We did not observe

an association between low BMI and baseline BMD values, although we did find an association between low BMI and subsequent decline in hip BMD. However, most patients were normal weight (BMI between 20 and 25), which Dinaciclib concentration may have diminished the influence of BMI. Of interest, Aukrust et al. found markedly decreased levels of bone formation markers and increased levels of bone resorption markers in untreated patients with advanced HIV infection and also found indications of normalization of the bone remodelling process during HAART [15]. The decrease in BMD observed during the initial 24 to 48 weeks of therapy could also partly be due to an ongoing BMD loss in untreated HIV-infected individuals, which do not reverse immediately after initiation of HAART. However, bone loss of the magnitude we observed in the 24–48-week period after HAART initiation

could not have taken place Lumacaftor nmr during the often many years of asymptomatic HIV infection without producing more pronounced osteopenia than observed at baseline in this and other studies [25]. In our study, the factors associated with a low baseline BMD were different from the factors associated with bone loss after HAART initiation; most notably, there was no association between low baseline CD4 cell count and low baseline BMD. It is important to note that the between-patient variability and statistical power concerning baseline BMD in the cross-sectional analysis are different from those in the analyses concerning percentage change in BMD from baseline to 24 weeks, but different processes may also drive the bone loss before and Clomifene after HAART initiation. Data on the effect of different drug classes on

BMD have not been consistent. While some observational studies found that PIs increased the risk of BMD decline, others did not confirm these results. In particular, studies that controlled for HIV-related and traditional risk factors for osteoporosis did not find an independent effect of PIs [26,27]. A randomized French study (n=71) of three different class-sparing strategies found a more pronounced decrease in lumbar spine BMD in the two PI-containing arms compared with the PI-sparing arm; however, no differences were found at the hip [6]. In contrast, recent results from a Dutch study including 50 patients indicated a role of zidovudine/lamivudine, as patients randomized to zidovudine/lamivudine and lopinavir/ritonavir had more pronounced bone loss than patients randomized to lopinavir/ritonavir and nevirapine [17].