A recent study by Cerezo et al. has provided valuable insights into the use of the CAIDE (Cardiovascular Risk Factors, Aging, and Incidence of Dementia) score for stratifying dementia risk in hypertensive patients. The findings suggest a potential avenue for early intervention, but several statistical and methodological aspects require further refinement.
The study analyzed a sample of 1279 hypertensive patients from Argentina and found an association between the CAIDE score and current cognitive status. However, to fully understand the generalizability of the results, a comparison with other demographic groups, particularly in terms of age distribution and educational levels, would be beneficial.
Incorporating advanced statistical techniques, such as non-parametric tests or generalized additive models (GAMs), could strengthen the analysis by capturing non-linear patterns. A longitudinal analysis would also allow for the observation of changes in cognitive function over time, potentially revealing early warning signs of dementia.
A sensitivity analysis around the cut-off point of 9 points for the CAIDE score would be valuable to assess its robustness and potential impact on sensitivity and specificity. Employing methods such as Bonferroni correction or False Discovery Rate (FDR) control could help adjust for multiple comparisons.
The study also identified an inverse association between the CAIDE score and cognitive test results, highlighting the need for a mediation analysis to explore potential mediators that might explain this relationship. Including covariates such as genetic predispositions, lifestyle factors, and access to healthcare could provide a more comprehensive understanding of dementia risk in hypertensive patients.
Overall, while the study by Cerezo et al. is a significant step towards understanding the predictive value of the CAIDE score, further research is needed to refine the predictive model and contribute to the prevention of dementia.
Source: https://www.nature.com/articles/s41440-025-02117-y