are responsible
for ethanol oxidation, is induced by high blood ethanol concentrations and chronic ethanol consumption (Lieber and DeCarli 1968, Salaspuro and Lieber 1978, Asai et al. 1996, Lieber 2004). The
catalase pathway requires hydrogen peroxide, which is produced only in small amounts under normal circumstances (Thurman and McKenna 1975, Thurman and Handler 1989, Molina et al. 2003).
More recently, however, Bradford and co-workers (1999) have reported that the catalase pathway
appears to be ethanol- and methanol-inducible and that the Kupffer cells in the liver participate in
the regulation of the hydrogen peroxide supply. It has also been suggested that the catalase pathway
may account for most of the ethanol oxidation occurring in brain (Aragon et al. 1991, McBride et al.
2002). This is of importance because acetaldehyde, unlike intact ethanol, cannot readily pass from
the blood to the brain because of the abundance of ALDH in the blood-brain barrier (Zimatkin and
Acetaldehyde is further oxidized to acetate in liver by hepatic ALDH. The resulting acetate is
transported to the muscles and heart and finally converted to carbon dioxide and water and eliminated from body.
2.3. Effects of ethanol on health
The health hazards associated with excessive alcohol consumption are numerous and well known.
Almost all tissues in the body are affected by ethanol and it is related to more than 60 medical conditions (Rehm et al. 2003). The adverse health effects of excessive or even moderate ethanol consumption include both physiological and mental problems (Lieber 1995, Damström Thakker 1998,
Adrian and Barry 2003, Corrao et al. 2004). Some beneficial effects of moderate drinking have also
been suggested recently. The adverse and beneficial effects of alcohol consumption habits on health
are summarized in Table 1.
When evaluating the health effects of ethanol, attention should be paid both to the actual amount
consumed and to drinking patterns, as some adverse effects are related to occasional heavy drinking
and intoxication and some to chronic consumption (Niemelä 2002, Rehm et al. 2003). Social harm
and the possible health benefits seem to be more closely linked to drinking patterns than are the
adverse health effects (Damström Thakker 1998). There are also several factors that affect individual susceptibility to the effects of alcohol, such as age, gender, race, body mass and genetic factors
(Lieber 1995, Damström Thakker 1998).
Table 1. Possible effects of various levels of ethanol intake on health.
0 or +
0 or +
Accidents and injuries
Coronary heart disease
+ protective effects
– harmful effects
0 no effects
(Medical Research Council 1998, Damström Thakker 1998, Cook 1998, Eckardt
et al. 1998, Rodgers et al. 2000, Diaz et al. 2002, Sareen et al. 2004, Watzl et al.
2004, Meyerhoff et al. 2005, Borges et al. 2006, Heng et al. 2006)
2.3.1. Ethanol and the liver
Since the liver accounts almost exclusively of ethanol metabolism, it is also the main target for
ethanol toxicity (Lieber 1977, Jaeschke et al. 2002, Niemelä 2002). Many studies have reported a
dose-response relationship between alcohol consumption and various types of liver damage (Coates
et al. 1986, Norton et al. 1987, Corrao et al. 1999, White et al. 2002). Some of the adverse effects of
ethanol on the liver are likely to be caused by acetaldehyde and free radical production rather than
the ethanol itself (Eriksson 2001, Molina et al. 2003, Chase et al. 2005, Dey and Cederbaum 2006),
and it has also become evident that the immune system plays a key role in the pathology of liver
diseases (Leevy and Elbeshbeshy 2005).
The earliest abnormality to appear in the liver due to excessive alcohol consumption is the accumulation of lipids, which may lead to steatosis (fatty liver). This condition is transient and usually
reversible if the alcohol intake ceases. Upon continuing ethanol intake, steatosis may develop into
an inflammation process, hepatitis, which involves cell death and massive migration of immune
cells into the liver. This is a potentially lethal condition and may easily recur in patients who re18
cover and resume drinking. Alcoholic cirrhosis is a condition in which the normal liver structure has
gradually become fibrotic as a consequence of persistent inflammation and abnormal connective
tissue metabolism. The normal functions of liver become impaired as fibrogenesis proceeds and
eventually this process may lead to liver failure and death. It has been estimated that 10–15 % of all
chronic alcoholics will eventually develop cirrhosis (Niemelä 2002, Mann et al. 2003).
Advanced liver disease may also contribute to brain function. When the venous blood flow to
the liver is obstructed due to extensive cirrhosis, toxic substances and metabolic by-products may
leak into the systemic circulation and pass to the brain, interfering in the actions of neurotransmitters (Oscar-Berman et al. 1997).
2.3.2. Ethanol and extrahepatic tissues
Even a moderate level of ethanol intake has been found to be associated with several types of cancer, including cancers of the upper aerodigestive tract, colorectal area and liver, and breast cancer in
women. There seems to be a dose-response relationship between ethanol intake and the related cancer risk (Corrao et al. 1999, Boffetta and Hashibe 2006). Genetic factors have also been found to
affect individual susceptibility to development of alcohol-related cancer, mainly via polymorphisms
of enzymes associated with ethanol metabolism, folate metabolism and DNA repair (Boffetta and
Hashibe 2006). The exact mechanisms by which ethanol induces cancer have not been fully defined
but several possibilities have been discussed (Wight and Ogden 1998, Molina et al. 2003, Dumitrescu and Shields 2005, Boffetta and Hashibe 2006). Alcohol has been estimated to cause 3 % of
all cancers worldwide (Medical Research Council 1998).
Continuous ethanol consumption is known to have adverse effects on the nervous system, particularly in the brain, including brain shrinkage, emotional and personality changes and learning and
memory defects. It is also known to affect sleep patterns, muscular coordination and the regulation
of body temperature (Oscar-Berman et al. 1997). The mechanisms by which ethanol affects the
nervous system are complex and not yet fully understood, and there appear to be additional factors
such as age, gender, family history, coexisting health problems and nutritional deficiencies that play
a role in the development of neurological defects (Oscar-Berman et al. 1997). It is well established
that ethanol can activate the dopamine system in the brain, which contributes to its rewarding effects (Di Chiara 1997, Tupala and Tiihonen 2004) and it also interferes with the actions of various
other neurotransmitters (Oscar-Berman et al. 1997, Eckardt et al. 1998). Unlike other addictive
drugs, however, it does not have specific receptors on cell surfaces.
Ethanol is known to penetrate the placental barrier and cause adverse effects on the growing fetus but the exact mechanisms are still unclear. Fetal alcohol syndrome (FAS), or its milder form,
fetal alcohol effects (FAE), may result from alcohol consumption during pregnancy and is associated with growth and mental retardation and certain facial features and also with abnormalities in
various organs (Jones and Smith 1973, Chaudhuri 2000, Jacobson and Jacobson 2002). FAS seems
to be related to relatively high levels of ethanol consumption but milder symptoms such as reduced
birth weight and learning problems have already been observed at moderate levels of ethanol intake
(Ogston and Parry 1992, Allebeck and Olsen 1998, Eckardt et al. 1998, Sood et al. 2001). Fetal
damage has been shown to be dose-dependent (Allebeck and Olsen 1998, Sood et al. 2001), al-
though high occasional blood alcohol levels have been reported to be a greater risk for the fetus
than lower continuous consumption (Maier and West 2001). However, not all heavy-drinking mothers give birth to a FAS/FAE child, so that additional risk factors have been suggested to exist (Jacobson et al. 1996, Maier and West 2001). The number of children born with FAS/FAE has been
increasing with the growing percentage of alcohol-consuming pregnant women and this problem
needs special attention in current health care (Chaudhuri 2000, Warren and Foudin 2001, Eustace et
Excessive drinking may also lead to disorders in the gastrointestinal, cardiovascular, respiratory,
reproductive and endocrine systems (Medical Research Council 1998, Damström Thakker 1998,
Adrian and Barry 2003). High alcohol intake has been reported to cause damage to heart muscle
and to increase the risk of arrhythmias (Piano and Schwertz 1994), and also to elevate blood pressure, which is a risk factor for brain infarction, for instance (Hillbom and Kaste 1990, Puddey and
Beilin 2006). Excessive alcohol consumption is a major cause of acute and chronic pancreatitis, its
relationship to the risk and severity of pancreatitis being dose-dependent (Durbec and Sarles 1978,
Jaakkola et al. 1994, Corrao et al. 1999, 2004). Alcohol is also known to cause various adverse effects in the gastrointestinal tract, such as disruptions in the gastric mucosal barrier, which may lead
to deficient absorption of many nutrients and vitamins (Bujanda 2000, Bode and Bode 2003).
2.3.3. Ethanol and the immune system
It is well known that the immune system is greatly influenced by excessive drinking (Cook 1998). It
was found in the early 20th century that alcoholics are more susceptible to pneumonia than nonalcoholics, and it was later observed that alcoholics are also more susceptible to other types of infection, such as septicaemia and tuberculosis. More recent studies on immunity and alcohol have
focused mostly on immune mechanisms in liver pathology although these mechanisms may also
play a part in alcohol-related cardiac, endocrine and neurological damage. Ethanol is known to interfere with the functioning of cells in the immune system such as B and T cells, macrophages,
monocytes and dendritic cells and may thus contribute to an increased susceptibility to infection
(Cook 1998, Saeed et al. 2004, Szabo et al. 2004, Chase et al. 2005). An imbalance in the T cell
subsets TH1 and TH2 has been suggested as a factor contributing to immunological disturbances in
alcoholics (Cook 1998). Decreased B-cell numbers have been reported in alcoholics with liver disease, but they often show elevated immunoglobulin levels, particularly IgA levels, and increased
IgA deposition in the kidneys, skin and liver (Swerdlow et al. 1983, van de Wiel et al. 1988, Amore
et al. 1994, Cook 1998). Excessive alcohol consumption may lead to the specific production of antibodies against ethanol-induced antigens (Israel et al. 1986, Koskinas et al. 1992, Tuma and Klassen 1992), and also to alterations in the cytokine balance, which both may contribute to the pathogenesis of tissue damage (McClain et al. 1993, Crews et al. 2006).
2.3.4. Ethanol, psychiatric disorders and accidents
The association between excessive drinking and psychiatric disorders has been studied since the
1950s. It has been reported that mental problems tend to emerge before the onset of excessive
drinking but may also follow excessive drinking or the two may coincide in onset (Kessler et al.
1997, Berglund and Öjehagen 1998). Alcohol abuse or dependence can worsen the prognosis for
psychiatric disorders (Berglund and Öjehagen 1998), and alcohol dependence is also a significant
risk factor for suicide, particularly if associated with mental disorders (Berglund and Öjehagen
1998, Pirkola et al. 2004, Preuss et al. 2006). Alcohol has been found in the blood of 20–55 % of
patients who have committed suicide (Öhberg et al. 1996, Crombie et al. 1998, Bedford et al. 2006).
At the population level, suicide rates are usually high in countries with a high alcohol consumption
and a rise in suicide rates is observed when total alcohol consumption increases (Berglund and Öjehagen 1998).
In addition to diseases, excessive alcohol consumption accounts for numerous cases of trauma,
violence and motor vehicle accidents (Brismar and Bergman 1998, Cunningham et al. 2002, Savola
et al. 2005). Alcohol-related accidents are common: it has been estimated that over 50 % of injured
emergency unit patients have alcohol in their blood (Savola et al. 2005). In connection with alcoholrelated violent behaviour, it has been speculated that alcohol may either weaken the normal inhibitory impulses or stimulate aggressive behaviour, but the theories are not conclusive (Källmén and
Gustafson 1998, Badawy 2003). Either way, the amount of alcohol consumed is clearly associated
with the number of reported cases of violence. Alcohol is also involved in significant proportions of
traffic accidents and the risk is known to increase with increasing blood alcohol levels (Hingson and
Winter 2003, Bedford et al. 2006, Heng et al. 2006). A recent report has also stated that even moderate drinking (blood alcohol concentration within the legal limits) may significantly increase the
risk of motor vehicle accidents (Heng et al. 2006).
2.3.5. Suggested positive effects of ethanol
Although the positive effects of ethanol consumption are usually thought to lie in psychological
benefits such as mood elevation, stress relief and more social behaviour, reports concerning the
beneficial effects of moderate drinking on somatic health have also been published recently. Studies
on coronary heart disease (CHD) have suggested that people drinking approximately 20 g of ethanol
per day have the lowest relative risk of CHD, although there is a rapidly increasing risk at higher
consumption levels (Doll 1997, Corrao et al. 2000, Agarwal 2002). It has been suggested that red
wine may confer a greater protective effect than other types of alcoholic beverage (Grønbæk 2006).
Ethanol is known to elevate high-density lipoprotein (HDL) levels and this, along with other possible factors, has been suggested to contribute to a reduction in the CHD risk (De Oliveira e Silva et
al. 2000, Agarwal 2002). Moderate drinking may also be associated with a decreased risk of diabetes (Howard et al. 2004). The thresholds for "healthy" drinking are not definitive, however, as there
is wide variation in individual susceptibility to the harmful effects of alcohol (Damström Thakker
1998, Enoch 2003, Fromme et al. 2004). Whether clinicians should advise patients with a high risk
of cardiovascular disease, for example, to have one drink per day remains a controversial issue
(Ellison 2002, Fisher Wilson 2003), especially as the same beneficial effects could in most cases be
achieved with regular exercise and a low-fat diet, without the risks associated with alcohol (Damström Thakker 1998, Goldberg 2003). The benefits of moderate drinking with respect to cardiovascular disease, for instance, may quickly be outweighed by increases in liver disease and cancer, and
the net effect is likely to be negative (Friedman and Klatsky 1993, Damström Thakker 1998, Goldberg 2003).
2.4. Assessing alcohol consumption
2.4.1. Amount and pattern of drinking
The limits for moderate ethanol intake and the definition of a standard drink show variation depending on the source and country. The upper limits quoted here for moderate drinking, 280 g of ethanol
per week for men and 190 g for women, are those defined for Finland (Sillanaukee et al. 1992a),
while in the United Kingdom, for instance, the weekly limits are set at 168 g for men and 112 g for
women (Sharpe 2001). One unit or drink is defined as containing ethanol in amounts ranging from 8
g to as much as 20 g, so that ethanol consumption should preferably be assessed in grams in order
to avoid confusion (Turner 1990, Dufour 1999, Brick 2006).
The methods for identifying excessive drinking include verbal reports, questionnaires and laboratory tests. In clinical practice, interviews and questionnaires form the basis for detecting excessive
drinking in its early phase, whereas laboratory tests are most useful for monitoring the treatment of
alcoholism, although they can also be integrated into the detection of excessive drinking (Treatment
of alcohol abuse. Current Care Guideline 2005).
On the basis of their drinking habits, individuals can be divided into five groups based on limits that
are currently used in Finland.
1. Abstainers, who do not drink any alcohol,
2. Moderate drinkers, who have a weekly ethanol consumption below 280 g (men) and 190 g
(women) and do not drink over 80 g (men) and 60 g (women) on any single occasion,
3. Hazardous drinkers, who consume alcohol in amounts exceeding the limits of moderate drinking,
but do not show any obvious immediate disorders,
4. Alcohol abusers, who consume ethanol in large amounts and have various medical, social and
financial problems due to their drinking, although they do not fulfill the criteria for alcoholism, and
5. Alcoholics, alcohol abusers who meet the criteria for alcohol dependence including pathological
alcohol consumption, social impairment, and the presence of tolerance and withdrawal symptoms.
Alcoholic individuals may further be classified into two subtypes according to the onset of their
drinking (Buydens-Branchey et al. 1989, Johnson et al. 2000, Enoch 2003). If excessive drinking
has started before the age of 20, the drinker is classified as an early onset alcoholic (EOA), also
called type II alcoholism. It is often associated with antisocial behaviour and a high familial or genetic loading towards alcoholism. Late onset alcoholics (LOA) or type I alcoholics usually start
drinking later in life, often as a consequence of psychosocial stress.
2.4.2. Self-reported alcohol consumption
Since there are no exact symptoms of early-phase alcohol problems, clinicians have limited means
of assessing patients' drinking based on external evaluations (MacKenzie et al. 1996, Sharpe 2001).
Patients with alcohol problems often tend to deny having any, and it may be difficult to make a
clinical diagnosis when based solely on an interview.
Structured questionnaires can be of assistance in identifying alcohol abusers and are well suited
for screening purposes (Seppä et al. 1995, Ryb et al. 1999, Aertgeerts et al. 2001). The best-known
questionnaires are AUDIT, CAGE and MAST. AUDIT (Alcohol Use Disorders Identification Test)
is a quite recently developed questionnaire, which consists of 10 questions and focuses mainly on
the level and frequency of consumption and the adverse consequences of drinking (Saunders et al.
1993). It has been reported that AUDIT is the most sensitive among the current questionnaires for
detecting hazardous and harmful drinking (Seppä et al. 1995, MacKenzie et al. 1996, Reid et al.
1999). Shorter forms of this questionnaire (three to five questions) have also been developed and
tested in clinical populations more recently, and the results have so far been quite promising (Piccinelli et al. 1997, Aertgeerts et al. 2001, Gual et al. 2002, Hodgson et al. 2002). The CAGE questionnaire includes four questions concerning drinking habits and alcohol-related problems (Ewing
1984), its name being an acronym for the key words in the questions (Cut down, Annoyed, Guilty,
Eye-opener). The most extensive questionnaire is the Michigan Alcoholism Screening Test
(MAST), which consists of 25 questions directed at the recognition of drinking problems, helpseeking behaviour and alcohol-related disabilities (Selzer 1971). Both CAGE and MAST have been
found to perform well in detecting alcohol dependence but to be less effective in identifying hazardous drinkers (MacKenzie et al. 1996, Reid et al. 1999). Combinations of these questionnaires
and novel variations have also been introduced, with efficacies comparable or superior to the older
ones (Seppä et al. 1998, Merikallio-Pajunen et al. 2004, Patton et al. 2004).
A commonly used means of assessing the amounts of ethanol consumed is the timeline followback (TLFB) method, in which the patient is not asked to estimate the mean consumption of alcohol
but to fill in a calendar and report how many drinks he or she has consumed during a specified period of time (Allen et al. 1992). This method is used mostly for research purposes and to a lesser
extent in clinical interviews.
2.4.3. Laboratory markers of alcohol consumption
The introduction of laboratory tests for detecting excessive alcohol intake has improved the assessment of ethanol consumption. Laboratory tests and structured questionnaires may be used in combination in order to achieve the most efficient detection of excessive alcohol use (Sharpe 2001, Dolman and Hawkes 2005, Berner et al. 2006).
An ideal biomarker for detecting excessive drinking should be
1. Specific to ethanol, in order to avoid false positives
2. Sensitive enough to provide a useful screening tool for identifying excessive drinking also in its
3. Related to the amount of alcohol consumed.
4. Suitable for monitoring abstinence (e.g. decreasing consistently after cessation of drinking) and
5. Easy and cost-effective in routine use.
None of the existing biomarkers fulfills all these criteria. The most frequently used markers today
are carbohydrate-deficient transferrin, gamma-glutamyltransferase and the mean corpuscular volume of erythrocytes.
The development of a new marker usually begins by contrasting alcoholics with abstainers,
which shows whether the method gives any differences between these two groups in the first place.
The process continues with studies in clinical populations, since the ultimate purpose of a marker is
to identify excessive drinkers in a general population, where various sources of unspecificity may
occur. If the marker appears to be of value in these studies, its use may be considered in routine
Basically, biomarkers can be divided into markers of acute and chronic alcohol consumption
(Rosman and Lieber 1992). The former, which remain elevated for a few days, may be used for
detecting relapses in alcoholic patients undergoing treatment, for example. The latter, which are
usually elevated after a longer period of drinking or several shorter periods of binge drinking, are
useful for identifying alcohol abuse and to some extent for screening purposes. Some of these longterm markers can also be useful for detecting relapses during treatment.
Among the measures used to describe the diagnostic performance of a marker (Boyd 1997), sensitivity is the percentage of positive test results of all the positive patients and specificity is the percentage of negative test results from all the negative patients. Thus, if an alcohol marker has a sensitivity of 60 % and a specificity of 90 %, it means that 60 % of the alcoholics will have abnormal test
results and 90 % of the control population will have normal test results. An ideal assay should provide both a specificity and a sensitivity close to 100 %. The other commonly used measures are
predictive values. The positive predictive value (PPV) is the percentage of true positive results of all
positive results (including true and false positives), and the negative predictive value (NPV) is the
percentage of true negative results of all negative results (including true and false negatives). The
closer these values are to 100 %, the better the predictive value of the marker.
The cut-off values, sensitivities and specificities of markers can be assessed with the receiveroperating characteristic (ROC) curves (Robertson and Zweig 1981, Zweig and Campbell 1993,
Boyd 1997). A cut-off value is what separates a normal from an abnormal (usually elevated) result.
ROC analysis gives sensitivity and specificity values for all possible cut-offs, and a measure called
the area under the curve (AUC), which describes the accuracy of the test. With ROC analysis, it is
easy to compare the performance of different markers and cut-off values.
One important aspect of the use of laboratory markers is the definition of "normal" values and
reference limits. Reference intervals are defined based on values for apparently healthy individuals.
This reference population should be carefully selected and the effects of age, gender, race, nutritional status and preferably alcohol drinking habits must be considered in order to avoid misleading
reference limits and future misinterpretation of laboratory results. The most common methods for
determining reference intervals are the mean ± 2SD for populations following a Gaussian distribution (or a Gaussian distribution is obtained by transformation of the data) and the non-parametric
method, which does not require Gaussian distributions but uses 2.5 and 97.5 percentiles as the limits (for a review, see Horn and Pesce (2003)).
2.4.4. Laboratory markers of ethanol-induced tissue injury
Alcohol-induced tissue injury may be examined with blood tests, although a tissue biopsy is the
only confirmative test (Niemelä 2002, Phillips et al. 2003). The liver enzymes commonly used for
measuring alcohol consumption (GGT, AST, ALT, see next section) reflect the status of the liver
and may also be used to assess liver damage. Serum albumin, bilirubin and alkaline phosphatase
may also be used for this purpose. Markers of the progression of liver disease are usually associated
with collagen and connective tissue metabolism, the most commonly used being the aminoterminal
propeptide of type III procollagen (PIIINP). Other postulated fibrosis markers include the aminoterminal propeptide of type I procollagen (PINP) and hyaluronic acid (HA).
2.4.5. Gender issues
Gender has been reported to influence both the effects of alcohol consumption on health and the
detection of alcohol abuse. Women are known to be more susceptible to the effects of alcohol than
men, particularly to liver disease, for several physiological and metabolic reasons (Lieber 1995,
Schenker 1997, Damström Thakker 1998). Immune responses have also been observed to vary
greatly between the genders, probably due to hormonal factors (Kovacs and Messingham 2002).
When assessing individuals' alcohol consumption, questionnaires appear to be more effective for
men than for women, with AUDIT being the most accurate (Seppä et al. 1995, Aertgeerts et al.
2001). Laboratory tests seem in general to be more accurate for detecting alcohol abuse in men than
in women, and significant gender-dependent differences in the diagnostic usefulness of markers
have been observed (Chalmers et al. 1980, Anton and Moak 1994, Löf et al. 1994, Allen et al. 2000,
Sillanaukee et al. 2001, Neumann and Spies 2003).
2.5. Routinely used biomarkers of ethanol consumption
2.5.1. Ethanol concentration in blood, breath or urine
Measurement of the ethanol concentration in the blood, breath or urine is suitable for acute health
care purposes and provides a specific marker for alcohol intake. It can also give information on the
patient's drinking habits, as blood ethanol levels exceeding 150 mg/l (1.5 ‰ or 33 mM) without
obvious signs of intoxication or 300 mg/l (3.0 ‰ or 65 mM) on any occasion indicate significant
tolerance, which is typically associated with alcohol dependence (Niemelä 2002). The blood alcohol
concentration at the time of admission has been reported to be the best indicator of hazardous drinking practices and alcohol dependence in trauma patients (Ryb et al. 1999, Savola et al. 2004).
2.5.2. Gamma-glutamyl transferase
The liver enzyme gamma-glutamyl transferase (GGT) is one of the most commonly used laboratory
tests for detecting excessive alcohol intake, although it may also be elevated by other causes of liver
damage, e.g. drugs or non-alcoholic liver diseases. Coffee, smoking, obesity and age have also been
found to influence GGT values (Cushman 1992, Aubin et al. 1998, Daeppen et al. 1998, Puukka et
al. 2006a, 2006b). There have been several reports of a positive correlation between the amount of
ethanol ingested and serum GGT levels, but the sensitivity of GGT has varied greatly between
populations, from 15 to 85 %, being lower in samples from the general practice (Bagrel et al. 1979,
Chick et al. 1981, Papoz et al. 1981, Anton and Moak 1994, Yersin et al. 1995, Sillanaukee et al.
1998, Hock et al. 2005, Berner et al. 2006). GGT tends to have a relatively low specificity as compared with carbohydrate-deficient transferrin (CDT), for example, because of many sources of unspecificity (Reynaud et al. 2000, Niemelä 2002, Miller and Anton 2004, Hock et al. 2005).
The ethanol consumption needed for an elevation in GGT has been estimated to be over 40 g per
day (Sharpe 2001, Miller and Anton 2004) although no clear threshold has been observed (Schellenberg et al. 2005). GGT values have been estimated to normalize in 2–5 weeks in alcoholics, depending on the original level and the presence or absence of liver pathology (Orrego et al. 1985,
Cushman 1992, Conigrave et al. 2003).
2.5.3. Carbohydrate-deficient transferrin
The first reports on the existence of carbohydrate-deficient transferrin (CDT) in alcoholics were
published almost 30 years ago (Stibler and Kjellin 1976, Stibler et al. 1978, 1979), since when CDT
has emerged as a highly useful marker of excessive ethanol consumption (Stibler 1991, Allen et al.
1994, Golka and Wiese 2004, Bortolotti et al. 2006, Niemelä 2007). Its sensitivity has varied from
20 to 90 %, depending on the population, being higher for men than for women in most studies, and
lower in detecting hazardous drinking than in identifying alcoholism (Bell et al. 1994, Anton and
Moak 1994, Anton and Bean 1994, Litten et al. 1995, Lesch et al. 1996b, Sillanaukee et al. 1998,
Allen et al. 2000, Anton et al. 2001, Berner et al. 2006). The advantage of CDT over other markers
is its high specificity (Stibler 1991). False positives have been reported to occur in rare cases of
genetic transferrin variants (Stibler et al. 1988, Helander et al. 2001) and in certain other conditions
(Sillanaukee et al. 2001, Fleming et al. 2004). Some variation in the results was also observed with
earlier methods that measure the absolute amounts of CDT (U/l) due to variations in total transferrin
levels (Sorvajärvi et al. 1996, Keating et al. 1998, Viitala et al. 1998, Anton et al. 2001). More recently developed methods, which express the CDT results as a percentage of total transferrin, have
improved the performance of this marker (Keating et al. 1998, Helander 1999, Anton et al. 2001,
Helander et al. 2005).
CDT includes isoforms that have fewer carbohydrate (sialic acid) moieties than normal transferrin (Stibler 1991). The predominant isoform in a healthy individual is tetrasialotransferrin, with four
sialic acid moieties, and excessive alcohol consumption has been shown to increase the proportion
of the asialotransferrin and disialotransferrin isoforms, which are generally considered to constitute
CDT. The trisialo fraction is not usually included in CDT measurements and has actually been
found to reduce the diagnostic accuracy of CDT (Mårtensson et al. 1997, Helander et al. 2001,
Arndt et al. 2002, Legros et al. 2002). It is not yet clear how ethanol induces the elevation of CDT,
but it has been suggested that acetaldehyde might either inhibit the enzymes that participate in
transferrin sialylation or activate the deglycosylating enzymes, or both (Stibler and Borg 1991, Xin
et al. 1995, Arndt 2001, Sillanaukee et al. 2001). CDT has been reported to become elevated at
daily ethanol consumptions ranging from 40 to 80 g, with duration of 2 to 3 weeks (Stibler 1991,
Schellenberg et al. 2005) although it has been suggested that even higher consumption may be
needed to increase CDT in the general population (Lesch et al. 1996a). The half-life of CDT has
been estimated to be approximately two weeks (Behrens et al. 1988, Stibler 1991, Allen et al.
