FormalPara Key Summary Points

The overall prevalence of autism spectrum disorder (ASD) among US children aged 3–17 years was 2.94%, with males showing a significantly higher prevalence than females.

Significant racial-ethnic disparities were identified, with Hispanic children having the highest ASD prevalence, followed by non-Hispanic Black and non-Hispanic White children.

Approximately two-thirds of children with ASD received at least one type of treatment, with 39% receiving only behavioral treatment and 21% receiving both behavioral and medication treatment.

Insurance coverage and socioeconomic status were strong determinants of treatment access and ASD prevalence.

The study highlights the ongoing challenges in ASD diagnosis and treatment, particularly regarding racial/ethnic and socioeconomic disparities.

Introduction

Autism spectrum disorder (ASD) is a neurodevelopmental disability stemming from distinct brain differences [1, 2]. Common signs and symptoms of ASD include difficulties in social communication and interaction and repetitive behaviors or interests. Regrettably, there is currently no definitive cure available, with existing treatments focused primarily on enhancing the quality of life by mitigating or improving the symptoms [3]. The complexity of treating ASD arises from its unique impact on each individual patient. Consequently, tailoring a personalized treatment plan becomes imperative, often involving a collaborative effort among healthcare professionals from various medical and psychological disciplines [4]. Some individuals with ASD may require lifelong support in their daily lives, particularly due to their difficulties in grasping social norms and effectively communicating with others. Such challenges also make them more prone to conditions like anxiety, depression, or attention-deficit/hyperactivity disorder, further underscoring the need for comprehensive medical and healthcare support. As a result, the financial implications of ASD-related healthcare are substantial, with individuals in the USA potentially facing lifetime expenses ranging from $1.4 million to $2.4 million, as indicated by reference [5, 6]. Therefore, comprehensively understanding the prevalence and evolving trends in ASD is of paramount importance.

The exact cause of ASD is not yet understood, but prior research has delved into its prevalence, evolving trends, and racial or ethnic disparities among patients with ASD in the USA [7,8,9,10]. In a study by Maenner and colleagues [11], findings from 2016 revealed that non-Hispanic White children exhibited a higher incidence of ASD than non-Hispanic Black children, while the overall prevalence of ASD among Hispanic children was lower than that observed in non-Hispanic Black and White children. Additionally, non-Hispanic Black children with ASD displayed a higher likelihood of having mental disabilities than their non-Hispanic White counterparts and received their first diagnosis at a later age. However, research by Yuan et al. [10] suggests that this gap may be gradually narrowing. Their study showed that between 2014 and 2019, the prevalence of ASD among non-Hispanic Black children increased by 43%, rising from 2.21% to 3.16%, while Hispanic children experienced a 40% increase, from 1.49% to 2.08%. These findings indicate evolving trends in ASD prevalence among different racial and ethnic groups.

There are several treatment options available to support individuals with ASD [3, 12]. Behavioral intervention aims to help individuals with ASD modify their behaviors [12, 13]. Medications, as an intervention approach, may be prescribed to address coexisting physical or mental health conditions often associated with ASD [14, 15]. These medications can help manage symptoms like hyperactivity, inattention, anxiety, aggression, and self-injurious behavior. However, they do not address the core symptoms of ASD. Psychological intervention refers to the psychological support that is valuable for individuals with ASD who may experience mental health issues like depression, anxiety, or other emotional challenges. Other treatment options include developmental intervention, educational intervention, social-relational intervention, and other complementary and alternative interventions. Among these treatment options, behavioral interventions have gained widespread acceptance within the healthcare community due to the robust evidence supporting their effectiveness [16].

This article examines the prevalence and trends of ASD, patterns in treatment options, gender differences, and racial-ethnic disparities in the USA from 2017 to 2020. We achieved this goal by analyzing a nationally representative dataset obtained from the National Survey of Children's Health (NSCH).

Methods

The NSCH (https://mchb.hrsa.gov/data-research/national-survey-childrens-health) is a leading national survey conducted annually by the US Census Bureau that produces national and state-level data for key measures of child health and well-being in the USA. The NSCH provides rich data on multiple, intersecting aspects of children’s lives, which are essential to understanding the health status and health service needs of children across the USA. We have obtained permission to use the NSCH survey data. This article was based on a previously conducted National Survey of Children's Health (NSCH) study in the USA, and did not contain any new studies with human participants or animals performed by any of the authors.

