Secondary Logo

Journal Logo

Clinical Science

Alcohol Consumption and CD4 T-Cell Count Response Among Persons Initiating Antiretroviral Therapy

Kowalski, Stefan*; Colantuoni, Elizabeth PhD; Lau, Bryan PhD, MHS; Keruly, Jeanne NP, MHS§; McCaul, Mary E. PhD; Hutton, Heidi E. PhD; Moore, Richard D. MD, MHS§; Chander, Geetanjali MD, MPH§

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: December 1st, 2012 - Volume 61 - Issue 4 - p 455-461
doi: 10.1097/QAI.0b013e3182712d39



Alcohol use is prevalent among HIV-infected individuals1,2 and is associated with decreased adherence to combination antiretroviral therapy (ART),3,4 decreased viral suppression on ART,5 and increased mortality.6 In addition, chronic alcohol use and HIV infection are both associated with immune suppression.7 HIV infection results in CD4 T-lymphocyte depletion, whereas heavy alcohol use is associated with defects in both cell-mediated and humoral immunity.7,8

Given that both alcohol use and HIV suppress the immune system, investigators have sought to determine if alcohol use furthers and hastens HIV disease progression through its effect on CD4 T-cell count (CD4). Samet et al9 investigated the association of alcohol use and CD4 T-cell count among 595 individuals with a history of alcohol problems. Among persons not on ART, heavy alcohol use, defined as >14 drinks per week or >4 drinks per occasion in men and >7 drinks per week or >3 drinks per occasion in women, was associated with a lower CD4 T-cell count compared with no alcohol use. Limiting the analysis to only those on ART, heavy alcohol use was not associated with a lower CD4 T-cell count after adjusting for medication adherence. In a study of 130 individuals with HIV and a history of alcohol and drug use who had a CD4 T-cell count > 200 cells per cubic millimeter, Baum et al10 found that 2 or more drinks per day was associated with a 2–3 times increased risk in decline of CD4 ≤ 200 cells per cubic millimeter compared with 1 or fewer drinks per day or no alcohol use. When they limited their analysis to persons not on ART, 2 or more drinks daily was associated with 7 times increased risk in decline of CD4 to < 200 cells per cubic millimeter compared to those with 1 or fewer daily drinks. Neither study stratified their analyses by sex to assess if alcohol’s effect on immune response to therapy differed between men and women.

Although it seems that there is a relationship between alcohol use and lower CD4 T-cell count among persons not receiving ART, it is less clear if (1) this association persists among persons on ART; (2) the effect of alcohol use on CD4 T-cell count response to therapy varies by whether viral suppression is achieved; (3) the effect of alcohol on immune response varies by sex, potentially important as women experience the adverse consequences of alcohol use at lower levels compared with men11; and (4) immunologic response to ART varies by the patterning of alcohol use. Thus, we examined the effects of the quantity and frequency of alcohol use on immunologic response to ART, whether alcohol use differentially affects CD4 T-cell count response in people who achieve viral suppression compared with those who do not, and finally, the effect of alcohol on immune response stratified by sex.


Study Design

This is a prospective cohort study of individuals enrolled in the Johns Hopkins HIV Clinical Cohort (JHHCC). The JHHCC is a longitudinal cohort of approximately 6000 HIV-infected adults receiving care in the Johns Hopkins HIV Clinic. All patients receiving care are eligible to participate. Data collected on enrollees include demographic, clinical, diagnostic, laboratory, and pharmacy data. Information from the clinical record is abstracted by trained staff. Laboratory data are obtained electronically. A description of the data collection methods for the Johns Hopkins HIV cohort has been published elsewhere.12


In July 2000, an Audio Computer-Assisted Self-interview (ACASI) collecting patient-reported outcomes was added to the data collection procedures of JHHCC. The survey takes about 15 minutes to complete and collects patient-reported outcomes including alcohol use, illicit drug use, depressive symptoms, ART use, and ART adherence over the prior 6 months. A description of the ACASI methods has been published elsewhere.13 Written informed consent is obtained from the participants. This study has been approved by the Johns Hopkins University School of Medicine’s Institutional Review Board.

Study Inclusion

We included HIV-infected individuals who participated in an ACASI within 6 months of ART initiation, who had a CD4 T-cell count measured before or on the start of ART, and who were not virologically suppressed at the time from a prior regimen. ART was defined as any regimen that included a combination of the following medications: nucleoside reverse transcriptase inhibitors, protease inhibitors (PI), a nonnucleoside reverse transcriptase inhibitor (NNRTI), an integrase strand transfer inhibitor, or a CD4 T-cell entry inhibitor.