2.5.4. Mean corpuscular volume
An increase in the mean corpuscular volume of erythrocytes (E-MCV), or macrocytosis, is typical
of people with chronic excessive ethanol intake, having sensitivities ranging from 20 to 90 %
(Cushman 1992, Bell et al. 1994, Yersin et al. 1995, Reynaud et al. 2000, Mundle et al. 2000).
MCV has proved to be diagnostically more sensitive among women, in whom it may even be superior to other markers (Chalmers et al. 1980, Morgan et al. 1981, Seppä and Sillanaukee 1994, Sillanaukee et al. 1998, Mundle et al. 2000). Several haematological diseases, hypothyroidism, reticulocytosis and vitamin B12 or folic acid deficiency (often related to excessive drinking) and smoking
have been shown to elevate MCV and thus reduce its specificity as an alcohol marker (Sharpe 2001,
Niemelä 2002). Alcohol abuse is frequently the underlying cause of elevated MCV in hospital patients with non-anaemic macrocytosis (Seppä et al. 1991).
The normalization time for MCV is longer than that for any other commonly employed marker,
2 to 4 months, which places limitations on its use for the short-term monitoring of abstinence but
offers a possibility for supervising abstinence over longer periods (Morgan et al. 1981, Mundle et
al. 1999, Niemelä 2002).
Other liver enzymes commonly used for detecting excessive drinking are the aminotransferases
(aspartate aminotransferase, AST and alanine aminotransferase, ALT). Although these are frequently elevated in patients with excessive alcohol consumption, they are more directly related to
liver status, so that increased AST and ALT values can be found in other conditions where the liver
is damaged, e.g. due to viral hepatitis or medications (Pratt and Kaplan 2000, Niemelä 2002, Conigrave et al. 2003). It has been suggested that the ratio of AST to ALT may indicate the aetiology of
liver disease: a ratio over 2 often being seen in alcohol-related liver disease, particularly in advanced states (Rosman and Lieber 1994, Nyblom et al. 2004). Aminotransferase levels have been
reported to normalize in 2 to 3 weeks after cessation of drinking, depending on the original level
(Niemelä 2002, Chrostek et al. 2006).
2.6. New biomarkers of ethanol consumption
2.6.1. Acetaldehyde adducts and anti-adduct antibodies
Acetaldehyde adducts are formed when acetaldehyde, the first metabolite of ethanol, reacts with
proteins and cellular constituents (Gaines et al. 1977, Stevens et al. 1981, Niemelä 2001) and may
persist in the blood for 1 to 3 weeks after the last dose of ethanol (Niemelä and Israel 1992). The
measuring of these adducts in either erythrocytes or plasma proteins could be of use for determining
recent alcohol consumption (Niemelä and Israel 1992, Sillanaukee et al. 1992b, Lin et al. 1993,
Niemelä 2002). Elevated levels have been reported in moderate drinkers as compared with abstainers (Niemelä and Israel 1992). The specificity of this approach may nevertheless be reduced to
some extent by endogenous acetaldehyde production (Rosman and Lieber 1992).
Acetaldehyde adducts have been shown to induce an immune response, and circulating IgA, IgG
and IgM antibodies, which recognize sequential and conformational epitopes in adducts, have been
found in alcohol-consuming populations (Israel et al. 1986, Koskinas et al. 1992, Tuma and Klassen
1992). Of these, anti-adduct IgAs have emerged as a possible indicator of excessive drinking (Worrall et al. 1994, 1996, 1998, Niemelä 2007).
The ratio of two serotonin metabolites, 5-hydroxytryptophol (5HTOL) and 5-hydroxyindole-3acetic acid (5HIAA), has been reported to increase in the urine after heavy drinking (Helander et al.
1992, Voltaire et al. 1992) and to remain elevated for 6–20 h after the disappearance of ethanol (Helander et al. 1993, Carlsson et al. 1993, Helander et al. 1996a). This marker has been suggested as
being of use for assessing recent ethanol ingestion e.g. for monitoring treatment or in forensic investigations, and also for verifying subjects' self-reports in alcohol studies (Helander et al. 1996a,
Helander and Eriksson 2002, Beck and Helander 2003).
2.6.3. Ethyl glucuronide
Ethyl glucuronide (EtG) is a direct non-oxidative metabolite of ethanol that has been reported to
increase in amount in the urine after ethanol ingestion (Dahl et al. 2002, Sarkola et al. 2003). Levels
of EtG may remain detectable for up to 5 days (Wurst et al. 2003) and as an intermediary marker it
could fill up the gap between the known short-term and long-term markers of alcohol consumption
(Wurst et al. 1999, Wurst and Metzger 2002, Wurst et al. 2003). Use of the EtG to creatinine ratio
in the urine has also been suggested, because drinking large volumes of water can lower the EtG
concentration but does not affect the EtG/creatinine ratio (Dahl et al. 2002, Bergström et al. 2003).
EtG has been found in serum, urine, tissues and even hair and may have significant applications in
both clinical and forensic medicine (Wurst et al. 1999, 2003, Bergström et al. 2003).
2.6.4. Fatty acid ethyl esters
Fatty acid ethyl esters (FAEEs), which are formed when ethanol reacts with fatty acids, were first
found in organs damaged by ethanol and also suggested as mediators of organ damage (Laposata
and Lange 1986, Laposata 1997). Later, their presence in serum and hair after ethanol intake was
also confirmed (Doyle et al. 1994, Pragst et al. 2001). FAEEs have been proposed as markers of
alcohol consumption, both in short term (measurements in serum) and long term (hair analysis)
(Doyle et al. 1996, Laposata 1997, Auwärter et al. 2001). They have been reported to remain de-
tectable in serum for up to 24 h after ethanol ingestion (Doyle et al. 1994), and for even longer periods in heavy drinkers (Borucki et al. 2004).
Phosphatidylethanol (PEth) is formed from phospholipides when ethanol is present, in a reaction
catalyzed by phospholipase D (Gustavsson 1995). Blood PEth has been reported to correlate closely
with ethanol consumption, but detectable PEth levels may not occur until ethanol consumption is
relatively high (Hansson et al. 1997, Varga et al. 1998, Aradottir et al. 2006). PEth may be detected
for more than two weeks after last dose of ethanol (Hansson et al. 1997, Gunnarsson et al. 1998) so
that it has been suggested as a possible marker of excessive drinking (Varga et al. 1998, Aradottir et
2.6.6. Sialic acid
Studies have suggested that concentrations of sialic acid (SA), a carbohydrate predominantly found
in the oligosaccharide chains on the surface of cell membranes and macromolecules (Traving and
Schauer 1998), are often elevated in serum and saliva of alcoholics and might serve as an alcohol
marker (Pönniö et al. 1999a, Sillanaukee et al. 1999b, Romppanen et al. 2002). However, SA concentrations seem to increase in a number of other conditions as well, e.g. in inflammatory and cardiovascular diseases, diabetes and cancer, which limits its use as a marker of excessive drinking
(Stefenelli et al. 1985, Sillanaukee et al. 1999a). In addition, body mass index (BMI), blood pressure, ageing, hormonal factors and smoking may cause variations in SA levels (Pönniö et al.
1999b). The presence of SA in the saliva of alcoholics could be of use when considering noninvasive methods of assessing alcohol intake (Pönniö et al. 1999a). SA concentrations also appear
to decrease during abstinence with a time range of several weeks (Pönniö et al. 1999a, Sillanaukee
et al. 1999b, Anttila et al. 2005).
Another application involving sialic acid is the plasma sialic acid index (SIJ) of apolipoprotein J
(ApoJ), referring to the number of sialic acid moieties of ApoJ. This index has been found to decrease as a consequence of chronic ethanol intake and could possibly serve as a marker of excessive
drinking (Ghosh et al. 2001).
2.6.7. Other postulated markers
In addition to other direct ethanol metabolites (EtG, FAEEs, PEth), ethyl sulphate (EtS) in the urine
has been introduced as a potential marker of alcohol consumption. It appears to have quite similar
characteristics to EtG (Wurst et al. 2006).
Other suggested markers of alcohol intake include hepatic lysosomal enzyme β-hexosaminidase,
which has been shown to increase in alcoholics (Isaksson et al. 1985, Kärkkäinen et al. 1990, Hultberg et al. 1991), but also in non-alcoholic liver disease and in some other conditions, which limits
its usefulness (Rosman and Lieber 1992, Javors and Johnson 2003).
Urinary dolichols are polyprenol compounds, which have been reported to increase in chronic
alcoholics (Pullarkat and Raguthu 1985, Roine et al. 1987). Their use as a marker, however, suffers
from a lack of both sensitivity (Stetter et al. 1991) and specificity (Roine et al. 1989, 1991).
In addition, mitochondrial AST (mAST) (Nalpas et al. 1986), erythrocyte ALDH (Agarwal et al.
1983), urinary salsolinol (Collins et al. 1979, Haber et al. 1995) and urinary alanine aminopeptidase
(Taracha et al. 2004) have also been proposed as indicators of excessive drinking. More research
into these postulated markers is needed in order to establish their clinical value and usefulness.
2.7. Marker combinations
Despite the abundance of potential markers, none of them has so far achieved perfect diagnostic
accuracy. Some researchers during the past 20 years have suggested that a diagnostic improvement
could be achieved by combining two or more alcohol markers (Monteiro and Masur 1985,
Salaspuro 1987, Niemelä and Israel 1992, Yersin et al. 1995, Anton et al. 2001, Sillanaukee and
Olsson 2001, Hock et al. 2005). An ideal combination would consist of markers which are affected
in different ways by alcohol intake. GGT and CDT, for example, do not correlate in the alcoholic
population and seem to be elevated in partially different populations, possibly reflecting independent mechanisms by which ethanol induces their elevation (Anton and Moak 1994, Helander et al.
1996b, Huseby et al. 1997, Anton et al. 2001, Conigrave et al. 2002).
The first researchers who studied the use of test results in combination integrated a large panel
of laboratory tests into a single score by mathematical methods (Beresford et al. 1982, Hillers et al.
1986, Hartz et al. 1997, Harasymiw et al. 2000, Harasymiw and Bean 2001). The use of large test
panels with dozens of assays would not, however, be either rational or cost-effective, and thus more
interest has been shown in combining two or three markers of alcohol consumption. The conventional manner of combining markers is to see whether either is elevated. This approach has been
found to give improved assay sensitivity but is frequently associated with a decrease in specificity
(Rosman and Lieber 1992, Anton and Moak 1994, Helander et al. 1996b, Huseby et al. 1997, Anton
et al. 2001, Schwan et al. 2004). Researchers have also introduced the idea of combining markers
by means of mathematical equations (Sillanaukee 1992, Sillanaukee and Olsson 2001), whereupon
the sensitivities appear to increase without significant loss of specificity (Sillanaukee and Olsson
2001, Anttila et al. 2003, Chen et al. 2003). The most frequently used components of such combinations are CDT and GGT, and some researchers have also included MCV as a third component (Sillanaukee et al. 1998, Sillanaukee and Olsson 2001, Anttila et al. 2003, Hock et al. 2005). A promising approach was introduced by Sillanaukee and Olsson (2001), who combined GGT and CDT by
means of the equation 0.8 x ln(GGT) + 1.3 x ln(CDT) and reported sensitivities of 79 % (men) and
72 % (women) for this marker. A further improvement in this combination was achieved by replacing absolute CDT values with %CDT (Anttila et al. 2003). The marker has been reported to be more
sensitive for detecting excessive drinking in men than in women (Sillanaukee and Olsson 2001,
Anttila et al. 2003, Chen et al. 2003). Chen and co-workers (2003) also found that integrating clinical information with the combined marker would be useful for the accuracy of detection.
More research into combinations of markers is needed in order to establish their usefulness in
screening for excessive drinking. Their cost-effectiveness would also need to be evaluated in order
to assess whether the improvement in detection is worth the increase in assay costs. The long-term
effects of any improvement in early detection and treatment should also be taken into consideration
in cost-effectiveness analyses.
2.8. Laboratory markers for the follow-up of alcoholics
Laboratory markers play an important role in the monitoring of abstinence in treatment units and in
health care, and they are also used for legal purposes, e.g. for the supervision of drunken drivers. A
suitable marker for follow-up purposes should normalize at a steady rate with as little interindividual variation as possible and should also react to relapses of drinking during follow-up.
For the moment, CDT is the most favoured marker for follow-up purposes, as it has been observed to decrease steadily during abstinence and to be suitable for detecting relapses (Schmidt et
al. 1997, Allen et al. 2001, Sharpe 2001, Anton et al. 2002). Determination of the patient's CDT
baseline could be of use in the monitoring process, as a 30 % elevation from the baseline may indicate a relapse (Borg et al. 1995, Anton et al. 2002). CDT appears to be especially suitable for follow-up in the case of men and of patients with liver disease (Bell et al. 1993, Anton et al. 1996, Allen et al. 1999, Anton et al. 2002), since the use of GGT and other liver enzymes in follow-up may
be complicated by the fact that they are affected by liver status (Rosman and Lieber 1994). Thus
they have usually not performed as well as CDT in follow-up studies (Bell et al. 1993, Salaspuro
1999, Anttila et al. 2004).
Markers of recent alcohol consumption (particularly the direct metabolites of ethanol) may
prove to be of significant value for follow-up purposes (Hansson et al. 1997, Wurst et al. 1999, Helander and Eriksson 2002, Bisaga et al. 2005, Wurst et al. 2005), and as these markers have different time frames with respect to detection (from hours to weeks) it might be possible to select suitable markers for different situations.
Combinations of markers have also been proposed for the follow-up of treatment. CDT together
with the 5HTOL/5-HIAA ratio in urine could serve as an accurate marker for detecting a recent
relapse (Carlsson et al. 1993), and another useful combination for this purpose could be GGT and
CDT (Allen et al. 1999, Anton et al. 2002).
2.9. Immune responses related to alcohol consumption
2.9.1. Immune responses to ethanol metabolites
Acetaldehyde is a highly reactive molecule and can form stable condensates with various proteins
and cellular constituents (Gaines et al. 1977, Stevens et al. 1981, Wehr et al. 1993, Svegliati-Baroni
et al. 1994, Braun et al. 1997, Niemelä 2001), and other ethanol metabolites such as malondialde31
hyde and 4-hydroxynonenal are also capable of generating protein adducts (Niemelä 2001). Acetaldehyde adducts may result in altered protein function and play a part in the stimulation of fibrogenesis and the induction of immune responses (Tuma and Sorrell 1987, Niemelä 1999). They have
been found in both the circulation and tissues of alcoholics in increased amounts as compared with
control populations (Stevens et al. 1981, Peterson and Polizzi 1987, Niemelä and Israel 1992, Sillanaukee et al. 1992b, Niemelä 2001). Haemoglobin-acetaldehyde adduct levels seem to correlate
with self-reported ethanol consumption (Sillanaukee et al. 1991a, 1991b, Hazelett et al. 1998) and
have thus been suggested as markers of alcohol intake (Niemelä et al. 1990, Sillanaukee et al.
1991b, Gross et al. 1992, Niemelä and Israel 1992, Sillanaukee et al. 1992b, Lin et al. 1993, Hazelett et al. 1998). They have been found to persist in the blood for 1 to 3 weeks (Niemelä and Israel
1992). The earliest methods for measuring haemoglobin-acetaldehyde adducts were based on chromatography and isoelectric focusing (Stevens et al. 1981, Huisman et al. 1983, Nguyen and Peterson 1984, Gordis and Herschkopf 1986, Sillanaukee and Koivula 1990), and antibody-based assays
have been developed more recently (Lin et al. 1990, Niemelä et al. 1990, Niemelä and Israel 1992).
Of the adducts localized in tissues, liver adducts have been the most intensively studied, and
their association with alcohol-induced liver damage has been well established (Niemelä et al. 1991,
Holstege et al. 1994, Paradis et al. 1996, Li et al. 1997, Niemelä 2001). There is only limited information on adducts in other tissues, but other potential sites of adduct formation are the alimentary
tract and pancreas (Iimuro et al. 1996, Salmela et al. 1997, Biewald et al. 1998, Niemelä 2001). Adducts have also been localized in the muscles and brain (Rintala et al. 2000, Worrall et al. 2000,
Niemelä 2001, Worrall et al. 2001) and in erythrocyte cell membranes (Niemelä and Parkkila
Acetaldehyde adducts are immunogenic and may trigger antibody responses that appear to be
specific to acetaldehyde-modified epitopes and not to depend on the carrier protein (Niemelä et al.
1987). Class IgG, A and M antibodies against acetaldehyde adducts have been found in the
circulation of alcohol consumers (Israel et al. 1986, Niemelä et al. 1987, Hoerner et al. 1988, Israel
et al. 1988, Izumi et al. 1989, Worrall et al. 1991, Koskinas et al. 1992, Tuma and Klassen 1992,
Worrall et al. 1994, Viitala et al. 1997). Class IgA antibodies against acetaldehyde adducts have
proved to be the most promising for differentiating between excessive drinkers and control individuals, the highest anti-adduct IgA values being observed in alcoholics with liver disease (Hoerner
et al. 1988, Izumi et al. 1989, Worrall et al. 1991, Koskinas et al. 1992, Viitala et al. 1997), although those without liver disease also show elevated values as compared with non-drinkers (Worrall et al. 1994, 1996, Viitala et al. 1997, Worrall et al. 1998). IgG and IgM responses also show
changes after ethanol ingestion, but not as distinctly as IgA (Worrall et al. 1991, 1994, Viitala et al.
It is still unclear whether the antibody responses to acetaldehyde adducts reflect protective or
harmful effects. It has been suggested that the generation of these responses may contribute to the
pathogenesis of alcoholic liver disease (Niemelä et al. 1987, Israel et al. 1988, Izumi et al. 1989),
but it is also possible that the antibodies may serve as neutralizing factors by binding and removing
acetaldehyde adducts from the circulation (Israel et al. 1988). It has been speculated that immunological damage to tissues may develop in the presence of both high antibody levels and continuous adduct formation i.e. continuous alcohol intake (Israel et al. 1988).
2.9.2. Alcohol-induced cytokine responses
The involvement of cytokine responses in alcohol-related liver injury has been widely studied in
recent years (Khoruts et al. 1991, Deviere et al. 1992, McClain et al. 1993, Border and Noble 1994,
Peters 1996, Yin et al. 1999, Neuman 2003, Bode and Bode 2005). Cytokines are a group of proteins that modulate immune responses and participate in cell growth and differentiation and they
can be classified as either pro-inflammatory or anti-inflammatory, although they may have overlapping or multiple functions depending on the cell type. An important function of cytokines is to recruit the cells of the immune system to a site of inflammation (McClain et al. 1997). Cytokines include several classes of molecules: interleukins (IL), tumor necrosis factor (TNF), transforming
growth factor (TGF) and interferons (IFN). The currently known family of cytokines has over 100
members altogether, working in concert to produce either anti-inflammatory or pro-inflammatory
effects depending on the target cells and organs.
Under normal circumstances the liver tissue produces only minimal levels of cytokines (Neuman 2003), but when the liver suffers damage, e.g. due to excessive alcohol consumption, their production increases and the resulting cytokines mediate the regeneration of liver tissue (Peters 1996).
This process includes an inflammatory response, which is required for normal tissue healing, and
the cytokine levels return to baseline levels when the inflammation is under control (Neuman 2003).
It appears that acute moderate alcohol intake leads to inhibition of this inflammatory process
(through increases in anti-inflammatory and decreases in pro-inflammatory cytokine production),
while chronic alcohol consumption shifts the cytokine balance towards persistent inflammation
(Crews et al. 2006). Patients with alcoholic liver disease appear to have elevated pro-inflammatory
(e.g. IL-1, IL-6, TNF-α, and IL-8) and lowered anti-inflammatory (IL-10, IL-4) cytokine levels
(McClain et al. 1997, Cook 1998, Crews et al. 2006). The continuous presence of pro-inflammatory
cytokines and inflammation in the liver (alcoholic hepatitis) may ultimately lead to scar tissue formation (fibrosis) and cirrhosis (Neuman 2003). Apoptosis has also been shown to contribute to the
tissue damage that occurs in alcoholic liver disease and this process may be induced by ethanol (Casey et al. 2001, Neuman 2003).
The increased intestinal permeability observed in alcoholics has been shown to lead to elevated
bacterial lipopolysaccharide (LPS) concentrations in the circulation, whereupon the resulting endotoxaemia may lead to the activation of liver macrophages and induces them to produce proinflammatory cytokines. This cascade leads to inflammation in the liver and may initiate or promote
the pathogenesis of alcoholic liver disease (Wheeler 2003, Bode and Bode 2005). The LPS pathway
may in fact represent a significant contributor to the onset of alcoholic liver disease (Bode and Bode
3. Aims of the present research
Despite advances in methods of detecting excessive alcohol consumption, the incidence of alcoholism and its associated medical, social and economic problems continue to increase in many Western
societies. Accurate screening tools and a better understanding of the primary mechanisms involved
in the adverse health effects of alcohol are clearly needed.
The aims of the present work were as follows:
To study the influence of the reference population on the reference intervals of gammaglutamyltransferase and other markers of excessive alcohol consumption.
To study the clinical behaviour of the marker combination GGT-CDT in assessing alcohol
To examine the association between immune responses to acetaldehyde adducts, alcoholinduced liver disease and alcohol consumption.
To develop a new marker for excessive drinking based on a specific immune response to acetaldehyde adducts.
4. Materials and methods
The excessive drinkers included in these studies were alcoholics who were recruited from a detoxification and alcoholism treatment unit. They had all been classified for alcohol dependence by reference to the DSM-IV criteria and had a well-documented history of excessive drinking. Detailed
interviews on their alcohol consumption were carried out using a timeline follow-back method, according to which the patients were asked how many drinks of alcohol (standard drink = 12 g of
ethanol) they had consumed during the past 24 h, the past week and the past 4 weeks before admission. The mean duration of abstinence prior to sampling was 2 ± 2 days. The patients with alcoholic
liver disease had all had a history of continuous ethanol consumption for at least five years (> 80
g/day) and the liver disease had been assessed by a previously established combined morphological
index (CMI) or a combined clinical and laboratory index (CCLI) (Orrego et al. 1983, Blake and
Orrego 1991). All the excessive drinkers were negative for hepatitis B antigen or hepatitis C serology. The reference population consisted of apparently healthy abstainers and moderate drinkers
who had mostly been recruited from hospital personnel and who had no previous social or medical
history of excessive drinking. The mean daily alcohol consumption of the moderate drinkers, as
assessed with questionnaires, was 1–40 g. Blood sampling was performed by trained laboratory
personnel. The analyses of liver enzymes and blood cell counts were performed immediately and
the serum samples separated by centrifugation were then stored at –70 °C until used for additional
analyses. The participants gave their informed consent and the research was carried out in accordance with the provisions of the Declaration of Helsinki.
The subjects in paper I included 195 individuals, comprising 103 alcoholics (90 men, 13
women) and 92 reference individuals (54 men, 38 women). The population of alcoholics had no
signs of liver damage and their mean consumption of ethanol had been 40–539 g per day during the
past 4 weeks. The reference population included 30 abstainers and 62 moderate drinkers with a
mean daily ethanol consumption of 1–40 g. For the comparisons between moderate drinkers and
abstainers, additional data were included from a survey on 2485 apparently healthy individuals
(1174 men, 1311 women) collected for the Nordic Reference Interval Project, as kindly provided by
the project coordinator, Professor Pål Rustad, Fürst Medical Laboratory, Oslo, Norway. These subjects were either abstainers (n = 1156: 471 men, 685 women) or moderate drinkers (n = 1329: 703
men, 626 women), the maximum alcohol consumption during the twenty-four hours prior to sampling having been 24 g (two standard drinks). Weekly alcohol consumption was assessed by means
of questionnaires employing categories of a) 0 drinks, b) 1–21 drinks and c) over 21 drinks, and
category c was excluded from the analyses. The survey also excluded individuals who had clinical
or laboratory evidence of current or recent illnesses or infections, were pregnant, had donated blood
during the past five months or had taken any prescription drugs during the preceding week. Smoking had not been allowed for one hour prior to sampling.
Paper II reports on a sample of 165 alcoholics (140 men, 25 women) with a mean daily ethanol
consumption of 40–540 g during the past month. They were further classified according to liver
status, with 51 patients (38 men, 13 women) designated as having liver disease. In order to assess
marker normalization rates, follow-up tests with supervised abstinence periods of 11 ± 4 days (confirmed with repeated breath analyses) were carried out on 44 alcoholics (39 men, 5 women). The
reference population comprised 86 healthy volunteers (49 men, 37 women), including 35 abstainers
and 51 moderate drinkers.
Paper III consisted of 86 male alcoholic patients (mean age 48 ± 12 years), of whom 54 had biopsy-proven liver disease and 32 had no clinical or laboratory evidence of significant liver disease.
A follow-up with supervised abstinence for 8 ± 2 days, controlled by means of hospitalization and
repeated breath analyses was carried out on 17 patients. The reference population consisted of 20
apparently healthy male volunteers who were either abstainers (n = 6) or moderate drinkers (n = 14)
with a mean daily ethanol consumption of 20 g (a maximum of 60 grams on any one occasion).
The subjects in paper IV consisted of 40 male alcoholics who had no clinical or laboratory signs
of liver disease, but had a history of continuous ethanol consumption or binge drinking, the mean
recent consumption having been 40–540 g/day for a period of 4 weeks prior to sampling. 19 alcoholics (mean age 42 ± 12 years) volunteered for a follow-up, which was carried out with supervised
abstinence over a period of 8 ± 3 days. In addition, there were 41 male reference individuals who
were either abstainers (n = 16) or moderate drinkers (n = 25), none of whom had any history or
clinical evidence of alcohol abuse, recent illnesses or immunological disorders.
4.2. Measurements of laboratory markers (I-II, IV)
CDT concentrations were measured using a turbidimetric immunoassay (TIA) after ion exchange
chromatography (%CDT, Axis-Shield, Oslo, Norway). This assay detects primarily asialo-,
monosialo- and disialotransferrin, although there may be some reactivity to the trisialo fraction of
CDT, as recently reported by Aldén and co-workers (2005). The %CDT results are expressed as
percentages of total transferrin. The measurements were carried out on a Behring Nephelometer II
(Dade Behring, Behring Diagnostics GmbH, Marburg, Germany). The within-run precision of the
assay was 4.7 %, day-to-day variation 6.0 % and accuracy 12.7 %.
Serum gamma glutamyl transferase (GGT), serum aspartate aminotransferase (AST), serum
alanine aminotransferase (ALT), alkaline phosphatase, albumin, bilirubin and mean corpuscular
volume (MCV) of erythrocytes were measured by standard clinical chemical methods in an accredited (SFS-EN ISO/IEC 17025) laboratory at Seinäjoki Central Hospital, Finland. The assay characteristics for these were as follows (assay; within-run variation, day-to-day variation, accuracy):
GGT; 0.85 %, 0.54 %, 5.0 %, AST; 0.90 %, 1.5 %, 10.6 %, ALT; 0.87 %, 1.1 %, 12.1 %, alkaline
phosphatase; 2.0 %, 1.9 %, 10.1 %, albumin; 2.3 %, 2.0 %, 11.2 %, bilirubin; 2.2 %, 2.4 %, 13.4 %
and MCV; 0.41 %, 0.37 %, 3.7 %. The cut-offs for these parameters were as follows: GGT < 80 U/l
(men), < 50 U/l (women); AST and ALT < 50 U/l (men), < 35 U/l (women); alkaline phosphatase
60–275 U/l; albumin 36–50 g/l; bilirubin 2–20 µmol/l; MCV 76–96 fl.
4.3. Determination of marker combinations (II, IV)
4.3.1. GGT-CDT (II)
The original equation for combining GGT and CDT in the present manner was suggested by Sillanaukee and Olsson (2001), who used absolute CDT values (U/l). These were later replaced with
%CDT results (Anttila et al. 2003). In the present work the combined GGT-CDT marker was calculated in the latter manner (Anttila et al. 2003): GGT-CDT = 0.8 x ln(GGT) + 1.3 x ln(%CDT). The
assay cut-off values were determined with ROC analyses using Analyse-It for Microsoft Excel
software, which yielded a GGT-CDT cut-off of 4.18 for men and 3.81 for women.
Several equations for combining %CDT and anti-adduct IgA results were tested for differentiating
between alcoholics and the control population. The IgA-CDT marker was eventually calculated
according to an equation logCDT + (IgA/103), which appeared to be superior to the others tested.
The cut-off value was 0.412, as determined by ROC analysis.
