Adolescent Alcohol Consumption: Biomarkers PEth ...

COMASCO ET AL.

1

Adolescent Alcohol Consumption: Biomarkers PEth and FAEE in Relation to Interview and Questionnaire Data* ERIKA COMASCO, M.SC.,† NIKLAS NORDQUIST, PH.D.,† JERZY LEPPERT, M.D., PH.D.,† LARS ORELAND, M.D., PH.D.,† ROBERT KRONSTRAND, PH.D.,† CHRISTER ALLING, M.D., PH.D.,† AND KENT W. NILSSON, PH.D.† Department of Neuroscience, Unit of Pharmacology, Uppsala University, BMC, Box 593, 751 24 Uppsala, Sweden

ABSTRACT. Objective: The aim of this study was to investigate the congruence of biomarkers, questionnaires, and interviews as instruments to assess adolescent alcohol consumption. Method: The methodology used was a cross-sectional study with a randomized sample. Four different methods were used to estimate high adolescent alcohol consumption. The concordance of the results was investigated. Surveys were performed, and biological specimens were collected at all schools in the county of Västmanland, Sweden in 2001. Eighty-one boys and 119 girls from a population of 16- and 19-year-old adolescents were randomly selected from quartiles of volunteers representing various degrees of psychosocial risk behaviors. Using a questionnaire (for a 1-hour session) and in-depth interviews, subjects were assessed regarding their alcohol-use habits. Blood and hair samples were analyzed for phosphatidylethanol

(PEth) and fatty acid ethyl esters (FAEEs), respectively. Results: High alcohol consumption was underreported in the questionnaire compared with the interviews. PEth and FAEE analyses weakly confirmed the self-reports, and the results of the two biochemical tests did not overlap. The PEth blood test was the most specific but the least sensitive, whereas the FAEE hair test revealed low specificity and an overrepresentation of positive results in girls. Conclusions: The expected higher self-report of high alcohol consumption by interview rather than by questionnaire was confirmed partly because of the influence of a bogus pipeline procedure. The absence of overlap between PEth and FAEE results and their poor agreement with self-reports suggested that biomarkers are unsuitable as screening tools for alcohol consumption in adolescents. (J. Stud. Alcohol Drugs 70: 000-000, 2009)

T

HE WORLD HEALTH ORGANIZATION (2004) Global Status Report on Alcohol revealed that alcohol is first misused between the ages of 12 and 13. A trend toward a more hedonistic attitude for drinking (consciously using alcohol for its pleasurable psychological effects) leading to an increased consumption among youth, such as heavy episodic drinking and drunkenness, has been observed (Andersson et al., 2002; Miller et al., 2007). Several studies have attempted to assess the proportion of alcohol misuse among adolescents using different methods, such as validated questionnaires and interviews; but, it is still under debate which is the most suitable. Anonymous self-reports from adolescents are generally valid if confidentiality is stressed (Campanelli et al., 1987). Methods that inquire about the frequency and amount consumed for beer, wine, and distilled spirits separately yield the most realistic levels of intake (Feunekes et al., 1999) and show good reliability

in test-retest (Hibell et al., 1997) and parallel test procedures (Andersson et al., 1998). Questionnaires can be commonly used for large study groups but with the inherent problem of a decrease in answer accuracy and participation rate, thus resulting in statistical noise and possible sample bias. Another argument for self-administered questionnaires is that they tend to produce more valid data than verbal interviews (Harrison, 1997). On the other hand, an interview can be seen as an opportunity to gather unexpected and more detailed findings and can be especially informative of the interviewee’s social situation (e.g., excessive alcohol consumption might evoke feelings of shame, anger, and/or guilt) (Scheff, 1988). Interviews are generally better suited for small study groups because of time and cost limitations but might, in this setting, not have sufficient statistical power to identify small, but important, differences.

Received: December 2, 2008. Revision: June 25, 2009. *This study was supported by grants from the following organizations: VR (Vetenskapsrådet—Swedish Research Council) (6072), SRA(Systembolagets Råd för Alkoholforskning—Sweden’s Alcohol Monopoly Research Council), Swedish Brain Foundation, AFA (Arbetsmarknadens Försäkrings Aktiebolag—Sweden’s labor market insurance company), Fredrik and Ingrid Thurings Foundation, the Regional Research Council of the Uppsala-Örebro Region, and the County Council of Västmanland. †Correspondence may be sent to Erika Comasco at the above address or via email at: [email protected]. She is also with the Centre for

Clinical Research, Uppsala University, Central Hospital, Västerås, Sweden. Niklas Nordquist and Lars Oreland are with the Section of Pharmacology, Department of Neuroscience, Uppsala University, Uppsala, Sweden. Jerzy Leppert and Kent W. Nilsson are with the Centre for Clinical Research, Uppsala University, Central Hospital, Västerås, Sweden. Robert Kronstrand is with the National Board of Forensic Medicine, Department of Forensic Toxicology, University Hospital, Linköping, Sweden. Christer Alling is with the Department of Laboratory Medicine, Division of Clinical Chemistry and Pharmacology, Lund University Hospital, Lund, Sweden. The sponsors of this study had no role in the study design, data collection, data analysis, data interpretation, or writing of this report.