Outcome Description

Our outcome was immunologic response to ART, defined in 2 ways: (1) CD4 T-cell count after ART initiation and (2) time from ART initiation to an increase of 100, 150, and 200 CD4 cells per cubic millimeter. Different levels of CD4 cell count response to therapy were selected to determine if there was a threshold for the effect of alcohol use on immune response.

Independent Variables

Our primary independent variables were the self-reported average number of alcoholic drinks consumed per typical drinking day (quantity of drinking, continuous) and number of days of alcohol consumption in a typical week (frequency of drinking, categorized as <1, 1–3, and >3 days) over the past 6 months. Alcohol use may vary over time. We assessed the variation in alcohol use over time within a subject by calculating the standard deviation of reported drinks per week across the subjects’ available ACASIs (all ACASIs 6 months before initiation of ART to the end of follow-up). We found that the majority of patients did not change their reported drinks per week over time (mean of SD: 1.87; median: 0), which lead to our decision to utilize alcohol use reported from the ACASI closest to ART initiation, within a 6 month window, as our main exposure variable.

Additional independent variables were chosen a priori and were either demonstrated to be associated with immunologic response to ART in previous literature or thought to be clinically important. These included age, race, HIV transmission risk factor, baseline CD4 T-cell count as a marker of disease severity, baseline HIV-1 RNA, PI use, NNRTI use, cocaine use, and depressive symptoms. HIV transmission risk was obtained from patient self-report; patients were able to report multiple risk factors. Cocaine use was ascertained using the ACASI and was defined as any self-reported cocaine use in the 6 months before the interview. We categorized depressive symptoms using the 5-item Centers for Epidemiological Studies Depression Scale (CESD-5) (low symptoms, score of 0–3; moderate symptoms, score 4–7; or high symptoms, 8–14).14 Collection of the CESD-5 did not begin until 2004; therefore, we assessed sensitivity of our finding to this variable in the subset of individuals who initiated ART after 2004. Analyses were stratified by sex as women experience biologic effects of alcohol at lower levels of use compared with men.11 A time-varying indicator variable was created for HIV RNA suppression and was used as a biomarker for adherence. Viral suppression was defined as an HIV RNA ≤ 1000 copies per milliliter. CD4 T-cell count and HIV-1 RNA measurements were not necessarily captured at concurrent time points. The HIV-1 RNA at the closest time to CD4 measurement was used to determine whether individuals had viral suppression. The median time lag in absolute value between CD4 T-cell count and HIV-1 RNA measurements after ART initiation was 0 days with a mean of 3.7 days (range, 0–180 days). We performed a sensitivity analysis limiting the range between CD4 T-cell count and HIV-1 RNA measurements to 90 days and found no difference between a maximum of a 90-day time lag and 180-day time lag.

Statistical Methods

Descriptive statistics were calculated for the outcome variables and independent variables for the entire cohort and separately by sex. Graphical displays of the outcome variables included scatter plots of CD4 T-cell count as a function of time and Kaplan–Meier curves for time from ART initiation to an increase of 100, 150, and 200 CD4 T cells per cubic millimeter.

Linear mixed effects models were used to quantify the association between CD4 T-cell count and alcohol drinking frequency (days per week, categorical) and consumption (drinks per drinking day) after adjusting for time from initiation of ART (linear spline with 4 df) and the a priori identified potential confounding variables. The models included a random intercept for the subject to account for the within-subject correlation over time in CD4 T-cell count and a random slope for time. Sex, viral suppression, and alcohol drinking frequency were treated as effect modifiers so that the models estimated the adjusted linear association between CD4 T-cell count and alcohol consumption separately for men and women, viral suppression, and alcohol drinking frequency. Interaction between this variable and the drinking exposure variables was examined to estimate a separate drinking dose effect with and without HIV RNA suppression. Likelihood ratio tests were used to test if alcohol drinking frequency modified the CD4 T-cell count and alcohol consumption association. Standard linear regression diagnostic procedures including residual plots and normal probability plots were used to confirm the validity of linearity and residual distribution assumptions.

