Abstract

Objective To compare approaches for adjusting serum ferritin concentrations for inflammation in Chilean adolescents with overweight and obesity. Study design Cross-sectional data from 518 adolescents (aged 16-17 years; 48% females) from Santiago, Chile were analyzed. Several approaches were compared for estimating the prevalence of depleted iron stores (defined as serum ferritin 15 mu g/L), including unadjusted prevalence and higher cutoffs for various subgroups (excluding participants with inflammation), correction factors, and regression corrections. A "reference" prevalence estimate was calculated as the prevalence of serum ferritin 15 mu g/L in normal weight individuals without inflammation. Each adjustment approach was compared with this reference prevalence. Results The sample comprised 61.2% normal weight, 23.7% overweight, and 15.1% obese individuals. The prevalence of inflammation (marked by C-reactive protein level >5.0 mg/L) was 6.3%, 8.1%, and 14.1% in the 3 groups, respectively. The correction factor approaches produced adjusted estimates closest to the reference estimate (24.1%-24.7% vs 22.9%), followed by the regression corrections (24.7%-25.1% vs 22.9%). Applying a higher serum ferritin cutoff (30 mu g/L) to all participants or to participants with overweight/obesity produced adjusted estimates farthest from the reference (59.5% and 35.3%, respectively). Conclusions Adjusting serum ferritin concentration may be necessary when assessing iron status in populations with high rates of overweight/obesity. After reviewing 6 approaches for adjusting for the influence of inflammation, this study suggests that using correction factors may be the most appropriate approach for adjusting serum ferritin in Chilean adolescents. Further research is needed to determine the optimal approach for adjusting serum ferritin concentrations for weight-related inflammation in broader populations.