Data Collection

NSCH surveys were conducted with an identical design and administration from 2017 to 2020. Households are randomly sampled and contacted by a mailed invitation, asking an adult in the household who is familiar with the child’s health and healthcare to complete a short screener questionnaire. The NSCH used a national sample of physical addresses to recruit participants [17]. The participants were asked whether one or more children aged 0–17 years were living in the household. If yes, then the responder was asked to fill out an initial screener with the age and sex information of all children in the household. Only one child from each household was randomly selected to be the subject for the age-specific main topical questionnaire. The survey questions included child and family demographics, child’s current health status, child’s health status as an infant, healthcare access and utilization, insurance coverage, early childhood (0–5 years) issues, issues specific to middle childhood and adolescence (6–17 years), family functioning, parental health status and family, and neighborhood and community characteristics [18]. The survey oversampled children with special healthcare needs and children 0–5 years of age. A total of 21,599, 30,530, 29,433, and 42,777 surveys were completed nationally from 2017 to 2020, respectively. The overall weighted response rate from 2017 to 2020 ranged from 37.4% to 42.4% [19].

Study Data

In the NSCH questionnaire, the participant was asked “Has a doctor or other health care provider ever told you that this child has autism or autism spectrum disorder (include diagnoses of Asperger’s disorder or pervasive developmental disorder)?” If the answer was yes, then the participant was further asked whether the child currently had the condition. In addition, the participant was asked whether the child was currently taking any medication for the condition and if the child had received behavioral treatment at any time during the past 12 months.

Here, we included all children aged 3–17 years with information on their ASD status in the study data. The weighted prevalence of ASD was estimated. Furthermore, weighted ASD prevalence was evaluated by individual characteristics, including age (3–5, 6–11, and 12–17 years), sex (male, female), self-reported race/ethnicity (Hispanic, non-Hispanic White, non-Hispanic Black, non-Hispanic Asian, and other), family income-to-poverty ratio (< 1.0, 1.0–1.9, 2.0–3.9, ≥ 4.0), and insurance type (private only, public only, both, not insured). We specifically investigated (1) race/ethnicity-specific prevalence for any indication of racial/ethnic disparities in the prevalence of ASD and (2) the pattern in types of treatment of ASD (medication treatment, behavior treatment, receiving neither medication treatment nor behavior treatment).

Statistical Analysis

Survey weights, strata, and primary sampling units that were provided along with the NSCH data were used to estimate the prevalence of ASD and treatment options and their 95% confidence intervals (CI). A chi-square test was performed to assess whether there was any difference in overall prevalence or difference in prevalence across strata. Weighted logistic regressions utilizing sampling weights were performed to examine treatment options for ASD across strata. All analyses were performed using the survey procedure in SAS version 9.4 to account for the complex sampling design of the NSCH. A two-sided test with a p-value of <0.05 is considered statistically significant.

Results

Prevalence Trends and Gender Differences of Autism Spectrum Disorder

Between 2017 and 2020, a total of 108,142 children and adolescents aged 3–17 years participated in the NSCH, with 3287 of them reported having ASD. The median age of first ASD diagnosis was 4.0 years, with an interquartile range of 3.0–7.0. Notably, 64.0% of those diagnosed received their ASD diagnosis at age 5 or younger. The overall prevalence of ASD was 2.94% (CI 2.68–3.18) (Table 1). Further analysis revealed that older age was significantly linked with a higher prevalence of ASD (p < 0.0001).

Table 1 Weighted prevalence (counts [N], percentage [%] with confidence interval [CI]) of autism spectrum disorder in US children and adolescents aged 3–17 by year from 2017 to 2020
Full size table

In addition, males had a significantly higher prevalence of ASD than females (4.54% vs. 1.26%, p < 0.0001), showing significant gender differences. Additionally, a significant decreasing trend in ASD prevalence (from 3.83% to 2.06%) was observed with an increase in the family income-to-poverty ratio (from < 1.0 to ≥ 4.0, p < 0.0001). Regarding insurance coverage, children with both public and private insurance had the highest prevalence of ASD (8.97%, CI 7.00–10.94), significantly higher than those with public insurance only (3.93%, CI 3.31–4.55) and private insurance only (2.07%, CI 1.86–2.28, p < 0.0001). Children with no insurance coverage had the lowest prevalence (1.88%, CI 0.99–2.77). Among the 50 states and the District of Columbia, the overall prevalence of ASD over the 4-year period ranged from the lowest in North Dakota (1.60%, CI 0.96–2.24) to the highest in Delaware (4.16%, CI 2.76–5.56) (Fig. 1).

