19456
Three-Generation Family-Wide Morbidity Patterns in Autism Spectrum Disorder: A Danish Population-Based Cohort Study
Many associations between specific family morbidity histories and the risk for autism spectrum disorder (ASD) have been reported, although studies have been limited in the range of family members and morbidities under investigation. Better understanding of family morbidity and its role in ASD etiology, e.g., delineating subgroups with different underlying pathogenesis, requires a more comprehensive perspective.
Objectives:
Conduct a family-wide morbidity association study: population-based cohort analysis of morbidity across three-generation families and their associations with ASD.
Methods:
The cohort comprised all Danish births from 1980 through 2010 (N=1,934,672 births) identified in the Medical Birth and Central Person Registers and linked to parents, grandparents and their descendants. Cohort and family members were followed through 2013 for any diagnoses reported to the Psychiatric Central Research Register and National Patient Register. Adjusted hazard ratios (aHR) of the relative risk for ASD in cohort members from family morbidity at the time of ASD diagnosis, calculated for each family member type and each morbidity diagnosis, were estimated with Cox regression methods using separate baseline ASD diagnostic rates in 3-year strata of birth year, adjusting for sex, birth weight, gestational age, parental age. Family history scores (FamHx) by morbidity for each cohort member were based on the sum of the log-aHRs per morbidity and family member type with the morbidity divided by the total number of family members (with and without the morbidity) for the cohort member. Bivariate correlations between family history scores were computed for cohort members. Due to the volume of results, for this abstract example results are reported for 12 morbidities (ASD, attention deficit hyperactivity disorder, intellectual disability (ID), obsessive compulsive disorder, affective disorder, schizophrenia, epilepsy, thyroid disorders, asthma, non-infective enteritis/colitis, dermatitis/eczema, hypertensive disorders of pregnancy) among parents, full/half siblings, and grandparents of cohort families with at least 2 children, only considering the first 2 children and the cohort member was the second child.
Results:
These example analyses and results included 693,331 cohort children (0.9% with ASD). Significant aHRs for ASD were observed across all 12 morbidities and family member types, from 16.4 (fathers with ASD; 95% CI 12.7-21.2) to 1.1 (maternal grandfathers with asthma; 95% CI 1.0-1.3). In ASD, 1%-24% (1%-17% of non ASD) had a morbidity family history. FamHx correlations between psychiatric disorders or between psychiatric and non-psychiatric disorders tended to be stronger in ASD than non-ASD and the correlation patterns suggest hypotheses for further study. For example, in ASD, the correlations between FamHX of ASD and ID, and FamHx of ID and epilepsy, were stronger than the correlation between FamHx of ASD and epilepsy, suggesting consideration of FamHx of ID in investigations of the link between ASD and epilepsy.
Conclusions:
Family morbidity history may pose a significant ASD risk but the majority of persons may lack a family history per morbidity. The family-wide morbidity association approach reveals a comprehensive underlying ASD family morbidity structure and specific morbidity patterns potentially applicable in etiologic analyses.