Share this article on:

Low Alcohol Consumption as a Predictor of Higher CD4+ Cell Count in HIV-Treated Patients: A French Paradox or a Proxy of Healthy Behaviors? The ANRS APROCO-COPILOTE CO-08 Cohort

Carrieri, Maria Patrizia PhD*,†,‡; Protopopescu, Camelia PhD*,†,‡; Raffi, François PhD, MD§; March, Laura MSc*,†,‡; Reboud, Philippe PhD; Spire, Bruno PhD, MD*,†,‡; Leport, Catherine PhD, MD¶,#the ANRS CO8 APROCO-COPILOTE Study Group

JAIDS Journal of Acquired Immune Deficiency Syndromes: April 1st, 2014 - Volume 65 - Issue 4 - p e148–e150
doi: 10.1097/QAI.0000000000000087
Letters to the Editor

*INSERM, UMR912 (SESSTIM), Marseille, France

Aix Marseille Université, UMR_S912, IRD, Marseille, France

ORS PACA, Observatoire Régional de la Santé Provence-Alpes-Côte d'Azur, Marseille, France

§CHU Hôtel-Dieu, Nantes, France

INSERM, U897, Université Bordeaux Segalen, ISPED, Bordeaux, France

Université Paris Diderot, Sorbonne Paris Cité, UMR 738, Paris, France

#INSERM, UMR 738, Paris, France

The authors have no funding or conflicts of interest to disclose.

To the Editors:

We have read with great interest the article published by Conen et al1 using 7-year longitudinal data from the Swiss HIV cohort. The objective of this study was to explore the relationship between alcohol consumption and HIV surrogate markers, that is virological failure, change in CD4 cell count over time, and ART interruption. The authors did not find any significant association between alcohol consumption, either with virological failure or with change in CD4 cell count over time. In line with previous literature,2 they found that elevated alcohol consumption is a major correlate of antiretroviral treatment interruptions.

The role that alcohol consumption plays in HIV outcomes is controversial. Excessive consumption is detrimental to sustained adherence to treatment but may not have quite such a strong influence on health care utilization.2 It has also been shown that nonadherence and treatment interruption in individuals reporting excessive alcohol consumption may be intentional to avoid potential interactions with alcohol.3 One systematic review2 highlights that several studies have reported a relationship between elevated alcohol consumption and poor HIV outcomes in patients receiving antiretrovirals and that this relationship is mediated by poor adherence. Although alcohol consumption can predict HIV progression in patients receiving antiretrovirals by poor adherence,4–7 it is important to note that its impact on HIV outcomes remains limited in untreated HIV-infected individuals.8 One possible explanation for this is that ART may have such a major impact on virological and immunological response that it hides any potential toxic effect of excessive alcohol consumption.9 Elevated alcohol consumption is also regarded as a risk factor for comorbidities that are frequent in HIV-infected patients, such as liver injury and cancer.

However, moderate alcohol consumption has been shown to reduce the risk of myocardial infarction10 and of cardiovascular mortality in the general population.11 It also seems to protect HIV-infected patients from major coronary and other arterial disease events.12

A previous study comparing non–HIV-infected French and Irish individuals showed that the effect of alcohol consumption on cardiovascular risk depends on the individual's pattern of alcohol use, with an increased protective effect of moderate drinking of red wine being observed in the French population.13

To verify whether the results found by our Swiss colleagues were consistent with those obtained in a comparable cohort in the French context, we conducted a similar longitudinal analysis to explore the association between alcohol consumption and both virological failure and CD4+ cell count.

We used an 11-year follow-up of the French national APROCO-COPILOTE (ANRS CO8) cohort, which enrolled 1281 HIV-1 positive individuals who started the first generation of protease inhibitor containing ART in 1997–1999. Follow-up was characterized by simultaneous collection of medical and socio-behavioral data using a self-administered questionnaire.