Fig. 1
figure 1

State-level weighted prevalence of autism spectrum disorder in US children and adolescents aged 3–17 from 2017 to 2020

Full size image

Regarding the time trend, the overall prevalence of ASD remained relatively stable from 2017 to 2020, showing no significant downward or upward trend (p = 0.7044) (Fig. 2). Among the 6–11-year and 12–17-year age groups, the prevalence also remained relatively stable (p = 0.1681 and 0.8252, respectively), while there was a 47% increase in the 3–5-year age group, rising from 1.56% (CI 0.84–2.28) to 2.29% (CI 1.37–3.21) (p = 0.0106). The prevalence in males and females remained stable as well (p = 0.9161 and p = 0.5495, respectively). However, there was a significant increasing trend in ASD prevalence among the non-Hispanic White population (p = 0.0256), increasing from 2.51% (CI 2.10–2.91) to 2.96% (CI 2.59–3.34). As for treatment trends among patients with ASD, there were some fluctuations in the percentage of individuals receiving behavioral treatment, but no significant linear trend was detected (p = 0.4387). Similarly, the percentage of individuals receiving medication treatment (p = 0.5160) and neither treatment (p = 0.5627) also showed no significant changes over the study period.

Fig. 2
figure 2

Trends in the prevalence of autism spectrum disorder among US children and adolescents aged 3–17 from 2017 to 2020

Full size image

Racial-Ethnic Disparities in Autism Spectrum Disorder

The overall prevalence of ASD showed significant disparities among different racial/ethnic groups. The highest prevalence was observed in the Hispanic population (3.44%, CI 2.65–4.23), followed by non-Hispanic Black individuals (3.11%, CI 2.49–3.74) and non-Hispanic White individuals (2.77%, CI 2.56–2.97). The lowest prevalence was found in the non-Hispanic Asian population (1.97%, CI 1.37–2.56). This racial/ethnic disparity was statistically significant (p = 0.0292) (Table 1). When analyzing different age groups, the prevalence of ASD did not differ significantly by race/ethnicity in either the 3–5-year or 6–11-year age groups. However, in the 12–17-year age group, the highest prevalence was observed in the Hispanic population (3.60%, CI 2.32–4.88, p < 0.0001) (Table 2).

Table 2 Weighted prevalence (counts [N], percentage [%] with confidence interval [CI]) of autism spectrum disorder in US children and adolescents aged 3–17 by race/ethnicity from 2017 to 2020
Full size table

No significant race/ethnicity disparities were found among different gender groups. However, when considering the family income-to-poverty ratio, Hispanic individuals showed significantly higher prevalence than others when the ratio was between 1.0 and 1.9 (4.30%, CI 2.53–6.07, p = 0.002). Furthermore, among those with both public and private insurance coverage, Hispanic individuals also exhibited a significantly higher prevalence of ASD (9.93%, CI 3.91–15.95, p < 0.0001). When children had no insurance coverage, non-Hispanic individuals from other racial/ethnic groups showed a significantly higher prevalence of ASD than their counterparts (3.18%, CI 0.00–6.53, p = 0.03). Otherwise, no significant difference in ASD prevalence by race/ethnicity was detected.

Treatment Use in Autism Spectrum Disorder

Table 3 presents the overall and characteristic-specific percentages of treatment use for ASD patients, categorized into non-mutually exclusive treatment options: behavioral treatment, medication treatment, and neither treatment. However, the table does not include the following important findings: nearly two-thirds of patients with ASD received at least one type of treatment (66.9%, CI 65.3–68.5); 39.0% (CI 37.3–40.7) received behavioral treatment only; 6.9% (CI 6.1–7.9) received medication treatment only; and 21.0% (CI 19.6–22.4) received both behavioral and medication treatment. Over the period from 2017 to 2020, there was no significant change in the rates of receiving behavior treatment, medication treatment, or neither for children with ASD. However, the percentage of individuals receiving behavioral treatment decreased significantly with age, declining from 74.64% in the 3–5-year age group to 51.20% in the 12–17-year age group (p = 0.0004). Conversely, the utilization of medication treatment showed an increasing pattern (from 5.44% to 32.56%, p < 0.0001), as did the group receiving neither treatment (from 25% to 40%, p = 0.037). Older age was associated with a higher percentage of individuals receiving neither treatment (from 24.61% to 40.24%, p = 0.0371).

