Background An association has been described between elevated ferritin (the major iron storage protein) levels and metabolic syndrome (MetS) in several populations. However there are limited data on this association in the UK population. We investigated whether ferritin levels were associated with MetS and its components in participants in the Viking Health Study – Shetland (VHSS).
Methods The analysis was conducted using data collected between 2013 and 2015 for 2047 individuals from the Shetland Islands (589 premenopausal women, 625 postmenopausal women, and 833 men) of 18–93 years of age. Ferritin levels were categorised into sex/menopausal specific quartiles and cut- points from the consensus definition of MetS and the interim joint statement definition of MetS was used to define risk factors. Logistic regression models were used to adjust for age, fibrinogen levels, smoking, alcohol consumption and body mass index.
Results Median (inter-quartile ranges) were 51 (38.9–63.2) years for age, and 56 (29–97) µg/L for ferritin. Prevalence of risk factors were: high waist circumference (HWC) 66.7 %, low HDL-cholesterol (LHDL) 17.1%, high glucose (HG) 7.2%, high triglycerides (HT) 11.9% and high blood pressure (HBP), 50.4% and prevalence of MetS (having three or more of the previous risk factors) was 18%. In un-adjusted models, the highest compared with the lowest quartile of ferritin was associated with HT and HWC in post-menopausal women (P < 0.05) and with all of the risk factors (except HBP) in men (P < 0.05). In men the crude OR for MetS for the highest compared to the lowest quartile of ferritin was 2.15 (1.36–3.41) (P < 0.001) but the fully adjusted OR was 1.43 (0.83–2.46) (P = 0.207). The association between ferritin and MetS was not statistically significant in either group of women in crude or adjusted analyses.
Conclusion Ferritin levels were not independently associated with MetS in this population. The analysis confirms our previous findings of the lack of an independent association between increased ferritin and some MetS components (LHDL, HBP and HWC) among participants in Scottish Health Surveys 1995–1998. The contrasting findings with previous studies, suggest that the ferritin-MetS association could differ between populations, possibly related to differences in distribution of ferritin and a threshold effect.