Comorbidities in autism spectrum disorder and their etiologies

Abstract

Individuals with autism spectrum disorder (ASD), in addition to the core features of the disease, experience a higher burden of co-occurring medical conditions. This study sought to describe the frequency and distribution of comorbidit conditions in individuals with ASD, and systematically evaluate the possibility that pre- and postnatal exposures (e.g., preterm birth, hypoxia at birth, traumatic brain injury, and fetal alcohol syndrome) associated with ASD may also be linked with distinct comorbidities. We used the SPARK study database, launched by the Simons Foundation Autism Research Initiative (SFARI). Comorbidities considered in the study included neurological, cognitive, psychiatric, and physical conditions. The study sample consisted of 42,569 individuals with ASD and their 11,389 non-ASD siblings (full and half siblings). Majority (74%) of individuals with ASD had at least one comorbidity, and had a greater average number of comorbidities than their non-ASD siblings. Preterm birth and hypoxia at birth were the most common peri-natal exposures in the sample. In logistic regression models adjusted for covariates, these exposures were associated with several distinct comorbidities in ASD cases, including attention and behavior problems, psychiatric and neurological disorders, and growth conditions. A similar pattern of association was also observed in non-ASD siblings. Our findings underscore that individuals with ASD experience a greater burden of comorbidities, which could be partly attributable to the higher rates of perinatal exposures compared to their non-ASD siblings. Study findings, if replicated in other samples, can inform the etiology of comorbidity in ASD.

Fig. 1. Barplots illustrating the prevalence (%) of pre- and postnatal exposures by ASD status.

Prevalence estimates in unaffected siblings (grey bars) were standardized to year of evaluation, age and sex distribution in ASD cases (blue bars).

Fig. 2. Barplots illustrating the prevalence of comorbid conditions by ASD status.

Prevalence estimates in unaffected siblings (grey bars) were standardized to year of evaluation, age and sex distribution in ASD cases (blue bars).

Fig. 3. Burden of comorbid conditions in ASD cases and their unaffected siblings.

A Prevalence of individuals with >1 comorbidity by ASD status. B Distribution of the number of comorbidities among those with at least one comorbidity by ASD status.

Fig. 4. Associations between pre- and postnatal exposures and comorbidity in individuals with ASD.

A The model for each comorbidity was adjusted for maternal and paternal education and age at delivery, child’s gender and year of birth, age of the child at evaluation, survey version, race, and annual household income. B In addition to the covariates above, the model for each comorbidity was adjusted for other comorbidities within the same, SPARK-defined disorder domain (domains are demarcated with solid black lines). Blue tiles indicate lower and red tiles higher probability of the diagnosis of the given comorbid condition (y-axis) in individuals with certain environmental exposure (x-axis). Grey tiles indicate the associations that could not be estimated due to insufficient number of exposed individuals with the comorbidity to allow reliable coefficient estimate. ^xp value < 0.05.

Fig. 5

Schematic illustrating potential relationships between genetic vectors, environmental exposures, ASD and its comorbid conditions.