Content uploaded by Howard B Moss
Author content
All content in this area was uploaded by Howard B Moss on Aug 02, 2019
Content may be subject to copyright.
Content uploaded by Howard B Moss
Author content
All content in this area was uploaded by Howard B Moss on Aug 02, 2019
Content may be subject to copyright.
Alcohol Biomarkers in Applied Settings: Recent Advances
and Future Research Opportunities
Raye Z. Litten, Ann M. Bradley, and Howard B. Moss
During the past decade, advances have been made in the identification, development, and applica-
tion of alcohol biomarkers. This is important because of the unique functions that alcohol bio-
markers can serve in various applied settings. To carry out these functions, biomarkers must
display several features including validity, reliability, adequacy of temporal window of assessment,
reasonable cost, and transportability. During the past two decades, several traditional alcohol bio-
markers have been studied in multiple human studies. Meanwhile, several new, promising biomar-
kers, including various alcohol metabolites and alcohol biosensors, are being explored in human
studies. In addition, researchers have explored using biomarkers in combination and using bio-
markers in combination with self-reports, resulting in increased sensitivity with little sacrifice in
specificity. Despite these advances, more research is needed to validate biomarkers, especially the
new ones, in humans. Moreover, recent advances in high-throughput technologies for genomics,
proteomics, and metabolomics offer unique opportunities to discover novel biomarkers, while
additional research is needed to perfect newly developed alcohol sensors. Development of more
accurate biomarkers will help practicing clinicians to more effectively screen and monitor individ-
uals who suffer from alcohol use disorders.
Key Words: Alcohol Biomarkers, Alcohol Sensors, Alcohol Use Disorders.
THE ALCOHOL USE disorders (AUDs), alcohol abuse,
and alcohol dependence are complex, sometimes devas-
tating, disorders responsible for a host of medical, psychologi-
cal, and social problems (Li, 2007). During the past decade,
new behavioral and pharmacological interventions have been
introduced that can help to arrest AUDs and related harms
before they become severe (Fuller and Hiller-Sturmhofel,
1999; Johnson, 2008; Litten et al., 2005; Miller et al., 2001).
In particular, the use of medications in alcohol-dependent
patients can help to reduce the likelihood and severity of early
relapse (Johnson, 2008; Litten et al., 2005). With these
advances in treatment, biomarkers can be instrumental to
identify persons with alcohol problems who may be in need
of treatment and to monitor a patient’s progress during treat-
ment. Indeed, the potential uses of alcohol biomarkers extend
beyond clinical settings to encompass multiple public safety,
criminal justice, and research applications.
While considerable progress has been made in the discovery
and development of alcohol biomarkers (Hannuksela et al.,
2007; Litten and Fertig, 2003; Niemela, 2007), efforts to
improve existing biomarkers and identify new biomarkers
continue to be an important National Institute on Alcohol
Abuse and Alcoholism (NIAAA) research priority, as evi-
denced by the 2005 Request for Applications Identification of
Alcohol Biomarker Signatures (RFA-AA-06-002) and Genom-
ic, Proteomic, and Metabolomic Fingerprints as Alcohol Bio-
markers (SBIR ⁄STTR) (RFA-AA-06-001), and the June 2008
conference ‘‘Workshop on Alcohol Biomarkers.’’ This article
reviews current and potential uses of alcohol biomarkers in
applied and research settings with a focus on chemical com-
pounds in the body that are part of either normal biological
processes or a pathogenic process. In addition, we describe
desirable features of biomarkers, summarize research knowl-
edge about traditional, new, and promising biomarkers, and
update knowledge of their clinical utility. The article is
intended to highlight future research opportunities in this
rapidly evolving component of the alcohol research field.
POTENTIAL USES OF ALCOHOL BIOMARKERS IN
APPLIED SETTINGS
Alcohol biomarkers perform several vital functions in med-
icine, public safety, and research (Litten and Fertig, 2003). In
clinical settings, they serve as objective means of identifying
problem drinking and estimating the extent of alcohol-related
tissue damage, thereby enabling clinicians to target treatments
to AUD severity. In addition, alcohol biomarkers can be used
to diagnose and monitor alcohol-related medical conditions
From the Division of Treatment and Recovery Research (RZL),
Office of Science Policy and Communications (AMB), and Office of
the Director (HBM), National Institute on Alcohol Abuse and
Alcoholism, Bethesda, Maryland.
Received for publication August 20, 2009; accepted January 6, 2010.
Reprint requests: Raye Z. Litten, PhD, Associate Director,
Division of Treatment and Recovery Research, National Institute on
Alcohol Abuse and Alcoholism, 5635 Fishers Lane, Room 2041,
Bethesda, MD 20892-9304; Fax: 301-443-8774; E-mail: rlitten@mail.
nih.gov (Rockville, MD 20852-1705 for FedEx)
Copyright 2010 by the Research Society on Alcoholism.
No claim to original U.S. government works
DOI: 10.1111/j.1530-0277.2010.01170.x
Alcoholism: Clinical and Experimental Research Vol. 34, No. 6
June 2010
Alcohol Clin Exp Res, Vol 34, No 6, 2010: pp 955–967 955
(e.g., cardiovascular and liver disease, pancreatitis, HIV ⁄
AIDS, and certain cancers) in which alcohol use has either a
causal or exacerbating role in the disorder or impacts treat-
ment adherence.
Increasingly, alcohol biomarkers also serve as objective
measures of AUD treatment outcome. Whereas the accuracy
of patient self-reports is well supported by research evidence
(Del Boca and Darkes, 2003) and, at present, the most com-
monly used means to ascertain patient relapse, biomarkers
could be used either to verify or, if sufficiently accurate, replace
patient self-reports altogether. Clinicians also use biomarkers
to provide patients with feedback about drinking effects and
to foster motivation to cut back or refrain from drinking. In
settings that also conduct clinical trials, biomarkers may be
used to gauge the effectiveness of experimental interventions.
In such applications, the use of valid, objective markers of
treatment outcome can contribute to both clinical trial effi-
ciency and confidence in outcome reports (Anton et al., 2002).
In individual and public safety applications, biomarkers are
used to monitor abstinence in high-risk individuals (e.g.,
actively drinking pregnant women, persons previously
convicted of alcohol-related offenses) and situations (e.g.,
medical, transportation, other occupations that affect public
well-being) (Litten and Fertig, 2003).
In alcohol research, biomarkers have the potential to serve
as trait markers of AUD phenotypes (i.e., observable
physical and biochemical characteristics of an individual
determined by genetic makeup and environmental factors).
The identification and categorization of alcohol dependence
phenotypes is expected to provide models for the multiple
subtypes of alcohol dependence (Ait-Daoud et al., 2009;
Mayfield and Harris, 2009; Oslin et al., 2003), enabling clini-
cians to identify patients who are likely to respond positively
or negatively to specific treatments, especially medications.
Thus, biomarkers will have a central role in personalized
medicine for AUDS.
DESIRABLE PSYCHOMETRIC CHARACTERISTICS
OF ALCOHOL BIOMARKERS
To perform these functions successfully, alcohol biomar-
kers must possess certain attributes including a high degree of
validity, i.e., the ability to measure accurately the condition of
interest (Litten and Fertig, 2003). Biomarker validity is
commonly determined by its criterion validity,thatis,thecor-
relation between the biomarker and a criterion variable (or
variables) taken as representative of the construct of interest.
These criterion variables may include concurrent measures of
alcohol consumption, indicators of alcohol dependence, or
indicators of organ damage. In addition, certain alcohol bio-
markers may be examined for predictive validity by testing the
manner in which the current biomarker level reflects some
future biological process, such as advanced organ damage or
severity of AUD. Last, useful alcohol biomarkers should
demonstrate ecological validity, wherein they demonstrate
utility in real-world situations. Biomarkers must be easily
implemented in and transported to a variety of settings,
including primary care, medical specialty, and workplace
settings as needed. They should be easily administered, with
samples easily collected and analyzed, and affordable, with
reasonable costs for instruments, other materials, and labor.
Above all, they must be acceptable to the clinicians who use
them and the patients to whom tests are administered.
Finally, a desirable test should display a practicable window
of assessment. It is important to determine how long a
biomarker will remain in the body as a function of the
amount of alcohol consumed after drinking is stopped.
Ideally, a biomarker should persist for at least several days
after cessation of drinking and return to normal values only
after a period of abstinence that affords sufficient time to
screen for the condition.
Test accuracy is another important characteristic of a bio-
marker that reflects the quality and usefulness of the biomar-
ker test. Tests for alcohol biomarkers must accurately
measure the marker of interest. For example, if a milliliter of
blood contains 30 units of a marker, a valid test would iden-
tify proportionally the same or a similar quantity.
The accuracy of biomarker is reflected in its sensitivity (i.e.,
the probability that the biomarker will produce a true positive
result when used on an affected population) and specificity
(i.e., the probability that a test will produce a true negative
result when used in a nonaffected population) (see Table 1).
