Assessment of Educational Attainment and Employment Among Individuals With Autism Spectrum Disorder in Denmark | Adolescent Medicine | JAMA Pediatrics | JAMA Network
[Skip to Navigation]
Sign In
Table 1.  Characteristics of Individuals With ASD and Without ASD
Characteristics of Individuals With ASD and Without ASD
Table 2.  Characteristics of Parents of Individuals With and Without Autism Spectrum Disorder
Characteristics of Parents of Individuals With and Without Autism Spectrum Disorder
Table 3.  Educational Attainment and Employment at Age 25 Years Among Individuals With and Without Autism Spectrum Disorder
Educational Attainment and Employment at Age 25 Years Among Individuals With and Without Autism Spectrum Disorder
Table 4.  Age at Completion of Ninth Grade and Upper Secondary School and Time to Employment Among Individuals With and Without Autism Spectrum Disorder
Age at Completion of Ninth Grade and Upper Secondary School and Time to Employment Among Individuals With and Without Autism Spectrum Disorder
Table 5.  Employment Status and Job Types at Age 25 Years Among Individuals With and Without Autism Spectrum Disorder
Employment Status and Job Types at Age 25 Years Among Individuals With and Without Autism Spectrum Disorder
1.
Lai  MC, Lombardo  MV, Baron-Cohen  S.  Autism.   Lancet. 2014;383(9920):896-910. doi:10.1016/S0140-6736(13)61539-1 PubMedGoogle ScholarCrossref
2.
Steinhausen  HC, Mohr Jensen  C, Lauritsen  MB.  A systematic review and meta-analysis of the long-term overall outcome of autism spectrum disorders in adolescence and adulthood.   Acta Psychiatr Scand. 2016;133(6):445-452. doi:10.1111/acps.12559 PubMedGoogle ScholarCrossref
3.
Keen  D, Webster  A, Ridley  G.  How well are children with autism spectrum disorder doing academically at school? an overview of the literature.   Autism. 2016;20(3):276-294. doi:10.1177/1362361315580962 PubMedGoogle ScholarCrossref
4.
Shattuck  PT, Carter Narendorf  S, Cooper  B, Sterzing  PR, Wagner  M, Lounds Taylor  J.  Postsecondary education and employment among youth with an autism spectrum disorder.   Pediatrics. 2012;129(6):1042-1049. doi:10.1542/peds.2011-2864 PubMedGoogle ScholarCrossref
5.
Dalsgaard  S, McGrath  J, Ostergaard  SD,  et al.  Association of mental disorder in childhood and adolescence with subsequent educational achievement.   JAMA Psychiatry. 2020;77(8):797-805. doi:10.1001/jamapsychiatry.2020.0217 PubMedGoogle ScholarCrossref
6.
Pedersen  CB, Gotzsche  H, Moller  JO, Mortensen  PB.  The Danish Civil Registration System. a cohort of eight million persons.   Dan Med Bull. 2006;53(4):441-449.PubMedGoogle Scholar
7.
Austin  PC.  An introduction to propensity score methods for reducing the effects of confounding in observational studies.   Multivariate Behav Res. 2011;46(3):399-424. doi:10.1080/00273171.2011.568786 PubMedGoogle ScholarCrossref
8.
Mors  O, Perto  GP, Mortensen  PB.  The Danish Psychiatric Central Research Register.   Scand J Public Health. 2011;39(7)(suppl):54-57. doi:10.1177/1403494810395825 PubMedGoogle ScholarCrossref
9.
United Nations Educational, Scientific and Cultural Organization. International Standard Classification of Education (ISCED) 2011. UNESCO Institute for Statistics; 2012. Accessed June 29, 2020. http://uis.unesco.org/sites/default/files/documents/international-standard-classification-of-education-isced-2011-en.pdf
10.
Jensen  VM, Rasmussen  AW.  Danish education registers.   Scand J Public Health. 2011;39(7)(suppl):91-94. doi:10.1177/1403494810394715 PubMedGoogle ScholarCrossref
11.
International Labour Organization. International Standard Classification of Occupations (ISCO). Updated June 9, 2010. Accessed June 29, 2020. http://www.ilo.org/public/english/bureau/stat/isco/index.htm
12.
Brookhart  MA, Sturmer  T, Glynn  RJ, Rassen  J, Schneeweiss  S.  Confounding control in healthcare database research: challenges and potential approaches.   Med Care. 2010;48(6)(suppl):S114-S120. doi:10.1097/MLR.0b013e3181dbebe3 PubMedGoogle ScholarCrossref
13.
Spiegelman  D, Hertzmark  E.  Easy SAS calculations for risk or prevalence ratios and differences.   Am J Epidemiol. 2005;162(3):199-200. doi:10.1093/aje/kwi188 PubMedGoogle ScholarCrossref
14.
Barros  AJD, Hirakata  VN.  Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio.   BMC Med Res Methodol. 2003;3:21. doi:10.1186/1471-2288-3-21 PubMedGoogle ScholarCrossref
15.
Howlin  P, Moss  P, Savage  S, Rutter  M.  Social outcomes in mid- to later adulthood among individuals diagnosed with autism and average nonverbal IQ as children.   J Am Acad Child Adolesc Psychiatry. 2013;52(6):572-581. doi:10.1016/j.jaac.2013.02.017 PubMedGoogle ScholarCrossref
16.
Farley  MA, McMahon  WM, Fombonne  E,  et al.  Twenty-year outcome for individuals with autism and average or near-average cognitive abilities.   Autism Res. 2009;2(2):109-118. doi:10.1002/aur.69 PubMedGoogle ScholarCrossref
17.
Weaver  LL. Effectiveness of work, activities of daily living, education, and sleep interventions for people with autism spectrum disorder: a systematic review. Am J Occup Ther. 2015;69(5):6905180020. doi:10.5014/ajot.2015.017962
18.
Hirvikoski  T, Boman  M, Chen  Q,  et al.  Individual risk and familial liability for suicide attempt and suicide in autism: a population-based study.   Psychol Med. 2020;50(9):1463-1474. doi:10.1017/S0033291719001405 PubMedGoogle ScholarCrossref
19.
Schendel  DE, Thorsteinsson  E.  Cumulative incidence of autism into adulthood for birth cohorts in Denmark, 1980-2012.   JAMA. 2018;320(17):1811-1813. doi:10.1001/jama.2018.11328 PubMedGoogle ScholarCrossref
20.
Magiati  I, Tay  XW, Howlin  P.  Cognitive, language, social and behavioural outcomes in adults with autism spectrum disorders: a systematic review of longitudinal follow-up studies in adulthood.   Clin Psychol Rev. 2014;34(1):73-86. doi:10.1016/j.cpr.2013.11.002 PubMedGoogle ScholarCrossref
21.
Petersson  F, Baadsgaard  M, Thygesen  LC.  Danish registers on personal labour market affiliation.   Scand J Public Health. 2011;39(7)(suppl):95-98. doi:10.1177/1403494811408483 PubMedGoogle ScholarCrossref
22.
Lauritsen  MB, Jorgensen  M, Madsen  KM,  et al.  Validity of childhood autism in the Danish Psychiatric Central Register: findings from a cohort sample born 1990-1999.   J Autism Dev Disord. 2010;40(2):139-148. doi:10.1007/s10803-009-0818-0 PubMedGoogle ScholarCrossref
23.
Rothman  KJ.  Six persistent research misconceptions.   J Gen Intern Med. 2014;29(7):1060-1064. doi:10.1007/s11606-013-2755-z PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    April 5, 2021

