Risk of Alzheimer’s disease nearly doubles in older adults with Covid-19! Like many of you, several weeks ago we have also been alerted by such headlines seen across multiple lay and semi-professional media (e.g., LINK).
We have retrieved the original paper (LINK) that claimed to provide evidence and that indeed made a clear statement in the Abstract: “… people with COVID-19 were at significantly increased risk for new diagnosis of Alzheimer’s disease within 360 days after the initial COVID-19 diagnosis (hazard ratio or HR:1.69, 95% CI: 1.53–1.72) …”.
We felt, however, much relieved when reading the article itself. The overall risk for new diagnosis of Alzheimer’s disease (over 360 days) was increased from 0.47% in the non-covid cohort to just 0.68% post-covid. Sounds less dramatic than the abstract or the headlines, doesn’t it?
So, we wanted first to discuss the Wang et al paper from the perspective of how scientific findings are sometimes misinterpreted when judged solely based on what is included in abstracts that are often written with an intent to inflate the “significance” of the findings.
But then, on a second look, another sentence in the Abstract triggered more thoughts: “Our findings call for research to understand the underlying mechanisms … of long-term impacts of COVID-19 on Alzheimer’s disease”.
Are the findings in a paper by Wang et al. robust enough to trigger this new line of research? Wang et al. acknowledge some generic limitations of their study such as “potential biases introduced by the observational and retrospective nature of this study”. It is indeed worth discussing numerous factors that could have introduced a greater risk of Alzheimer’s disease diagnosis in the post-covid cohort. The two cohorts compared in the study were those that had and had no documented COVID-19 infection but both had medical encounters with healthcare organizations between 2/2/2020–5/30/2021. Could the medical encounters triggered by COVID-19 have lasted longer and thereby increased the chances for having the Alzheimer’s disease diagnosed? Were these encounters more likely to be associated with a hospital visit or stay?
Are the findings in a paper by Wang et al. meaningful enough to trigger a new line of research? Over the past months, multiple publications reported on COVID-19 infection increasing the risk of atrial fibrillation (LINK), other cardiovascular outcomes (LINK), schizophrenia (LINK) and diabetes (LINK), etc. It is important to know whether analyses such as the one in the Wang et al paper follow a pre-specified hypothesis. Have Wang et al looked specifically at the risk of Alzheimer’s disease? Or, have they look at multiple outcomes and report only what came out as “statistically significant”?
Communication of scientific findings with an extra bit of positive spin can mislead general public with no major consequences. However, it is a completely different situation when the misled individuals or institutions start to spend time and resources, especially if the latter are ethically sensitive (Note: a quick search NIH RePORTER reveals a fair number of funded projects on mice infected with SARS-CoV-2).