RESUMEN
BACKGROUND: A high prevalence of hypovitaminosis D (hypD) in HIV-infected patients has been reported, but reasons are unclear. METHODS: The 25 hydroxy vitamin D (vitD) concentration was measured in a sample of HIV-positive patients from Italy enrolled in the Icona Foundation Study. The change in absolute levels of vitD pre/post combination antiretroviral treatment was modelled by linear regression controlling for confounders and seasonality. Factors associated with hypD were identified using logistic regression analysis, and survival analysis was employed to evaluate the prognostic value of vitD concentration to predict severe diseases (diabetes, cardiovascular, renal), AIDS, and death. RESULTS: We studied 810 patients contributing 1408 vitD measures. Median age was 36 years (range: 20-69). VitD insufficiency (30-75 nmol/L) and deficiency (<30 nmol/L) were found in 47% and 6% of the measures. Factors independently associated with vitD deficiency were African or Centre/South American nationality [odds ratio (OR): 4.16 vs. European, P = 0.04], the sample being collected in spring (OR: 11.27, P = 0.001) or in winter (OR: 4.22, P = 0.03) vs. summer, and a previous history of severe diseases (OR: 5.43, P = 0.03) or AIDS (OR: 2.44, P = 0.04). Over a median follow-up of 6.3 years, patients with vitD insufficiency were at higher risk of subsequent severe diseases than those with normal levels (relative hazard = 1.60, P = 0.05). CONCLUSIONS: Our analysis shows that despite the relatively young age of our HIV-infected population, the prevalence of hypD was high. Classic risk factors for hypD in the general population were confirmed in this setting. HypD seems to be moderately associated with the risk of severe disease, AIDS, and death.