825    

Background:  The etiology of bone loss in HIV disease is unclear. Osteopenia has been frequently observed in the era of highly active antiretroviral (HAART) use. We evaluated the association between time on specific ART and change in total bone mineral density over time in HIV-infected men and women.

Methods:  Total bone mineral density was measured at annual clinic visits, in 302 men and 101 women in the Nutrition for Healthy Living Cohort using dual energy absorptiometry (DXA) on the Hologic QDR 2000. Concurrent data was collected on medical and behavior history, CD4 count, HIV viral load, time known HIV⁺, liver function, weight change, and duration of specific ART. The median time of follow-up from first to last DXA was 3.1 years (25th 1.6, 75th 4.5). On each person, the percent change in total bone mineral density from first to last visit was divided by the total follow-up time to obtain the annualized percent change in total bone mineral density per year of follow-up. Using a repeated measures regression model, we evaluated the relationship between duration of anti-retroviral therapies and the percent change in bone mineral density between consecutive DXA exams (n = 902 intervals), adjusted for viral load, CD4, age, gender, albumin, menopause, smoking, and time between visits.

Results:  At baseline, the median age was 42 years old, median CD4 was 372 cells/mm³ and median viral load was 2.7 log₁₀ copies/mL. The median annualized change in total bone mineral density was –0.0187% per month (p < 0.0001); a loss of 0.22% over 1 year (25th –0.86%, 75th 0.41%) and a loss of 0.68% over 3 years (25th –2.6%, 75th 1.2%). Longer durations of didanosine (ddI) use (p = 0.0043), tenofovir (TDF) use (p < 0.0001) and years known HIV⁺ (p = 0.016), and occurrence of bilirubin > 2 (p = 0.01) were independently associated with greater loss of total bone mineral density over time. Longer duration of stavudine (d4T) (p = 0.016) was independently associated with less loss of total bone mineral density over time. Duration of indinavir (IDV) was only marginally associated with less loss of total bone mineral density (p = 0.08).

Conclusions: Specific ART may accelerate bone loss (ddI, TDF) or be protective (d4T, IDV). Duration of HIV and elevated bilirubin are associated with more bone loss.

Keywords: bone loss; anti-retroviral therapy; longitudinal