* A study examined the association between weight change and a number of aspects of eating, which were recorded on an app. It received much media coverage.
* The study was a population study, not an intervention trial. It involved 547 US adults – mostly women – followed for an average of 6.3 years.
* The factors of interest were how many meals were consumed a day, the approximate size of those meals (small, medium, and large) and when those meals were consumed. The timing of meals was connected to sleep patterns.
* The app also asked participants to record snacks. Snack data were then excluded in the paper and study findings. I asked the corresponding author why. The answer was “because it was not frequently reported so we think it was often missing.” i.e., the reported data were unreliable. This undermines any findings in the research.
* There were other indicators of unreliable data. Participants were included if they had even one weight and one height measurement recorded in the two years before study enrollment. The before enrollment weight measurement was taken at a clinic. Baseline and follow-up weight was self-reported. Participants were included if they used the app even once in the few months of observation post enrollment.
* The weight changes during the study were tiny. Averaged across all people, in all BMI groups, before enrollment, an average of 0.1kg was lost per year. After enrollment, an average of 0.1kg was gained per year. An increase of one daily meal was associated with (we can’t say caused) an annual average weight gain of 0.28 kg.
* Notwithstanding the data unreliability, the study claimed that the number of meals each day was associated with weight change (no surprise). It also claimed that there was no support for time-restricted eating. There were too many flaws in this study to make any such claim.
The rest of this article is available to site subscribers, who get access to all articles plus a weekly newsletter.
To continue reading, please login below or sign up for a subscription. Thank you.