Objectives To investigate association between ferritin and 5-year risk of developing metabolic syndrome (MetS) among apparently healthy middle-aged male workers.

Methods We established a prosepective cohort study in an electronic-manufacturing factory by using a detailed medical checkup program in 2002, and followed up them with a health examination in 2007. Each individual underwent physical examination and blood biochemistry tests; body mass index (BMI), blood pressure, and waist circumference were measured by the registered nurses. We collected data from self-reported questionnaires, ferritin, insulin resistance estimated by homeostasis model assessment (HOMA), and fatty liver revealed by abdominal ultrasound in 2002, and aimed to explore their association with risk of metabolic abnormalities in 2007. MetS was diagnosed according to the modified National Cholesterol Education Program Adult Treatment Panel III criteria accounting for Asia Pacific/Taiwanese population. Cox proportional hazard models were applied to discover if ferritin is a predictor for development of MetS.

Results A total of 1493 workers were recruited in this study. Most subjects were males (73.3%) with a baseline mean (SD) age of 32.5 (6.0). Baseline MetS was diagnosed in 21.5% of males, and 14.8% of females. The prevalence of MetS after 5 years significantly increased with the tertiles of baseline ferritin for both genders. In males, ferritin >200 mcg/L was associated with increased risk of MetS. Within 5-year follow-up, incident 114 cases of MetS developed among 877 MetS-free males. Among the non-obese males (BMI< 25), ferritin >200 mcg/L may predict MetS with a hazard ratio (HR) of 2.23 (95% C. I. 1.02–4.89) compared to the first tertile (<123 mcg/L) after controlling for age in the Cox models; non-alcoholic fatty liver diseases (NAFLD) was significantly related to new-onset MetS with a HR of 4.83 while a positively increased trend of higher tertiles of ferritin associated with MetS was observed.