Back

Clinician Article

---

Effect of Aspirin on All-Cause Mortality in the Healthy Elderly.

---

---

• McNeil JJ
• Nelson MR
• Woods RL
• Lockery JE
• Wolfe R
• Reid CM, et al.

Read abstract Read full text

---

---

---

Clinical Comments

General Internal Medicine-Primary Care(US)

The recently published ASPREE study of the effects of low-dose aspirin on disability-free survival in older adults demonstrated a borderline statistically significant increased risk for death in the aspirin group. The number-needed-to-treat to cause harm was quite high though at 143. Cancer deaths accounted for most of this increase. No particular cancer predominated. The authors rightly note that other aspirin studies have not demonstrated this result. The number of clinical endpoints that occurred was much less than the general population rate. This reduced the power of the study overall for a number of the measured endpoints. It does seem, though, that a takeaway message from the recent aspirin studies is that aspirin does not add much in primary prevention for most patients with controlled blood pressure and with good lipid control on statins who are otherwise fairly healthy.

Geriatrics

This is a thought-provking article that raises questions about the efficacy and safety of low-dose ASA in relatively healthy older persons at a time when low-dose ASA has been recommended for nearly all healthy elderly. As pointed out by the authors, the article's strength compared with previous studies is that it is based on an exclusively older population. The finding by McNeil et al that ASA appears to increase the risk for both all-cause mortality and cancer-related mortality irrespective of site appears to contradict previous meta-analyses, and may reflect a different biology in an older population that was poorly represented in previous studies. The authors acknowledge that the principal limitation is they did not statistically take into account their multiple comparisons, so the association of ASA with a small but statistically significant increased risk for cancer after ~3 years cannot be taken as definitive. Interestingly, in their supplementary analyses (Forest plot S2), being a white Australian but not being a white American was associated with a significantly increased risk for cancer-related mortality. This suggests that there may be an environment-ASA interaction that requires further exploration.

Internal Medicine

Even if one doesn`t believe the cancer-related excess mortality is causal, some of my primary care colleagues would do well to recognize that, at a minimum, there was no benefit of low-dose aspirin therapy in otherwise healthy older adults. There is the pervasive faith that something shown to be beneficial in patients with disease should also be beneficial in patients without disease - ergo the widespread belief that a baby aspirin "can`t hurt and might help" healthy adults, even not-so-old adults. Studies like this ought to help restrain that kind of thinking.

Public Health

The majority of participants were overweight or obese, which is a serious risk factor for developing cancer. Also, in different categories (diabetes, history of smoking, former cancer history), there were small differences between groups that might have influenced the result.

Public Health

This is definitely an important paper that could change a well entrenched paradigm for a hospitalist.