4.4. Measurement of antibodies against acetaldehyde adducts (III-IV)
The test antigen was prepared by purifying a human erythrocyte fraction from the EDTA-blood of
an abstainer. The cells were separated by centrifugation and washed three times with phosphatebuffered saline (PBS: 7.9 mM Na2HPO4, 1.5 mM KH2PO4, 137 mM NaCl, 2.7 mM KCl, pH 7.4).
They were then haemolyzed with polyoxyethylene ether, 0.1 % V/V in borate buffer (Hemolysis
Reagent, DIAMATTM Analyzer system, Bio-Rad) and incubated for 35 min at 37 °C to remove unstable Schiff bases. The haemolyzate was diluted with PBS to contain 12 mg/ml of protein, dialysed
twice against PBS and then mixed with acetaldehyde (in PBS) so that the final concentration of the
latter in the reaction mixture was 10 mM. The mixture was allowed to react overnight (18 h) at 4 °C
in a tightly sealed container and the adducts generated were reduced by the addition of sodium
cyanoborohydride (10 mM) and mixed for 5 h at 4 °C. The reduced protein solutions were then dialysed twice against PBS at 4 ºC and stored in aliquots for single use at –70 ºC. The unmodified
proteins were prepared and treated similarly to the modified proteins except for the addition of acetaldehyde.
For the measurement of antibody titres, microtitre plates (Nunc-Immuno Plate, MaxisorpTM, InterMed, Denmark) were coated with acetaldehyde-modified red cell protein or corresponding unmodified protein in PBS (3 µg protein in 100 µl/well) and incubated for 1½ h at +37 °C. Nonspecific binding was blocked by incubation with 0.2 % gelatin in PBS (150 µl/well) for 1 h at +37
°C. The samples were diluted (1:40) in PBS containing 0.04 % Tween 20 (PBS-Tween) and 50 µl
of the diluted serum was allowed to react with the coated proteins (1 h, +37 °C). The plates were
then washed extensively with PBS-Tween. Alkaline phosphatase -linked goat anti-human immunoglobulins IgA, IgG, or IgM (Jackson ImmunoResearch Laboratories, Inc., West Grove) were
used to detect antibody-antigen complexes. The immunoglobulins (50 µl/well) were diluted in PBSTween containing 8 mM MgCl2 and a small amount of dithiothreitol (DTT). The plates were incubated overnight at +4 °C and washed with PBS-Tween. After washing, 100 µl of pnitrophenylphosphate solution was added as a colour reaction substrate (Alkaline Phosphatase Substrate Kit, Bio-Rad Laboratories, Hercules, CA). This reaction was stopped by adding 100 µl NaOH
(0.4 M) and the optical densities (ODs) were read at 405 nm with an Anthos HTII microplate reader
(Anthos Labtec Instruments, Salzburg, Austria). The anti-adduct IgA results are expressed as
units/litre (U/l), corresponding to OD405 nm x 103.
4.5. Serum cytokines
The concentrations of serum cytokines (IL-2, IL-6, IL-8, IL-10, TNF-α and TGF-β1) were determined using Quantikine high sensitivity ELISA kits (R&D Systems Inc., Minneapolis, USA). The
results are presented in U/l units, which correspond to OD(450 nm–540 nm) x 103. The analytical characteristics, as given by the assay manufacturer, were as follows (assay; average within-run variation,
average day-to-day variation, detection limit): IL-2; 3.1 %, 4.6 %, 7 pg/ml, IL-6; 2.6 %, 4.5 %, 0.70
pg/ml, IL-8; 5.8 %, 7.7 %, 3.5 pg/ml, IL-10; 3.7 %, 6.9 %, 3.9 pg/ml, TNF-α; 4.7 %, 5.8 %, 1.6
pg/ml, TGF-β1; 5.3 %, 11.0 %, 7 pg/ml. All the assays were linear within their dynamic range.
4.6. Statistical methods
The statistical analyses were carried out using GraphPad Prism (GraphPad Software Inc., San
Diego, CA, USA) and SPSS 13.0 (SPSS Inc., Chicago, IL, USA) and the ROC analyses with Analyse-it software (Analyse-it Software Ltd, Leeds, U.K.).
The t-test was used for comparisons between two groups following a Gaussian distribution, and
the Mann-Whitney test was used for non-Gaussian populations. For comparisons of three or more
groups, a one-way analysis of variance (ANOVA) with Bonferroni’s post hoc test was employed
when the data followed Gaussian distribution and passed the homogeneity of variance test, while
the Kruskal-Wallis test with an appropriate post hoc test was used if a Gaussian distribution or
equal variances in the groups could not be obtained from the original data or from the logarithmically transformed data.
Correlations were calculated using Pearson product-moment correlation coefficients for continuous non-skewed parameters or Spearman’s rank correlations for non-continuous or skewed parameters, as required. A p-value of less than 0.05 was considered statistically significant.
Cut-off values for new markers were either determined with ROC-analyses or calculated as
mean + 2 SD of the control group.
5.1. Influence of drinking levels on markers (I, II, IV)
In paper I, GGT was measured in a large population of alcoholics, moderate drinkers and abstainers.
The concentration in alcoholics consuming 40–80 g (68 ± 54 U/l) or > 80 g (167 ± 254 U/l) of ethanol per day significantly exceeded those in both the abstainers (p < 0.001) and the moderate drinkers (p < 0.001) consuming ethanol in amounts ranging from 1 to 40 g per day. GGT activities were
also found to be significantly higher in the moderate drinkers (28 ± 23 U/l) than in the abstainers
(24 ± 17 U/l) (p < 0.001). When experimental reference intervals were calculated from these populations, the upper limits were found to be on average 43 % higher when the moderate drinkers were
taken as the source than when calculated for the abstainers. Gender and age were also found to have
an effect on the reference intervals; the limits for men were higher than for women and the age
group > 40 years had higher levels than the age group 18–39 years in both genders. The fluctuation
in reference intervals due to the choice of reference population was found to have a significant effect on the sensitivity of GGT, so that 69 % of the heavy drinkers were identified when contrasted
with the abstainers, as opposed to 56 % with the moderate drinkers.
The inclusion of moderate drinkers in the reference population was also found to lead to decreases in the sensitivity of MCV and AST whereas GGT-CDT and CDT were not affected. The
anti-adduct IgA responses studied in paper IV were also found to be affected by moderate drinking,
as these values were significantly higher in the moderate drinkers than in the abstainers (p < 0.05).
5.2. Clinical characteristics of marker combinations (II, IV)
5.2.1. GGT-CDT (II)
The performance of a mathematical combination of GGT and CDT, assessed in paper II, was found
to be superior to that of the conventional laboratory markers in detecting excessive drinking. GGTCDT reached a sensitivity of 90 % (for a specificity of 98 %), whereas the sensitivity of CDT alone
remained at 63 % and that of GGT at 58 %. The sensitivities and specificities of GGT-CDT were
high for both men and women. Combining GGT and CDT in a manner which gave a positive result
when either of the markers was positive obviously yielded a higher sensitivity (85 %) than either
assay alone, but did not reach the sensitivity of the mathematically formulated combination. Liver
status was not found to complicate the assessment of alcohol consumption with GGT-CDT, the sensitivities being quite similar for both alcoholics with (93 %) and without (88 %) liver disease, while
the performance of GGT, MCV, AST and ALT was clearly dependent on liver status. Elevation of
the GGT-CDT marker was estimated to require a threshold ethanol consumption of 40 g per day,
and GGT-CDT was found to correlate more closely with self-reported ethanol consumption (r =
0.76, p < 0.001) than either GGT (r = 0.71, p < 0.001) or CDT (r = 0.59, p < 0.001) alone or any of
the other markers.
5.2.2. IgA-CDT (IV)
When the possibility of using anti-adduct IgA results in marker combinations was examined in paper IV, the highest sensitivities and specificities were obtained by combining these with CDT in a
mathematical equation. This approach yielded a sensitivity of 90 %, a specificity of 98 % and an
AUC of 0.966. Preliminary analyses suggest that IgA-CDT may to be comparable or even superior
to GGT-CDT (Figure 1).
Figure 1. Sensitivity of laboratory markers and their combinations in detecting excessive drinking.
IgA-CDT, a combination derived from the equation logCDT + (IgA/103); GGT-CDT, a combination
derived from the equation 0.8 x ln(GGT) + 1.3 x ln(%CDT); IgAs, immunoglobulin A against acetaldehyde adducts; CDT, carbohydrate-deficient transferrin; GGT, gamma-glutamyl transferase;
MCV, mean corpuscular volume of erythrocytes; AST, aspartate aminotransferase.
5.3. Antibody and cytokine responses in alcoholics (III–IV)
5.3.1. Alcohol-related changes in cytokine and antibody production (III-IV)
Alterations in anti-adduct immunoglobulin levels were found in alcoholics, particularly in those
with liver disease. The anti-adduct titres were measured for immunoglobulin classes IgA, IgG and
IgM in paper III, while paper IV focused on the characteristics of anti-adduct IgA as a marker of
excessive drinking. The highest anti-adduct IgA and IgG titres were found in the patients with alcoholic liver disease (paper III), the differences relative to the moderate drinkers and the abstainers
being significant (p < 0.001), while the alcoholics without any apparent liver disease also showed
elevated IgA levels as compared with both control groups (p < 0.01). IgM titres were highest in the
alcoholics without liver disease, and the values for the moderate drinkers were also higher than for
the abstainers (p < 0.05), as was also the case for class IgA antibodies. There were no significant
differences in total serum immunoglobulin levels between the groups.
Cytokine profiles were measured in paper III in order to assess the effects of alcohol intake and
liver disease on the regulation of inflammation. The alcoholics with liver disease showed markedly
increased levels of pro-inflammatory cytokines IL-2, IL-6, IL-8 and TNF-α, and decreased TGF-β1,
while those without liver disease had normal or slightly elevated pro-inflammatory cytokine levels,
except for IL-6, which was high in both alcoholic groups as compared with the control groups (p <
0.001). IL-10 levels were higher in all the alcohol-consuming groups than in the abstainers.
Anti-adduct IgA values were found to correlate significantly (p < 0.001) with recent ethanol
consumption (r = 0.77), as also were IL-6 levels (r = 0.83, p < 0.001).
5.3.2. Anti-adduct IgAs as a marker of alcohol consumption (IV)
When anti-adduct IgAs and various traditional alcohol markers were measured in a population
comprising alcoholics, moderate drinkers and abstainers, the alcoholics showed elevated levels as
compared with either the moderate drinkers (p < 0.001) or the abstainers (p < 0.001). The diagnostic
characteristics of anti-adduct IgAs for detecting excessive drinking were compared with those of the
conventional markers CDT, GGT, MCV and AST by contrasting the data obtained for the alcoholics with those obtained for the abstainers and/or moderate drinkers. The sensitivity of anti-adduct
IgAs using abstainers as the reference population was 73 % for a specificity of 94 %, which exceeded the corresponding figures obtained for CDT, GGT, MCV and AST. When moderate drinkers
were also included in the reference population, the sensitivity of the anti-adduct IgA assay was 65
% for a specificity of 88 %, showing essentially similar characteristics to CDT while remaining
superior to GGT, MCV or AST. The sensitivity of anti-adduct IgAs can be further improved by
combining it with CDT (see section 5.2.2.).
The anti-adduct IgA values showed significant correlations with all the conventional markers of
alcohol consumption and also with actual recent consumption. Anti-adduct IgAs were found to cor-
relate significantly with serum IL-6 (r = 0.33, p < 0.05) and TNF-α (r = 0.31, p < 0.05), but not with
IL-2, IL-8, IL-10 or TGF-β1.
5.4. Markers in the follow-up of alcoholics (II-IV)
Marker normalization was examined in papers II and IV, and changes in cytokine levels were also
addressed in paper III. GGT-CDT was found to decrease in 93 % of alcoholics during 11 ± 4 days
of supervised abstinence, and the normalization rate was estimated to be 18 ± 9 days, depending on
the initial value. GGT-CDT and CDT showed more consistent declines during abstinence than did
GGT, MCV, AST or ALT.
Anti-adduct IgG ad IgM values in the alcoholics did not show any significant changes upon abstinence, but a significant decrease was observed in anti-adduct IgA levels (papers III and IV),
which also remained higher in the alcoholics after abstinence than in the moderate drinkers or the
abstainers. Assessment of the normalization rate of anti-adduct IgAs in paper IV showed it to be
about 3 % per day, with a mean normalization time of about four weeks, depending on the original
Significant changes were also noted in the levels of cytokines during abstinence, the changes in
IL-6 (–47 %), IL-10 (–82 %) and TNF-α (–41 %) being significant, whereas IL-2 (± 0 %), IL-8 (–26
%) and TGF-β1 (+3 %) remained relatively stable. IL-6 levels still remained higher after abstinence
than those in the control groups (p < 0.05 for moderate drinkers; p < 0.001 for abstainers).
6.1. Influence of moderate drinking on marker levels (I-IV)
The present data indicate that moderate drinking has an influence on several biomarkers of alcohol
consumption, which could lead at the population level to changes in reference intervals and subsequently to problems in recognizing excessive alcohol consumption in its early phase. It would therefore be essential to define reference intervals on the basis of abstainers rather than moderate drinkers. In order to obtain universally comparable policies, the concept of moderate drinker should also
be defined more accurately in terms of both drinking pattern and alcohol intake.
Serum GGT concentrations were observed to be significantly elevated in alcoholics and also in
moderate drinkers as compared with the abstainers (paper I), and the estimated upper normal limits
for GGT would be approximately 40 % higher if moderate drinkers were used as the reference
population instead of abstainers. This situation was also reflected in a recent NORIP survey produced in the Nordic countries, which showed markedly increased GGT reference values as compared with the limits used previously (Stromme et al. 2004). Such a change in GGT limits would
obviously lead to decreased sensitivity in the detection of alcoholics, as the present data suggest that
13 % of them would escape detection if moderate drinkers were taken as the reference population
instead of abstainers. At the same time, it can be argued that with lower GGT limits the specificity
of the assay would decrease, leading to false positive values, as approximately 11 % of the moderate drinkers studied here would have been labelled as having elevated values with lower limits. This
aspect may also be viewed, of course, as a possibility for detecting excessive drinking in an earlier
phase, for these reference individuals with slightly elevated GGT may report moderate drinking but
actually represent drinkers who are on the edge of excessive alcohol consumption or have other
confounding factors such as obesity and diabetes, which can both cause elevated GGT values
(Kornhuber et al. 1989, Cushman 1992, Aubin et al. 1998, Daeppen et al. 1998, Puukka et al.
The differences between moderate drinkers and abstainers were also investigated in the case of
the GGT-CDT, CDT, MCV, AST and ALT (paper II), and the inclusion of moderate drinkers in the
reference population was found to affect the diagnostic performance of GGT, AST and MCV but
not GGT-CDT. This may be an important consideration when screening for excessive alcohol consumption.
Papers III and IV focused on the immunological responses to alcohol and its metabolites. These
responses were also found to show variation as a result of moderate drinking. Anti-adduct IgA values were found to be significantly elevated in moderate drinkers by comparison with abstainers, and
the other anti-adduct immunoglobulins, IgG and IgM, were also found to show variation between
moderate drinkers and abstainers, the difference in IgM being significant. For cytokines, anti43
inflammatory IL-10 values appeared to be slightly elevated in moderate drinkers. It could be speculated that both the IgM response and IL-10 elevation indicate an early response against the adverse
effects of alcohol.
6.2. Marker combinations (II, IV)
Although various markers for detecting excessive alcohol consumption have emerged during past
decades, clinicians continue to lack a sensitive tool for the identification of excessive drinking, especially in its early phase. Since the passage of a new marker molecule from its discovery to routine
use can take years or even decades, increasing interest has been shown in the idea of combining
existing markers. Various marker combinations have been introduced during the past two decades
(Salaspuro 1987, Anton et al. 2001, Sillanaukee and Olsson 2001, Anttila et al. 2003, Hock et al.
2005). Traditionally, combinations have included CDT, GGT and/or MCV, usually combined in a
manner in which a case is deemed positive when either of the markers is positive. The combining of
GGT and CDT in this way, for instance, has been shown to yield a sensitivity of 90 % for a specificity of 81 % in men and a sensitivity of 75 % for a specificity of 87 % in women (Anton et al.
2001). More recently, Schwan and co-workers (2004) have shown that combining GGT and CDT as
independent parameters provides a sensitivity of 90 % in alcohol abusers and 99 % in alcoholdependent subjects, whereas the specificity remains at a level of only 63 %. A further improvement
in marker combinations has been achieved by using mathematical equations, the most promising
combination of this kind so far being GGT and CDT with an equation of 0.8 x ln(GGT) + 1.3 x
ln(%CDT) (Sillanaukee and Olsson 2001, Anttila et al. 2003). In line with this view, Hock and coworkers (2005) recently reported a sensitivity of 83 % with a specificity of 95 % using a simple
combination of log GGT and CDT. When MCV was included as a third component in these analyses, the sensitivity was reported to increase to 88 %.
The data presented in papers II and IV clearly support the idea of using marker combinations for
detecting hazardous drinking. The mathematically formulated equation based on serum GGT and
CDT results, referred to in paper II as GGT-CDT was first introduced by Sillanaukee and Olsson
(2001), who used absolute CDT values (U/l) in their equation. The replacement of these with
%CDT results has proved to improve the performance of this approach still further (Anttila et al.
2003). The present comparison of the diagnostic characteristics of GGT-CDT with those of various
other biochemical markers used in the diagnosis of excessive drinking (II) showed this combination
to be superior to all the conventional markers. One remarkable thing is that this improvement in
sensitivity was achieved without sacrificing the specificity. The use of MCV or any other parameter
as a third component in the combination was not found to lead to any additional improvement over
GGT-CDT alone. GGT-CDT also appears to be relatively independent of liver status, despite having GGT as one of its components. Earlier observations suggest that using %CDT in the equation
instead of absolute CDT further improves this property (Anttila et al. 2003). On the other hand, the
presence of liver pathology was found to affect the performance of GGT, AST, ALT and MCV in
the assessment of excessive drinking.
Paper IV introduces briefly a marker combination consisting of anti-adduct IgAs and %CDT,
combined by means of a mathematical equation logCDT + (IgA/103). The resulting marker, IgA-
CDT, was found to provide improved sensitivity as compared with its parent components or other
alcohol markers, without any loss in assay specificity. The diagnostic performance of IgA-CDT also
seems to be similar to or even better than that of GGT-CDT.
Ideal markers for combinations should have different mechanisms of induction, so that the combination will gain the diagnostic benefits of both components. GGT and CDT, for instance, frequently increase in different individuals and may represent different types of ethanol-induced
pathophysiological processes (Anton et al. 2002, Anttila et al. 2003, Neumann and Spies 2003).
This combination of different diagnostic windows could partly explain the distinct improvement
that can be obtained with marker combinations. Taken together, this work supports the idea of employing marker combinations to achieve a more sensitive diagnosis of excessive drinking. Antiadduct IgA measurements are not currently available for clinical use, but since GGT and CDT already belong to the set of routine assays, GGT-CDT would be cost-effective and easy to manage in
6.3. Antibody and cytokine responses in alcohol abusers (III-IV)
An understanding of the pathophysiological processes related to alcohol is equally as important as
the discovery of a reliable laboratory marker. If the mechanisms of tissue pathology could be clarified, it might provide new tools for effective treatment and enable the use of immune parameters as
biomarkers of alcohol consumption and alcoholic liver disease. Alcohol affects almost every tissue
in the body, and our knowledge of the mechanisms of these actions is limited. The present study of
the antibody and cytokine responses related to alcohol intake (III) and the possibility of using antibodies as markers of alcoholism (IV) was focused on male subjects because the immunological responses in vivo may show significant gender dependence, possibly due to differences in sex hormones (Kovacs and Messingham 2002). Women generally show stronger immune responses, and
thus further studies with female populations are clearly warranted.
6.3.1. Changes in antibody and cytokine production attributable to alcohol and liver
Some of the adverse consequences of alcohol consumption are clearly mediated by immunological
mechanisms. The antibody and cytokine responses studied here seem to be associated with inflammatory processes in tissues, particularly in the liver. The generation of immune responses to acetaldehyde-modified epitopes appears to occur early in the sequence of events leading from excessive
alcohol consumption to clinical signs of alcoholic liver injury. Both anti-adduct IgA and IgM already showed an elevation in alcoholics without liver disease, and IgG antibodies also parallelled
the evolution of liver injury. These findings, together with the observation that cytokine levels also
respond to ethanol intake, suggest that the balance in the immune regulation of tissues may be disturbed by ethanol.
Previous studies focusing on alcoholic liver disease have reported an increase in circulating total
IgA concentrations together with increased deposition of IgA in tissues (Swerdlow et al. 1983, van
de Wiel et al. 1987, 1988, Tuma and Klassen 1992, Amore et al. 1994). It has been suggested that
this may result from either decreased IgA catabolism or excretion or increased IgA production (van
de Wiel et al. 1987, Koskinas et al. 1992, Tuma and Klassen 1992, Viitala et al. 1997). The present
findings support the theory that the increase in total IgA is due to antigen-driven antibody production, because the levels of specific antibodies become elevated before any generalized increase in
serum total IgA levels.
The specific IgAs against ethanol metabolites may be derived from intestinal B-cells, because
IgA is the predominant antibody in the gastrointestinal tract, and gastric immunity may readily respond to environmental and dietary antigens (Kerr 1990, Amore et al. 1994). The epithelial tissues,
which are rich in enzymes capable of metabolizing ethanol to acetaldehyde, are continuously exposed to ethanol in excessive drinkers, thus enabling the formation of acetaldehyde adducts and
antibodies against them (Seitz et al. 1994, Salaspuro 1996, Salmela et al. 1997, Visapää et al. 1998).
This is also supported by the present observations that there is a close correlation between antiadduct IgA levels and recent ethanol intake, and also by the presence of anti-adduct IgAs in individuals reporting moderate drinking. Excessive ethanol intake has previously been found to increase
intestinal permeability and disrupt mucosal barriers (Bode and Bode 2003). This could further enhance the antibody responses to various gut-derived antigens and also increase the absorption of
IgA. Alcoholics with liver disease often have high endotoxin levels and they have been reported to
produce IgA antibodies to endotoxin (Nolan et al. 1986, Parlesak et al. 2002, Wheeler 2003) and
also to human gut luminal aspirates (Douds et al. 1998). Anti-endotoxin antibodies as well as antiadduct antibodies may contribute to the formation of alcoholic liver disease (Klassen et al. 1995,
Parlesak et al. 2002, Leevy and Elbeshbeshy 2005).
The involvement of cytokine signalling in the progression of alcoholic liver disease has become
more and more evident in the past decades, and the present studies also show a distinctly different
cytokine profiles for alcoholics and for healthy controls. It is not known, however, whether the immune responses in excessive drinkers represent protective or harmful mechanisms for the liver. The
positive correlation found between anti-adduct IgAs and the pro-inflammatory cytokines TNF-α and
IL-6 may be an indication of early-phase inflammatory response to ethanol-derived antigens, as IL6 has been reported to participate in the acute phase hepatic response (Khoruts et al. 1991) and also
in the control of immunoglobulin production (Deviere et al. 1992). The increased amount of antiadduct IgM and IL-10 in moderate drinkers and in alcoholics without liver disease may represent an
early-phase immune response and the regulation of protective immune mechanisms. IgA antibodies
may also contribute to immune protection by excluding and neutralizing the altered protein structures resulting from acetaldehyde modification (Israel et al. 1988). When excessive drinking continues, the anti-adduct IgA increases and the pro-inflammatory cytokines become even more elevated,
while anti-inflammatory cytokines show a decline. When antigenic stimulation is excessive, IgA
immune complexes may be damaging, since these can cause monocytes to release mediators of tissue damage (Deviere et al. 1991). IgG antibodies are known as mediators of several immunopathogenic consequences, including complement activation and the induction of cytotoxic reactions.
These immune responses are likely to play an important role in the progression of liver pathology.
Alcoholics with liver disease clearly have a skewed balance in cytokine levels, with distinct elevations in IL-2, IL-6, IL-8 and TNF-α, while TGF-β1 levels are actually decreased, although TGF-β1
has previously been associated with the progression of fibrosis in alcoholics (Chen et al. 2002). An
excessive release of pro-inflammatory cytokines can induce the activation of inflammatory cells,
increase the production of reactive oxygen species from hepatocytes and induce apoptosis (Neuman
2003). According to several studies, increased TNF-α production and Kupffer cell activation may
play key roles in hepatic inflammation (Iimuro et al. 1997, Yin et al. 1999).
6.3.2. Anti-adduct IgAs as a marker of alcohol consumption
It is shown in paper IV that the specific IgA response to acetaldehyde adducts may serve as a clinically useful tool for diagnosing excessive ethanol consumption. A high sensitivity and specificity
was observed for the adduct-specific IgA measurements in differentiating between alcoholics and
non-alcoholics, and a close correlation between alcohol intake and anti-adduct IgA levels was also
found. The diagnostic usefulness of IgA antibodies has previously been studied with a different test
antigen (acetaldehyde-modified bovine serum albumin) and these investigations also gave a significant correlation (r = 0.44) between anti-adduct IgAs and alcohol intake (Worrall et al. 1996, 1998).
Although the anti-acetaldehyde antibody responses are generated independently of the carrier protein, there may be differences in the characteristics of the test antigen and in the efficiency of the
assay in differentiating alcoholics from the control population. The red cell protein (mainly consisting of haemoglobin) appears to suit well for anti-adduct antibody measurements, which may be due
to an efficient modification of the carrier protein, perhaps leading to an adduct resembling the naturally occurring ones.
It should be noted that moderate drinkers show elevated IgA values as compared with abstainers. This has to be taken into consideration when defining normal values for such measurements.
The diagnostic sensitivity of anti-adduct IgAs exceeds that of the conventional markers if abstainers
are used as the reference individuals, but the inclusion of moderate drinkers elevates the reference
limit and causes a slight decrease in assay sensitivity. Nevertheless, measurements of specific IgAs
against acetaldehyde-modified epitopes in proteins could be used as a sensitive tool for detecting
6.4. Follow-up studies
Of the markers studied here, GGT-CDT was found to be most suitable for the follow-up of alcoholics during treatment, since it was observed to decrease in 93 % of the subjects and the mean normalization rate was estimated to be 2–3 weeks. The time required for the normalization of GGTCDT was slightly longer than for either of its components separately, suggesting that this mathematical combination follows a slightly different kinetics in its normalization. In line with previous
studies, CDT performed well in the follow-up, while the normalization of GGT, AST and ALT was
not as consistent as that of either GGT-CDT or CDT alone (Bell et al. 1993, Salaspuro 1999, Anttila
et al. 2004). This is perhaps due to the induction of liver pathology, which may be reflected in
marker levels even after the cessation of drinking.
Anti-adduct IgA antibodies and IL-6 levels were found to decrease significantly in the follow-up
of abstaining alcoholics, and they still remained elevated by comparison with those in healthy controls. Significant changes during abstinence were also observed for IL-10 and TNF-α. The follow-
up studies showed that anti-adduct IgA levels normalize at an average rate of 3 % per day, the mean
time required for normalization being 29 days. Since the average half-life of a single IgA molecule
is 5–6 days, the longer appearance of anti-adduct IgAs in serum may be explained by the fact that
acetaldehyde adducts in the erythrocytes of alcoholics persist in the circulation for 1 to 3 weeks
after the last dose of ethanol (Niemelä and Israel 1992). This consistent decrease in anti-adduct
IgAs during abstinence suggests, however, that this marker could also be used for follow-up purposes.
6.5. Possible limitations of this study
It should be noted that this study contrasted apparently healthy controls with alcoholics, while in
clinical practice the hospital population may comprise patients with various medical conditions and
drinking habits, which may complicate the assessment of alcohol consumption by means of laboratory markers. Data on possible refusers among alcoholics were not collected, but it can be assumed
that the significance of this factor is relatively small in large populations. The control individuals
used here also entail some limitations. The permitted alcohol intake for women was similar to that
for men, which may have led to the inclusion of some female individuals who were in fact on the
edge of excessive consumption rather than true moderate drinkers. Also, as the alcohol consumption
data are based on individuals' own reports, the possibility of having some excessive drinkers among
the male control population as well cannot be excluded.