Reliable cutoff values (i.e., threshold values that define the
presence of a condition) are essential in determining sensitivity
and specificity and depend on the condition being measured
and the setting in which it is measured. For example, a high
cutoff value will make it more difficult to detect the condition
(decreasing the sensitivity) but will reduce the number of false
Table 1. Diagnostic Test Characteristics and Definitions
Condition or disease
+)
Test result + True positive (TP) False positive (FP) All positive
tests = TP+FP
Positive predictive
value = TP ⁄(TP+FP)
)False negative (FN) True negative (TN) All negative
tests = TN+FN
Negative predictive
value = TN ⁄(TN+FN)
All with condition = TP+FN All without condition = FP+TN Prevalence = (TP + FN)/(TP+FN+FP+TN)
Sensitivity = TP ⁄(TP+FN) Specificity = TN ⁄(FP+TN) Accuracy = (TP + TN)/(TP+FN+FP+TN)
956 LITTEN ET AL.
positives (increasing the specificity). In contrast, a low cutoff
value results in a higher sensitivity and a lower specificity. In
settings such as primary care screening for problematic drink-
ing where high sensitivity is desirable and false positives are
acceptable, a lower cutoff value is used. By contrast, in crimi-
nal justice settings, a higher cutoff value may be prudent
because false positives can have devastating effects. The deter-
mination of cutoff values should be validated by rigorous
research studies.
Positive and negative predictive value of a given biomarker
test is another measure of test accuracy. When using alcohol
biomarkers, professionals must take into account the positive
predictive value (PPV) (i.e., the percentage of positive tests in
which the condition occurs) and the negative predictive value
(NPV) (i.e., the percentage of negative tests in which the con-
dition does not occur) for the biomarker used. Predictive val-
ues depend not only on the sensitivity and specificity of a test,
but also on the prevalence of the disorder in the population
tested. A population with low prevalence of the disorder
resultsinanincreaseinfalsepositive tests, whereas a higher
prevalence rate yields more false-negative tests. Table 2 pre-
sents examples of this concept: assuming that a biomarker
has a sensitivity and specificity of 90%, if the prevalence of
heavy drinkers is 10% (as in a primary care screening situa-
tion), the PPV is only 50%, while the NPV value is 99%. On
the other hand, if the prevalence of heavy drinkers is 90% (as
in an addiction clinic), then the PPV is 99%, while the NPV is
only 50%. Thus, the interpretation of tests will vary from one
setting to another depending on the prevalence of the problem
being screened. This concept is not only a concern for screen-
ing alcohol problems, but for all biomarkers used to screen
any medical disorder.
Useful alcohol biomarkers must also demonstrate a high
degree of reliability. Reliability of a biomarker is only of
interest when the biomarker has first demonstrated its valid-
ity. Biomarker reliability refers to the degree to which the bio-
marker is internally consistent and stable in measuring that
whichitisintendedtomeasureovertime.Thereareseveral
general types of test reliability in the psychometrics literature,
but only a few are applicable to biomarker analysis. Perhaps,
the most important of these for biomarkers is test–retest
reliability. Here, the biomarker is evaluated in the same
sample of individuals on two different occasions. This
approach assumes that there is no substantial change in the
underlying condition of interest (e.g., level of alcohol con-
sumption, magnitude of organ damage) between the two
occasions. Because these conditions may vary over time, the
amount of time allowed between measures is critical. Thus,
theshorterthetimegap,thehigher the expected test–retest
reliability; the longer the time gap, the lower the expected
reliability.
UPDATE ON ALCOHOL BIOMARKERS
During the past decade, many alcohol biomarkers have been
studied in humans. Among clinical applications, alcohol
biomarkers have been used primarily to detect alcohol con-
sumption and less frequently to detect tissue damage. This sec-
tion reviews traditional, new, and promising alcohol
biomarkers that have been used in various human studies (see
Table 3).
Traditional Alcohol Biomarkers
The most direct way to ascertain alcohol consumption is
to measure alcohol’s presence in blood, urine, or breath.
Although such measures may be useful under certain condi-
tions (e.g., to test for recent consumption in the workplace),
the relatively short half-life of alcohol prohibits their utility
in most settings (Swift, 2003). To address this problem,
Table 2. Effects of High and Low Prevalence on Diagnostic Test Results
+)
High prevalence
Test result +81 (TP) 1 (FP) All + tests = TP+FP = 82 Positive predictive
value TP ⁄(TP+FP) = 99%
)9 (FN) 9 (TN) All – tests = TN+FN = 18 Negative predictive
value TN ⁄(TN+FN) = 50%
All with condition – 90 All without the condition – 10
Sensitivity 90% Specificity 90%
Condition or disease
+-
Low prevalence
Test result +9 (TP) 9 (FP) All + tests = TP+FP = 18 Positive predictive
value TP ⁄(TP+FP) = 50%
-1 (FN) 81 (TN) All – tests = TN+FN = 82 Negative predictive
value TN ⁄(TN+FN) = 99%
All with condition – 10 All without the condition – 90
Sensitivity 90% Specificity 90%
Sensitivity and specificity of the test are set at 90% for 100 subjects screened. For high prevalence, 90% are assumed to have the condition.
TP, true positive; FP, false positive; TN, true negative; FN, false negative.
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 957
Table 3. Characteristics of Traditional, New, and Promising Alcohol Biomarker Devices
Alcohol
biomarker ⁄
device
Drinking
behavior
targeted
Sample
source
Window of
assessment
a
Primary
indication Sensitivity False positives Population Cost ⁄type of testing
GGT Chronic heavy
drinking
Blood 2 to 3 weeks Screening Moderate
(low to high)
Many sources of
false positives
Adults
(ages 30 to 60)
Low ⁄routine testing
AST and
ALT
Chronic heavy
drinking
Blood 2 to 3 weeks Screening Lower than GGT Many sources of
false positives
Adults
(ages 30 to 60)
Low ⁄routine testing
MCV Chronic heavy
drinking
Blood Up to several
months
Screening Lower than GGT Several conditions of
false positives
Adults
(ages 30 to 60)
Low ⁄routine testing
% CDT Moderate to high
(heavy drinking
for 7 to 10 days)
Blood 2 to 3 weeks Screening ⁄relapse Similar to GGT Few conditions of
false positives
Adults (influenced
by gender)
Moderate to high ⁄
specialized testing
5-HTOL ⁄
5-HIAA
4 drinks Urine 1 day Screening ⁄relapse High Few conditions of
false positives
More research on
patient variability
High ⁄specialized
testing
EtG 1 to 2 drinks Urine Several days Abstinence ⁄relapse Few sources of
false negatives
Few sources of false
positives
More research on
individual variability
Moderate ⁄
specialized testing
Unknown Hair Up to several months Screening ⁄Abstinence Unknown Unknown
EtS 1 to 2 drinks Urine 1 to 2 days Abstinence ⁄Relapse High Few sources of
false positives
More research on
individual variability
Moderate ⁄
specialized testing
FAEE Unknown Hair Up to several
months
Screening ⁄Abstinence Unknown Unknown More research on
individual variability
High ⁄specialized
testing
At least several
drinks
Blood 2 days Abstinence ⁄relapse Unknown Unknown
PEth Heavy drinking
for 5 days
Blood 1 to 2 weeks Screening ⁄relapse High for
screening ⁄
unknown for
relapse
Unknown More research on
individual variability
Moderate ⁄
specialized testing
SCRAM 1 to 5 drinks Transdermal
alcohol vapor
Continuously
records
Abstinence ⁄relapse Moderate False positives
approach zero
More research on
individual variability
High ⁄high-tech
device worn
around ankle
GGT, gamma-glutamyl transpeptidase; AST, aspartate aminotransferase; ALT, alanine aminotransferase; MCV, mean corpuscular volume; EtS, ethyl sulfate; CDT, carbohydrate-
deficient transferrin; EtG, ethyl glucuronide; FAEE, fatty acid ethyl esters; PEth, phosphatidyl ethanol; HIAA, hydroxyindoleacetic; HTOL, hydroxytryptophol; SCRAM, secure continuous
remote alcohol monitor.
a
Window of assessment: time biomarker remains measureable in body after drinking is stopped.
Primary indications: Abstinence – detecting any alcohol intake; Relapse – monitoring drinking, especially heavy drinking during treatment; Screening – detecting drinking, especially
high-risk heavy drinking.
Sources: Substance Abuse and Mental Health Services Administration, 2006; Kissack et al., 2008; Conigrave et al., 1995, 2003; Litten et al., 1995; Niemela, 2007; Javors et al., 1997;
Fleming et al., 2004; Jeppsson et al., 2007.
958 LITTEN ET AL.
researchers have identified biomarkers of alcohol consump-
tion with longer windows of assessment than direct alcohol
measures of breath and body fluids. These so-called tradi-
tional biomarkers measure alcohol consumption indirectly by
detecting tissue damage or other physiological reactions to
heavy drinking over time.
