Getting Personal: Personalized Nutrition
“I was not born to be forced. I will breathe after my own fashion. Let us see who is the strongest.”
― Henry David Thoreau, On the Duty of Civil Disobedience
Although historians have fairly well debunked the story of Marie-Antoinette (Queen to France’s King Louis XVI) supposedly remarking, “Qu’ils mangent de la brioche”— “Let them eat cake”—when told that the French peasants had no bread, the tale not only survives but thrives to this very day. Perhaps it is because such a cruel and uncaring remark seems to reap karmic just desserts when she was beheaded during the French Revolution on 16 October 1793. Then again, it could simply point to a more basic human truism that no one likes being told what they have to eat, whether it be elites pushing bugs or royalty recommending cakes and jams no average Joe can afford.
This basic human response doesn’t seem to shift significantly when the guidelines are proclaimed by health experts and agencies. And really, when we think about it, why should they? Standard dietary parameters are typically based on population averages, which means that the popularly promoted one-size-fits-all nutritional recommendations all too often fit no one. A fact reflected in the “variable efficacy of tightly controlled lifestyle intervention trials.”[1] In one study examining postprandial glycemic responses, one individual had an exaggerated glycemic response to a banana but not a cookie, whilst a different participant had the exact opposite reaction.[2]
The literature is rich in evidence, showing that people differ greatly in their response to various dietary interventions. There are even considerable inter-individual differences in postprandial metabolic responses to the same meals. This, prima facie alone, argues against generalized standardized dietary and nutritional recommendations. Yet, there is abundant evidence that certain dietary practices and patterns and health are intimately linked.
Adding to the difficulty is the fact that individual food choices are influenced as significantly by perceived enjoyment and satiation as they are by perceived health effects. It turns out that these choices, things like meal composition, meal timing, exercise, and sleep, along with our gut microbiota and circadian rhythm, play a more significant role than even our genetics. It may be why direct-to-consumer nutrigenomic companies that base their recommendations on DNA variance alone often have less than robust results.
Such findings raise the question: could predictive modeling based on such multiple inputs allow for a successful population-wide personalized nutrition strategy?
This week’s study spotlight examines just such a personalized dietary program (PDP) versus standard general dietary guideline recommendations (control).
- The study was a randomized clinical trial with a total of 347 participants aged 41-70 years old.
- The mean age was 52 ± 7.5 years old; 86% of the participants were female, and the mean body mass index (BMI) was 34 ± 5.8 kilograms/meter2.
- There were 177 people in the PDP group and 170 participants in the control group.
- Results were reported utilizing intention-to-treat and per-protocol analysis.
- The intervention period was 18 weeks.
The Take-Away:
- There was a significant reduction in triglycerides between the two groups.
- There was no difference in LDL cholesterol between the two groups.
- There were statistically significant reductions in body weight, waist circumference, and hemoglobin A1c in the PDP group.
- There were statistically significant improvements in diet quality and gut microbiota diversity in the PDP group.
The Caveat:
Individual variances in the response to different foods can lead to failures in a one-size-fits-all approach to dietary recommendations. Ultimately, this leads to failure when attempting to apply diet and lifestyle strategies to reduce the risk and burden of chronic disease within a given population. Individual frustration can lead to resistance to adopting and implementing any dietary guidelines at all. Indeed, less than 1% of the UK population follows all nine core dietary recommendations.[1]
This study is one of the first attempts to implement a personalized nutrition approach utilizing multiple inputs and predictive analysis compared to traditional blanket recommendations. The personalized method utilized a system developed by Zoe (https://zoe.com), and the control group utilized dietary and lifestyle advice recommended by the United States Department of Agriculture (USDA).
Interestingly, although one might assume higher adherence to a personalized nutrition approach, only 61% of the PDP group completed the 18-week protocol compared to 69% of the control group. Furthermore, adherence to the recommendations was self-reported, which introduces additional room for error and/or bias in the results. Despite these limitations, the study reported modest success with the reduction in triglycerides in the PDP group. The other primary outcome measure, LDL cholesterol concentrations, were not different between the two treatment groups at the study’s end. Secondary endpoint measures revealed significant reductions for the PDP group in body weight, waist circumference, and hemoglobin A1c. The PDP group likewise had significant increases in diet quality and gut microbiota diversity. There were no differences in the measured secondary endpoints of hip circumference, blood pressure, insulin, glucose, C-peptide, apolipoproteins A1 and B, total protein, albumin, globulin, bilirubin, alkaline phosphatase, aspartate aminotransferase, alanine aminotransferase, C-reactive protein, and tumor necrosis factor-alpha.
The PDP group also reported significant improvements in energy level, sleep quality, general mood, and reduced hunger compared to controls. However, it is important to recognize that this was not a blinded trial, and the participants were aware of the treatment arm to which they had been randomized. Additionally, although multiple inputs were used in developing the personalized recommendations (PDP group), there was no accounting for potentially important variables such as ultra-processing (UPF designation), which has been shown to significantly impact health, particularly with respect to chronic diseases. Finally, as with any measure of public health, there has to be an accounting for the return on investment. Currently, the initial cost of testing to implement the Zoe method is roughly US $375. There is also a monthly fee that can run as high as $75 per month.
Is the modest success obtained in this study worth the cost? Will the improved parameters measured actually impact morbidity and mortality associated with those chronic diseases linked to poor dietary choices? The results and conclusions in this research paper provide no answers to these important questions. But they do open the door of possibilities, and that should spur our curiosity and exploration. Because if we continue an incurious status quo, that open door might as well be a wall.
The Study:
Additional Resources:
[1] (Scheelbeek, 2020)
[1] (Berry, 2021)
[2] (Zeevi, 2015)
