Associations of Serum Carotenoids With Risk of Cardiovascular Mortality Among Individuals With Type 2 Diabetes: Results From NHANES
Zixin Qiu,Xue Chen,Tingting Geng,Zhenzhen Wan,Qi Lu,Lin Li,Kai Zhu,Xuena Zhang,Yujie Liu,Xiaoyu Lin,Liangkai Chen,Zhilei Shan,Liegang Liu,An Pan,Gang Liu
DOI: https://doi.org/10.2337/dc21-2371
2022-06-02
Diabetes Care
Abstract:Objective: Although carotenoids have been suggested to exhibit antioxidant properties, some experimental studies reported that β-carotene may show pro-oxidant effects under certain conditions. Current evidence regarding the cardiovascular effects of carotenoids among patients with type 2 diabetes (T2D) is scarce. This study aimed to prospectively examine the associations of individual serum carotenoid concentrations with cardiovascular mortality among adults with T2D. Research design and methods: This analysis included 3,107 individuals with T2D from the Third National Health and Nutrition Examination Survey (NHANES III) and NHANES 2001-2006. Cardiovascular mortality was ascertained by linkage to National Death Index records through 31 December 2015. Cox proportional hazards regression models were used to estimate hazard ratios (HRs) and 95% CIs. Results: During an average of 14 years of follow-up, 441 cardiovascular deaths occurred. After multivariate adjustment including lifestyles, dietary factors, glucose control, and other major carotenoids, higher serum β-carotene concentrations were significantly associated with an elevated risk of cardiovascular mortality in a dose-response manner. When extreme quartiles of β-carotene were compared, the multivariable-adjusted HR was 2.47 (95% CI 1.62, 3.76) for cardiovascular mortality (Ptrend = 0.002); and per one-unit increment in natural log-transformed serum β-carotene was associated with a 46% higher risk of cardiovascular mortality (P = 0.001). Other individual carotenoids (α-carotene, β-cryptoxanthin, lycopene, and lutein/zeaxanthin) were not significantly associated with the risk of cardiovascular mortality. Consistent results were observed when stratifying by age, sex, race, BMI, smoking status, diabetes duration, and glycated hemoglobin A1c levels. Conclusions: Higher concentrations of serum β-carotene, but not other individual carotenoids, were significantly associated with an increased risk of cardiovascular mortality among individuals with T2D. Our findings, if replicated, underscore the need to estimate the optimal serum β-carotene concentrations in individuals with T2D.