The most common traditional marker is gamma glutamyl
transferase (GGT). GGT is found in the cell membranes of
multiple tissues, including the liver, kidney, pancreas, spleen,
heart, and prostate (Javors et al., 1997). Chronic heavy drink-
ing elevates GGT levels in the blood because of increased
GGT synthesis and ⁄or GGT leakage from liver cells that have
been damaged or destroyed by alcohol. GGT has a long win-
dow of assessment, with values that remain elevated for 2 to
3 weeks after heavy-drinking cessation (Javors et al., 1997).
GGT usually is not elevated in heavy-drinking adolescents
and young adults. Although the reason for this is still
unknown, it may be related to the need for individuals to
reach a certain physiological age before noticeable GGT
changes occur or to drink heavily for many years to elevate
GGT (Conigrave et al., 2003). Overall, GGT sensitivity for
screening heavy drinking is moderate, ranging from low to
high values depending on the population and setting where it
is used (Conigrave et al., 1995, 2003; Litten et al., 1995).
Although GGT is measured routinely at reasonable cost in
most clinical laboratories, its clinical utility is limited by a
high rate of false positives because of non-alcohol-related liver
diseases, obesity, diabetes, smoking, and such medications as
anticonvulsants, anticoagulants, and barbiturates (Conigrave
et al., 2003).
Aspartate aminotransferase (AST) and alanine aminotrans-
ferase (ALT), liver enzymes often measured in routine
screening for liver damage, also are used as biomarkers of
heavy alcohol consumption. AST and ALT have characteris-
tics similar to those of GGT but with lower sensitivity
(Conigrave et al., 2003). Consequently, AST and ALT are
used less frequently than GGT to screen for heavy alcohol
consumption.
Another traditional alcohol biomarker is mean corpuscular
volume (MCV), a measurement of the size of red blood
cells. Chronic heavy drinking increases the size of red blood
cells. As with the liver enzymes, MCV is routinely measured
in most laboratories and works best to detect heavy drinking
in adults aged 30 to 60. Overall, for reasons that are still
unclear, MCV appears to have lower sensitivity in males and
higher sensitivity in females than GGT and is more specific
than GGT (Conigrave et al., 2003). Because red blood cells
exhibit a long half-life (e.g., 13 to 27 weeks), MCV values can
remain elevated up to several months after cessation of drink-
ing (Niemela, 2007). Numerous sources of false positives
includefolateandB
12
deficiencies, nonalcoholic liver diseases,
hemolysis, bleeding disorders, hypothyroidism, medications
that produce marrow toxicity, and bone marrow disorders
(Conigrave et al., 2003; Niemela, 2007).
Most recent among traditional biomarkers is carbohydrate-
deficient transferrin (CDT) (Anton, 2001; Niemela, 2007).
Transferrin, a glycoprotein synthesized and secreted by the
liver, transports iron throughout the body. Heavy drinking
(50 to 80 g alcohol ⁄day) for 7 to 10 days decreases the
carbohydrate content of transferrin, including sialic acid,
galactose, and N-acetylglucosamine, by a mechanism that
remains unclear but most likely involves decreased activity
of glycosyltransferases and increased activity of sialidases,
enzymes that add and remove carbohydrate groups, respec-
tively (Litten and Allen, 1998; Stibler, 1991). After cessation
of drinking, 2 to 3 weeks are required for serum CDT val-
ues to return to normal. CDT has been tested as a screening
marker of alcohol intake and recently has been demon-
strated to detect relapse (Anton et al., 2002). Whereas CDT
sensitivity is similar to that of GGT (Litten et al., 1995), its
real strength is its high specificity. The few identified sources
of false positives include rare genetic transferrin variants,
primary biliary cirrhosis, chronic end-stage liver disease, and
hepatocarcinoma (Fleming et al., 2004; Javors and Johnson,
2003). In addition, CDT is influenced by smoking, body
weight, and female gender (Fleming et al., 2004). During
the past several years, methods to measure CDT have
improved, resulting in a further increase in sensitivity and
specificity (Bergstrom and Helander, 2008). These improve-
ments have also minimized the effects of smoking, body
weight, and gender on CDT values. As a result of these
improvements, the International Federation of Clinical
Chemistry and Laboratory Medicine Working Group made
the following recommendations (Jeppsson et al., 2007): (i) to
normalize the variability of the transferrin content, espe-
cially in women, CDT should be expressed as relative
amount (CDT ⁄total transferrin); (ii) CDT is a complex
compound consisting of several glycoforms, among which
the disialotransferrin glycoform displays the best correlation
with alcohol intake (Bergstrom and Helander, 2008; Jepps-
son et al., 2007). Therefore, the disialotransferrin glycoform
is recommended as the primary target for CDT measure-
ment.Thisglycoformcanbemeasuredwithhighperfor-
mance liquid chromatography (HPLC), which has been
suggestedtobethereferencemethodfor%CDT
measurement (Jeppsson et al., 2007). Finally, % CDT has
been approved by the FDA in the United States as a mar-
ker of heavy alcohol consumption.
New Biomarkers
A serotonin metabolite that is elevated during alcohol con-
sumption, 5-hydroxytryptophol (5-HTOL) is elevated at
approximately 50+ g of alcohol during a drinking occasion
(Beck and Helander, 2003). 5-HTOL displays high sensitivity
and specificity and appears uninfluenced by age, gender,
liver diseases, or medications other than disulfiram (Beck
and Helander, 2003). 5-HTOL is expressed as a ratio to
5-hydroxyindole-3-acetic acid (5-HIAA), another metabolite
of serotonin, an attribute that reduces false positives,
especially from foods rich in serotonin such as bananas,
pineapples, and tomatoes (Borg et al., 1992). This ratio
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 959
measurement is important because increases in serotonin
from these foods elevate the serotonin metabolites 5-HTOL
and 5-HIAA equally. Limitations of 5-HTOL include 1-day
window of assessment (Beck and Helander, 2003; Hoiseth
et al., 2008). In addition, cumbersome methodologies, such as
gas chromatography–mass spectroscopy (GS–MS), liquid
chromatograph-mass spectroscopy (LCMS), or HPLC, must
be employed to measure the metabolites (Beck and Helander,
2003). Recently, the glucuronidated form of 5-HTOL,
glucuronidated 5-hydroxytryptophol (GTOL), has been used
as a biomarker (Beck et al., 2007). GTOL is the predominant
metabolite of 5-HTOL in the urine and may be measured
using simpler methods such as the enzyme-linked immuno-
sorbent assay (ELISA) (Dierkes et al., 2007).
Recent research has focused onthemoredirectmeasure-
ment of alcohol metabolites, an important advance in alcohol
biomarkers. The most advanced such biomarker is ethyl
glucuronide (EtG), formed from a conjugation reaction of
alcohol with glucuronic acid in the presence of the enzyme
uridine diphosphate glucuronyl transferase (UGT) (Wurst
et al., 2003). This metabolite is measurable in blood, hair, and
urine, most commonly urine. There are 16 functional human
genes of UGT, raising the possibility that the formation rate
of EtG may vary by genotype (Miners et al., 2002; Wurst
et al., 2003).
At present, EtG usually is measured by LC–MS, although
less costly methods are in development (Bottcher et al.,
2008). Detection time for EtG in the urine is longer than in
the blood (14 to 24 hours) (Borucki et al., 2007; Hoiseth
et al., 2007), so that urine measures have potential to both
screen for problematic drinking and monitor relapse in treat-
ment and recovery programs (Kissack et al., 2008; Wurst
et al., 2004). The detection time of EtG in the urine varies
ranging from 20 hours up to 80 to 102 hours (Kissack et al.,
2008; Wojcik and Hawthorne, 2007; Wurst et al., 2004),
depending on the amount of alcohol consumed and individ-
ual variability. Methodology also is in development to
measure EtG in the hair. So far, EtG can be detected in the
hair for several months after drinking is stopped. This has
implications for forensic purposes as well as for monitoring
abstinence among individuals convicted of driving while
intoxicated and women who are at risk for drinking during
pregnancy (Bendroth et al., 2008; Pragst and Yegles, 2008;
Wurst et al., 2008a,b).
EtG appears to be highly sensitive and specific to alcohol
intake. In fact, it is so sensitivethatitcanmeasureincidental
alcohol exposure including alcohol in foods (e.g., cooking
wines, flavoring extracts), over-the-counter cold medications,
mouthwash, hand sanitizer gel, and other hygiene products
(Costantino et al., 2006; Rosano and Lin, 2008). In addition,
some urine samples may contain yeast that can convert urine
glucose to alcohol and subsequently EtG if stored at room
temperature for more than 12 hours (Kissack et al., 2008;
Saady et al., 1993); this may lead to be problems when EtG is
used with to test persons who are diabetic and have high levels
of glucose in the urine. False negatives also can arise from
E. coli hydrolysis of EtG in urinary tract infections (Helander
and Dahl, 2005) or from ingestion of chloral hydrate medica-
tions (Arndt et al., 2009). Another source of false negatives is
urine dilution. To counter this, it is recommended that either
urinary creatinine be measured with a cutoff of 25 mg ⁄dl to
indicate dilution (Goll et al., 2002) or EtG be expressed in
ratio to creatinine (Dahl et al., 2002).
