Abstract
BACKGROUND: Modifiable risk factors play a central role in the development and course of neurodegenerative disorders of later life, including dementias. Although past research has focused on independent associations of modifiable risk factors, including cardiovascular disease risk factors using Framingham cardiovascular risk score, physical activity, dietary quality, body mass index, and sleep, on neurodegeneration, the impact of all 5 factors simultaneously in a multimodal model has not been studied. We examined independent associations and an overall combined model with 5 modifiable risk factors with white matter injury, a recognized risk factor for dementia, ≈10 years later in a diverse Hispanic/Latino population. METHODS: Participants were from the HCHS/SOL (Hispanic Community Health Study/Study of Latinos) Investigation of Nerocognitive Aging-Magnetic Resonance Imaging longitudinal study (n=2667; clinical visit 1 mean age, 52.01 [8.90] years). We conducted path and mediation analyses across 5 modifiable risk factors obtained at clinical visit 1 (2008-2011) and 2 measures of white matter injury (free water and white matter hyperintensity volume) obtained at the magnetic resonance imaging visit (2018-2022; mean age, 62.37 [9.23] years). We controlled for age at the time of the dependent variable, sex, education, Hispanic/Latino heritage, overall cognitive status, and b-value when free water was included. RESULTS: We observed 11 significant independent associations across modifiable risk factors and white matter injury measures. The association of Framingham cardiovascular risk score to white matter hyperintensity volume was mediated by free water (indirect mediation: β=2.473; SE=0.207; P<0.001), and the association of physical activity to body mass index was mediated by sleep duration (indirect mediation: β=-0.038; SE=0.019; P=0.050). We combined the 2 mediational pathways and added diet associated with white matter hyperintensity volume (β=0.008; SE=0.007; P=0.004) and physical activity associated with Framingham cardiovascular risk score (β=-0.007; SE=0.002; P=0.001) in the overall model. CONCLUSIONS: Our findings identified how modifiable risk factors synergistically influence future white matter injury in Hispanic/Latino populations. Such multimodal models may lead to the development of novel and personalized lifestyle clinical interventions for adults at risk for dementia due to elevated white matter hyperintensity volume.