    Assessment of Educational Attainment and Employment Among Individuals With Autism Spectrum Disorder in Denmark

    Author Affiliations
    • 1Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark
    • 2Steno Diabetes Center, Aarhus University Hospital, Aarhus, Denmark
    • 3Personalized Health Care Data Science, F. Hoffmann-La Roche, Basel, Switzerland
    • 4Lundbeck Foundation Initiative for Integrative Psychiatric Research, Aarhus, Denmark
    • 5National Centre for Register-Based Research, Department of Economics and Business, Aarhus University, Aarhus, Denmark
    • 6Department of Public Health, Aarhus University, Aarhus, Denmark
    JAMA Pediatr. 2021;175(6):601-608. doi:10.1001/jamapediatrics.2021.0124
    Key Points

    Question  Is diagnosis of autism spectrum disorder in childhood associated with educational attainment and employment?

    Findings  In this cross-sectional study of 8910 individuals with and without autism spectrum disorder identified through population-based registers in Denmark, the proportion of those with autism spectrum disorder who completed compulsory primary and lower secondary education (ninth grade) was similar to that for individuals without autism spectrum disorder. At age 25 years, individuals with autism spectrum disorder had a lower prevalence of educational attainment beyond ninth grade and lower employment compared with those without autism spectrum disorder.