Most of the methods used in these studies are already in routine use and thus quite well characterized in terms of variation and accuracy as well as clinical usefulness. The anti-adduct IgA assay
has so far been used for research purposes only and corresponding figures for its precision have not
yet been established. We cannot exclude some additional variation in results due to this unstandardized assay method, and this work should therefore be continued by standardization of the assay.
6.6. Future considerations
The markers and combinations presented here appear to provide some improvements in the diagnosis of excessive alcohol consumption. Since the prevalence of excessive drinking continues to increase and frequently still remains undetected in health care, there is an obvious need for new objective tools. The markers that are already in routine use could also be used in combinations for
clinical purposes. In the case of new markers, assay methods need to be standardized both within
and between laboratories, and further studies are needed in clinical settings in order to establish
their diagnostic value.
These findings indicate that the assessment of alcohol consumption in clinical practice could be
improved by using laboratory markers and their combinations in a carefully standardized manner.
Moderate drinking at the population level may induce changes in some of the commonly used
biomarkers of alcohol consumption, such as GGT. This should be taken into consideration in the
clinical use of such markers and when selecting individuals for reference populations. Any ethanolrelated deviation in GGT reference intervals will obviously reduce the sensitivity of this marker in
detecting excessive drinking.
The diagnosis of excessive drinking can be improved by using combinations of markers. The
present data suggest that the diagnostic characteristics of the combined marker GGT-CDT appear to
exceed those of the traditional markers and both of its components separately.
Immune responses to alcohol-induced antigens, together with disturbances in cytokine balance,
may contribute to the development of tissue injury in alcoholics. Changes in ethanol-induced immune responses may already occur in alcoholics without liver disease, however, and even in moderate drinkers. Anti-acetaldehyde adduct IgAs may be suitable for assessing alcohol consumption,
particularly when combined with CDT.
This research was carried out at the Department of Clinical Chemistry and Medical Research Unit,
Seinäjoki Central Hospital, during the years 2004–2007.
I would like to express my sincere gratitude to my supervisor, Professor Onni Niemelä, M.D.,
Ph.D., for the opportunity to work in his research group and for his encouragement during these
years. This thesis would not have come into existence without his wide knowledge of alcohol research and his guidance throughout the work.
Professor Kaija Seppä and Docent Kari Pulkki, the reviewers appointed by the Medical Faculty,
are gratefully acknowledged for their valuable comments and constructive criticism on the thesis. It
was a great honour for me to hear that Docent Arja Rautio had agreed to be the official opponent for
the discussion of the thesis. I also wish to thank Mr. Malcolm Hicks for his revision of the English
language of the manuscript.
I thank Dr. Matti Rekiaro and Docent Hannu Puolijoki for their interest in my thesis and in our
research group, and Jaakko Pihlajamäki, Head of the Hospital District, for his positive attitude towards the research at our department.
I would like to express my sincere thanks to my closest co-workers, Heidi Koivisto and Taana
Sandman, who, in addition to professional collaboration, have supported me with their friendship
during these years. I also wish to express my warmest thanks to my other co-workers and coauthors, Katri Puukka, Jaana Latvala, Petra Anttila, Katja Viitala and Kimmo Järvi, and to all the
members of the staff of Seinäjoki Central Hospital laboratory. Especially I would like to thank my
colleagues Aira Sorto, Kristina Eichholz and Markku Latva-Nevala for their support to a young
specializing hospital chemist.
I owe my deepest gratitude to my parents, who have always believed in me and supported me in
every way throughout my life. I am sincerely thankful to my beloved fiancé, Pekka, for his love and
understanding and for reminding me that there is life outside the academic world. I would also like
to thank all my friends and relatives for their support and for the interest they have shown in my
thesis and in alcohol research in general.
Parts of this research were supported financially by the Finnish Foundation for Alcohol Studies
and the Foundation for the Promotion of Laboratory Medicine.
Seinäjoki, January 2007
Adrian M and Barry SJ (2003): Physical and mental health problems associated with the use of alcohol and drugs. Subst Use Misuse 38:1575-1614.
Aertgeerts B, Buntinx F, Ansoms S and Fevery J (2001): Screening properties of questionnaires and
laboratory tests for the detection of alcohol abuse or dependence in a general practice population. Br
J Gen Pract 51:206-217.
Agarwal DP (2002): Cardioprotective effects of light-moderate consumption of alcohol: a review of
putative mechanisms. Alcohol Alcohol 37:409-415.
Agarwal DP, Tobar-Rojas L, Harada S and Goedde HW (1983): Comparative study of erythrocyte
aldehyde dehydrogenase in alcoholics and control subjects. Pharmacol Biochem Behav 18 Suppl
Aldén A, Ohlson S, Påhlsson P and Rydén I (2005): HPLC analysis of carbohydrate deficient transferrin isoforms isolated by the Axis-Shield %CDT method. Clin Chim Acta 356:143-146.
Allebeck P and Olsen J (1998): Alcohol and fetal damage. Alcohol Clin Exp Res 22:329S-332S.
Allen JP, Litten RZ and Anton R (1992): Measures of alcohol consumption in perspective. In:
Measuring Alcohol Consumption. Psychosocial and Biochemical Methods, pp. 205-226. Eds. RZ
Litten and JP Allen, The Humana Press Ins., Totowa, New Jersey.
Allen JP, Litten RZ, Anton RF and Cross GM (1994): Carbohydrate-deficient transferrin as a measure of immoderate drinking: remaining issues. Alcohol Clin Exp Res 18:799-812.
Allen JP, Sillanaukee P and Anton R (1999): Contribution of carbohydrate deficient transferrin to
gamma glutamyl transpeptidase in evaluating progress of patients in treatment for alcoholism. Alcohol Clin Exp Res 23:115-120.
Allen JP, Litten RZ, Fertig JB and Sillanaukee P (2000): Carbohydrate-deficient transferrin,
gamma-glutamyltransferase, and macrocytic volume as biomarkers of alcohol problems in women.
Alcohol Clin Exp Res 24:492-496.
Allen JP, Litten RZ, Fertig JB and Sillanaukee P (2001): Carbohydrate-deficient transferrin: an aid
to early recognition of alcohol relapse. Am J Addict 10 Suppl:24-28.
Amore A, Coppo R, Roccatello D, Piccoli G, Mazzucco G, Gomez-Chiarri M, Lamm ME and
Emancipator SN (1994): Experimental IgA nephropathy secondary to hepatocellular injury induced
by dietary deficiencies and heavy alcohol intake. Lab Invest 70:68-77.
Anderson P (1993): Management of alcohol problems: the role of the general practitioner. Alcohol
Anton R and Bean P (1994): Two methods for measuring carbohydrate-deficient transferrin in inpatient alcoholics and healthy controls compared. Clin Chem 40:364-368.
Anton RF and Moak DH (1994): Carbohydrate-deficient transferrin and gammaglutamyltransferase as markers of heavy alcohol consumption: gender differences. Alcohol Clin
Exp Res 18:747-754.
Anton RF, Moak DH and Latham P (1996): Carbohydrate-deficient transferrin as an indicator of
drinking status during a treatment outcome study. Alcohol Clin Exp Res 20:841-846.
Anton RF, Dominick C, Bigelow M and Westby C (2001): Comparison of Bio-Rad %CDT TIA and
CDTect as laboratory markers of heavy alcohol use and their relationships with gammaglutamyltransferase. Clin Chem 47:1769-1775.
Anton RF, Lieber C and Tabakoff B (2002): Carbohydrate-deficient transferrin and gammaglutamyltransferase for the detection and monitoring of alcohol use: results from a multisite study.
Alcohol Clin Exp Res 26:1215-1222.
Anttila P, Järvi K, Latvala J, Blake JE and Niemelä O (2003): A new modified gamma-%CDT
method improves the detection of problem drinking: studies in alcoholics with or without liver disease. Clin Chim Acta 338:45-51.
Anttila P, Järvi K, Latvala J and Niemelä O (2004): Method-dependent characteristics of carbohydrate-deficient transferrin measurements in the follow-up of alcoholics. Alcohol Alcohol 39:59-63.
Anttila P, Järvi K, Latvala J, Romppanen J, Punnonen K and Niemelä O (2005): Biomarkers of alcohol consumption in patients classified according to the degree of liver disease severity. Scand J
Clin Lab Invest 65:141-151.
Aradottir S, Asanovska G, Gjerss S, Hansson P and Alling C (2006): Phosphatidylethanol (Peth)
concentrations in blood are correlated to reported alcohol intake in alcohol-dependent patients. Alcohol Alcohol 41:431-437.
Aragon CM, Stotland LM and Amit Z (1991): Studies on ethanol-brain catalase interaction: evidence for central ethanol oxidation. Alcohol Clin Exp Res 15:165-169.
Arndt T (2001): Carbohydrate-deficient transferrin as a marker of chronic alcohol abuse: a critical
review of preanalysis, analysis, and interpretation. Clin Chem 47:13-27.
Arndt T, Korzec A, Bar M and Kropf J (2002): Further arguments against including trisialo-Fe2transferrin in carbohydrate-deficient transferrin (CDT): a study on male alcoholics and hazardous
drinkers. Med Sci Monit 8:CR411-CR418.
Asai H, Imaoka S, Kuroki T, Monna T and Funae Y (1996): Microsomal ethanol oxidizing system
activity by human hepatic cytochrome P450s. J Pharmacol Exp Ther 277:1004-1009.
Aubin HJ, Laureaux C, Zerah F, Tilikete S, Vernier F, Vallat B and Barrucand D (1998): Joint influence of alcohol, tobacco, and coffee on biological markers of heavy drinking in alcoholics. Biol
Auwärter V, Sporkert F, Hartwig S, Pragst F, Vater H and Diefenbacher A (2001): Fatty acid ethyl
esters in hair as markers of alcohol consumption. Segmental hair analysis of alcoholics, social
drinkers, and teetotalers. Clin Chem 47:2114-2123.
Badawy AA (2003): Alcohol and violence and the possible role of serotonin. Crim Behav Ment
Bagrel A, d'Houtaud A, Gueguen R and Siest G (1979): Relations between reported alcohol consumption and certain biological variables in an "unselected" population. Clin Chem 25:1242-1246.
Beck O and Helander A (2003): 5-hydroxytryptophol as a marker for recent alcohol intake. Addiction 98 Suppl 2:63-72.
Bedford D, O'Farrell A and Howell F (2006): Blood alcohol levels in persons who died from accidents and suicide. Ir Med J 99:80-83.
Behrens UJ, Worner TM and Lieber CS (1988): Changes in carbohydrate-deficient transferrin levels after alcohol withdrawal. Alcohol Clin Exp Res 12:539-544.
Bell H, Tallaksen C, Sjaheim T, Weberg R, Raknerud N, Orjasaeter H, Try K and Haug E (1993):
Serum carbohydrate-deficient transferrin as a marker of alcohol consumption in patients with
chronic liver diseases. Alcohol Clin Exp Res 17:246-252.
Bell H, Tallaksen CM, Try K and Haug E (1994): Carbohydrate-deficient transferrin and other
markers of high alcohol consumption: a study of 502 patients admitted consecutively to a medical
department. Alcohol Clin Exp Res 18:1103-1108.
Beresford T, Low D, Hall RC, Adduci R and Goggans F (1982): A computerized biochemical profile for detection of alcoholism. Psychosomatics 23:713-720.
Berglund M and Öjehagen A (1998): The influence of alcohol drinking and alcohol use disorders on
psychiatric disorders and suicidal behavior. Alcohol Clin Exp Res 22:333S-345S.
Bergström J, Helander A and Jones AW (2003): Ethyl glucuronide concentrations in two successive
urinary voids from drinking drivers: relationship to creatinine content and blood and urine ethanol
concentrations. Forensic Sci Int 133:86-94.
Berner MM, Bentele M, Kriston L, Manz C, Clement HW, Harter M and Mundle G (2006): DOVER and QUVER-new marker combinations to detect and monitor at-risk drinking. Alcohol Clin
Exp Res 30:1372-1380.
Biewald J, Nilius R and Langner J (1998): Occurrence of acetaldehyde protein adducts formed in
various organs of chronically ethanol fed rats: An immunohistochemical study. Int J Mol Med
Bisaga A, Laposata M, Xie S and Evans SM (2005): Comparison of serum fatty acid ethyl esters
and urinary 5-hydroxytryptophol as biochemical markers of recent ethanol consumption. Alcohol
Blake J and Orrego H (1991): Monitoring treatment of alcoholic liver disease: evaluation of various
severity indices. Clin Chem 37:5-13.
Blocker JS, Jr. (2006): Did prohibition really work? Alcohol prohibition as a public health innovation. Am J Public Health 96:233-243.
Bode C and Bode JC (2003): Effect of alcohol consumption on the gut. Best Pract Res Clin Gastroenterol 17:575-592.
Bode C and Bode JC (2005): Activation of the innate immune system and alcoholic liver disease:
effects of ethanol per se or enhanced intestinal translocation of bacterial toxins induced by ethanol?
Alcohol Clin Exp Res 29:166S-171S.
Boffetta P and Hashibe M (2006): Alcohol and cancer. Lancet Oncol 7:149-156.
Border WA and Noble NA (1994): Transforming growth factor beta in tissue fibrosis. N Engl J
Borg S, Helander A, Voltaire CA and Hogstrom Brandt AM (1995): Detection of relapses in alcohol-dependent patients using carbohydrate-deficient transferrin: improvement with individualized
reference levels during long-term monitoring. Alcohol Clin Exp Res 19:961-963.
Borges G, Cherpitel C, Orozco R, Bond J, Ye Y, Macdonald S, Rehm J and Poznyak V (2006):
Multicentre study of acute alcohol use and non-fatal injuries: data from the WHO collaborative
study on alcohol and injuries. Bull World Health Organ 84:453-460.
Bortolotti F, De Paoli G and Tagliaro F (2006): Carbohydrate-deficient transferrin (CDT) as a
marker of alcohol abuse: a critical review of the literature 2001-2005. J Chromatogr B Analyt
Technol Biomed Life Sci 841:96-109.
Borucki K, Kunstmann S, Dierkes J, Westphal S, Diekmann S, Bogerts B and Luley C (2004): In
heavy drinkers fatty acid ethyl esters in the serum are increased for 44 hr after ethanol consumption.
Alcohol Clin Exp Res 28:1102-1106.
Boyd JC (1997): Mathematical tools for demonstrating the clinical usefulness of biochemical markers. Scand J Clin Lab Invest Suppl 227:46-63.
Bradford BU, Enomoto N, Ikejima K, Rose ML, Bojes HK, Forman DT and Thurman RG (1999):
Peroxisomes are involved in the swift increase in alcohol metabolism. J Pharmacol Exp Ther
Braun KP, Pavlovich JG, Jones DR and Peterson CM (1997): Stable acetaldehyde adducts: structural characterization of acetaldehyde adducts of human hemoglobin N-terminal beta-globin chain
peptides. Alcohol Clin Exp Res 21:40-43.
Brick J (2006): Standardization of alcohol calculations in research. Alcohol Clin Exp Res 30:12761287.
Brismar B and Bergman B (1998): The significance of alcohol for violence and accidents. Alcohol
Clin Exp Res 22:299S-306S.
Bujanda L (2000): The effects of alcohol consumption upon the gastrointestinal tract. Am J Gastroenterol 95:3374-3382.
Buydens-Branchey L, Branchey MH and Noumair D (1989): Age of alcoholism onset. I. Relationship to psychopathology. Arch Gen Psychiatry 46:225-230.
Carlsson AV, Hiltunen AJ, Beck O, Stibler H and Borg S (1993): Detection of relapses in alcoholdependent patients: comparison of carbohydrate-deficient transferrin in serum, 5-hydroxytryptophol
in urine, and self-reports. Alcohol Clin Exp Res 17:703-708.
Caro AA and Cederbaum AI (2004): Oxidative stress, toxicology, and pharmacology of CYP2E1.
Annu Rev Pharmacol Toxicol 44:27-42.
Casey CA, Nanji A, Cederbaum AI, Adachi M and Takahashi T (2001): Alcoholic liver disease and
apoptosis. Alcohol Clin Exp Res 25:49S-53S.
Chalmers DM, Chanarin I, Macdermott S and Levi AJ (1980): Sex-related differences in the haematological effects of excessive alcohol consumption. J Clin Pathol 33:3-7.
Chaloupka FJ, Grossman M and Saffer H (1998): The effects of price on the consequences of alcohol use and abuse. Recent Dev Alcohol 14:331-346.
Chase V, Neild R, Sadler CW and Batey RG (2005): The medical complications of alcohol use:
understanding mechanisms to improve management. Drug Alcohol Rev 24:253-265.
Chaudhuri JD (2000): Alcohol and the developing fetus--a review. Med Sci Monit 6:1031-1041.
Chen J, Conigrave KM, Macaskill P, Whitfield JB and Irwig L (2003): Combining carbohydratedeficient transferrin and gamma-glutamyltransferase to increase diagnostic accuracy for problem
drinking. Alcohol Alcohol 38:574-582.
Chen WX, Li YM, Yu CH, Cai WM, Zheng M and Chen F (2002): Quantitative analysis of transforming growth factor beta 1 mRNA in patients with alcoholic liver disease. World J Gastroenterol
Chick J, Kreitman N and Plant M (1981): Mean cell volume and gamma-glutamyl-transpeptidase as
markers of drinking in working men. Lancet 1:1249-1251.
Chrostek L, Cylwik B, Szmitkowski M and Korcz W (2006): The diagnostic accuracy of carbohydrate-deficient transferrin, sialic acid and commonly used markers of alcohol abuse during abstinence. Clin Chim Acta 364:167-171.
Coates RA, Halliday ML, Rankin JG, Feinman SV and Fisher MM (1986): Risk of fatty infiltration
or cirrhosis of the liver in relation to ethanol consumption: a case-control study. Clin Invest Med
Collins MA, Nijm WP, Borge GF, Teas G and Goldfarb C (1979): Dopamine-related tetrahydroisoquinolines: significant urinary excretion by alcoholics after alcohol consumption. Science
Conigrave KM, Degenhardt LJ, Whitfield JB, Saunders JB, Helander A and Tabakoff B (2002):
CDT, GGT, and AST as markers of alcohol use: the WHO/ISBRA collaborative project. Alcohol
Clin Exp Res 26:332-339.
Conigrave KM, Davies P, Haber P and Whitfield JB (2003): Traditional markers of excessive alcohol use. Addiction 98 Suppl 2:31-43.
Cook RT (1998): Alcohol abuse, alcoholism, and damage to the immune system--a review. Alcohol
Clin Exp Res 22:1927-1942.
Corrao G, Bagnardi V, Zambon A and Arico S (1999): Exploring the dose-response relationship
between alcohol consumption and the risk of several alcohol-related conditions: a meta-analysis.
Corrao G, Rubbiati L, Bagnardi V, Zambon A and Poikolainen K (2000): Alcohol and coronary
heart disease: a meta-analysis. Addiction 95:1505-1523.
Corrao G, Bagnardi V, Zambon A and La Vecchia C (2004): A meta-analysis of alcohol consumption and the risk of 15 diseases. Prev Med 38:613-619.
Crews FT, Bechara R, Brown LA, Guidot DM, Mandrekar P, Oak S, Qin L, Szabo G, Wheeler M
and Zou J (2006): Cytokines and alcohol. Alcohol Clin Exp Res 30:720-730.
Crombie IK, Pounder DJ and Dick PH (1998): Who takes alcohol prior to suicide? J Clin Forensic
Cunningham RM, Maio RF, Hill EM and Zink BJ (2002): The effects of alcohol on head injury in
the motor vehicle crash victim. Alcohol Alcohol 37:236-240.
Cushman P (1992): Blood and liver markers in the estimation of alcohol consumption. In: Measuring Alcohol Cunsumption. Psychosocial and Biochemical Methods, pp. 135-147. Eds. RZ Litten
and JP Allen, The Humana Press Inc., Totowa, New Jersey.
Daeppen JB, Smith TL and Schuckit MA (1998): Influence of age and body mass index on gammaglutamyltransferase activity: a 15-year follow-up evaluation in a community sample. Alcohol Clin
Exp Res 22:941-944.
Dahl H, Stephanson N, Beck O and Helander A (2002): Comparison of urinary excretion characteristics of ethanol and ethyl glucuronide. J Anal Toxicol 26:201-204.
Damström Thakker K (1998): An overview of health risks and benefits of alcohol consumption.
Alcohol Clin Exp Res 22:285S-298S.
De Oliveira e Silva E, Foster D, McGee HM, Seidman CE, Smith JD, Breslow JL and Brinton EA
(2000): Alcohol consumption raises HDL cholesterol levels by increasing the transport rate of apolipoproteins A-I and A-II. Circulation 102:2347-2352.
Deviere J, Vaerman JP, Content J, Denys C, Schandene L, Vandenbussche P, Sibille Y and Dupont
E (1991): IgA triggers tumor necrosis factor alpha secretion by monocytes: a study in normal subjects and patients with alcoholic cirrhosis. Hepatology 13:670-675.
Deviere J, Content J, Denys C, Vandenbussche P, Le Moine O, Schandene L, Vaerman JP and Dupont E (1992): Immunoglobulin A and interleukin 6 form a positive secretory feedback loop: a
study of normal subjects and alcoholic cirrhotics. Gastroenterology 103:1296-1301.
Dey A and Cederbaum AI (2006): Alcohol and oxidative liver injury. Hepatology 43:S63-S74.
Di Chiara G (1997): Alcohol and dopamine. Alcohol Health Res World 21:108-114.
Diaz LE, Montero A, Gonzalez-Gross M, Vallejo AI, Romeo J and Marcos A (2002): Influence of
alcohol consumption on immunological status: a review. Eur J Clin Nutr 56 Suppl 3:S50-S53.
Doll R (1997): One for the heart. BMJ 315:1664-1668.
Dolman JM and Hawkes ND (2005): Combining the audit questionnaire and biochemical markers
to assess alcohol use and risk of alcohol withdrawal in medical inpatients. Alcohol Alcohol 40:515519.
Douds AC, Lewis DJ, Lim AG, Maxwell JD and Poulton TA (1998): Serum IgA antibodies to human gut luminal aspirates and human liver in alcoholic liver disease. Alcohol Clin Exp Res
Doyle KM, Bird DA, al Salihi S, Hallaq Y, Cluette-Brown JE, Goss KA and Laposata M (1994):
Fatty acid ethyl esters are present in human serum after ethanol ingestion. J Lipid Res 35:428-437.
Doyle KM, Cluette-Brown JE, Dube DM, Bernhardt TG, Morse CR and Laposata M (1996): Fatty
acid ethyl esters in the blood as markers for ethanol intake. JAMA 276:1152-1156.
Dufour MC (1999): What is moderate drinking? Defining "drinks" and drinking levels. Alcohol Res
Dumitrescu RG and Shields PG (2005): The etiology of alcohol-induced breast cancer. Alcohol
Durbec JP and Sarles H (1978): Multicenter survey of the etiology of pancreatic diseases. Relationship between the relative risk of developing chronic pancreaitis and alcohol, protein and lipid consumption. Digestion 18:337-350.
Eckardt MJ, File SE, Gessa GL, Grant KA, Guerri C, Hoffman PL, Kalant H, Koob GF, Li TK and
Tabakoff B (1998): Effects of moderate alcohol consumption on the central nervous system. Alcohol Clin Exp Res 22:998-1040.
Ellison RC (2002): Balancing the risks and benefits of moderate drinking. Ann N Y Acad Sci
Enoch MA (2003): Pharmacogenomics of alcohol response and addiction. Am J Pharmacogenomics
Eriksson CJP (2001): The role of acetaldehyde in the actions of alcohol (update 2000). Alcohol Clin
Exp Res 25:15S-32S.
Eustace LW, Kang DH and Coombs D (2003): Fetal alcohol syndrome: a growing concern for
health care professionals. J Obstet Gynecol Neonatal Nurs 32:215-221.
Ewing JA (1984): Detecting alcoholism. The CAGE questionnaire. JAMA 252:1905-1907.
Fisher Wilson J (2003): Should doctors prescribe alcohol to adults? Ann Intern Med 139:711-714.
Fleming MF, Mundt MP, French MT, Manwell LB, Stauffacher EA and Barry KL (2002): Brief
physician advice for problem drinkers: long-term efficacy and benefit-cost analysis. Alcohol Clin
Exp Res 26:36-43.
Fleming MF, Anton RF and Spies CD (2004): A review of genetic, biological, pharmacological,
and clinical factors that affect carbohydrate-deficient transferrin levels. Alcohol Clin Exp Res
Frezza M, di Padova C, Pozzato G, Terpin M, Baraona E and Lieber CS (1990): High blood alcohol
levels in women. The role of decreased gastric alcohol dehydrogenase activity and first-pass metabolism. N Engl J Med 322:95-99.
Friedman GD and Klatsky AL (1993): Is alcohol good for your health? N Engl J Med 329:18821883.
Fromme K, de Wit H, Hutchison KE, Ray L, Corbin WR, Cook TA, Wall TL and Goldman D
(2004): Biological and behavioral markers of alcohol sensitivity. Alcohol Clin Exp Res 28:247-256.
Gaines KC, Salhany JM, Tuma DJ and Sorrell MF (1977): Reaction of acetaldehyde with human
erythrocyte membrane proteins. FEBS Lett 75:115-119.
Ghosh P, Hale EA and Lakshman MR (2001): Plasma sialic-acid index of apolipoprotein J (SIJ): a
new alcohol intake marker. Alcohol 25:173-179.
Goldberg IJ (2003): To drink or not to drink? N Engl J Med 348:163-164.
Golka K and Wiese A (2004): Carbohydrate-deficient transferrin (CDT)--a biomarker for long-term
alcohol consumption. J Toxicol Environ Health B Crit Rev 7:319-337.
Gordis E and Herschkopf S (1986): Application of isoelectric focusing in immobilized pH gradients
to the study of acetaldehyde-modified hemoglobin. Alcohol Clin Exp Res 10:311-319.
Grønbæk M (2006): Factors influencing the relation between alcohol and cardiovascular disease.
Curr Opin Lipidol 17:17-21.
Gross MD, Gapstur SM, Belcher JD, Scanlan G and Potter JD (1992): The identification and partial
characterization of acetaldehyde adducts of hemoglobin occurring in vivo: a possible marker of
alcohol consumption. Alcohol Clin Exp Res 16:1093-1103.
Gunnarsson T, Karlsson A, Hansson P, Johnson G, Alling C and Odham G (1998): Determination
of phosphatidylethanol in blood from alcoholic males using high-performance liquid chromatography and evaporative light scattering or electrospray mass spectrometric detection. J Chromatogr B
Biomed Sci Appl 705:243-249.
Gual A, Segura L, Contel M, Heather N and Colom J (2002): Audit-3 and audit-4: effectiveness of
two short forms of the alcohol use disorders identification test. Alcohol Alcohol 37:591-596.
Gustavsson L (1995): ESBRA 1994 Award Lecture. Phosphatidylethanol formation: specific effects
of ethanol mediated via phospholipase D. Alcohol Alcohol 30:391-406.
Haber H, Putscher I, Georgi M and Melzig MF (1995): Influence of ethanol on the salsolinol excretion in healthy subjects. Alcohol 12:299-303.
Hansson P, Caron M, Johnson G, Gustavsson L and Alling C (1997): Blood phosphatidylethanol as
a marker of alcohol abuse: levels in alcoholic males during withdrawal. Alcohol Clin Exp Res
Harasymiw J and Bean P (2001): The combined use of the early detection of alcohol consumption
(EDAC) test and carbohydrate-deficient transferrin to identify heavy drinking behaviour in males.
Alcohol Alcohol 36:349-353.
Harasymiw JW, Vinson DC and Bean P (2000): The early detection of alcohol consumption
(EDAC) score in the identification of Heavy and at-risk drinkers from routine blood tests. J Addict
Hartz AJ, Guse C and Kajdacsy-Balla A (1997): Identification of heavy drinkers using a combination of laboratory tests. J Clin Epidemiol 50:1357-1368.
Hazelett SE, Liebelt RA, Brown WJ, Androulakakis V, Jarjoura D and Truitt EB, Jr. (1998):
Evaluation of acetaldehyde-modified hemoglobin and other markers of chronic heavy alcohol use:
effects of gender and hemoglobin concentration. Alcohol Clin Exp Res 22:1813-1819.
Helander A (1999): Absolute or relative measurement of carbohydrate-deficient transferrin in serum? Experiences with three immunological assays. Clin Chem 45:131-135.
