*Department of Medicine, University of California, San Francisco, CA
†Department of Epidemiology and Biostatistics, University of California, San Francisco, CA
‡Department of Community Health, Mbarara University of Science and Technology, Uganda
§Massachusetts General Hospital Center for Global Health, Boston, MA Ragon Institute of MGH, MIT and Harvard, Charlestown, MA Department of Global Health and Social Medicine, Harvard Medical School, Boston, MA Faculty of Medicine, Mbarara University of Science and Technology, Mbarara, Uganda
Supported by grants from the National Institutes of Health, (University of California San Francisco-Gladstone Institute of Virology & Immunology Center for AIDS Research P30 AI027763), P30 DK026743, R01 MH54907, K-24 MH87227, and R01 AA018631.
The authors have no conflicts of interest to disclose.
To the Editors:
There is a significant overlap between the problem of heavy alcohol consumption and the HIV epidemic in sub Saharan Africa (SSA). Heavy alcohol consumption has direct health consequences and increases the risk of HIV transmission, incomplete adherence, and disease progression in SSA.1 However, the criterion validity of patient-reported alcohol consumption by HIV-infected persons in SSA is unknown. Valid report is necessary for implementing interventions to reduce alcohol consumption that could in turn reduce the burden of HIV disease.2
We investigated the effect of introducing biologic measures on self-report of alcohol consumption among persons on HIV antiretroviral therapy (ART) in Mbarara, Uganda. In Uganda, the adult HIV prevalence is 6.5% and 47% with CD4 counts <350 cells per cubic millimeter are on ART.3 Participants of a prospective cohort study of HIV ART treatment outcomes were recruited into a nested study. Roughly equal numbers of participants with fully and incompletely suppressed HIV RNA (> or ≤400 copies/mL) were selected (29 incompletely suppressed, 32 suppressed) matched on sex (52.5% female) and duration on ART (within six months); all had been enrolled in the parent study for at least 6 months.
Parent study activities included baseline and quarterly structured interviews and blood draws. Participants were asked at baseline when, if ever, they last consumed any alcohol, and at follow-up, how often they consumed alcohol in the prior 3 months. We recategorized these as never, ever but not within the prior 3 months, and within the prior 3 months, using the most current survey and all previous surveys.
The nested study activities, which were conducted within a median of 1 day (interquartile range: 0–30) after a parent study visit, included a structured interview, breath analysis, and blood and urine specimen collection. The nested study survey asked when the participants last consumed alcohol, if ever. The biomarkers examined were phosphatidylethanol (PEth) in whole blood and ethyl glucuronide/ethyl sulfate in urine (UEtG/UEtS). We previously found that PEth is 88.0% sensitive and 88.5% specific for detecting any alcohol consumption in the prior 21 days, and 76.4% sensitive and 100% specific for detecting any alcohol consumption in the prior 3 months in persons infected with HIV in Uganda.4 We conducted breath analysis to determine that breath alcohol was zero, to avoid in vitro formation of PEth.5 UEtG has been 89.3% sensitive and 98.9% specific for detecting very recent (prior 1–3 days) alcohol consumption,6 and sensitivity may be increased when combined with UEtS.7
All protocols were approved by the Institutional Review Boards of the Mbarara University of Sciences and Technology, the Uganda National Council for Science and Technology, and the University of California, San Francisco. All participants gave informed consent for the parent study and, subsequently, for the nested study.
In the parent study, 20% of participants reported drinking any alcohol in the prior 3 months, 39% reported drinking alcohol more than 3 months prior, and 41% reported never drinking alcohol (Table 1). In the nested study that included concurrent biomarkers, there was a 2-fold increase in reporting any alcohol consumption in the prior 3 months, from 20% to 41% (McNemar P < 0.01). The results were similar when the sample was limited to those who completed both the parent and nested study interviews on the same day (n = 30) and when stratified by viral suppression status (data not shown). We were unable to determine whether self-report validity varied by religion; 10 participants were of religions that prohibit alcohol (Moslems and “Saved” Christians). Self-report differed by sex (Table 1), with self-reported 3-month alcohol consumption increasing from 6% to 34% among the women (P < 0.01) and from 35% to 48% among the men (P = 0.13).