1

2

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2009

Validation of self-reports requires comparison with another method that is likely to be equally or more accurate. Some studies have assessed a pipeline procedure (Murray et al., 1987). In this procedure, subjects are told that their self-reported alcohol use can be independently verified by an investigator using a biochemical measure. This is designed to increase the accuracy of self-reports. However, the results of such a procedure have been questioned (Campanelli et al., 1987; Murray and Perry, 1987; Wagenaar et al., 1993; Werch et al., 1987). Over the past few decades, new methods have been developed for the analysis of alcohol metabolites from biological specimens, so-called biomarkers, to indicate past ethanol consumption (Stibler, 1991). Phosphatidylethanol (PEth) and fatty acid ethyl esters (FAEEs) are suggested as direct alcohol markers because the traditional ones—such as carbohydrate-deficient transferrin, gamma-glutamyltransferase, and mean corpuscular volume—have limitations, either in sensitivity or specificity. The blood biomarker PEth is a phospholipid formed in cell membranes only in the presence of ethanol (Alling et al., 1983, 1984) via the action of the enzyme phospholipase D (Alling et al., 1984; Aradottir et al., 2002; Gustavsson and Alling, 1987; Hansson et al., 1997; Ishikawa, 1999; Lundqvist et al., 1994; Varga et al., 2000). It is a promising indicator of alcohol consumption because it has a relatively long half-life (Varga and Alling, 2002), being detectable up to 14 days after sobriety (Aradottir et al., 2004a, 2006; Hansson et al., 1997; Hartmann et al., 2007; Varga et al., 1998; Wurst et al., 2005). Alcohol intake can also be identified through FAEEs (ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate) in plasma or serum to detect ethanol consumption for the past day and in hair to detect chronic ethanol consumption. Only produced when alcohol is present in the bloodstream (Auwarter et al., 2001; Brown, 1985; Doyle et al., 1994, 1996; Klein et al., 1999; Laposata and Lange, 1986; Pragst et al., 2001; Wurst et al., 2004), these are the endproducts of nonoxidative metabolism that are derived from the esterification of ethanol with endogenous free fatty acids. The reaction is catalyzed primarily by acyl-coenzyme A:ethanol O-acyltransferase and FAEE synthase (Dan and Laposata, 1997; Laposata, 1998; Laposata and Lange, 1986; Laposata et al., 1987). The mechanism by which FAEEs are incorporated into hair has not yet been precisely defined (Pragst and Balikova, 2006). One possibility is that the biomarkers are transported into the growing hair cells via simple passive diffusion from the bloodstream, whereas another possibility is that they are deposited into the hair shaft through the sebaceous glands (Auwarter et al., 2001; Brown, 1985). Hair analysis, in particular, is considered a promising method for providing standardized validity criterion measurements, because alcohol use can be detected over a long period, and hair samples can be easily obtained (Harrison, 1997).

The present study investigated the congruence of biomarkers, questionnaires, and interviews as instruments to assess adolescent alcohol consumption. Method Population Participants were recruited from 2,611 students age 16 and from 1,649 students age 19 in the Swedish county of Västmanland who answered the Survey of Adolescent Life in Vestmanland during 2001. The questionnaire covered alcohol consumption, risk behavior, and a wide range of psychosocial variables. It was completed in the classroom during a 1-hour session with the research assistant. Subjects were classified according to a risk index, depending on the risk behavior reported in the questionnaire relating to alcohol, drug, property, and violent offenses. Four hundred students matched for age, gender, and risk behavior were randomly selected. The initial risk survey was performed to ensure that enough participants from both ends of the deviant behavior continuum were included in the study. Informed consent was obtained from 200 individuals, and they agreed to give biological samples and to take part in an interview. Seventy-eight girls and 57 boys were age 16, and 41 girls and 24 boys were age 19. For a more detailed description of the population, see Nilsson et al. (2005). All participants were asked to give a venous blood sample and three hair samples from different areas of the scalp. They were told that these samples would be analyzed for tobacco, alcohol, and other drugs by the forensic department (Wagenaar et al., 1993). Blood was obtained from all participants; 196 participants gave hair samples, and 4 participants could not provide hair samples because they were bald. The study was approved by the Human Ethics Committee of the Faculty of Medicine of Uppsala University. Instrumentation PEth was analyzed by high-performance liquid chromatography (Waters Alliance 2690 system; Milford, MA) and evaporation light-scattering detection (PL-ELS 2100 Polymer Laboratories, Shropshire, UK). The equipment was operated as previously described (Varga et al., 1998). FAEE was analyzed by headspace solid-phase microextraction and gas chromatography (Auwarter et al., 2004). PEth analysis Blood samples have been stored at -80º C until tested in 2007. They were analyzed together with control samples at two levels—(1) low: 0.92 µmol/L and (2) high: 3.2 µmol/ L—using phosphatidylbutanol as an internal standard. A special evaluative algorithm was applied (Aradottir and Ols-

