L-shaped association of serum 25-hydroxyvitamin D concentrations with cardiovascular and all-cause mortality in individuals with osteoarthritis: results from the NHANES database prospective cohort study

Abstract

Abstract Background The relationship between vitamin D status and mortality in patients with osteoarthritis (OA) is unknown. This study investigated the associations of serum 25-hydroxyvitamin D [25(OH)D] concentrations with all-cause and cause-specific mortality among American adults with OA. Methods This study included 2556 adults with OA from the National Health and Nutrition Examination Survey (2001–2014). Death outcomes were ascertained by linkage to National Death Index (NDI) records through 31 December 2015. Cox proportional hazards model and two-piecewise Cox proportional hazards model were used to elucidate the nonlinear relationship between serum 25(OH)D concentrations and mortality in OA patients, and stratified analyses were performed to identify patients with higher mortality risk. Results During 16,606 person-years of follow-up, 438 all-cause deaths occurred, including 74 cardiovascular disease (CVD)-related and 78 cancer deaths. After multivariable adjustment, lower serum 25(OH)D levels were significantly and nonlinearly associated with higher risks of all-cause and CVD mortality among participants with OA. Furthermore, we discovered L-shaped associations between serum 25(OH)D levels and all-cause and CVD mortality, with mortality plateauing at 54.40 nmol/L for all-cause mortality and 27.70 nmol/L for CVD mortality. Compared to participants with 25(OH)D levels below the inflection points, those with higher levels had a 2% lower risk for all-cause mortality (hazard ratio [HR] 0.98, 95% confidence interval [CI] 0.96–0.99) and 17% lower risk for CVD mortality (HR 0.83, 95% CI 0.72–0.95). Conclusions Nonlinear associations of serum 25(OH)D levels with all-cause and CVD mortality were observed in American patients with OA. The thresholds of 27.70 and 54.40 nmol/L for CVD and all-cause mortality, respectively, may represent intervention targets for lowering the risk of premature death and cardiovascular disease, but this needs to be confirmed in large clinical trials.