Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jan 1.
Published in final edited form as: Autism. 2022 May 24;27(1):226–243. doi: 10.1177/13623613221091316

Suicidal Ideation and Intentional Self-Inflicted Injury in Autism Spectrum Disorder and Intellectual Disability: An Examination of Trends in Youth Emergency Department Visits in the US from 2006–2014

Paige E Cervantes a, Derek S Brown b, Sarah M Horwitz a
PMCID: PMC9684352  NIHMSID: NIHMS1790064  PMID: 35608134

Abstract

Substantial efforts have been dedicated to understanding, assessing, and managing suicide risk in youth broadly. However, little attention has been focused specifically on autistic youth and youth with intellectual disability (ID). Because emergency departments (EDs) are an essential point of suicide-related care, we used the National Emergency Department Sample databases to explore differences in prevalence of US ED visits with a suicidal ideation or intentional self-inflicted injury ICD-9 diagnostic code by autistic youth, youth with ID, and youth without these diagnoses (i.e., the comparison group). ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more prevalent in autistic youth (5.1%) and youth with ID (6.6%) than in the comparison group (1.2%). Similar results were found when examining visits with a suicidal ideation diagnosis and with an intentional self-inflicted injury diagnosis separately. Prevalence of these ED visits increased more from 2006–2014 in autistic youth and in youth with ID than in the comparison group and were correlated with common and distinct sociodemographic and clinical factors across groups. Results suggest autistic youth and youth with ID may be uniquely vulnerable to suicide risk, highlighting the urgency of addressing suicidality and self-harm in these groups, particularly within ED settings.

Keywords: Autism Spectrum Disorder, Intellectual Disability, Youth Suicide, Emergency Department, NEDS

Lay Summary

Youth suicide is a major problem in the US and globally, but little is known about suicide risk in autistic youth and youth with intellectual disability specifically. Using data from the National Emergency Department Sample, which is the largest database of emergency department visits in the US, we found that emergency department visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more common in autistic youth and youth with intellectual disability than in youth without these diagnoses (i.e., the comparison group). This was true when examining both suicidal ideation diagnoses and intentional self-inflicted injury diagnoses at emergency department visits. In addition, the number of emergency department visits with a suicidal ideation or intentional self-inflicted injury diagnosis increased more from 2006–2014 in autistic youth and youth with intellectual disability compared to the comparison group. We also found both similarities and differences when examining factors, such as age, sex, and co-occurring mental health conditions, related to emergency department visits with a suicidal ideation or intentional self-inflicted injury diagnosis across groups that may be helpful for understanding suicide risk. It is urgent that we improve our understanding, assessment, and treatment of suicidality and self-harm in these groups through more research and clinical efforts.

Over the past decade, the suicide rate has increased 56% and is now the second leading cause of death in youth ages 10–24 years (Curtin & Heron, 2019). This increase in the prevalence of suicide co-occurs with a 60% increase in US emergency department (ED) visits related to mental health disorders and a 329% increase in visits related to deliberate self-harm in individuals under 18 years (Lo et al., 2020). These increases in major mental health challenges have prompted numerous published recommendations for assessing and managing youth suicide risk across clinical settings (Brahmbhatt et al., 2019; Horowitz, Boudreaux, et al., 2018; Shain & AAP Committee On Adolescence, 2016), the development and implementation of several community-based suicide prevention programs (Schilling et al., 2016; Wyman et al., 2010), and increases in the National Institutes for Health (NIH) research funding for suicide and suicide prevention (NIH, 2020).

While this focus on addressing child and adolescent suicide broadly is important, youth who might be particularly vulnerable to suicidality, such as autistic youth and youth with intellectual disability (ID), have received much less attention. A recent body of literature suggests that suicidal ideation and self-harm are more prevalent (Cassidy et al., 2018; Cassidy & Rodgers, 2017; Chen et al., 2017; Demirkaya et al., 2016; Horowitz, Thurm, et al., 2018; Kõlves et al., 2021; McDonnell et al., 2019; Oliphant et al., 2020; Segers & Rawana, 2014) and the incidence of suicide death is higher (Hirvikoski et al., 2016; Kirby et al., 2019; Kõlves et al., 2021) in the autism spectrum disorder (ASD) population than in the general population. Further, once erroneously thought to be protected from suicide because of an inability to comprehend the concept, it is now recognized that individuals with ID experience suicidality (Dodd et al., 2016; Ludi et al., 2012; Merrick et al., 2005; Mollison et al., 2014). However, findings on the co-occurrence of ID and suicide risk are inconsistent. While some research has demonstrated that rates are lower in the ID population, several studies have found that rates in ID are similar to or higher than in the general population (Dodd et al., 2016; Erlangsen et al., 2020; King et al., 2019; Nagraj & Omar, 2017; Salvatore et al., 2016). These inconsistent findings may be attributable to sampling differences, including the fact that different levels of ID may be associated with different aspects of suicidality (e.g., suicidal ideation versus self-inflicted injury), as well as methodological differences, including the assessment of intentionality to determine the relationship of self-injury to suicidality (Dodd et al., 2016). It is also important to note that ASD and ID commonly co-occur, and emerging research suggests that autistic individuals with ID experience increased suicide risk as well. For instance, Hirvikoski et al. (2016) found that while rates were lower compared to autistic individuals without ID, autistic individuals with ID had significantly higher rates of suicide death than non-autistic individuals.

Because suicide risk may present differently in autistic individuals and individuals with ID compared to the general population, it is possible that findings from research with the general population may not be generalizable to the ASD and ID populations. Standard assessment and management practices likely need to be adapted for the ASD and ID populations to increase acceptability and effectiveness (Cassidy & Rodgers, 2017; Dodd et al., 2016; International Society for Autism Research, 2021; Jachyra et al., 2021; Ludi et al., 2012). To inform the development of clinical adaptations for these populations, we must improve our understanding of suicide risk presentation in ASD and ID. Much of the current research focuses on adults, which is an issue because findings may not be generalizable to children and adolescents and because identifying suicide risk in youth is important for early intervention. Further, as described in Kirby et al. (2019), most of these studies use convenience or clinical samples. While there is emerging research using similar methods to understand suicidality in autistic adults (Hand et al., 2020), no study has employed a nationally representative sample to assess suicidal ideation and self-harm in youth using US healthcare claims data. These studies are imperative, as the sample would represent those most likely to seek care, and thus results hold clinical utility for providers.

Therefore, the present study used a nationally representative database of US emergency department (ED) visits to: (1) examine differences in the prevalence of ICD-9 diagnoses of suicidal ideation and intentional self-inflicted injury in ED visits by autistic youth, youth with ID, and youth without ASD or ID diagnoses (i.e., the comparison group); (2) compare changes in the rate of suicidal ideation and intentional self-inflicted injury diagnoses in ED visits across groups over time; and (3) identify and compare sociodemographic and clinical correlates for ED visits with these diagnoses within and across groups. The National Emergency Department Sample (NEDS), a collection of hospital-based ED databases developed through the Healthcare Cost and Utilization Project by the Agency for Healthcare Research and Quality, was used to examine these issues because the ED is a point of frequent contact for youth with psychiatric concerns (Lo et al., 2020), and without validated instruments (Howe et al., 2020; Ludi et al., 2012), medical record review offers a way to examine suicidal ideation and self-harm in the ASD and ID populations, although it is likely to be a conservative estimate in these groups. Given that autistic individuals and individuals with ID are often more likely to experience common risks for suicidality (e.g., trauma and bullying, mental health conditions, social isolation; Dodd et al., 2016; Ludi et al., 2012; McDonnell et al., 2019; Mollison et al., 2014), and that unique factors may also exist and influence risk in these groups (e.g., cognitive level, deficits in emotion awareness, attempts to camouflage; Cassidy et al., 2018; Costa et al., 2020; Dodd et al., 2016; Ludi et al., 2012), we expect the prevalence of suicidal ideation and intentional self-inflicted injury diagnoses to be higher in ED visits by autistic youth and youth with ID than in ED visits by the comparison group, and the rate of these diagnoses at ED visits to have increased similarly to or more than the rate in the comparison group. Correlates of ED visits with suicidal ideation and intentional self-inflicted injury diagnoses (e.g., patient demographics, clinical presentation, hospital characteristics) may be similar or distinct across groups.

Methods

Data Source

We used data from the NEDS databases, developed for the Healthcare Cost and Utilization Project by the Agency for Healthcare Research and Quality. The NEDS is the largest all-payer ED database in the US and contains data from over 30 million ED visits each year. Data includes ED visit details, patient demographics and clinical diagnoses, and hospital characteristics, excluding elements that could directly or indirectly identify patients, hospitals, or sites (Healthcare Cost and Utilization Project NEDS, 2006–2014). In the current study, we analyzed select data from ED visits of youth aged 7–25 years old in the 2006 to 2014 databases.

