Autism Spectrum Disorder: DSM-5 Changes, Epidemiology, and OutcomesThe definition of autism has evolved over the seven decades since Leo Kanner first described this unique syndrome in 1943. Variously referred to as “childhood schizophrenia” in the 1950s and 1960s, and “infantile autism” in the 1980s, it was ultimately classified as a “pervasive developmental disorder” (PDD) in DSM-IV-TR. DSM-5, published in 2013, includes a new, all-encompassing category, “autism spectrum disorder” (ASD), which incorporates autism and a number of autism-related disorders, including Asperger’s syndrome (Singer E, Nature 2012;491(7422):S12–S13).

ASD (a term which replaces pervasive developmental disorders or PDDs in DSM-5) is a lifelong neurodevelopmental disorder appearing during early childhood. Typically occurring before three years of age, ASD is characterized by significant social difficulties associated to varying degrees with communication and behavioral problems. In DSM-IV-TR, autistic disorder was classified into three distinct domains:

1. Impairment in social interaction skills: deficits in nonverbal behaviors, peer relationships, and interest in one’s environment.

2. Impairment in communication: delay in development of spoken language, idiosyncratic language use, or absence of spontaneous play.

3. Restrictive or repetitive behavior or interests: inflexible adherence to routines, stereotyped motor activities, and preoccupation with parts of objects.

In DSM-IV-TR, the PDDs included autistic disorder; Asperger’s syndrome; childhood disintegrative disorder; Rett’s disorder; and pervasive developmental disorder not otherwise specified. The introduction of the ASD category is an attempt to improve both sensitivity—diagnosing those who have the disorder, and specificity—excluding those who don’t. While DSM-IV included the three domains listed previously, the new classification in DSM-5 will lump social and communication deficits together as problems in “social communication and interaction” (SCI) (Tanguay PE, Am J Psychiatry 2011; 168(11):1142–1144). The criteria for both “social communication and interaction” and “fixated interests and repetitive behavior” must be met to be diagnosed with ASD.

The merging of distinct diagnostic categories has raised concerns that some children may no longer meet criteria for ASD according to DSM-5, and, as such, might not receive access to services. Some studies have shown that while the proposed criteria entail excellent specificity, they may be too restrictive, with only 60% or fewer of those meeting DSM-IV criteria receiving a diagnosis under DSM-5 (Gibbs V et al, J Autism Devel Dis 2012;42(8):1750–1756). Reports indicate particularly low sensitivity for individuals diagnosed with Asperger’s syndrome or PDD-NOS (Mazefsky CA, et al, J Autism Devel Dis 2013;43(5):1236).

On the other hand, supporters of DSM-5 counter that the changes are based on a better understanding of the condition, and that preliminary studies indicate that few will be excluded (Singer E, op.cit). However, comparison studies will be needed to ascertain the true effect.

In addition to ASD, a separate diagnosis of social (pragmatic) communication disorder (SCD) has been added to DSM-5. This diagnosis will include some individuals who were vaguely diagnosed as PDD-NOS under DSM-IV. Individuals with social communication disorder have deficits in the “social uses of verbal and nonverbal communication” but do not exhibit the repetitive behavior patterns and fixated interests required for ASD (Lohr WD and Le JF, Pediatric Annals 2012; 41(10):1–3).

Finally, DSM-5 introduces levels of severity for the dimensions of ASD, in addition to the categorical (ie, checklist-based) system from DSM-IV. This reflects observations from population and family studies of a broad distribution of autism symptoms in the population, and elevated levels of autistic traits in siblings and other family members of affected cases (Frazier TW et al, J Am Acad Child Adol Psychiatry 2012;51(1):28–40). Whether a dimensional system will affect rates of ASD diagnosis will be determined by studies evaluating subjects with both DSM-IV and DSM-5 criteria. Preliminary results from one such study funded by APA suggest that the new DSM-5 criteria are unlikely to exclude many people with ASD (Huerta M et al, Am J Psychiatry 2012;169(10):1056–1064).


There has been a rapid increase in the prevalence of autism and related disorders. A 1966 English study estimated the prevalence of autism at 4.5 per 10,000 using Kanner’s criteria (Lotter V, Social Psychiatry 1966;1:124–137). After DSM-IV was introduced in 1994, studies found rates of autistic disorder as high as 15 to 40 per 10,000 (Wing L and Potter D, in Assessment of Autism Spectrum Disorders, 2009; New York, NY: Guilford Press, 18–54). Recent research shows a prevalence of 70 to 113 per 10,000, or roughly 1% of the population, a tenfold increase in last 40 years (Fombonne E, Pediatr Res 2009;65(6):591–598). And a 2013 CDC study shows parent-reported rates of 2% of children ages 6 to 17 (Blumberg SJ et al, National Center for Health Statistics. National Health Statistics Reports. March 20, 2013;

After intellectual disability, ASD is the second most common and fastest growing developmental disability in the US. Each year, more children are diagnosed with ASD than AIDS, diabetes, and cancer, combined. Some factors contributing to this increase in prevalence may include the broadening of diagnostic criteria and practice, increased awareness of ASD among the professional community and the public, and greater availability of treatment services.

