A Nested Case-Control Study of Stimulant and Amyl Nitrite (“poppers”) Use Among Men Who Have Sex with Men Pre and Post HIV Seroconversion

Background:  Numerous studies have shown men who have sex with men (MSM) disproportionately use stimulants (e.g., cocaine, methamphetamine) and amyl nitrites (i.e., “poppers”). These drugs, either singly or in combination, have been strongly associated with increased risk for HIV infection, as use often occurs in the context of unprotected and more frequent sex with higher numbers of partners. Stimulant use is particularly elevated among HIV-infected MSM. However, these findings are based on cross-sectional studies leaving open to question whether the high rates of use among HIV-positive MSM reflect a pattern that precedes and contributes to their HIV-infection risk, or whether stimulant use increases post-seroconversion owing possibly to depression or decreased concern about sexual risk. This study analyzed longitudinal data collected from MSM, some of whom seroconverted over the course of the study, to examine stimulant/amyl nitrite use pre- and post-seroconversion to better understand patterns of use over time.

Methods: This study was a nested case-control, secondary analysis of data collected for the Multicenter AIDS Cohort Study (MACS), a bi-annual, ongoing longitudinal panel study of the natural history of AIDS begun in 1984 (cohort 1) with recruitment of additional cohorts in 1987-1991 and 2001-2003 (total N = 6,972). Participants that seroconverted at any time between enrollment and mid-2009 (N = 631) were matched on demographics, recruitment cohort, and available interviews using iterative propensity score matching. This yielded 398 seroconverters (SC) matched with 398 persistently seropositive (PSP) and 398 persistently seronegative (PSN) participants. Interview times for matched participants were centered two years prior and two years post the seroconversion interview of the SC participants for a maximum of 9 time points per participant. Binary logistic and negative binomial generalized estimating equations with unstructured correlations estimated the effects of cohort, serostatus, and time on past 6-month: 1) use of stimulant drugs/poppers (yes/no); and 2) number of male sex partners. Covariates included alcohol use, other drug use, and demographics.

Results:  Relative to PSN participants, SC participants had the highest odds of stimulant drug use (OR = 3.27, p < .001) and rates of male sex partners (IRR = 2.83, p < .001). These differences were especially pronounced pre-seroconversion. Drug use and number of sex partners declined over time for all groups but most precipitously for SC participants. Nevertheless, their rates of use and number of sexual partners as well as those of the PSP group remained elevated in the two years post-seroconversion relative to PSN participants.

Implications: We did not find evidence for increases in either stimulant use or sex partners post-seroconversion. Instead, the higher rates of both for MSM who seroconverted or were seropositive appear to reflect long-standing patterns rather than behavioral change in response to infection. In fact, post-seroconversion was associated with decreased albeit still elevated stimulant use and sex partners. The temporal pattern of HIV-risk, even in the face of HIV-infection, provides new insight into the persistent challenge stimulant use poses to HIV long-term prevention efforts for ‘high-risk’ MSM, particularly with regard to issues like medication adherence and secondary HIV transmission.