It is difficult to see a week peacefully pass by without headlines pointing out how terribly wrong our diet or eating patterns have been so far. This week, stern.de joined the clickbait race by asking whether eating breakfast is the new smoking. Have we been all misled for decades? Fooled by froot loops? The post is based on a new book titled “Breakfast is a Dangerous Meal” by Terence Kealey. In his book, Kealey suggests that eating breakfast has a devastating effect on insulin sensitivity in the long run. My first thought was: If this is true, how could I miss the emerging evidence while doing research on insulin sensitivity and diabetes?
This section about the book on the publisher’s website tells us what is his novel take on skipping breakfast:
Breakfast may be the most important meal of the day, but only if we skip it. Since Victorian times, we have been told to breakfast like kings and dine like paupers. In the wake of his own type 2 diabetes diagnosis, Professor Terence Kealey was given the same advice. He soon noticed that his glucose levels were unusually high after eating first thing in the morning. But if he continued to fast until lunchtime they fell to a normal level. Professor Kealey began to question how much evidence there was to support the advice he’d been given, and whether there might be an advantage for some to not eating breakfast after all.
Well, yes, for some, it might be an advantage to not eat breakfast and this is perhaps true for Professor Kealey. At the same time, for most of us this is unlikely to be true. To the best of my knowledge, many large cohort studies have in fact shown the opposite: skipping breakfast increases the risk for diabetes (Mekary et al., 2013; Mekary, Giovannucci, Willett, van Dam, & Hu, 2012; Odegaard et al., 2013) and is associated with higher BMI (Cho, Dietrich, Brown, Clark, & Block, 2003) compared to controls. But Kealey has an explanation for the discrepancy in the literature as well. Many of the studies suggesting that breakfast improves metabolic health are funded by the cereal industry, which certainly has an interest in proclaiming breakfast as the most important meal of the day. Indeed, two authors of the Odegaard et al. paper report a potential conflict of interest due to their association with the Bell Institute of Health and Nutrition, which is financed by General Mills. No such conflict of interest were reported in the other representative papers I cited though.
Do such dangerous ties between the industry and researchers perhaps invalidate a whole body of the current literature? That’s a tough question that highlights a fundamental problem in nutrition research. Most studies are only observational. It is very challenging to do controlled trials that enroll thousands of participants, in particular without industry funding backing it. In observational studies, there are many opportunities to fiddle with the data to obtain a desired effect by p-hacking your way forward. Direct replication is often not feasible leaving “conceptual” replications that, if they fail, can often be “reconciled” by some subtle differences in study cohorts or design. The list could go on and on.
Updating beliefs based on incidental evidence?
Good research in nutrition is tough and cumbersome. But that complexity is often not reflected in the media coverage of new studies. Sadly, many readers are aware of the limitations. Nevertheless, I have heard many times how friends and family who follow these science news sense the disagreement of the “experts” on what is a good diet, for example to effectively lose weight. Some of them go with the flow and try the promised superfoods for a while before the next big thing comes along in the food trend aisle. Others get resistant to any kind of new evidence. If experts are uncertain, why should I change my ways?
Both strategies appear to be perfectly rationale from a Bayesian perspective. If you use a wide prior on your current diet, a single study reported in the media might convince you to try something new. Over time, you will probably realize that one superfood is barely enough to make a difference, creating the right state of mind for the next trend to sub in. On the other hand, a sharp “old ways are the best” prior will hardly be influenced by weak and fluctuating evidence coming in. The true risk I see is that both strategies will likely fail to pick up the valid and consensual evidence in the field. And yes, there is some converging evidence.
What we know for certain (almost)
After complaining about the stern.de headline referring to the book, I want to add something positive about the coverage. In the last section, it is acknowledged that Kealey’s theory is controversial. The nutritionist Dagmar von Cramm is cited speaking words of wisdom: It is probably more about what you have for breakfast than if you have breakfast at all.
If you check the nutrition facts labels in the cereal aisle, you will see that most of the common breakfast items should be classified as candy instead. There is growing evidence suggesting that sugar sweetened beverages increase the risk of developing diabetes. The evidence for carbohydrate intake in general as a risk factor is more heterogenous for many reasons, but points to the same direction. I think this is literally the grain of truth in the cereal killer story. Skipping breakfast may help individuals who make very unhealthy choices for breakfast, but find it difficult to eat something better instead. I doubt though that a down-to-earth statement like that would catch any attention in the media. This is why we can count on the next great diet advice to save our poor souls. While I was typing this post, I heard the low carb story in the radio again as if it was something brand new. Sigh.
Ironically, whereas many people associate the health risk of smoking primarily to a heightened incidence of cancer, smoking is a well-established risk factor for the development of diabetes. Breakfast, as we know it today, probably dates back to the rise of agriculture almost 10,000 years ago and I don’t think it is likely that we’ll see fundamental changes in the next decades. If Kealey’s book manages to change public opinion based on very limited evidence in favor of his theory, I will be happily joining the remaining smokers in front of the building with my bowl of ostracized cereals.
Cho, S., Dietrich, M., Brown, C. J. P., Clark, C. A., & Block, G. (2003). The Effect of Breakfast Type on Total Daily Energy Intake and Body Mass Index: Results from the Third National Health and Nutrition Examination Survey (NHANES III). Journal of the American College of Nutrition, 22(4), 296-302. doi:10.1080/07315724.2003.10719307
Mekary, R. A., Giovannucci, E., Cahill, L., Willett, W. C., van Dam, R. M., & Hu, F. B. (2013). Eating patterns and type 2 diabetes risk in older women: breakfast consumption and eating frequency. Am J Clin Nutr, 98(2), 436-443. doi:10.3945/ajcn.112.057521
Mekary, R. A., Giovannucci, E., Willett, W. C., van Dam, R. M., & Hu, F. B. (2012). Eating patterns and type 2 diabetes risk in men: breakfast omission, eating frequency, and snacking. Am J Clin Nutr, 95(5), 1182-1189. doi:10.3945/ajcn.111.028209
Odegaard, A. O., Jacobs, D. R., Jr., Steffen, L. M., Van Horn, L., Ludwig, D. S., & Pereira, M. A. (2013). Breakfast frequency and development of metabolic risk. Diabetes Care, 36(10), 3100-3106. doi:10.2337/dc13-0316