We examined the associations between the biomarkers and self-reported recent alcohol consumption. Few, 6 in the parent study and 10 in the nested study, reported alcohol consumption in the prior 3 days; therefore, we do not report UEtG/UEtS results. PEth was detectable (>8 ng/mL) for 83% (10/12) and 80% (20/25) of those reporting any alcohol consumption in the prior three months in the parent and nested studies, respectively. In addition, PEth was detectable for 27% (13/49) and 8% (3/36) of those denying prior 3 month alcohol consumption in the parent and nested studies, respectively. Furthermore, PEth was detectable in 10 of the 13 (77%) persons who denied prior 3-month alcohol consumption in the parent study but who later reported drinking in the nested study, corroborating evidence of under-report in the parent study.
Under-report of alcohol consumption may reflect fear that ART will be denied if alcohol consumption is admitted8 or may reflect social desirability bias. The greater under-report by women is consistent with increased stigma for drinking incurred by women.9 Alternatively, repeated alcohol consumption questions may increase recall. However, there is a trend in the parent cohort toward decreasing self-reported alcohol consumption over time rather than the opposite (data not shown).
Biomarker procedures discussed during the informed consent process just before the nested study survey may have had an impact on self-report. Biological measures have been used to improve the accuracy of self-report previously10–14 with mixed results. To our knowledge, this is the first study in SSA to show that self-reported alcohol consumption on a survey is increased when biomarker measurement of alcohol exposure is conducted concurrently.
There are important implications of under-reported alcohol consumption for persons infected with HIV. On the individual level, nondisclosure of alcohol use prevents health care providers from providing interventions to protect patients from the direct negative effects of alcohol or indirect effects on HIV treatment adherence and treatment outcomes. On the population level, under-reported alcohol consumption hampers research of the public health consequences of alcohol.
More work is needed to determine the context in which alcohol consumption is under-reported and to develop methods to improve assessment of alcohol consumption. Research is needed to determine the number and type of biological specimens needed to gain valid self-report; it is possible that a minimally invasive and inexpensive measure such as a breath test may suffice.
1. Hahn JA, Woolf-King SE, Muyindike W. Adding fuel to the fire: alcohol's effect on the HIV epidemic in Sub-Saharan Africa. Curr HIV/AIDS Rep. 2011;8:172–180.
2. Parry C, Rehm J, Poznyak V, et al.. Alcohol and infectious diseases: an overlooked causal linkage? Addiction 2009;104:331–332.
3. UNAIDS. 2010 Report on the Global AIDS Epidemic. UNAIDS/10.11E JC1958E, 2010.
4. Hahn JA, Dobkin LM, Mayanja B, et al.. Phosphatidylethanol (PEth) as a biomarker of alcohol consumption in HIV-Positive patients in sub-Saharan Africa. Alcohol Clin Exp Res. 2012;36:854–862.
5. Aradottir S, Seidl S, Wurst FM, et al.. Phosphatidylethanol in human organs and blood: a study on autopsy material and influences by storage conditions. Alcohol Clin Exp Res. 2004;28:1718–1723.
6. Staufer K, Andresen H, Vettorazzi E, et al.. Urinary ethyl glucuronide as a novel screening tool in patients pre- and post-liver transplantation improves detection of alcohol consumption. Hepatology 2011;54:1640–1649.
7. Wurst FM, Dresen S, Allen JP, et al.. Ethyl sulfate: A direct ethanol metabolite reflecting recent alcohol consumption. Addiction; 2006:204–211.
8. Morris CN, Levine B, Goodridge G, et al.. Three-country assessment of alcohol-HIV related policy and programmematic responses in Africa. Afr J Drug Alcohol Stud. 2006;5:170–184.
9. Wilsnack RW, Wilsnack SC, Obot IS. Why study gender, alcohol and culture? In: Obot IS, Room R, eds. Alcohol, Gender and Drinking Problems: Perspectives From Low and Middle Income Countries. Geneva, Switzerland: World Health Organization; 2005:1–23.
10. Roese NJ, Jamieson DW. Twenty years of bogus pipeline research: a critical review and meta-analysis. Psychol Bull. 1993;114:363–375.
11. Wagenaar AC, Komro KA, McGovern P, et al.. Effects of a saliva test pipeline procedure on adolescent self-reported alcohol use. Addiction 1993;88:199–208.
12. Campanelli PC, Dielman TE, Shope JT. Validity of adolescents' self-reports of alcohol use and misuse using a bogus pipeline procedure. Adolescence. 1987;22:7–22.
13. Werch CE, Gorman DR, Marty PJ, et al.. Effects of the bogus-pipeline on enhancing validity of self-reported adolescent drug use measures. J Sch Health 1987;57:232–236.
14. Werch CE, Lundstrum RH, Moore A. Bogus-pipeline effects on self-reported college student drug use, problems, and attitudes. Int J Addict. 1989;24:1003–1010.