Food Allergy and Mental Health in Children and Adolescents-The Role of Shared Familial Environment

BackgroundEvidence suggests a link between food allergy and poor mental health, however, this may be explained by shared genetic and environmental factors. We aimed to investigate the association between food allergy of different severity and mental health in children, and the role of familial factors.MethodsThis population-based, longitudinal cohort study is based on the Child and Adolescent Twin Study in Sweden with questionnaire data reported by parents and/or children. Food allergy 'ever' and doctor's diagnosis were reported at age 9-12 years, and >= 1 recent dispensation of adrenaline was used as a marker for current severe food allergy. Outcomes were identified using validated questionnaires for anxiety; Screen for Child Anxiety Related Disorders (SCARED); Strength and Difficulties Questionnaire (SDQ) and depression; Short Mood and Feelings Questionnaire (SMFQ), Center for Epidemiological Studies Depression Scale (CES-D) and Diagnostic and Statistical manual of Mental Disorders (DSM-IV-MDE) and reported at 9-12, 15 and 18 years of age. Multivariate linear and logistic modelling was applied to the whole cohort and a co-twin control approach to remove confounding by familial factors.ResultsIn total, 3039 (8.9%) children had a parent-reported food allergy. Among these, 1292 (43.5%) had non-severe food allergy without diagnosis, 1490 (49%) had non-severe food allergy with diagnosis and 257 (8.5%) had severe food allergy. Compared to children with no food allergy, non-severe food allergy with diagnosis by 9-12 years was associated with parent-reported anxiety/depression; SCARED (adjOR 2.10, 95% CI 1.48-2.98), SMFQ (adjOR 1.92, 95% CI 1.19-3.10) at 9-12 years and SDQ (adj beta 0.2, 95% CI 0.0-0.4) at 15 years. All other associations were null including for those with severe food allergy. All positive estimates in the full cohort were attenuated using co-twin controls.ConclusionEvidence associating paediatric food allergy severity and poor mental health was weak, and positive associations observed were likely due to familial confounding.