Abstract
Objectives. To examine epidemiological patterns of injury fatalities in individuals with a diagnosis of autism.
Methods. We identified individuals with a diagnosis of autism who died between 1999 and 2014 by screening causes of death in the multiple cause-of-death data files in the National Vital Statistics System based on the International Classification of Diseases, 10th Revision, code F84.0. We used the general US population as the reference to calculate proportionate mortality ratios (PMRs) and 95% confidence intervals (CIs).
Results. During the study period, 1367 deaths (1043 males and 324 females) in individuals with autism were recorded in the United States. The mean age at death for individuals with autism was 36.2 years (SD = 20.9 years), compared with 72.0 years (SD = 19.2 years) for the general population. Of the deaths in individuals with autism, 381 (27.9%) were attributed to injury (PMR = 2.93; 95% CI = 2.64, 3.24), with suffocation (n = 90; PMR = 31.93; 95% CI = 25.69, 39.24) being the leading cause of injury mortality, followed by asphyxiation (n = 78; PMR = 13.50; 95% CI = 10.68, 16.85) and drowning (n = 74; PMR = 39.89; 95% CI = 31.34, 50.06).
Conclusions. Individuals with autism appear to be at substantially heightened risk for death from injury.
Autism spectrum disorder is a developmental disability that can cause significant social communication and behavioral challenges.1 Patterns of impairment in social interactions and communication and restricted, repetitive, and stereotyped behaviors may emerge in early childhood, usually between age 2 and 3 years, and last throughout adulthood.1–4 Examples of such maladaptive behaviors include self-harm and forms of aggression with various degrees of severity.5,6
The reported prevalence of autism spectrum disorder has increased markedly in the past 2 decades.7 The estimated prevalence of autism spectrum disorder from the Autism and Developmental Disabilities Network was 1 in 68 children aged 8 years, and this rate has more than doubled between 2000 and 2012.7 The disorder is about 4 times as common in males as in females.2 Prevalence tends to be higher among non-Hispanic White children and among children of highly educated parents.2
Excess mortality has been reported in the autism spectrum disorder population. Previous studies have suggested that the risk of premature death is 2- to 10-fold higher in the autism spectrum disorder population than in the general population, particularly among females.8–10 However, autism spectrum disorder alone rarely causes death. Thus, factors contributing to the excess mortality are likely related to comorbid conditions, such as schizophrenia, attention-deficit/hyperactivity disorder, epilepsy, and depression.11,12 Aside from the natural disease causes of death, autistic individuals appear to be at heightened risk for death from unintentional and intentional injuries.8–10 Previous studies were limited by small numbers of deaths and to select age groups and geographic locations. The purpose of this study was to document the epidemiological patterns of injury fatalities in individuals with a recorded diagnosis of childhood autism in the United States from 1999 through 2014.
METHODS
Data for this study came from the multiple cause-of-death data files in the National Vital Statistics System from 1999 to 2014. The multiple cause-of-death data files are a census of death certificates in the United States and contain information about the causes of death and demographic characteristics. During the 16-year study period, a total of 39 524 849 deaths were recorded in the National Vital Statistics System. We identified decedents with a diagnosis of autism by screening the causes of death and conditions contributing to the death in the multiple cause-of-death data files based on the International Classification of Diseases, 10th Revision (ICD-10),13 code F84.0 (designated for autism or autistic disorder). Up to 20 ICD-10 codes could be entered into the death certificate, with the first code being the underlying cause of death and other codes being contributory causes of death. Injury causes were identified by ICD-10 codes S00 to Y99. For death certificates that did not list any injury as the underlying cause of death, the first injury code listed as the condition contributing to the death was used as the injury cause in the analysis.
We used the general US population as the reference to calculate proportionate mortality ratios (PMRs) and 95% confidence intervals (CIs) for specific causes of death, assuming that the mortality data followed a Poisson distribution. We used PMRs instead of standardized mortality ratios because population data for those with autism were unavailable.
RESULTS
During 1999 through 2014, a total of 1367 deaths (1043 males and 324 females) in individuals with a diagnosis of autism were recorded in the United States. The annual number of decedents with a diagnosis of autism was 27 (accounting for 0.001% of all deaths) in 1999, which increased to 184 (0.007% of all deaths) in 2014. The mean age at death for individuals with autism was 36.2 years (SD = 20.9 years; median = 34.0 years), whereas the mean age at death for the general population was 72.0 years (SD = 19.2 years; median = 77.0 years).