COMASCO ET AL. son, 2005). Precision of the method was determined by the coefficient of variation (CV) for peak areas at low and high levels. These CVs were 13 CV% (total) for the low level and 6 CV% (total) for the high level. The limit of quantification was fixed at 0.25 µmol/L. Concentrations below that level were not present; therefore, results were dichotomized into negative (0) and positive (1) values. FAEE analysis Three hair samples were collected from the back, lateral, and frontal parts of the head by cutting as close as possible to the scalp to obtain proximal 0-6 cm segment. Samples were wrapped in tinfoil and saved in dry storage at room temperature, and analyses were conducted in 2004. Extraction and analysis of FAEEs in hair were performed as described by Auwärter et al. (2004). The amount of FAEEs found in hair was measured in nanograms. The sum of the FAEE concentrations (CFAEE) of the four esters—(1) ethyl myristate, (2) ethyl palmitate, (3) ethyl oleate, and (4) ethyl stearate—was used for evaluation of alcohol use. This helped distinguish between the following four intervals: (1) CFAEE ≤ 0.12 ng/mg for teetotalers, (2) CFAEE = 0.13-0.39 ng/mg for social drinkers, (3) CFAEE = 0.40-0.66 ng/mg for heavy drinkers, and (4) CFAEE ≥ 0.67 ng/mg for alcoholics (Yegles et al., 2004). Two cutoffs were established in the present study: (1) CFAEE ≥ 0.40 ng/mg and (2) CFAEE ≥ 0.67 ng/mg (Auwarter et al., 2001; Hartwig et al., 2003; Kulaga et al., 2006; Pragst and Balikova, 2006; Yegles et al., 2004). Hair samples could not be obtained from four individuals, and chemical analysis failed in 43 cases. Definition of alcohol drinking patterns by questionnaire Questions about the amount and frequency of alcohol intake were used as described in a report by the European School Survey Project on Alcohol and Other Drugs (Laposata et al., 1987). The frequency variable was calculated by the addition of frequencies of consumption for each of the alcoholic beverages, ranging from 0 to 365 times a year. The frequency of recurring intoxication was measured using the question: “How often do you get drunk when you consume alcohol?” On a six-point scale, possible answers included the following: (1) do not drink alcohol, (2) never, (3) seldom, (4) occasionally, (5) almost always, and (6) always. A new variable for high versus low alcohol consumption was created by combining the following two variables: (1) frequency of alcohol consumption and (2) frequency of recurring intoxication. This new variable was dichotomized into the following: those who consumed alcohol twice a month or more and always or almost always got drunk (“high alcohol consumption”); and those who consumed alcohol less frequently and never, seldomly, or occasionally became intoxicated, or inter-

3

mediate-frequency alcohol consumers who never or seldomly became intoxicated (“low alcohol consumption”). Interview The semistructured interview was adapted to the situation regarding formulation, sequence, and extent to reflect alcohol habits in a qualitative dimension. The interview answers were used to code the amount and frequency of different alcoholic beverages (Nilsson et al., 2005). Those adolescents who drank two times or more per month and always, or almost always, became drunk were coded as “high alcohol consumers.” Those individuals who consumed alcohol less frequently or frequently, but who were never or seldomly drunk, were coded as “low alcohol consumers.” Statistical analysis To measure gender differences in high alcohol consumption, we used the chi-square test (χ2). Spearman’s rho was performed to study the relationship between self-reported alcohol consumption in grams of alcohol per year, PEth, and CFAEE. Cohen’s kappa was applied to analyze the intermeasurement agreement of alcohol consumption for the different methods. Results The mean (SD) age at first intoxication for both boys and girls was 14 (1.75) years. Of 200 adolescents, 104 (52%) were classified as high alcohol consumers based on interview answers. However, based on the self-report questionnaire, 70 of 200 (35%) adolescents were classified as high alcohol consumers. Blood analysis of PEth identified 13 individuals as high alcohol consumers using a cutoff value of 0.25 µmol/L. A chi-square test for group differences revealed that boys were significantly overrepresented among individuals positive for PEth (χ2 = 4.69, 1 df, p = .03). According to hair FAEE quantification, 20 and 52 individuals exceeded the high- and low-cutoff levels, respectively. A nonsignificant trend toward an overrepresentation of girls in the high-level group (CFAEE ≥ 0.67 ng/mg) was observed (χ2 = 3.06, 1 df, p = .08). Further details are presented in Table 1. The self-reported alcohol consumption in grams per year was positively correlated to PEth (r = .154, p = .030) and negatively correlated to CFAEE (r = -.211, p = .009). There was no correlation between PEth and CFAEE (r = .107, p = .207) (Table 1). The rate of congruence for questionnaire versus interview was 89%; that is, of the 70 individuals who were high alcohol consumers, according to the questionnaire, 62 of these individuals were recognized by the interview. The corresponding figure for interview versus questionnaire was 64%. There was a moderate agreement between interview and questionnaire (κ = 0.506, p < .001). As can be seen in

4

JOURNAL OF STUDIES ON ALCOHOL AND DRUGS / SEPTEMBER 2009 TABLE 1. Frequency of high alcohol consumption according to interview, self-report questionnaire, blood (PEth) and hair (FAEE) data, stratified by gender. Chi-square values and p values are given for group differences between boys and girls

Variable

High alcohol consump.