Finally, EtG varies among individuals (Sarkola et al.,
2003). Factors that may underlie this variability include gen-
der, age, ethnic group, medical or psychiatric comorbidity,
genetic polymorphism of UGT, and possible others. Charac-
terizing this variation will be important in establishing reliable
cutoff values to determine sensitivity and specificity in differ-
ent populations and settings.
Promising Biomarkers
In addition to EtG, researchers are exploring several other
promising alcohol metabolites, including phosphatidylethanol
(PEth), ethyl sulfate (EtS), fatty acid ethyl esters (FAEEs),
andacetaldehydeadducts.PEth,whichisformedbythe
action of phospholipase D, is a group of phospholipids with a
common nonpolar phosphoethanol head group (Helander
and Zheng, 2009). PEth is detected in the blood after consum-
ing approximately 1000 g of alcohol, usually over about a
2-week period (Varga et al., 1998). PEth has a 1- to 2-week
window of assessment, depending on the level of drinking
(Stewart et al., 2009; Varga et al., 2000). In addition, its
sensitivity and specificity appear better than for traditional
biomarkers (Aradottir et al., 2006; Hartmann et al., 2006).
Because PEth appears not to be influenced by liver diseases
(Stewart et al., 2009), it may be promising in detecting and
monitoring heavy drinkers with hepatic pathology because
most traditional biomarkers are elevated by non-alcohol-
induced liver disease. PEth can be measured using high-
performance LC and evaporative light scattering and
electrospray MS technologies (Gunnarsson et al., 1998).
Nonetheless, new methods are being developed to measure
PEth. PEth has recently been measured by combining LC–
ESI-MS ⁄MS which is more sensitive than previous method of
combining HPLC with evaporative light scattering detection
(ELSD), allowing a >2 weeks of window assessment (Gnann
et al., 2009). Helander and Zheng (2009) recently identified
PEth species in the blood by using an electrospray ionization
(ESI) LC–MS approach. More researchisrequiredtoclini-
cally characterize PEth, in particular to determine its variabil-
ity among individuals (Varga et al., 2000). EtS is another
minor alcohol metabolite formed from the sulfate conjugation
of alcohol from 3¢-phosphoadenosine 5¢-phosphosulfate alco-
hol, a reaction catalyzed by sulfotransferase (Helander and
Beck, 2005). EtS is detectable in urine around 30 hours after
the last drink, a short window of assessment. However,
because EtS and EtG appear to be highly correlated and LC–
MS can simultaneously measure each (Helander and Beck,
2005; Wurst et al., 2006), measuring both may serve to
corroborate results. Junghanns and colleagues (2009) recently
960 LITTEN ET AL.
reported that all patients who were positive for urinary EtG
were also positive for urinary EtS.
FAEEs, such as ethyl palmitate, ethyl oleate, and ethyl
stearate, are nonoxidative metabolites of alcohol measurable
by GC–MS (Pragst et al., 2001). In very heavy drinkers,
FAEEs can be detected up to 99 hours in the blood
(Borucki et al., 2007). FAEEs also are present in meconium
fluid, enabling detection of alcohol use during pregnancy
(Bearer et al., 2003; Moore et al., 2003; Swift, 2003). A limita-
tion of this practice is the inability to accurately determine
when and how much alcohol consumption occurred. Recent
studies (Pragst and Yegles, 2008; Pragst et al., 2001; Wurst
et al., 2008a) also have measured FAEEs in the hair. The
advantage of this procedure is that it can detect alcohol in
the hair up to 2 months after abstinence (Pragst et al., 2001),
useful in monitoring drinking in pregnant women (Pragst and
Yegles, 2008; Wurst et al., 2008a) and offenders convicted of
driving under the influence (DUI) (Wurst et al., 2008b). EtG
in hair often is measured simultaneously with FAEEs because
GC–MS is employed for the measurement of both (Pragst
and Yegles, 2008). Further research is needed to determine
the relationship between drinking and alcohol metabolite
levels in the hair, rate of change in levels in the hair following
abstinence, distribution of alcohol metabolites along hair
shaft, and sources of false positives and false negatives, such
as use of hair lotions, and racial or ethnic variations in FAEE
deposition.
Acetaldehyde adducts are formed by the reaction of
acetaldehyde, the major metabolite of alcohol, with proteins.
Acetaldehyde forms both stable and unstable adducts with
proteins. The stable ones remain in the blood for weeks, ren-
dering acetaldehyde adducts a promising marker of alcohol
consumption (Litten and Allen, 1998). Sensitivity and specific-
ity appear to be as least as accurate as with traditional bio-
markers (Niemela, 2007; Swift, 2003). Other than drinking,
additional sources of acetaldehyde, especially from smoking
(Salaspuro, 2007), must be identified and characterized.
Although, to this time, researchers have been unable to
develop a routine method to accurately measure these
adducts, new approaches are still being tested, including the
use of the ELISA assay to measure specific immunoglobulin
A (IgAs) against acetaldehyde–protein adducts (Hietala et al.,
2006a).
Finally, other biomarkers have been explored, including
sialic acids, sialic acid index of plasma apolipoprotein J, beta-
hexosaminidase, platelet MAO B activity, high density lipo-
protein (HDL), and dolichol, but these appear no better than
the traditional biomarkers (Hannuksela et al., 2007). The
measurement of platelet MAO B protein levels rather than
activity is also being investigated and early results appear
promising (Tabakoff et al., 2009).
Biomarker Combinations
Because all biomarkers exhibit some limitations, one
approach to improve accuracy is to use them in combination.
The most common combination studied, so far, is CDT mea-
sured in conjunction with GGT (CDT + GGT) (Litten
et al., 1995). Employing the rule that subjects are labeled posi-
tive if either test exceeds its cutoff values, CDT + GGT con-
sistently and often impressively increases sensitivity, with little
loss of specificity, compared with using either biomarker
alone (Litten et al., 1995). This finding was consistent in
screening for problematic drinking in multiple populations,
including alcoholics with and without liver disease, heavy
drinkers, different ethnic groups, males and females, and col-
lege students (Litten et al., 1995). Some investigators have
used a mathematically formulated equation of CDT + GGT
and reported higher sensitivities for the combined markers
(Anttila et al., 2003; Hietala et al., 2006b; Sillanaukee and
Olsson, 2001). Lastly, Anton and colleagues (2002) found that
CDT + GGT performed better than either biomarker alone
in detecting relapse to heavy drinking.
Recent studies have used the traditional markers CDT and
GGT and the newer markers EtG, EtS, PEth, and FAEEs in
combination with the self-report Alcohol Use Disorders
Identification Test (AUDIT) questionnaire. Using alcohol
metabolite biomarkers with CDT or GGT, in combination
with the AUDIT, improved detection of drinking in emer-
gency rooms and workplace settings and with pregnant
women (Hermansson et al., 2000; Kip et al., 2008; Neumann
et al., 2008; Wurst et al., 2008a). Finally, several investigators
have developed a program that incorporates the combination
of many clinical tests (Harasymiw and Bean, 2001; Korzec
et al., 2005). For example, the Bayesian Alcoholism Test
(BAT) consists of 8 to 15 clinical and biochemical indicators
including GGT, AST, and CDT. Korzec and colleagues
(2009) recently reported that BAT was more sensitive than
CDT, GGT, and AST alone in distinguishing harmful alcohol
users from controls. Another example is the Early Detection
of Alcohol Consumption (EDAC), a program that uses a lin-
ear discriminant function analysis on 10 to 36 routine labora-
tory tests (Harasymiw and Bean, 2001, 2007). Harasymiw
and Bean (2007) found that EDAC test performed better than
GGT in detecting heavy drinking in subjects from treatment
and community settings.
Alcohol Sensor Devices
From biochemical markers of alcohol consumption (except
for direct measurement of blood alcohol levels), it is impossi-
ble to determine precisely when drinking occurred, the exact
amount of alcohol consumed, how many drinking episodes
were required to produce the value rendered, or whether that
value has increased or decreased from the last drink. Nor do
biochemical compounds correlate accurately with amount of
alcohol intake. For these reasons, alcohol sensor devices pro-
vide important alternative. With alcohol sensor devices, one
could determine the time that drinking occurred; such devices
also have the potential to measure accurately the amount of
alcohol intake during each drinking episode and provide an
estimate of blood alcohol concentration.
