Background
People with HIV (PWH) have higher osteoporosis prevalence vs people without HIV. Bone loss is multifactorial and accelerated by chronic inflammation, ART, particularly TDF, and ageing. Bisphosphonates (BP) are recommended for osteoporosis treatment, but evidence on the extent to which TDF discontinuation and/or BP initiation improve bone health in PWH is limited.
Methods
In the cohort who had dual energy X-ray absorptiometry (DXA) at Chelsea and Westminster Hospital, London, we identified PWH diagnosed with osteoporosis (T-score ≤-2.5) who had at least one follow-up DXA between 2009-2024. The impact of TDF exposure (never, continued, stopped at DXA, past) and BP initiation on the annualized rate of change in femoral and spinal T-scores (delta standard deviation [SD] units/year) was assessed by mixed-effects models adjusted for demographics, HIV and lifestyle covariates, with interaction terms for drug effects. P values were adjusted for False Discovery Rate (FDR).
Results
907 PWH had DXAs, 107 had osteoporosis (11.8%), 86% were male, 94% virally suppressed, 55% on TDF (25% with PI/b). Mean age was 56 years, and median CD4+ count 659 cells/µL. At follow-up DXA (mean duration 4 years), 40 people discontinued TDF (37.4%) and 44 started BP (41.1%). Femoral T-scores improved in all groups, with the largest gains in those who discontinued TDF, either with (+0.20 SD/year, FDR<0.001) or without BP (+0.21 SD/year, FDR=0.029). A similar trajectory was observed in PWH who continued TDF (+0.19 SD/year, FDR<0.001 with BP vs +0.16 SD/year, FDR<0.001 without BP). Point estimates suggested slightly higher gains with BP in those continuing TDF. Spinal T-scores improved in most groups, except when not starting BP, and never exposed or previously exposed to TDF. Significant gains were seen if continuing (+0.20 SD/year, FDR=<0.001 with BP vs +0.19 SD/year, FDR=0.024 without BP) or discontinuing TDF (+0.14 SD/year, FDR=0.045 with BP vs +0.10 SD/year, FDR=0.027 without BP). BP use was associated with more positive slopes, but the added effect on TDF discontinuation was modest.
Conclusions
In this retrospective cohort of PWH, discontinuing TDF was associated with improvements in bone DXA T-scores (especially femoral T-scores), while BP initiation added modest gains. The broad upward trend across most groups likely reflects concurrent, unmeasured interventions such as lifestyle, vitamin D supplementation, and age-related plateauing of bone loss, underscoring the need for prospective, controlled studies.