Comparing the prevalence of substance use disorders between persons with and without autism spectrum disorders

Abstract

Recent research has suggested that people with autism spectrum disorder may be disproportionately at risk of substance use disorders. This study analyzed national-level Medicaid Claims data to compare substance use disorder prevalence among Medicaid enrollees with autism spectrum disorder (N = 388,426) and a random sample of enrollees without autism spectrum disorder (n = 745,699) and to examine whether this association differs across sex and age groups and changes after adjusting for co-occurring mental health conditions. We also examined how the association between autism spectrum disorder and substance use disorder is moderated by co-occurring non-autism spectrum disorder mental health conditions and by community-level social determinants of health by merging Medicaid Claims data with zip code-level US Census data on socioeconomic deprivation. By 2016, 7% of Medicaid beneficiaries with autism spectrum disorder and no intellectual disability had at least one substance use disorder diagnosis, up from 1.75% USD prevalence among enrollees with autism spectrum disorder (no intellectual disability) in 2012 Medicaid data. Individuals with autism spectrum disorder aged 30–64 years were at an elevated risk of cannabis and hallucinogen disorders; this risk is likely compounded by co-occurring mental health conditions, which affect a half of all individuals with autism spectrum disorder and only 23% of individuals without autism spectrum disorder. Research and policy implications are discussed in turn.

Lay abstract

Recent research has suggested that people with autism spectrum disorder may be disproportionately at risk of substance use disorders. The present study analyzed national-level Medicaid Claims data to compare substance use disorder prevalence among Medicaid beneficiaries with autism spectrum disorder and without autism spectrum disorder and to examine whether this association differs across sex and age groups and depends on mental health conditions besides autism. We also examined how the association between autism spectrum disorder and substance use disorder is moderated by co-occurring non-autism spectrum disorder mental health conditions and by community-level social determinants of health. For the latter purpose, Medicaid Claims data were merged with zip code-level US Census data on socioeconomic deprivation. Our analyses demonstrated that, by 2016, 7% of Medicaid enrollees with autism spectrum disorder and no intellectual disability had at least one substance use disorder diagnosis, up from 1.75% USD prevalence among enrollees with autism spectrum disorder (no intellectual disability) in 2012 Medicaid data. Individuals with autism spectrum disorder aged 30–64 years are at an elevated risk of cannabis and hallucinogen use disorders, while this risk is likely compounded by co-occurring mental health conditions, which affect a half of all individuals with autism spectrum disorder and only 23% of individuals without autism spectrum disorder. Research and policy implications are discussed in turn.