Cox proportional hazards regression models were used to estimate the relative hazard of an increase in the CD4 T-cell count of 100, 150, and 200 CD4 cells per cubic millimeter per additional alcoholic drink per drinking day. The relative hazards were estimated separately by sex and alcohol drinking frequency after adjusting for the potential confounding variables. Separate baseline hazard functions were estimated for days with and without viral suppression. An interaction between alcohol-related covariates and viral suppression strata was assessed but found to be nonsignificant (P > 0.05) and thus not included in the final model. Likelihood ratio tests were used to determine if alcohol drinking frequency modified the relationship between an increase in the CD4 T-cell count of 100, 150, or 200 CD4 cells per cubic millimeter and alcohol consumption. Martingale and Schoenfeld residual plots were used to visually inspect the linearity and proportional hazards assumptions within the model. Models were adjusted for potential confounders, including age, race, HIV transmission risk factor, baseline CD4 T-cell count, baseline HIV-1 RNA, PI use, NNRTI use, cocaine use, and depressive symptoms.

Because we were interested in determining whether the effect of alcohol on CD4 T-cell response was driven by daily quantity, weekly frequency, or a combination of both, we did not initially create a summary measure of alcohol use. However, to allow us to compare our results with prior studies examining alcohol use and CD4 T-cell count, we created a summary measure for alcohol use using the same categories used by Samet and colleagues,9 following guidelines provided by the National Institute on Alcohol Abuse and Alcoholism: Heavy alcohol use was defined as >14 drinks per week or >4 drinks per occasion in men and >7 drinks per week in women and >3 drinks per occasion in women, and moderate alcohol use was any use below these levels.15 We then ran the same analyses outlined above.

Missing Data

The drinks per drinking day and frequency variables were incomplete for 4% (n = 67) of the subjects. The missing values were imputed by examining the 2 closest ACASIs (on either side of the regimen start date) and averaging the values for the variable(s) with missing data. If the 2 closest ACASIs contained only a single nonmissing value for the missing variable, then that value was used to impute the missing value. If both values were missing for the missing variable, then the patient was removed from the study (n = 11, <1%).


The initial population included 1695 subjects with ACASIs from the JHHCC who initiated ART between January 16, 2000, and July 15, 2008. Twenty-three individuals were excluded due to a lack of initial CD4 T-cell count (n = 12) or consumption/frequency of alcohol use (n = 11), and 22 individuals were excluded due to a lack of CD4 T-cell counts and HIV-1 RNA measurements after ART initiation. An additional 543 patients with suppressed viral loads at the time of ART initiation were excluded, leaving 1107 subjects in the study population.

Table 1 provides descriptive statistics for the study population. The median age was 42 years (range, 20–77 years), 61% were male, and 85% were African American. Sixty percent had an initial CD4 T-cell count < 200 cells per cubic millimeter. Sixty percent of the subjects reported no alcohol use, and 17%, 13%, and 10% of the subjects reported <1, 1–3, and >3 days of alcohol use on average per week, respectively. Among subjects who reported consuming alcohol, the median number of drinks consumed per drinking day was 2 (interquartile range, 1–4), with the median number of drinks per drinking day increasing with increasing frequency of use.

Demographic and HIV Clinical Characteristics of 1107 Persons Initiating ART, Overall and by Sex

Sixty-two percent of viral loads over time were <1000, and 94% of which were <400. Overall, patients with suppressed viral loads had higher average CD4 T-cell count compared with nonsuppressed subjects. The mean CD4 T-cell count changed very little over time for subjects without viral load suppression.

Tables 2 and 3 summarize the linear mixed effects model for males and females, respectively. Neither in men nor in women was there a statistically significant difference in CD4 T-cell count by average drinks per drinking day at any frequency of use, irrespective of virologic suppression. Among the potential confounding variables, the estimated mean CD4 T-cell count increased/decreased as a function of the baseline CD4 T-cell count (P < 0.05) for both males and females. Wider confidence intervals (CIs) for the estimates among the females were likely driven by their smaller sample size compared with men. In addition, within models that estimated the effect of drinks per drinking day stratified only by sex and viral load suppression but adjusted for drinking frequency, we found no difference in the estimated effects of drinks per drinking day (P = 0.54 and 0.58 among males; P = 0.75 and 0.48 among females), for suppressed and nonsuppressed, respectively.