    Meaning  This study’s findings indicated that a diagnosis of autism spectrum disorder was not associated with a lower prevalence of compulsory educational attainment but was associated with a lower prevalence of further educational attainment and employment.

    Abstract

    Importance  Population-based data on educational and employment outcomes in adulthood among individuals diagnosed with autism spectrum disorder (ASD) in childhood are currently limited.

    Objective  To evaluate educational attainment and employment among individuals with and without a diagnosis of ASD before age 12 years in Denmark.

    Design, Setting, and Participants  This nationwide cross-sectional prevalence study was conducted using data from Danish registers. Individuals with a diagnosis of ASD recorded before age 12 years were identified among all individuals born in Denmark between January 1, 1989, and December 31, 1991, who were alive at age 25 years. Individuals with ASD were then matched on a 10:1 ratio by age, sex, and region of residence with a comparison population of individuals without a diagnosis of ASD at age 12 years. Data were analyzed from March 2019 to December 2020.

    Exposures  Autism spectrum disorder diagnosis and diagnostic subtype recorded before age 12 years.

    Main Outcomes and Measures  Adjusted prevalence ratios (aPRs) with 95% CIs for the completion of compulsory primary and lower secondary school (grade 9), upper secondary school (grades 10-12 or vocational), and tertiary school (university) and for employment by age 25 years were estimated using log-binomial regression analysis.

    Results  A total of 810 individuals with a diagnosis of ASD before age 12 years were matched with a comparison population of 8100 individuals without ASD. The prevalence of ninth-grade completion was similar among those with and without ASD (785 individuals [96.9%] and 7982 individuals [98.5%], respectively; aPR, 0.98; 95% CI, 0.97-1.00). Compared with those without ASD, persons with ASD had a lower prevalence of completing upper secondary school (6338 individuals [78.2%] vs 286 individuals [35.3%], respectively; aPR, 0.46; 95% CI, 0.42-0.50) and tertiary school (2185 individuals [27.0%] vs 70 individuals [8.6%]; aPR, 0.33; 95% CI, 0.26-0.41) and obtaining employment (4284 individuals [77.7%] vs 177 individuals [27.2%]; aPR, 0.35; 95% CI, 0.31-0.40) at age 25 years. A ninth-grade final examination score was available for 394 individuals (48.6%) with ASD and 7417 individuals (91.6%) without ASD. In an analysis stratified by ASD subtype, individuals diagnosed with childhood autism had lower educational attainment and employment than those diagnosed with Asperger syndrome or pervasive developmental disorder not otherwise specified. A total of 461 individuals (56.9%) with ASD were receiving public assistance or a pension (ie, disability benefits) at age 25 years compared with 1094 individuals (13.5%) without ASD in the comparison population.

    Conclusions and Relevance  In this population-based cross-sectional study, a diagnosis of ASD in childhood was not associated with the completion of compulsory primary and lower secondary education (ninth grade). An ASD diagnosis before age 12 years was associated with a lower prevalence of attaining education beyond ninth grade and obtaining employment by age 25 years, indicating a substantially higher risk of reliance on public assistance in young adulthood.

    Introduction

    Autism spectrum disorder (ASD) is a heritable condition comprising a broad range of disability levels from mild to severe and is characterized by deficits in social communication and unusually restricted repetitive behavior and interests. The worldwide prevalence of ASD is approximately 1%.1 Data regarding long-term educational attainment and employment among persons with ASD are sparse and often based on selected or small populations.

    A 2016 systematic review and meta-analysis of overall outcomes among individuals with ASD identified 15 studies from the United Kingdom, Scandinavia, North America, and Japan, with a mean age at follow-up ranging from 12 to 44 years but comprising only 828 total individuals with ASD.2 The overall outcome of those with ASD was reported to be good (normal or near-normal social life and satisfactory functioning at school or work) for 20% of individuals, fair (social and educational progress despite social abnormalities) for 31% of individuals, and poor (severely impaired but with some potential for social progress) or very poor (unable to lead any kind of independent existence) for 48% of individuals.2 Another review reported considerable variability in school performance among persons with ASD.3

    The largest US study of 500 individuals with ASD who were receiving special education services indicated that 35% had attended college and 55% had paid employment during the first 6 years after high school completion.4 A large Danish population-based study of children with mental disorders recently found that children with childhood autism or Asperger syndrome (2 diagnostic subtypes of ASD) completed the ninth-grade final examination at a lower proportion than children without mental disorders, and fewer children with childhood autism completed the final ninth-grade examination compared with children with Asperger syndrome.5 To extend the knowledge of educational and employment outcomes among young adults with ASD, we aimed to examine the association of ASD with educational attainment and employment in a population-based study of individuals with and without ASD in Denmark.

