What's the Relation between Iron Deficiency (ID) and Febrile Seizure (FS)? A Case Control Study in Tehran, Iran

Abstract

Objective: To evaluate the role of ID in the pathogenesis of FS. Methods: In this case-control study (2014-2016), 70 children were studied, 35 children with FS and 35 (controls)children with febrile diseases without convulsion (The mean age of cases was 2.191 ± 0946 vs. 1.93 ± 1.433 years in controls). Serum ferritin was estimated by the EIAS test. Data were compared between 2 groups, The ROC (receiver-operating-characteristic) curve was illustrated. The sensitivity, specificity, PPV, and NPV of the test, were calculated. Results: Serum ferritin levels had no significant difference between the 2 groups. The ferritin level (36ng/ml) had 74.3% sensitivity, 20% specificity, 56% PPV, and 52% NPV, with a Positive likelihood Ratio being 1.3 and a Negative likelihood Ratio: 0.93 to discriminate the 2 groups. Conclusion: Here the ferritin level (cut-off=36ng/ml) has an acceptable sensitivity (74.3%) but poor specificity (20%) and just 56% PPV and 52% NPV to differentiate the FS cases from non-convulsive febrile children. Although a different cut-off value 21.50 ng/ml provides 91% sensitivity and very low specificity . This lower threshold cut-off might have clinically relevant outcomes in FS children if considering the other comorbidities. In our opinion, ID could not lead to FS in all children, but in some cases, with a genetic basis; ID raises the threshold for seizures. The ferritin levels as an acute phase reactant are acceptable in every febrile case. The ferritin base level in each child (case /control) before infection was unknown, but in the present study, both groups were febrile in contrast to previous studies in which ferritin levels were compared with afebrile children. Due to the high prevalence of ID (26%), especially in the young Iranian population, adding iron to the diet might help decrease FS in susceptible cases. We recommend in the future study the FS cases selected with known iron levels before convulsion.