Hidden risk: Latent cognitive profiles and structural brain age reveal vulnerability in midlife metabolic syndrome

Objective: Metabolic syndrome (MetS) is linked to later-life cognitive decline and brain aging, but early detection of vulnerability in midlife remains challenging. This study applied two methods to detect subtle changes in midlife adults with MetS: (1) latent profile analysis (LPA) to identify cognitive performance patterns and (2) an MRI-derived brain-predicted age metric to assess structural brain aging. Method: Participants were cognitively unimpaired, community-dwelling adults from prior studies on metabolic and brain health ( N = 230; ages 40 – 65). MetS status was assigned using clinical criteria based on cardiovascular indicators and medical history. Cognitive test scores, adjusted for age, sex, and education, were analyzed using LPA, identifying four cognitive subgroups: High Memory, Low Executive, Global Average, and Low Memory. T1-weighted MRI scans were processed with brainageR to compute brain-predicted age difference (PAD). Analyses were conducted in R using chi-square tests, ANCOVA, regression, and nonparametric methods, with appropriate covariates and effect size estimates. Results: MetS prevalence differed across cognitive profiles (χ 2 = 10.99, p = .012, V = 0.22), with higher rates in the Low Memory and Global Average groups than in the High Memory group. Individuals without MetS had younger brain ages than those with MetS ( p = 0.003, η 2 = 0.03). Only elevated triglycerides were associated with greater PAD ( p = 0.012, η 2 = 0.02). A Johnson–Neyman analysis showed the MetS–PAD association was significant between ages 40.0 and 54.6. PAD did not differ by cognitive profile. Conclusions: Cognitive profiles and brain-predicted age metrics identify early vulnerability in midlife MetS, underscoring the importance of early intervention.