    Methods
    Study Design and Population

    This cross-sectional prevalence study used routinely collected data from population-based nationwide registries in Denmark. Data were accessed via researcher services on Statistics Denmark. The study was registered at the Danish Data Protection Agency from Aarhus University and conducted in accordance with the Danish Act on Processing of Personal Data, which specifies that informed consent is not required for register-based studies in Denmark. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.

    The study population comprised individuals in Denmark who were aged 12 years between January 1, 2001, and December 31, 2003 (born from January 1, 1989, to December 31, 1991), and who had a diagnosis of ASD recorded before age 12 years. A comparison group without an ASD diagnosis before age 12 years was obtained from the general Danish population6 and matched on a 10:1 ratio, with replacement,7 on birth year, sex, and region of residence with each person with an ASD diagnosis. Both groups were observed until age 25 years. We excluded individuals who died or emigrated before age 25 years and individuals who did not have information on mother, father, sex, or region of residence (additional inclusion and exclusion criteria are available in the eFigure in the Supplement).

    Definitions, Outcomes, and Covariates

    Autism spectrum disorder was defined as any ASD diagnosis recorded in the Danish Psychiatric Central Register8 using International Classification of Diseases, Eighth Revision (ICD-8) codes 299.00, 299.01, 299.02, and 299.03 for diagnoses before 1994 and International Classification of Diseases, Tenth Revision (ICD-10) codes F84.0, F84.1, F84.5, F84.8, and F84.9 for diagnoses in 1994 and later. The Danish Psychiatric Central Register includes discharge diagnoses from inpatient admissions through 1994 and from both inpatient and outpatient hospital encounters beginning in 1995. In the stratified analysis, ASD subtypes included childhood autism (ICD-8 code 299.00 and ICD-10 code F84.0), Asperger syndrome (ICD-8 code 299.02 and ICD-10 code F84.5), or pervasive developmental disorder not otherwise specified (ICD-8 codes 299.01 and 299.03 and ICD-10 codes F84.1, F84.8, and F84.9).

    Educational attainment was classified according to the framework of the International Standard Classification of Education (ISCED), 2011 version.9 Data on educational attainment (yes or no) at age 25 years (2014-2016) were obtained from the Population Education Register of Denmark. Data were collected for the completion of 3 educational levels: compulsory primary and lower secondary school (grade 9; ISCED levels 1-2), upper secondary school (grades 10-12 or vocational; ISCED level 3), and tertiary school (university; ISCED levels ≥4).10 In addition, mean test scores at the end of ninth grade were obtained from the Danish Academic Achievement Register.

    Five distinct categories were used for employment status: unemployed, student, receiving public assistance, employed, receiving a pension (ie, disability benefits), and self-employed (including working at spouse’s company). Employment at age 25 years was defined as having a paid job for more than 50% of the year (yes or no). Employment status was also categorized by skill level and skills specialization based on 1-digit codes from the Danish version of the International Standard Classification of Occupations published in 2008 (DISCO-08),11 via the Employment Classification Module.

    The following covariates were assessed separately for parents of the study population: (1) age (in years) at the child’s birth (continuous variables), which was obtained from the Danish Civil Registration System; (2) recorded psychiatric disorders (schizophrenia, schizotypal, and delusional disorders; affective disorders; neurotic, stress-related, and somatoform disorders; disorders of adult personality and behavior; and ASD) that were diagnosed from 5 years before the child’s birth to the beginning of follow-up at the child’s twelfth birthday, which were obtained from the Danish Psychiatric Central Register (eTable 1 in the Supplement); and (3) socioeconomic status in the year before the child’s birth (representing prenatal exposure), defined by employment status and DISCO-08 job codes obtained from the Employment Classification Module.

    Statistical Analysis

    We selected potential confounding variables based on previously published data.12 Covariates associated with the outcomes were included to increase the precision of the estimates. In addition to an unadjusted model, we used a model adjusted for parental characteristics that included age, socioeconomic status, and psychiatric disorders.

