Early cognitive ability and dementia risk: Response to Chen and colleagues

1.

We thank Chen and colleagues for their interesting and useful comments regarding our article. ¹ They note that there were patterns in the data that we did not address. Given space limitations in the journal, we chose to keep our main focus on the issue of general cognitive ability (GCA) versus education in predicting dementia because so many articles focus on education while ignoring prior GCA.

They note that those with lower GCA may die before dementia onset, suggesting that GCA also contributes to survival. This possible survival bias suggests that our results regarding the effect of GCA may actually be relatively conservative. We also appreciate their noting our finding that physical activity was also protective against dementia in the mortality‐adjusted models (our Table S3), again suggesting possible premature death in inactive participants. This finding may warrant greater attention, although we note that the effect of GCA remained stronger than the effect of physical activity. Also, we have findings currently under review from a different sample showing that young‐adult GCA is also predictive of later‐life physical activity level.

We thank Chen et al. for their suggestion – after observing hazard ratios for low GCA and low education – that there might be an overlooked interaction. However, our formal testing has shown that there was not a significant GCA × education interaction in this sample. In this context, they also point to the importance of focusing on more than just single variables. We certainly agree, as we have shown in our own work (e.g., Franz et al. ² ), but again, the main focus of this article was the primary direction of effect between GCA and education.

Although, as indicated by Chen et al., more dementia diagnoses in those born between 1954 and 1958 compared to those born after 1969 may suggest generational differences in dementia prevalence, it is important to note that that cannot be formally tested based on our data. They also stated that follow‐up durations were comparable, but that is not the case since follow‐up was essentially from date of conscription to last International Classification of Diseases code. What we see as the likely strongest factor is that later born participants would have been considerably younger at follow‐up, which would rather substantially reduce the likelihood of dementia.

Chen et al. also note the limitations of an all‐male, military conscription sample for generalizability. We noted the limitation of the male sample in the article but also pointed out that our findings were consistent with those from the Lothian Birth Cohort and Project Talent, both non‐military, mixed‐sex samples. ³ , ⁴ We also acknowledged limitations of the diagnoses, and for that reason we looked at dementia without specifying subtype.

Chen et al. state that factors prior to GCA assessment at age 18 are important and that GCA and education likely influence each other. This, in fact, was a major point that we made in the article, particularly in the section on implications for social policy. Citing a meta‐analysis as well as our previous work, ⁵ , ⁶ we explained how education improves GCA (and probably brain development as well) up until late adolescence/early adulthood. At that point, the effect levels off, and it is then GCA that is the primary driver of later cognitive function. That was the basis for our position that early (childhood) intervention to improve cognitive and brain development may be important for reducing dementia risk. It would, of course, be ideal to have all the early childhood data, but having both early childhood and late‐life data is extremely rare.

Finally, we thank Chen et al. for their valuable observations. And again, we wish to emphasize our key conclusion that although education is often viewed as a causal factor in reducing dementia risk, that association is largely a downstream effect of prior GCA rather than education being a causal factor in its own right.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.

Supporting information

Supporting Information

Kremen WS, Ericsson M, Gatz M, et al. Early cognitive ability and dementia risk: Response to Chen and colleagues. Alzheimer's Dement. 2025;21:e70618. 10.1002/alz.70618