Linear Mixed Effects Model Among Males, Examining Alcohol Consumption (Drinks per Drinking Day, Overall and Within Weekly Frequency of Alcohol Use) and CD4 Response to Therapy
Linear Mixed Effects Model Among Females Examining Alcohol Consumption and CD4 Response to Therapy

Table 4 summarizes the Cox proportional hazards model estimating the relative hazard of achieving a CD4 T-cell count increase of 100, 150, and 200 cells per cubic millimeter for each additional drink per drinking day, stratified by weekly drinking frequency and sex. For example, we estimate that the hazard of an increase of 100 cells per cubic millimeter increases by roughly 1% per additional drink per drinking day among males who report <1 day of drinking per week (hazard ratio, 1.01; 95% CI: 0.96 to 1.07). Overall, we found no difference in the hazard ratio for drinks per drinking day within the categories of drinking frequency (likelihood ratio test for drinks per drinking day and drinking frequency interaction: P = 0.53, 0.81, and 0.27 among males; P = 0.33, 0.86, and 0.45 among females for time to CD4 T-cell count increase of 100, 150, and 200 cells/mm3, respectively).

Cox Proportional Hazards Model Estimating the Relative Hazard of Achieving a CD4 Cell Increase of 100, 150, and 200 Cells per Cubic Millimeter for Each Additional Drink per Drinking Day, Stratified by Weekly Frequency of Consumption and Sex

We performed a sensitivity analysis limiting our sample to only those individuals initiating ART after 2004 who also completed the CESD-5. CESD-5 score was not significant in the models, and the results were similar to results in the overall models.

We then examined the relationship between alcohol use and CD4 T-cell response to therapy using categories for weekly consumption: heavy, moderate, and none. Using linear mixed effects models and combining both men and women, there was no significant difference in CD4 cell count between heavy drinkers and abstainers (difference 1.45; 95% CI: −26.89 to 27.97) and moderate drinkers and abstainers (difference 3.93; 95% CI: −15.63 to 23.48) among those with viral suppression and those without viral suppression (heavy vs. abstinent: 0.76, 95% CI: −27.16 to 26.68; moderate vs. abstinent −1.65, 95% CI: −21.13 to 17.83). Similarly, there was no difference in time to CD4 T-cell increase of 100 (adjusted hazard ratio heavy vs. no alcohol use: 0.94, 95% CI: 0.74 to 1.20; for increase of 150 cells/mm3: adjusted hazard ratio 0.90, 95% CI: 0.69 to 1.17; for increase of 200 cells/mm3: adjusted hazard ratio 0.95, 95% CI: 0.72 to 1.25).


In this cohort of HIV-infected individuals, level of alcohol use at the time of ART initiation was not associated with CD4 cell count response to therapy irrespective of whether viral suppression was achieved and irrespective of sex. Among those who suppressed their viral load, immunologic response to therapy was robust and associated significantly with CD4 cell count and viral load at the time of ART initiation. Similarly, among those who did not suppress their viral load, there was no differential increase or decline in CD4 cell count by either daily quantity or weekly frequency of alcohol use. These results demonstrate that the benefit of viral suppression on ART outweighs the potential detrimental effects of alcohol use on immunologic response to therapy in both men and women.

Our results support and extend findings from 2 earlier studies in the ART era examining associations between alcohol use and immunologic response to therapy.9,16 Samet et al prospectively assessed CD4 T-cell count among 595 HIV-infected individuals, of whom 354 were on ART. Using linear mixed models, they found that among persons on ART, heavy alcohol use was not associated with CD4 cell count. Similarly, in a longitudinal study of 516 HIV-infected women using data from the HIV Epidemiologic Research Study cohort, alcohol use was not statistically associated with CD4 T-cell count, irrespective of ART use.16 Our results add to the current evidence that alcohol use does not significantly affect CD4 T-cell count among persons on ART, and extends this finding to individuals initiating ART who achieve viral suppression.

Our results are in contrast to a recent study by Baum et al10 who prospectively followed HIV-infected individuals with alcohol and or/illicit drug use. Among 130 individuals with a baseline CD4 > 200 cells per milliliter, they examined the association of alcohol use and time to decline of CD4 ≤ 200 cells per milliliter. Defining frequent alcohol use as ≥2 alcoholic drinks daily, and adjusting for ART use as a time-varying covariate, they found that frequent alcohol use was associated with an increased risk in decline of CD4 cell count. When they combined alcohol and cocaine use, they found that the risk of CD4 cell count decline persisted, although it was not greater than frequent alcohol use alone. There are a number of potential explanations for the difference in findings between our study and those of Baum et al. Our sample was limited to individuals initiating ART, whereas Baum et al included both individuals on and off of ART. Although they adjusted for use of ART, it may be that the increased risk in decline of CD4 cell count in their sample was driven by those not on ART. A second possible explanation is that concurrent cocaine use may be responsible for the association between alcohol use and CD4 cell count in their study. In a previous study by Baum et al,17 using the same cohort of drug users, crack cocaine use was associated with a decline of CD4 cell count to <200 cells per cubic millimeter independent of ART use and controlling for any alcohol use, suggesting a direct link between crack cocaine and HIV disease progression. In the Women’s Interagency HIV Study, both intermittent and persistent crack use was associated with greater CD4 cell count decline, controlling for ART use and adherence. Alcohol use was not significantly associated with lower CD4 cell count in Women’s Interagency HIV Study, although it was associated with increased HIV viral load.18

