Whether specific drugs such as tenofovir increase the risk, reduce the risk, or have a variable effect over time remains unclear.
ART Reduces Risk
- Exposure to antiretroviral therapy overall was associated with significantly reduced risk of bone fractures (odds ratio [OR] 0.64).
- Use of nucleoside/nucleotide reverse transcriptase inhibitor (NRTI) and non-nucleoside reverse transcriptase inhibitor (NNRTI) drug classes was associated with reduced fracture risk.
- Exposure to protease inhibitors (PIs) was associated with a "null effect" that became slightly reduced among people with the longest duration of exposure (> 18 months).
- Use of darunavir (Prezista), delavirdine (Rescriptor), and saquinavir (Invirase or Fortovase) were associated with increased fracture risk over time.
- Efavirenz (Sustiva), emtricitabine (Emtriva), lamivudine (3TC; Epivir), tenofovir (Viread), and zidovudine (AZT; Retrovir) were associated with reduced fracture risk.
- Abacavir (Ziagen), didanosine (ddI; Videx), nelfinavir (Viracept), ritonavir (Norvir), and stavudine (d4T; Zerit) were initially associated with increased risk, but this became null with longer duration of use.
- Nevirapine (Viramune) was initially associated with a null risk that became reduced risk with longer use.
- Amprenavir (Agenerase), atazanavir (Reyataz), enfuvirtide (T-20; Fuzeon), fosamprenavir (Lexiva), indinavir (Crixivan), lopinavir/ritonavir (Norvir), tipranavir (Aptivus), and zalcitabine (ddC; Hivid) were all associated with null or uncertain fracture risk.
- Other significant fracture risk factors included history of prior fractures, heavy alcohol use, low physical activity, and low body weight.
Some Drugs Raise Risk
- Here, tenofovir exposure was associated with a small but significant increase in fracture risk in a univariate analysis (hazard ratio [HR] 1.08; P < 0.001), falling to 1.06 in both multivariate models.
- Among 32, 439 patients who entered the cohort during the highly active antiretroviral therapy (HAART) era starting in the mid-1990s, tenofovir was associated with a yearly hazard ratio of 1.16 in a univariate analysis (P < 0.001), 1.13 in model MV1 (P = 0.001), and 1.12 in model MV2 (P = 0.011).
- Boosted protease inhibitor exposure was also associated with a small increase in fracture risk: HR 1.11 (P = 0.001) in a univariate analysis, 1.08 (P = 0.026) in model MV1, and 1.05 (non-significant) in model MV2.
- Within this class, lopinavir/ritonavir had an HR of 1.09, which was borderline significant (P = 0.051) in model MV2.
- Osteoporotic fractures were also independently associated with older age, non-black race, low body mass index, and smoking.