Definition of outcome

The biomarkers used to define cardiometabolic risk were defined a priori and represent markers of inflammation, hypertension, glucose metabolism and lipid status and included: High-sensitive CRP, interleukin-6, fibrinogen, systolic and diastolic blood pressure, insulin, glucose, high-density lipoprotein cholesterol and triglycerides. To include potential synergistic effects of different biological domains influencing disease risk, a continuous scale of cardiometabolic risk (CMR) was constructed. A continuous scale is statistically more sensitive and less prone to error compared to dichotomous data [28].

The nine cardiometabolic biomarkers used in the CMR score were standardized (inflammatory markers on the log-scale) to eliminate risk of unequal variance, and sample-weights, represented by latest self-reported BMI-group, were applied. The standardized scores were generated for each sex separately and summarized within each biological domain. The mean values of the four domains were then summarized and standardized to create CMR. Prior to standardization, we multiplied the values of high-density lipoprotein cholesterol by − 1 to account for the inverse association with disease risk. Two participants with diabetes mellitus type 1 were excluded from the glucose metabolism domain but included in the overall CMR score.