Variables of Interest

Diagnostic groups.

We divided youth ED visits into three groups, the ASD group, the ID group, and the comparison group (i.e., no ASD or ID diagnoses) using ICD-9 codes. Visits with any of the four pervasive developmental disorders (PDD) diagnoses (i.e., autistic disorder [299.0], childhood disintegrative disorder [299.1], as well as specific [299.8] and nonspecific [299.9] pervasive developmental disorders) present among the 15 potential diagnoses coded were assigned to the ASD group, regardless of ID status. ED visits by autistic youth with co-occurring ID were maintained in the ASD group because these visits represented a small minority of the group (~9%) and because maintaining these youth visits was considered a more cautious approach than excluding them. As prevalence of co-occurring ASD and ID would be expected to be closer to 30% based on data from the Autism and Developmental Disabilities Monitoring (ADDM) Network in the US (Maenner et al., 2020), it is possible that ED clinicians failed to code the co-occurring diagnosis of ID in some cases of ASD. Visits with any of the ID codes (i.e., 317 [mild ID], 318 [other specified ID], and 319 [other, unspecified ID]) present among all potential diagnoses coded, without any coded PDD diagnoses (i.e., no ASD), were assigned to the ID group. Visits that did not have any of the PDD or ID codes present were assigned to the comparison group.

Youth Demographics.

Youth variables included age, sex, insurance type, and median household income by zip code. For purposes of our analyses, we grouped patient age into 7–12-year-old, 13–18-year-old, and 19–25-year-old cohorts. Insurance type was collapsed into private insurance, Medicaid, and self-pay/other. Race/ethnicity is not available in the NEDS databases.

Youth Clinical Presentation.

We used Clinical Classifications Software (CCS) codes to indicate the presence of suicidal ideation and intentional self-inflicted injury, differentiate between psychiatric and non-psychiatric presenting concerns, and code for co-occurring mental health conditions. The CCS is a tool developed by the Agency for Healthcare Research and Quality used to cluster ICD diagnoses and procedures into smaller, clinically meaningful categories. The CCS code 662 was used to indicate a suicidal ideation or intentional self-inflicted injury diagnosis, as this code captures all intentional self-inflicted injury E codes as well as suicidal ideation (Table 1). It is important to note that E codes do not specify intent to die; therefore, it is not possible to differentiate between suicidal and non-suicidal intentional self-inflicted injury.

Table 1.

ICD-9 diagnostic codes included in CCS 662

CCS 662: Suicide and intentional self-inflicted injury
V6284 Suicidal ideation
E950 Suicide and self-inflicted poisoning by solid or liquid substances
E951 Suicide and self-inflicted poisoning by gases in domestic use
E952 Suicide and self-inflicted poisoning by other gases or vapors
E953 Suicide and self-inflicted injury by hanging, strangulation, and suffocation
E954 Suicide and self-inflicted injury by submersion (drowning)
E955 Suicide and self-inflicted injury by firearms, air guns, and explosives
E956 Suicide and self-inflicted injury by cutting and piercing instrument
E957 Suicide and self-inflicted injury by jumping from high places
E958 Suicide and self-inflicted injury by other and unspecified means
E959 Late effects of self-inflicted injury
Open in a new tab

Replicating the procedures used by Kalb et al. (2012), we used first listed diagnosis to identify the presenting concern for each ED visit. Psychiatric visits were defined as those that have a psychiatric disorder listed as primary diagnosis. If ASD or ID was coded as the primary diagnosis, the secondary diagnosis was used as the presenting concern, as ASD and ID alone would not be reason for an ED visit. To characterize co-occurring psychiatric conditions in the sample, the presence of common mental health diagnoses was extracted using CCS categories, including adjustment disorders (CCS: 650), mood disorders (CCS: 657), anxiety disorders (CCS: 651), behavioral disorders (CCS: 652), and psychotic disorders (CCS: 659). Last, for those cases where ID was present, level of ID was indicated using ICD-9 codes (i.e., [317] mild ID, [318.0] moderate ID, [318.1] severe ID, [318.2] profound ID, and [319] unspecified ID).

Hospital Characteristics.

Hospital variables included region of hospital (i.e., Northeast, Midwest, West, and South) and teaching status/rurality of hospital (i.e., Metropolitan, teaching, Metropolitan, non-teaching, and Non-metropolitan).

Statistical Analysis

All statistical analyses were weighted, using the NEDS discharge weights to produce nationwide visit-level statistics using the ED visit as the unit of analysis. Variance adjustments were made for the NEDS design, which is based on a stratified, single-stage cluster sample of 20% of hospital-owned EDs in the US according to five strata (region, urban/rural, teaching status, ownership, and trauma level). This project was declared exempt from IRB review, as it uses de-identified public use data. Similarly, consent of participants for retrospective data analysis was not required.

Prevalence of suicidal ideation and intentional self-inflicted injury diagnoses across groups.

Linear probability model regressions and Wald tests were performed to compare the prevalence of suicidal ideation and intentional self-inflicted injury diagnoses across ASD, ID, and comparison groups overall and within each age group (i.e., 7–12 years, 13–18 years, and 19–25 years). Prevalence of diagnoses using the combined CCS code 662 and of suicidal ideation and intentional self-inflicted injury diagnoses separately were evaluated.

Rates of suicidal ideation and intentional self-inflicted injury diagnosis across time.

We evaluated change in the rate of suicidal ideation and intentional self-inflicted injury diagnoses across time within the ASD, ID, and comparison groups overall and within each age group using linear probability model regressions and Wald tests. Because prevalence estimates of suicidal ideation and intentional self-inflicted injury diagnoses separately followed similar patterns, we evaluated change in the combined CCS code 662.

Correlates.

Differences across a number of potential correlates, including youth demographics (i.e., age, sex, insurance type, and median household income by zip code) and clinical presentation (i.e., presence of psychiatric presenting concern, co-occurring mental health diagnoses, level of ID where applicable), and hospital characteristics (i.e., region and teaching status/rurality) were examined. First, the odds of a suicidal ideation or intentional self-inflicted injury diagnosis were examined within groups using bivariate logistic regression. Then, limiting analysis to only cases with a suicidal ideation or intentional self-inflicted injury diagnosis, the odds of group membership (i.e., ID or ASD relative to the comparison group) were examined using bivariate multinomial logistic regression. For this portion, Wald tests of the differences in the relative odds of ASD and the relative odds of ID were conducted. Again, because prevalence estimates of suicidal ideation and intentional self-inflicted injury diagnoses separately followed similar patterns, we examined these diagnoses together using the combined CCS code 662.

Community Involvement

There was no community involvement in this study.

Results

Of the 59,437,675 ED visits of youth aged 7–25 years old included in NEDS between 2006–2014, 145,929 had a diagnosis of ASD coded, with or without co-occurring ID (i.e., the ASD group), and 90,372 had a diagnosis of ID coded, without ASD (i.e., the ID group). All other youth visits were placed in the comparison group (n=59,201,374).

Prevalence across groups

Evaluated using linear probability model regressions and Wald tests, the overall prevalence of ED visits with a CCS code 662 diagnosis (i.e., suicidal ideation or intentional self-inflicted injury) in the ID group (6.6%) was statistically significantly greater than in the ASD group (5.1%) and in the comparison group (1.2%). Prevalence in the ASD group was also statistically significantly greater than in the comparison group (all p<0.001). However, results varied by age group. In 7–12-year-olds, the prevalence of visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the ASD group (2.0%) was statistically significantly greater than in the ID group (1.3%) and the comparison group (0.3%), and prevalence in the ID group was statistically significantly greater than in the comparison group (all p<0.001). Similar results were found in the 13–18-year-old cohort, with the ASD group having statistically significantly more visits with a suicidal ideation or intentional self-inflicted injury diagnosis (7.2%) compared to the ID group (5.8%) and the comparison group (1.9%) and the ID group having statistically significantly more visits with these diagnoses compared to the comparison group (all p<0.001). In 19–25-year-olds, the prevalence of visits with a suicidal ideation or intentional self-inflicted injury diagnosis was statistically significantly greater in the ID group (8.3%) compared to the ASD group (7.4%) and the comparison group (1.2%), and prevalence in the ASD group was also statistically significantly greater than in the comparison group (all p<0.001).

Suicidal Ideation.

Prevalence of visits with a suicidal ideation diagnosis followed an identical pattern. Within the full sample, the prevalence of ED visits with a suicidal ideation diagnosis coded in the ID group (5.0%) was statistically significantly greater than in the ASD group (3.7%) and in the comparison group (0.7%), and prevalence in the ASD group was statistically significantly greater than in the comparison group (all p<0.001). In 7–12-year-olds and 13–18-year-olds, prevalence of ED visits with a suicidal ideation diagnosis in the ASD group (7–12: 1.6%; 13–18: 5.2%) was statistically significantly greater than in the ID group (7–12: 1.1%, p<0.01; 13–18: 4.4%, p=0.011) and in the comparison group (7–12: 0.2%, p<0.001; 13–18: 1.1%, p<0.001), and prevalence in the ID group was also statistically significantly greater than in the comparison group (both p<0.001). In 19–25-year-olds, the ID group had statistically significantly more visits with a suicidal ideation diagnosis coded (6.3%) than the ASD group (5.2%) and the comparison group (0.7%), and prevalence in the ASD group was statistically significantly greater than in the comparison group (all p<0.001).