We did not find a difference in CD4 cell count response to ART by sex. We hypothesized that women consuming alcohol would experience decreased immune response to ART at lower levels of alcohol use compared with men; however, neither men nor women showed an effect of level of alcohol use on CD4 cell count. This is likely explained by the generally equal potency of ART between men and women.19

Although we did not find a blunted CD4 cell count response after ART initiation among persons with alcohol use, alcohol consumption does negatively affect HIV disease progression and transmission through other mechanisms: decreased medication adherence4,20–22 and increased transmission risk behaviors.23,24 Hendershot et al21 synthesized literature on the association between alcohol use and antiretroviral adherence. Across 40 studies, with 25,000 participants, any alcohol use was associated with a 40% decreased odds of adherence and at-risk drinking or an alcohol use disorder was associated with 53% lower odds of antiretroviral adherence. Alcohol use is also associated with risky sexual behaviors.24 In a meta-analysis of 27 studies, any alcohol use, problem drinking, and alcohol use in sexual contexts were all significantly associated with unprotected sex among HIV-infected individuals.24 With alcohol’s negative effect on HIV medication adherence and viral suppression and its association with increased transmission behaviors, screening for alcohol use and brief interventions encouraging reduction or abstinence in alcohol use is essential to optimize the management of HIV.

Given that individuals with alcohol use can have a robust immunologic response to ART and that prior studies have demonstrated lower CD4 cell count among persons not on ART who consume alcohol, early initiation of ART is essential to achieve optimal CD4 cell count. With more durable ART regimens, which are more forgiving to medication nonadherence, initiation of ART should be considered irrespective of alcohol use.

Our study has potential limitations. The median of drinks per drinking day among the highest frequency drinkers was 4 (interquartile range, 2–6). Thus, our sample may not have included the heaviest drinkers. As a result, we cannot generalize our findings to the subset of individuals who drink outside the range of what was reported in this study. Our generalizability is also limited to populations similar to our study sample, which includes largely urban, African American, HIV-infected individuals. Also, this is a longitudinal cohort study, and although we adjusted for potential confounders, there was likely residual confounding. In addition, alcohol use was measured by self-report, which may have led to underreporting of alcohol use. Finally, the CESD-5 was used as our measure of depressive symptoms, chosen specifically in this cohort for its brevity. However, this 5-question measure is not used routinely in research and may not have adequately identified individuals with depression.

In conclusion, among persons initiating ART, alcohol use is not associated with decreased immunologic response to ART, irrespective of viral suppression. People who drink alcohol and achieve viral suppression can reap the same immunologic benefits of those who do not drink alcohol. Although there are a number of other important medical and behavioral reasons that a patient’s alcohol consumption should be addressed, an immunologic effect on the CD4 cell count response to ART at levels of alcohol use in this study is not among them. Aggressive treatment for HIV, coupled with adherence interventions to ensure viral suppression, should be considered for people who consume alcohol within the range of use studied in this cohort.


