Serum phthalates and risks of incident cardiovascular disease among individuals with type 2 diabetes in a prospective cohort study

Prospective epidemiological evidence about the associations between phthalates (PAEs) and incident cardiovascular disease (CVD) risk among type 2 diabetes mellitus (T2DM) population has been urgently needed. A total of 2806 participants with T2DM were recruited based on the baseline of the Dongfeng-Tongji cohort study (2008-2010) and followed up to 2018. Serum di-n-butyl phthalate (DnBP), Σdibutyl phthalate (ΣDBP), and Σlow-molecular-weight phthalate (ΣLMW) levels were associated with incident stroke risk, whether their levels were treated as continuous variables [HR (95%CI): 1.36 (1.02, 1.80), 1.35 (1.01, 1.81), and 1.50 (1.07, 2.10), respectively] or divided by quartiles [HR (95%CI): 1.41 (1.01, 1.98), 1.48 (1.05, 2.08), and 1.53 (1.08, 2.15) in the highest quartile with <i>P</i> _trend ​= ​0.03, 0.02, and 0.02, respectively] after multiple adjustment in Cox proportional hazard models. Meanwhile, linear associations for serum di-iso-butyl phthalate (DiBP), DnBP, ΣDBP, and ΣLMW, and a nonlinear U-shaped dose-response association for serum butyl-benzyl phthalate (BBP) with incident stroke risk were proved by restricted cubic spline regression analyses. Furthermore, a positive correlation between PAE mixture and incident stroke risk was identified in Quantile-based g-computation [HR (95%CI): 1.24 (1.05, 1.46)]. However, serum PAEs, either as individuals or as a mixture, were not associated with incident CVD or coronary heart disease risk. Our study revealed that serum PAEs were positively associated with increased incident stroke risk in the T2DM population. More prospective cohort and mechanism studies are warranted to validate our findings.