Table 3 Weighted percentage of treatment use (counts [N], percentage [%] with confidence interval [CI]) among US children and adolescents aged 3–17 with autism spectrum disorder from 2017 to 2020
Full size table

Regarding gender differences, males were more likely to receive behavioral treatment (61.25% vs. 48.83%, p = 0.0145), while females were more likely to receive neither treatment (45.57% vs. 33.09%, p = 0.0152). However, the percentages of individuals receiving behavioral treatment or neither treatment did not differ significantly among different race/ethnic groups. On the other hand, for medication treatment, non-Hispanic Asians had the lowest percentage (8.94%), while non-Hispanic Blacks had the highest (33.84%, p = 0.0003). The binary logistic regression analysis (Table 4) confirmed the association of treatment options with age, sex, and race/ethnicity, as indicated by the odds ratios. Furthermore, the family income-to-poverty ratio was significantly associated with receiving behavior treatment and neither treatment. Individuals in families with a ratio greater than 4 were more likely to receive behavior treatment and less likely to receive neither treatment than those with a ratio less than 1. Insurance type did not show a significant impact on receiving behavioral treatment or neither treatment. However, individuals in families with private insurance only or both private and public insurance were more likely to receive medication treatment than those with public insurance only.

Table 4 Weighted binary logistic regression for dichotomized treatment use (odds ratio with confidence intervals [CI]) among US children and adolescents aged 3–17 with autism spectrum disorder from 2017 to 2020
Full size table

Discussion

This study delves into the trends in prevalence and racial-ethnic disparities among children and adolescents aged 3 to 17 with ASD in the USA between 2017 and 2020. It provides a valuable comparison to prior research. Previously, Xu and colleagues [20, 21] examined the prevalence of ASD in the USA from 2014 to 2016 and reported a rate of less than 2.50%. However, our analysis reveals a significant increase in the prevalence of ASD from 2017 to 2020, with rates ranging from 2.74% to 3.13%, indicating a stable upward trend. Previous research consistently indicates gender differences in the prevalence of ASD. These studies typically reveal a higher proportion of male ASD cases than female cases, with gender ratios typically ranging from 4:1 to 3:1 [22,23,24,25,26]. In our study, we observed that the incidence rate among male children remained more than three times higher than that among females. This discrepancy is notable, as it suggests that female children with ASD continue to receive less treatment, indicating that this issue may not have been effectively addressed.

Xu et al. [21] reported in 2016 that the prevalence of ASD was highest among non-Hispanic Black children in the USA, at 2.79%. In our study, we observed a different trend, with the highest prevalence among Hispanic individuals, reaching 3.44%, closely followed by non-Hispanic Black children with a prevalence rate of 3.11%. This suggests that racial-ethnic disparities in ASD prevalence have not shown significant improvement. Furthermore, Zhang et al. conducted an analysis of NSCH data spanning from 2011 to 2012 [27]. Their findings revealed that families with ASD children who relied on private health insurance faced out-of-pocket expenses that were 11 times higher than those with public health insurance. While some states have introduced exemptions to enhance access to public health insurance for individuals with ASD and have reformed private health insurance to improve ASD service coverage [28], our analysis of 2019 data indicates that families with private insurance were still less likely to have their necessary services covered than those with public insurance. We also noted a significantly lower prevalence rate of ASD among individuals without insurance, suggesting that these families may lack adequate access to medical services for diagnosis and treatment, contributing to this disparity. Collectively, these studies underscore the critical role of public insurance in the diagnosis and treatment of children with ASD.

The racial-ethnic disparities in ASD prevalence in the USA can be attributed to several interrelated factors. One major reason is inequitable access to healthcare, where minority populations, particularly Black, Hispanic, and Native American communities, often face barriers such as lower insurance coverage, fewer available specialists, and geographical limitations in accessing services. Additionally, cultural differences and language barriers can influence how symptoms are perceived by parents and healthcare providers, potentially leading to misdiagnosis or delayed identification of ASD in minority children. Implicit bias in the healthcare system can also result in unequal treatment, with clinicians potentially underestimating or overlooking ASD symptoms in minority children compared to their White peers. Furthermore, socioeconomic disparities contribute significantly, as families from lower-income backgrounds may not have the resources to seek specialized care or advocate for their children in complex medical systems. Together, these factors create significant gaps in early diagnosis and intervention, ultimately contributing to the observed racial disparities in ASD prevalence.

