Abstract
OBJECTIVE: Although the effect of vitamins on the risk of mortality in diabetic patients has been reported, most studies focus on individual vitamins. However, humans are often exposed to multiple vitamins simultaneously in daily life. Therefore, it is worth exploring the effects of co-exposure to multiple vitamins on the risk of mortality in diabetic patients. METHODS: This study included diabetic patients aged ≥20WD years who participated in NHANES from 2003 to 2006. An unsupervised K-means clustering method was used to cluster eight vitamins in serum into several patterns of co-exposure to multiple vitamins, and the Cox proportional hazards model was used to evaluate the impact of different patterns of co-exposure to multiple vitamins on the risk of all-cause mortality in diabetic patients. RESULTS: Three patterns of co-exposure to multiple vitamins were generated based on K-means clustering, namely, low-level, moderate-level, and high-level. Among the 484 diabetic patients, with a median follow-up of 13.7 years, a total of 211 deaths occurred. After adjusting for covariates, the individual vitamins had varying effects on the risk of all-cause mortality in diabetic patients. Compared to the low-level group of co-exposure to multiple vitamins, the high-level group significantly reduced the risk of all-cause mortality in diabetic patients, with a HR of 0.42 (95% CI: 0.20, 0.87). Subgroup analysis demonstrated that high levels of co-exposure to multiple vitamins significantly reduced the risk of all-cause mortality in males, individuals aged ≥ 60 years, and non-Hispanic White people with diabetes compared to the low-level group, with HR of 0.42 (95% CI: 0.18, 0.98), 0.53 (95% CI: 0.26, 0.98), and 0.26 (95% CI: 0.12, 0.58) respectively. CONCLUSION: While individual vitamins had different effects on the risk of all-cause mortality in patients with diabetes, high-level co-exposure to multiple vitamins significantly reduced the risk of all-cause mortality in patients with diabetes, with differences observed among genders, ages, and race. This suggests that when developing vitamin intervention strategies for patients with diabetes, consideration should be given not only to the dosage of individual vitamins but also to the variations between different population groups.