Investigating the joint structure and age-invariance of psychopathology and cognition in 112,712 older adults from the UK Biobank

Aims This study examined the joint structure of psychopathology and cognition in older adulthood. Methods Confirmatory factor analysis (CFA) was used to examine the latent structure of psychopathology and cognition in participants aged 55-78 years old from the UK Biobank (N = 112,712; 44.6% male; M = 65.05 years). We tested four commonly studied CFA models, including: one-factor, correlated-factors, bi-factor, and higher-order models. The best-fitting model was selected based on traditional model-fit indices, model-based reliability estimates, and evaluation of model parameters. After determining the best-fitting model, multigroup CFA was used to examine measurement invariance across four age groups (i.e., 55-59, 60-64, 65-69, and 70-78 years old). Results A joint higher-order model of psychopathology and cognition was selected as the best-fitting model, including a general higher-order factor and lower-order factors of internalizing, addictions and substance use, thought disorder, and cognitive dysfunction. This model demonstrated acceptable fit to the data (CFI = 0.936; TLI = 0.933; RMSEA = 0.042) and superior reliability of the lower-order factors (H = 0.850-0.974) compared to specific factors of the bi-factor model (H = 0.676-0.924). This model also demonstrated configural, lower-order metric/scalar, and higher-order metric/scalar invariance across age groups (change in CFI values ≤ 0.002). Conclusions Our results suggest that cognition can be incorporated into the higher-order structure of psychopathology in older adulthood and that the joint structure of psychopathology and cognition is age-invariant from 55 to 78 years old.