Recent headlines would have us believe that catching Covid makes you less intelligent, with even mild infections knocking several points off your IQ. All these headlines refer to a study by Adam Hampshire and colleagues published in the New England Journal of Medicine.
The study was based on the Real-Time Assessment of Community Transmission (REACT) cohort in England – a sample of 2.4 million people who were randomly selected to take a Covid test during the pandemic. Of the original participants, about 110,000 subsequently completed an online questionnaire and cognitive test. And of these, about 10,000 people were excluded because their symptoms began less than 12 weeks before taking the cognitive test.
The authors therefore had data on cognitive ability and Covid infection for about 100,000 people. As for Covid infection, they knew both the date of infection and its severity, measured by the number and duration of various symptoms. They also had data on people’s demographic characteristics and pre-existing health conditions.
Their main findings are shown below. The top three charts plot the distribution of cognitive test scores by variant, illness duration and hospitalisation. All three comparisons are adjusted for demographic characteristics and pre-existing health conditions. The bottom three plot average differences based on a model that estimates the effect of each variable controlling for the other two, along with demographic characteristics and pre-existing health conditions.
The two blue charts on the left show that people who caught the original virus or the Alpha variant scored lower than those who caught the Delta or Omicron variants or who didn’t catch Covid during the study period. The red charts in the centre show that people whose symptoms persisted scored lower than those whose symptoms were short-lived. And the green charts on the right show that people who were hospitalised scored lower than those who weren’t.
The first thing to say is that Hampshire and colleagues didn’t actually look at change in test scores, which means that headlines referring to “drops in IQ” or “falls in IQ” are highly misleading. Rather, they looked at the association between test scores and illness duration/severity across individuals. As the authors clearly state, “we could not assess cognitive change” and “we could not infer causality”.
However, it’s true that one possible explanation for their findings is that Covid infection has a causal effect on cognitive ability – reducing it by 42% of a standard deviation (6 IQ points) among those with symptoms lasting more than 12 weeks. How likely is this? Before getting to an alternative explanation, I will highlight several things in the paper that caught my eye.
First, if you go to Table S6 in the Supplementary Appendix, you find the exact distributions of cognitive scores corresponding to the top three charts in the image above. These confirm that individuals who caught the Omicron variant, individuals who had asymptomatic infections, and individuals who required no medical treatment, had a marginally better (i.e., more right-shifted) distribution of scores than those who did not catch Covid during the study period.
Second, the authors estimated the effects of variant, illness duration and hospitalisation on a cognitive score variable that was standardised after adjustment for demographic characteristics and pre-existing health conditions. Depending on the interrelationships among the variables, this may have resulted in larger effects of variant, illness duration and hospitalisation than would have been observed if standardisation had been done before adjustment.
Third, if you go to Table S7 in the Supplementary Appendix, you find the output from a multivariate model of cognitive test scores that includes a predictor for number of vaccine doses. The output seems to imply that vaccination had a statistically significant negative effect on cognitive ability. This is contrary to a Swedish study which found a strong positive association between vaccination and cognitive ability. I say “seems to imply” because the output is rather unclear. It’s possible that I am misinterpreting how the vaccine doses variable was coded. (The authors did not respond to my request for clarification.)
But let’s take the main findings at face value. Could they be explained by something other than Covid infection having a causal effect on cognitive ability? Yes: self-selection.
We know that people with higher cognitive ability tend to work in professions that had lower risk of exposure to Covid. Hence they may have been less likely to get infected during the early months of the pandemic and more likely to catch the milder Omicron variant. We also know that people with higher cognitive ability tend to be healthier. So they may have been less likely to become seriously ill when they did catch Covid.
In other words, rather than early and more-serious infections causing lower cognitive ability, it may be that higher cognitive ability caused people to get later and less-serious infections. This explanation is consistent with the first of the three observations I made above.
Overall, the Hampshire and colleagues’ findings are interesting and they clearly warrant further scrutiny. But I’m not yet convinced that Covid infection has a large negative impact on cognitive ability. Having said that, I could certainly believe that an infection bad enough to send someone to the ICU might knock a few points off their IQ – at least temporarily.
To join in with the discussion please make a donation to The Daily Sceptic.
Profanity and abuse will be removed and may lead to a permanent ban.