Girls (n = 119/200)

Boys (n = 81/200)

Failed

χ2 (1 df)

p

Interview Questionnaire Blood (PEth) Hair (CFAEE ≥ 0.40ng/mg) Hair (CFAEE ≥ 0.67ng/mg)

104 (52%) 70 (35%) 13 (6.5%) 52 (31%) 20 (13%)

65 (55%) 46 (39%) 4 (3%) 38 (37%) 17 (16%)

39 (48%) 24 (30%) 9 (11%) 14 (29%) 3 (6%)

– – 1 (0.5%) 47 (24%)a 47 (24%)a

0.81 0.92 4.69* 0.94 3.06

.37 .34 .03 .33 .08

Notes: PEth = phosphatidylethanol; CFAEE = fatty acid ethyl ester concentration; consump. = consumption. aFour individuals were bald, and therefore no hair sample was taken. *p < .05.

Figure 1, no agreement was found between interview and PEth (κ = 0.043, p = .205) or CFAEE (CFAEE ≥ 0.40 ng/mg: κ = -0.048, p = .426; CFAEE ≥ 0.67 ng/mg: κ = -0.030, p = .548). None of the individuals who were classified as high alcohol consumers by PEth was simultaneously classified by CFAEE or vice versa. Only 9% of those who were classified as high alcohol consumers in the interview tested positive for PEth in blood analysis (mean: 0.40 µmol/L; range: 0.25-0.71 µmol/L). Among individuals who tested positive for PEth, 46% were confirmed by questionnaire and 69% by interview, although they all scored in the lowest interval for CFAEE (0-0.12 ng/mg). Twelve individuals were found within the CFAEE interval of 0-0.12 ng/mg, 89 were found within 0.13-0.39 ng/mg, 32 were found within 0.40-0.66 ng/mg, and 20 were found above 0.67 ng/mg. Fifty-two individuals scored above the cutoff point of 0.40 ng/mg. In the group that had the highest levels of CFAEE (20 of 153 individuals with CFAEE ≥ 0.67

PEth

ng/mg), 50% were classified as high alcohol consumers according to interview and 30% according to questionnaire. Five individuals in this group (45%), all of whom were girls, were teetotalers according to interview and questionnaire. Sensitivity and specificity of the biochemical markers for the detection of high alcohol consumption, as established by interview, are presented in Figure 2. Given that PEth in blood can provide information of high alcohol intake only for a limited period over the previous 2 weeks, the test nevertheless correctly identified 9 of 13 individuals, who, according to interview, had high alcohol consumption. The remaining four individuals had been classified as low alcohol consumers, and 95 individuals defined as high alcohol consumers by interview were not detected by PEth (Table 2). The positive predictive value of the PEth test was 69%, which refers to the likelihood that a positive test result will be correct, whereas the negative predictive value was 49%, with 92 of 187 negative test results being correct. The positive and negative predictive values for the cutoff CFAEE

CFAFE ≥ 0.40 ng/mg

PEth

CFAFE ≥ 0.67 ng/mg

κ = .04 p = .20

κ = .04 p = .20

κ = .03 p = .55 κ = .05 p = .43

FIGURE 1. Venn diagrams illustrating agreement between a) interview, phosphatidylethanol (PEth), and CFAEE ≥ 0.40 ng/mg; and b) interview, PEth, and CFAEE ≥ 0.67 ng/mg among adolescents classified as high alcohol consumers according to the different methods. Numbers of individuals are indicated in the diagram, together with Cohen’s kappa (κ), as a measure of interrater agreement. CFAEE = fatty acid ethyl ester concentration; FAEE = fatty acid ethyl esters.

COMASCO ET AL.

FIGURE 2. Sensitivity and specificity of blood (phosphatidylethanol [PEth] ≥ 0.25 µmol/L) and hair analyses (CFAEE ≥ 0.67 ng/mg and CFAEE ≥ 0.40 ng/mg) in relation to interview. CFAEE = fatty acid ethyl ester concentration. Δ PEth ≥ 0.25 µmol/L; ■ CFAEE ≥ 0.67 ng/mg; ❍ CFAEE ≥ 0.40 ng/mg.

≥ 0.40 ng/mg were 50% and 41%, respectively, according to interview. Discussion In the present study, there was a moderate agreement for questionnaire versus interview (κ = .506, p < .001), identifying 70 of 104 high alcohol consumers. The PEth test had a high specificity (96%) but a low sensitivity (9%). The FAEE test (CFAEE ≥ 0.40 ng/mg) had a low specificity (30%)

5

and a positive predictive value of 50%, despite its higher sensitivity (61%). The number of high alcohol consumers identified by FAEE analysis, yet not confirmed by interview, was high (50%), particularly among girls. The relationship between PEth and CFAEE was contradictory; high consumption according to PEth levels was not associated with high consumption according to FAEEs levels and vice versa. The suggestion that self-reports in questionnaires from adolescents are generally valid (Campanelli et al., 1987) and highly reliable (Andersson et al., 1998; Hibell et al., 1997), if frequency and consumption levels of different beverages are inquired about separately (Feunekes et al., 1999), was not confirmed by the present study. The higher self-reported alcohol use from interview, when the subjects knew that their answers would be checked against biochemical analyses— rather than from questionnaire—confirms the existence of a “pipeline procedure” effect, as previously observed (Murray et al., 1987); nonetheless, some research has questioned the existence of such an effect (Campanelli et al., 1987; Murray and Perry, 1987; Wagenaar et al., 1993; Werch et al., 1987). This indicates that adolescents underreport their alcohol consumption in questionnaires in a similar way to adults (Hoyer et al., 1995). Therefore, a semistructured interview method should provide more reliable measurements than a questionnaire, especially if combined with a biological sample test. There has been little uniformity in the tools used to monitor alcohol consumption among youth. Most research conducted on the validation of self-reports has focused on particular sample populations or on different age groups without proving generalizability of the results. Several biomarkers have been evaluated for alcoholics in different clinical settings and applied in forensic investigations, legal cases, or doping controls but never in an adolescent population. During the last decade, there has been a strong focus on direct ethanol metabolites, but these new methodologies have yet to be formally contrasted and compared. Crucial points still remain, such as the identification of those samples that exceed the cutoff levels as false positives and the correct

