59 (0.39, 0.88) EMD 1214063 and 0.52 (0.37, 0.72), respectively, after adjusting for age, gender and calendar year of starting HAART. When the effect of HBV or HCV coinfection on the probability of developing elevated levels of individual lipids was examined, HBV was found to have a stronger effect and HCV had a somewhat weaker effect than in the analysis classifying patients as HBV- or HCV-coinfected if they had a positive laboratory test

or a note in the medical chart. It was not possible to assess the effects of all individual antiretroviral medications in these analyses, as a consequence of the fact that the outcome was the occurrence of a grade 3 or 4 elevation in lipid measurements or use of lipid-lowering drugs at any time during follow-up and antiretroviral use changed over time. However, we did specifically assess the effects of ever having used tenofovir given the dual antiviral activity against HIV and HBV and of ever having used nevirapine given the relatively ‘lipid friendly’ characteristics of this medication. The proportions of participants who had ever used tenofovir was greater in HIV/HBV-coinfected patients (64%) compared with HIV-monoinfected patients (47%) (P<0.0001) but similar in HIV/HCV-coinfected individuals (51%) compared with

monoinfected individuals (P=0.22). Tenofovir use was not associated with hyperlipidaemia or lipid-lowering drug use (unadjusted OR 1.05; 95% CI 0.88, 1.24; P=0.62). The proportions of participants who had ever used nevirapine was lower in HIV/HBV-coinfected patients (19%) compared with HIV-monoinfected patients (27%) (P=0.02) but similar in Crizotinib HIV/HCV-coinfected individuals (24%) compared with monoinfected individuals (P=0.27). Nevirapine use was associated with hyperlipidaemia or lipid-lowering drug use (adjusted OR 1.41; 95% CI 1.14, 1.74; P<0.01). Methocarbamol This may reflect a selection bias or the concurrent nucleosides administered with nevirapine. Other previously reported predictors remained unchanged

with the inclusion of nevirapine in our models. Chronic HCV infection has been associated with lower total and LDL cholesterol levels in patients with and without advanced liver disease [8–13,15]. Lower serum triglyceride and cholesterol levels have been reported in those with chronic HCV infection [16]. Our analysis suggests that this perturbation of the lipid profile extends to HAART-treated, HIV/HCV-coinfected patients. This is consistent with our previous work [8] and an analysis specifically focused on an HIV/HCV-coinfected Hispanic population [17]. HBV may have a much smaller effect on lipid profile. However, this effect was inconsistently demonstrated by our analysis. HIV/HCV coinfection was found to protect against grade 3 and 4 lipid events following the initiation of HAART. This effect was consistent over the entire period of evaluation.