Causal Associations Between Age at Diagnosis of Diabetes and Cardiovascular Outcomes: A Mendelian Randomization Study.

Chaojie Ye,Lijie Kong,Yiying Wang,Hong Lin,Shuangyuan Wang,Zhiyun Zhao,Mian Li,Yu Xu,Jieli Lu,Yuhong Chen,Min Xu,Weiqing Wang,Guang Ning,Yufang Bi,Tiange Wang

DOI: https://doi.org/10.1210/clinem/dgac658

2022-01-01

Abstract:CONTEXT:Whether diabetes diagnosed at different age groups is causally associated with cardiovascular diseases (CVDs) is unknown. OBJECTIVE:We conducted 2-sample Mendelian randomization analyses to investigate the causal associations of diabetes by age at diagnosis with 5 type-specific CVDs and 11 cardiometabolic traits. METHODS:We selected 208 single nucleotide polymorphisms (SNPs) for diabetes and 3, 21, 57, and 14 SNPs for diabetes diagnosed at <50, 50-60, 60-70, and >70 years, respectively, based on the genome-wide association study (GWASs) (24 986 cases/187 130 controls) in the UK Biobank, and extracted genetic associations with stroke, myocardial infarction, heart failure, atrial fibrillation, and CVD mortality, as well as blood pressures, adiposity measurements, and lipids and apolipoproteins from corresponding European-descent GWASs. The inverse variance-weighted method was used as the main analysis with several sensitivity analyses. RESULTS:Diabetes diagnosed at all 4 age groups was causally associated with increased risks of stroke (5-8%) and myocardial infarction (8-10%), higher systolic blood pressure (0.56-0.94 mmHg) and waist to hip ratio (0.003-0.004), and lower body mass index (0.31-0.42 kg/m2), waist circumference (0.68-0.99 cm), and hip circumference (0.57-0.80 cm). Diabetes diagnosed at specific age groups was causally associated with increased risks of heart failure (4%) and CVD mortality (8%), higher diastolic blood pressure (0.20 mmHg) and triglycerides (0.06 SD), and lower high-density lipoprotein cholesterol (0.02 mmol/L). The effect sizes of genetically determined diabetes on CVD subtypes and cardiometabolic traits were comparable and the corresponding 95% confidence intervals largely overlapped across the 4 age groups. CONCLUSION:Our findings provide novel evidence that genetically determined diabetes subgroups by age at diagnosis have similar causal effects on CVD and cardiometabolic risks.