Factors that affect serum levels of ferritin in Australian adults and implications for follow-up.

BACKGROUND & AIMS: Serum levels of ferritin are commonly measured to assess iron stores but are affected by factors such as obesity and chronic disease. Published reference ranges have not changed in decades, and the number of patients whose levels exceed the upper limits has been increasing. As a result, more patients are evaluated for iron overload. METHODS: We compared serum levels of ferritin in 1188 Australian adults who participated in the 2005 Busselton Population Survey with levels from the 1995 survey. Parametric regression was used to assess the effects of body weight and biochemical parameters on serum level of ferritin to derive contemporary population-appropriate reference ranges. RESULTS: In 2005, age-adjusted levels of ferritin were 21% higher in men (P < .0001) and 10% higher in women (P = .01) than in 1995; 31% of men exceeded levels of 300 µg/L, compared with 23% in 1995. Body mass index (BMI) ≥25 kg/m(2) was associated with higher levels of ferritin in men ≥35 years old and in postmenopausal women (P ≤ .002). Serum level of γ-glutamyltransferase (GGT) correlated with serum level of ferritin (P < .0001). In men, the estimated 95th percentiles ranged from 353 to 495 µg/L (<35 years), from 350 to 511 µg/L (≥35 years, BMI <25 kg/m(2)), and from 413 to 696 µg/L (≥35 years, BMI ≥25 kg/m(2)) when GGT levels were 10-75 IU/L. In women, the 95th percentiles ranged from 106 to 235 µg/L (premenopausal), from 222 to 323 µg/L (postmenopausal, BMI <25 kg/m(2)), and from 249 to 422 µg/L (postmenopausal, BMI ≥25 kg/m(2)) when GGT levels were 8-45 IU/L. CONCLUSION: Serum levels of ferritin increased significantly between 1995 and 2005. Reference ranges that accommodate demographic and biomedical variations will assist clinicians in identifying individuals who require further evaluation for iron overload.