Back

Faithful Diagnoses: Association of Socioeconomic Status with Diagnoses of ASD and ADHD in Girls

Mehta, A. K.

2025-05-12 epidemiology
10.1101/2025.05.11.25327397 medRxiv
Show abstract

PurposeAutism (ASD) and attention-deficit/hyperactivity disorder (ADHD) are chronic developmental disorders that have drastically increased in prevalence and co-occurrence over the past two decades. In previous research, ADHD has been correlated with low income and autism with high income, but it is uncertain how gender affects this. This analysis tests each of these income correlations, socioeconomic status (SES) confounders, and income modality against both diagnoses in girls. MethodsCross-sectional data from the 2020 National Survey of Childrens Health was analyzed. Multiple logistic regressions tested for correlations between each diagnosis and income, bimodal diagnosis patterns were tested at low and high income, and posttest analyses compared the respective odds ratios against both diagnoses. ResultsThis analysis was unable to detect any of the previously researched relationships between diagnosis in girls and income, nor a significant bimodal correlation of either diagnosis with income. Low income was linearly associated with autism diagnosis (odds ratio: 0.997, p = 0.011) while ADHD was not correlated with income. A peak in both diagnoses was observed near 138% of the federal poverty limit (FPL). Compared with White girls, only Hispanic Asian girls showed higher odds of autism in this analysis (OR: 8.2, p = 0.023). ConclusionPreviously observed trends between SES and ASD or ADHD did not appear specifically in girls. Instead, girls experience income, SES, and ethnoracial trends unique to them. A peak in diagnoses near 138% FPL suggests that expanded Medicaid eligibility may play a larger role for autism diagnosis in girls than expected.

Matching journals

The top 2 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.