    Log-binomial regression analysis was used to estimate prevalence ratios (PRs) with 95% CIs (unadjusted and adjusted for the covariates) for educational level (completion of ninth grade, upper secondary, or tertiary education) and employment status (yes or no) among individuals with and without ASD. The ASD subtypes (childhood autism, Asperger syndrome, and pervasive developmental disorder not otherwise specified) were also compared.13 Students were excluded from the comparison of employment status because not all students had completed their educations by age 25 years. Supplementary analyses, stratified by sex and restricted to individuals without intellectual disabilities, were performed to obtain unadjusted results on educational attainment and employment for individuals with and without ASD. For some categories, exact numbers and percentages were not calculated per the Statistics Denmark requirement that exact counts not be reported for groups comprising less than 5 people.

    Ordinal logistic regression analysis was used for calculating prevalence odds ratios for different levels of ninth-grade final examination scores among individuals with and without ASD and among the ASD subtypes, with the highest grade levels (10-12) used as the reference category. Hazard ratios (HRs) were calculated using Cox regression analysis to compare age (in years) at the completion of ninth grade with age (in years) at the completion of upper secondary school among individuals with and without ASD and among the ASD subtypes, adjusted for a similar set of covariates.14

    Multinomial logistic regression analysis was used to calculate odds ratios comparing employment status and job types among individuals with and without ASD and among the ASD subtypes, with employed status and professional job type (DISCO-08 group 2) used as reference categories. Data were analyzed using SAS software, version 9.4 (SAS Institute), from March 2019 to December 2020.

    Results

    A total of 810 individuals with a diagnosis of ASD before age 12 years were matched with a comparison population of 8100 individuals (Table 1). Maternal and paternal age distributions and socioeconomic status were similar between the 2 groups. The mothers of individuals with ASD had a higher prevalence of psychiatric disorders compared with the mothers of those without ASD (eg, 63 mothers [7.8%] vs 274 mothers [3.4%] had neurotic, stress-related, or somatoform disorders, respectively) (Table 2). With regard to paternal mental disorders, the numbers were low, but no major differences were observed. A total of 785 individuals (96.9%) with ASD and 7982 individuals (98.5%) without ASD attained a ninth-grade education (adjusted PR [aPR], 0.98; 95% CI, 0.97-1.00) (Table 3). However, compared with individuals without ASD, those with ASD had a substantially lower prevalence of completing upper secondary school (6338 individuals [78.2%] vs 286 individuals [35.3%], respectively; aPR, 0.46; 95% CI, 0.42-0.50) and tertiary school (2185 individuals [27.0%] vs 70 individuals [8.6%]; aPR, 0.33; 95% CI, 0.26-0.41) and of obtaining employment (4284 individuals [77.7%] vs 177 individuals [27.2%]; aPR, 0.35; 95% CI, 0.31-0.40) by age 25 years.

    The analysis stratified by ASD subtype indicated that individuals diagnosed with childhood autism had a lower prevalence of educational attainment after ninth grade (eg, 21 individuals [14.4%] completed upper secondary school; aPR, 0.18; 95% CI, 0.12-0.27) and a lower prevalence of employment (10 individuals [7.9%] were employed >50% of the year; aPR, not estimated because of small numbers) compared with individuals with Asperger syndrome (eg, 130 individuals [52.2%] completed upper secondary school [aPR, 0.66; 95% CI, 0.59-0.75], and 57 individuals [33.1%] were employed >50% of the year [aPR, 0.42; 95% CI, 0.34-0.52]) and pervasive developmental disorder not otherwise specified (eg, 135 individuals [32.5%] completed upper secondary school [aPR, 0.43; 95% CI, 0.37-0.49], and 110 individuals [31.2%] were employed >50% of the year [aPR, 0.41; 95% CI, 0.35-0.48]). Unadjusted analyses, stratified by sex, among individuals with and without ASD indicated higher PRs for upper secondary and tertiary educational attainment and employment among men compared with women (eTable 2 in the Supplement). When restricted to individuals without intellectual disabilities, those with ASD were more likely to complete their educations and obtain a job (eTable 2 in the Supplement) than those in the nonrestricted comparison (Table 3).

    A total of 394 individuals (48.6%) with ASD and 7417 individuals (91.6%) without ASD completed the ninth-grade final examination. However, fewer than 27.0% of individuals with childhood autism completed the test. Among those with available data on final test scores, a larger proportion of individuals with ASD achieved higher scores than those without ASD (eTable 3 in the Supplement). The stratified analyses of ASD subtypes indicated that individuals with Asperger syndrome had the highest ninth-grade final examination scores.

