Ferritin levels throughout childhood and metabolic syndrome in adolescent stage

Abstract

Background and aim: Increased ferritin levels have been widely associated with cardiovascular risk in adults. Whether ferritin levels and their changes during childhood are related to metabolic syndrome (MetS) at adolescence is unknown. We aimed to evaluate these associations using levels of ferritin at 5, 10 and 16 years and their linear increases and patterns of sustained increased levels across childhood.

Methods and results: There were four samples evaluated according to non-missing values for study variables at each stage (5 years: 562; 10 years: 381; and 16 years: 567 children; non-missing values at any stage: 379). MetS risk was evaluated as a continuous Z score. Patterns of sustained increased ferritin (highest tertile) and slope of the change of ferritin per year across the follow-up were calculated. Ferritin levels in the highest versus lowest tertile at five and 16 years were significantly positively associated with MetS risk Z score at adolescence in boys and these associations were unaffected by adjustment for covariates. Having high, compared to low/moderate ferritin level at 2 or more time periods between 5 and 16 years was related to higher Mets Z-score in boys only [e.g. 5-10 years adjusted-beta (95 %CI):0.26 (0.05-0.48),P < 0.05]. In girls, ferritin Z score at 10 and 16 years was positively and independently associated with HOMA-IR Z score. In girls, the slope of ferritin per year in the highest tertile was positively associated with MetS risk Z-score [adjusted-beta (95 %CI):0.21 (0.05-0.38),P < 0.05].

Conclusions: Ferritin levels throughout childhood are positively related to cardiometabolic risk in adolescence, with associations varying by sex.

Keywords: Children; Ferritin; Iron; Metabolic syndrome.

Figure 1.. Linear regression for longitudinal (A and B) and cross-sectional (C) associations between tertiles of ferritin at the three points of follow-up and the MetS Z score.

A, n= 565. B= n=381. C, n=567. Reference. Unadjusted. Adjusted: Longitudinal associations with ferritin at 5 and 10 years were adjusted for the respective baseline BMI Z score and haemoglobin level, and tanner stages for 10 year’s evaluation; Cross-sectional association at 16–17 years were adjusted for BMI Z score, CRP level, haemoglobin level and tanner stage. Means (SD) of MetS Z score by tertiles are also provided.

Figure 2.. Linear regressions associations between MetS Z score at 16/17 years and patterns of combined high(H) and low/moderate(L/M) ferritin concentration throughout the follow-up.

Reference. Unadjusted. Adjusted: aassociations were adjusted for the values of BMI Z score and haemoglobin level and tanner stage(if applicable) at the end of the interval (e.g interval 5–10 years included covariates of BMI Z score, haemoglobin levels, and tanner stage at 10 years old). For major comparability of the intervals of follow-up, this analysis used cases with no missing values for all of the three points of measurements (n=379 at each stage).