We selected 1108 individuals who matched the following criteria: (1) they provided complete data in their self-administered questionnaire at enrollment and (2) they had at least 1 assessment of alcohol consumption during follow-up.12

We used the same definition of virological failure as in the Swiss study,1 that is either failure to achieve virological suppression (2 consecutive viral load measurements >400 copies/mm3 after 6 months or >50 copies/mm3 after 12 months) or viral rebound after viral suppression (2 consecutive detectable viral load measurements after 2 consecutive undetectable viral load measurements). We also used CD4 cell count as the outcome, measured on a logarithmic scale for reasons of comparability with the Swiss study. Daily alcohol consumption was defined at any visit as follows: abstinent: reference category; low: <10 g, moderate: 10–30 g for women and 10–40 g for men; and elevated: >30 g for women and >40 for men.

The relationship between daily alcohol consumption and virological failure was assessed using a Cox model, whereas a linear mixed model was used to assess the association between alcohol consumption and CD4 cell count over time. We also adjusted for a similar pattern of correlates used in multivariate analysis in the Swiss study. The studied population was predominantly men (78%), with a median age at enrollment of 36 years [interquartile range (IQR), 32–42] and had been diagnosed with HIV for a median of 3.8 years. Seventeen percent of them had become infected through injecting drug use and 41% through homosexual transmission. At enrollment, approximately half of the patients (45%) were treatment naive, 21% had experienced progression to AIDS, the prevalence of HCV coinfection was 22%, and the median (IQR) of the logarithm of viral load and CD4 cell count were 4.5 (3.7–5.2) log (copies/mL) and 280 (130–424)/mm3, respectively. At enrollment, the prevalence of alcohol consumption using the categories indicated above was as follows: 19%—abstinent, 54%—low consumption, 21%—moderate consumption, and 6%—elevated alcohol consumption. The prevalence of moderate/elevated alcohol consumption was higher than that observed in the Swiss cohort reflecting a similar difference when comparing alcohol consumption in the general population in France and Switzerland (

The 1108 patients had a median (IQR) follow-up of 6.9 (1.8–11.1) years. Among them, 555 cases of virological failure were identified during the M6-M144 follow-up period, representing 4218 person-years of follow-up and yielding an overall virological failure incidence of 13.1 per 100 person-years of follow-up (95% confidence interval: 12.1 to 14.3).

In line with the results of the Swiss study, after adjustment for a similar pattern of predictors (irrespective of whether these predictors were significant or not in the final model), no significant association was found between alcohol consumption and virological failure. This was true after adjustment for significant correlates only, and also after adjustment for nonadherence to ARV. By contrast, a significant association was found between alcohol consumption and CD4 cell count; compared with abstinent patients, those reporting low alcohol consumption were more likely to have significantly higher CD4 cell count [coefficient (95% confidence interval): 0.06 (0.02 to 0.09), P = 0.006], whereas those drinking more than 10 g/day had CD4 cell counts not significantly different from those of abstinent patients, after adjustment for a pattern of correlates similar to that used in the Swiss study. This association was confirmed even when using as the reference group only those patients who remained abstinent over the whole follow-up, to control for possible misclassification of ex-users as alcohol abstainers. When removing variables that did not significantly contribute to the multivariate model, low alcohol consumption was still significantly associated with CD4 outcome. We did not find any significant effect of age on CD4 cell count, probably because of different characteristics between the 2 cohorts, related to the study period analyzed (from 1997 onward for the French cohort and from 2005 onward for the Swiss cohort).

The association between low alcohol consumption and higher CD4 cell count levels is rather intriguing. It is also possible that this association was not found in the Swiss study because the “low consumption” and “moderate consumption” categories were collapsed together. However, we cannot assess the possible hypothesis of a French paradox, that is a true “protective” effect of low alcohol consumption arising either from the pattern of alcohol use in France or from the fact that low alcohol consumption may be a proxy of low food intake14 as the self-administered questionnaire did not assess such specificities. It is also likely that this association is significant because low alcohol consumption can be regarded more as a proxy of healthier behaviors (high-quality food, exercise, and others) and high social status, which are not completely captured by the variables used for adjustment in the model.