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 961
Several new noninvasive alcohol devices are now in devel-
opment, the most advanced of which is the Secure Continu-
ous Remote Alcohol Monitor (SCRAM). The SCRAM is a
bracelet worn around the ankle that measures alcohol vapor
electrochemically in a semi-quantitative manner (Sakai et al.,
2006). A unique property of the SCRAM is that it measures
alcohol intake over a 24-hour period, enabling one to deter-
mine the approximate time that drinking occurred. So far, its
specificity appears to be excellent and its sensitivity moderate
for the detection of drinking (Wojcik, 2008). However, a 2007
National Highway Traffic Safety Administration report
identified several problems. For example, water accumulates
in the device, resulting in reduced sensitivity and lower signals
in females than males. In addition, the SCRAM should be
more precise in measuring alcohol intake, less expensive, and
more comfortable to wear. Further research is needed to
determine interindividual variability and to test SCRAM
utility in various settings.
Another device in development is the Giner WrisTAS. In
contrast to the SCRAM, this device is worn around the wrist.
Like the SCRAM, the Giner WrisTAS measures alcohol
vapor electrochemically and measures drinking continuous
with time (Swift, 2003). It is still at the research stage and is
not ready for commercial use (National Highway Traffic
Safety Administration, 2007).
BIOMARKERS IN SPECIFIC SETTINGS
Although there is, at present, no ideal biomarker, biomar-
kers, nevertheless, are in use in multiple applied settings. This
section reviews some of these applications and reports
current knowledge of and strategies for using alcohol
biomarkers (Table 4). In all such applications, it is critical
that biomarkers be validated by adequately controlled
research studies.
Recovery Programs
Many U.S. recovery programs are expressly geared to treat
health care professionals, while others monitor airline and
maritime pilots, air traffic controllers, and railway engineers.
For example, physician health programs currently monitor
abstinence in over 9,000 physicians who have been identified
as having alcohol problems (Skipper et al., 2004). A recent
survey of physician health programs in 46 states found that
many use traditional and new alcohol biomarkers to monitor
abstinence (Jansen et al., 2004). Determining whether pro-
gram participants have relapsed to drinking is a special con-
cern for such programs, given the public service performed by
these professionals. In such settings, biochemical tests that
monitor abstinence must be sensitive to alcohol consumption
while keeping the number of false positives at a minimum.
This is important because positive tests can lead to severe con-
sequences. Whereas, at present, many of these programs use
EtG, its sensitivity to incidental alcohol exposure remains
problematic. Accordingly, a 2006 Substance Abuse and
Mental Health Services Administration (2006) included a
warning about using EtG test in recovery programs. More
research is needed to accurately determine reliable cutoff
values that account for individual variability and incidental
alcohol exposure.
Possible future research for this setting includes exploring
the combination of biomarkers with alcohol sensor devices,
such as the SCRAM. EtG and EtS can detect one to two
drinks, whereas the SCRAM appears sensitive to several
drinks with few false positives. In addition, the SCRAM can
record the approximate time that drinking occurs.
Criminal Justice Settings
Forty percent of violent crimes committed in the United
States involve the consumption of alcohol (U.S. Department
of Justice, 1998). Moreover, 4 in 10 fatal motor vehicle acci-
dents involve alcohol, while approximately 1.5 million DUIs
occur each year (U.S. Department of Justice, 1998; Wojcik,
2008). Total alcohol abstinence is a common pretrial, proba-
tion, or parole order for many of these offenders. Because
alcohol is rapidly eliminated from the body, monitoring
required abstinence is a challenge to multiple criminal justice
settings.
Table 4. Use of Biomarkers in Applied Settings
Setting Primary indication Type of alcohol biomarker needed Possible alcohol biomarkers
Recovery program Abstinence Sensitive to low amounts of alcohol;
no false positives
EtG, EtS, SCRAM
Criminal justice Abstinence Sensitive to low amounts of alcohol;
no false positives
EtG, EtS, SCRAM
Primary care Screening ⁄relapse Sensitive to heavy drinking CDT+GGT
Alcohol ⁄drug specialty treatment Relapse Sensitive to drinking–especially
heavy drinking
CDT+GGT EtG, EtS, SCRAM
Other specialty treatment Screening Sensitive to any drinking–especially
to heavy drinking
CDT+GGT EtG, EtS
Workplace: employee assistance programs Abstinence Sensitive to low levels of alcohol EtG, EtS SCRAM
Workplace: health and safety screening Screening Sensitive to drinking–especially
heavy drinking
EtG, EtS CDT, GGT
GGT, gamma-glutamyl transpeptidase; EtG, ethyl glucuronide; EtS, ethyl sulfate; CDT, carbohydrate-deficient transferring; SCRAM, secure
continuous remote alcohol monitor.
962 LITTEN ET AL.
During recent years, more than 1600 courts in 45 states
have monitored more than 65,000 offenders using the
SCRAM (Wojcik, 2008). Although the SCRAM is not suf-
ficiently sensitive to measure total abstinence, early results
suggest that it is sensitive in detecting five to six drinks
(Wojcik, 2008). In addition, Wurst and colleagues (2008b)
suggestthatmeasuringEtG and FAEE in hair may be a
reasonable approach to monitor drinking in this population.
Although determining alcohol metabolites in hair samples
will not reveal amount or duration of drinking, it does
reveal whether any drinking occurred over the past few
months. Similar to the recovery programs, alcohol biomar-
kers in criminal justice applications should be sensitive
to alcohol intake with the number of false positives
approaching zero.
Primary Care Settings
Although only 10 to 13% of patients with AUDs seek alco-
hol ⁄drug specialty treatment (McLellan, 2007; Willenbring,
2009), it is likely that many more are seen by primary care
physicians for health problems related to drinking. Most
primary care clinicians, however, do not screen for alcohol
problems (Dawson et al., 2005; Willenbring, 2009). However,
primary care physicians often conduct a battery of clinical
tests, including tests of the liver enzymes GGT, AST, and
ALT, to evaluate liver function. Elevation of these enzymes
could alert physicians to possible drinking problems. As more
effective medications receive FDA approval, physicians are
more likely to screen and more confidently offer treatment to
this population. To that purpose, traditional biomarkers,
such as GGT and CDT alone and in combination, may be
useful tools to screen for patients who drink heavily over a
prolonged period, although these biomarkers without self-
reports may not be very sensitive in screening for alcoholism
(Conigrave et al., 1995). Encouragingly, in recent studies, Dil-
lie and colleagues (2005) and Kapoor and colleagues (2009)
found that adding CDT screening to patient self-report in pri-
mary care settings results in significant savings in medical and
legal costs. Alcohol biomarkers used in such settings should
be sensitive to heavy at-risk drinking, defined for men as 5 or
more and for women as 4 or more drinks per occasion
(National Institute on Alcohol Abuse and Alcoholism, 2007).
For example, Anton and Youngblood (2006) found that
CDT could detect heavy drinking that occurred 4 to 5 days a
week. The addition of a biomarker not only would confirm
the self-report but also would provide results from an objec-
tive biochemical test to help physicians to motivate patients
to either stop drinking or cut backtolow-risklevels(Miller
et al., 2004).
Alcohol ⁄Drug Specialty Treatment Settings
Although most alcohol ⁄drug specialty facilities do not, at
present, use biomarkers for screening, some existing biomar-
kers could be used for that purpose. For example, research-
ers (Anton et al., 2002) reported that the combination of
CDT and GGT was successful inmonitoringrelapseto
heavy drinking in inpatient alcoholics. Biomarkers, such as
EtG and EtS, also have the potential to monitor relapse to
any drinking while patients are in treatment. Wurst and
colleagues (2008c) reported that EtG measurement in hair
and urine of methadone maintenance patients detected
alcohol use in those who scored negatively on the AUDIT.
As in other settings, the combined use of ethanol metabo-
lites and traditional biomarkers with patient self-reports
might be most efficient in monitoring drinking in this popu-
lation.
Other Specialty Treatment Settings
Because alcohol affects every system in the body, many
problematic drinkers present themselves in other specialty
treatment settings with physical symptoms including liver,
cardiovascular, gastroenterological, and sleep disorders (Li,
2007). Reports (Kip et al., 2008; Neumann et al., 2008)
suggest that the alcohol metabolites EtG and PEth are useful
in emergency medicine to detect drinking in injured and
clinically nonintoxicated patients, as well as in those with tho-
racic or gastrointestinal complaints. In contrast, Neumann
and colleagues (2009) found that adding the traditional bio-
markers % CDT, MCV, and GGT to the AUDIT was not
useful for screening trauma patients with alcohol problems.