Helander A and Eriksson CJP (2002): Laboratory tests for acute alcohol consumption: results of the
WHO/ISBRA Study on State and Trait Markers of Alcohol Use and Dependence. Alcohol Clin Exp
Helander A, Wikstrom T, Lowenmo C, Jacobsson G and Beck O (1992): Urinary excretion of 5hydroxyindole-3-acetic acid and 5-hydroxytryptophol after oral loading with serotonin. Life Sci
Helander A, Beck O, Jacobsson G, Lowenmo C and Wikstrom T (1993): Time course of ethanolinduced changes in serotonin metabolism. Life Sci 53:847-855.
Helander A, Beck O and Jones AW (1996a): Laboratory testing for recent alcohol consumption:
comparison of ethanol, methanol, and 5-hydroxytryptophol. Clin Chem 42:618-624.
Helander A, Carlsson AV and Borg S (1996b): Longitudinal comparison of carbohydrate-deficient
transferrin and gamma-glutamyl transferase: complementary markers of excessive alcohol consumption. Alcohol Alcohol 31:101-107.
Helander A, Eriksson G, Stibler H and Jeppsson JO (2001): Interference of transferrin isoform types
with carbohydrate-deficient transferrin quantification in the identification of alcohol abuse. Clin
Helander A, Dahl H and Bergstöm JP (2005): Comparison of Dade Behring N Latex CDT immunoassay with high-pressure liquid chromatography for testing of carbohydrate-deficient transferrin.
Alcohol Alcohol 40:i51.
Heng K, Hargarten S, Layde P, Craven A and Zhu S (2006): Moderate alcohol intake and motor
vehicle crashes: the conflict between health advantage and at-risk use. Alcohol Alcohol 41:451-454.
Hillbom M and Kaste M (1990): Alcohol abuse and brain infarction. Ann Med 22:347-352.
Hillers VN, Alldredge JR and Massey LK (1986): Determination of habitual alcohol intake from a
panel of blood chemistries. Alcohol Alcohol 21:199-205.
Hingson R and Winter M (2003): Epidemiology and consequences of drinking and driving. Alcohol
Res Health 27:63-78.
Hock B, Schwarz M, Domke I, Grunert VP, Wuertemberger M, Schiemann U, Horster S, Limmer
C, Stecker G and Soyka M (2005): Validity of carbohydrate-deficient transferrin (%CDT), gammaglutamyltransferase (gamma-GT) and mean corpuscular erythrocyte volume (MCV) as biomarkers
for chronic alcohol abuse: a study in patients with alcohol dependence and liver disorders of nonalcoholic and alcoholic origin. Addiction 100:1477-1486.
Hodgson R, Alwyn T, John B, Thom B and Smith A (2002): The FAST Alcohol Screening Test.
Alcohol Alcohol 37:61-66.
Hoerner M, Behrens UJ, Worner TM, Blacksberg I, Braly LF, Schaffner F and Lieber CS (1988):
The role of alcoholism and liver disease in the appearance of serum antibodies against acetaldehyde
adducts. Hepatology 8:569-574.
Holstege A, Bedossa P, Poynard T, Kollinger M, Chaput JC, Houglum K and Chojkier M (1994):
Acetaldehyde-modified epitopes in liver biopsy specimens of alcoholic and nonalcoholic patients:
localization and association with progression of liver fibrosis. Hepatology 19:367-374.
Horn PS and Pesce AJ (2003): Reference intervals: an update. Clin Chim Acta 334:5-23.
Howard AA, Arnsten JH and Gourevitch MN (2004): Effect of alcohol consumption on diabetes
mellitus: a systematic review. Ann Intern Med 140:211-219.
Huisman TH, Henson JB and Wilson JB (1983): A new high-performance liquid chromatographic
procedure to quantitate hemoglobin A1c and other minor hemoglobins in blood of normal, diabetic,
and alcoholic individuals. J Lab Clin Med 102:163-173.
Hultberg B, Isaksson A, Berglund M and Moberg AL (1991): Serum beta-hexosaminidase isoenzyme: a sensitive marker for alcohol abuse. Alcohol Clin Exp Res 15:549-552.
Huseby NE, Nilssen O and Kanitz RD (1997): Evaluation of two biological markers combined as a
parameter of alcohol dependency. Alcohol Alcohol 32:731-737.
IAS – Institute of Alcohol Studies (2006): Alcohol Consumption and Harm in the UK and EU,
Iimuro Y, Bradford BU, Gao W, Kadiiska M, Mason RP, Stefanovic B, Brenner DA and Thurman
RG (1996): Detection of alpha-hydroxyethyl free radical adducts in the pancreas after chronic exposure to alcohol in the rat. Mol Pharmacol 50:656-661.
Iimuro Y, Gallucci RM, Luster MI, Kono H and Thurman RG (1997): Antibodies to tumor necrosis
factor alfa attenuate hepatic necrosis and inflammation caused by chronic exposure to ethanol in the
rat. Hepatology 26:1530-1537.
Isaksson A, Blanche C, Hultberg B and Joelsson B (1985): Influence of ethanol on the human serum level of beta-hexosaminidase. Enzyme 33:162-166.
Israel Y, Hurwitz E, Niemelä O and Arnon R (1986): Monoclonal and polyclonal antibodies against
acetaldehyde-containing epitopes in acetaldehyde-protein adducts. Proc Natl Acad Sci U S A
Israel Y, Orrego H and Niemelä O (1988): Immune responses to alcohol metabolites: pathogenic
and diagnostic implications. Semin Liver Dis 8:81-90.
Izumi N, Sakai Y, Koyama W and Hasumura Y (1989): Clinical significance of serum antibodies
against alcohol-altered hepatocyte membrane in alcoholic liver disease. Alcohol Clin Exp Res
Jaakkola M, Sillanaukee P, Löf K, Koivula T and Nordback I (1994): Amount of alcohol is an important determinant of the severity of acute alcoholic pancreatitis. Surgery 115:31-38.
Jacobson JL and Jacobson SW (2002): Effects of prenatal alcohol exposure on child development.
Alcohol Res Health 26:282-286.
Jacobson JL, Jacobson SW and Sokol RJ (1996): Increased vulnerability to alcohol-related birth
defects in the offspring of mothers over 30. Alcohol Clin Exp Res 20:359-363.
Jaeschke H, Gores GJ, Cederbaum AI, Hinson JA, Pessayre D and Lemasters JJ (2002): Mechanisms of hepatotoxicity. Toxicol Sci 65:166-176.
Javors MA and Johnson BA (2003): Current status of carbohydrate deficient transferrin, total serum
sialic acid, sialic acid index of apolipoprotein J and serum beta-hexosaminidase as markers for alcohol consumption. Addiction 98 Suppl 2:45-50.
Johnson BA, Cloninger CR, Roache JD, Bordnick PS and Ruiz P (2000): Age of onset as a discriminator between alcoholic subtypes in a treatment-seeking outpatient population. Am J Addict
Jones KL and Smith DW (1973): Recognition of the fetal alcohol syndrome in early infancy. Lancet
Julkunen RJ, Tannenbaum L, Baraona E and Lieber CS (1985): First pass metabolism of ethanol: an
important determinant of blood levels after alcohol consumption. Alcohol 2:437-441.
Källmén H and Gustafson R (1998): Alcohol and disinhibition. Eur Addict Res 4:150-162.
Kärkkäinen P, Poikolainen K and Salaspuro M (1990): Serum beta-hexosaminidase as a marker of
heavy drinking. Alcohol Clin Exp Res 14:187-190.
Keating J, Cheung C, Peters TJ and Sherwood RA (1998): Carbohydrate deficient transferrin in the
assessment of alcohol misuse: absolute or relative measurements? A comparison of two methods
with regard to total transferrin concentration. Clin Chim Acta 272:159-169.
Kerr MA (1990): The structure and function of human IgA. Biochem J 271:285-296.
Kessler RC, Crum RM, Warner LA, Nelson CB, Schulenberg J and Anthony JC (1997): Lifetime
co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the
National Comorbidity Survey. Arch Gen Psychiatry 54:313-321.
Khoruts A, Stahnke L, McClain CJ, Logan G and Allen JI (1991): Circulating tumor necrosis factor, interleukin-1 and interleukin-6 concentrations in chronic alcoholic patients. Hepatology 13:267276.
Klassen LW, Tuma D and Sorrell MF (1995): Immune mechanisms of alcohol-induced liver disease. Hepatology 22:355-357.
Koivisto T and Salaspuro M (1996): Aldehyde dehydrogenases of the rat colon: comparison with
other tissues of the alimentary tract and the liver. Alcohol Clin Exp Res 20:551-555.
Kornhuber J, Kornhuber HH, Backhaus B, Kornhuber A, Kaiserauer C and Wanner W (1989): [The
normal values of gamma-glutamyltransferase are falsely defined up to now: on the diagnosis of hypertension, obesity and diabetes with reference to "normal" consumption of alcohol]. Versicherungsmedizin 41:78-81.
Koskinas J, Kenna JG, Bird GL, Alexander GJ and Williams R (1992): Immunoglobulin A antibody
to a 200-kilodalton cytosolic acetaldehyde adduct in alcoholic hepatitis. Gastroenterology
Kovacs EJ and Messingham KA (2002): Influence of alcohol and gender on immune response. Alcohol Res Health 26:257-263.
Laposata EA and Lange LG (1986): Presence of nonoxidative ethanol metabolism in human organs
commonly damaged by ethanol abuse. Science 231:497-499.
Laposata M (1997): Fatty acid ethyl esters: short-term and long-term serum markers of ethanol intake. Clin Chem 43:1527-1534.
Latt N and Saunders JB (2002): Alcohol misuse and dependence. Assessment and management.
Aust Fam Physician 31:1079-1085.
Leevy CB and Elbeshbeshy HA (2005): Immunology of alcoholic liver disease. Clin Liver Dis
Legros FJ, Nuyens V, Minet E, Emonts P, Boudjeltia KZ, Courbe A, Ruelle JL, Colicis J, de L'Escaille F and Henry JP (2002): Carbohydrate-deficient transferrin isoforms measured by capillary
zone electrophoresis for detection of alcohol abuse. Clin Chem 48:2177-2186.
Lesch OM, Walter H, Antal J, Heggli DE, Kovacz A, Leitner A, Neumeister A, Stumpf I, Sundrehagen E and Kasper S (1996a): Carbohydrate-deficient transferrin as a marker of alcohol intake: a
study with healthy subjects. Alcohol Alcohol 31:265-271.
Lesch OM, Walter H, Freitag H, Heggli DE, Leitner A, Mader R, Neumeister A, Passweg V, Pusch
H, Semler B, Sundrehagen E and Kasper S (1996b): Carbohydrate-deficient transferrin as a screening marker for drinking in a general hospital population. Alcohol Alcohol 31:249-256.
Li CJ, Nanji AA, Siakotos AN and Lin RC (1997): Acetaldehyde-modified and 4-hydroxynonenalmodified proteins in the livers of rats with alcoholic liver disease. Hepatology 26:650-657.
Lieber CS (1977): The metabolic basis of alcohol's toxicity. Hosp Pract 12:73-80.
Lieber CS (1995): Medical disorders of alcoholism. N Engl J Med 333:1058-1065.
Lieber CS (2004): The discovery of the microsomal ethanol oxidizing system and its physiologic
and pathologic role. Drug Metab Rev 36:511-529.
Lieber CS (2005): Metabolism of alcohol. Clin Liver Dis 9:1-35.
Lieber CS and DeCarli LM (1968): Ethanol oxidation by hepatic microsomes: adaptive increase
after ethanol feeding. Science 162:917-918.
Lin RC, Lumeng L, Shahidi S, Kelly T and Pound DC (1990): Protein-acetaldehyde adducts in serum of alcoholic patients. Alcohol Clin Exp Res 14:438-443.
Lin RC, Shahidi S, Kelly TJ, Lumeng C and Lumeng L (1993): Measurement of hemoglobinacetaldehyde adduct in alcoholic patients. Alcohol Clin Exp Res 17:669-674.
Litten RZ, Allen JP and Fertig JB (1995): Gamma-glutamyltranspeptidase and carbohydrate deficient transferrin: alternative measures of excessive alcohol consumption. Alcohol Clin Exp Res
Löf K, Seppä K, Itälä L, Koivula T, Turpeinen U and Sillanaukee P (1994): Carbohydrate-deficient
transferrin as an alcohol marker among female heavy drinkers: a population-based study. Alcohol
Clin Exp Res 18:889-894.
MacKenzie D, Langa A and Brown TM (1996): Identifying hazardous or harmful alcohol use in
medical admissions: a comparison of audit, cage and brief mast. Alcohol Alcohol 31:591-599.
Maier SE and West JR (2001): Drinking patterns and alcohol-related birth defects. Alcohol Res
Mann RE, Smart RG and Govoni R (2003): The epidemiology of alcoholic liver disease. Alcohol
Res Health 27:209-219.
Mårtensson O, Härlin A, Brandt R, Seppä K and Sillanaukee P (1997): Transferrin isoform distribution: gender and alcohol consumption. Alcohol Clin Exp Res 21:1710-1715.
McBride WJ, Li TK, Deitrich RA, Zimatkin S, Smith BR and Rodd-Henricks ZA (2002): Involvement of acetaldehyde in alcohol addiction. Alcohol Clin Exp Res 26:114-119.
McClain C, Hill D, Schmidt J and Diehl AM (1993): Cytokines and alcoholic liver disease. Semin
Liver Dis 13:170-182.
McClain CJ, Shedlofsky S, Barve S and Hill DB (1997): Cytokines and alcoholic liver disease. Alcohol Health Res World 21:317-320.
Medical Research Council (1998): Individual harmful and protective effects of alcohol. Medical
Research Council, Spring 1997. Alcohol Clin Exp Res 22:365S-373S.
Merikallio-Pajunen A, Strid N, Suokas A, Podlketnova I and Alho H (2004): Comparison of new
alcohol use test, the Helsinki Alcohol Use Test questionnaire, and Alcohol Use Disorders Identification Test and laboratory markers serum gamma-glutamyl transferase and carbohydrate-deficient
transferrin. Addict Behav 29:717-722.
Meyerhoff DJ, Bode C, Nixon SJ, de Bruin EA, Bode JC and Seitz HK (2005): Health risks of
chronic moderate and heavy alcohol consumption: how much is too much? Alcohol Clin Exp Res
Miller PM and Anton RF (2004): Biochemical alcohol screening in primary health care. Addict
Millikan LE (1999): History and epidemiology of alcohol use and abuse. Clin Dermatol 17:353356.
Molina PE, Hoek JB, Nelson S, Guidot DM, Lang CH, Wands JR and Crawford JM (2003):
Mechanisms of alcohol-induced tissue injury. Alcohol Clin Exp Res 27:563-575.
Monteiro MG and Masur J (1985): Diagnostic of alcoholism: how useful is the combination of
gamma glutamyl transferase with different biochemical markers? Drug Alcohol Depend 16:31-37.
Moore RD, Bone LR, Geller G, Mamon JA, Stokes EJ and Levine DM (1989): Prevalence, detection, and treatment of alcoholism in hospitalized patients. JAMA 261:403-407.
Morgan MY, Camilo ME, Luck W, Sherlock S and Hoffbrand AV (1981): Macrocytosis in alcoholrelated liver disease: its value for screening. Clin Lab Haematol 3:35-44.
Mundle G, Ackermann K, Munkes J, Steinle D and Mann K (1999): Influence of age, alcohol consumption and abstinence on the sensitivity of carbohydrate-deficient transferrin, gammaglutamyltransferase and mean corpuscular volume. Alcohol Alcohol 34:760-766.
Mundle G, Munkes J, Ackermann K and Mann K (2000): Sex differences of carbohydrate-deficient
transferrin, gamma-glutamyltransferase, and mean corpuscular volume in alcohol-dependent patients. Alcohol Clin Exp Res 24:1400-1405.
Nalpas B, Vassault A, Charpin S, Lacour B and Berthelot P (1986): Serum mitochondrial aspartate
aminotransferase as a marker of chronic alcoholism: diagnostic value and interpretation in a liver
unit. Hepatology 6:608-614.
Neuman MG (2003): Cytokines--central factors in alcoholic liver disease. Alcohol Res Health
Neumann T and Spies C (2003): Use of biomarkers for alcohol use disorders in clinical practice.
Addiction 98 Suppl 2:81-91.
Nguyen LB and Peterson CM (1984): Modification of hemoglobin by acetaldehyde: a time course
study by high pressure liquid chromatography. Alcohol Clin Exp Res 8:516-521.
Niemelä O (1999): Aldehyde-protein adducts in the liver as a result of ethanol-induced oxidative
stress. Front Biosci 4:D506-D513.
Niemelä O (2001): Distribution of ethanol-induced protein adducts in vivo: relationship to tissue
injury. Free Radic Biol Med 31:1533-1538.
Niemelä O (2002): Serum Diagnosis of Alcoholic Liver Disease and Markers of Ethanol Intake. In:
Ethanol and the Liver, pp. 411-449. Eds. DIN Sherman, V Preedy and RR Watson, Taylor and
Francis, New York and London.
Niemelä O (2007): Biomarkers in alcoholism. Clin Chim Acta 377:39-49.
Niemelä O and Israel Y (1992): Hemoglobin-acetaldehyde adducts in human alcohol abusers. Lab
Niemelä O and Parkkila S (2004): Alcoholic macrocytosis--is there a role for acetaldehyde and adducts? Addict Biol 9:3-10.
Niemelä O, Klajner F, Orrego H, Vidins E, Blendis L and Israel Y (1987): Antibodies against acetaldehyde-modified protein epitopes in human alcoholics. Hepatology 7:1210-1214.
Niemelä O, Israel Y, Mizoi Y, Fukunaga T and Eriksson CJP (1990): Hemoglobin-acetaldehyde
adducts in human volunteers following acute ethanol ingestion. Alcohol Clin Exp Res 14:838-841.
Niemelä O, Juvonen T and Parkkila S (1991): Immunohistochemical demonstration of acetaldehyde-modified epitopes in human liver after alcohol consumption. J Clin Invest 87:1367-1374.
Nolan JP, DeLissio MG, Camara DS, Feind DM and Gagliardi NC (1986): IgA antibody to lipid A
in alcoholic liver disease. Lancet 1:176-179.
Norton R, Batey R, Dwyer T and MacMahon S (1987): Alcohol consumption and the risk of alcohol related cirrhosis in women. Br Med J (Clin Res Ed) 295:80-82.
Nyblom H, Berggren U, Balldin J and Olsson R (2004): High AST/ALT ratio may indicate advanced alcoholic liver disease rather than heavy drinking. Alcohol Alcohol 39:336-339.
Ogston SA and Parry GJ (1992): EUROMAC. A European concerted action: maternal alcohol consumption and its relation to the outcome of pregnancy and child development at 18 months. Results--strategy of analysis and analysis of pregnancy outcome. Int J Epidemiol 21 Suppl 1:S45-S71.
Öhberg A, Vuori E, Ojanperä I and Lönngvist J (1996): Alcohol and drugs in suicides. Br J Psychiatry 169:75-80.
Orrego H, Israel Y, Blake JE and Medline A (1983): Assessment of prognostic factors in alcoholic
liver disease: toward a global quantitative expression of severity. Hepatology 3:896-905.
Orrego H, Blake JE and Israel Y (1985): Relationship between gamma-glutamyl transpeptidase and
mean urinary alcohol levels in alcoholics while drinking and after alcohol withdrawal. Alcohol Clin
Exp Res 9:10-13.
Oscar-Berman M, Shagrin B, Evert DL and Epstein C (1997): Impairments of brain and behavior:
the neurological effects of alcohol. Alcohol Health Res World 21:65-75.
Österberg E (2006): High alcohol-related burden of disease as basis for controlling total alcohol
consumption. Addiction 101:1070.
Papoz L, Warnet JM, Pequignot G, Eschwege E, Claude JR and Schwartz D (1981): Alcohol consumption in a healthy population. Relationship to gamma-glutamyl transferase activity and mean
corpuscular volume. JAMA 245:1748-1751.
Paradis V, Scoazec JY, Kollinger M, Holstege A, Moreau A, Feldmann G and Bedossa P (1996):
Cellular and subcellular localization of acetaldehyde-protein adducts in liver biopsies from alcoholic patients. J Histochem Cytochem 44:1051-1057.
Parlesak A, Schäfer C and Bode C (2002): IgA against gut-derived endotoxins: does it contribute to
suppression of hepatic inflammation in alcohol-induced liver disease? Dig Dis Sci 47:760-766.
Patton R, Hilton C, Crawford MJ and Touquet R (2004): The Paddington Alcohol Test: a short report. Alcohol Alcohol 39:266-268.
Peters M (1996): Actions of cytokines on the immune response and viral interactions: an overview.
Peterson CM and Polizzi CM (1987): Improved method for acetaldehyde in plasma and hemoglobin-associated acetaldehyde: results in teetotalers and alcoholics reporting for treatment. Alcohol
Phillips MG, Preedy VR and Hughes RD (2003): Assessment of prognosis in alcoholic liver disease: can serum hyaluronate replace liver biopsy? Eur J Gastroenterol Hepatol 15:941-944.
Piano MR and Schwertz DW (1994): Alcoholic heart disease: a review. Heart Lung 23:3-17.
Piccinelli M, Tessari E, Bortolomasi M, Piasere O, Semenzin M, Garzotto N and Tansella M
(1997): Efficacy of the alcohol use disorders identification test as a screening tool for hazardous
alcohol intake and related disorders in primary care: a validity study. BMJ 314:420-424.
Pirkola SP, Suominen K and Isometsä ET (2004): Suicide in alcohol-dependent individuals: epidemiology and management. CNS Drugs 18:423-436.
Pönniö M, Alho H, Heinälä P, Nikkari ST and Sillanaukee P (1999a): Serum and saliva levels of
sialic acid are elevated in alcoholics. Alcohol Clin Exp Res 23:1060-1064.
Pönniö M, Alho H, Nikkari ST, Olsson U, Rydberg U and Sillanaukee P (1999b): Serum sialic acid
in a random sample of the general population. Clin Chem 45:1842-1849.
Pragst F, Auwaerter V, Sporkert F and Spiegel K (2001): Analysis of fatty acid ethyl esters in hair
as possible markers of chronically elevated alcohol consumption by headspace solid-phase microextraction (HS-SPME) and gas chromatography-mass spectrometry (GC-MS). Forensic Sci Int
Pratt DS and Kaplan MM (2000): Evaluation of abnormal liver-enzyme results in asymptomatic
patients. N Engl J Med 342:1266-1271.
Preuss UW, Koller G, Barnow S, Eikmeier M and Soyka M (2006): Suicidal behavior in alcoholdependent subjects: the role of personality disorders. Alcohol Clin Exp Res 30:866-877.
Puddey IB and Beilin LJ (2006): Alcohol is bad for blood pressure. Clin Exp Pharmacol Physiol
Pullarkat RK and Raguthu S (1985): Elevated urinary dolichol levels in chronic alcoholics. Alcohol
Clin Exp Res 9:28-30.
Puukka K, Hietala J, Koivisto H, Anttila P, Bloigu R and Niemelä O (2006a): Additive effects of
moderate drinking and obesity on serum gamma-glutamyl transferase activity. Am J Clin Nutr
Puukka K, Hietala J, Koivisto H, Anttila P, Bloigu R and Niemelä O (2006b): Age-related changes
on serum ggt activity and the assessment of ethanol intake. Alcohol Alcohol 41:522-527.
Pyörälä E (1990): Trends in alcohol consumption in Spain, Portugal, France and Italy from the
1950s until the 1980s. Br J Addict 85:469-477.
Rehm J, Room R, Graham K, Monteiro M, Gmel G and Sempos CT (2003): The relationship of
average volume of alcohol consumption and patterns of drinking to burden of disease: an overview.
Rehm J, Taylor B and Patra J (2006): Volume of alcohol consumption, patterns of drinking and
burden of disease in the European region 2002. Addiction 101:1086-1095.
Reid AL, Webb GR, Hennrikus D, Fahey PP and Sanson-Fisher RW (1986): Detection of patients
with high alcohol intake by general practitioners. Br Med J (Clin Res Ed) 293:735-737.
Reid MC, Fiellin DA and O'Connor PG (1999): Hazardous and harmful alcohol consumption in
primary care. Arch Intern Med 159:1681-1689.
Reynaud M, Schellenberg F, Loisequx-Meunier MN, Schwan R, Maradeix B, Planche F and Gillet
C (2000): Objective diagnosis of alcohol abuse: compared values of carbohydrate-deficient transferrin (CDT), gamma-glutamyl transferase (GGT), and mean corpuscular volume (MCV). Alcohol
Clin Exp Res 24:1414-1419.
Rintala J, Jaatinen P, Parkkila S, Sarviharju M, Kiianmaa K, Hervonen A and Niemelä O (2000):
Evidence of acetaldehyde-protein adduct formation in rat brain after lifelong consumption of ethanol. Alcohol Alcohol 35:458-463.
Robertson EA and Zweig MH (1981): Use of receiver operating characteristic curves to evaluate the
clinical performance of analytical systems. Clin Chem 27:1569-1574.
Rodgers B, Korten AE, Jorm AF, Jacomb PA, Christensen H and Henderson AS (2000): Non-linear
relationships in associations of depression and anxiety with alcohol use. Psychol Med 30:421-432.
Roine RP, Turpeinen U, Ylikahri R and Salaspuro M (1987): Urinary dolichol--a new marker of
alcoholism. Alcohol Clin Exp Res 11:525-527.
Roine R, Humaloja K, Hämäläinen J, Nykänen I, Ylikahri R and Salaspuro M (1989): Significant
increases in urinary dolichol levels in bacterial infections, malignancies and pregnancy but not in
other clinical conditions. Ann Med 21:13-16.
Roine R, Heinonen T, Salmela K, Heikkonen E, Suokas A, Luurila OJ, Koskinen P, Palo J and
Salaspuro M (1991): Strenuous physical activity, aspirin and heat stress increase urinary dolichols:
evidence for lysosomal origin of urinary dolichols. Clin Chim Acta 204:13-21.
Romppanen J, Punnonen K, Anttila P, Jakobsson T, Blake J and Niemelä O (2002): Serum sialic
acid as a marker of alcohol consumption: effect of liver disease and heavy drinking. Alcohol Clin
Exp Res 26:1234-1238.
Room R (1997): Alcohol, the individual and society: what history teaches us. Addiction 92 Suppl
Room R, Babor T and Rehm J (2005): Alcohol and public health. Lancet 365:519-530.
Rosman AS and Lieber CS (1992): On overview of current and emerging markers of alcoholism. In:
Measuring Alcohol Consumption. Psychosocial and Biochemical Methods, pp. 99-134. Eds. RZ
Litten and JP Allen, The Humana Press Inc., Totowa, New Jersey.
Rosman AS and Lieber CS (1994): Diagnostic utility of laboratory tests in alcoholic liver disease.
Clin Chem 40:1641-1651.
Ryb GE, Söderström CA, Kufera JA, Dischinger PC and Ho SM (1999): Use of blood alcohol concentration and laboratory tests to detect current alcohol dependence in trauma center patients. J
Saeed RW, Varma S, Peng T, Tracey KJ, Sherry B and Metz CN (2004): Ethanol blocks leukocyte
recruitment and endothelial cell activation in vivo and in vitro. J Immunol 173:6376-6383.
Salaspuro M (1987): Use of enzymes for the diagnosis of alcohol-related organ damage. Enzyme
Salaspuro M (1996): Bacteriocolonic pathway for ethanol oxidation: characteristics and implications. Ann Med 28:195-200.
Salaspuro M (1999): Carbohydrate-deficient transferrin as compared to other markers of alcoholism: a systematic review. Alcohol 19:261-271.
Salaspuro MP and Lieber CS (1978): Non-uniformity of blood ethanol elimination: its exaggeration
after chronic consumption. Ann Clin Res 10:294-297.
Salmela KS, Sillanaukee P, Itälä L, Väkeväinen S, Salaspuro M and Roine RP (1997): Binding of
acetaldehyde to rat gastric mucosa during ethanol oxidation. J Lab Clin Med 129:627-633.
Sareen J, McWilliams L, Cox B and Stein MB (2004): Does a U-shaped relationship exist between
alcohol use and DSM-III-R mood and anxiety disorders? J Affect Disord 82:113-118.
Sarkola T, Dahl H, Eriksson CJP and Helander A (2003): Urinary ethyl glucuronide and 5hydroxytryptophol levels during repeated ethanol ingestion in healthy human subjects. Alcohol Alcohol 38:347-351.
