Milton Fabian Suárez-Ortegón1, Eduardo Ensaldo-Carrasco2, Ting Shi2, Stela McLachlan2, José Manuel Fernández-Real3, Sarah H Wild2. 1. Centre for Population Health Sciences, University of Edinburgh, Edinburgh, Scotland, United Kingdom; Nutrition Group, Universidad del Valle, Cali, Colombia. Electronic address: Milton.Suarez@ed.ac.uk. 2. Centre for Population Health Sciences, University of Edinburgh, Edinburgh, Scotland, United Kingdom. 3. Department of Diabetes, Endocrinology and Nutrition, Institut d'Investigació Biomèdica de Girona, CIBEROBN (CB06/03/010) and Instituto de Salud Carlos III (ISCIII), Girona, Spain. Electronic address: jmfreal@idibgi.org.
Abstract
BACKGROUND AND AIMS: Mechanisms for the association between iron stores and risk factors for diabetes and cardiovascular disease, such as metabolic syndrome (MetS) and its components, are still not clear. We evaluated the associations between ferritin levels, MetS and its individual components, and potential role of confounding, in a meta-analysis. METHODS: We searched articles in MEDLINE and EMBASE until February 14th, 2018. There were two approaches: meta-analysis of 1) cross-sectional and longitudinal studies and 2) only cross-sectional studies. Meta-regressions were conducted to identify sources of heterogeneity in the associations of ferritin with MetS and its individual components. RESULTS: Information from 26 studies (5 prospective) was systematically reviewed and 21 studies were meta-analysed. The pooled OR for MetS by increased ferritin was 1.78 (95%CI: 1.60-1.97) in the meta-analysis 1, and 1.70 (95%CI: 1.49-1.95) in the meta-analysis 2. The pooled association was weaker in studies adjusted for hepatic injury markers (meta-regression coefficient (95% CI): -0.34 (-0.60,-0.09) p = 0.008) and body mass index (BMI) (meta-regression coefficient (95% CI): -0.27 (-0.53,-0.01) p = 0.039). Among MetS components, the pooled association with increased ferritin was strongest with high triglycerides [OR (95%CI): 1.96 (1.65-2.32)] and high glucose levels [OR 95%CI: 1.60 (1.40-1.82)]. Higher cut-off points used to define high ferritin concentrations were more strongly associated with high triglycerides [meta-regression coefficient (95% CI): 0.22 (0.03, 0.041), p = 0.023]. CONCLUSIONS: High triglycerides and glucose are the components more strongly associated with ferritin. Hepatic injury and BMI appear to influence the ferritin-MetS association, and a threshold effect of high ferritin concentration on the ferritin-high triglycerides association was observed.
BACKGROUND AND AIMS: Mechanisms for the association between iron stores and risk factors for diabetes and cardiovascular disease, such as metabolic syndrome (MetS) and its components, are still not clear. We evaluated the associations between ferritin levels, MetS and its individual components, and potential role of confounding, in a meta-analysis. METHODS: We searched articles in MEDLINE and EMBASE until February 14th, 2018. There were two approaches: meta-analysis of 1) cross-sectional and longitudinal studies and 2) only cross-sectional studies. Meta-regressions were conducted to identify sources of heterogeneity in the associations of ferritin with MetS and its individual components. RESULTS: Information from 26 studies (5 prospective) was systematically reviewed and 21 studies were meta-analysed. The pooled OR for MetS by increased ferritin was 1.78 (95%CI: 1.60-1.97) in the meta-analysis 1, and 1.70 (95%CI: 1.49-1.95) in the meta-analysis 2. The pooled association was weaker in studies adjusted for hepatic injury markers (meta-regression coefficient (95% CI): -0.34 (-0.60,-0.09) p = 0.008) and body mass index (BMI) (meta-regression coefficient (95% CI): -0.27 (-0.53,-0.01) p = 0.039). Among MetS components, the pooled association with increased ferritin was strongest with high triglycerides [OR (95%CI): 1.96 (1.65-2.32)] and high glucose levels [OR 95%CI: 1.60 (1.40-1.82)]. Higher cut-off points used to define high ferritin concentrations were more strongly associated with high triglycerides [meta-regression coefficient (95% CI): 0.22 (0.03, 0.041), p = 0.023]. CONCLUSIONS: High triglycerides and glucose are the components more strongly associated with ferritin. Hepatic injury and BMI appear to influence the ferritin-MetS association, and a threshold effect of high ferritin concentration on the ferritin-high triglycerides association was observed.