Aim: Some employees have difficulty undergoing health checkups in the workplace in a fasting state. However, hypertriglyceridemia is usually diagnosed based on fasting triglyceride (TG) measurements. The current study investigated the performance of non-fasting health checkup data for predicting hypertriglyceridemia in a Japanese population.
Methods: We recruited a total of 1,959 Japanese employees who had their fasting TG levels reexamined after undergoing initial health checkups under either a fasting (the fasting population; n= 856) or non-fasting state (the non-fasting population; n= 1103). Hypertriglyceridemia was defined as a fasting TG level of ≥ 1.7 mmol/l.
Results: The area under the receiver operating characteristic (ROC) curve of the initial TG measurements for reexamination-detected hypertriglyceridemia was 0.85 in the fasting population and 0.83 in the non-fasting population. The area under the ROC curve of the initial TG measurements in the non-fasting population was not inferior to that of the multivariate model where other non-fasting health checkup data were added. The optimal non-fasting TG cutoff point was 2.0 mmol/l. The cutoff point was further lowered when the population was limited to patients undergoing health checkups four or more hours after their last meal and when the prevalence of hypertriglyceridemia in the population was simulated to be reduced.
Conclusions: The non-fasting workplace TG measurements by themselves exhibited a tolerable performance for predicting hypertriglyceridemia. The optimal cutoff point in Japanese employees appears to be lower than 2.3 mmol/l, the recently proposed Western cutoff point.