Abstract

Background/aims: Particularly among the oldest old, there is restricted knowledge regarding the factors associated with mild cognitive impairment and dementia using data from large, nationally representative samples. Thus, our aim was to address this knowledge gap.

Methods/design: We used data from the nationally representative "Old Age in Germany (D80+)" study covering community-dwelling and institutionalized individuals in the entire country (n = 2,555). Mean age was 85.5 years (SD: 4.2), ranging from 80 to 100 years (61.7% of the participants were female). The DemTect was used to measure cognitive impairment in terms of probable mild cognitive impairment and probable dementia. Sociodemographic, lifestyle-related, psychosocial and health-related independent variables were included in the multinomial regression analysis.

Results: In the analytic sample, 57.8% of the individuals did not have cognitive impairment, 24.2% of the individuals had mild cognitive impairment and 18.0% had probable dementia. Regression analysis identified some sociodemographic (e.g., advanced age, being male, lower education), lifestyle-related (lower cognitive activities), psychosocial (higher loneliness and absence of meaning in life), as well as health-related (e.g., functional impairment) factors associated with probable MCI and probable dementia. Loneliness was only associated with these outcomes among women, but not men.

Discussions: Based on large, nationally representative data, this study revealed several factors associated with probable MCI and dementia - which enhances our current understanding mainly based on small or selective samples.

Conclusion: Such knowledge may help to address those at risk for cognitive impairment. Longitudinal studies are required to gain further insights.