Intentional Self-Inflicted Injury.

Results from examining prevalence of intentional self-inflicted injury diagnoses at ED visits were also similar. Within the full sample, the prevalence of ED visits with an intentional self-inflicted injury diagnosis coded in the ID group (1.8%) was statistically significantly greater than in the ASD group (1.6%) and in the comparison group (0.6%). Prevalence in the ASD group was also statistically significantly greater than in the comparison group (all p<0.001). In 7–12-year-olds and 13–18-year-olds, prevalence of ED visits with an intentional self-inflicted injury diagnosis in the ASD group (7–12: 0.4%; 13–18: 2.3%) was statistically significantly greater than in the ID group (7–12: 0.2%; 13–18: 1.7%; both p<0.001) and the comparison group (7–12: 0.1%; 13–18: 0.9%; both p<0.001), and prevalence in the ID group was statistically significantly greater than in the comparison group (7–12: p=0.019, 13–18: p<0.001). In 19–25-year-olds, prevalence of visits with an intentional self-inflicted injury diagnosis did not differ between the ID group (2.4%) and the ASD group (2.6%). Both groups had statistically significantly more visits with an intentional self-inflicted injury diagnosis than the comparison group (0.6%; p<0.001; Table 2).

Table 2.

Prevalence of CCS 662: Suicide and Intentional Self-Inflicted Injury Diagnostic Codes in ASD, ID, and Comparison groups from 2006–2014: Weighted

Prevalence of CCS 662 Overall Prevalence of Suicidal Ideation Prevalence of Intentional Self-Inflicted Injury
N (unweighted) 95% CI N (unweighted) 95% CI N (unweighted) 95% CI
7–12 years old
ASD 1,194 2.0 (1.7–2.2)bc 988 1.6 (1.4–1.9)bc 223 0.4 (0.3–0.4)bc
ID 173 1.3 (0.8–1.7)ac 155 1.1 (0.7–1.6)ac 22 0.2 (0.1–0.2)ac
Comparison 36,115 0.3 (0.3–0.3)ab 28,607 0.2 (0.2–0.3)ab 8,666 0.1 (0.1–0.1)ab
13–18 years old
ASD 3,772 7.2 (6.5–7.9)bc 2,696 5.2 (4.6–5.8)bc 1,219 2.3 (2.1–2.5)bc
ID 1,517 5.8 (5.0–6.6)ac 1,153 4.4 (3.7–5.1)ac 426 1.7 (1.4–1.9)ac
Comparison 318,036 1.9 (1.9–2.0)ab 181,665 1.1 (1.0–1.2)ab 158,048 0.9 (0.9–1.0)ab
19–25 years old
ASD 2,433 7.4 (6.9–8.0)bc 1,705 5.2 (4.8–5.7)bc 852 2.6 (2.4–2.8)c
ID 4,319 8.3 (7.7–9.0)ac 3,291 6.3 (5.7–7.0)ac 1,221 2.4 (2.2–2.6)c
Comparison 378,904 1.2 (1.2–1.3)ab 226,258 0.7 (0.7–0.8)ab 175,255 0.6 (0.6–0.6)ab
Overall (7–25 years old)
ASD 7,339 5.1 (4.7–5.5)bc 5,389 3.7 (3.3–4.1)bc 2,294 1.6 (1.5–1.7)bc
ID 6,009 6.6 (5.9–7.2)ac 4,599 5.0 (4.4–5.6)ac 1,669 1.8 (1.7–2.0)ac
Comparison 733,055 1.2 (1.2–1.3)ab 436,530 0.7 (0.7–0.8)ab 341,969 0.6 (0.6–0.6)ab
Open in a new tab
a

Statistically significantly different than ASD group

b

Statistically significantly different than ID group

c

Statistically significantly different than Comparison group

Rates across time

Increases in the rates of ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis from 2006–2014, analyzed using linear probability model regressions and Wald tests, were statistically significant in all groups (all p<0.001; Figure 1). Apart from visits by 7–12-year-olds in the ID group, where change in the rate of diagnosis across time did not reach statistical significance (p=0.053), significant increases in the number of ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis over time were seen across age groups (i.e., 7–12 years old; 13–18 years old; 19–25 years old) within all groups as well (all p<0.01).

Figure 1.

Figure 1.

Open in a new tab

Rates of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis from 2006–2014 Across Age Groups

Statistically significant between group differences were also present. Across ages, increases in visits with a suicidal ideation or intentional self-inflicted injury diagnosis from 2006–2014 were larger in the ASD group compared to the comparison group (p<0.001) and in the ID group compared to the comparison group (p<0.001). Further, increases in the ASD group were larger than in the ID group (p=0.008). Examining each age group independently, several differences were observed. Increases in the rate of visits with a suicidal ideation or intentional self-inflicted injury diagnosis were larger in the ASD group than in the comparison group in 7–12-year-olds (p=0.002), 13–18-year-olds (p<0.001), and 19–25-year-olds (p<0.001). Increases in the rate of visits with these diagnoses were also larger in the ID group compared to the comparison group within the 13–18-year-old age group (p=0.013) and the 19–25-year-old group (p<0.001). No differences were found between ASD and ID groups within each age group or between the ID group and comparison group within the 7–12-year-old age group. See Figures 2, 3 and 4.

Figure 2.

Figure 2.

Open in a new tab

Rates of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis from 2006–2014 in Youth Aged 7–12 Years Old

Figure 3.

Figure 3.

Open in a new tab

Rates of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis from 2006–2014 in Youth Aged 13–18 Years Old

Figure 4.

Figure 4.

Open in a new tab

Rates of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis from 2006–2014 in Youth Aged 19–25 Years Old

Correlates

As described, results for all within group correlate analyses were obtained from bivariate logistic regression models. Then, evaluating only cases with a suicidal ideation or intentional self-inflicted injury diagnosis, bivariate multinomial logistic regression were used to examine the odds of group membership (i.e., ID or ASD relative to the comparison group), and Wald tests of the differences in the relative odds of ASD and the relative odds of ID were conducted.

ASD Group.

Of all visits by autistic youth, autistic youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were less likely to be male than female (OR=0.78, 95% CI=0.73–0.83) and were more likely to be 13–18 years old (OR=3.86, 95% CI=3.51–4.24) and 19–25 years old (OR=4.00, 95% CI=3.53–4.52) than 7–12 years old. They were also less likely to have Medicaid than self-pay/other (OR=0.63, 95% CI=0.57–0.70) and were more likely to have lived in an area of higher median household income (26th-50th Percentile: OR=1.20, 95% CI=1.08–1.33; 51st-75th Percentile: OR=1.33, 95% CI=1.20–1.48; 76th-100th Percentile: OR=1.42, 95% CI=1.25–1.62) than in an area with median household income in the 25th percentile or lower. Clinically, youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to present to the ED with a psychiatric concern than a non-psychiatric concern (OR=11.65, 95% CI=10.68–12.71) and were more likely to receive each of the psychiatric discharge diagnoses examined, including adjustment disorders (OR=3.52, 95% CI=2.89–4.28), mood disorders (OR=9.82, 95% CI=9.14–10.55), anxiety disorders (OR=1.99, 95% CI=1.84–2.16), behavioral disorders (OR=1.13, 95% CI=1.05–1.21), and psychotic disorders (OR=2.20, 95% CI=1.95–2.48), compared to receiving no psychiatric diagnosis. Regarding co-occurring ID, youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to have mild ID (OR=2.68, 95% CI=2.27–3.17) and less likely to have severe (OR=0.34, 95% CI=0.22–0.53) and unspecified ID (OR=0.64, 95% CI=0.52–0.77) compared to having no co-occurring ID diagnosis. Hospital characteristics were not different (Table 3).

Table 3.

