Modifiable dementia risk factors associated with distinct subjective and objective cognitive outcomes

Abstract

AbstractBackgroundA range of potentially modifiable environmental, social, and lifestyle factors are estimated to contribute to over a third of dementia cases worldwide. Leading theories propose that these risk factors have different risk‐conferring mechanisms. For example, social isolation, hypertension, and exercise may respectively impact most on cognitive reserve, neurodegeneration, and inflammation. If true, this may support the development of more personalised preventative strategies, depending on risk profiles. Here, we aimed to address this by gathering multiple forms of cognitive assessment in a large sample, across lifespan.MethodN = 2831 users of a free smartphone app, Neureka (aged 18‐82, M(SD) = 46.84(14.7)), completed questionnaires about their background, health, and lifestyle. Additionally, they completed subjective and objective assessments of their cognitive status, via questionnaires (subjective memory problems, attitude to own ageing, and cognitive fluctuations) and cognitive tests (working memory, processing speed, cognitive flexibility, and model‐based planning), respectively. We tested the differential cognitive correlates of previously established risk factors using univariate regressions, controlling for age and gender.ResultLower education and socio‐economic status (SES) were associated with worse cognitive status across all domains (all p<.05). However, the association between SES and subjective cognition (ß = 0.38–0.51) was more than twice that of objective cognition (ß = 0.05–0.09), while this was not the case for education (ß = 0.15–0.32 for subjective and ß = 0.10–0.16 for objective cognition). Subjective measures were also more sensitive to depression, loneliness, hearing handicap, age‐related hearing loss, and tinnitus (all p<.01 for subjective and .001<p<.75 for objective cognition). In contrast, cardiovascular factors like diabetes, hypertension, stroke, and smoking were associated with comparable impacts on objective and subjective cognition. Overall, there was little evidence for risk factor correlates of different aspects of objective cognition (e.g., working memory vs. processing speed).ConclusionObjective and subjective cognition have dissociable associations with certain potentially modifiable risk factors for dementia. This may aid in our understanding of the distinct mechanisms through which they confer risk for dementia. The objective cognitive assessments gathered, model‐based planning, memory, processing speed and flexibility, yielded remarkably similar associations with risk factors, suggesting a common mechanism through which they are linked to risk factors for dementia.