Vitamin D and incidence of mortality, morbidity, and growth failure among a cohort of HIV‐infected and HIV‐exposed Tanzanian infants (804.5)

Methods: Serum 25(OH)D was quantified by HPLC‐MS/MS at 5‐7 weeks of age for 253 HIV‐infected and 948 HIV‐exposed (uninfected) infants enrolled in a trial of multivitamins (not including vitamin D) conducted in Tanzania. Infants were prospectively followed at monthly clinic visits for 24 months. Results: The association of 25(OH)D with mortality appeared to be U‐shaped for both HIV‐infected and HIV‐exposed infants. After adjustment, 25(OH)D 蠅30 ng/mL (classically defined as sufficient) was significantly associated with increased mortality as compared to 25(OH)D 蠅20 and <30 ng/mL for HIV‐infected (HR: 2.47; 95% CI: 1.13‐5.44; p=0.02) and HIV‐exposed infants (HR: 4.00; 95% CI: 1.67‐9.58; p<0.01). 25(OH)D levels <10 ng/mL also appeared to be associated with increased risk of mortality as compared to 蠅20 and <30 ng/mL in both groups, but results were not statistically significant. Among HIV‐exposed infants, 25(OH)D <10 ng/mL was significantly associated with incidence of wasting (p<0.01) and oral candidiasis (p=0.04), while 25(OH)D 蠅30 ng/mL was associated with increased diagnosis of malaria (p<0.01). Conclusion: These results suggest a complex relationship of vitamin D with infant mortality and morbidity in the context of HIV. Trials of vitamin D supplements appear to need rigorous monitoring of safety, particularly for malaria, among those who reach high vitamin D levels in this setting. Grant Funding Source : Supported by NICHD, NIAID and Harvard