TABLE 2. Frequency of high alcohol consumption according to the different scoring methods, stratified by gender, and discordance frequency in relation to interview Discordances vs interview Total High alcohol consumers Girls High alcohol consumers Positive discordancea Negative discordanceb Boys High alcohol consumers Positive discordancea Negative discordanceb

PEth (≥ 0.25 Interview Questionnaire µmol/L)

CFAEE (≥ 0.40 ng/mg)

CFAEE (≥ 0.67 ng/mg)

104

70

13

52

20

65 – –

46 6 25

4 1 62

38 17 37

17 9 50

39 – –

24 2 17

9 3 33

14 9 23

3 1 26

Notes: CFAEE = fatty acid ethyl ester concentration. aPositives results not confirmed by interview; bnegative results that are positive according to interview.

6


interpretation of the results below the cutoff level, which represents light or no drinking. PEth formation occurs only after ethanol intake, which should give it a biological specificity close to 100%. However, a few studies have concluded that a single dose of ethanol, even as high as 50 g, does not produce measurable amounts of PEth in the blood (Hansson et al., 1997; Varga et al., 1998; Wurst et al., 2005). This finding is clinically advantageous because it implies that the assay will neither detect single drinking episodes nor, presumably, the majority of social drinkers, which might encompass a large part of the sample in this study. Moreover, PEth in the blood can give information on alcohol consumption only for the last few weeks (Hartmann et al., 2007) and at clinical levels; only values above 0.70 µmol/L are considered because the correlation for values between 0.25 µmol/L and 0.70 µmol/L and the amount of alcohol consumed is still uncertain (Varga et al., 1998). The present data, therefore, do not support the use of PEth as a potential sensitive and specific marker for alcohol misuse among adolescents. It is suitable for reflecting a long-lasting intake of higher amounts of alcohol, which is what it was originally developed to detect. Hair is recognized as an alternative biological specimen for alcohol testing, having the practical advantages that sample collection is noninvasive and the sample itself easy to handle and store. Hair FAEEs can be traced retrospectively for months and can, therefore, be used as indicators of chronic alcohol consumption, although the interpretation of the dose–concentration relationship is unclear. It is also still uncertain how alcohol is incorporated into the hair and which factors influence its deposition (hair thickness and porosity; hair growth and sebaceous secretions; metabolic activity; and the individual’s ethnicity, age, gender, and genetic variability) (Auwarter et al., 2004). Reliable reference ranges for FAEEs and a cutoff level to distinguish nondrinkers or light drinkers (CFAEE < 0.40 ng/mg) from excessive drinkers (CFAEE > 0.40 ng/mg) have been established (Yegles et al., 2004). However, the method was not developed to detect teetotalers or social drinkers, and a positive result is significant only if it is above the cutoff level for excessive alcohol consumption. Furthermore, the origin of low hair FAEE concentrations in teetotalers has not yet received a clear explanation. False-positive results may be the result of external sources of FAEEs (Auwarter et al., 2001; Hartwig et al., 2003), which could explain the high frequency of elevated FAEE results in girls compared with boys, because the use of hair care products is more common among girls. It has been suggested that the concentration of FAEEs is not significantly affected by cosmetic hair treatment and normal hair care, but other studies have indicated that such treatments, as well as environmental conditions, may indeed alter the presence of FAEEs (Hartwig et al., 2003; Kulaga et al., 2006).