The treatment of ASD remains a significant challenge, and to date, there have been no groundbreaking developments in ASD treatment. Behavioral therapy continues to be the primary and most crucial treatment option, whereas pharmacological interventions are typically utilized to address comorbid conditions rather than the core deficits of ASD [29, 30]. Nationally, data from 2016 indicate that approximately 70% of children with ASD receive only behavioral treatment, while 6.9% receive only medication [21]. About 20.3% of children receive a combination of both behavioral treatment and medication, leaving a substantial 29.5% of children with current ASD without access to either behavioral treatment or medication. Our study's findings align with these statistics. Our research also reveals that the utilization of behavioral treatment tends to decrease as individuals with ASD age. This decline may be attributed to the effectiveness of early interventions, which can contribute to a reduction in core ASD symptoms over time [31,32,33,34]. As individuals with ASD age, they often encounter additional comorbidities, such as attention deficit hyperactivity disorder and depression [35,36,37,38]. This increase in comorbid conditions may explain the elevated use of medication treatment observed in our study. In summary, the treatment landscape for ASD remains a complex issue, with behavioral therapy at the forefront and medication primarily addressing associated conditions. The shift towards medication treatment as individuals with ASD age may be attributed to the emergence of these comorbidities.

As of April 2021, the US Food and Drug Administration (FDA) had approved two antipsychotic drugs to treat ASD-associated irritability in children between certain ages. Risperidone is the first FDA-approved medication for the treatment of symptoms associated with ASD in children and adolescents aged 5–16, including aggressive behavior, deliberate self-injury, and temper tantrums [39, 40]. For patients aged 6–17, aripiprazole is another FDA-approved medication for the treatment of irritability in children and adolescents with ASD. Owen et al. (2009) [41] found that when a group of 98 children were randomized to treatment and placebo groups, those taking aripiprazole experienced a 25% or greater reduction in autism-related irritability symptoms by week 8 of the trial, compared with 14% of those who took the placebo. However, while these drugs can help with related symptoms, they do not target the underlying causes of autism. Other drugs or therapies may also be used to treat behaviors associated with ASD, or seizures, which are common in children with autism. However, many of these drugs are not approved by the FDA for treating autism specifically, and some are not approved for people under 18 years old [42].

The racial-ethnic disparity in ASD treatment use may stem from a variety of socioeconomic, cultural, and systemic factors. Cultural stigmas around mental health and developmental disorders may also deter some families from seeking diagnosis or treatment. Additionally, biases within the healthcare system, including lower expectations or misdiagnosis of autism in minority children, contribute to delayed diagnoses and underutilization of treatment options. In addition, gender disparities in autism treatment use are largely driven by differences in diagnosis rates and the ways autism presents in males versus females, as ASD has historically been seen as more common in males. This gender bias in diagnosis has a direct impact on treatment access. The implications of this disparity are profound. Children from minority backgrounds may miss crucial early interventions that significantly improve long-term outcomes. Without equitable access to behavioral therapies or appropriate medications, these children are at risk for greater social, emotional, and cognitive difficulties, perpetuating cycles of disadvantage and reinforcing health inequalities in the broader population.

Our study drew upon nationally representative health survey data collected over four consecutive years, offering a substantial sample size and a diverse range of racial representation. This diversity enhances the representativeness of our analysis results. Nevertheless, it is important to acknowledge certain limitations. Firstly, the information regarding physician-diagnosed cases of ASD relied on self-reporting from parents or guardians. This methodology introduces the potential for misreporting and recall bias. Secondly, while standardized diagnostic scales were employed for ASD diagnoses, the possibility of underdiagnosis or misdiagnosis remains, especially when healthcare professionals or institutions vary in their levels of experience and expertise. However, the extensive sample size of our study is likely to mitigate the impact of these issues on our results. Lastly, we were unable to examine the presence of several other comorbid conditions, such as anxiety, aggressive behavior, or deliberate self-injury, due to the limitations of the NSCH survey design, which did not account for these aspects. However, the NSCH does include survey questions related to attention-deficit/hyperactivity disorder (ADHD). The comorbid presence of ADHD in a population of children diagnosed with ASD has been studied intensively [43, 44]. Its estimates range widely from 37% to 78% [45].

Conclusions

Our study reveals a consistent overall trend in the prevalence of ASD in the USA from 2017 to 2020, with a significant increase noted among younger children, hinting at a possible upturn in ASD incidence. Together with the report from Yan et al. [46] for the two years 2021 and 2022, it underscores the importance of closely monitoring and understanding this trend. Gender differences and racial-ethnic disparities in ASD prevalence underscore the need for a more comprehensive understanding of ASD in females and in minorities to expedite timely diagnosis and treatment. Although there has been some headway in healthcare reforms related to ASD, the outcomes may still fall short, potentially causing delays in diagnosis and treatment for affected individuals. Enhancing access to public health insurance may be a necessary step to effectively addressing these challenges and ensuring that individuals with ASD receive the care and support they require.