Odds of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis across All Visits in the ASD Group: Weighted

Odds Ratio 95% CI
Age
7–12 years [Reference]
13–18 years 3.86 (3.51–4.24)*
19–25 years 4.00 (3.53–4.52)*
Sex
Female [Reference]
Male 0.78 (0.73–0.83)*
Insurance type
Self Pay/Other [Reference]
Medicaid 0.63 (0.57–0.70)*
Private Insurance 1.03 (0.93–1.16)
Median HH income for patient’s ZIP code
0–25th Percentile [Reference]
26th–50th Percentile 1.20 (1.08–1.33)*
51st–75th Percentile 1.33 (1.20–1.48)*
76th–100th Percentile 1.42 (1.25–1.62)*
Clinical Factors
No psychiatric presenting concern [Reference]
Any psychiatric concern 11.65 (10.68–12.71)*
No psychiatric presenting concern [Reference]
Adjustment disorders 3.52 (2.89–4.28)*
Mood disorders 9.82 (9.14–10.55)*
Anxiety disorders 1.99 (1.84–2.16)*
Behavioral disorders 1.13 (1.05–1.21)*
Psychotic disorders 2.20 (1.95–2.48)*
Region of Hospital
Northeast [Reference]
Midwest 1.27 (0.95–1.70)
South 0.94 (0.73–1.21)
West 0.93 (0.71–1.21)
Teaching status of hospital
Non-metropolitan [Reference]
Metropolitan, non-teaching 0.95 (0.81–1.12)
Metropolitan, teaching 1.09 (0.90–1.32)
Intellectual Disability
No ID [Reference]
Co-occurring ID (any level) 1.00 (0.88–1.13)
No ID [Reference]
Mild ID 2.68 (2.27–3.17)*
Moderate ID 1.01 (0.73–1.39)
Severe ID 0.34 (0.22–0.53)*
Profound ID 0.50 (0.24–1.03)
Unspecified ID 0.64 (0.52–0.77)*
Open in a new tab
*

Statistically significant odds ratio

ID Group.

Of all visits by youth with ID, youth with ID at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were less likely to be male than female (OR=0.68, 95% CI=0.63–0.73) and more likely to be in the 13–18- (OR=4.76, 95% CI=3.57–6.35) and 19–25-year-old age groups (OR=7.04, 95% CI=5.06–9.78) than in the 7–12-year-old age group. These youth were also less likely to have private insurance (OR=0.49, 95% CI=0.41–0.59) and Medicaid (OR=0.77, 95% CI=0.68–0.89) than self-pay/other. Unlike the ASD group, youth with ID at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were less likely to have lived in areas of higher median household income (51st-75th Percentile: OR=0.79, 95% CI=0.69–0.90; 76th-100th Percentile: OR=0.64, 95% CI=0.54–0.75) than in an area with median household income in the 25th percentile or lower. As in the ASD group, they were more likely to have presented to the ED with a psychiatric concern than a non-psychiatric concern (OR=12.40, 95% CI=10.75–14.30), and were more likely to receive each of the psychiatric discharge diagnoses examined, including adjustment disorders (OR=3.40, 95% CI=2.84–4.07), mood disorders (OR=8.29, 95% CI=7.45–9.21), anxiety disorders (OR=1.96, 95% CI=1.78–2.15), behavioral disorders (OR=1.20, 95% CI=1.10–1.30), and psychotic disorders (OR=2.90, 95% CI=2.61–3.21), compared to receiving no psychiatric diagnosis. Youth with ID at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were less likely to have moderate (OR=0.57, 95% CI=0.49–0.66), severe (OR=0.04, 95% CI=0.03–0.05), profound (OR=0.02, 95% CI=0.01–0.04), and unspecified ID (OR=0.22, 95% CI=0.20–0.25) compared to mild ID, and were more likely seen in a hospital in the Midwest (OR=1.60, 95% CI=1.22–2.09) and were less likely seen in a hospital in the West (OR=0.47, 95% CI=0.33–0.68) than in a hospital in the Northeast (Table 4).

Table 4.

Odds of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis across All Visits in the ID Group: Weighted

Odds Ratio 95% CI
Age
7–12 years [Reference]
13–18 years 4.76 (3.57–6.35)*
19–25 years 7.04 (5.06–9.78)*
Sex
Female [Reference]
Male 0.68 (0.63–0.73)*
Insurance type
Self Pay/Other [Reference]
Medicaid 0.77 (0.68–0.89)*
Private Insurance 0.49 (0.41–0.59)*
Median HH income for patient’s ZIP code
0–25th Percentile [Reference]
26th–50th Percentile 0.93 (0.81–1.06)
51st–75th Percentile 0.79 (0.69–0.90)*
76th–100th Percentile 0.64 (0.54–0.75)*
Clinical Factors
No psychiatric presenting concern [Reference]
Any psychiatric concern 12.40 (10.75–14.30)*
No psychiatric presenting concern [Reference]
Adjustment disorders 3.40 (2.84–4.07)*
Mood disorders 8.29 (7.45–9.21)*
Anxiety disorders 1.96 (1.78–2.15)*
Behavioral disorders 1.20 (1.10–1.30)*
Psychotic disorders 2.90 (2.61–3.21)*
Region of Hospital
Northeast [Reference]
Midwest 1.60 (1.22–2.09)*
South 1.25 (1.00–1.55)
West 0.47 (0.33–0.68)*
Teaching status of hospital
Non-metropolitan [Reference]
Metropolitan, non-teaching 0.95 (0.77–1.16)
Metropolitan, teaching 0.85 (0.67–1.08)
Intellectual Disability
Mild ID [Reference]
Moderate ID 0.57 (0.49–0.66)*
Severe ID 0.04 (0.03–0.05)*
Profound ID 0.02 (0.01–0.04)*
Unspecified ID 0.22 (0.20–0.25)*
Open in a new tab
*

Statistically significant odds ratio

Comparison Group.

Unlike the ASD and ID groups, no significant sex differences were found in the comparison group. Youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to be 13–18 (OR=6.32, 95% CI=6.05–6.60) and 19–25 years old (OR=4.01, 95% CI=3.75–4.26) than 7–12 years old. While no significant differences were evident in insurance type, youth with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to have resided in areas of higher median household income (26th-50th Percentile: OR=1.22, 95% CI=1.18–1.26; 51st-75th Percentile: OR=1.40, 95% CI=1.34–1.45; 76th-100th Percentile: OR=1.59, 95% CI=1.50–1.68) than in an area with median household income in the 25th percentile or lower. As in the ASD and ID groups, they were more likely to present to the ED with a psychiatric concern than a non-psychiatric concern (OR=44.79, 95% CI=42.79–46.89). They were also more likely to have received each of the psychiatric discharge diagnoses examined, including adjustment disorders (OR=17.65, 95% CI=15.86–19.65), mood disorders (OR=53.70, 95% CI=51.44–56.05), anxiety disorders (OR=1.59, 95% CI=1.51–1.68), behavioral disorders (OR=1.12, 95% CI=1.08–1.17), and psychotic disorders (OR=2.83, 95% CI=2.70–2.96), compared to receiving no psychiatric diagnosis. Several differences in hospital characteristics were found. Youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely seen in hospitals in the Midwest (OR=1.15, 95% CI=1.02–1.29) and West (OR=1.16, 95% CI=1.05–1.29) and less likely in hospitals in the South (OR=0.87, 95% CI=0.78–0.96) compared to hospitals in the Northeast. Further, they were more likely to be seen in metropolitan, teaching hospitals (OR=1.28, 95% CI=1.19–1.38) and metropolitan, non-teaching hospitals (OR=1.12, 95% CI=1.05–1.20) compared to non-metropolitan hospitals (Table 5).

Table 5.

Odds of Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis across All Visits in the Comparison Group: Weighted

Odds Ratio 95% CI
Age
7–12 years [Reference]
13–18 years 6.32 (6.05–6.60)*
19–25 years 4.01 (3.75–4.26)*
Sex
Female [Reference]
Male 1.01 (1.00–1.03)
Insurance type
Self Pay/Other [Reference]
Medicaid 0.99 (0.96–1.02)
Private Insurance 1.08 (1.05–1.12)
Median HH income for patient’s ZIP code
0–25th Percentile [Reference]
26th–50th Percentile 1.22 (1.18–1.26)*
51st–75th Percentile 1.40 (1.34–1.45)*
76th–100th Percentile 1.59 (1.50–1.68)*
Clinical Factors
No psychiatric presenting concern [Reference]
Any psychiatric concern 44.79 (42.79–46.89)*
No psychiatric presenting concern [Reference]
Adjustment disorders 17.65 (15.86–19.65)*
Mood disorders 53.70 (51.44–56.05)*
Anxiety disorders 1.59 (1.51–1.68)*
Behavioral disorders 1.12 (1.08–1.17)*
Psychotic disorders 2.83 (2.70–2.96)*
Region of Hospital
Northeast [Reference]
Midwest 1.15 (1.02–1.29)*
South 0.87 (0.78–0.96)*
West 1.16 (1.05–1.29)*
Teaching status of hospital
Non-metropolitan [Reference]
Metropolitan, non-teaching 1.12 (1.05–1.20)*
Metropolitan, teaching 1.28 (1.19–1.38)*
Open in a new tab
*

Statistically significant odds ratio

Differences Among Groups.

