BackgroundOur study aimed to investigate the pathways by which socio-demographic factors, modifiable health and lifestyle risk factors influence each other, and subsequently, lead to dementia.MethodsWe used data from the Well-being of the Singapore Elderly study, a nationally representative survey of the older adult population aged 60 years and above in Singapore. Dementia diagnosis was established using 10/66 dementia criteria. Structural equation modelling (SEM) without latent variable was applied to confirm the hypothesized model.ResultsThe results of SEM supported the hypothesized model (χ
2 = 14.999, df = 10, p = 0.132). The final model showed that those aged 75–84 years and 85 years and over (vs. 60–74 years), having no formal education, who had completed primary or secondary education (vs. completed tertiary), who were homemakers and retired (vs. paid work), and with a history of stroke were directly associated with higher odds of having dementia, while those who had more frequent contact with friends and neighbors as well as being physically active were directly associated with lower odds of having dementia diagnosis. The study also found that physical activity, more frequent contact with friends and stroke played a significant role as mediators in these relationships. The overall pathways model explained 57.7% of the variance in dementia.ConclusionOur results suggest that physical activity, social contact and stroke were potential mediators in the relationship between age, education, employment and dementia. Intervention programmes focusing on physical activity such as exercise and social contact may be useful in reducing the risk of dementia among older adults.