1. Chander G, Josephs J, Fleishman JA, et al.. Alcohol use among HIV-infected persons in care: results of a multi-site survey. HIV Med. 2008;9:196–202.
2. Galvan FH, Bing EG, Fleishman JA, et al.. The prevalence of alcohol consumption and heavy drinking among people with HIV in the United States: results from the HIV Cost and Services Utilization Study. J Stud Alcohol. 2002;63:179–186.
3. Braithwaite RS, McGinnis KA, Conigliaro J, et al.. A temporal and dose-response association between alcohol consumption and medication adherence among veterans in care. Alcohol Clin Exp Res. 2005;29:1190–1197.
4. Samet JH, Horton NJ, Meli S, et al.. Alcohol consumption and antiretroviral adherence among HIV-infected persons with alcohol problems. Alcohol Clin Exp Res. 2004;28:572–577.
5. Chander G, Lau B, Moore RD. Hazardous alcohol use: a risk factor for non-adherence and lack of suppression in HIV infection. J Acquir Immune Defic Syndr. 2006;43:411–417.
6. DeLorenze GN, Weisner C, Tsai AL, et al.. Excess mortality among HIV-infected patients diagnosed with substance use dependence or abuse receiving care in a fully integrated medical care program. Alcohol Clin Exp Res. 2011;35:203–210.
7. Szabo G, Mandrekar P. A recent perspective on alcohol, immunity, and host defense. Alcohol Clin Exp Res. 2009;33:220–232.
8. Ghare S, Patil M, Hote P, et al.. Ethanol inhibits lipid raft-mediated TCR signaling and IL-2 expression: potential mechanism of alcohol-induced immune suppression. Alcohol Clin Exp Res. 2011;35:1435–1444.
9. Samet JH, Cheng DM, Libman H, et al.. Alcohol consumption and HIV disease progression. J Acquir Immune Defic Syndr. 2007;46:194–199.
10. Baum MK, Rafie C, Lai S, et al.. Alcohol use accelerates HIV disease progression. AIDS Res Hum Retroviruses. 2010;26:511–518.
11. Frezza M, di PC, Pozzato G, et al.. High blood alcohol levels in women. The role of decreased gastric alcohol dehydrogenase activity and first-pass metabolism. N Engl J Med. 1990;322:95–99.
12. Moore RD. Understanding the clinical and economic outcomes of HIV therapy: the Johns Hopkins HIV clinical practice cohort. J Acquir Immune Defic Syndr Hum Retrovirol. 1998;17(suppl 1):S38–S41.
13. Lucas GM, Gebo KA, Chaisson RE, et al.. Longitudinal assessment of the effects of drug and alcohol abuse on HIV-1 treatment outcomes in an urban clinic. AIDS. 2002;16:767–774.
14. Lewinsohn PM, Seeley JR, Roberts RE, et al.. Center for Epidemiologic Studies Depression Scale (CES-D) as a screening instrument for depression among community-residing older adults. Psychol Aging. 1997;12:277–287.
15. NIAAA Alcoholism. The Physician’s Guide to Helping Patients With Alcohol Problems. Washington, DC: National Institutes of Health; 2005.
16. Ghebremichael M, Paintsil E, Ickovics JR, et al.. Longitudinal association of alcohol use with HIV disease progression and psychological health of women with HIV. AIDS Care. 2009;21:834–841.
17. Baum MK, Rafie C, Lai S, et al.. Crack-cocaine use accelerates HIV disease progression in a cohort of HIV-positive drug users. J Acquir Immune Defic Syndr. 2009;50:93–99.
18. Cook JA, Burke-Miller JK, Cohen MH, et al.. Crack cocaine, disease progression, and mortality in a multicenter cohort of HIV-1 positive women. AIDS. 2008;22:1355–1363.
19. Nicastri E, Leone S, Angeletti C, et al.. Sex issues in HIV-1-infected persons during highly active antiretroviral therapy: a systematic review. J Antimicrob Chemother. 2007;60:724–732.
20. Hahn JA, Samet JH. Alcohol and HIV disease progression: weighing the evidence. Curr HIV/AIDS Rep. 2010;7:226–233.
21. Hendershot CS, Stoner SA, Pantalone DW, et al.. Alcohol use and antiretroviral adherence: review and meta-analysis. J Acquir Immune Defic Syndr. 2009;52:180–202.
22. Parsons JT, Rosof E, Mustanski B. The temporal relationship between alcohol consumption and HIV-medication adherence: a multilevel model of direct and moderating effects. Health Psychol. 2008;27:628–637.
23. Hutton HE, McCaul ME, Chander G, et al.. Alcohol use, anal sex, and other risky sexual behaviors among HIV-infected women and men. AIDS Behav. 2012. Epub ahead of print, doi: 10.1007/s10461-012-0191-4.
24. Shuper PA, Joharchi N, Rehm J. Alcohol as correlated of unprotected sexual behavior among people living with HIV/AIDS: review and meta-analysis. AIDS Behav. 2009;13:1021–1036.

HIV; alcohol; immune response; CD4 T-cell count; antiretroviral therapy

© 2012 Lippincott Williams & Wilkins, Inc.