The present study has some limitations. First, the biochemical tests have been developed for alcoholics, whereas adolescents can be classified only as high alcohol consumers; they do not drink heavily and regularly compared with alcoholic adults. Hence, these methods could be considered inappropriate for measuring adolescent alcohol consumption in a population in which primarily episodic use is assumed. A second limitation is the difference in the detection time spans covered by the different measurements. PEth has a half-life of only a few weeks, whereas FAEEs are detectable for approximately 6 months (for a hair length of 0-6 cm). However, interview data estimated the long-term mean alcohol consumption (grams per year) and classified high versus low alcohol consumption. Even so, it was PEth that correlated positively to the self-report of alcohol intake in grams per year, whereas CFAEE was negatively correlated. A third limitation concerning the markers is the low number of positive PEth results and the high frequency of failed FAEE analyses, thus making it difficult to draw general conclusions from the study because of low statistical power. A fourth limitation is the possible influence of cosmetics on the hair analysis, particularly for girls, because the majority of adolescents in the present study might have used hair care products containing alcohol (Yegles et al., 2004). A fifth limitation results from the sample determination time: the study was performed in 2001 but the blood and hair analyses were conducted, respectively, in 2007 and 2004. There is no study indicating stability of PEth and FAEE, even at -80º C for several years (Aradottir et al., 2004b). The last limitation concerns the small size of the study population. To confirm the results, a larger sample size would be needed in a replication study. However, this is a random sample of adolescents from 4,260 individuals in which we investigated self-reported alcohol consumption with a well-established questionnaire, in-depth interviews, and blood and hair analyses. This study provides a portrait of adolescent drinking patterns highlighting how common periodic heavy episodic drinking is in Nordic countries (Donovan et al., 2004; Hibell et al., 2004). However, the authors do not want to misstate the potential of PEth and FAEEs, because both are described in the literature as direct ethanol metabolites, thus reflecting longer lasting intake of larger amounts of ethanol. The reader should not draw the incorrect conclusion that these markers lack sensitivity and specificity if used in the correct context. In conclusion, alcohol is the most misused drug in Western society with an increasing heavy episodic drinking culture—creating a growing alcohol-related crisis, especially among adolescents—that requires adequate screening tools. Validity research is still in its early stages, and some discrepancies between analysis results cannot be easily explained at present. Further investigations are required to examine the applicability of different methods for the detection of social drinkers as opposed to heavy drinkers.

COMASCO ET AL. Acknowledgments The authors thank Professor Dr. Fritz Pragst for his assistance with FAEE analysis and Dr. Therese Hansson for her assistance with PEth analysis.

References ALLING, C., GUSTAVSSON, L., AND ANGGARD, E. An abnormal phospholipid in rat organs after ethanol treatment. FEBS Lett. 152: 24-28, 1983. ALLING, C., GUSTAVSSON, L., MANSSON, J.E., BENTHIN, G., AND ANGGARD, E. Phosphatidylethanol formation in rat organs after ethanol treatment. Biochim. Biophys. Acta 793: 119-122, 1984. ANDERSSON, B., GRÖNBERG, K., AND HIBELL, B. Skolelever drogvanor: Tabeller med sammanfattande kommentarer, 1997 (School Pupils’ Drug Habits, 1997), Stockholm, Sweden: Centralförbundet för alkohol-och narkotikaupplysning (CAN) 1998. ANDERSSON, B., HANSAGI, H., DAMSTROM THAKKER, K., AND HIBELL, B. Longterm trends in drinking habits among Swedish teenagers: National School Surveys 1971-1999. Drug Alcohol Rev. 21: 253-260, 2002. ARADOTTIR, S., ASANOVSKA, G., GJERSS, S., HANSSON, P., AND ALLING, C. PHosphatidylethanol (PEth) concentrations in blood are correlated to reported alcohol intake in alcohol-dependent patients. Alcohol Alcsm 41: 431-437, 2006. ARADOTTIR, S., LUNDQVIST, C., AND ALLING, C. Phosphatidylethanol in rat organs after ethanol exposure. Alcsm Clin. Exp. Res. 26: 514-518, 2002. ARADOTTIR, S., MOLLER, K., AND ALLING, C. Phosphatidylethanol formation and degradation in human and rat blood. Alcohol Alcsm 39: 8-13, 2004a. ARADOTTIR, S. AND OLSSON, B.L. Methodological modifications on quantification of phosphatidylethanol in blood from humans abusing alcohol, using high-performance liquid chromatography and evaporative light scattering detection. BMC Biochem. 6: 18, 2005. ARADOTTIR, S., SEIDL, S., WURST, F.M., JONSSON, B.A.G., AND ALLING, C. Phosphatidylethanol in human organs and blood: A study on autopsy material and influences by storage conditions. Alcsm Clin. Exp. Res. 28: 1718-1723, 2004b. AUWARTER, V., KIESSLING, B., AND PRAGST, F. Squalene in hair—a natural reference substance for the improved interpretation of fatty acid ethyl ester concentrations with respect to alcohol misuse. Forensic Sci. Int. 145: 149-159, 2004. AUWARTER, V., SPORKERT, F., HARTWIG, S., PRAGST, F., VATER, H., AND DIEFENBACHER, A. 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, 2001. BROWN, D.J. The pharmacokinetics of alcohol excretion in human perspiration. Meth. Find. Exp. Clin. Pharmacol. 7: 539-544, 1985. CAMPANELLI, P.C., DIELMAN, T.E., AND SHOPE, J.T. Validity of adolescents’ self-reports of alcohol use and misuse using a bogus pipeline procedure. Adolescence 22: 7-22, 1987. DAN, L. AND LAPOSATA, M. Ethyl palmitate and ethyl oleate are the predominant fatty acid ethyl esters in the blood after ethanol ingestion and their synthesis is differentially influenced by the extracellular concentrations of their corresponding fatty acids. Alcsm Clin. Exp. Res. 21: 286-292, 1997. DONOVAN, J.E., LEECH, S.L., ZUCKER, R.A., LOVELAND-CHERRY, C.J., JESTER, J.M., FITZGERALD, H.E., PUTTLER, L.I., WONG, M.M., AND LOOMAN, W.S. Really underage drinkers: Alcohol use among elementary students. Alcsm Clin. Exp. Res. 28: 341-349, 2004. DOYLE, K.M., BIRD, D.A., AL-SALIHI, S., HALLAQ, Y., CLUETTE-BROWN, J.E., GOSS, K.A., AND LAPOSATA, M. Fatty acid ethyl esters are present in human serum after ethanol ingestion. J. Lipid Res. 35: 428-437, 1994. DOYLE, K.M., CLUETTE-BROWN, J.E., DUBE, D.M., BERNHARDT, T.G., MORSE,

