Autism incidence and spatial analysis in more than 7 million pupils in English schools: a retrospective, longitudinal, school registry study
Bibliographic record
Abstract
BACKGROUND: Understanding how certain factors affect autism incidence can help to identify inequities in diagnostic access. We aimed to investigate the incidence of autism in England as a function of geography and sociodemographics, examining spatial distribution across health service boundaries. METHODS: In this retrospective, longitudinal, school registry study, we sourced data for the years 2014-17 from the summer school census, which is a component of the National Pupil Database, a government registry of pupils under state education in England. Our main outcome was the incidence of autism in the English state-funded education system, defined by the amount of new autism-specific Education, Health and Care Plans or autism-specific special education needs and disability support recorded during each summer school census year since the 2014 baseline. After excluding prevalent cases in 2014, we calculated unadjusted incidence and age-adjusted, sex-adjusted incidence per 100 000 person-years per subsequent school year and by various sociodemographic categories and local authority districts. We report spatial effects using local indicators of spatial association. We used a three-level mixed-effects logistic regression model with two random intercepts (lower-layer super output area [a geographical area in England containing 1000-3000 residents] and pupil identifier) to calculate odds ratios (ORs) for autism incidence, adjusting for age, sex, ethnicity, claimed eligibility for free school meals, ethnic density quintile, Index of Multiple Deprivation quintile, first language spoken at home, and year, with our reference category being White girls without claimed eligibility for free school meals who speak English as their first language. FINDINGS: Between 2014 and 2017, our total sample included 31 580 512 person-years and 102 338 newly diagnosed autistic pupils, corresponding to an unadjusted annual autism incidence of 429·1 cases per 100 000 person-years (95% CI 426·4-431·7) and an age-adjusted, sex-adjusted annual incidence of 426·9 cases per 100 000 person-years (423·5-430·4). The adjusted incidence of autism was slightly higher in 2014-15 than in 2015-16 or 2016-17, and, of the age groups, pupils aged 1-3 years, 4-6 years, and 10-12 years had the highest incidence of autism. Adjusted autism incidence in boys was 3·9-times the incidence in girls (668·6 cases per 100 000 person-years [95% CI 662·5-674·6] vs 173·2 cases per 100 000 person-years [170·1-176·3]). Across ethnic groups, adjusted incidence was highest in pupils who had an unclassified ethnicity (599·4 cases per 100 000 person-years [574·5-624·3]) or were Black (466·9 cases per 100 000 person-years [450·8-483·0]). However, in our fully adjusted mixed-effects logistic regression model, we observed lower odds of autism among Asian (OR 0·65 [0·59-0·71]), Black (0·84 [0·77-0·92]), and Chinese (0·62 [0·42-0·92]) girls compared with White girls when these groups had not claimed free school meals and spoke English as a first language. Boys from all ethnicities irrespective of first language spoken and free school meals status had increased odds of autism compared with White girls with no claimed eligibility for free school meals who spoke English as their first language. We also found that claimed free school meal eligibility, first language spoken, sex, and ethnicity differentially impacted the odds of autism. Our spatial analysis showed significant spatial autocorrelation across lower-layer super output areas in England, with 2338 hotspots (high-incidence areas surrounded by other high-incidence areas). INTERPRETATION: The incidence of autism varies across sex, age, ethnicity, and geographical location. Environmental and social factors might interact with autism aetiology. Speaking a language other than English and economic hardship might increase access barriers to autism diagnostic services, autism-specific Education, Health and Care Plans, and school-level support. FUNDING: The Commonwealth Fund, the Institute for Data Valorization, the Fonds de recherche du Québec-Santé, Calcul Quebec, the Digital Research Alliance of Canada, the Wellcome Trust, the Innovative Medicines Initiative, the Autism Centre of Excellence, the Simons Foundation Autism Research Initiative, the Templeton World Charitable Fund, the Medical Research Council, the National Institute for Health and Care Research Cambridge Biomedical Research Centre, and the National Institute for Health and Care Research Applied Research Collaboration East of England-Population Evidence and Data Science.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.004 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.004 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".