Cleanliness is next to…Alzheimer’s?

Based on some recent media reports, you’d be forgiven for thinking that washing your hands was now considered a major risk to your cognitive health. The Telegraph headlined its piece “Good hygiene may be to blame for soaring Alzheimer’s”, with similar though slightly less definitive statements in the Guardian, Independent, and Daily Mail, to name a few. The headlines were based on a recent study that reported an association between hygiene levels and the incidence of Alzheimer’s disease using national-level data from almost 200 countries.

What’s the science behind the story?

Being exposed to microbes, bugs and germs helps with the development of a healthy immune system, which is then better adapted and able to cope with future exposures to the same or more virulent beasties. However, the ‘hygiene hypothesis’ posits that modern, developed countries with clean water, widespread antibiotic use, paved roads, etc. have reduced exposure to these microbes, with subsequent negative consequences for immune development in those populations.

A good overview of the hygiene hypothesis is available at NHS Choices. That overview highlights that, although cleaner environments are one of the commonly cited reasons for the increasing prevalence of allergies, there are other factors likely to play a much larger role (including diet, exercise and a family history). The current study uses the hygiene hypothesis as a foundation, and examines how it might partly account for the rates of Alzheimer’s across different countries.

What about the actual research?

The study, led by PhD Student Molly Fox at the University of Cambridge, looked at data from countries in the developed and developing world (ranging from 148 to 191 countries, depending on the availability of the specific measures being analysed). Fox and colleagues collected information on “age-standardized disability-adjusted life year rates” as a proxy for the prevalence of Alzheimer’s (though they noted the data also included other forms of dementia). The measure is said to indicate the “number of years lost due to premature mortality and years spent in disability”. They also accessed a range of survey data in order to extract proxies for exposure to microbes. Again, they noted that they did not have actual measures of microbial exposure, but were using indicators such as access to improved sanitation facilities and clean drinking water.

The researchers reported “highly significant” associations between their measures of hygiene and Alzheimer’s, always in the direction that better hygiene was associated with a higher risk. Although the results were frequently referred to as correlations (a measure of association, basically how one variable is related to the presence or absence of another), the researchers did not actually report them as correlations. Furthermore, based on the type of analyses performed, the measure of association reported should have been something called beta coefficients. These come from a type of statistical analysis called regression, which is also a way of understanding how one variable might be associated with others, but allows you to consider more than one predictor at a time. However, the only figures provided were the percentage of common variance between variables, which on its own says nothing about the direction of the association (the directionality of the associations is shown in the figures in the paper). It also isn’t clear from the description of the results if any important covariates were also included in the analyses, such as other environmental or sociodemographic factors. That latter point is much more critical to a full understanding of any associations reported.

What did the media say?

To be fair to the media reports of the study, while they described the findings much as they appeared in the main press release, they did at least include input from those not involved in the study. The Telegraph, for example, quoted Dr James Pickett of the Alzheimer’s Society: “it is always difficult to pin causality to one factor and this study does not cancel out the role of the many other lifestyle differences such as diet, education and wider health which we know can also have a role to play”. The same article also highlights the crucial problem in attributing causal claims when national-level data are used, where different countries will have different standards and clinical guidelines in terms of their reporting structures.

The bottom line.

Importantly, the results reported were at the level of comparing country to country, rather than individuals within one country or area. In doing so, there are so many potential confounders of the associations (unmeasured factors and differences across countries in disease reporting) that it cannot be definitively stated why these variables might be associated. For those interested in their health and well-being, maintaining a good level of hygiene is likely to be more beneficial than not.

Fox, M., Knapp, L. A., Andrews, P. W., & Fincher, C. L. (2013). Hygiene and the world distribution of Alzheimer’s disease: epidemiological evidence for a relationship between microbial environment and age-adjusted disease burden. Evolution, Medicine, and Public Health. doi: 10.1093/emph/eot015