Fleming and colleagues (2009) reported that measuring blood
alcohol levels in patients admitted for trauma was helpful in
identifying subjects who denied excessive drinking as well as
predicting the development of alcohol withdrawal and other
adverse health events. The usefulness of CDT in these patients
was, however, unclear and required larger samples. In another
medical setting, Wurst and colleagues (2008a) demonstrated
that, in a second-trimester ultrasound screening at a hospital
setting, pregnant women were most likely to be detected for
drinking if a combination of alcohol metabolites (EtG,
FAEE, EtS) and the AUDIT were employed. Recently,
Pichini and colleagues (2009) found, for the first time, EtG
and EtS in the meconium. This raises the possibility of mea-
suring the combination of EtG and EtS along with FAEEs,
which has already been measured in the meconium to moni-
tordrinkinginpregnantwomenwhoareatahighriskfor
drinking. As in primary care settings, it is important that
biomarkers used in other specialty settings identify patients
engaged in high-risk drinking. A prime example is patients
with severe liver disease who may be awaiting transplantation
or other treatments where it is crucial to differentiate recent
alcohol use from underlying liver pathology.
The Workplace
Monitoring abstinence in workers who are problematic
drinkers is essential in high-risk work settings, especially in
settings that affect public safety such as the transportation
sector. For this and less critical workplace purposes,
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 963
combining biomarkers with self-reports can be a practical
monitoring method. For example, Hermansson and col-
leagues (2000, 2002) found that the CDT and the AUDIT
combined are complementary in identifying risky drinking for
employees in the transportation sector. Ideally, in most work-
place settings, biomarkers should be sensitive to low levels of
alcohol consumption and demonstrate high specificity.
RESEARCH DIRECTIONS
Although progress has been made in alcohol biomarker dis-
covery and development, a need remains for relatively inex-
pensive tests that more accurately measure alcohol intake. All
biomarkers developed for clinical use must be fully character-
ized, with strengths and limitations clearly delineated, so that
they can be optimally used in applied settings. This includes
identifying factors that can affect a biomarker response to
alcohol, including severity of problematic drinking, pattern
and amount of drinking, genetic differences, race, gender,
age, body weight, and physical diseases (Fleming et al., 2004).
In addition, reliable cut off values must be established that
reflect optimal sensitivity and specificity as it relates to differ-
ent populations and settings. Because, at present, no biomar-
ker is ideal, it is essential to explore and characterize
combinations of biochemical markers, alcohol devices, and
self-reports in research studies.
Aided by recent advances in high-throughput technologies
for genomics, proteomics, and metabolomics, opportunities
abound to discover novel biomarkers (Hodgkinson et al.,
2008; Mayfield and Harris, 2009). To identify individuals who
are sensitive to alcohol intake, new bioinformatic approaches
and computational tools should be developed that will inte-
grate thousands of RNAs, proteins, and metabolites. This is
necessary to be able to identify those that are sensitive to alco-
hol intake and alcohol-induced tissue damage. For the devel-
opment of biomarkers to detect alcohol-induced issue
damage, it is essential to define the degree of tissue damage
being assessed and also its specificity to alcohol. Finally, it is
also important to develop appropriate in vitro and animal
models to detect alcohol biomarker signatures of acute and
chronic alcohol consumption and alcohol-induced tissue
damage.
Research also should continue to explore new technologies,
including noninvasive imaging techniques and alcohol biosen-
sors, the latter with a focus on developing more comfortable
devices that also are less expensive and more quantitative in
measuring alcohol intake.
FINAL COMMENTS
Although more research is needed for validation, alcohol
biomarkers already serve several vital functions in the preven-
tion, screening, and treatment of AUDs, primarily in recovery
and criminal justice settings where monitoring abstinence is
the primary goal. In various treatment settings, including pri-
mary care settings and trauma services, biomarkers are either
not being used or being used only sparingly. This can be
expected to change as more effective treatments, especially
medications, emerge. However, before any biomarkers are
used in applied settings, it is essential that they are character-
ized rigorously in research studies to better gauge their
strengths and weaknesses.
Overall, the future of biomarkersisbright.Newhigh-
throughput technologies increase the possibility of discovering
biomarker panels or signatures with the potential to be more
sensitive and specific. Biomarker signatures, composed of
multiple parameters, are sought to monitor either alcohol
consumption or alcohol-induced organ damage. Also,
advances in technology have been made in developing new
alcohol sensors that have the potential not only to measure
quantitatively the amount of drinking but also to determine
when drinking occurred. Development of more accurate bio-
markers will allow clinicians to better identify and monitor
individuals who suffer from problematic drinking.
ACKNOWLEDGMENTS
The authors appreciate the review of the manuscript by
Drs. Kathy Jung and Daniel Falk and the preparation of the
tables by Megan Ryan.
REFERENCES
Ait-Daoud N, Roache JD, Dwes MA, Liu L, Wang XQ, Javors MA,
Seneviratne C, Johnson BA (2009) Serotonin transporter genotype predict
craving in alcoholism? Alcohol Clin Exp Res 33:1–7.
Anton RF (2001) Carbohydrate-deficient transferring for detection and moni-
toring of sustained heavy drinking: what have we leaned? Where do we go
from here? Alcohol 25:185–188.
Anton RF, Lieber C, Tabakoff B (2002) Carbohydrate-deficient transferring
and gamma-glutamyltransferase for the detection and monitoring of alcohol
use: results from a multisite study. Alcohol Clin Exp Res 26:1215–1222.
Anton RF, Youngblood M (2006) Factors affecting %CDT status at entry
into a multisite clinical treatment trial: experience from the COMBINE
study. Alcohol Clin Exp Res 30:1878–1883.
Anttila P, Jarvi K, Latvala J, Blake JE, Niemela 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.
Aradottir S, Asanovska G, Gjerss S, Hansson AP, Alling C (2006) Phosphati-
dylethanol (PEth) concentrations in blood are correlated to reported alcohol
intake in alcohol-dependent patients. Alcohol Alcohol 41:431–437.
Arndt T, Gierten B, Gussregen B, Werle A, Gruner J (2009) False-positive
ethyl glucuronide immunoassay screening associated with chloral hydrate
medication as confirmed by LC-MS ⁄MS and self-medication. Forensic Sci
Int 184:e27–e29.
Bearer C, Jacobson JL, Jacobson SW, Barr D, Croxford J, Molteno CD,
Viljoen DL, Marais AS, Chiodo LM, Cwik A (2003) Validation of a new
biomarker of fetal exposure to alcohol. J Pediatr 143:463–469.
Beck O, Helander A (2003) 5-Hydroxytryptophol as a marker for recent alco-
hol intake. Addiction 98(Suppl 2):63–72.
Beck O, Stephanson N, Bottcher M, Dahmen N, Fehr C, Helander A (2007)
Biomarkers to disclose recent intake of alcohol: potential of 5-hydroxytry-
ptophol glucuronide testing using new direct UPLC-tandem MS and
ELISA methods. Alcohol Alcohol 42:321–325.
Bendroth P, Kronstrand R, Helander A, Greby J, Stephanson N, Krantz P
(2008) Comparison of ethyl glucuronide in hair with phosphatidylethanol in
964 LITTEN ET AL.
whole blood as post-mortem markers of alcohol abuse. Forensic Sci Int
176:76–81.
Bergstrom JP, Helander A (2008) Clinical characteristics of carbohydrate-
deficient transferring (% disialotransferrin) measured by HPLC: sensitivity,
specificity, gender effects, and relationship with other alcohol biomarkers.
Alcohol Alcohol 43:436–441.
Borg S, Beck O, Helander A, Voltaire A, Stibler H (1992) Carbohydrate-
deficient transferring and 5-hydroxytryptophol, in Measuring Alcohol
Consumption (Litten RZ, Allen JP eds), pp 149–159. Humana Press,
Totowa, NJ.
BoruckiK,DierkesJ,WartbergJ,WestphalS,GenzA,LuleyC(2007)In
heavy drinkers, fatty acid ethyl esters remain elevated for up to 99 hours.
Alcohol Clin Exp Res 31:423–427.
Bottcher M, Beck O, Helander A (2008) Evaluation of a new immunoassay
for urinary ethyl glucuronide testing. Alcohol Alcohol 43:46–48.
Conigrave KM, Davies P, Haber P, Whitfield JB (2003) Traditional markers
of excessive alcohol use. Addiction 98 (Suppl 2):31–43.
Conigrave KM, Saunders JB, Whitfield JB (1995) Diagnostic tests for alcohol
consumption. Alcohol Alcohol 30:13–26.
Costantino A, DiGregorio EJ, Korn W, Spayd S, Rieders F (2006) The effect
of the use of mouthwash on ethylglucuronide concentrations in urine.
J Anal Toxicol 30:659–662.
Dahl H, Stephanson N, Beck O, Helander A (2002) Comparison of urinary
excretion characteristics of ethanol and ethyl glucuronide. J Anal Tox
26:201–204.
Dawson DA, Grant BF, Stinson FS, Chou PS, Huang B, Ruan WJ (2005)
Recovery from DSM-IV alcohol dependence: U.S. 2001-2002. Addiction
100:281–292.
