The Association between Autism Spectrum Traits and Age-Related Spatial Working Memory Decline: A Large-Scale Longitudinal Study.

Background And Objectives: Based on mixed findings from previous research, researchers have hypothesised autism may be a protective or risk factor for age-related cognitive decline/dementia, or that autism does not influence it (parallel ageing). To differentiate between hypotheses, longitudinal studies that account for autism underdiagnosis, are needed and lacking. This study examined if higher autistic traits in adults aged 50+ are associated with a greater risk of spatial working memory (SWM) decline, a key cognitive domain affected in both healthy aging and autism.

Research Design And Methods: Participants from the online PROTECT cohort (n = 13,390) were classified into three groups based on autism spectrum traits (AST): high (H-AST, n = 205), intermediate (I-AST, n = 589), and no traits (COA, n = 12,451). SWM performance was captured annually across 7 years. Growth mixture models (GMM) and latent growth curve models (LGCMs) were estimated to examine the relationship between AST and SWM.

Results: GMMs revealed an optimal 1-class quadratic solution, consistent across groups. There were no significant differences between AST groups in baseline SWM (p = 0.837). AST were not associated with SWM at baseline (B = 0.01, SE = 0.05, p = 0.901) or SWM trajectory (B = 0.00, SE = 0.01, p = 0.856), regardless of accounting for covariates.

Discussion And Implications: Findings suggest a single SWM trajectory in middle-aged/older adults with higher autistic traits and no autistic traits. Future research should address if these broader autism phenotype results are replicated in diagnosed autism groups.