Abstract
BACKGROUND: The association of vitamin D with all-cause mortality remains controversial and longitudinal evidence exploring the potential effects of change in vitamin D status is limited in the oldest old (aged ≥ 80 years old). We aimed to study the relationship between vitamin D change and all-cause mortality among older Chinese adults including the oldest old. METHODS: The data of Chinese Longitudinal and Health Longevity Study in 2012 and 2014 wave was used for baseline data. Mortality was assessed in the subsequent 2018 survey waves. Cox proportional hazard regression models were used to calculate hazard ratios (HRs) and 95% confidence interval (CI) of all-cause mortality related to vitamin D change, including maintaining deficiency or no deficiency, deficiency to no deficiency, and no deficiency to deficiency, using below 50 nmol/L as definition of deficiency. RESULTS: The mean age of the total 1362 participants was 84.4 ± 12.1(60-113) years. The prevalence of vitamin D deficiency was 67.5% and 68.4% in 2012 and 2014 wave respectively, and significantly differed by sex and age at baseline. Cox regression showed that participants with deficiency to no deficiency and maintaining no deficiency of vitamin D status had decreased HR for all-cause mortality, compared to the maintaining deficiency group. The HRs for mortality were 0.70(95%CI: 0.50-0.96, p = 0.028) and 0.47(95%CI: 0.33-0.68, p < 0.001) respectively in the adjusted model. Also, females and the oldest old had a greatest reduction in mortality risk. And no significant difference in mortality in the no deficiency to deficiency group. CONCLUSIONS: Not only maintaining no deficiency, but also the change from deficiency to no deficiency of vitamin D status were associated with lower risk of all-cause mortality, especially in the female and oldest-old participants initially with low vitamin D level.