Finally, we examined differences among youth visits with a suicidal ideation or intentional self-inflicted injury diagnosis across ASD, ID, and comparison groups. Of all youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis, those in the ASD group were less likely to be 13–18 (OR=0.36, 95% CI=0.33–0.39) and 19–25 years old (OR=0.20, 95% CI=0.18–0.22) and those in the ID group were more likely to be 19–25 years old (OR=2.38, 95% CI=1.80–3.15) compared to 7–12 years old. These odds were significantly different between ASD and ID groups (p<0.001). Further, visits in the ASD and ID groups were more likely to be by male than female youth (ASD: OR=4.00, 95% CI=3.74–4.27; ID: OR=1.12, 95% CI=1.03–1.20), and the odds of being male were significantly higher in the ASD group than ID group (p<0.001). Compared to all youth visits with a suicidal ideation or intentional self-inflicted injury diagnosis, those in the ASD group were more likely to have both Medicaid (OR=2.46, 95% CI=2.22–2.73) and private insurance (OR=2.56, 95% CI=2.30–2.85), whereas, youth in the ID group were more likely to have Medicaid (OR=2.45, 95% CI=2.18–2.75) and less likely to have private insurance (OR=0.38, 95% CI=0.33–0.45) compared to self-pay/other. The odds of having private insurance was significantly different across ASD and ID groups (p<0.001). Of all visits with a suicidal ideation or intentional self-inflicted injury diagnosis, autistic youth were more likely to have resided in areas of higher median household income (26th-50th Percentile: OR=1.25, 95% CI=1.14–1.37; 51st-75th Percentile: OR=1.49, 95% CI=1.34–1.67; 76th-100th Percentile: OR=1.90, 95% CI=1.68–2.15) than in an area with median household income in the 25th percentile or lower. Youth with ID were less likely to have resided in areas of higher median household income (26th-50th Percentile: OR=0.78, 95% CI=0.68–0.90; 51st-75th Percentile: OR=0.64, 95% CI=0.57–0.73; 76th-100th Percentile: OR=0.47, 95% CI=0.40–0.55) than in an area with median household income in the 25th percentile or lower, and these odds were significantly different from the ASD group (all p<0.001).

Clinically, autistic youth and youth with ID at visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to have a psychiatric presenting concern (ASD: OR=1.27, 95% CI=1.19–1.35; ID: OR=2.28, 95% CI=2.07–2.51), and the odds were significantly higher in the ID group than the ASD group (p<0.001). Youth at visits in the ID group were more likely to have received each of the psychiatric discharge diagnoses examined, including adjustment disorders (OR=1.73, 95% CI=1.49–2.00), mood disorders (OR=1.39, 95% CI=1.28–1.51), anxiety disorders (OR=1.21, 95% CI=1.11–1.31), behavioral disorders (OR=3.33, 95% CI=3.08–3.59), and psychotic disorders (OR=5.92, 95% CI=5.44–6.43), compared to receiving no psychiatric diagnosis. Youth at visits in the ASD group were more likely to have been diagnosed with anxiety disorders (OR=1.57, 95% CI=1.48–1.67), behavioral disorders (OR=5.45, 95% CI=5.13–5.78), and psychotic disorders (OR=1.62, 95% CI=1.46–1.79), compared to receiving no psychiatric diagnosis. The odds of psychotic disorder, adjustment disorder, and mood disorder diagnoses were significantly higher in the ID group than the ASD group (all p<0.001), while the odds of anxiety disorder and behavioral disorder diagnoses were significantly higher in the ASD group than the ID group (all p<0.001).

Finally, of all youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis, autistic youth were less likely to be seen at hospitals in the South (OR=0.70, 95% CI=0.54–0.90) and West (OR=0.72, 95% CI=0.55–0.94) than the Northeast, and youth with ID were less likely to be seen at hospitals in the West (OR=0.34, 95% CI=0.26–0.46) than the Northeast. Odds of being seen at a hospital in the West were significantly different between the ASD and ID group (p<0.001). Further, youth in the ASD and ID groups were more likely to be seen at a metropolitan, teaching hospitals than non-metropolitan hospitals (ASD: OR=1.86, 95% CI=1.53–2.26; ID: OR=1.49, 95% CI=1.19–1.86), and the odds were significantly higher in the ASD group than the ID group (p=0.029). Youth in the ASD group were also more likely to have been seen at a metropolitan, non-teaching hospital than a non-metropolitan hospital (OR=1.19, 95% CI=1.01–1.41; Table 6).

Table 6.

Odds of Being in the ASD or ID Group Across All Visits with a Suicidal Ideation or Intentional Self-Inflicted Injury Diagnosis: Weighted

ASD Group ID Group
Odds Ratio 95% CI Odds Ratio 95% CI
Age
7–12 years [Reference]
13–18 years 0.36 (0.33–0.39)*b 0.98 (0.75–1.27)a
19–25 years 0.20 (0.18–0.22)*b 2.38 (1.80–3.15)*a
Sex
Female [Reference]
Male 4.00 (3.74–4.27)*b 1.12 (1.03–1.20)*a
Insurance type
Self Pay/Other [Reference]
Medicaid 2.46 (2.22–2.73)* 2.45 (2.18–2.75)*
Private Insurance 2.56 (2.30–2.85)*b 0.38 (0.33–0.45)*a
Median HH income for patient’s ZIP code
0–25th Percentile [Reference]
26th–50th Percentile 1.25 (1.14–1.37)*b 0.78 (0.68–0.90)*a
51st–75th Percentile 1.49 (1.34–1.67)*b 0.64 (0.57–0.73)*a
76th–100th Percentile 1.90 (1.68–2.15)*b 0.47 (0.40–0.55)*a
Clinical Factors
No psychiatric presenting concern [Reference]
Any psychiatric concern 1.27 (1.19–1.35)*b 2.28 (2.07–2.51)*a
No psychiatric presenting concern [Reference]
Adjustment disorders 0.92 (0.80–1.07) 1.73 (1.49–2.00)*a
Mood disorders 1.01 (0.94–1.08) 1.39 (1.28–1.51)*a
Anxiety disorders 1.57 (1.48–1.67)*b 1.21 (1.11–1.31)*a
Behavioral disorders 5.45 (5.13–5.78)*b 3.33 (3.08–3.59)*a
Psychotic disorders 1.62 (1.46–1.79)*b 5.92 (5.44–6.43)*a
Region of Hospital
Northeast [Reference]
Midwest 0.93 (0.70–1.22) 1.10 (0–84–1.44)
South 0.70 (0.54–0.90)* 0.83 (0.64–1.06)
West 0.72 (0.55–0.94)*b 0.34 (0.26–0.46)*a
Teaching status of hospital
Non-metropolitan [Reference]
Metropolitan, non-teaching 1.19 (1.01–1.41)* 0.99 (0.81–1.22)
Metropolitan, teaching 1.86 (1.53–2.26)*b 1.49 (1.19–1.86)*a
Open in a new tab
*

Statistically significant odds ratio

a

Odds ratio statistically significantly different than odds ratio of the ASD group

b

Odds ratio statistically significantly different than odds ratio of the ID group

Discussion

While attention to autistic youth and youth with ID has been limited in the suicide literature, results of the current study support the urgency of understanding and addressing the unique vulnerability of these groups to suicidality and self-harm, as significantly more ED visits were coded with a suicidal ideation or intentional self-inflicted injury diagnosis in these groups compared to ED visits by youth without ASD or ID diagnoses (comparison group) overall and in each age range. Notably, when examining suicidal ideation and intentional self-inflicted injury diagnoses separately, the results remained the same. Further, while the rate of suicidal ideation or intentional self-inflicted injury diagnosis at youth ED visits increased from 2006–2014 in all groups, increases in the ASD and ID groups were significantly larger than in the comparison group overall and specifically in visits by 13–18-year-olds and 19–25-year-olds. These findings are concerning, because the ED workforce is often underprepared to provide care for youth with developmental disabilities, such as autistic youth or youth with ID (Nicholas et al., 2016; Spassiani et al., 2020), and because the assessment and management of suicide risk in community EDs has been inconsistent and under-studied for children and adolescents broadly (Cervantes et al., 2021). Taken together, these findings suggest that the co-occurrence of suicidality and ASD or ID would serve as a sizeable barrier to adequate clinical care.

Importantly, differences were found across autistic youth and youth with ID. While the ID group had more ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis than the ASD group overall, the ASD group had significantly more in 7–12- and 13–18-year-olds and the ID group had significantly more in 19–25-year-olds. These findings were largely maintained when examining suicidal ideation alone and intentional self-inflicted injury alone. These differences suggest that there may be different pathways to suicidality among autistic youth, youth with ID, and youth without diagnoses of ASD or ID, and highlights the importance of examining sociodemographic and clinical correlates both within and between youth groups to identify potential factors for suicide risk.

