---

A March 2022 study claimed to have found a significantly higher rate of new-onset dementia in patients with pneumonia associated with covid than the rate seen with other cases of pneumonia. 

Professor Norman Fenton reviewed the study and exposed the elephant in the room: the study ignored the covid vaccination status of the patients.

There is another elephant in the room: the lead author of the study is a consultant for AstraZeneca.  So, Prof. Fenton has written to this author to ask some pertinent questions.

About the study claiming covid leads to 3% increase in early onset dementia

By Where Are the Numbers

Energy Drink Analogy

Imagine you are a sports scientist and have some data about 1000 young people (aged between 14 and 35) who have broken a leg in the last year. You have a theory that a key risk factor contributing to such people getting broken legs is that they regularly have energy drinks (yes – it sounds daft, but bear with us on this). The 1,000 with broken legs is your “case” group.

To test your theory you find some data on a set of young people who have NOT broken a leg in the last year. This is your “control” group. To ensure they are a good “match” to your case group – and hence to avoid the possibility that there could be missing “confounding” variables that could compromise your results – you make sure that all key physiological and demographic attributes of the control group people are similar to those of the case group. Specifically:

• The proportion of females in the control group is the same as the proportion in the case group
• The distribution of ages in the control group is the same in the control group as in the case group
• The proportion of people with pre-existing medical conditions in the control group is the same as the proportion in the case group
• The distribution of socio-economic classes in the control group is the same in the control group as in the case group
• etc.

Having done that you can now simply record the number of people in each group who report that they regularly have energy drinks.

After tallying the data, you find that a significantly higher proportion of people in the case group regularly take energy [drinks] compared to the control group. You report this and it is widely accepted as evidence that energy drinks cause broken legs.

But there was something rather important missing in this study. We know that a large proportion of broken legs in young people are caused by sporting activities (football, rugby, athletics etc). We also know that young people who do such sporting activities regularly have energy drinks. So, while it is the sporting activities that lead to the increase in broken legs and NOT the high energy drinks, it is still inevitable that you will find a higher proportion of energy drinkers among those with broken legs.

As a sports scientist, failing to consider participation in sports, was an unforgivable flaw in your study. You were missing the elephant in the room.

Flaw in the Dementia Study

But what has this got to do with the effect of covid on early onset dementia?

It’s because of the direct analogies here with a recent study claiming that the 3% of patients with pneumonia who developed early onset dementia after covid infection was significantly more than those with pneumonia who did not have covid infection. This study also managed similarly to miss an elephant in the room.

I was alerted to this study on 18 January by Mike (a German reader) who sent me this tweet by a leading German doctor about the study. The doctor asserts that we can conclude from the study that about 50,000-100,000 pneumonia patients in Germany were affected by covid induced dementia. Apparently, others in Germany have gone further and concluded that the vaccines must therefore be saving hundreds of thousands from dementia.

The study is flawed (even ignoring the usual problem that the raw data is not provided for independent review). As in the imaginary broken legs study, it uses the strange ‘case-control’ method, whereby you focus on a set of ‘cases’ being people who have a positive PCR test and then find a set of ‘matching’ control patients who did not test positive. Even though this strange method has been increasingly used in covid observational studies, I believe the results of such studies are unreliable (something we will be covering in a forthcoming article about why all the well-known observational studies claiming to show vaccine efficacy or safety are fundamentally flawed).

But the major problem with the study is the failure to notice the “elephant in the room.” While the authors considered all kinds of physiological and demographic factors, they seem to have forgotten to consider the covid vaccination status of the patients, which is quite an omission given that the study was carried out after the initial vaccine rollout.

Study Author is a Consultant for Big Pharma

And it’s even stranger given that the first named author Adnan I Qureshi also just happens to be a consultant for AstraZeneca – who you might have thought would be interested in the impact of the vaccines in all of this:

I therefore sent the following self-explanatory email to Dr Qureshi on 20 January 2023 asking simple questions that should not be difficult to answer.

Dear Dr Qureshi

I have only recently seen your very interesting March 2022 paper “New-Onset Dementia Among Survivors of Pneumonia Associated With Severe Acute Respiratory Syndrome Coronavirus 2 Infection”

https://doi.org/10.1093/ofid/ofac115 for which you are the lead and corresponding author.

I note that, while the study considered all kinds of demographic and physiological factors in the ‘case’ and matched ‘control’ participants, one crucial factor missing is their covid vaccination status. This is especially curious given that the study was carried out after the initial vaccine rollout, and hence that vaccine status is a potential key confounder for the results.

I note that you have received consultation fees from AstraZeneca who I would have thought would have been especially interested in knowing the extent to which the vaccine impacts the overall results.

I would therefore be interested to know the following:

1. Was covid vaccination status data ever part of the study and if not, why not?
2. If covid vaccination status data was part of the study, then why was it excluded from the reporting and results?
3. Is there any plan to follow up with results that include the covid vaccination status?

Yours

Norman Fenton

I have yet to receive a reply from Dr Qureshi. If and when I do I will happily update this article and make any corrections if we have misunderstood his article or data.

About the Author

Norman Fenton and Martin Neil are two academic professors who, between them, have authored hundreds of scientific papers and numerous books on statistics, decision making, risk and uncertainty systems and software engineering, and have consulted commercially to scores of commercial organisations.  Together, they publish articles on a Substack page titled ‘Where are the numbers?’. If you are interested in science and statistics in the post-Covid era, you can subscribe to and follow their Substack HERE.