Aims/Hypothesis The increasing number of people with dementia and cognitive impairments makes it essential to detect and prevent modifiable risk factors of dementia. This study focuses on type 2 diabetes mellitus, especially on undiagnosed cases and their increased risk of cognitive impairment. Furthermore, the potential of physical activity and social integration to moderate the relation between diabetes and cognitive impairment is assessed. Methods We used cross-sectional data from 1299 participants of the Berlin Aging Study II (BASE-II) aged between 60 to 84 years and performed logistic regression models to analyze the association of diabetes status, physical activity, and cohabitation status with poor cognitive performance. Cognitive performance was measured with the Consortium to Establish a Registry for Alzheimer's Disease (CERAD)-Plus test battery. Results Undiagnosed diabetes (odds ratio (OR) = 2.12, p = 0.031), physical inactivity (OR = 1.43, p = 0.008) and non-cohabiting (OR = 1.58, p = 0.002) were associated with an increased likelihood of poor cognitive performance. The highest odds were observed in participants who suffered from undiagnosed or insulin-dependent diabetes and, in addition, were inactive (undiagnosed diabetes: OR = 3.44, p = 0.003; insulin-dependent diabetes: OR = 6.19, p = 0.019) or lived alone (undiagnosed diabetes: OR = 4.46, p = 0.006; insulin-dependent diabetes: OR = 6.46 p = 0.052). Physical activity and cohabiting appeared to be beneficial. Conclusions/Interpretation Physical activity and cohabitation status moderate the link between diabetes mellitus and cognitive performance. Special attention should be paid to undiagnosed and insulin-dependent diabetes cases, which have a particularly high risk of poor cognitive performance.