Within group comparisons revealed some commonalities in correlates across ASD, ID, and comparison groups. Across groups, youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to be older, have a psychiatric presenting concern, and have a higher rate of additional mental health diagnoses coded. Unlike the comparison group, where no sex differences were identified, youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the ASD and ID groups were less likely to be male. This aligns with previous ASD research (Hirvikoski et al., 2019; Kõlves et al., 2021), suggesting female sex may be a risk factor for suicidality and self-harm in these groups. Though, more research is needed. Again, in line with previous research in the ASD and ID literatures (Dodd et al., 2016; McDonnell et al., 2019), youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the ID group and in the ASD group when co-occurring ID was present were also more likely to be of higher intellectual functioning. Interestingly, autistic youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were nearly three times more likely to have mild ID than have no co-occurring ID diagnosis. This contrasts with emerging research suggesting autistic individuals without co-occurring ID experience increased suicidality. It is important to note that previous studies did not examine each level of ID independently but rather examined intellectual functioning continuously (McDonnell et al., 2019) or collapsed ID levels into one category (Hand et al., 2020; Hirvikoski et al., 2019; Kõlves et al., 2021). As mild ID is not typically associated with the presentation differences that may be characteristic of individuals with more severe cognitive impairments, it is possible that co-occurring ASD and mild ID is associated with increased suicide risk, and future studies should examine this further.

In the ID group, youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely to have lived in areas of lower median household income. This is in line with a recent study on the general youth population which found increased suicide rates were associated with increased county-level poverty concentration (Hoffmann et al., 2020). In contrast, youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the comparison group were more likely to have lived in areas of higher median household income than youth at ED visits without these diagnoses. While these data may reflect the mixed associations between socioeconomic status and suicidality found in research (Hoffmann et al., 2020), this finding may also reflect who is accessing ED services for suicidality and self-harm. More ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the ASD group were also by youth who lived in areas of higher median household income, conflicting with the sparse ASD literature on the association between socioeconomic status and suicidality, which found that lower parental education was associated with increased risk of talking about suicide (McDonnell et al., 2019). While it is possible that suicidality and self-harm or ED visits for suicidality and self-harm are more frequent in autistic youth of higher socioeconomic status, this finding may also be associated with existing disparities in the recognition of ASD. Previous research has found that ASD service availability is limited in areas of greater socioeconomic disadvantage (Drahota et al., 2020) and that individuals of lower socioeconomic status often fail to be diagnosed or are diagnosed with ASD at later ages (Fountain et al., 2011; Thomas et al., 2012), whereas individuals of higher socioeconomic status may access diagnostic services earlier and at higher rates. Therefore, because ED providers are not specialized in identifying and diagnosing ASD, there may be ED visits included in the comparison group by youth of lower socioeconomic status who meet criteria for a diagnosis of ASD but whose symptoms remained unidentified.

Comparing sociodemographic and clinical factors of youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis across groups, youth in the ASD group were much more likely to be male; this finding is unsurprising given the sex ratio found in population-level ASD prevalence studies (Maenner et al., 2020) and existing barriers to ASD diagnosis for females (Lockwood Estrin et al., 2020). Youth in the ASD group were significantly more likely to reside in areas of higher median household income, and youth in the ID group were significantly less likely to reside in these areas. Of all youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis, autistic youth were more likely to be younger, and youth with ID were more likely to be in the 19–25-year-old age range. Although adapting to the many changes associated with the transition from adolescence to adulthood is difficult for all youth, and particularly for autistic youth and youth with ID (Biggs & Carter, 2016), these results suggest that there may be different periods of vulnerability for autistic youth and youth with ID. As youth with ID appear particularly vulnerable during young adulthood, perhaps this is associated with a reduction in opportunities for social interaction, support, and community involvement and an increase in experiences of failure to achieve the same level of autonomy as their peers after leaving high school. Together with findings related to socioeconomic status, it is possible that youth with ID living in areas of lower median household income have less opportunity and receive fewer services to aid in the transition, leading to poorer mental health outcomes. In contrast, autistic youth appear more vulnerable at younger ages. This may reflect the impact of increasingly complex social demands and the growing importance of peer relationships during adolescence; unsuccessful attempts by autistic adolescents to fit in and compensate for inherent social differences with more nuanced skills required, in addition to potential increases in bullying at this age (Hebron & Humphrey, 2014), may contribute to feelings of loneliness and social isolation, which have been associated with increased risk for suicidality (Hedley, Uljarević, Foley, et al., 2018; Hedley, Uljarević, Wilmot, et al., 2018; Holden et al., 2020). Of course, additional research is necessary to explore these hypotheses.

Clinically, youth at ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis had high rates of mood disorder diagnoses across ASD, ID, and comparison groups. However, youth visits with a suicidal ideation or intentional self-inflicted injury diagnosis in the ASD and ID groups were more likely to have had additional mental health diagnoses coded. Autistic youth were more likely to be diagnosed with anxiety, behavioral, and psychotic disorders at discharge, and the odds of having anxiety and behavioral disorder diagnoses were significantly higher than in the ID group. Youth in the ID group were more likely to have each of the psychiatric diagnoses examined at discharge, and the odds of having mood, adjustment, and psychotic disorder diagnoses were significantly higher than in the ASD group. This suggests that suicidality and self-harm may co-occur with varying mental health presentations in the ASD and ID populations, perhaps complicating detection of suicide risk and illustrating the importance of screening for suicide risk in all patients presenting to the ED. Further illustrating the importance of suicide risk screening for all patients in the ED, most but not all youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis had a psychiatric presenting concern across groups.

Limitations

There are limitations to this study. First, as a study of suicidal ideation and intentional self-inflicted injury diagnoses at ED visits, findings are specific to the population of youth seeking emergency services rather than the youth population more broadly. In addition, as is the case in all large national-level medical record databases, assessment and subsequent coding of diagnoses across EDs were not standardized. Because ED clinicians are not often specialized in the evaluation of neurodevelopmental and psychiatric diagnoses, ASD, ID, and suicidal ideation and intentional self-inflicted injury diagnoses coded at ED visits likely represent underestimations of true prevalence. This may be particularly true for suicidal ideation and self-harm at visits by autistic youth and youth with ID, as clinicians may be less confident assessing and managing suicide risk in these groups (Cervantes et al., 2022; Jager-Hyman et al., 2020) and validated instruments to screen for suicide risk in autistic youth and youth with ID do not yet exist (Howe et al., 2020; Ludi et al., 2012). Next, as discussed, a minority of ED visits in the ASD group include visits by youth with both ASD and ID diagnoses (~9%), which is less than would be expected based on data from the ADDM Network and may point to under-identification, suggesting that maintaining these visits in the sample would be a cautious approach to analyses. Nevertheless, basing the probable prevalence of co-occurring ASD and ID on results from the ADDM Network is imperfect, as these data are collected at age 8 (Maenner et al., 2020) and thus fail to capture individuals diagnosed with ASD later in life who may be less likely to have co-occurring ID. Importantly though, by maintaining these youth and examining each level of ID independently in within group analyses, we found that ED visits with a suicidal ideation or intentional self-inflicted injury diagnosis were more likely in autistic youth with mild ID than without co-occurring ID. While more research is needed, this is a novel finding that may advance our understanding of suicide risk factors in ASD. Further, the comparison group may have included ED visits by youth with other psychiatric or neurodevelopmental diagnoses that may influence rates of suicidal ideation or intentional self-inflicted injury diagnoses. As mentioned, diagnostic E codes do not differentiate between suicidal and non-suicidal self-inflicted injury, which may have different characteristics. Further, because the pattern of results was similar across suicidal ideation and intentional self-inflicted injury diagnoses among groups, we used the combined CCS code 662 for correlate and change over time analyses. As research suggests suicidal ideation and self-harm may have both distinct and shared features (Hand et al., 2020; McDonnell et al., 2019), further investigation is needed. It is also difficult to tease apart findings attributable to true differences versus existing disparities in care. More research in this area is needed. Subsequent research on disposition decisions for youth at visits with a suicidal ideation or intentional self-inflicted injury diagnosis across ASD, ID, and comparison groups would also be informative, as discrepancies may point to inequities in mental health care accessibility and quality for youth with disabilities. Next, NEDS is de-identified, so recidivism is not accounted for in these analyses. NEDS also does not include data on race/ethnicity, which may be a particularly important variable to consider given the substantial increases in suicide among Black youth over the past decade (Sheftall & Miller, 2021). There may also be cohort effects that are not considered in these analyses. For instance, the prevalence of ASD diagnosis has increased over the previous several decades (Maenner et al., 2020), and therefore, autistic youth at ED visits in the younger age groups may be more likely to be formally diagnosed with ASD than in the 19–25-year-old age group. Finally, we only examined trends until 2014. Because Curtin and Heron (2019) found that the pace of increase in youth suicide deaths was greater from 2013–2017 compared to 2007–2013, it will be important to replicate these analyses using more recent data to explore differences across groups.

