Investigating the association between cancer and dementia risk: a longitudinal cohort study

Background Previous studies found that cancer survivors had a reduced risk of dementia compared with the general population. However, these findings were uncertain because of survivor bias and a lack of stratification by cancer types. This current cohort study used data from the UK Biobank to explore these associations. Methods Multivariable Cox regression analyses were used to examine the association of cancer status and the risk of dementia with its subtypes after adjusting for age and sex. Hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated as a measure of relative risk by comparing observed dementia incidence among cancer patients. Results We included 263,151 participants in the observational analysis. During a median follow-up of 9.18 years, dementia was diagnosed in 472 individuals with cancer and 3685 individuals without cancer, respectively. Cancer patients had lower risks of dementia (hazard ratio: 0.89, confidence interval: 0.81-0.98) and its subtypes (Alzheimer's disease [AD]: 0.85 [0.74-0.98]; vascular dementia [VD]: 0.81 [0.66-0.99]) in the Cox regression adjusted for age and sex. Individuals with cancers in the male genital system had substantially reduced risks of dementia (0.66 [0.46-0.93]) and AD (0.53 [0.29-0.97]) than those with cancers in other systems. Moreover, non-melanoma skin cancer and prostate cancer were associated with a reduced risk of dementia (0.79 [0.62-0.99]; 0.69 [0.49-0.97]), but not with AD or VD (P>0.05). Conclusions The current study supported a negative association between cancer and dementia risk, and encourages further exploration of the mechanistic basis of this inverse relationship to improve understanding.