Table 4.
Summary of studies assessing bone mineral density (BMD) and osteoporosis in women with HIV and menopause
| Authors | Study design | Location | Population | Findings |
|---|---|---|---|---|
| Yin et al144 | Cross-sectional | USA | 31 HIV-positive postmenopausal women compared to 186 historical-matched (age, ethnicity) HIV-negative controls | Mean BMD lower in HIV-positive group vs controls |
| Prevalence of osteoporosis 42% (vs 23%) at lumbar spine (P=0.003) | ||||
| Prevalence of osteoporosis 10% (vs 1%) at total hip (P=0.003) | ||||
| Correlates of low BMD were: | ||||
| Years since menopause (P=0.02) | ||||
| Lowest weight (P=0.001) | ||||
| No association with duration of HIV infection, ART, CD4 count | ||||
| Arnsten et al133 | Cross-sectional | USA | 263 HIV-positive and 232 HIV-negative women ≥40 years of age | HIV-positive women had lower BMD at femoral neck (P=0.001) and lumbar spine (P=0.04) vs HIV-negative women |
| HIV was an independent predictor of low BMD (especially in non-Black women) | ||||
| CD4 count, ART and use of PIs not associated with low BMD | ||||
| Anastos et al155 | Cross-sectional | USA | 274 HIV-positive and 152 HIV-negative women | Prevalence of low BMD was higher in ART-naïve HIV-positive women compared to HIV-negative women (OR 4.36), indicating an independent association with HIV infection |
| Prevalence of low BMD higher in women receiving PI-therapy (OR 3.72) compared to HIV-negative women | ||||
| PI-based ART associated with lower BMD than non-PI-ART (P=0.014) | ||||
| Longer lopinavir use correlated with lower BMD (P=0.006) and longer efavirenz use associated with higher BMD (P=0.004) | ||||
| Yin et al132 | Cross-sectional analysis of data from a prospective cohort study | USA | 92 HIV-positive and 95 HIV-negative women ≥40 years of age who were postmenopausal and not on hormone therapy | HIV-positive women were more likely to have low BMD at lumbar spine, femoral neck and total hip |
| BMD was 5.9% lower in HIV-positive vs HIV-negative women at the total hip | ||||
| HIV was an independent predictor of BMD at the lumbar spine and total hip | ||||
| No difference between HIV-positive and HIV-negative for fragility fractures | ||||
| No difference by ART status | ||||
| Yin et al140 | Cross-sectional analysis of data from a prospective cohort study | USA | 92 HIV-positive and 95 HIV-negative women ≥40 years of age who were postmenopausal and not on hormone therapy | HIV-positive women had increased bone turnover markers compared to HIV-negative women |
| No difference between: | ||||
| ART vs no ART | ||||
| Ritonavir vs no ritonavir in ART regimen | ||||
| Greater induction of peripheral blood mononuclear cells into osteoclast-like cells when exposed to autologous HIV-positive serum (vs HIV-negative) and serum containing ritonavir (vs non-ritonavir-based ART) | ||||
| Pinto Neto et al141 | Cross-sectional | Brazil | 300 HIV-positive patients with median age 46 years (57.7% male and 42.3% female) | Low BMD in 54.7% of patients (95% CI 49.1%–60.3%) |
| Independent predictors of low BMD were: | ||||
| BMI <25 kg/m2 (OR 2.9) | ||||
| Menopause (OR 13.4) | ||||
| Undetectable viral load (OR 7.9) | ||||
| Zidovudine use (OR 0.2) and nevirapine use (OR 0.1) were protective against low BMD | ||||
| Sharma et al145 | Longitudinal | USA | 245 HIV-positive and 219 HIV-negative women | HIV-positive women had lower baseline BMD at femoral neck (P=0.02) and total hip (P<0.01) than HIV-negative women, but there was no difference in prevalence of osteoporosis between HIV-positive and HIV-negative women |
| No difference between HIV-positive and HIV-negative women in rate of BMD decline over time | ||||
| Also no difference in BMD decline between: | ||||
| Tenofovir vs no tenofovir use | ||||
| PI vs no-PI use | ||||
| In multivariate analysis, menopause and chronic HCV infection associated with greater decline in BMD | ||||
| Gomes et al120 | Cross-sectional | Brazil | 273 HIV-positive women aged 40–60 years; 206/273 answered questions related to study | 33.5% had low BMD at lumbar spine and 33.1% at femoral neck |
| Low BMD at lumbar spine and femoral neck associated with: | ||||
| Age >50 years (P<0.01 at LS and FN) | ||||
| Menopause (P<0.01 at LS and FN) | ||||
| Increased FSH >40 mIU/mL (P<0.001 at LS and P<0.01 at FN) | ||||
| Menopause associated with 23-fold increased risk of low BMD at lumbar spine (P<0.01) and 57-fold increased risk of low BMD at femoral neck (P<0.01) | ||||
| No association between low BMD and smoking, alcohol, nadir CD4 count, viral load, tenofovir use, protease inhibitors, or duration of HIV infection | ||||
| Dravid et al147 | Cross-sectional | India | 536 HIV-positive patients (66% men and 34% women) with median age 42 years | In ART-naïve patients: |
| 67% had low BMD and 29.6% had osteoporosis | ||||
| In ART-experienced patients: | ||||
| 80.4% had low BMD and 36.6% had osteoporosis | ||||
| Low BMD associated with: | ||||
| Increased age (P<0.001) | ||||
| Reduced BMI (P<0.001) | ||||
| Smoking (P=0.05) | ||||
| Menopause (P=0.03) | ||||
| Prior et al146 | Case-control | Canada | 138 HIV-positive women and 402 HIV-negative controls matched for age and geographic region | HIV-positive women more likely to have other osteoporosis risk factors such as smoking, menstrual irregularity, weight loss, glucocorticoid use |
| No difference in BMD between HIV-positive and HIV-negative women | ||||
| HIV-positive cases had higher rates of lifetime fragility fractures compared to HIV-negative controls (26.1% vs 17.7%, OR 1.7); however, HIV status itself not associated with fracture in multivariable analysis when other osteoporosis risk factors taken into account | ||||
| Yin et al149 | Prospective cohort study – Women’s Interagency Health Study (WIHS) | USA | 1,728 HIV-positive and 663 HIV-negative women followed for median 5.4 years | One-third of new fractures were fragility fractures |
| No difference between HIV-positive and HIV-negative women in history of self-reported fracture | ||||
| In multivariate analysis, new fracture was associated with older age, white race, HCV infection, and increased creatinine, but not HIV status | ||||
| In HIV-positive patients, older age, white race, smoking, and history of ADI were associated with new fracture, while NNRTIs were slightly protective (OR 0.92); there was no association with ART, CD4 count or cumulative tenofovir exposure |
Abbreviations: ART, antiretroviral therapy; BMI, body mass index; PI, protease inhibitor; HCV, hepatitis C virus; IDU, injection drug use; FSH, follicle-stimulating hormone; OR, odds ratio; BMD, bone mineral density; CI, confidence interval; LS, lumbar spine; FN, femoral neck; ADI, AIDS-defining illness; NNRTI, non-nucleoside reverse transcriptase inhibitor.