Del Boca FK, Darkes J (2003) The validity of self-reports of alcohol consump-
tion: state of the science and challenges for research. Addiction 98 (Suppl
2):1–12.
Dierkes J, Wolfersdorf M, Borucki K, Weinmann W, Wiesbeck G, Beck O,
Borg S, Wurst FM (2007) Determination of glucuronidated 5-hydroxytry-
ptophol (GTOL), a marker of recent alcohol intake, by ELISA technique.
Clin Biochem 40:128–131.
Dillie KS, Mundt M, French MT, Fleming MF (2005) Cost-benefit analysis of
a new alcohol biomarker, carbohydrate deficient transferring, in a chronic
illness primary care sample. Alcohol Clin Exp Res 29:2008–2014.
Fleming MF, Anton RF, Spies CD (2004) A review of genetic, biological,
pharmacological, and clinical factors that affect carbohydrate-deficient
transferring levels. Alcohol Clin Exp Res 28:1347–1355.
Fleming M, Bhamb b, Schurr M, Mundt M, Williams A (2009) Alcohol bio-
markers in patients admitted for trauma. Alcohol Clin Exp Res 33:1777–
1781.
Fuller RK, Hiller-Sturmhofel S (1999) Alcoholism treatment in the United
States. Alcohol Res Health 23:69–77.
Gnann H, Weinmann W, Engelmann C, Wurst FM, Skopp G, Winkler M,
Thierauf A, Auwarter V, Dresen S, Bouzas NF (2009) Selective detection of
phosphatidylethanol homologues in blood as biomarkers for alcohol con-
sumption by LC-ESI-MS ⁄MS. J Mass Spectrom 44:1293–1299.
Goll M, Schmitt G, Ganbmann B, Aderjan RE (2002) Excretion profiles of
ethyl glucuronide in human urine after internal dilution. J Anal Toxicol
26:262–266.
Gunnarsson T, Karlsson A, Hansson P, Johnson G, Alling C, Odham G
(1998) Determination of phosphatidylethanol in blood from alcoholic males
using high-performance liquid chromatography and evaporative light scat-
tering or electrospray mass spectrometric detection. J Chromatogr B
705:243–249.
Hannuksela ML, Liisanantti MK, Nissinen AET, Savolainen MJ (2007) Bio-
chemical markers of alcoholism. Clin Chem Lab Med 45:953–961.
Harasymiw J, Bean P (2001) The combined use of the early detection of
alcohol consumption (EDAC) test and carbohydrate-deficient transferring
to identify heavy drinking behavior in males. Alcohol Alcohol 36:349–
353.
Harasymiw J, Bean P (2007) The Early Detection of Alcohol Consumption
(EDAC) test shows better performance than gamma-glutamyltransferase
(GGT) to detect heavy drinking in a large population of males and females.
Med Sci Monit 13:PI19–PI24.
Hartmann S, Aradottir S, Graf M, Wiesbeck G, Lesch O, Ramskogler K,
Wolfersdorf M, Alling C, Wurst FM (2006) Phosphatidylethanol as a sensi-
tive and specific biomarker – comparison with gamma-glutamyl transpepti-
dase, mean corpuscular volume and carbohydrate-deficient transferrin.
Addict Biol 12:81–84.
Helander A, Beck O (2005) Ethyl sulfate: a metabolite of ethanol in humans
and a potential biomarker of acute alcohol intake. J Anal Toxicol 29:270–
274.
Helander A, Dahl H (2005) Urinary tract infection: a risk factor for false-
negative urinary ethyl glucuronide but not ethyl sulfate in the detection of
recent alcohol consumption. Clin Chem 51:1728–1730.
Helander A, Zheng Y (2009) Molecular species of the alcohol biomarker phos-
phatidylethanol in human blood measured by LC-MS. Clin Chem 55:1395–
1405.
Hermansson U, Helander A, Brandt L, Huss A, Ronnberg S (2002) The
alcohol use disorders identification test and carbohydrate-deficient trans-
ferring in alcohol-related sickness absence. Alcohol Clin Exp Res 26:28–
35.
Hermansson U, Helander A, Huss A, Brandt L, Ronnberg S (2000) The
Alcohol Use Disorders Identification Test (AUDIT) and carbohydrate-
deficient transferring (CDT) in a routine workplace health examination.
Alcohol Clin Exp Res 24:180–187.
Hietala J, Koivisto H, Anttila P, Niemela O (2006b) Comparison of the com-
bined marker GGT-CET and the conventional laboratory markers of alco-
hol abuse in heavy drinkers, moderate drinkers and abstainers. Alcohol
Alcohol 41:528–533.
Hietala J, Koivisto H, Latvala J, Anttila P, Niemela O (2006a) 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.
Hodgkinson CA, Yuan Q, Xu K, Shen PH, Heinz E, Lobos E, Binder EB,
Cubells J, Ehlers CL, Gelernter J, Mann J, Riley B, Roy A, Tabakoff B,
Todd RD, Zhou Z, Goldman D (2008) Addictions biology: haplotype-
based analysis for 130 candidate genes on a single array. Alcohol Alcohol
43:505–515.
Hoiseth G, Bernard JP, Karinen R, Johnsen L, Helander A, Christophersen
AS, Morland J (2007) A pharmacokinetic study of ethyl glucuronide in
blood and urine: applications to forensic toxicology. Forensic Sci Int
172:119–124.
Hoiseth G, Bernard JP, Stephanson N, Normann PT, Christophersen AS,
Morland J, Helander A (2008) Comparison between the urinary alcohol
markers EtG, EtS, and GTOL ⁄5-HIAA in a controlled drinking experi-
ment. Alcohol Alcohol 43:187–191.
Jansen AM, Bell LB, Sucher MA, Stoehr JD (2004) Detection of alcohol use
in monitored aftercare programs: a national survey of state physician health
programs. J Med Licens Discipl 90:8–13.
Javors MA, Johnson BA (2003) Current status of carbohydrate deficient
transferring, total serum sialic acid, sialic acid index of apolipoprotein J and
serum b-hexosaminidase as markers for alcohol consumption. Addiction 98
(Suppl 2):45–50.
Javors MA, Pichot JT, King TS, Anton RF (1997) Search for biological mark-
ers, in Drug Addiction and Its Treatment: Nexus of Neuroscience and Beha-
vior (Johnson BA, Roache JD eds), pp 205–230. Lippincott-Raven
Publishers, Philadelphia, PA.
Jeppsson JO, Arndt T, Schellenberg F, Wielders JPM, Anton RF, Whitfield
JB, Helander A (2007) Toward standardization of carbohydrate-deficient
transferring (CDT) measurements: I. Analyte definition and proposal of a
candidate reference method. Clin Chem Lab Med 45:558–562.
Johnson BA (2008) Update on neuropharmacological treatments for alco-
holism: scientific basis and clinical findings. Biochem Pharmacol 75:34–
56.
Junghanns K, Graf I, Pfluger J, Wetterling G, Ziems C, Ehrenthal D, Zollner
M, Dibbelt L, Bachaus J, Weinmann W, Wurst FM (2009) Urinary ethyl
glucuronide (EtG) and ethyl sulphate (EtS) assessment: valuable tools to
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 965
improve verification of abstention in alcohol-dependent patients during in-
patient treatment and at follow-ups. Addiction 104:921–926.
Kapoor A, Kraemer KL, Smith KJ, Roberts MS, Saitz R (2009) Cost-
effectiveness of screening for unhealthy alcohol use with % carbohydrate-
deficient transferrin: results from a literature-based decision analytic
computer model. Alcohol Clin Exp Res 33:1–10.
Kip MJ, Spies CD, Neumann T, Nachbar Y, Alling C, Aradottir S,
Weinmann W, Wurst FM (2008) The usefulness of direct ethanol
metabolites in assessing alcohol intake in nonintoxicated male patients in an
emergency room setting. Alcohol Clin Exp Res 32:1284–1291.
Kissack JC, Bishop J, Leatherwood A (2008) Ethylglucuronide as a biomarker
for ethanol detection. Pharmacotherapy 28:769–781.
Korzec A, de Bruijn C, van Lambalgen M (2005) The Bayesian Alcoholism
Test had better diagnostic properties for confirming diagnosis of hazardous
and harmful alcohol use. J Clin Epidemiol 58:1024–1032.
Korzec S, Korzed A, Conigrave K, Gisolf J, Tabakoff B (2009) Validation of
the Bayesian Alcoholism Test compared to single biomarkers in detecting
harmful drinking. Alcohol Alcohol 44:398–402.
Li TK (2007) Quantifying the risk for alcohol-use and alcohol-attributable
health disorders: present findings and future research needs. J Gastroenterol
Hepatol 23(Suppl 1):S2–S8.
Litten RZ, Allen JP (1998) Clinical application of biochemical markers of
alcohol consumption, in Use of Carbohydrate Deficient Transferrin Among
General Practitioners (Allen JP, Poldrugo F eds), pp 7–26. ICCA Publica-
tion, Lausanne, Switzerland.