Conclusion

The current study clearly demonstrates that continued research and clinical efforts are urgently needed to address the co-occurrence of suicidality and self-harm in ID and ASD. It is essential that autistic youth and youth with ID are identified as at-risk populations for suicide in the education of ED providers and that validated suicide risk assessment tools and evidence-based adaptations to care in the ED be developed, evaluated, and disseminated. ED providers should be mindful of the age trends identified in this study, with younger autistic youth and older youth with ID demonstrating vulnerability. Clinicians should also consider sex and cognitive level when assessing suicide risk and understand that autistic youth and youth with ID may present with varying psychiatric presentations. We should continue to evaluate differences in risk and protective factors across disability groups to help inform predictions of risk level and important factors to consider in conducting safety assessments. To inform efforts to improve conditions for youth, it will also be important to understand the shared suicide risk factors experienced by the disability community. Further, it is essential but likely insufficient to improve availability of quality mental health care for autistic youth and youth with ID without efforts to address systems on a larger scale to indirectly decrease suicide risk. We must also address unmet service and support needs across domains (e.g., educational, social, occupational, recreational), especially for transition-aged youth, and work to develop more accommodating and accepting environments (Jones et al., 2021; South et al., 2021). Considerable work remains to be done to address suicidality and self-harm in ID and ASD in the ED and more broadly, and it is essential that attention in research, clinical care, and policy be devoted to the unique experience of these youth.

Acknowledgements:

This study was funded by the Organization for Autism Research (OAR) and the National Institute of Mental Health (P50MH113662) and was made possible with data from many state HCUP Data Partners (https://www.hcup-us.ahrq.gov/db/hcupdatapartners.jsp). The authors thank OAR and the NIMH for their support and acknowledge that the findings and conclusions are those of the authors and do not necessarily reflect the opinions of OAR or the NIMH.

