Multiple vitamin co-exposure and mortality risk: A prospective study

Abstract

Background & aims: Existing epidemiological studies explored the associations of circulating vitamins and mortality focusing on individual vitamin effects, and controversial findings were obtained. The joint effects of multiple vitamin co-exposure are worth studying. The study aimed to elucidate the associations of circulating vitamins and the joint effects of these vitamins' co-exposure with all-cause and cause-specific mortality risks.

Methods: We prospectively evaluated the associations of the concentrations of six kinds of vitamins (A, D, E, C, B12 and B9) in serum with risks for all-cause and cause-specific mortalities among U.S. adults. Mortality status and cause of death were determined by NHANES-linked public available files dated up to 31 December 2015. An unsupervised K-means clustering method was used to cluster the participants into several vitamin co-exposure patterns. The Cox proportional hazards model was used for statistical analysis.

Results: A total of 1404 deaths occurred during a median of 10.9 years follow-up among 8295 participants. In multivariable adjustment, increasing levels of vitamin D were associated with reduced all-cause and cause-specific mortality risks. A J-shaped nonlinear exposure-response relationship was observed between all studied vitamins (except for vitamin D) and all-cause mortality risk. Four co-exposure patterns were generated based on the studied vitamins, as follows: low-level exposure (cluster 1), vitamin A/D exposure (cluster 2), water-soluble vitamin exposure (cluster 3) and high-level exposure (cluster 4). Compared with those in cluster 1, participants in cluster 2 had lower all-cause and cancer mortality risks, with hazard ratios (95% confidence intervals [CIs]) of 0.67 (0.53, 0.85) and 0.45 (0.29, 0.71), respectively.

Conclusions: The findings in this study indicated that high circulating vitamin D levels were associated with reduced mortality risk among U.S. adults. Vitamin co-exposure at moderate levels appropriately contributed to low all-cause and cancer mortality risks. Our findings provided a novel perspective for exploring the joint health effects of multivitamin co-exposure. Future investigations are needed to further unravel the underlying mechanisms of possible vitamin interactions.

Keywords: Co-exposure; K-means clustering; Mortality; Vitamin.