ASD tends to run in families. There is a 20- to 50-fold higher risk of recurrence in future births when one child has been diagnosed with ASD (O’Roak BJ and State MW, Autism Res 2008;1:4–17). Concordance rates are estimated to be 60% to 90% in monozygotic twins, compared to 10% in dizygotic twins (Bailey et al. 1995). The mode of inheritance seems to be complex, with environmental factors like paternal age and maternal infections during pregnancy also playing a role. Males generally have a higher prevalence rate of ASD than females, with a male-to-female ratio of 4:1 for autistic disorder and 9:1 for Asperger’s syndrome (Centers for Disease Control and Prevention, MMWR 2012;61:1–19). Females with ASD tend to have greater social, communicative, and cognitive impairments than males (Banach R, et al, J Autism Devel Disorders 2009;39:188–193), although the core symptoms of ASD present similarly in both men and women (Rivet TT and Matson JL, J Devel Phys Disabilities 2011;23:399–420).

Clinicians can identify ASD in children as young as two years old, although children from ethnic minority groups are usually diagnosed at a later age than their Caucasian counterparts. ASD is commonly comorbid with attention-deficit hyperactivity disorder, anxiety disorders, intellectual disability, epilepsy and other genetic conditions like fragile X syndrome, tuberous sclerosis, neurofibromatosis, congenital rubella syndrome, Down syndrome, Prader-Willi syndrome, and Angelman syndrome. Until recently, there was little, if any, epidemiological research focusing on the prevalence of ASD in adults. In 2011, one study reported the prevalence of ASD in an adult sample to be 1%, with higher rates for men (1.8%) than women (0.2%) (Brugha T et al, Arch Gen Psych 2011;68:459–466).

Similarly, there are few studies evaluating outcomes and prognosis for adults with ASD. Given current prevalence rates, the population of adults with autism is expected to rise 625% by the year 2030, and the estimated lifetime cost per individual with autism, including caregiving costs and lost productivity, can reach up to $3.2 million (Ganz M, Arch Ped Adol Med 2007;161(4):343–354).

Longitudinal Course

As psychiatrists, we should know what to tell patients and their families when they ask about the course of ASD. Unfortunately, the disease course varies widely, and we need to acquire further knowledge of its naturalistic course to gauge the effect of current treatment modalities. Current research suggests that the severity of core symptoms of ASD declines over time (Shattuck PT et al, J Autism Devel Disord 2007;37:1735–1747), with restricted and repetitive behaviors less frequent and less severe among older than younger individuals. Adolescents with ASD tend to demonstrate better social function compared to adults. However, measures related to quality of life remain poor throughout the lifespan.

Which factors might predict outcomes for individuals with ASD in adulthood? Academic performance and cognitive ability are two such factors, but existing evidence suggests that they are not good predictors for better functioning or quality of life (McGovern C and Sigman M, J Child Psychol and Psychiatry 2005;46:401–408). A 10-year study of patients with ASD and normal IQ showed that daily living skills improve during adolescence but plateau around late 20s and start to decline in early 30s (Smith L et al, J Am Acad Child Adol Psychiatry 2012;51(6):622–631).

A meta-analysis of 23 outcome studies of ASD published between the late 1960s and 2011 indicates that the majority of people with ASD, including those with normal IQs, were at a disadvantage regarding employment, social relationships, physical and mental health and quality of life (Howlin P and Moss P, Can J Psych 2012;57(5):275–283). A follow-up study on one population from this study found that the majority of 60 adults with ASD (mean age 44 years) continued to have problems with social integration, and became more socially isolated as they grew older. They also suffered from greater rates of unemployment and lower incomes, and most were living with parents or in other specialist accommodations. Only a minority of individuals had ever been able to make close friends or develop intimate relationships (Howlin P et al, J Am Acad Child Adol Psychiatry, 2013, in press).

Psychiatric and neurologic outcomes in old age for individuals with ASDs remain unknown. There is, therefore, a great need for research into effective interventions to improve outcomes for adults with ASD. Childhood factors that might predict outcome and prognosis of ASD in early and mid-adulthood remain unidentified.

TCPR’s VERDICT: ASD includes a heterogeneous group of diagnoses that are widespread and carry a high public health cost. It’s essential that we advocate and search for new treatment modalities to improve the quality of life for individuals with ASDs throughout the lifespan.


This article originally appeared in:

The Carlat Psychiatry Report
Click on the image to learn more or subscribe today!

This article was published in print 6/2013 in Volume:Issue 11:6.