Litten RZ, Allen JP, Fertig JB (1995) Gamma-glutamyltranspeptidase and
carbohydrate deficient transferring: alternative measures of excessive alcohol
consumption. Alcohol Clin Exp Res 19:1541–1546.
Litten RZ, Fertig J (2003) Self-report and biochemical measures of alcohol
consumption. Addiction 98(Suppl 2):iii–iv.
Litten RZ, Fertig J, Mattson ME, Egli M (2005) Development of medications
for alcohol use disorders: recent advances and ongoing challenges. Expert
Opin Emerg Drugs 10:323–343.
Mayfield RD, Harris RA (2009) Gene expression profiling in blood: new diag-
nostics in alcoholism and addiction? Neuropsychopharmacology 34:250–251.
McLellan AT (2007) Reducing heavy drinking: a public health strategy and a
treatment goal? J Subst Abuse Treat 33:81–83.
Miller PM, Ornstein SM, Nietert PJ, Anton RF (2004) Self-report and bio-
marker alcohol screening by primary care physicians: the need to translate
research into guidelines and practice. Alcohol Alcohol 39:325–328.
Miller WR, Walters ST, Bennett ME (2001) How effective is alcoholism treat-
ment in the United States? J Stud Alcohol 62:211–220.
Miners JO, McKinnon RA, Mackenzie PI (2002) Genetic polymorphisms of
UDP-glucuronosyltransferases and their functional significance. Toxicology
181–182:453–456.
Moore C, Jones J, Lewis D, Buchi K (2003) Prevalence of fatty acid ethyl
esters in meconium specimens. Clin Chem 49:133–136.
National Highway Traffic Safety Administration (2007) Evaluating Trans-
dermal Alcohol Measuring Devices: Final Report. NHTSA, Washington,
DC.
National Institute on Alcohol Abuse and Alcoholism (2007) Helping Patients
Who Drink too Much: A Clinician’s Guide, Updated 2005 edition, US.
Department of Health & Human Services, National Institutes of Health,
Bethesda, Maryland.
Neumann T, Gentilello LM, Neuner B, Weiss-Gerlach E, Schurmann H,
Schroder T, Muller C, Haas NP, Spies CD (2009) Screening trauma patients
with the alcohol use disorders identification test and biomarkers of alcohol
use. Alcohol Clin Exp Res 33:970–976.
Neumann T, Helander A, Dahl H, Holzmann T, Neuner B, Weib-Gerlach E,
Muller C, Spies C (2008) Value of ethyl glucuronide in plasma as a biomar-
ker for recent alcohol consumption in the emergency room. Alcohol Alco-
hol 43:431–435.
Niemela O (2007) Biomarkers in alcoholism. Clin Chim Acta 377:39–49.
Oslin DW, Berrettini W, Kranzler HR, Pettinati H, Gelernter J, Volpicelli JR,
O’Brien CP (2003) A functional polymorphism of the l-opioid receptor
gene is associated with naltrexone response in alcohol-dependent patients.
Neuorpsychopharmacology 28:1546–1552.
PichiniS,MoriniL,MarcheiE,PalmiI,RotoloMC,VagnarelliF,Garcia-
Algar O, Vall O, Zuccaro P (2009) Ethylglucuronide and ethylsulfate in
meconium to assess gestational ethanol exposure: preliminary results in two
Mediterranean cohorts. Can J Clin Pharmacol 16:e370–e375.
Pragst F, Auwaerter V, Sporkert F, 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 chromatogra-
phy-mass spectrometry (GC-MS). Forensic Sci Int 121:76–88.
Pragst F, Yegles M (2008) Determination of fatty acid ethyl esters (FAEE)
and ethyl glucuronide (EtG) in hair: a promising way for retrospective
detection of alcohol abuse during pregnancy? Ther Drug Monit 30:255–263.
Rosano TG, Lin J (2008) Ethyl glucuronide excretion in humans following
oral administration of and dermal exposure to ethanol. J Anal Toxicol
32:594–600.
Saady JJ, Poklis A, Dalton HP (1993) Production of urinary ethanol after
sample collection. J Forensic Sci 38:1467–1471.
Sakai Jt, Mikulich-Gilbertson SK, Long RJ, Crowley TJ (2006) Validity of
transdermal alcohol monitoring: fixed and self-regulated dosing. Alcohol
Clin Exp Res 30:26–33.
Salaspuro M (2007) Interrelationship between alcohol, smoking, acetaldehyde
and cancer. Novartis Found Symp 285:80–89.
Sarkola T, Dahl H, Eriksson CJP, Helander A (2003) Urinary ethyl glucuro-
nide and 5-hydroxytryptophol levels during repeated ethanol ingestion in
healthy human subjects. Alcohol Alcohol 38:347–351.
Sillanaukee P, Olsson U (2001) Improved diagnostic classification of alcohol
abusers by combing carbohydrate-deficient transferring and gamma-
glutamyltransferase. Clin Chem 47:681–685.
Skipper G, Weinmann W, Wurst FM (2004) Ethylglucuronide (EtG): a new
marker to detect alcohol use in recovering physicians. J Med Licens Discipl
90:14–17.
Stewart SH, Reuben A, Brzezinski WA, Koch DG, Basile J, Randall PK,
Miller PM (2009) Assessment and detection: preliminary evaluation of phos-
phatidylethanol and alcohol consumption in patients with liver disease and
hypertension. Alcohol Alcohol 44:464–467.
Stibler H (1991) Carbohydrate-deficient transferring in serum: a new marker
of potentially harmful alcohol consumption reviewed. Clin Chem 37:2029–
2037.
Substance Abuse and Mental Health Services Administration (2006) The role
of biomarkers in the treatment of alcohol use disorders. Subst Abuse Treat
Advisory 5:1–7.
Swift R (2003) Direct measurement of alcohol and its metabolites. Addiction
98(Suppl 2):73–80.
Tabakoff B, Snell L, Pridzun L, Herion D, Heilig M (2009) Platelets proteins
as markers for hazardous ⁄harmful alcohol consumption. Alcohol Clin Exp
Res 33:304A.
U.S. Department of Justice (1998) Bureau of Justice Statistics: Alcohol and
Crime. U.S. Government, Washington D.C.
Varga A, Hannson P, Johnson G, Alling C (2000) Normalization rate and
cellular localization of phosphatidylethanol in whole blood from chronic
alcoholics. Clin Chim Acta 299:141–150.
Varga A, Hansson P, Lundqvist C, Alling C (1998) Phosphatidylethanol in
blood as a marker of ethanol consumptioninhealthyvolunteers: compari-
son with other markers. Alcohol Clin Exp Res 22:1832–1837.
Willenbring ML (2009) Helping patients who drink too much: an evidence-
based guide for primary care physicians. Am Fam Physician 80:1–7.
Wojcik M (2008) A continuous alcohol monitoring device and its use in crimi-
nal justice settings. Alcohol Clin Exp Res 32:312A.
Wojcik MH, Hawthorne JS (2007) Sensitivity of commercial ethyl glucuronide
(EtG) testing in screening for alcohol abstinence. Alcohol Alcohol 42:317–
320.
Wurst FM, Dresen S, Allen JP, Wiesbeck G, Graf M, Weinmann W (2006)
Ethyl sulphate: a direct ethanol metabolite reflecting recent alcohol con-
sumption. Addiction 101:204–211.
Wurst FM, Dursteler-MacFarland AKM, Auwaeter V, Ergovic S, Thon N,
Yegles M, Halter C, Weinmann W, Wiesbeck G (2008c) Assessment of alco-
hol use among methadone maintenance patients by direct ethanol metabo-
lites and self-reports. Alcohol Clin Exp Res 32:1–6.
966 LITTEN ET AL.
Wurst FM, Kelso E, Weinmann W, Pragst F, Yegles M, Poromaa IS (2008a)
Measurement of direct ethanol metabolites suggests higher rate of alcohol
use among pregnant women than found with the AUDIT – a pilot study in
a population-based sample of Swedish women. Am J Obstet Gynecol
198:407.e1–407.e5.
Wurst FM, Skipper GE, 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, Wiesbeck GA, Metzger JW, Weinmann W, Graf M (2004) On
sensitivity, specificity, and the influence of various parameters on ethyl glu-
curonide levels in urine – results from the WHO ⁄ISBRA study. Alcohol
Clin Exp Res 28:1220–1228.
Wurst FM, Yegles M, Alling C, Aradottir S, Dierkes J, Wiesbeck GA, Halter
CC, Pragst F, Auwaerter V (2008b) Measurement of direct ethanol metabo-
lites in a case of a former driving under the influence (DUI) of alcohol offen-
der, now claiming abstinence. Int J Legal Med 122:235–239.
ALCOHOL BIOMARKERS IN APPLIED SETTINGS 967
