AbstractBackgroundLifestyle health behaviors, such as physical activity, cognitive activity, healthy diet, and social engagement, can improve cognition in older adults. We aim to identify profiles of late‐life lifestyle health behaviors related to subsequent maintenance of cognition and explore sociodemographics and health characteristics as effect modifiers.MethodAnalyses were conducted using data from 715 older adult participants from the Rush Memory and Aging Project who were dementia‐free at the first concurrent assessment of self‐reported lifestyle health behaviors (physical activity, cognitive activity, dietary habits, and social engagement) and with ³2 annual cognitive tests to assess changes in cognition. Neuropsychological test scores were grouped to form summary measures of global cognition, and five domains (episodic memory, semantic memory, working memory, perceptual speed, visuospatial memory). First, each participant was assigned to a specific group based on the initial levels of lifestyle behaviors using a latent profile analysis. Then, we assessed change in cognition among groups using linear mixed‐effects models adjusted for sex, education, and age at baseline. In additional models, we examined potential effect modifiers, including motor function, vascular risk factors (hypertension, diabetes, smoking), and vascular diseases (stroke, cardiac disease, congestive heart failure, claudication).ResultThree groups were identified: high engagement in behaviors (n = 183), moderate engagement (n = 441), and low engagement (n = 91). Participants in the high group were more likely to be women, younger, and with higher educational level, compared to the low group. Compared to the high group, the low group had a faster annual rate of decline in global cognition (mean difference [MD] = ‐0.06 [‐0.08;‐0.03], p<.0001) after full adjustment. In addition, compared to the high group, the low group showed a faster rate of decline in episodic memory, semantic memory, and perceptual speed, while the moderate group had a faster rate of decline for episodic memory only. There were no significant effects modifiers by any sociodemographic, motor function, or vascular health factor.ConclusionClassification in the moderate or high engagement group was related to better maintenance of cognition over time. Future research should examine multidomain intervention strategies that promote adoption of higher engagement in health behaviors targeting at least moderate levels.