Saunders JB, Aasland OG, Babor TF, de lF, Jr. and Grant M (1993): Development of the Alcohol
Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption--II. Addiction 88:791-804.
Savola O, Niemelä O and Hillbom M (2004): Blood alcohol is the best indicator of hazardous alcohol drinking in young adults and working-age patients with trauma. Alcohol Alcohol 39:340-345.
Savola O, Niemelä O and Hillbom M (2005): Alcohol intake and the pattern of trauma in young
adults and working aged people admitted after trauma. Alcohol Alcohol 40:269-273.
Schellenberg F, Schwan R, Mennetrey L, Loiseaux MN, Pages JC and Reynaud M (2005): Doseeffect relation between daily ethanol intake in the range 0-70 grams and %CDT value: validation of
a cut-off value. Alcohol Alcohol 40:531-534.
Schenker S (1997): Medical consequences of alcohol abuse: is gender a factor? Alcohol Clin Exp
Schmidt LG, Schmidt K, Dufeu P, Ohse A, Rommelspacher H and Muller C (1997): Superiority of
carbohydrate-deficient transferrin to gamma-glutamyltransferase in detecting relapse in alcoholism.
Am J Psychiatry 154:75-80.
Schwan R, Loiseaux MN, Schellenberg F, Albuisson E, Favre JD, Rigaud A, Llorca PM, Gillet C
and Reynaud M (2004): Multicenter validation study of the %CDT TIA kit in alcohol abuse and
alcohol dependence. Alcohol Clin Exp Res 28:1331-1337.
Seitz HK and Pöschl G (1997): The role of gastrointestinal factors in alcohol metabolism. Alcohol
Seitz HK, Gartner U, Egerer G and Simanowski UA (1994): Ethanol metabolism in the gastrointestinal tract and its possible consequences. Alcohol Alcohol Suppl 2:157-162.
Selzer ML (1971): The Michigan alcoholism screening test: the quest for a new diagnostic instrument. Am J Psychiatry 127:1653-1658.
Seppä K and Sillanaukee P (1994): Women, alcohol, and red cells. Alcohol Clin Exp Res 18:11681171.
Seppä K, Laippala P and Saarni M (1991): Macrocytosis as a consequence of alcohol abuse among
patients in general practice. Alcohol Clin Exp Res 15:871-876.
Seppä K, Mäkelä R and Sillanaukee P (1995): Effectiveness of the Alcohol Use Disorders Identification Test in occupational health screenings. Alcohol Clin Exp Res 19:999-1003.
Seppä K, Lepistö J and Sillanaukee P (1998): Five-shot questionnaire on heavy drinking. Alcohol
Clin Exp Res 22:1788-1791.
Sharpe PC (2001): Biochemical detection and monitoring of alcohol abuse and abstinence. Ann
Clin Biochem 38:652-664.
Sillanaukee P (1992): The diagnostic value of a discriminant score in the detection of alcohol abuse.
Arch Pathol Lab Med 116:924-929.
Sillanaukee P and Koivula T (1990): Detection of a new acetaldehyde-induced hemoglobin fraction
HbA1ach by cation exchange liquid chromatography. Alcohol Clin Exp Res 14:842-846.
Sillanaukee P and Olsson U (2001): Improved diagnostic classification of alcohol abusers by combining carbohydrate-deficient transferrin and gamma-glutamyltransferase. Clin Chem 47:681-685.
Sillanaukee P, Seppä K and Koivula T (1991a): Association of a haemoglobin-acetaldehyde adduct
with questionnaire results on heavy drinkers. Alcohol Alcohol 26:519-525.
Sillanaukee P, Seppä K and Koivula T (1991b): Effect of acetaldehyde on hemoglobin: HbA1ach as
a potential marker of heavy drinking. Alcohol 8:377-381.
Sillanaukee P, Kiianmaa K, Roine R and Seppä K (1992a): Alkoholin suurkulutuksen kriteerit.
Suom Lääkäril 47:2919-2921.
Sillanaukee P, Seppä K, Koivula T, Israel Y and Niemelä O (1992b): Acetaldehyde-modified hemoglobin as a marker of alcohol consumption: comparison of two new methods. J Lab Clin Med
Sillanaukee P, Aalto M and Seppä K (1998): Carbohydrate-deficient transferrin and conventional
alcohol markers as indicators for brief intervention among heavy drinkers in primary health care.
Alcohol Clin Exp Res 22:892-896.
Sillanaukee P, Pönniö M and Jääskeläinen IP (1999a): Occurrence of sialic acids in healthy humans
and different disorders. Eur J Clin Invest 29:413-425.
Sillanaukee P, Pönniö M and Seppä K (1999b): Sialic acid: new potential marker of alcohol abuse.
Alcohol Clin Exp Res 23:1039-1043.
Sillanaukee P, Strid N, Jousilahti P, Vartiainen E, Poikolainen K, Nikkari S, Allen JP and Alho H
(2001): Association of self-reported diseases and health care use with commonly used laboratory
markers for alcohol consumption. Alcohol Alcohol 36:339-345.
Sood B, Delaney-Black V, Covington C, Nordstrom-Klee B, Ager J, Templin T, Janisse J, Martier
S and Sokol RJ (2001): Prenatal alcohol exposure and childhood behavior at age 6 to 7 years: I.
dose-response effect. Pediatrics 108:E34.
Sorvajärvi K, Blake JE, Israel Y and Niemelä O (1996): Sensitivity and specificity of carbohydratedeficient transferrin as a marker of alcohol abuse are significantly influenced by alterations in serum
transferrin: comparison of two methods. Alcohol Clin Exp Res 20:449-454.
Stakes – National Research and Development Centre for Welfare and Health (2005): Yearbook of
Alcohol and Drug Statistics 2005. Gummerus Kirjapaino Oy, Vaajakoski.
Stakes – National Research and Development Centre for Welfare and Health (2006): Quick Facts
2006 about Alcohol and Drugs, http://www.stakes.fi/NR/rdonlyres/CA7D94B0-7EF0-411A-9FA354ADCC7AF6AB/0/taskumatti2006englanti.pdf (4.10.2006).
Stefenelli N, Klotz H, Engel A and Bauer P (1985): Serum sialic acid in malignant tumors, bacterial
infections, and chronic liver diseases. J Cancer Res Clin Oncol 109:55-59.
Stetter F, Gaertner HJ, Wiatr G, Mann K and Breyer-Pfaff U (1991): Urinary dolichol--a doubtful
marker of alcoholism. Alcohol Clin Exp Res 15:938-941.
Stevens VJ, Fantl WJ, Newman CB, Sims RV, Cerami A and Peterson CM (1981): Acetaldehyde
adducts with hemoglobin. J Clin Invest 67:361-369.
Stibler H (1991): Carbohydrate-deficient transferrin in serum: a new marker of potentially harmful
alcohol consumption reviewed. Clin Chem 37:2029-2037.
Stibler H and Borg S (1991): Glycoprotein glycosyltransferase activities in serum in alcoholabusing patients and healthy controls. Scand J Clin Lab Invest 51:43-51.
Stibler H and Kjellin KG (1976): Isoelectric focusing and electrophoresis of the CSF proteins in
tremor of different origins. J Neurol Sci 30:269-285.
Stibler H, Allgulander C, Borg S and Kjellin KG (1978): Abnormal microheterogeneity of transferrin in serum and cerebrospinal fluid in alcoholism. Acta Med Scand 204:49-56.
Stibler H, Borg S and Allgulander C (1979): Clinical significance of abnormal heterogeneity of
transferrin in relation to alcohol consumption. Acta Med Scand 206:275-281.
Stibler H, Borg S and Beckman G (1988): Transferrin phenotype and level of carbohydratedeficient transferrin in healthy individuals. Alcohol Clin Exp Res 12:450-453.
Stromme JH, Rustad P, Steensland H, Theodorsen L and Urdal P (2004): Reference intervals for
eight enzymes in blood of adult females and males measured in accordance with the International
Federation of Clinical Chemistry reference system at 37 degrees C: part of the Nordic Reference
Interval Project. Scand J Clin Lab Invest 64:371-384.
Svegliati-Baroni G, Baraona E, Rosman AS and Lieber CS (1994): Collagen-acetaldehyde adducts
in alcoholic and nonalcoholic liver diseases. Hepatology 20:111-118.
Swerdlow MA, Chowdhury LN, Mishra V and Kavin H (1983): IgA deposits in skin in alcoholic
liver disease. J Am Acad Dermatol 9:232-236.
Szabo G, Catalano D, White B and Mandrekar P (2004): Acute alcohol consumption inhibits accessory cell function of monocytes and dendritic cells. Alcohol Clin Exp Res 28:824-828.
Taracha E, Habrat B, Lehner M, Wislowska A, Woronowicz BT, Bogulas M, Charewicz J, Markuszewski C and Plaznik A (2004): Alanine aminopeptidase activity in urine: a new marker of
chronic alcohol abuse? Alcohol Clin Exp Res 28:729-735.
Thurman RG and Handler JA (1989): New perspectives in catalase-dependent ethanol metabolism.
Drug Metab Rev 20:679-688.
Thurman RG and McKenna WR (1975): Pathways of ethanol metabolism in perfused rat liver. Adv
Exp Med Biol 56:57-76.
Traving C and Schauer R (1998): Structure, function and metabolism of sialic acids. Cell Mol Life
Treatment of alcohol abuse. Current Care Guideline. Working group appointed by the Finnish Society of Addiction Medicine. Finnish Medical Society Duodecim, 2005. Available in Internet:
Tuma DJ and Klassen LW (1992): Immune responses to acetaldehyde-protein adducts: role in alcoholic liver disease. Gastroenterology 103:1969-1973.
Tuma DJ and Sorrell MF (1987): Functional consequences of acetaldehyde binding to proteins. Alcohol Alcohol Suppl 1:61-66.
Tupala E and Tiihonen J (2004): Dopamine and alcoholism: neurobiological basis of ethanol abuse.
Prog Neuropsychopharmacol Biol Psychiatry 28:1221-1247.
Turner C (1990): How much alcohol is in a 'standard drink'? An analysis of 125 studies. Br J Addict
van de Wiel A, Delacroix DL, van Hattum J, Schuurman HJ and Kater L (1987): Characteristics of
serum IgA and liver IgA deposits in alcoholic liver disease. Hepatology 7:95-99.
van de Wiel A, van Hattum J, Schuurman HJ and Kater L (1988): Immunoglobulin A in the diagnosis of alcoholic liver disease. Gastroenterology 94:457-462.
Varga A, Hansson P, Lundqvist C and Alling C (1998): Phosphatidylethanol in blood as a marker
of ethanol consumption in healthy volunteers: comparison with other markers. Alcohol Clin Exp
Viitala K, Israel Y, Blake JE and Niemelä O (1997): Serum IgA, IgG, and IgM antibodies directed
against acetaldehyde-derived epitopes: relationship to liver disease severity and alcohol consumption. Hepatology 25:1418-1424.
Viitala K, Lähdesmäki K and Niemelä O (1998): Comparison of the Axis %CDT TIA and the
CDTect method as laboratory tests of alcohol abuse. Clin Chem 44:1209-1215.
Visapää JP, Jokelainen K, Nosova T and Salaspuro M (1998): Inhibition of intracolonic acetaldehyde production and alcoholic fermentation in rats by ciprofloxacin. Alcohol Clin Exp Res
Voltaire A, Beck O and Borg S (1992): Urinary 5-hydroxytryptophol: a possible marker of recent
alcohol consumption. Alcohol Clin Exp Res 16:281-285.
Walsh B (1997): Trends in alcohol production, trade and consumption. Addiction 92 Suppl 1:S61S66.
Warren KR and Foudin LL (2001): Alcohol-related birth defects--the past, present, and future. Alcohol Res Health 25:153-158.
Watzl B, Bub A, Pretzer G, Roser S, Barth SW and Rechkemmer G (2004): Daily moderate
amounts of red wine or alcohol have no effect on the immune system of healthy men. Eur J Clin
Wehr H, Rodo M, Lieber CS and Baraona E (1993): Acetaldehyde adducts and autoantibodies
against VLDL and LDL in alcoholics. J Lipid Res 34:1237-1244.
Wheeler MD (2003): Endotoxin and Kupffer cell activation in alcoholic liver disease. Alcohol Res
White IR, Altmann DR and Nanchahal K (2002): Alcohol consumption and mortality: modelling
risks for men and women at different ages. BMJ 325:191.
Wight AJ and Ogden GR (1998): Possible mechanisms by which alcohol may influence the development of oral cancer--a review. Oral Oncol 34:441-447.
Wilson DM (1995): Problem drinkers: can we identify and help them? CMAJ 152:825-828.
Worrall S, de Jersey J, Shanley BC and Wilce PA (1991): Antibodies against acetaldehydemodified epitopes: an elevated IgA response in alcoholics. Eur J Clin Invest 21:90-95.
Worrall S, Jersey JD, Wilce PA, Seppä K and Sillanaukee P (1994): Studies on the usefulness of
acetaldehyde-modified proteins and associated antibodies as markers of alcohol abuse. Alcohol Alcohol Suppl 2:503-507.
Worrall S, de Jersey J, Wilce PA, Seppä K, Hurme L and Sillanaukee P (1996): Relationship between alcohol intake and immunoglobulin a immunoreactivity with acetaldehyde-modified bovine
serum albumin. Alcohol Clin Exp Res 20:836-840.
Worrall S, de Jersey J, Wilce PA, Seppä K, Hurme L and Sillanaukee P (1998): Comparison of carbohydrate-deficient transferrin, immunoglobulin A antibodies reactive with acetaldehyde-modified
protein and acetaldehyde-modified albumin with conventional markers of alcohol consumption.
Alcohol Clin Exp Res 22:1921-1926.
Worrall S, Richardson PJ and Preedy VR (2000): Experimental heart muscle damage in alcohol
feeding is associated with increased amounts of reduced- and unreduced-acetaldehyde and
malondialdehyde–acetaldehyde protein adducts. Addict Biol 5:421-427.
Worrall S, Niemelä O, Parkkila S, Peters TJ and Preedy VR (2001): Protein adducts in type I and
type II fibre predominant muscles of the ethanol-fed rat: preferential localisation in the sarcolemmal
and subsarcolemmal region. Eur J Clin Invest 31:723-730.
Wurst FM and Metzger J (2002): The ethanol conjugate ethyl glucuronide is a useful marker of recent alcohol consumption. Alcohol Clin Exp Res 26:1114-1119.
Wurst FM, Kempter C, Seidl S and Alt A (1999): Ethyl glucuronide--a marker of alcohol consumption and a relapse marker with clinical and forensic implications. Alcohol Alcohol 34:71-77.
Wurst FM, Skipper GE and Weinmann W (2003): Ethyl glucuronide--the direct ethanol metabolite
on the threshold from science to routine use. Addiction 98 Suppl 2:51-61.
Wurst FM, Alling C, Aradottir S, Pragst F, Allen JP, Weinmann W, Marmillot P, Ghosh P,
Lakshman R, Skipper GE, Neumann T, Spies C, Javors M, Johnson BA, Ait-Daoud N, Akhtar F,
Roache JD and Litten R (2005): Emerging biomarkers: new directions and clinical applications.
Alcohol Clin Exp Res 29:465-473.
Wurst FM, Dresen S, Allen JP, Wiesbeck G, Graf M and Weinmann W (2006): Ethyl sulphate: a
direct ethanol metabolite reflecting recent alcohol consumption. Addiction 101:204-211.
Xin Y, Lasker JM and Lieber CS (1995): Serum carbohydrate-deficient transferrin: mechanism of
increase after chronic alcohol intake. Hepatology 22:1462-1468.
Yersin B, Nicolet JF, Dercrey H, Burnier M, van Melle G and Pecoud A (1995): Screening for excessive alcohol drinking. Comparative value of carbohydrate-deficient transferrin, gammaglutamyltransferase, and mean corpuscular volume. Arch Intern Med 155:1907-1911.
Yin M, Wheeler MD, Kono H, Bradford BU, Gallucci RM, Luster MI and Thurman RG (1999):
Essential role of tumor necrosis factor alpha in alcohol-induced liver injury in mice. Gastroenterology 117:942-952.
Zimatkin SM and Deitrich RA (1997): Ethanol metabolism in the brain. Addict Biol 2:387-400.
Zweig MH and Campbell G (1993): Receiver-operating characteristic (ROC) plots: a fundamental
evaluation tool in clinical medicine. Clin Chem 39:561-577.
Alcohol and Alcoholism Advance Access published August 30, 2005
Alcohol & Alcoholism Page 1 of 4
doi:10.1093/alcalc/agh201, available online at www.alcalc.oxfordjournals.org
SERUM GAMMA-GLUTAMYL TRANSFERASE IN ALCOHOLICS, MODERATE DRINKERS
AND ABSTAINERS: EFFECT ON GT REFERENCE INTERVALS AT POPULATION LEVEL
JOHANNA HIETALA, KATRI PUUKKA, HEIDI KOIVISTO, PETRA ANTTILA
and ONNI NIEMELÄ*
Department of Laboratory Medicine and Medical Research Unit, Seinäjoki Central Hospital, and University of Tampere,
FIN-60220 Seinäjoki, Finland
(Received 2 June 2005; first review notified 15 July 2005; accepted in final revised form 8 August 2005)
Abstract — Aims: To clarify in the association between amount of ethanol consumption and serum gamma-glutamyl transferase (GT)
levels. Methods: GT values were measured from 195 individuals with a wide variety of well-documented ethanol consumption assessed
by detailed personal interviews using a time-line follow-back technique. These included 103 heavy drinkers (90 men, 13 women) and 92
healthy volunteers (54 men, 38 women) who were either abstainers (n = 30) or moderate drinkers (n = 62). For comparisons, data were
collected from GT measurements for establishing GT reference intervals from 2485 healthy volunteers including 1156 abstainers and
1329 moderate drinkers. Results: GT values in the individuals whose mean ethanol consumption exceeded 40 g of ethanol per day
were significantly higher than those in the moderate drinkers with a mean consumption of 1–40 g/day (P < 0.001) or in abstainers
(P < 0.001). The GT values in the group of moderate drinkers also exceeded those of the abstainers (P < 0.001). The upper normal
GT limits obtained from the data from abstainers were markedly lower (men 45 U/l, women 35 U/l) than those obtained from the population of moderate drinkers (men 66 U/l, women 40 U/l). Conclusions: Serum GT concentrations may respond to relatively low levels of
ethanol consumption, which should be considered when defining GT reference intervals. The continuous increase in alcohol consumption at population level may lead to increased GT cut-off limits and hamper the detection of alcohol problems and liver affection in their
or abstainers: 54 men, mean age 41 ± 16 years; 38 women,
mean age 44 ± 19 years) who underwent detailed personal
interviews using a time-line follow-back technique (interview
sample). The heavy drinkers had a history of continuous ethanol consumption or binge drinking, the mean consumption
being in the range of 40–539 g/day during the period of
4 weeks prior to sampling. In addition, this interview sample
included 30 abstainers and 62 moderate drinkers with a
mean daily ethanol consumption between 1 and 40 g/day.
Measurements of GT levels were carried out using standard
clinical chemical methods in an accreditated (SFS-EN
45001, ISO/IEC Guide 25) laboratory of Seinäjoki Central
Hospital, Finland. For comparisons, data from a survey on
2485 apparently healthy individuals (1174 men, age 47 ±
18 years; 1311 women, age 47 ± 18 years) collected for establishing reference intervals in Nordic countries were also
used as kindly provided by the project coordinator, Professor
Pål Rustad, Fürst Medical Laboratory, Oslo, Norway. These
subjects were classified as either abstainers (n = 1156:
471 men, age 49 ± 19 years; 685 women, age 49 ± 19 years)
or moderate drinkers (n = 1329: 703 men, age 46 ± 17 years;
626 women, age 45 ± 16 years). In these subjects, the maximum amount of alcohol consumption during the 24 h period
prior to sampling had been 24 g (two standard drinks). The
survey excluded individuals who had clinical or laboratory
evidence of current or recent illnesses or infections, who
were pregnant, had donated blood during the past 5 months,
or had used any prescription drugs during the preceding
1 week. Smoking was not allowed for 1 h prior to sampling.
All GT measurements were carried out with homogeneous
International Federation of Clinical Chemistry (IFCC)
compatible measuring systems.
Gamma-glutamyl transferase (GT) is a commonly used laboratory parameter for detecting excessive alcohol consumption
(Zein and Discombe, 1970; Reyes and Miller, 1980). Although
several studies have reported a positive correlation between
the amount of alcohol consumed and serum GT levels, the
reported sensitivities of this marker in previous literature
have varied greatly, from 15 to 85% (Bagrel et al., 1979;
Chick et al., 1981; Papoz et al., 1981; Persson et al., 1990;
Leino et al., 1995; Anttila et al., 2004).
Over the past decades, both the total ethanol consumption
per capita and associated medical disorders have continued
to increase. Simultaneously, the percentage of individuals
fully abstaining from ethanol has decreased. In previous studies and in routine health care, reference intervals for GT determinations have been based on values obtained from mixed
populations of apparently healthy moderate drinkers and
abstainers, whereas only limited attention has been paid on the
exact amounts of ethanol consumption in these individuals.
In this work we explored the relationship between ethanol
consumption and GT values in individuals with a wide variety
of ethanol consumption. Our data indicate distinct effects of
mild to moderate ethanol consumption on serum GT levels,
which should be considered in the clinical use of GT measurements as a marker of ethanol abuse and liver status.
Serum GT was first measured from a sample of 195 individuals (103 heavy drinkers: 90 men, mean age 42 ± 10 years;
13 women, mean age 40 ± 7 years, and 92 moderate drinkers
The procedure was approved by the institutional review board.
Informed consent was obtained from the participants and the
*Author to whom correspondence should be addressed at: Tel.: +358 6 415
4719; Fax: +358 6 415 4924; E-mail: email@example.com
Ó The Author 2005. Published by Oxford University Press on behalf of the Medical Council on Alcohol. All rights reserved
J. HIETALA et al.
study was carried out according to the provisions of the
Declaration of Helsinki.
Values are expressed as mean ± SD. Comparisons were made
with Kruskal-Wallis test and Dunn’s Multiple Comparison
Test or Mann–Whitney test when comparing two groups.
Correlations were calculated with Pearson product–moment
correlation coefficients. Reference intervals were calculated
after logarithmic transformation as previously described
(Horn and Pesce, 2003). A P-value <0.05 was considered
In the total population, serum GT concentrations (mean ± SD)
in the groups of heavy drinkers drinking 40–80 g (68 ± 54 U/l)
or >80 g (167 ± 254 U/l) of ethanol per day significantly
exceeded the levels of both abstainers (P < 0.001) and moderate drinkers (P < 0.001) (Fig. 1). Male alcoholics had slightly
higher GT values (166 ± 267 U/l) than female alcoholics
(130 ± 163 U/l), although the difference between genders did
not reach significance. Interestingly, GT values in the group of
moderate drinkers with a daily consumption of 1–40 g
(28 ± 23 U/l) also exceeded the values obtained from the
group of abstainers (24 ± 17 U/l) (P < 0.001). The correlation
between ethanol consumption and GT values, as calculated
from the individuals interviewed with the time-line followback method, was significant (r = 0.35, P < 0.001).
Figure 2 demonstrates the previously established changes in
national GT reference intervals in Finland in comparison with
the yearly changes in ethanol consumption at population level.
The data on GT reference intervals, as calculated from the
individuals classified according to ethanol consumption data
in the detailed personal interviews, are summarized in
Table 1. The upper normal limits were found to be on average
43% higher, when the individuals with moderate drinking are
contrasted with the population of abstainers. The upper normal
limits for men were higher than those for women and the age
group >40 years had higher levels than the age group 18–
39 years in both genders. However, the correlation between
GT levels and age per se did not reach significance (r = 0.097).
The effect of the choice of the reference population on the
estimated diagnostic sensitivity of GT as a marker of excessive ethanol consumption is demonstrated in Fig. 3. When
the heavy drinkers are contrasted with abstainers, 69% of
heavy drinkers become correctly classified. If the reference
interval and definition of normal values would be based on
moderate drinkers, the sensitivity remains at a level of 56%.
In separate analyses of men and women, the corresponding
percentages were 68 and 54% for men and 77 and 69% for
women, showing essentially similar changes.
Fig. 1. GT values in individuals with different levels of ethanol consumption.
The values in the groups with ethanol consumption over 80 g (167 ± 254 U/l)
or 40–80 g (68 ± 54 U/l) are significantly higher than those in the abstainers
(24 ± 17 U/l) or moderate drinkers consuming a mean of 1–40 g/day
(28 ± 23 U/l). The difference between the group of moderate drinkers and
abstainers was also significant (P < 0.001).
Alcohol abuse and alcoholism rank as one of the most serious
health problems in most Western countries (Room et al.,
2005). Therefore, a high priority should be given to reduction
in the prevalence of alcoholism through more effective diagnosis and early intervention. Objective methods for detecting
excessive ethanol consumption in health care are necessary
for a majority of heavy drinkers who have not self-identified
as having alcohol problems.
Studies in the past have shown that a number of biochemical
parameters are altered in alcoholics, of which serum GT has
emerged as one of the most efficient tests (Bagrel et al.,
1979; Chick et al., 1981; Papoz et al., 1981; Bernadt et al.,
1982; Leino et al., 1995; Anttila et al., 2004). The present
study indicates that even moderate amounts of ethanol consumption influence serum GT concentrations at population
level and this phenomenon may significantly affect the interpretation and the establishment of common reference intervals
for GT measurements in health care. The data support the view
that in order to improve the diagnostic potential of laboratory
markers of excessive ethanol consumption and liver status, the
reference intervals of each test should be based on healthy
Table 1. GT reference intervals based on the data from the groups of individuals classified according to ethanol intake in the interview sample
Alcohol consumption (1–40 g/day)
Moderate drinkers and abstainers
Alcohol consumption (0–40 g/day)
Alcohol consumption (0 g/day)
GT AND ALCOHOL CONSUMPTION
Liters of 100 % ethanol per capita / year
Yearly ethanol consumption (Finland)
Recommendations for national reference intervals for
GT measurements (Finland)
≥ 40 yrs
Percentages of alcoholics correctly
Fig. 2. (A) Mean ethanol consumption in Finland over the years 1975–2004. (B) The changes in recommended reference intervals for GT measurements based on
surveys on apparently healthy individuals.
alcohol consumption 1–40 g/day
alcohol consumption 0 g/day
Source of reference interval
Fig. 3. Effect of the source of reference intervals on the sensitivity of detecting problem drinking with GT measurements. When reference intervals are
based on values from abstainers, 69% of the heavy drinkers are detected.
When the reference population consists of moderate drinkers, only 56% of
the heavy drinkers become correctly identified.
individuals who abstain from ethanol. Since a gold standard
for a bona fide social drinker currently does not exist, the concepts of moderate drinking and social drinking should also be
defined more accurately and the proportions of abstainers and
moderate drinkers considered separately when selecting reference individuals in future studies.
The present data indicate that the estimated upper normal
limits for GT measurements would be 40% higher if the
data based on moderate drinkers would be used as the basis
of the reference population instead of abstainers. In accordance with this view, a recent NORIP survey from the Nordic
countries showed markedly increased GT reference values
(Stromme et al., 2004). The diagnostic sensitivity of GT measurements as a marker of excessive ethanol consumption
would obviously improve if reference intervals would be
based on the data from abstainers. This work indicates that
13% of alcoholics would escape detection if moderate drinkers
are used as the reference population, instead of abstainers.
Thus, there may be a need for revising the reference range
downwards. It may, however, be argued that setting a lower
limit could worsen the specificity of GT assays and lead to a
high number of false positive values. According to this work,
11% of the moderate drinkers would have shown increased
values. However, there may be individuals who are in the
upper range of the limits of social drinking. Since the data
are based on self-reports, we cannot rule out occult alcohol
abuse in these subjects. It should be noted, however, that
future studies are clearly warranted to explore the independent
effect of various possible sources of unspecificity on GT
J. HIETALA et al.
values, such as obesity or diabetes in individuals reporting
either moderate drinking or no drinking. Our preliminary
analyses on moderate drinkers with different degrees of
obesity have indicated potentiation of GT activities in individuals with significant obesity (data not shown). The associations between GT, moderate drinking, and obesity have
previously been examined by Kornhuber et al. (1989) who
also concluded that the definition of GT normal values may
need to be readdressed. The correlation (r = 0.35) between
alcohol consumption per se and GT values in this study is consistent with previous observations (Bagrel et al., 1979; Chick
et al., 1981; Papoz et al., 1981; Leino et al., 1995; Anttila
et al., 2004). The correlation was, however, essentially similar
in women (r = 0.36) and men (r = 0.32), though some earlier
studies have reported higher correlations in populations consisting of men only (Papoz et al., 1981; Anton and Moak,
1994; Sillanaukee et al., 1998). The diagnostic sensitivity of
GT has usually been shown to be lower for women than for
men (Anton and Moak, 1994; Yersin et al., 1995; Mundle
et al., 2000). Based on the present data, these findings may
in part be explained by the definition of reference intervals.
