Breakfast is often considered the most important meal of the day, and at first glance, science appears to back this up. Multiple studies have even found that missing breakfast is associated with an increased risk of all kinds of serious diseases, from diabetes to cancer, leading to an increased probability of an early death. One study published earlier this year suggested that those who rarely consumed breakfast were 60% more likely to die from any cause during the study period than those who ate breakfast every day. To put that into perspective, the increased risk associated with smoking 10 cigarettes per day is about 70%. Now if that raises your eyebrows, you’re not the only one. Let’s take a closer look at this study and talk about why we can’t always take these numbers at face value.
The researchers looked at data from 7007 adults, aged 40 or older, who participated in the United States’ National Health and Nutrition Examination Survey between 1988–1994. Specifically, they looked at how respondents answered the question ‘how often do you eat breakfast?’ They then checked the death certificate records in 2015 and compared the death rates (both from all causes and from cancer specifically) between those who said they always ate breakfast to those who said they ate breakfast only sometimes or not at all. They then attempted to control for 14 factors that could confound the results, such as differences in average age, diet quality and physical activity between the two groups.
After controlling for confounding factors, the data showed that people who consumed breakfast some days or rarely were 52% more likely to have died of cancer and 69% more likely to have died from any cause when compared to those who ate breakfast every day. The authors of the study conclude ‘This study provides evidence for the benefits of regular breakfast consumption in reducing the risk of all-cause and cancer mortality.’
This study, just like others before it, shows very clearly that people who skip breakfast are more likely to die. That’s not the same as showing that skipping breakfast makes you more likely to die – in other words, these studies show correlation but not causation. In order to provide evidence of causation, you need a randomised, placebo-controlled trial. In practice, it is not possible to make a trial of breakfast consumption placebo controlled – you can’t really trick someone into thinking they ate breakfast. Having a randomised trial is more attainable. In a randomised trial, participants would be randomly assigned to either a breakfast-eating or breakfast -skipping group. This is a good thing, because it separates the variable of eating breakfast from any other confounding variables. For example, it doesn’t matter if smokers are less likely to eat breakfast, because in a randomised trial, you would expect to have the same number of smokers in the breakfast-eating group as in the breakfast-skipping group.
Unfortunately, randomised trials are more costly, harder to organise, and harder to recruit for than studies that simply observe the data that is already there – especially when you want to run the study for a long time (would you sign up for a trial where you might be randomly asked to skip breakfast for several years?) Consequently, most studies on breakfast consumption are not randomised. Why is this a problem? Because now you are comparing two groups that are not identical:
You can see from the above table from Peter Attia’s article that breakfast eaters in this study were older on average, less likely to be male, less likely to smoke, and less likely to be overweight, among other differences. In order to isolate the effects of breakfast consumption from these other factors, researchers take the death rates they actually measured, and use statistical techniques to estimate what the death rates would have been had the two groups been identical.
You can see from the graph above that in the crude data (the unadjusted values) on the left, the people who consumed breakfast every day were actually more likely to die compared to those who did not, probably in large part due to the greater average age among breakfast-eaters (age being the most important risk factor for cancer). Only after adjustment (model 4, on the right) is the risk higher among breakfast-skippers. These adjustments are made based on our understanding of the risks associated with the confounding factors mentioned in the table, which are themselves imperfect (it may come as no surprise that there are no randomised, placebo-controlled trials of diabetes). There are also any number of factors that were not controlled for. What if some people don’t eat breakfast because they work night shifts, sleep late or are hungover? The point is that it is impossible to perfectly control for all possible confounders in a study that is not randomised.
I wouldn’t go that far – studies like this are suggestive of a potential effect, and are worthy of our attention, but not our alarm. They are not designed to show a causal effect, and the authors admit as much, saying: ”the seemingly reduced all-cause and cancer mortality risk among persons who consume breakfast every day may simply be the reflection of habitual breakfast consumption being a proxy for a health-conscious lifestyle”. I, for one, would sooner skip breakfast for a week than smoke a single cigarette, let alone 10 a day…
The association of skipping breakfast with cancer-related and all-cause mortality in a national cohort of United States adults: https://doi.org/10.1007/s10552-021-01401-9
Does skipping breakfast increase the risk of an early death? Part I: https://peterattiamd.com/does-skipping-breakfast-increase-the-risk-of-an-early-death-part-i/
Cigarette Smoking and Mortality Risk Twenty-five–Year Follow-up of the Seven Countries Study: doi:10.1001/archinte.159.7.733