Of all the recorded deaths in individuals with autism, 381 (27.9%) were attributed to injury. Of the 381 injury fatalities in individuals with a diagnosis of autism, 40.4% occurred in homes or residential institutions. The mean age of injury deaths in individuals with autism was 29.1 years (SD = 19.0 years; median = 26.0 years) compared with 54.7 years (SD = 24.6 years; median = 54.0 years) in the general population.
When adjusted for age, sex, and calendar year, autism was associated with a significantly increased proportionate mortality from unintentional injury (PMR = 2.93; 95% CI = 2.64, 3.24; Table 1). The excess proportionate mortality from unintentional injury was particularly pronounced in autistic children younger than 15 years (PMR = 41.87; 95% CI = 34.33, 50.59). The most common injury cause of death in individuals with a diagnosis of autism was suffocation (n = 90), followed by asphyxiation (n = 78) and drowning (n = 74; Table 1). These 3 injury causes were each associated with substantially heightened proportionate mortality in both sexes (Table 1).
TABLE 1—
Age-Adjusted Proportionate Mortality Ratios (PMRs) and 95% Confidence Intervals (CIs), by Cause of Death, for Individuals With a Diagnosis of Autism, by Sex: United States, 1999–2014
| Males |
Females |
Total |
|||||||
| Cause of Death | Observed | Expected | PMR (95% CI) | Observed | Expected | PMR (95% CI) | Observed | Expected | PMR (95% CI) |
| Injury | 296 | 82.97 | 3.57 (3.17, 4.00) | 85 | 47.12 | 1.80 (1.44, 2.23) | 381 | 130.08 | 2.93 (2.64, 3.24) |
| Asphyxiation | 60 | 3.04 | 19.73 (15.07, 25.38) | 18 | 2.74 | 6.58 (3.90, 10.34) | 78 | 5.78 | 13.50 (10.68, 16.85) |
| Drowning | 57 | 1.47 | 38.66 (29.30, 50.05) | 17 | 0.38 | 44.68 (26.07, 71.18) | 74 | 1.85 | 39.89 (31.34, 50.06) |
| Suffocation | 67 | 1.56 | 42.82 (33.20, 54.35) | 23 | 1.25 | 18.34 (11.64, 27.43) | 90 | 2.82 | 31.93 (25.69, 39.24) |
| Other | 112 | 75.49 | 1.48 (1.22, 1.79) | 27 | 56.43 | 0.42 (0.32, 0.48) | 139 | 131.91 | 1.05 (0.89, 1.24) |
| All other causes of death (excluding injury) | 747 | 572.59 | 1.30 (1.21, 1.40) | 239 | 633.62 | 0.38 (0.33, 0.43) | 986 | 1206.21 | 0.82 (0.77, 0.87) |
DISCUSSION
Although previous research has reported increased mortality in individuals with autism, injury mortality in the autism spectrum disorder population is understudied. We examined national mortality data and found that individuals with a diagnosis of autism died on average 35.8 years younger than individuals in the general population. For deaths caused by injury, the difference was almost as striking. We also found that the annual number of recorded deaths for individuals with a diagnosis of autism has risen nearly 7 times from 1999 to 2014. This number has likely increased because of the increasing prevalence of autism resulting from improved detection and diagnosis of the condition and changes in the diagnostic criteria. Despite the marked increase in the annual number of deaths occurring in individuals with autism, autism-related deaths still may be severely underreported, particularly deaths from intentional injury such as assaults, homicide, and suicide.
We also found that deaths in individuals with autism were nearly 3 times as likely as were deaths in the general population to be caused by unintentional injury. The excess risk of unintentional injury associated with autism was particularly high for children younger than 15 years and for 3 specific causes: drowning, suffocation, and asphyxiation. Together, these 3 causes accounted for 79.4% of the total injury mortality in children with autism.
This study had several notable limitations. First, in the absence of population data on autism, we were unable to calculate the death rates and cause-specific standardized mortality ratios. Instead, we relied on proportionate mortality analysis to explore the relative burden of injury mortality in individuals with autism. In proportionate mortality analysis, a cause-specific PMR is influenced by changes in the number of deaths from the specific cause or the number of deaths from all other causes. Consequently, the heightened PMR from injury observed in individuals with autism may not necessarily measure the excess risk of injury mortality in individuals with autism as compared with the general population. Second, our study was limited to death certificate data. The accuracy of data on death certificates filed by medical examiners and coroners varies with the cause of death and the types of disease. Therefore, autism as a contributing cause of death is likely underreported.