7

C.R., AND LAPOSATA, M. Fatty acid ethyl esters in the blood as markers for ethanol intake. JAMA 276: 1152-1156, 1996. FEUNEKES, G.I., VAN’T VEER, P., VAN STAVEREN, W.A., AND KOK, F.J. Alcohol intake assessment: The sober facts. Amer. J. Epidemiol. 150: 105-112, 1999. GUSTAVSSON, L. AND ALLING, C. Formation of phosphatidylethanol in rat brain by phospholipase D. Biochem. Biophys. Res. Commun. 142: 958-963, 1987. HANSSON, P., CARON, M., JOHNSON, G., GUSTAVSSON, L., AND ALLING, C. Blood phosphatidylethanol as a marker of alcohol abuse: Levels in alcoholic males during withdrawal. Alcsm Clin. Exp. Res. 21: 108-110, 1997. HARRISON, L. The validity of self-reported drug use in survey research: An overview and critique of research methods. In: HARRISON, L. AND HUGHES, A. (Eds.) The Validity of Self-Reported Drug Use: Improving the Accuracy of Survey Estimates. NIDA Research Monograph No. 167, NIH Publication No. 97-4147, Washington: Government Printing Office, 1997, pp. 17-36. HARTMANN, S., ARADOTTIR, S., GRAF, M., WIESBECK, G., LESCH, O., RAMSKOGLER, K., WOLFERSDORF, M., ALLING, C., AND WURST, F.M. Phosphatidylethanol as a sensitive and specific biomarker: Comparison with gamma-glutamyl transpeptidase, mean corpuscular volume and carbohydrate-deficient transferrin. Addict. Biol. 12: 81-84, 2007. HARTWIG, S., AUWARTER, V., AND PRAGST, F. Effect of hair care and hair cosmetics on the concentrations of fatty acid ethyl esters in hair as markers of chronically elevated alcohol consumption. Forensic Sci. Int. 131: 90-97, 2003. HIBELL, B., ANDERSSON, B., BJARNASON, T., AHLSTRÖM, S., BALAKIREVA, O., KOKKEVI, A., AND MORGAN, M. The ESPAD Report 2003. Alcohol and Other Drug Use Among Students in 35 European Countries, Stockholm, Sweden: The Swedish Council for Information on Alcohol and Other Drugs (CAN), Council of Europe, Co-operation Group to Combat Drug Abuse and Illicit Trafficking in Drugs (the Popmpidou Group), 2004. HIBELL, B., ANDERSSON, B., BJARNASON, T., KOKKEVI, A., MORGAN, M., AND NARUSK, A. The 1995 ESPAD Report. Alcohol and Other Drug Use Among Students in 26 European Countries, Stockholm, Sweden: The Swedish Council for Information on Alcohol and Other Drugs (CAN), Council of Europe, Co-operation Group to Combat Drug Abuse and Illicit Trafficking in Drugs (the Popmpidou Group), 1997. HOYER, G., NILSSEN, O., BRENN, T., AND SCHIRMER, H. The Svalbard Study 1988-89: A unique setting for validation of self-reported alcohol consumption. Addiction 90: 539-544, 1995. ISHIKAWA, H., OHTSUKI, T., ISHIGURO, H., YAMAKAWA-KOBAYASHI, K., ENDO, K., LIN, Y.-L., YANAGI, H., TSUCHIYA, S., KAWATA, K.-I, HAMAGUCHI, H., AND ARINAMI, T. Association between serotonin transporter gene polymorphism and smoking among Japanese males. Cancer Epidemiol. Biomark. Prev. 8: 831-833, 1999. KLEIN, J., KARASKOV, T., AND KORENT, G. Fatty acid ethyl esters: A novel biologic marker for heavy in utero ethanol exposure: a case report. Therapeut. Drug Monit. 21: 644-646, 1999. KULAGA, V., CAPRARA, D., IQBAL, U., KAPUR, B., KLEIN, J., REYNOLDS, J., BRIEN, J., AND KOREN, G. Fatty acid ethyl esters (FAEE): Comparative accumulation in human and guinea pig hair as a biomarker for prenatal alcohol exposure. Alcohol Alcsm 41: 534-539, 2006. LAPOSATA, M. Fatty acid ethyl esters: Ethanol metabolites which mediate ethanol-induced organ damage and serve as markers of ethanol intake. Prog. Lipid Res. 37: 307-316, 1998. LAPOSATA, E.A. AND LANGE, L.G. Presence of nonoxidative ethanol metabolism in human organs commonly damaged by ethanol abuse. Science 231: 497-499, 1986. LAPOSATA, E.A., SCHERRER, D.E., MAZOW, C., AND LANGE, L.G. Metabolism of ethanol by human brain to fatty acid ethyl esters. J. Biol. Chem. 262: 4653-4657, 1987. LUNDQVIST, C., ARADOTTIR, S., ALLING, C., BOYANO-ADANEZ, M.C., AND GUSTAVSSON, L. Phosphatidylethanol formation and degradation in brains of acutely and repeatedly ethanol-treated rats. Neurosci. Lett. 179: 127-131, 1994.