    The median difference in age at completion of ninth grade between individuals with and without ASD was 0.3 years (among those with ASD, median age, 15.3 years; interquartile range [IQR], 14.8-16.0 years; among those without ASD, median age, 15.1 years; IQR, 14.8-15.4 years) (Table 4). For completion of upper secondary education, the difference was more substantial, with a median difference of 1.1 years between individuals with ASD (median age, 21.3 years; IQR, 20.1-22.8 years) and without ASD (median age, 20.2 years; IQR, 19.5-21.3 years). The median difference in time to obtain employment after the completion of education was small. No substantial differences were observed between ASD subtypes, although this finding was based on small numbers.

    The prevalence of receiving a pension or other type of public assistance was higher among persons with ASD (461 individuals [56.9%]) compared with those without ASD (1094 individuals [13.5%]) (Table 5). The prevalence of employment by age 25 years was lower among individuals with ASD (177 individuals [21.9%]) compared with those without ASD (4284 individuals [52.9%]). Among those employed, a higher proportion of individuals with ASD vs without ASD were working as technicians or associate professionals (DISCO-08 group 3)11 or as plant or machine operators and assemblers (DISCO-08 group 8) relative to professionals (DISCO-08 group 2); however, the numbers were small and the estimates imprecise (Table 4).

    Discussion

    This population-based cross-sectional study examined educational attainment and employment among young adults diagnosed with ASD before age 12 years. The prevalence of ninth-grade completion was similar between individuals with and without ASD, but a substantially lower proportion of persons with ASD completed the ninth-grade final examination. The prevalence of upper secondary and tertiary educational attainment and employment at age 25 years was more than 50% lower among individuals with ASD, with longer median educational completion times, compared with individuals without ASD.

    Another recent Danish study found lower ninth-grade examination scores among children with childhood autism but no association among children with Asperger syndrome,5 suggesting that those with childhood autism perform worse in primary and lower secondary school. The results of the present study indicate that it may be more difficult for individuals with ASD to attain education after completing compulsory primary and lower secondary school, and individuals with childhood autism appear to attain additional education and obtain employment at substantially lower rates than those with other ASD subtypes.

    Information on educational attainment and employment in adulthood among individuals with ASD has been previously assessed in a few selected study populations.15,16 In the largest previous study conducted in the US,4 the proportion of individuals with ASD who attended college was similar to that of the present study, with 35.3% of individuals completing upper secondary school. However, we found that only 27.2% of individuals with ASD were employed for more than 50% of the year by age 25 years, which is lower than the 55% of individuals who had paid employment during the first 6 years after high school in the previous study.4 Differences in inclusion criteria between the US study and the present study (the US study included only individuals receiving special education, and the present study included all individuals with an ASD diagnosis) and differences in the availability of public assistance between the US and Denmark may partially explain the difference in results. A majority of individuals with ASD (56.9%) were receiving public assistance or a pension at age 25 years compared with 13.5% of those in the comparison population. Several intervention programs have been initiated to help individuals with ASD improve their skills for educational and employment functioning, but these programs have had limited success to date.17

    Limitations

    This study has several limitations. The study was restricted to persons born in Denmark who were alive at age 25 years. Mortality may be slightly higher among individuals with ASD compared with the general population because of an increased risk of suicide among those with ASD.18 In addition, this study excluded emigrants and immigrants because of incomplete medical histories, which could have slightly biased our results. In Denmark, health care and education, including university education, is publicly funded; therefore, economic barriers are limited. Some misclassification or underreporting of ASD diagnosis may have occurred, given that we excluded diagnoses after age 12 years to ensure that the ASD diagnosis was present before the examined outcomes, and girls often receive ASD diagnoses at older ages. The examined outcomes and covariates in the Danish registers may also have been misclassified, but this is unlikely to produce differential misclassification.

    The number of individuals identified with ASD in this study was similar to the number expected based on national prevalence rates of ASD in Denmark.19 A systematic review including 25 studies of 2022 individuals with ASD, with follow-up on adult cognitive, social, and behavioral outcomes among those with an ASD diagnosis before age 16 years, reported that ASD was generally persistent from childhood to adulthood; however, the severity of autism-associated behavioral symptoms generally improved in adulthood.20 We did not follow up on ASD diagnosis in our study, but it can be assumed that ASD is persistent into adulthood. The Population Education Register contains information on the highest completed educational level for 97% of the Danish population born between 1945 and 1990, and Statistics Denmark reports 0% to 3% misclassification in the register.10 The validity of the Employment Classification Module is also considered high.21 Childhood autism among Danish children born between 1990 and 1999 has previously been validated, with a positive predictive value of 94% based on medical records review of 499 cases22; however, overall ASD has not, to our knowledge, been validated.

