Objectives: Clusters of low fitness and high obesity in childhood are associated with poorer health outcomes in later life, however their relationship with cognition is unknown. Identifying such profiles may inform strategies to reduce risk of cognitive decline. This study examined whether specific profiles of childhood fitness and obesity were associated with midlife cognition.

Design: Prospective study.

Methods: In 1985, participants aged 7–15 years from the Australian Childhood Determinants of Adult Health study were assessed for fitness (cardiorespiratory, muscular power, muscular endurance) and anthropometry (waist-to-hip ratio). Participants were followed up between 2017 and 2019 (aged 39–50). Composites of psychomotor speed-attention, learning-working memory and global cognition were assessed using CogState computerised battery. Latent profile analysis was used to derive mutually exclusive profiles based on fitness and anthropometry. Linear regression analyses examined associations between childhood profile membership and midlife cognition adjusting for age, sex and education level.

Results: 1244 participants were included [age: 44.4 ± 2.6 (mean ± SD) years, 53% female]. Compared to those with the highest levels of fitness and lowest waist-to-hip ratio, three different profiles characterised by combinations of poorer cardiorespiratory fitness, muscular endurance and power were associated with lower midlife psychomotor-attention [up to −1.09 (−1.92, −0.26) SD], and lower global cognition [up to −0.71 (−1.41, −0.01) SD]. No associations were detected with learning-working memory.

Conclusions: Strategies that improve low fitness and decrease obesity levels in childhood could contribute to improvements in cognitive performance in midlife.