The strength of our analysis is the longer duration of follow-up with respect to that in the Swiss study [with a median (IQR) duration of follow-up of 6.9 (1.8–11.1) years versus 30 (14–51) months for the Swiss study]. However, our study has the same limitations as the Swiss one that is missing data were probably not completely at random, as individuals who did not answer the alcohol questionnaire or missed the follow-up visits were, perhaps, more likely to belong to the at-risk population. Nevertheless, the mode used to gather data from the French cohort (self-administered questionnaires instead of face-to-face interviews) may tend to minimize both social desirability bias and underestimation of alcohol use consumption. It is therefore possible that although the prevalence of alcohol consumption was higher in the French cohort, it was underestimated less than in the Swiss study group. To what extent this difference may affect the consistency of the results is difficult to assess, but we believe that individuals with low alcohol consumption are adequately represented in this analysis.

Now that people living with HIV are increasingly similar to the general population in terms of life expectancy and quality of life, there is a need to promote research on the influence that moderate alcohol consumption, various healthy behaviors, and eating habits may all have on HIV progression, immunological recovery, and survival.

Back to Top | Article Outline


1. Conen A, Wang Q, Glass TR, et al.. Association of alcohol consumption and HIV surrogate markers in participants of the Swiss HIV Cohort Study. J Acquir Immune Defic Syndr. 2013;64:472–478.
2. Azar MM, Springer SA, Meyer JP, et al.. A systematic review of the impact of alcohol use disorders on HIV treatment outcomes, adherence to antiretroviral therapy and health care utilization. Drug Alcohol Depend. 2010;112:178–193.
3. Shuper PA, Joharchi N, Irving H, et al.. Alcohol as a correlate of unprotected sexual behavior among people living with HIV/AIDS: review and meta-analysis. AIDS Behav. 2009;13:1021–1036.
4. Braithwaite RS, Conigliaro J, Roberts MS, et al.. Estimating the impact of alcohol consumption on survival for HIV+ individuals. AIDS Care. 2007;19:459–466.
5. Galai N, Vlahov D, Bareta JC, et al.. Prognostic factors for survival differ according to CD4+ cell count among HIV-infected injection drug users: pre-HAART and HAART eras. J Acquir Immune Defic Syndr. 2005;38:74–81.
6. Hooshyar D, Hanson DL, Wolfe M, et al.. Trends in perimortal conditions and mortality rates among HIV-infected patients. AIDS. 2007;21:2093–2100.
7. Samet JH, Walley AY, Bridden C. Illicit drugs, alcohol, and addiction in human immunodeficiency virus. Panminerva Med. 2007;49:67–77.
8. Samet JH, Horton NJ, Traphagen ET, et al.. Alcohol consumption and HIV disease progression: are they related? Alcohol Clin Exp Res. 2003;27:862–867.
9. Kowalski S, Colantuoni E, Lau B, et al.. Alcohol consumption and CD4 T-cell count response among persons initiating antiretroviral therapy. J Acquir Immune Defic Syndr. 2012;61:455–461.
10. Mukamal KJ, Conigrave KM, Mittleman MA, et al.. Roles of drinking pattern and type of alcohol consumed in coronary heart disease in men. N Engl J Med. 2003;348:109–118.
11. Mukamal KJ, Chen CM, Rao SR, et al.. Alcohol consumption and cardiovascular mortality among U.S. adults, 1987 to 2002. J Am Coll Cardiol. 2010;55:1328–1335.
12. Carrieri MP, Protopopescu C, Le Moing V, et al.. Impact of immunodepression and moderate alcohol consumption on coronary and other arterial disease events in an 11-year cohort of HIV-infected patients on antiretroviral therapy. BMJ Open. 2012;2:e001155.
13. Marques-Vidal P, Montaye M, Arveiler D, et al.. Alcohol consumption and cardiovascular disease: differential effects in France and Northern Ireland. The PRIME study. Eur J Cardiovasc Prev Rehabil. 2004;11:336–343.
14. Rozin P, Kabnick K, Pete E, et al.. The ecology of eating: smaller portion sizes in France than in the United States help explain the French paradox. Psychol Sci. 2003;14:450–454.
© 2014 by Lippincott Williams & Wilkins