Stable bone density in HAART-treated individuals with HIV: a meta-analysis.

CONTEXT Longitudinal studies of bone mineral density (BMD) in HIV have reported conflicting results. OBJECTIVE We investigated whether temporal changes in BMD differ by highly active antiretroviral therapy (HAART) status at baseline. DATA SOURCES Data sources included MEDLINE, EMBASE, and the Web of Science for English language studies (1966 to September 2010) and conference abstracts (1997-2010). STUDY SELECTION Longitudinal studies reporting BMD at least 48 wk apart in adult patients with HIV with a comparable uninfected control group were eligible. Uncontrolled studies were included in secondary analyses. DATA EXTRACTION Data were independently extracted by two researchers. DATA SYNTHESIS Data were pooled using random-effects models. In the primary analysis of six controlled studies (follow-up 1.5-2.7 yr), there were no significant differences in the percent change from baseline in BMD at the total hip or femoral neck between HIV cohorts and controls and a decrease of 0.6% (95% confidence interval = -1.1 to -0.1) at the spine in the HIV cohorts. In the secondary analysis of 37 studies (31 uncontrolled, six controlled), cohorts treated with HAART at baseline had stable or slight increases in BMD at 1 yr, stable or slight decreases in BMD at 2 yr, and stable BMD at 2.5 yr or later. In cohorts that were HAART-naive/untreated at baseline, there was accelerated loss of BMD at all time points, and the annualized rate of BMD change was greatest at 1 yr, but thereafter decreased. CONCLUSION BMD is stable in HIV cohorts established on HAART, whereas cohorts initiating HAART have short-term accelerated BMD loss followed by a longer period of BMD stability/increases. Routine monitoring of BMD in many HAART-treated patients may not be necessary.