Older adults from racial and ethnic minorities carry a disproportionately high burden of chronic disease and disability, yet they are underrepresented in clinical trials. The ASPREE trial previously showed that aspirin does not improve disability-free survival in the general population, but separate estimates for minorities hinted at possible benefit. Alongside this work, the archive holds studies searching for early markers of dementia and cognitive decline: through eye movements, brain folds, and blood proteins. All of them try to catch the disease before it becomes irreversible, and all face the same problem: average effects hide differences between people.
The analysis included 1,270 ASPREE participants from the US who identified as non-Hispanic Black or Latino. All were over 70 years old (or over 65 for Black and Latino participants), and none had cardiovascular disease, dementia, or persistent disability at baseline. The primary endpoint was loss of disability-free survival, defined as the first occurrence of dementia, persistent physical disability, or death. To estimate individual treatment effects, an X-learner was used: first, survival models were trained separately in the aspirin and placebo groups, then the risk difference was modeled for each participant. Five-fold cross-validation and weighting were applied to account for censoring.
On average, aspirin reduced the risk of losing disability-free survival: hazard ratio 0.65 (95% CI 0.45–0.93), log-rank P = 0.018. However, the effect was heterogeneous. The X-learner model divided participants into three groups based on predicted benefit. In the group with the highest predicted benefit (T1), aspirin reduced the risk by 64% (HR 0.36; 95% CI 0.19–0.71; P = 0.003) and by 11.1 percentage points over 5 years (ARD −11.1; 95% CI −22.0 to −0.1; P = 0.048). In the middle group, the effect was weaker and did not reach significance, and in the group with the lowest predicted benefit (T3), there was no benefit at all (HR 1.26; 95% CI 0.70–2.27). Participants in T1 were younger, more often lived alone, smoked, had diabetes or obesity, higher blood pressure and grip strength, but lower cholesterol.
The work shows that average drug effects can mask clinically meaningful differences between people, especially in groups that are underrepresented in research. Unlike neighboring studies that search for biomarkers of dementia and cognitive decline, this one uses a different tool: not a marker of disease, but a model of individual response to a preventive intervention. This is the first test of the idea that aspirin may benefit not all older minorities, but only those with higher cardiometabolic risk. Such an approach could change the way clinical trials are analyzed: instead of looking for an average effect, look for those whom the treatment helps.
The field may move toward validating these groups in independent data and clarifying which specific characteristics determine aspirin's benefit. Tools for clinicians may emerge to help assess the individual benefit of aspirin for a given patient. It would also be interesting to test whether this approach works for other drugs and other populations.
The results could influence recommendations for preventive aspirin use in older adults from minority groups, if confirmed. They also encourage the use of individual treatment effect methods in gerontology and cardiology.
External validation of the model in independent cohorts is needed, for example, using data from other clinical trials. It is also worth checking whether the benefit of aspirin in the T1 group persists with longer follow-up.
The work addresses the unresolved problem of heterogeneity of treatment effects in geriatrics: how to determine which older patients will benefit from a preventive intervention and which will be harmed. It also touches on the issue of representativeness in clinical trials and the need to account for racial and ethnic differences in treatment response.
In the group with the highest predicted benefit from aspirin, participants were more likely to live alone but had stronger grip strength. Grip strength is considered a marker of overall health and survival in older adults.
Not published on arXiv. Here are its page, the PDF and the formalization: you can check it yourself, not only read our account.