You may have seen the headlines from a new study published in the Canadian Medical Association Journal. The study suggested that household cleaners could be contributing to obesity by disrupting the gut bacteria that make up your microbiome.
In reality though, the paper didn’t quite make that claim. What the study actually did was send questionnaires to new mothers and then measured the bacteria in their baby’s stool when they were 3 months of age. They then checked in to see if their babies were overweight or obese at 3 years of age.
The first thing to realize is that given that your microbiome is constantly changing and being influenced by the things you eat as well as host of other factors, is it really valid to use the microbiome at age 3 months to infer something about obesity at age 3 years. Given that the microbiome does change over time, using a single measurement and one far distant from the outcome of interest, may not be entirely valid.
Another thing to keep in mind is that in non-randomized trials, the people themselves choose what they do rather than researchers telling them what to do. Therefore, we must ask, are people who clean their house more frequently different in some way from the people who clean less frequently? It is in fact easy to imagine that they probably are. In this study, a number of variables seemed to differ between those who had more exposure to cleaning products. Whether the mother got antibiotics during delivery, whether the mother had a c-section, whether the mother herself was overweight prior to the pregnancy, where the family lived, whether anyone in the household smoked, and whether the baby was breastfed all differed between groups. It is not hard to see how these variables could have affected the results. For example, breastfeeding is known to change the composition of a baby’s microbiome (whether these changes are clinically significant or not is another question for another day). So one could reasonably ask, is the cleaning products or the differences in breastfeeding that are driving the results in this study. There are ways to try to adjust for these differences using statistical analysis, but these attempts are never perfect and we should always acknowledge that non-randomized trials have the potential to have alternative explanations for their findings.
Finally, we should remember that this study relied to questionnaires filled out by the parents. Questionnaires are not particularly reliable mechanisms for measuring patient behavior given that your cleaning practices probably vary quite a bit over time. Directly measuring the exposure to cleaning products would have been better, although one suspects fewer parents would have volunteered for the study if the requirements were that researchers had to come and look through their bathroom.
In the end, this is not a bad study but it is too preliminary to be too excited about, and there are some pretty compelling alternative theories that explain the findings. As with most things, the simplest explanation is generally the most likely one.
Want to comment on this post? Here it is on our Facebook page!