All covariates were assessed at baseline following standard protocols. BMI was calculated from measured weight and height at both visits and obesity was defined as a BMI ≥ 30 kg/m2. Diabetes was defined as fasting (≥8 hours) blood glucose ≥126 mg/dL, non-fasting blood glucose ≥ 200 mg/dL, self-reported physician diagnosed diabetes or “sugar in blood”, or use of medication for diabetes in the past two weeks. Hypertension was defined as mean systolic blood pressure (BP) ≥ 140 mmHg, mean diastolic BP ≥ 90 mmHg, or use of medication for high blood pressure in the past 2 weeks. Atherosclerotic cardiovascular disease (ASCVD) was defined as history of coronary heart disease and/or stroke.
Assays of NT-proBNP levels were conducted in serum (visit 2) and plasma (visit 4) samples that had been stored at −70°C. Measurements were made using a sandwich immunoassay method on a Roche Elecsys 2010 Analyzer at visit 2, and an ECLIA immunoassay on an automated Cobas e411 analyzer at visit 4. The lower detection limit for both assays was 5 pg/mL, and participants with unmeasurable levels were assigned a value of 2.5 pg/mL. NT-proBNP, glucose, total cholesterol, HDL- and LDL-cholesterol, triglycerides and hs-CRP were re-calibrated based on published equations to minimize any systematic differences across study visits (23 (link)).
Assays of NT-proBNP levels were conducted in serum (visit 2) and plasma (visit 4) samples that had been stored at −70°C. Measurements were made using a sandwich immunoassay method on a Roche Elecsys 2010 Analyzer at visit 2, and an ECLIA immunoassay on an automated Cobas e411 analyzer at visit 4. The lower detection limit for both assays was 5 pg/mL, and participants with unmeasurable levels were assigned a value of 2.5 pg/mL. NT-proBNP, glucose, total cholesterol, HDL- and LDL-cholesterol, triglycerides and hs-CRP were re-calibrated based on published equations to minimize any systematic differences across study visits (23 (link)).