8


MILLER, J.W., NAIMI, T.S., BREWER, R.D., AND JONES, S.E. Binge drinking and associated health risk behaviors among high school students. Pediatrics 119: 76-85, 2007. MURRAY, D.M., O’CONNELL, C.M., SCHMID, L.A., AND PERRY, C.L. The validity of smoking self-reports by adolescents: A reexamination of the bogus pipeline procedure. Addict. Behav. 12: 7-15, 1987. MURRAY, D.M. AND PERRY, C.L. The measurement of substance use among adolescents: When is the “bogus pipeline” method needed? Addict. Behav. 12: 225-233, 1987. NILSSON, K.W., SJOBERG, R.L., DAMBERG, M., ALM, P.O., OHRVIK, J., LEPPERT, J., LINDSTROM, L., AND ORELAND, L. Role of the serotonin transporter gene and family function in adolescent alcohol consumption. Alcsm Clin. Exp. Res. 29: 564-570, 2005. PRAGST, F., AUWAERTER, V., SPORKERT, F., AND SPIEGEL, K. Analysis of fatty acid ethyl esters in hair as possible markers of chronically elevated alcohol consumption by headspace solid-phase microextraction (HSSPME) and gas chromatography-mass spectrometry (GC-MS). Forensic Sci. Int. 121: 76-88, 2001. PRAGST, F. AND BALIKOVA, M.A. State of the art in hair analysis for detection of drug and alcohol abuse. Clin. Chim. Acta 370: 17-49, 2006. SCHEFF, T. Shame and conformity: The Deference-Emotion System. Amer. Sociol. Rev. 53: 395-406, 1988. STIBLER, H. Carbohydrate-deficient transferrin in serum: A new marker of potentially harmful alcohol consumption reviewed. Clin. Chem. 37: 2029-2937, 1991. VARGA, A. AND ALLING, C. Formation of phosphatidylethanol in vitro in red blood cells from healthy volunteers and chronic alcoholics. J. Lab. Clin. Med. 140: 79-83, 2002. VARGA, A., HANSSON, P., JOHNSON, G., AND ALLING, C. Normalization rate and cellular localization of phosphatidylethanol in whole blood from chronic alcoholics. Clin. Chim. Acta 299: 141-150, 2000.

VARGA, A., HANSSON, P., LUNDQVIST, C., AND ALLING, C. Phosphatidylethanol in blood as a marker of ethanol consumption in healthy volunteers: Comparison with other markers. Alcsm Clin. Exp. Res. 22: 1832-1837, 1998. WAGENAAR, A.C., KOMRO, K.A., MCGOVERN, P., WILLIAMS, C.L., AND PERRY, C.L. Effects of a saliva test pipeline procedure on adolescent self-reported alcohol use. Addiction 88: 199-208, 1993. WERCH, C.E., GORMAN, D.R., MARTY, P.J., FORBESS, J., AND BROWN, B. Effects of the bogus-pipeline on enhancing validity of self-reported adolescent drug use measures. J. School Hlth 57: 232-236, 1987. WORLD HEALTH ORGANIZATION. Global Status Report on Alcohol 2004, Geneva, Switzerland: Department of Mental Health and Substance Abuse, World Health Organization, 2004. WURST, F.M., ALEXSON, S., WOLFERSDORF, M., BECHTEL, G., FORSTER, S., ALLING, C., ARADOTTIR, S., JACHAU, K., HUBER, P., ALLEN, J.P., AUWARTER, V., AND PRAGST, F. Concentration of fatty acid ethyl esters in hair of alcoholics: Comparison to other biological state markers and self reported-ethanol intake. Alcohol Alcsm 39: 33-38, 2004. WURST, F.M., TABAKOFF, B., ALLING, C., ARADOTTIR, S., WIESBECK, G.A., MULLER-SPAHN, F., PRAGST, F., JOHNSON, B., JAVORS, M., AIT-DAOUD, N., SKIPPER, G.E., SPIES, C., NACHBAR, Y., LESCH, O., RAMSKOGLER, K., HARTMANN, S., WOLFERSDORF, M., DRESEN, S., WEINMANN, W., HINES, L., KAISER, A., LU, R.-B., KO, H.-C., HUANG, S.-Y., WANG, T.-J., WU, Y.-S., WHITFIELD, J., SNELL, L.D., WU, C., AND HOFFMAN, P.L. World Health Organization/International Society for Biomedical Research on Alcoholism study on state and trait markers of alcohol use and dependence: Back to the future. Alcsm Clin. Exp. Res. 29: 1268-1275, 2005. YEGLES, M., LABARTHE, A., AUWARTER, V., HARTWIG, S., VATER, H., WENNIG, R., AND PRAGST, F. Comparison of ethyl glucuronide and fatty acid ethyl ester concentrations in hair of alcoholics, social drinkers and teetotallers. Forensic Sci. Int. 145: 167-173, 2004.