References

  1. Biggs EE, & Carter EW (2016). Quality of life for transition-age youth with autism or intellectual disability. Journal of Autism and Developmental Disorders, 46(1), 190–204. 10.1007/s10803-015-2563-x [DOI] [PubMed] [Google Scholar]
  2. Brahmbhatt K, Kurtz BP, Afzal KI, Giles LL, Kowal ED, Johnson KP, Lanzillo E, Pao M, Plioplys S, & Horowitz LM (2019). Suicide risk screening in pediatric hospitals: Clinical pathways to address a global health crisis. Psychosomatics, 60(1), 1–9. 10.1016/j.psym.2018.09.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Cassidy S, Bradley L, Shaw R, & Baron-Cohen S (2018). Risk markers for suicidality in autistic adults. Molecular Autism, 9(1), 42. 10.1186/s13229-018-0226-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cassidy S, & Rodgers J (2017). Understanding and prevention of suicide in autism. The Lancet Psychiatry, 4(6), e11. 10.1016/S2215-0366(17)30162-1 [DOI] [PubMed] [Google Scholar]
  5. Cervantes PE, Seag DEM, Baroni A, Gerson R, Knapp K, Tay ET, Wiener E, & Horwitz SM (2021). Universal suicide risk screening for youths in the emergency department: A systematic review. Psychiatric Services, appi.ps.202000881. 10.1176/appi.ps.202000881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cervantes PE, Li A, Sullivan KA, Seag DEM, Baroni A, & Horwitz SM (2022). Assessing and managing suicide risk in autistic youth: Findings from a clinician survey in a pediatric psychiatric emergency setting. Journal of Autism and Developmental Disorders. 10.1007/s10803-022-05448-8 [DOI] [PubMed] [Google Scholar]
  7. Chen M-H, Pan T-L, Lan W-H, Hsu J-W, Huang K-L, Su T-P, Li C-T, Lin W-C, Wei H-T, & Chen T-J (2017). Risk of suicide attempts Among adolescents and young adults with autism spectrum disorder: A nationwide longitudinal follow-up study. The Journal of Clinical Psychiatry, 78(9), 1174–1179. 10.4088/JCP.16m11100 [DOI] [PubMed] [Google Scholar]
  8. Costa AP, Loor C, & Steffgen G (2020). Suicidality in adults with autism spectrum disorder: The role of depressive symptomatology, alexithymia, and antidepressants. Journal of Autism and Developmental Disorders. 10.1007/s10803-020-04433-3 [DOI] [PubMed] [Google Scholar]
  9. Curtin SC, & Heron M (2019). Death Rates Due to Suicide and Homicide Among Persons Aged 10–24: United States, 2000–2017. https://www.cdc.gov/nchs/data/databriefs/db352-h.pdf [PubMed]
  10. Demirkaya SK, Tutkunkardaş MD, & Mukaddes NM (2016). Assessment of suicidality in children and adolescents with diagnosis of high functioning autism spectrum disorder in a Turkish clinical sample. Neuropsychiatric Disease and Treatment, Volume 12, 2921–2926. 10.2147/NDT.S118304 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dodd P, Doherty A, & Guerin S (2016). A systematic review of suicidality in people with intellectual disabilities. Harvard Review of Psychiatry, 24(3), 202–213. 10.1097/HRP.0000000000000095 [DOI] [PubMed] [Google Scholar]
  12. Drahota A, Sadler R, Hippensteel C, Ingersoll B, & Bishop L (2020). Service deserts and service oases: Utilizing geographic information systems to evaluate service availability for individuals with autism spectrum disorder. Autism, 24(8), 2008–2020. 10.1177/1362361320931265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Erlangsen A, Stenager E, Conwell Y, Andersen PK, Hawton K, Benros ME, Nordentoft M, & Stenager E (2020). Association Between Neurological Disorders and Death by Suicide in Denmark. JAMA, 323(5), 444–454. 10.1001/jama.2019.21834 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Fountain C, King MD, & Bearman PS (2011). Age of diagnosis for autism: Individual and community factors across 10 birth cohorts. Journal of Epidemiology & Community Health, 65(6), 503–510. 10.1136/jech.2009.104588 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hand BN, Benevides TW, & Carretta HJ (2020). Suicidal ideation and self-inflicted injury in medicare enrolled autistic adults with and without co-occurring intellectual disability. Journal of Autism and Developmental Disorders, 50(10), 3489–3495. 10.1007/s10803-019-04345-x [DOI] [PubMed] [Google Scholar]
  16. HCUP Nationwide Emergency Department Sample (NEDS). Healthcare Cost and Utilization Project (HCUP). 2006–2014. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/nedsoverview.jsp [PubMed] [Google Scholar]
  17. Hebron J, & Humphrey N (2014). Exposure to bullying among students with autism spectrum conditions: A multi-informant analysis of risk and protective factors. Autism, 18(6), 618–630. 10.1177/1362361313495965 [DOI] [PubMed] [Google Scholar]
  18. Hedley D, Uljarević M, Foley K-R, Richdale A, & Trollor J (2018). Risk and protective factors underlying depression and suicidal ideation in autism spectrum disorder. Depression and Anxiety, 35(7), 648–657. 10.1002/da.22759 [DOI] [PubMed] [Google Scholar]
  19. Hedley D, Uljarević M, Wilmot M, Richdale A, & Dissanayake C (2018). Understanding depression and thoughts of self-harm in autism: A potential mechanism involving loneliness. Research in Autism Spectrum Disorders, 46, 1–7. 10.1016/j.rasd.2017.11.003 [DOI] [Google Scholar]
  20. Hirvikoski T, Boman M, Chen Q, D’Onofrio BM, Mittendorfer-Rutz E, Lichtenstein P, Bölte S, & Larsson H (2019). Individual risk and familial liability for suicide attempt and suicide in autism: A population-based study. Psychological Medicine, 1–12. 10.1017/S0033291719001405 [DOI] [PubMed] [Google Scholar]
  21. Hirvikoski T, Mittendorfer-Rutz E, Boman M, Larsson H, Lichtenstein P, & Bölte S (2016). Premature mortality in autism spectrum disorder. The British Journal of Psychiatry, 208(3), 232–238. 10.1192/bjp.bp.114.160192 [DOI] [PubMed] [Google Scholar]
  22. Hoffmann JA, Farrell CA, Monuteaux MC, Fleegler EW, & Lee LK (2020). Association of Pediatric Suicide With County-Level Poverty in the United States, 2007–2016. JAMA Pediatrics, 174(3), 287–294. 10.1001/jamapediatrics.2019.5678 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Holden R, Mueller J, McGowan J, Sanyal J, Kikoler M, Simonoff E, Velupillai S, & Downs J (2020). Investigating bullying as a predictor of suicidality in a clinical sample of adolescents with autism spectrum disorder. Autism Research, 13(6), 988–997. 10.1002/aur.2292 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Horowitz LM, Boudreaux ED, Schoenbaum M, Pao M, & Bridge JA (2018). Universal suicide risk screening in the hospital setting: Still a Pandora’s Box? The Joint Commission Journal on Quality and Patient Safety, 44(1), 1–3. 10.1016/j.jcjq.2017.11.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Horowitz LM, Thurm A, Farmer C, Mazefsky C, Lanzillo E, Bridge JA, Greenbaum R, Pao M, Siegel M… for the Autism and Developmental Disorders Inpatient Research Collaborative (ADDIRC). (2018). Talking about death or suicide: Prevalence and clinical correlates in youth with autism spectrum disorder in the psychiatric inpatient setting. Journal of Autism and Developmental Disorders, 48(11), 3702–3710. 10.1007/s10803-017-3180-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Howe SJ, Hewitt K, Baraskewich J, Cassidy S, & McMorris CA (2020). Suicidality among children and youth with and without autism spectrum disorder: A systematic review of existing risk assessment tools. Journal of Autism and Developmental Disorders. 10.1007/s10803-020-04394-7 [DOI] [PubMed] [Google Scholar]
  27. International Society for Autism Research. (2021). Autism Community Priorities for Suicide Prevention: An International Society for Autism Research Policy Brief. https://cdn.ymaws.com/www.autism-insar.org/resource/resmgr/files/policybriefs/2021-insar_policy_brief.pdf
  28. Jachyra P, Lai M-C, Zaheer J, Fernandes N, Dale M, Sawyer A, & Lunsky Y (2021). Suicidal thoughts and behaviours among autistic adults presenting to the psychiatric emergency department: An exploratory chart review. Journal of Autism and Developmental Disorders. 10.1007/s10803-021-05102-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Jager-Hyman S, Maddox BB, Crabbe SR, & Mandell DS (2020). Mental health clinicians’ screening and intervention practices to reduce suicide risk in autistic adolescents and adults. Journal of Autism and Developmental Disorders, 50(10), 3450–3461. 10.1007/s10803-020-04441-3 [DOI] [PubMed] [Google Scholar]
  30. Jones DR, DeBrabander KM, & Sasson NJ (2021). Effects of autism acceptance training on explicit and implicit biases toward autism. Autism, 25(5), 1246–1261. 10.1177/1362361320984896 [DOI] [PubMed] [Google Scholar]
  31. Kalb L, Stuart E, Freedman B, Zablotsky B, & Vasa R (2012). Psychiatric-related emergency department visits among children with an autism spectrum disorder. Pediatric Emergency Care, 28(12), 1269–1276. 10.1097/PEC.0b013e3182767d96 [DOI] [PubMed] [Google Scholar]
  32. King TL, Milner A, Aitken Z, Karahalios A, Emerson E, & Kavanagh AM (2019). Mental health of adolescents: Variations by borderline intellectual functioning and disability. European Child & Adolescent Psychiatry, 28(9), 1231–1240. 10.1007/s00787-019-01278-9 [DOI] [PubMed] [Google Scholar]
  33. Kirby AV, Bakian AV, Zhang Y, Bilder DA, Keeshin BR, & Coon H (2019). A 20-year study of suicide death in a statewide autism population. Autism Research, 12(4), 658–666. 10.1002/aur.2076 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kõlves K, Fitzgerald C, Nordentoft M, Wood SJ, & Erlangsen A (2021). Assessment of suicidal behaviors among individuals with autism spectrum disorder in Denmark. JAMA Network Open, 4(1), e2033565–e2033565. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Lo CB, Bridge JA, Shi J, Ludwig L, & Stanley RM (2020). Children’s Mental Health Emergency Department Visits: 2007–2016. Pediatrics, 145(6). 10.1542/peds.2019-1536 [DOI] [PubMed] [Google Scholar]
  36. Lockwood Estrin G, Milner V, Spain D, Happé F, & Colvert E (2020). Barriers to autism spectrum disorder diagnosis for young women and girls: A systematic review. Review Journal of Autism and Developmental Disorders. 10.1007/s40489-020-00225-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Ludi E, Ballard ED, Greenbaum R, Pao M, Bridge J, Reynolds W, & Horowitz L (2012). Suicide risk in youth with intellectual disabilities: The challenges of screening. Journal of Developmental & Behavioral Pediatrics, 33(5), 431–440. 10.1097/DBP.0b013e3182599295 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Maenner MJ, Shaw KA, Baio J, Washington A, Patrick M, DiRienzo M, Christensen DL, Wiggins LD, Pettygrove S, Andrews JG, Lopez M, Hudson A, Baroud T, Schwenk Y, White T, Rosenberg CR, Lee L-C, Harrington RA, Huston M, … Dietz PM (2020). P Prevalence of autism spectrum disorder among children aged 8 years—Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2016. MMWR Surveillance Summaries, 69(4), 1–12. 10.15585/mmwr.ss6904a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McDonnell CG, DeLucia EA, Hayden EP, Anagnostou E, Nicolson R, Kelley E, Georgiades S, Liu X, & Stevenson RA (2019). An exploratory analysis of predictors of youth suicide-related behaviors in autism spectrum disorder: Implications for prevention science. Journal of Autism and Developmental Disorders. 10.1007/s10803-019-04320-6 [DOI] [PubMed] [Google Scholar]
  40. Merrick J, Merrick E, Morad M, & Kandel I (2005). Adolescents with intellectual disability and suicidal behavior. The Scientific World Journal, 5, 724–728. 10.1100/tsw.2005.90 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Mollison E, Chaplin E, Underwood L, & McCarthy J (2014). A review of risk factors associated with suicide in adults with intellectual disability. Advances in Mental Health and Intellectual Disabilities, 8(5), 302–308. 10.1108/AMHID-05-2014-0021 [DOI] [Google Scholar]
  42. Nagraj D, & Omar HA (2017). Disability and suicide: A review. International Journal of Child Health and Human Development, 10(4), 345–354. [Google Scholar]
  43. National Institute for Health (2020). Estimates of Funding for Various Research, Condition, and Disease Categories (RCDC). Retrieved December 22, 2020, from https://report.nih.gov/funding/categorical-spending#/. [Google Scholar]
  44. Nicholas DB, Zwaigenbaum L, Muskat B, Craig WR, Newton AS, Kilmer C, Greenblatt A, Roberts W, & Cohen-Silver J (2016). Experiences of emergency department care from the perspective of families in which a child has autism spectrum disorder. Social Work in Health Care, 55(6), 409–426. 10.1080/00981389.2016.1178679 [DOI] [PubMed] [Google Scholar]
  45. Oliphant RYK, Smith EM, & Grahame V (2020). What is the prevalence of self-harming and suicidal behaviour in under 18s with ASD, with or without an intellectual disability? Journal of Autism and Developmental Disorders. 10.1007/s10803-020-04422-6 [DOI] [PubMed] [Google Scholar]
  46. Salvatore T, Brown J, Huntley D, Arndt C, & Wiley C (2016). ntellectual disability and suicide risk: An exploratory discussion. Journal of Special Populations, 1, 1. [Google Scholar]
  47. Schilling EA, Aseltine RH, & James A (2016). The SOS suicide prevention program: Further evidence of efficacy and effectiveness. Prevention Science, 17(2), 157–166. 10.1007/s11121-015-0594-3 [DOI] [PubMed] [Google Scholar]
  48. Segers M, & Rawana J (2014). What do we know about suicidality in autism spectrum disorders? A systematic review. Autism Research, 7(4), 507–521. 10.1002/aur.1375 [DOI] [PubMed] [Google Scholar]
  49. Shain B, & AAP Committee On Adolescence. (2016). Suicide and suicide attempts in adolescents. Pediatrics, 138(1). 10.1542/peds.2016-1420 [DOI] [PubMed] [Google Scholar]
  50. Sheftall AH, & Miller AB (2021). Setting a ground zero research agenda for preventing Black youth suicide. JAMA Pediatrics. 10.1001/jamapediatrics.2021.1112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. South M, Costa AP, & McMorris C (2021). Death by suicide among people with autism: beyond zebrafish. JAMA network open, 4(1), e2034018–e2034018. [DOI] [PubMed] [Google Scholar]
  52. Spassiani NA, Abou Chacra MS, Selick A, Durbin J, & Lunsky Y (2020). Emergency department nurses’ knowledge, skills, and comfort related to caring for patients with intellectual disabilities. International Emergency Nursing, 50, 100851. 10.1016/j.ienj.2020.100851 [DOI] [PubMed] [Google Scholar]
  53. Thomas P, Zahorodny W, Peng B, Kim S, Jani N, Halperin W, & Brimacombe M (2012). The association of autism diagnosis with socioeconomic status. Autism, 16(2), 201–213. 10.1177/1362361311413397 [DOI] [PubMed] [Google Scholar]
  54. Wyman PA, Brown CH, LoMurray M, Schmeelk-Cone K, Petrova M, Yu Q, Walsh E, Tu X, & Wang W (2010). A An outcome evaluation of the sources of strength suicide prevention program delivered by adolescent peer leaders in high schools. American Journal of Public Health, 100(9), 1653–1661. 10.2105/AJPH.2009.190025 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES