Different Associations Between HDL Cholesterol and Cardiovascular Diseases in People with Diabetes Mellitus and People Without Diabetes Mellitus: a Prospective Community-Based Study.

Zhijun Wu,Zhe Huang,Alice H. Lichtenstein,Cheng Jin,Shuohua Chen,Shouling Wu,Xiang Gao
DOI: https://doi.org/10.1093/ajcn/nqab163
IF: 8.472
2021-01-01
American Journal of Clinical Nutrition
Abstract:BACKGROUND Experimental studies have found that the functionality of HDL cholesterol may be lost in the presence of diabetes mellitus (DM). OBJECTIVES We aimed to elucidate whether DM modified the association between HDL-cholesterol concentrations and cardiovascular disease (CVD) outcomes. METHODS Included were 91,354 Chinese adults (8244 participants with DM and 83,110 participants without DM) free of CVD or cancer at baseline (2006) and without use of lipid-lowering drugs at baseline and during follow-up. The primary endpoint of interest was a composite of CVDs (myocardial infarction, ischemic stroke, and hemorrhagic stroke). Cumulative average HDL-cholesterol concentrations were calculated from all available HDL-cholesterol measures at baseline (2006) and during the follow-up period (2008, 2010, 2012, and 2014). RESULTS During a mean of 10.4 y of follow-up, there were 5076 CVD events identified. There was a significant interaction between DM and HDL-cholesterol concentrations on CVD risk (Pinteraction = 0.003). The association between HDL-cholesterol concentrations and CVD followed a U-shaped curve in individuals without DM (Pnonlinearity < 0.001). The adjusted HR of CVD was 1.26 (95% CI: 1.07, 1.48) for HDL-cholesterol concentrations < 1.04 mmol/L and 1.76 (95% CI: 1.53, 2.03) for HDL-cholesterol concentrations > 2.07 mmol/L, relative to the lowest-risk group (HDL-cholesterol concentrations of 1.30-1.42 mmol/L). In participants with DM, higher HDL-cholesterol concentrations were associated with a higher risk of CVD, in a dose-response manner (Pnonlinearity = 0.44; Ptrend < 0.001). The adjusted HR of CVD was 1.62 (95% CI: 1.19, 2.20) for HDL-cholesterol concentrations >2.07 mmol/L, relative to HDL-cholesterol concentrations of 1.30-1.42 mmol/L. CONCLUSIONS High HDL-cholesterol concentrations were paradoxically associated with high risk of composite CVD outcomes in individuals with or without DM. However, low HDL-cholesterol concentrations failed to predict future CVD risk in individuals with DM.
What problem does this paper attempt to address?