Claims-based data showed markedly higher exposure to medications with anticholinergic effects among the clinical groups studied. The study team reported exposure prevalences of 92.5% in the autism+ID group, 88.6% in the ID-only group, and 80.0% in the autism-only group. Controls — defined as non-autistic, non-ID beneficiaries — had a substantially lower exposure prevalence of 61.3%. The pattern is clear in the abstract: adults with autism and/or intellectual disability were prescribed AC medications more often than matched controls, across the 40-plus age range examined. Short, stark fact: exposure was common. Longer view: the investigators positioned AC exposure as a candidate modifiable risk factor for adverse aging-related outcomes in these populations.
Data indicated stepwise associations between AC potency and three aging-related outcomes: ADRD, falls, and mortality. The authors observed that higher AC potency levels tracked with increased associations with these outcomes. Effects were not uniform across groups: the autism and ID groups showed more prominent associations relative to the control sample, according to the abstract. The paper attributes outcome ascertainment to validated claims-based algorithms for ADRD and falls, with death determined by Medicare/Medicaid indicators. The investigators used those endpoints to link AC potency exposures to clinically meaningful events — cognitive disorder diagnoses, injurious events, and death — though the abstract does not present effect estimates, confidence intervals, or absolute risk differences. The absence of those specific statistics in the abstract means readers must consult the full text for magnitude measures.
Researchers performed a retrospective cohort analysis of linked US Medicare and Medicaid claims spanning 2017–2019; the sample included beneficiaries aged 40 years and older. Administrative claims algorithms were used to identify diagnostic groups (autism, ID, ADRD, falls). Medications were coded using a comprehensive AC potency scale; dosing details, duration thresholds, and specific drug lists are not reported in the abstract. The analytic strategy employed modified Poisson regressions to estimate associations between AC exposure (and potency strata) and the three outcomes, and the models tested interactions between AC potency and diagnostic group membership. That approach fits the objective of evaluating whether AC effects differ by clinical subgroup. The abstract does not provide model covariates, adjustment sets, or detailed regression outputs — the study team reported the analytic framework but did not include those numerical results within the abstract.
The authors acknowledged the potential for indication bias given psychiatric medication prescribing in these populations. To address this, they ran sensitivity analyses that excluded persons prescribed antidepressant and/or antipsychotic medications. The abstract reports that sensitivity analyses were performed but does not supply the specific outcomes or comparative effect sizes from those analyses. The study therefore attempted to separate AC exposure effects from confounding by indication for psychotropic treatment; however, readers will need the full manuscript to evaluate how these exclusions altered the associations, or whether residual confounding remained.
This study is a retrospective, claims-based preprint that has not been peer-reviewed. Claims analyses raise well-known limitations: diagnostic misclassification, residual confounding, and incomplete clinical detail — and the abstract does not provide granularity on how those concerns were mitigated. The authors confirmed institutional review board approval via Drexel University under a HIPAA waiver (protocol information is reported). Funding sources are declared in the abstract and include National Institute of Mental Health awards and HRSA cooperative agreement support. The data availability statement notes that individual-level data cannot be shared due to Centers for Medicare and Medicaid Services data use agreement constraints. The investigators presented AC exposure as a potentially modifiable risk factor for adverse aging-related outcomes among middle-aged and older adults with autism and/or intellectual disability, but the preprint format and the reliance on administrative data mean causal inference is limited.
Several specific unanswered questions remain in the abstract and require the full report: which medications comprised the highest potency categories, how potency was operationalized over time, what covariates and confounder adjustments were included in regression models, and the numerical effect estimates for associations and sensitivity analyses. Clinical audiences should interpret the reported associations with caution: the authors themselves note this work as a foundational examination rather than definitive causal proof. The paper concludes by identifying AC exposure as a candidate modifiable factor; additional peer-reviewed analysis and replication will be needed to inform prescribing practice and guideline development.