    The study population was limited to Danish children born from January 1989 through December 1991. Although the study’s findings may not be representative of other time periods and countries, scientific conclusions based on studies with limited variability in confounding factors are considered more valid than those based on studies with a highly representative sample but greater variability in confounding factors.23

    Conclusions

    In Denmark, a diagnosis of ASD before age 12 years was not associated with the prevalence of compulsory primary and lower secondary (ninth grade) educational attainment but was associated with a lower prevalence of completing ninth-grade final examinations. Persons diagnosed with childhood autism also had a substantially lower prevalence of educational attainment beyond ninth grade and a lower prevalence of employment by age 25 years, with more than 50% of individuals with ASD dependent on some type of public assistance in young adulthood.

    Back to top
    Article Information

    Accepted for Publication: December 31, 2020.

    Published Online: April 5, 2021. doi:10.1001/jamapediatrics.2021.0124

    Corresponding Author: Gunnar Toft, PhD, DMSc, Steno Diabetes Center, Aarhus University Hospital, Hedeager 3, 8200 Aarhus N, Denmark (guntof@rm.dk).

    Author Contributions: Drs Toft and Ehrenstein had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Toft, Loss, Ehrenstein.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Toft, Ehrenstein.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Menon, Loss.

    Obtained funding: Loss.

    Administrative, technical, or material support: Liu, Schendel.

    Supervision: Toft, Liu, Ehrenstein.

    Conflict of Interest Disclosures: Dr Liu reported being an employee of F. Hoffmann-La Roche Ltd. Dr Loss reported receiving personal fees from F. Hoffmann-La Roche Ltd as an employee during the conduct of the study. Dr Ehrenstein reported receiving grants from Roche during the conduct of the study; in addition, her organization provides research services to several private and public entities, including regulators and pharmaceutical companies. No other disclosures were reported.

    Funding/Support: This study was funded by F. Hoffmann-La Roche through institutional research funding that was provided to and administered by Aarhus University.

    Role of the Funder/Sponsor: The funding organization had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication. Aarhus University had the right to overrule the sponsor in all aspects of the study, including publication of results.

    Additional Contributions: Helene Matilde Lundsgaard Svane, MSc, of the Department of Clinical Epidemiology at Aarhus University provided input on data management and statistical analyses. She did not receive additional compensation for her assistance.

