Risk of hospitalization and death among autistic young people in England during the Covid-19 pandemic.

The sample was young, with a median age of 20 years, so it says little about middle-aged and older autistic adults. Deaths were rare, so mortality estimates used only birth year and gender and could not include the full adjustment variables. The model for deaths linked to a positive test was unstable, violated statistical assumptions, and had very wide uncertainty. The study did not record sex and gender separately, which may weaken matching for autistic people, who are more likely to be gender diverse. Many people from the original records were excluded because they lacked follow-up after 1 January 2020, and 81 were excluded for missing socioeconomic status data. Conditions were identified from GP codes, so people without a code may have had the condition but were counted as not having it. Vaccination status, residence type, and care-home living were not available, although these could affect hospitalization and death risk. The sample over-represented East Midlands GP practices and under-represented Greater London, South West, and West Midlands practices.

The supplied funding statement lists multiple funders, including charitable and government research funders; no author conflicts of interest are reported.

Do not conclude that autism itself independently causes higher risk of hospitalization or death from Covid-19. The all-cause hospitalization risk persisted after adjustment, but the Covid-19-specific hospitalization risk became non-significant after adjustment, and the death estimates were based on very few events.