Reduced bone density accrual among peripubertal boys with HIV in Zimbabwe.

Objective: To investigate bone density accrual over 1 year among peripubertal children with HIV (CWH) compared to children without infection (CWOH); and risk factors associated with bone density accrual among CWH.

Design: A prospective cohort study in urban Zimbabwe.

Methods: CWH on antiretroviral therapy aged 8-16 years, and CWOH, frequency-matched by age were recruited in Zimbabwe. Z -scores for height-adjusted total-body less-head bone mineral content for lean mass (TBLH-BMC LBM ) and size-adjusted lumbar spine bone mineral apparent density (LS-BMAD) were calculated from dual X-ray absorptiometry (DXA) scan measurements. Linear regression compared bone density accrual by HIV status.

Results: Of 609 participants, 492 (80.7%) completed a follow-up visit (50.2% boys, 49.6% CWH). Mean baseline age was 12.5 years. More girl CWH than CWOH were in Tanner stages I/II at baseline. Bone density accrual (Δ) adjusted for age, Tanner stage and baseline DXA Z -score was less in boy CWH than boy CWOH {adjusted mean (95% confidence interval (CI)] ΔLS-BMAD Z -score -0.14 (-0.25 to -0.02) vs. 0.01 (-0.09 to 0.12), P  = 0.020, and ΔTBLH-BMC LBMZ -score -0.19 (-0.33 to -0.04) vs. 0.07 (-0.07 to 0.20), P  = 0.015}, but similar in girls with and without HIV [ΔLS-BMAD Z -score 0.05 (-0.07 to 0.17) vs. -0.01 (-0.09 to 0.07), P  = 0.416, and ΔTBLH-BMC LBMZ -score 0.08 (-0.07 to 0.22) vs. -0.03 (-0.12 to 0.07), P  = 0.295]. Viral load greater than 1000 copies/ml and tenofovir disoproxil fumarate use were associated with less gain in LS-BMAD Z -score among boys, whereas Tanner stage IV and V were associated with greater gains in LS-BMAD and TBLH-BMC LBMZ -scores among CWH.

Conclusion: Among boys only, CWH had impaired bone accrual, associated with high viral load and tenofovir use. Bone density gains were greater in later puberty among CWH suggesting potential to correct deficits.