Autism spectrum disorder (hereafter referred to as autism) is a neurodevelopmental condition characterised by early-onset social-communication difficulties and repetitive or stereotypical behaviours. Autism has been increasingly recognised, with a current global prevalence of approximately 1%.1 Co-occurring mental health or psychiatric conditions (hereafter referred to as co-occurring mental health conditions) have been frequently reported in individuals with autism—we acknowledge there is no universally accepted single way to refer to individuals on the autism spectrum, and in the stakeholder communities preferences vary between person-first language (ie, individual with autism) and identity-first language (ie, autistic individual),2 herein we use the person-first language because of The Lancet Psychiatry's convention, rather than indicating any superiority over identity-first language—with as many as 70% of people with autism diagnosed with at least one, and almost 50% diagnosed with multiple co-occurring mental health conditions.3, 4 The evolution of classification systems pertaining to how autism and co-occurring mental health conditions are defined and diagnosed, and shifts in official exclusion criteria for diagnosing these conditions in the context of autism, play crucial roles in how the co-occurring conditions are recognised. Additionally, various patterns of co-occurrence might result from several factors, including overlap of specific symptoms of co-occurring mental health conditions and autism (according to current behaviour-based criteria),5 potentially shared underlying mechanisms cutting across diagnostic categories of autism and co-occurring mental health oconditions,6 and life experiences associated with being on the autism spectrum from early in life.7
Research in context
Evidence before this study
Before undertaking this systematic review and meta-analysis, we searched PubMed for observational studies, with no language restrictions or date restrictions on March 15, 2018, using combinations of search terms that included “autis*”, “mental health”, “psychiatr*”, “comorbid*”, “co-occurring”, and “disorder”. We identified a wide variety of observational studies reporting divergent prevalences of mental health symptoms or psychiatric diagnoses (using various tools and definitions) in the autism population, with substantial variability in study design (ie, from small-sample clinical studies to national population-based registry studies), across a wide range of ages, gender compositions, sample proportions with co-occurring intellectual disability, and countries of origin. Most commonly studied conditions include attention-deficit hyperactivity disorder (ADHD), anxiety, depression, bipolar disorder, obsessive-compulsive disorder (OCD), psychosis and sleep difficulties. However, the range of reported prevalence of co-occurring psychiatric diagnoses varies widely (eg, 1–86% for ADHD, 1–70% for depressive disorders). On the basis of these observations, we undertook a systematic review and meta-analysis to determine the best prevalence estimates of co-occurring psychiatric conditions in autism and to identify sources of heterogeneity among studies.
Added value of this study
We did a comprehensive systematic review and meta-analysis of the prevalence of often reported co-occurring mental health conditions in the autism population. Our findings provide prevalence estimates that are valid reflections of the current literature. We identified eight co-occurring psychiatric diagnoses most commonly reported in the literature among people with autism. From eight meta-analyses of 96 studies, we estimated that the pooled prevalences were 28% for ADHD; 20% for anxiety disorders; 13% for sleep–wake disorders; 12% for disruptive, impulse-control, and conduct disorders; 11% for depressive disorders; 9% for obsessive-compulsive disorder; 5% for bipolar disorders; and 4% for schizophrenia spectrum disorders. Estimates in clinical sample-based studies were higher than in population-based and registry-based studies. Age, gender, intellectual functioning, country of study, and publication year were associated with heterogeneity of prevalence estimates in the literature, yet after accounting for these sources, substantial unexplained heterogeneity remained.
Implications of all the available evidence
Co-occurring psychiatric diagnoses are common in people with autism across the lifespan. Mental health promotion, improved assessment, and adequate treatment of mental health concerns are essential for care and support for people with autism.
For people with autism, having co-occurring mental health conditions increases the possibility of worse long-term outcomes,8, 9, 10, 11 including increased mortality risk.12 For instance, co-occurrence of autism and attention-deficit hyperactivity disorder (ADHD) is associated with greater impairments in adaptive functioning, health-related quality of life, and executive functioning than having autism alone.13, 14, 15, 16 Similarly, co-occurring anxiety in individuals with autism amplifies autistic symptoms, including social impairments,17, 18, 19 sensory features, and repetitive behaviours,19, 20, 21, 22 and might be associated with the development of depression,23, 24 contributing to increased risk of suicide and early mortality.25 Aggression, self-injury, and oppositional behaviour might also begin or increase with the onset of depression.26 Autism diagnoses are also associated with increased risk of severe mental illnesses, including psychosis spectrum and bipolar spectrum disorders.27 Co-occurring mental health conditions in autism tend to persist from childhood into adolescence,28 and the prevalence of co-occurring psychopathology increases in adults with autism,29 contributing to substantial long-term negative effects on health and quality of life.
Understanding of the prevalence of co-occurring mental health conditions in people with autism provides crucial background information for adequate mental health evaluation and service delivery. However, prevalence estimates in the literature vary greatly between studies.3 Reported prevalences might differ because of different ascertainment methods or heterogeneity among subgroups of people with autism.30 Furthermore, whether intellectual functioning or gender influences the recognition and diagnosis of co-occurring mental health conditions and to what extent patterns of co-occurring mental health conditions are dependent on age are unclear. Prevalence variation might also result from sociodemographic differences between populations (eg, health-care accessibility or socioeconomic status). We aimed to determine the best estimates of the prevalences of co-occurring mental health conditions in people with autism and identify the extent to which the aforementioned moderators account for heterogeneity among studies.