PUBLIC HEALTH IMPLICATIONS
Findings from this study indicate that individuals with a diagnosis of autism appear to be at substantially heightened risk for death from unintentional injury, particularly for children younger than 15 years. Because suffocation, asphyxiation, and drowning are the leading injury causes of death for individuals with autism, prevention programs specifically targeting these risks must be developed and implemented. Given the exceptionally heightened risk of drowning for children with autism, clinicians, parents, and social workers should help provide swimming lessons to these children as soon as the diagnosis is made.
ACKNOWLEDGMENTS
This project was supported in part by the National Center for Injury Prevention and Control, Centers for Disease Control and Prevention (grant 1 R49 CE002096).
We thank Stanford Chihuri, MPH, and Margaret Warner, PhD, for their assistance with data management and statistical analysis and Barbara H. Lang, MPH, for administrative and editorial assistance.
Note. The contents of this article are solely the responsibility of the authors and do not represent the official views of the Centers for Disease Control and Prevention or the US Department of Health and Human Services.
HUMAN PARTICIPANT PROTECTION
This study meets the criteria for the Protection of Human Subjects exemption 4 (research involving preexisting data) of the US Code of Federal Regulations (45 CFR 46.101). The study was deemed exempt from review by the institutional review board’s Administrative Review Committee at Columbia University Medical Center.
REFERENCES
- 1.Centers for Disease Control and Prevention. Autism spectrum disorder (ASD): data & statistics. July 11, 2016. Available at: http://www.cdc.gov/ncbddd/autism/data.html. Accessed September 2, 2016.
- 2.Perou R, Bitsko RH, Blumberg SJ et al. Mental health surveillance among children—United States, 2005-2011. MMWR Suppl. 2013;62(2):1–35. [PubMed] [Google Scholar]
- 3.Baxter AJ, Brugha TS, Erskine HE, Scheurer RW, Vos T, Scott JG. The epidemiology and global burden of autism spectrum disorders. Psychol Med. 2015;45(3):601–613. doi: 10.1017/S003329171400172X. [DOI] [PubMed] [Google Scholar]
- 4.Autism Speaks. What Is Autism? 2016. Available at: https://www.autismspeaks.org/what-autism. August 9, 2016. Accessed September 2, 2016.
- 5.Carroll D, Hallett V, McDougle CJ et al. Examination of aggression and self-injury in children with autism spectrum disorders and serious behavioral problems. Child Adolesc Psychiatr Clin N Am. 2014;23(1):57–72. doi: 10.1016/j.chc.2013.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Soke GN, Rosenberg SA, Hamman RF et al. Factors associated with self-injurious behaviors in children with autism spectrum disorder: findings from two large national samples. J Autism Dev Disord. 2017;47(2):285–296. doi: 10.1007/s10803-016-2951-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Christensen DL, Baio J, Van Naarden Braun K et al. Prevalence and characteristics of autism spectrum disorder among children aged 8 years–Autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2012. MMWR Surveill Summ. 2016;65(3):1–23. doi: 10.15585/mmwr.ss6503a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Gillberg C, Billstedt E, Sundh V, Gillberg IC. Mortality in autism: a prospective longitudinal community-based study. J Autism Dev Disord. 2010;40(3):352–357. doi: 10.1007/s10803-009-0883-4. [DOI] [PubMed] [Google Scholar]
- 9.Mouridsen SE, Bronnum-Hansen H, Rich B, Isager T. Mortality and causes of death in autism spectrum disorders: an update. Autism. 2008;12(4):403–414. doi: 10.1177/1362361308091653. [DOI] [PubMed] [Google Scholar]
- 10.Pickett J, Xiu E, Tuchman R, Dawson G, Lajonchere C. Mortality in individuals with autism, with and without epilepsy. J Child Neurol. 2011;26(8):932–939. doi: 10.1177/0883073811402203. [DOI] [PubMed] [Google Scholar]
- 11.Schendel DE, Overgaard M, Christensen J et al. Association of psychiatric and neurologic comorbidity with mortality among persons with autism spectrum disorder in a Danish population. JAMA Pediatr. 2016;170(3):243–250. doi: 10.1001/jamapediatrics.2015.3935. [DOI] [PubMed] [Google Scholar]
- 12.Bilder D, Botts EL, Smith KR et al. Excess mortality and causes of death in autism spectrum disorders: a follow up of the 1980s Utah/UCLA autism epidemiologic study. J Autism Dev Disord. 2013;43(5):1196–1204. doi: 10.1007/s10803-012-1664-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.World Health Organization. International Statistical Classification of Diseases and Related Health Problems, 10th Revision. Geneva, Switzerland: World Health Organization; 1992. [Google Scholar]