Furthermore, GT values in women may increase at lower
levels of alcohol consumption as a result of women’s
increased vulnerability to the toxic effects of alcohol (Anton
et al., 1998).
The advent of carbohydrate-deficient transferrin (CDT)
testing has recently imposed a new challenge to the use of
GT measurements because CDT has shown higher specificities than GT in several trials. Although CDT has become
an increasingly important tool for assessing excessive ethanol
consumption, it appears that CDT and GT frequently increase
in different individuals (Anton et al., 2002; Anttila et al.,
2003; Neumann and Spies, 2003). Some studies have concluded that GT is more efficient in identifying female alcoholics than CDT (Anton and Moak, 1994, Anton et al.,
2002). Therefore, it seems at this time that CDT alone does
not cover all the needs for an alcohol marker in routine clinical
practice, and other markers, especially GT are needed as well.
Taken together, the present data indicate that the changes in
drinking behaviour at population level may parallel increases
in recommended GT cut-offs, which may subsequently lead
to problems in recognizing excessive alcohol consumption in
its early phase. Therefore, a critical re-evaluation of reference
intervals even in the use of the well-established biochemical
markers of alcohol consumption may be necessary in order
to improve the assessment and treatment of patients with
early-stage alcohol problems.
Acknowledgements — The help of professor Pål Rustad, Fürst Medical Laboratory, Oslo, Norway, for providing data on GT measurements in the Nordic
NORIP Survey for establishing reference intervals is gratefully acknowledged.
The studies were supported in part by a grant from the Finnish Foundation for
Anton, R. F. and Moak, D. H. (1994) Carbohydrate-deficient transferrin and gamma-glutamyltransferase as markers of heavy alcohol
consumption: gender differences. Alcoholism: Clinical and
Experimental Research 18, 747–754.
Anton, R. F., Stout, R. L., Roberts, J. S. et al. (1998) The
effect of drinking intensity and frequency on serum
carbohydrate-deficient transferrin and gamma-glutamyl transferase levels in outpatient alcoholics. Alcoholism: Clinical and
Experimental Research 22, 1456–1462.
Anton, R. F., Lieber, C. and Tabakoff, B. (2002) Carbohydratedeficient transferrin and gamma-glutamyltransferase for the
detection and monitoring of alcohol use: results from a multisite
study. Alcoholism: Clinical and Experimental Research 26,
Anttila, P., Järvi, K., Latvala, J. et al. (2003) A new modified
gamma-%CDT method improves the detection of problem
drinking: studies in alcoholics with or without liver disease.
Clinica Chimica Acta 338, 45–51.
Anttila, P., Järvi, K., Latvala, J. et al. (2004) Method-dependent characteristics of carbohydrate-deficient transferrin measurements in
the follow-up of alcoholics. Alcohol and Alcoholism 39, 59–63.
Bagrel, A., d’Houtaud, A., Gueguen, R. et al. (1979) Relations
between reported alcohol consumption and certain biological
variables in an ‘‘unselected’’ population. Clinical Chemistry 25,
Bernadt, M. W., Mumford, J., Taylor, C. et al. (1982) Comparison of
questionnaire and laboratory tests in the detection of excessive
drinking and alcoholism. The Lancet 1, 325–328.
Chick, J., Kreitman, N. and Plant, M. (1981) Mean cell volume and
gamma-glutamyl-transpeptidase as markers of drinking in
working men. The Lancet 1, 1249–1251.
Horn, P. S. and Pesce, A. J. (2003) Reference intervals: an update.
Clinica Chimica Acta 334, 5–23.
Kornhuber, J., Kornhuber, H. H., Backhaus, B. et al. (1989) The
normal values of gamma-glutamyltransferase are falsely defined
up to now: on the diagnosis of hypertension, obesity and diabetes
with reference to ‘‘normal’’ consumption of alcohol.
Versicherungsmedizin 41, 78–81.
Leino, A., Impivaara, O., Irjala, K. et al. (1995) Health-based reference intervals for ALAT, ASAT and GT in serum, measured
according to the recommendations of the European Committee
for Clinical Laboratory Standards (ECCLS). The Scandinavian
Journal of Clinical and Laboratory Investigation 55, 243–250.
Mundle, G., Munkes, J., Ackermann, K. et al. (2000) Sex
differences of carbohydrate-deficient transferrin, gammaglutamyltransferase, and mean corpuscular volume in alcoholdependent patients. Alcoholism: Clinical and Experimental
Research 24, 1400–1405.
Neumann, T. and Spies, C. (2003) Use of biomarkers for alcohol use
disorders in clinical practice. Addiction 98 (Suppl. 2), 81–91.
Papoz, L., Warnet, J. M., Pequignot, G. et al. (1981) Alcohol
consumption in a healthy population. Relationship to gammaglutamyl transferase activity and mean corpuscular volume. The
Journal of the American Medical Association 245, 1748–1751.
Persson, J., Magnusson, P. H. and Borg, S. (1990) Serum gammaglutamyl transferase (GGT) in a group of organized teetotallers.
Alcohol 7, 87–89.
Reyes, E. and Miller, W. R. (1980) Serum gamma-glutamyl transpeptidase as a diagnostic aid in problem drinkers. Addictive Behaviors
Room, R., Babor, T. and Rehm, J. (2005) Alcohol and public health.
The Lancet 365, 519–530.
Sillanaukee, P., Aalto, M. and Seppä, K. (1998) Carbohydratedeficient transferrin and conventional alcohol markers as indicators for brief intervention among heavy drinkers in primary health
care. Alcoholism: Clinical and Experimental Research 22,
Stromme, J. H., Rustad, P., Steensland, H. et al. (2004) Reference
intervals for eight enzymes in blood of adult females and males
measured in accordance with the International Federation of
Clinical Chemistry reference system at 37 C: part of the Nordic
Reference Interval Project. The Scandinavian Journal of Clinical
and Laboratory Investigation 64, 371–384.
Yersin, B., Nicolet, J. F., Dercrey, H. et al. (1995) Screening for
excessive alcohol drinking. Comparative value of carbohydratedeficient transferrin, gamma-glutamyltransferase, and mean
corpuscular volume. Archives of Internal Medicine 155,
Zein, M. and Discombe, G. (1970) Serum gamma-glutamyl transpeptidase as a diagnostic aid. The Lancet 2, 748–750.
Alcohol & Alcoholism Vol. 41, No. 5, pp. 528–533, 2006
Advance Access publication 23 June 2006
COMPARISON OF THE COMBINED MARKER GGT–CDT AND THE CONVENTIONAL
LABORATORY MARKERS OF ALCOHOL ABUSE IN HEAVY DRINKERS, MODERATE
DRINKERS AND ABSTAINERS
JOHANNA HIETALA, HEIDI KOIVISTO, PETRA ANTTILA and ONNI NIEMELÄ*
Department of Laboratory Medicine and Medical Research Unit, Seinäjoki Central Hospital and University of Tampere,
FIN-60220 Seinäjoki, Finland
(Received 6 January 2006; first review notified 19 April 2006; in revised form 5 May 2006; accepted 19 May 2006;
advance access publication 23 June 2006)
Abstract — Aims: A combined index based on g-glutamyltransferase (GGT) and carbohydrate-deficient transferrin (CDT) measurements (GGT–CDT) has been recently suggested to improve the detection of excessive ethanol consumption. The aim of this work
was to compare GGT–CDT with the conventional markers of alcohol abuse in individuals with a wide variety of alcohol consumption.
Methods: A cross-sectional and follow-up analysis was conducted in a sample of 165 heavy drinkers, consuming 40–540 g of ethanol
per day, and 86 reference individuals who were either moderate drinkers (n = 51) or abstainers (n = 35). Results: GGT–CDT
(5.35 ± 1.08) in the heavy drinkers was significantly higher than in the reference individuals (3.30 ± 0.37). The sensitivity of GGT–
CDT (90%) in correctly classifying heavy drinkers exceeded that of CDT (63%), GGT (58%), mean corpuscular volume (MCV)
(45%), aspartate aminotransferase (AST) (47%), and alanine aminotransferase (ALT) (50%), being also essentially similar for
alcoholics with (93%) or without (88%) liver disease. When comparing the data using either moderate drinkers or abstainers as
reference population, the sensitivity of GGT–CDT, CDT, and ALT remained unchanged whereas the sensitivity of GGT, MCV, and
AST was found to show variation. Conclusions: GGT–CDT improves the sensitivity of detecting excessive ethanol consumption as
compared with the traditional markers of ethanol consumption. These findings should be considered in the assessment of patients
with alcohol use disorders.
the CDTect assay. However, as yet, only limited information
has been available on the clinical performance of the new
GGT–CDT method and its comparisons with other markers
of ethanol intake.
This study was aimed at comparing the diagnostic value of
GGT–CDT with several other biomarkers of alcohol abuse in
a well-characterized population of heavy drinkers, moderate
drinkers, and abstainers.
Although the prevalence of alcoholism and associated medical
disorders are continuously growing in most Western countries,
patients with hazardous drinking practices continue to escape
detection in clinical work (Lieber, 1995; Conigrave et al.,
2002; Niemelä, 2002; Room et al., 2005). Therefore, the
need for objective laboratory tests, which respond to excessive
ethanol intake in a sensitive and specific manner, has been
widely recognized. In spite of the fact that a wide variety of
biochemical parameters in circulation are altered in alcoholics, none of them has so far provided enough diagnostic accuracy to meet the demands of clinicians in routine use for
differentiating between alcoholics and non-alcoholics (Allen
et al., 1994; Salaspuro, 1999; Scouller et al., 2000; Arndt,
2001; Conigrave et al., 2002; Niemelä, 2002).
Recent studies have suggested the possibility of using marker combinations, which could improve assay sensitivities
without sacrificing specificity (Sillanaukee and Olsson,
2001; Anttila et al., 2003a). A mathematically formulated
equation derived from the data of g-glutamyltransferase
(GGT) and carbohydrate-deficient transferrin (CDT) measurements appears to be elevated in a higher percentage of
alcoholics than either GGT or CDT alone. This approach,
when the latter component of the equation is replaced by the
results from %CDT assay, which expresses CDT data as
percentage of total transferrin, was recently found to further
improve this method as compared with the previous
CDTect-based equation (Anttila et al., 2003a). The main
advantage achieved by the %CDT method in this context
is its ability to avoid the interference of unexpected variation
in serum transferrin levels, which complicated the use of
The sample of alcohol abusers consisted of 165 heavy drinkers
(140 men, 25 women), age (mean ± SD) 46 ± 10 years (range
19–73 years). All patients showed a well-documented history
of excessive ethanol consumption, as assessed by detailed personal interviews using a time-line follow-back technique and
medical and social records. They had consumed ethanol in
amounts ranging from 40 to 540 g/day during the past 1 month
either continuously or during repeated episodes of binge
drinking. The patients also met the DSM-IV criteria of alcohol
dependence including pathological alcohol use, social impairment, presence of tolerance, and withdrawal symptoms. There
were 51 patients, (38 men, 13 women) age 49 ± 10 years,
range 30–67 years, who also showed evidence of liver disease,
as assessed by previously established combined morphological index (CMI) or a combined clinical and laboratory index
(CCLI) (Orrego et al., 1983; Blake and Orrego, 1991).
The mean duration of abstinence prior to sampling was
2 ± 2 days, range 0–6 days. All patients were negative for hepatitis B antigen or hepatitis C serology. Follow-up studies with
supervised abstinence for assessing marker normalization
rates were carried out in 44 alcoholics, age 42 ± 11 years
(range 19–59 years), who were monitored for a period of
11 ± 4 days, range 3–19 days.
*Author to whom correspondence should be addressed. Tel: +358 6 415 4719;
Fax: +358 6 415 4924; E-mail: firstname.lastname@example.org
The Author 2006. Published by Oxford University Press on behalf of the Medical Council on Alcohol. All rights reserved
BIOMARKERS OF ALCOHOL CONSUMPTION
Reference individuals were 86 healthy volunteers (49 men,
37 women), age 48 ± 17 years (range 19–84 years) who were
either abstainers (n = 35, age 54 ± 18 years, range 22–
84 years) or moderate drinkers (n = 51, age 43 ± 14 years,
range 19–77 years) whose mean ethanol consumption per
day, as also assessed from the past one month by questionnaires, was between 1 and 40 g. All of them were without
any previous social or medical history of alcohol abuse. The
weekly ethanol consumption in these individuals did not
exceed 280 g of alcohol (men) and 160 g (women) or 6 drinks
(men) and 4 drinks (women) on any single occasion.
All serum samples were stored at –70 C until analysis. All
participants gave their informed consent and the study was
carried out according to the provisions of the Declaration of
Helsinki. The study was approved by the hospital ethical
Microsoft Excel software, which yielded a cut-off of 4.18 for
men and 3.81 for women.
The values are expressed as means ± SD. The comparisons
between groups were carried out using the Kruskal–Wallis
test with the Dunn’s test for multiple comparisons. Correlations were calculated using the Pearson product-moment correlation coefficients for continuous non-skewed parameters
or the Spearman’s rank correlations for non-continuous
variables, as required. Statistical analyses were carried out
using GraphPad Prism, version 3.03 (GraphPad Software,
San Diego, CA) and Analyse-It for Microsoft Excel software
(version 1.68), Leeds, UK. A P-value < 0.05 was considered
The concentration of CDT was measured by a turbidimetric
immunoassay (TIA) after ion exchange chromatography
(%CDT, Axis-Shield, Oslo, Norway). The assay detects
primarily a-, mono- and disialotransferrin, although there
may be some reactivity towards the trisialofraction of CDT,
as recently reported by Aldén et al. (2005). The measurements
were carried out on Behring Nephelometer II (Dade Behring,
Behring Diagnostics GmbH, Marburg, Germany). The
within-run precision was 4.7%, day-to-day variation was
6.0%, and accuracy 12.7%. Serum GGT was measured using
enzymatic colorimetric assay, as standardized against IFCC
(International Federation of Clinical Chemistry and Laboratory Medicine). The imprecision within run was 0.85% and
the day-to-day variation was 0.54%. The accuracy of the
GGT method was found to be 5.0%. Serum aspartate aminotransferase (AST) and serum alanine aminotransferase (ALT)
were analysed using pyridoxal phosphate methods according
to IFCC. For AST, the within-run precision was 0.90%, the
day-to-day precision was 1.5%, and accuracy of the method
was 10.6%. For ALT, the corresponding values were: within
run 0.87%, day-to-day 1.1%, and accuracy 12.1%. The analyses of GGT, AST, and ALT were carried out with Cobas
Integra 800 analyser (Roche Diagnostics, Basel, Switzerland).
Mean corpuscular volume (MCV) of erythrocytes was
measured using the Sysmex XE-2100 hematology analyser
(Sysmex Corporation, Kobe, Japan). The imprecision within
run was 0.41% and the day-to-day variation was 0.37%. The
accuracy of the method was 3.7%. All assays were carried
out in an accredited (SFS-EN ISO/IEC 17025) laboratory of
the Seinäjoki Central Hospital, Finland. The cut-offs in the
above assays were as follows: GGT, men 80 U/l, women
50 U/l; CDT, 2.6%; MCV, 96 fl; AST, men 50 U/l, women
35 U/l; ALT, men 50 U/l, women 35 U/l.
Counting the combined marker
GGT–CDT was counted using an equation based on the data
derived from GGT and CDT measurements as follows:
GGT–CDT = 0.8 · ln(GT) + 1.3 · ln(%CDT) (Anttila et al.,
2003a). Analyses for assay cut-offs from the present reference
population were made with ROC analyses using Analyse-It for
The mean values for GGT–CDT, GGT, CDT, MCV, AST, and
ALT were all significantly higher in the heavy drinkers
than those in the moderate drinkers or abstainers (Fig. 1,
P < 0.001 for all comparisons). The sensitivities and specificities of the various markers in differentiating between the
heavy drinkers and the reference individuals are summarized
in Table 1. GGT–CDT reached a sensitivity of 90% for a specificity of 98%, which clearly exceeded the sensitivities
achieved by all the other markers in these comparisons. The
sensitivities and specificities of GGT–CDT were high for
both men and women (Table 1). Combining GGT and CDT
in a manner where either marker is positive obviously yielded
a higher sensitivity (85%) than the assays alone, but even this
approach did not reach the sensitivity of the mathematically
formulated combination (Table 1). When the alcoholic
patients were further classified according to the presence
or absence of liver disease, GGT–CDT showed essentially
similar diagnostic accuracies for both groups, whereas the
diagnostic characteristics of GGT, MCV, AST and ALT
were found to change as a function of liver status (Table 1).
The degree of liver disease severity, as assessed by the combined morphological index (CMI), was not found to correlate
significantly with GGT–CDT (r = 0.19).
When the marker data obtained from the heavy drinkers
were contrasted alternatively with either moderate drinkers
or abstainers, the sensitivity percentages of GGT (+4%),
MCV (+17%), and AST (+20%) all increased when abstainers
were used as the only control population. In contrast, GGT–
CDT and CDT sensitivities were not affected. The threshold
ethanol consumption for elevation of GGT–CDT values
when plotting daily ethanol consumption and GGT–CDT
was found to be 40 g of ethanol (Figure 2). GGT–CDT
also correlated more strongly with self-reported ethanol consumption (r = 0.76, P < 0.001) than either GGT (r = 0.71,
P < 0.001) or CDT (r = 0.59, P < 0.001) alone or any of the
other markers (Table 2).
In the follow-up of 44 alcoholic patients, GGT–CDT was
found to decrease in 93% of these individuals during
11 ± 4 days (range 3–19 days) of supervised abstinence
(Table 3). The estimated time for normalization for GGT–
CDT, depending on the initial value, was 18 ± 9 days, the
mean rate of decay being 1.5% of the initial value per day,
J. HIETALA et al.
Heavy drinkers Moderate drinkers
Heavy drinkers Moderate drinkers
Heavy drinkers Moderate drinkers
Heavy drinkers Moderate drinkers
Heavy drinkers Moderate drinkers
Heavy drinkers Moderate drinkers
Fig. 1. Box plots of various laboratory markers of alcohol consumption in heavy drinkers, moderate drinkers, and abstainers. Alcohol abusers show
significantly higher values than moderate drinkers or abstainers in all comparisons (P < 0.001). GGT–CDT, combined marker based on the data from GGT
and CDT measurements; GGT, g-glutamyltransferase; CDT, carbohydrate-deficient transferrin; MCV, mean corpuscular volume; AST, aspartate aminotransferase; ALT, alanine aminotransferase.
as compared with 3.4% for GGT and 3.7% for CDT. Nevertheless, the changes towards normalization were found to be more
consistent for GGT–CDT and CDT than those for the other
markers (Table 3).
In the diagnostics of alcohol use disorders, it is crucial that the
laboratory analyses are accurate. Such analyses are needed
BIOMARKERS OF ALCOHOL CONSUMPTION
Table 1. Sensitivities of laboratory markers of alcohol consumption in heavy drinkers, as also divided by gender and liver status, and the specificities, as
obtained from the current reference population
Heavy drinkers with
n = 165
n = 140
n = 25
n = 51
n = 38
n = 13
n = 114
n = 102
n = 12
GGT or CDT
GGT and CDT
Heavy drinkers without
both for detecting excessive ethanol consumption and for
monitoring abstinence. An ideal assay should provide
both specificity and sensitivity near 100%. However, to
date such assays have not emerged, although marker combinations have been suggested to open new possibilities for
improving the situation (Salaspuro, 1987; Anton et al., 2001;
Sillanaukee and Olsson, 2001; Anttila et al., 2003a; Hock
et al., 2005).
The present work compares a recently developed marker
combination, defined here as GGT–CDT, with several other
biochemical markers in the diagnosis of alcohol abuse. The
data show that the sensitivity of GGT–CDT, which relies on
a mathematically formulated equation based on serum GGT
and CDT results, exceeds the diagnostic sensitivity of all the
conventional markers of alcohol abuse. Interestingly, this
advantage is achieved without sacrificing assay specificity.
When using standardized methods for the GGT and CDT
measurements, the combination can well be standardized for
multi-laboratory use, too.
It appears that weighing GGT and CDT in the present manner is important for optimizing assay sensitivity. Previously,
combining independent measurements of GGT and CDT
(when either of the markers is positive) has been shown to
yield a sensitivity of 90% for a specificity of 81% in men
and a sensitivity of 75% for a specificity of 87% in women
(Anton et al., 2001). Schwan et al. (2004) recently showed
that combining GGT and CDT as independent parameters provides a sensitivity of 90% in alcohol abusers and 99% in the
alcohol-dependent group, whereas the specificity remains at
a level of only 63%. Hock et al. (2005) recently combined
log GGT and CDT and obtained a sensitivity of 83% with a
specificity of 95%. The sensitivity increased to 88% when
MCV was included as a third component in the analyses.
The use of MCV or any other parameter as a third component
in the present material was, however, not found to lead to any
additional improvement as compared with GGT–CDT alone
(data not shown).
The first studies employing the combination of GGT and
CDT used similar cut-offs for men and women (Sillanaukee
and Olsson, 2001; Anttila et al., 2003a). The present work
shows, however, that different cut-offs for men and women
may be necessary also for this marker. Previously, genderdependent analytical variation has been observed especially
in CDT assays with the CDTect method, which has been later
abolished by the use of %CDT assays (Anton and Moak, 1994;
Tsutsumi et al., 1994; Niemelä et al., 1995; Viitala et al.,
1998; Conigrave et al., 2003). Women are known to be more
sensitive to the hepatotoxic effects of alcohol, and it is possible that the activities of serum GGT may also respond to
alcohol consumption in a gender-dependent manner, which
could explain in part the need for separate cut-offs for men
and women. In this material, the presence of liver pathology
was found to also affect the interpretation of GGT, AST,
ALT, and MCV in the assessment of heavy drinking, whereas
GGT–CDT appears to be more resistant towards the variation
induced by liver pathology per se. The association between
GGT–CDT combination and liver disease may, however,
also depend on the method used for analysing the CDT component of the assay (Viitala et al., 1998; Anttila et al.,
2003b; Fleming et al., 2004).
The diagnostic potential of GGT–CDT is also supported by
its strong correlation with the actual amounts of ethanol consumption from the past 1 month prior to sampling. The values
appear to increase after the daily ethanol consumption exceeds
40 g. Previously, CDT has been reported to elevate with daily
ethanol consumption ranging from 40 to 80 g, possibly also
depending on the method used (Stibler, 1991; Schellenberg
et al., 2005). While GGT has also been suggested to increase
with a threshold consumption of over 40 g of ethanol per day
(Sharpe, 2001), even moderate drinkers may, however, show
increased GGT values more often than abstainers (Hietala
et al., 2005). Interestingly, recent data have suggested that
GGT could actually be considered a marker of oxidative stress
(Lim et al., 2004). It is, thus, possible that combining GGT and
J. HIETALA et al.
Table 2. Correlations between biochemical markers of ethanol
consumption, and markers and self-reported ethanol consumption
0.71*** 0.59*** 0.52*** 0.59*** 0.50***
0.33*** 0.85*** 1
0.37*** 0.38*** 0.33***
***P < 0.001.
EtOH g/day during previous month
Table 3. Normalization rates for alcohol markers based on follow-ups of
44 alcoholic patients with supervised abstinence for a period 11 ± 4 days
(days), mean ± SD
18 ± 9
16 ± 8
16 ± 11
13 ± 20
16 ± 19
N.D. not determined.
The sensitivities and specificities are expressed as percentages.
GGT–CDT, combined marker based on the data from GGT and
EtOH g/day during previous month
EtOH g/day during previous month
Fig. 2. Correlations of GGT–CDT, GGT, and CDT with daily ethanol consumption from the past 1 month prior to sampling. Marker values showed an
increase after a threshold consumption of 40 g/day, GGT–CDT showing
the strongest correlation with the amount of ethanol consumption.
CDT could provide new diagnostic windows with synergistic
benefits for the assessment of hazardous drinking practices.
GGT–CDT seems to recognize ethanol overconsumption
in a similar manner whether or not heavy drinkers are
contrasted with abstainers or moderate drinkers, which may
be a useful characteristic for instance in screening
programmes for excessive alcohol consumption. It should
be noted that in routine health care reference populations in
trials examining diagnostic tests have usually consisted of
combined populations of moderate drinkers and abstainers.
Here inclusion of moderate drinkers into the reference population was found to affect the diagnostic performance of GGT,
AST, and MCV although not GGT–CDT. The combined
marker was also found to be suitable for the follow-up of
abstinence, showing a rather consistent decline during supervised abstinence, with mean normalization rate of 2–3 weeks.
Interestingly, the time required for normalization for GGT–
CDT appeared slightly longer than for each marker separately,
suggesting that the mathematically formulated combination
also follows a slightly different kinetics in its decay.
Taken together, this work supports the idea of using GGT–
CDT for achieving a more sensitive diagnosis of alcohol
abuse. Since this approach is also cost effective and easy to
manage in hospital laboratories, it should be suitable for
routine clinical work.
Aldén, A., Ohlson, S., Påhlsson, P. et al. (2005) HPLC analysis of
carbohydrate deficient transferrin isoforms isolated by the AxisShield %CDT method. Clinica Chimica Acta 356, 143–146.
Allen, J. P., Litten, R. Z., Anton, R. F. et al. (1994) Carbohydratedeficient transferrin as a measure of immoderate drinking:
remaining issues. Alcoholism: Clinical and Experimental
Research 18, 799–812.
Anton, R. F., Dominick, C., Bigelow, M. et al. (2001) Comparison
of Bio-Rad %CDT TIA and CDTect as laboratory markers of
heavy alcohol use and their relationships with gammaglutamyltransferase. Clinical Chemistry 47, 1769–1775.
Anton, R. F. and Moak, D. H. (1994) Carbohydrate-deficient transferrin and gamma-glutamyltransferase as markers of heavy alcohol
BIOMARKERS OF ALCOHOL CONSUMPTION
consumption: gender differences. Alcoholism: Clinical and
Experimental Research 18, 747–754.
Anttila, P., Järvi, K., Latvala, J. et al. (2003a) A new modified
gamma-%CDT method improves the detection of problem drinking: studies in alcoholics with or without liver disease. Clinica
Chimica Acta 338, 45–51.
Anttila, P., Järvi, K., Latvala, J. et al. (2003b) Diagnostic characteristics of different carbohydrate-deficient transferrin methods in the
detection of problem drinking: effects of liver disease and alcohol
consumption. Alcohol and Alcoholism 38, 415–420.
Arndt, T. (2001) Carbohydrate-deficient transferrin as a marker of
chronic alcohol abuse: a critical review of preanalysis, analysis,
and interpretation. Clinical Chemistry 47, 13–27.
Blake, J. and Orrego, H. (1991) Monitoring treatment of alcoholic
liver disease: evaluation of various severity indices. Clinical
Chemistry 37, 5–13.
Conigrave, K. M., Degenhardt, L. J., Whitfield, J. B. et al. (2002)
CDT, GGT, and AST as markers of alcohol use: the WHO/ISBRA
collaborative project. Alcoholism: Clinical and Experimental
Research 26, 332–339.
Conigrave, K. M., Davies, P., Haber, P. et al. (2003) Traditional
markers of excessive alcohol use. Addiction 98 (Suppl. 2), 31–43.
Fleming, M. F., Anton, R. F. and Spies, C. D. (2004) A review of
genetic, biological, pharmacological, and clinical factors that
affect carbohydrate-deficient transferrin levels. Alcoholism:
Clinical and Experimental Research 28, 1347–1355.
Hietala, J., Puukka, K., Koivisto, H. et al. (2005) Serum gammaglutamyl transferase in alcoholics, moderate drinkers and abstainers: effect on gt reference intervals at population level. Alcohol
and Alcoholism 40, 511–514.
