Background:  Low bone density is prevalent among HIV⁺ women. Data on short term bone loss, fracture risk and effect of antiretrovirals are limited. 

Methods:  In 114 HIV⁺ and 74 HIV^– premenopausal women enrolled in the Women’s Interagency HIV Study (WIHS), bone density was measured by dual X-ray aborptiometry at the femoral neck and lumbar spine at baseline and at 2 years. Baseline serum levels of bone formation (osteocalcin and bone alkaline phosphatase) and bone resorption (N-telopeptide) markers, and pro-resorptive cytokines (interleukin-6 [IL-6], tumor necrosis factor –alpha [TNF-α], and receptor activator of NF-B ligand [RANKL]) were determined.

Results:  HIV⁺ women were older than HIV^– (41±5 vs 37±7 years, p = 0.0002) and had lower current body mass index (29±6 vs 31±6 kg/m², p = 0.04), but were similar with regard to race (81 vs 77% non-white, p = 0.81), and prevalence of smoking, alcohol use, heroin use, diabetes, calcium and vitamin D supplementation. HIV⁺ women had lower baseline lumbar spine bone density than HIV^– (1.26±0.15 vs 1.30±0.16 g/cm², p = 0.05) and lower femoral neck bone density (1.04±0.1 vs 1.09±0.2 g/cm², p = 0.02). femoral neck bone density was lowest in HIV⁺ women on protease inhibitor (PI) -based HAART (HIV⁺PI⁺) (0.99±0.14 g/cm²). Prevalence of femoral neck low bone density (T score <–1) was 17% in HIV⁺PI⁺, 6% in HIV⁺ on non-PI-based HAART (HIV⁺PI^–) and 7% in HIV^–. Baseline serum bone turnover marker and cytokine levels were similar between HIV⁺ and HIV^–. Annual percentage change in bone density was similar in HIV⁺ and HIV^– at femoral neck (–0.88±3.3 vs –0.61±1.9%, p = 0.76) and lumbar spine (–0.96±2.1 vs –0.45±2.2%, p = 0.26), after adjusting for age and body mass index in mixed model analysis. Among HIV⁺, N-telopeptide levels were higher among HIV⁺PI⁺ in comparison to ART-naive (14.33±8.6 vs 10.09±3.8 nM bone collagen equivalents/L, p = 0.04) and bone alkaline phosphatase lower among HIV⁺PI⁺ in comparison to HIV⁺PI^– (28.72±10.4 vs 36.87±13.7 U/L, p = 0.03); however annual percentage change in bone density was similar between all treatment groups. Lastly, over a 2-year follow-up, occurrence of new self-reported fragility fractures was similar between the 2 groups (3%).

Conclusions:  In premenopausal HIV⁺ women, baseline bone density was lower than comparable HIV^– women but rates of short term bone loss at the lumbar spine and femoral neck and fragility fracture were similar. In HIV⁺ women on PI-based HAART, elevated bone resorption and depressed bone formation markers did not translate to increased bone loss.