    References
    1.
    Lai  MC, Lombardo  MV, Baron-Cohen  S.  Autism.   Lancet. 2014;383(9920):896-910. doi:10.1016/S0140-6736(13)61539-1 PubMedGoogle ScholarCrossref
    2.
    Steinhausen  HC, Mohr Jensen  C, Lauritsen  MB.  A systematic review and meta-analysis of the long-term overall outcome of autism spectrum disorders in adolescence and adulthood.   Acta Psychiatr Scand. 2016;133(6):445-452. doi:10.1111/acps.12559 PubMedGoogle ScholarCrossref
    3.
    Keen  D, Webster  A, Ridley  G.  How well are children with autism spectrum disorder doing academically at school? an overview of the literature.   Autism. 2016;20(3):276-294. doi:10.1177/1362361315580962 PubMedGoogle ScholarCrossref
    4.
    Shattuck  PT, Carter Narendorf  S, Cooper  B, Sterzing  PR, Wagner  M, Lounds Taylor  J.  Postsecondary education and employment among youth with an autism spectrum disorder.   Pediatrics. 2012;129(6):1042-1049. doi:10.1542/peds.2011-2864 PubMedGoogle ScholarCrossref
    5.
    Dalsgaard  S, McGrath  J, Ostergaard  SD,  et al.  Association of mental disorder in childhood and adolescence with subsequent educational achievement.   JAMA Psychiatry. 2020;77(8):797-805. doi:10.1001/jamapsychiatry.2020.0217 PubMedGoogle ScholarCrossref
    6.
    Pedersen  CB, Gotzsche  H, Moller  JO, Mortensen  PB.  The Danish Civil Registration System. a cohort of eight million persons.   Dan Med Bull. 2006;53(4):441-449.PubMedGoogle Scholar
    7.
    Austin  PC.  An introduction to propensity score methods for reducing the effects of confounding in observational studies.   Multivariate Behav Res. 2011;46(3):399-424. doi:10.1080/00273171.2011.568786 PubMedGoogle ScholarCrossref
    8.
    Mors  O, Perto  GP, Mortensen  PB.  The Danish Psychiatric Central Research Register.   Scand J Public Health. 2011;39(7)(suppl):54-57. doi:10.1177/1403494810395825 PubMedGoogle ScholarCrossref
    9.
    United Nations Educational, Scientific and Cultural Organization. International Standard Classification of Education (ISCED) 2011. UNESCO Institute for Statistics; 2012. Accessed June 29, 2020. http://uis.unesco.org/sites/default/files/documents/international-standard-classification-of-education-isced-2011-en.pdf
    10.
    Jensen  VM, Rasmussen  AW.  Danish education registers.   Scand J Public Health. 2011;39(7)(suppl):91-94. doi:10.1177/1403494810394715 PubMedGoogle ScholarCrossref
    11.
    International Labour Organization. International Standard Classification of Occupations (ISCO). Updated June 9, 2010. Accessed June 29, 2020. http://www.ilo.org/public/english/bureau/stat/isco/index.htm
    12.
    Brookhart  MA, Sturmer  T, Glynn  RJ, Rassen  J, Schneeweiss  S.  Confounding control in healthcare database research: challenges and potential approaches.   Med Care. 2010;48(6)(suppl):S114-S120. doi:10.1097/MLR.0b013e3181dbebe3 PubMedGoogle ScholarCrossref
    13.
    Spiegelman  D, Hertzmark  E.  Easy SAS calculations for risk or prevalence ratios and differences.   Am J Epidemiol. 2005;162(3):199-200. doi:10.1093/aje/kwi188 PubMedGoogle ScholarCrossref
    14.
    Barros  AJD, Hirakata  VN.  Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio.   BMC Med Res Methodol. 2003;3:21. doi:10.1186/1471-2288-3-21 PubMedGoogle ScholarCrossref
    15.
    Howlin  P, Moss  P, Savage  S, Rutter  M.  Social outcomes in mid- to later adulthood among individuals diagnosed with autism and average nonverbal IQ as children.   J Am Acad Child Adolesc Psychiatry. 2013;52(6):572-581. doi:10.1016/j.jaac.2013.02.017 PubMedGoogle ScholarCrossref
    16.
    Farley  MA, McMahon  WM, Fombonne  E,  et al.  Twenty-year outcome for individuals with autism and average or near-average cognitive abilities.   Autism Res. 2009;2(2):109-118. doi:10.1002/aur.69 PubMedGoogle ScholarCrossref
    17.
    Weaver  LL. Effectiveness of work, activities of daily living, education, and sleep interventions for people with autism spectrum disorder: a systematic review. Am J Occup Ther. 2015;69(5):6905180020. doi:10.5014/ajot.2015.017962
    18.
    Hirvikoski  T, Boman  M, Chen  Q,  et al.  Individual risk and familial liability for suicide attempt and suicide in autism: a population-based study.   Psychol Med. 2020;50(9):1463-1474. doi:10.1017/S0033291719001405 PubMedGoogle ScholarCrossref
    19.
    Schendel  DE, Thorsteinsson  E.  Cumulative incidence of autism into adulthood for birth cohorts in Denmark, 1980-2012.   JAMA. 2018;320(17):1811-1813. doi:10.1001/jama.2018.11328 PubMedGoogle ScholarCrossref
    20.
    Magiati  I, Tay  XW, Howlin  P.  Cognitive, language, social and behavioural outcomes in adults with autism spectrum disorders: a systematic review of longitudinal follow-up studies in adulthood.   Clin Psychol Rev. 2014;34(1):73-86. doi:10.1016/j.cpr.2013.11.002 PubMedGoogle ScholarCrossref
    21.
    Petersson  F, Baadsgaard  M, Thygesen  LC.  Danish registers on personal labour market affiliation.   Scand J Public Health. 2011;39(7)(suppl):95-98. doi:10.1177/1403494811408483 PubMedGoogle ScholarCrossref
    22.
    Lauritsen  MB, Jorgensen  M, Madsen  KM,  et al.  Validity of childhood autism in the Danish Psychiatric Central Register: findings from a cohort sample born 1990-1999.   J Autism Dev Disord. 2010;40(2):139-148. doi:10.1007/s10803-009-0818-0 PubMedGoogle ScholarCrossref
    23.
    Rothman  KJ.  Six persistent research misconceptions.   J Gen Intern Med. 2014;29(7):1060-1064. doi:10.1007/s11606-013-2755-z PubMedGoogle ScholarCrossref
    ×