Hock, B., Schwarz, M., Domke, I. et al. (2005) Validity
of carbohydrate-deficient transferrin (%CDT), gammaglutamyltransferase (gamma-GT) and mean corpuscular
erythrocyte volume (MCV) as biomarkers for chronic alcohol
abuse: a study in patients with alcohol dependence and liver
disorders of non-alcoholic and alcoholic origin. Addiction 100,
Lieber, C. S. (1995) Medical disorders of alcoholism. The New
England Journal of Medicine 333, 1058–1065.
Lim, J. S., Yang, J. H., Chun, B. Y. et al. (2004) Is serum gammaglutamyltransferase inversely associated with serum antioxidants
as a marker of oxidative stress? Free Radical Biology and
Medicine 37, 1018–1023.
Niemelä, O. (2002) Serum diagnosis of alcoholic liver disease and
markers of ethanol intake. In Ethanol and the Liver, Sherman,
D. I. N., Preedy, V. and Watson, R. R. eds, pp. 411–449. Taylor
and Francis, New York and London.
Niemelä, O., Sorvajärvi, K., Blake, J. E. et al. (1995) Carbohydratedeficient transferrin as a marker of alcohol abuse: relationship to
alcohol consumption, severity of liver disease, and fibrogenesis.
Alcoholism: Clinical and Experimental Research 19, 1203–1208.
Orrego, H., Israel, Y., Blake, J. E. et al. (1983) Assessment of
prognostic factors in alcoholic liver disease: toward a global
quantitative expression of severity. Hepatology 3, 896–905.
Room, R., Babor, T. and Rehm, J. (2005) Alcohol and public health.
The Lancet 365, 519–530.
Salaspuro, M. (1987) Use of enzymes for the diagnosis of alcoholrelated organ damage. Enzyme 37, 87–107.
Salaspuro, M. (1999) Carbohydrate-deficient transferrin as compared
to other markers of alcoholism: a systematic review. Alcohol 19,
Schellenberg, F., Schwan, R., Mennetrey, L. et al. (2005) Dose–effect
relation between daily ethanol intake in the range 0–70 grams and
%CDT value: validation of a cut-off value. Alcohol and
Alcoholism 40, 531–534.
Schwan, R., Loiseaux, M. N., Schellenberg, F. et al. (2004) Multicenter validation study of the %CDT TIA kit in alcohol abuse
and alcohol dependence. Alcoholism: Clinical and Experimental
Research 28, 1331–1337.
Scouller, K., Conigrave, K. M., Macaskill, P. et al. (2000) Should we
use carbohydrate-deficient transferrin instead of gammaglutamyltransferase for detecting problem drinkers? A systematic
review and metaanalysis. Clinical Chemistry 46, 1894–1902.
Sharpe, P. C. (2001) Biochemical detection and monitoring of
alcohol abuse and abstinence. Annals of Clinical Biochemistry
Sillanaukee, P. and Olsson, U. (2001) Improved diagnostic classification of alcohol abusers by combining carbohydrate-deficient
transferrin and gamma-glutamyltransferase. Clinical Chemistry
Stibler, H. (1991) Carbohydrate-deficient transferrin in serum: a new
marker of potentially harmful alcohol consumption reviewed.
Clinical Chemistry 37, 2029–2037.
Tsutsumi, M., Wang, J. S. and Takada, A. (1994) Microheterogeneity
of serum glycoproteins in alcoholics: is desialo-transferrin the
marker of chronic alcohol drinking or alcoholic liver injury?
Alcoholism: Clinical and Experimental Research 18, 392–397.
Viitala, K., Lähdesmäki, K. and Niemelä, O. (1998) Comparison of
the axis %CDT TIA and the CDTect method as laboratory tests
of alcohol abuse. Clinical Chemistry 44, 1209–1215.
ALCOHOLISM: CLINICAL AND EXPERIMENTAL RESEARCH
Vol. 30, No. 10
IgAs Against Acetaldehyde-Modiﬁed Red Cell Protein as a
Marker of Ethanol Consumption in Male Alcoholic
Subjects, Moderate Drinkers, and Abstainers
Johanna Hietala, Heidi Koivisto, Jaana Latvala, Petra Anttila, and Onni Niemelä
Background: Alcohol abuse has been shown to result in the production of antibodies against
acetaldehyde-modiﬁed epitopes in proteins. However, as yet, only limited information has been available on the clinical usefulness of such responses as markers of hazardous drinking.
Methods: We developed an ELISA to measure specific IgAs against acetaldehyde-protein
adducts. This method was evaluated in cross-sectional and follow-up studies on male heavy drinkers
with a current ethanol consumption of 40 to 540 g/d (n 5 40), moderate drinkers consuming 1 to
40 g/d (n 5 25), and abstainers (n 5 16). The clinical assessments included detailed interviews on the
amounts and patterns of ethanol consumption and various biochemical markers of alcohol abuse and
Results: The mean antiadduct IgAs (198 28 U/L) in the alcohol abusers were significantly
higher than those in the moderate drinkers (58 11 U/L, po0.001) or abstainers (28 8 U/L,
po0.001). The values of moderate drinkers were also higher than those in abstainers (po0.05). The
amount of ethanol consumed during the period of 1 month preceding blood sampling correlated
strongly with antiadduct IgAs (r 5 0.67, po0.001). The sensitivity (73%) and speciﬁcity (94%) of this
marker were found to exceed those of the conventional laboratory markers of alcohol abuse in comparisons contrasting heavy drinkers with abstainers although not in comparisons contrasting heavy
drinkers with moderate drinkers. During abstinence, antiadduct IgAs disappeared with a mean rate
of 3% per day. In additional analyses of possible marker combinations, antiadduct IgAs, together
with CDT, were found to provide the highest sensitivity and speciﬁcity.
Conclusions: Measurements of antiadduct IgAs may provide a new clinically useful marker of
alcohol abuse, providing a close relationship between marker levels and the actual amounts of recent
Key Words: Ethanol, Marker, Acetaldehyde, Immune Responses.
XCESSIVE ALCOHOL CONSUMPTION has been
shown to induce the production of circulating antibodies, which recognize sequential and conformational
epitopes generated in covalent binding reactions between
proteins and ethanol metabolites (Israel et al., 1986;
Koskinas et al., 1992; Niemelä, 2001; Viitala et al., 1997;
Worrall et al., 1991; Xu et al., 1998). Previously, studies in
alcoholic patients with liver disease have indicated that in
such patients, there is often an increase in serum total IgAs
coinciding with abnormal IgA tissue deposition (Amore
et al 1994; van de Wiel et al 1988). Other studies have
From the Department of Laboratory Medicine and Medical Research
Unit, Seinäjoki Central Hospital, and University of Tampere, Tampere,
Received for publication March 21, 2006; accepted May 30, 2006.
Supported in part by a grant from the Finnish Foundation for Alcohol
Reprint requests: Prof. Onni Niemelä, Seinäjoki Central Hospital,
Laboratory, Seinäjoki, FIN-60220, Finland; Fax:1358-6-415-4924;
Copyright r 2006 by the Research Society on Alcoholism.
Alcohol Clin Exp Res, Vol 30, No 10, 2006: pp 1693–1698
further shown high levels of IgA antibodies directed against
acetaldehyde-derived neoantigens in alcoholic subjects,
suggesting antigen-driven IgA responses (Latvala et al.,
2005; Viitala et al., 1997; Worrall et al., 1991). Worrall
et al. (1996, 1998) have previously suggested the possibility of
using antiadduct IgAs also as markers of ethanol consumption. However, as yet, only a few studies have been
available on clinical evaluations of this approach.
The present study was set out to gain further insight on
the diagnostic utility of measuring antiadduct IgAs by
comparing the IgA marker with several traditional markers of ethanol consumption in male subjects with a wide
range of alcohol consumption.
MATERIALS AND METHODS
Patients and Control Subjects
The study population included 40 male heavy drinkers (mean age
45 11 years), who showed a well-documented history of excessive
ethanol consumption. All of these patients had been admitted for
detoxiﬁcation, but were devoid of clinical and laboratory signs of
significant liver disease. They showed a history of continuous
ethanol consumption or binge drinking, the mean recent consumption
HIETALA ET AL.
being 40 to 540 g/d from the period of 4 weeks before sampling. The
documentation of alcohol abuse was based on detailed personal
interviews using a time-line follow-back technique. The mean duration
of abstinence before sampling was 2 2 days. Nineteen alcoholic
subjects (mean age 42 12 years) volunteered for a follow-up, which
was carried out with supervised abstinence and repeated sampling
during hospitalization over a period of 8 3 days. Blood alcohol
concentrations during this time were controlled by repeated ethanol
analyses from breath air.
Reference individuals were healthy male volunteers who were
either abstainers (n 5 16, age 54 15 years) or moderate drinkers
(n 5 25, age 45 12 years), whose daily ethanol consumption did
not exceed 40 g. All of these subjects were devoid of any history or
clinical evidence of alcohol abuse, recent illnesses, or immunological
All serum samples were stored at 70 1C until analysis. All participants of the study gave their informed consent, and the study was
carried out according to the provisions of the Declaration of Helsinki.
Antiadduct IgA Analyses
Preparation of Acetaldehyde-Modiﬁed Red Cell Protein. For preparation of the test antigen, human erythrocyte protein fraction was
ﬁrst puriﬁed from EDTA blood of an abstainer (Latvala et al., 2005).
The cells were separated by centrifugation and washed 3 times with
phosphate-buffered saline (PBS, 7.9 mM Na2HPO4, 1.5 mM
KH2PO4, 137 mM NaCl, 2.7 mM KCl, pH 7.4). The washed cells
were lysed with polyoxyethylene ether, 0.1% v/v in borate buffer
(Hemolysis Reagent, DIAMATTM Analyzer system, Bio-Rad Laboratories, Hercules, CA), and incubated for 35 minutes at 37 1C to
remove unstable Schiff bases. The hemolyzate containing 12 mg of
red cell protein/mL (in PBS) was stored frozen in aliquots at 70 1C.
For acetaldehyde labeling, the hemolyzate was mixed with acetaldehyde (in PBS) to obtain a ﬁnal acetaldehyde concentration of
10 mM. After incubation for 18 hours at 4 1C in tightly sealed containers, the protein adducts generated were reduced by addition of
sodium cyanoborohydride (10 mM) and gentle mixing for 5 hours
at 4 1C. Protein solutions were then dialyzed exhaustively against
PBS at 4 1C and stored in aliquots for single use at 70 1C. The
unmodiﬁed proteins were obtained by hemolyzing and treating
EDTA blood of an abstainer without acetaldehyde.
Measurements of the Antibody Titers. Microtiter plates (NuncImmuno Plate, MaxisorbTM, InterMed, Denmark) were coated with
acetaldehyde-modiﬁed red cell protein, or corresponding unmodiﬁed
proteins in PBS (3 mg protein in 100 mL/well) and incubated for
112 hours at 137 1C. Nonspecific binding was blocked by incubation
(1 hour, 137 1C) with 0.2% gelatin in PBS (150 mL/well). The samples were diluted (1:40) in PBS, which contained 0.04% Tween-20.
The serum dilutions were allowed to react with the coated proteins
(1 hour, 137 1C), followed by extensive washing with PBS-Tween.
Antibody–antigen complexes were detected using alkaline
phosphatase–linked goat anti-human immunoglobulin IgA (Jackson
ImmunoResearch Laboratories Inc., West Grove, PA). The
immunoglobulins were diluted in PBS-Tween, containing 8 mM
MgCl2 and a small amount of dithiothreitol (DTT). The plates were
incubated overnight at 14 1C and washed with PBS-Tween and 100
mL of p-nitrophenylphosphate solution was added as a color reaction
substrate (alkaline phosphatase substrate kit, Bio-Rad). This reaction was stopped by adding 100 mL NaOH (0.4 M), and the optical
densities were read at 405 nm by an Anthos HTII microplate reader
(Anthos Labtec Instruments, Salzburg, Austria).
Serum carbohydrate-deﬁcient transferrin (CDT) was measured
by a turbidimetric immunoassay (TIA) after ion-exchange
chromatography (Axis-Shield, Oslo, Norway), which separates
transferrin variants with 0 to 2 sialic acid residues. The measurements were carried out on a Behring Nephelometer II (Dade
Behring, Behring Diagnostics GmbH, Marburg, Germany)
according to the instructions of the manufacturer. Serum g-glutamyl
transferase (GGT), aspartate aminotransferase (AST), and mean
corpuscular volume (MCV) of erythrocytes were measured by standard clinical chemical methods in an accredited (SFS-EN 45001,
ISO/IEC Guide 25) laboratory of EP Central Hospital, Seinäjoki,
Finland. Cytokines IL-2, IL-6, IL-8, IL-10, TNF-a, and TGF-b1
were measured using Quantikine high-sensitivity ELISA kits (R&D
Systems, Abington Science Park, UK).
Calculations and Statistical Methods
The data for the study variables are expressed as mean SEM. In
the assays for the antiadduct IgAs, the values (OD405) obtained in a
reaction with the sample and the unconjugated protein (background)
were subtracted from the corresponding values measured
from the reaction between the sample and the acetaldehydeprotein conjugate. The antiadduct IgA values are expressed as U/L,
corresponding to OD405103. The combination marker IgA-CDT
was calculated with an equation logCDT1(IgA/103). The combination of GGT and CDT (GGT-CDT) was counted using a previously
established equation GGT-CDT 5 0.8ln(GGT)11.3ln(%CDT)
(Anttila et al., 2003). The differences between the groups of heavy
drinkers, moderate drinkers, and abstainers were analyzed with an
ANOVA or a Kruskal–Wallis test, as required. A paired t-test was
used for comparing alcoholic subjects before and after abstinence,
and the difference between moderate drinkers and abstainers was
further examined with an unpaired t-test or a Mann–Whitney test, as
required. Correlations were calculated using the Pearson product–
moment correlation coefﬁcients for continuous nonskewed parameters or Spearman’s rank-correlations for noncontinuous or skewed
variables. A p-value o0.05 was considered statistically significant.
The mean antiadduct IgA levels (198 28 U/L) in the
alcohol abusers were significantly higher than those in the
moderate drinkers (58 11 U/L, po0.001) or abstainers
(28 8 U/L, po0.001) (Fig. 1). When comparing only
moderate drinkers and abstainers with each other, the values of moderate drinkers were also significantly higher
than those in abstainers (po0.05).
The diagnostic sensitivity and speciﬁcity of antiadduct
IgA measurements were subsequently compared with
those of the traditional markers of alcohol abuse by contrasting the data obtained from the heavy drinkers with
the data obtained from the abstainers and/or moderate
drinkers (Table 1). In comparison with the abstainers only,
the sensitivity of antiadduct IgAs was 73% for a speciﬁcity
of 94%, which exceeded the corresponding ﬁgures found
for CDT, GGT, MCV, and AST. When both moderate
drinkers and abstainers were included in the reference
population, the sensitivity of antiadduct IgA assay was
65% for a speciﬁcity of 88%, showing essentially similar
analytical characteristics to those of CDT although
remaining higher than those of GGT, MCV, or AST
(Table 1). The antiadduct IgAs correlated significantly to
all the conventional markers of alcohol abuse (Table 2)
and also highly significantly to the actual amount of recent
ethanol consumption (r 5 0.67, po0.001). In comparison
ANTIADDUCT IgAs AND ALCOHOL
P < 0.001***
P < 0.001***
P < 0.001***
P < 0.05*
Anti-adduct IgA (U /l)
P < 0.001***
P < 0.001***
P < 0.001***
P < 0.001***
P < 0.01***
P < 0.05*
Fig. 1. Markers of ethanol consumption in heavy drinkers, moderate drinkers, and abstainers. Antiadduct IgA, carbohydrate-deficient transferrin, and
g-glutamyl transferase values in heavy drinkers are all significantly higher than those in both moderate drinkers and abstainers (po0.001 for all comparisons).
Mean corpuscular volume (MCV) was significantly higher (po0.001) in alcoholic subjects than in abstainers, and aspartate aminotransferase was significantly
higher (po0.05) in alcoholic subjects than in moderate drinkers. Significant differences between moderate drinkers and abstainers were found in antiadduct IgA
and in MCV values (dotted lines).
with the major cytokines, antiadduct IgAs were found to
correlate significantly with serum IL-6 (r 5 0.33, po0.05)
and TNF-a (r 5 0.31, po0.05) levels, whereas not with
IL-2, IL-8, IL-10, or TGF-b1.
Figure 2 demonstrates the results of the follow-up of
heavy drinkers. The values at the initiation (160 31 U/L)
decreased significantly to 131 26 U/L (po0.01) during a
period of 8 3 days of supervised abstinence. After this
time, the levels were still higher than those of moderate
drinkers or abstainers, the difference to the latter group
remaining statistically significant (po0.001). The rate of
antiadduct IgA disappearance was estimated to be about
3% per day, and the mean normalization time to be about
In the analyses for possible marker combinations, the
highest sensitivities and speciﬁcities were obtained by
HIETALA ET AL.
Table 1. Sensitivity and Specificity Characteristics of Various Markers of
Ethanol Consumption When Either Abstainers or Moderate Drinkers Are
Used as the Control Population
Source of cutoff
Abstainers and moderate
Abstainers and moderate
Abstainers and moderate
Abstainers and moderate
Abstainers and moderate
CutoffSensitivity %Specificity %
Cut-offs were determined as mean12SD.
AUC, area under curve; SE, standard error of AUC; IgA, immunoglobulin A; CDT, carbohydrate-deficient transferrin; GGT, g-glutamyl
transferase; MCV, mean corpuscular volume of erythrocytes; AST,
combining IgAs with CDT, which yielded a sensitivity of
90%, a speciﬁcity of 98%, and an area under the curve of
0.966 (Table 3).
The present data indicate that the assessment of IgA
antibodies to acetaldehyde-modiﬁed protein epitopes may
provide a clinically useful tool for diagnosing excessive
ethanol consumption. As IgAs already among moderate
drinkers may exceed the levels observed in teetotallers, it
appears that the generation of such immune responses is
an early event in the physiological consequences induced
by ethanol ingestion, which should also be considered in
Table 2. Correlations Between the Different Markers of Ethanol Intake
po0.05; po0.01; po0.001.
CDT, carbohydrate-deficient transferrin; GGT,g-glutamyl transferase;
MCV, mean corpuscular volume of erythrocytes; AST, aspartate aminotransferase.
studies on the mechanisms of ethanol-induced tissue
Previous studies in patients with alcoholic liver disease
have demonstrated an increase in serum total IgAs and the
appearance of IgA deposits in liver and kidney tissue
(Amore et al., 1994; Tuma and Klassen, 1992; van de Wiel
et al., 1988). More recent evidence has suggested that the
generation of excess IgAs in such patients could in fact be
antigen-driven (Koskinas et al., 1992; Latvala et al., 2005;
Viitala et al., 1997; Worrall et al., 1991). Such specific
antibodies could originate from intestinally induced B-cells
under conditions where the gastrointestinal tract, which is
rich in enzymes capable of metabolizing ethanol to acetaldehyde, is repeatedly exposed to ethanol (Latvala et al.,
2005; Salaspuro, 1996; Seitz et al., 1994; Visapää et al.,
1998). This view is also consistent with the strong correlation between the specific IgA antibody levels and the
amounts of recent ethanol ingestion as well as with the
presence of antiadduct IgAs in individuals reporting
moderate drinking. As alcohol is also known to increase
intestinal permeability, this could further enhance the
immune responses toward the intestinal antigens. The positive correlation found between antiadduct IgAs and the
proinﬂammatory cytokines TNF-a and IL-6 indicates that
there is also an early-phase inﬂammatory response to
ethanol-derived neoantigens. Upon continuing heavy
drinking, the induction of the immune responses and the
cytokine cascades may also play an important role in the
sequence of events leading to liver pathology. Acetaldehyde modiﬁcation of proteins and cellular constituents has
been previously shown to disturb protein function in vivo,
and the generation of IgAs toward ethanol metabolites
could also promote the destruction of such protein modiﬁcations.
The present data indicate a high sensitivity and
speciﬁcity for the adduct-specific IgA measurements in
differentiating between alcoholic subjects and nonalcoholic subjects. Worrall et al. (1996, 1998) have previously
investigated antiadduct IgA responses as markers of alcohol
abuse using acetaldehyde-modiﬁed bovine serum albumin
as a test antigen. Despite the difference in the method and
the nature of protein adducts in these and the present studies, it should be noted that Worrall and coworkers also
found a significant correlation (r 5 0.44) between ethanol
intake and antiadduct IgAs among social drinkers and
alcoholic subjects. Previously, the control populations in
biomarker studies have usually consisted primarily of
social drinkers, and the possible differences occurring
between moderate drinkers and teetotallers have received
less attention. The present data show that the diagnostic
sensitivity of antiadduct IgAs exceeds that of the conventional markers especially when abstainers are used as the
reference individuals. Conversely, inclusion of moderate
drinkers in the reference population decreases the diagnostic sensitivity, indicating that future reference limits for
alcohol markers should perhaps preferably be based on
ANTIADDUCT IgAs AND ALCOHOL
p = 0.0028**
p = 0.0002***
Anti-adduct IgA (U/l)
p = 0.13
p = 0.024*
p = 0.24
Fig. 2. Laboratory parameters of alcoholic subjects before and after 8 3 days of abstinence. The decreases in antiadduct IgA (p 5 0.0028) and carbohydrate-deficient transferrin values (p 5 0.0002) were statistically significant. Mean corpuscular volume values showed a significant increase during abstinence
(p 5 0.024). (A) Alcoholic subjects before abstinence and (B) after abstinence.
abstainers. This approach would obviously lead to some
positive values in individuals with moderate drinking habits, but provide us with a new tool for raising the issue of
possible hazardous drinking practices in an earlier phase.
Current follow-up studies indicate that serum antiadduct IgA levels normalize at an average rate of 3% per
day, the mean time required for normalization being
29 days. As the average half-life of a single IgA molecule is
5 to 6 days, the relatively long appearance of antiadduct
IgAs in the serum may be explained by the fact that acetaldehyde adducts in erythrocytes of alcoholic subjects may
Table 3. Sensitivities and Specificities of Marker Combinations
CDT, carbohydrate-deficient transferrin.
persist in circulation for 1 to 3 weeks after the last dose of
ethanol (Niemelä and Israel, 1992). The consistent
decrease in the antiadduct IgAs upon abstinence indicates,
however, that IgA titers could also be used for monitoring
treatment in recovering alcoholic subjects.
Recent evidence has suggested that further improvement
in the clinical accuracy of alcohol markers may be
obtained through the development of ideal marker combinations. Such combinations have usually consisted of
CDT, GGT, and MCV. Recent studies have indicated the
highest diagnostic accuracy for a mathematically formulated combination of GGT and CDT (Anttila et al., 2003;
Sillanaukee and Olsson, 2001). The present work shows
that the combination of antiadduct IgAs and CDT also
provides improved sensitivity compared with its parent
components without sacriﬁcing assay speciﬁcity. The
diagnostic performance also seems to exceed that found
for the combination of GGT and CDT. It should be noted,
however, that before adopting the IgA-based assay into
HIETALA ET AL.
routine clinical and multilaboratory use, a careful standardization of the method will be needed. This could be
achieved by using internationally standardized antigen
preparations and test protocols including standard color
reaction times and the expression of the assay results using
international reference samples.
In this study, we chose to study men only, because sex
may be a significant confounding factor in studies on
immunological responses in vivo (Kovacs and Messingham,
2002; Makkonen et al., 2001). Although the specific mechanisms underlying these phenomena have remained
unclear, sex hormones appear to play a key role in the
regulation of immune responses. As women generally also
show stronger immune responses and higher serum
immunoglobulin levels, future studies in populations consisting of women only also appear warranted to evaluate
the diagnostic potential and pathogenic significance of the
Taken together, this work shows that measurements of
specific IgAs against acetaldehyde-modiﬁed epitopes in
proteins could be used as a sensitive tool for detecting
heavy drinking. Assays for such immune responses could
also be important in studies on the pathogenesis of
ethanol-induced tissue injury.
Amore A, Coppo R, Roccatello D, Piccoli G, Mazzucco G,
Gomez-Chiarri M, Lamm ME, Emancipator SN (1994) Experimental
IgA nephropathy secondary to hepatocellular injury induced by dietary deﬁciencies and heavy alcohol intake. Lab Invest 70:68–77.
Anttila P, Järvi K, Latvala J, Blake JE, Niemelä O (2003) A new
modiﬁed gamma-%CDT method improves the detection of problem
drinking: studies in alcoholics with or without liver disease. Clin Chim
Israel Y, Hurwitz E, Niemelä O, Arnon R (1986) Monoclonal and
polyclonal antibodies against acetaldehyde-containing epitopes in
acetaldehyde-protein adducts. Proc Natl Acad Sci U S A 83:7923–7927.
Koskinas J, Kenna JG, Bird GL, Alexander GJ, Williams R (1992)
Immunoglobulin A antibody to a 200-kilodalton cytosolic acetaldehyde adduct in alcoholic hepatitis. Gastroenterology 103:1860–1867.
Kovacs EJ, Messingham KA (2002) Inﬂuence of alcohol and gender on
immune response. Alcohol Res Health 26:257–263.
Latvala J, Hietala J, Koivisto H, Järvi K, Anttila P, Niemelä O (2005)
Immune responses to ethanol metabolites and cytokine proﬁles differentiate alcoholics with or without liver disease. Am J Gastroenterol
Makkonen K, Viitala KI, Parkkila S, Niemelä O (2001) Serum IgG and
IgE antibodies against mold-derived antigens in patients with symptoms of hypersensitivity. Clin Chim Acta 305:89–98.
Niemelä O (2001) Distribution of ethanol-induced protein adducts
in vivo: relationship to tissue injury. Free Radic Biol Med 31:
Niemelä O, Israel Y (1992) Hemoglobin-acetaldehyde adducts in human
alcohol abusers. Lab Invest 67:246–252.
Salaspuro M (1996) Bacteriocolonic pathway for ethanol oxidation:
characteristics and implications. Ann Med 28:195–200.
Seitz HK, Gartner U, Egerer G, Simanowski UA (1994) Ethanol
metabolism in the gastrointestinal tract and its possible consequences.
Alcohol Alcohol 2 (suppl): 157–162.
Sillanaukee P, Olsson U (2001) Improved diagnostic classiﬁcation of
alcohol abusers by combining carbohydrate-deﬁcient transferrin and
gamma-glutamyltransferase. Clin Chem 47:681–685.
Tuma DJ, Klassen LW (1992) Immune responses to acetaldehydeprotein adducts: role in alcoholic liver disease. Gastroenterology
van de Wiel A, van Hattum J, Schuurman HJ, Kater L (1988) Immunoglobulin A in the diagnosis of alcoholic liver disease. Gastroenterology
Viitala K, Israel Y, Blake JE, Niemelä O (1997) Serum IgA, IgG, and
IgM antibodies directed against acetaldehyde-derived epitopes: relationship to liver disease severity and alcohol consumption. Hepatology
Visapää JP, Jokelainen K, Nosova T, Salaspuro M (1998) Inhibition of
intracolonic acetaldehyde production and alcoholic fermentation in
rats by ciproﬂoxacin. Alcohol Clin Exp Res 22:1161–1164.
Worrall S, de Jersey J, Shanley BC, Wilce PA (1991) Antibodies against
acetaldehyde-modiﬁed epitopes: an elevated IgA response in alcoholics. Eur J Clin Invest 21:90–95.
Worrall S, de Jersey J, Wilce PA, Seppä K, Hurme L, Sillanaukee P
(1996) Relationship between alcohol intake and immunoglobulin a
immunoreactivity with acetaldehyde-modiﬁed bovine serum albumin.
Alcohol Clin Exp Res 20:836–840.
Worrall S, de Jersey J, Wilce PA, Seppä K, Hurme L, Sillanaukee P
(1998) Comparison of carbohydrate-deﬁcient transferrin, immunoglobulin A antibodies reactive with acetaldehyde-modiﬁed protein
and acetaldehyde-modiﬁed albumin with conventional markers of
alcohol consumption. Alcohol Clin Exp Res 22:1921–1926.
Xu D, Thiele GM, Beckenhauer JL, Klassen LW, Sorrell MF, Tuma DJ
(1998) Detection of circulating antibodies to malondialdehydeacetaldehyde adducts in ethanol-fed rats. Gastroenterology 115:686–692.