Examine publishes rigorous, unbiased analysis of the latest and most important nutrition and supplementation studies each month, available to all Examine Members. Click here to learn more or log in.

Quick Navigation

Issue #51 (January 2019)

From the Editor

Volume 1

To start of the new year, U.S. News and World Report released their ninth annual list of the best diets. They created this list through a multi-step process.

First, the editors looked at several lines of evidence for 41 diets to create profiles for each diet. These profiles were then reviewed by experts in nutrition and diet-associated diseases. The experts edited these profiles and then rated each diet according to seven different categories:

  1. How easy it is to follow

  2. How well it helps with short-term weight loss

  3. How well it helps with long-term weight loss

  4. Its nutritional value

  5. Its safety

  6. How well it can prevent or manage heart disease

  7. How well it can prevent or manage diabetes

The editors then took these expert-modified profiles and converted the experts’ input into scores ranging from one (the lowest) to five (the highest) and created nine “best diet” lists by category, including the easiest diets to follow, the best heart healthy diets, and the best diet overall.

That last category concerns me a little bit, since I’m a pretty big skeptic of anything that’s claimed to be the “best”, for reasons very similar to why I’m wary when people ask “should I take this supplement?” or “does this supplement work?” My reason for caution is because I don’t think these are complete questions. To answer them well, you need to know the whys and whats. Why would you want to take this supplement in the first place? What’s your aim? Does this supplement work for what? The same question arises for the best diet — the best diet for what?

U.S. News’ method for finding the best overall diet was to combine the seven categories I mentioned above, weighting some categories more than others. Their weighting strategy makes some sense — for instance, they rate safety highly, since a diet that can harm people isn’t really worth it even if it sheds a few extra pounds. What I’m questioning here is why they have a “best overall diet” category at all. If you’re going to change your diet, you are probably doing so for a specific reason: because you’re hoping for a certain outcome, or set of outcomes. So why not optimize for those outcomes rather than just going for “the best overall”, when that’s just an average of factors, some of which you may not actually care about?

For both nutrition and life, it’s helpful to have a goal in mind ahead of time. To quote the ancient Roman Seneca the Younger (who ate a vegetarian diet, in case you’re curious): If one does not know to which port one is sailing, no wind is favorable.

Gregory Lopez, MA, PharmD
Editor-in-chief, Nutrition Examination Research Digest


Volume 2

In NERD #30, volume 2, we covered the SMILES (Supporting the Modification of lifestyle In Lowered Emotional State) trial. While the tortured trial acronym should make you frown, the results seem like something to be happy about — it found a pretty large effect.

As a refresher, the SMILES trial involved people who were moderately to severely depressed and had poor quality diets. They were assigned to two conditions: a control consisting of a “befriending” condition where the researchers chatted with the participants in a friendly manner. The treatment was a modified Mediterranean-style diet.

After twelve weeks, the diet group improved much more than the control group — the effect size of the treatment was 1.2 compared to control, which is several times larger than the typical effect size for antidepressant medications. That’s a huge impact, if it’s to be believed.

In our review, we pointed out a few reasons for interpreting these results cautiously. One was that the treatment group started out with a lower quality diet after randomization than the control group. If the low-quality diet was contributing to the depression, then improving diet quality should have a relatively bigger impact on people with lower quality diets. Since the treatment group started out with lower quality diets by chance, this may account for some of huge difference compared to the control group. Another possible cause for caution was that there was a lot of error in the effect size estimate; while the average effect size was indeed large, the data were compatible with effect sizes much closer to antidepressants’ effects, too. The reason for the large error is due in part to the relatively small sample size of the study. A third problem was the large dropout rate in the control group; over a quarter of people there dropped out, compared to the single-digit percentage dropout rate of the treatment group. That’s a bit strange — I would expect that, all things being equal, more people would drop out from the diet group, since overturning your diet seems harder than showing up to chat with someone a few times over the course of three months.

A colleague recently pointed out a letter to me critiquing the SMILES trial that puts forth an argument that could explain the high dropout rate in the control group while providing another reason to take the results with a huge grain of salt. In short, the authors of the letter suggest that the recruitment methods used for the SMILES trial made it very clear to participants that diet can impact depression, thereby raising the participants’ expectancy that the treatment diet would help their depression. Since the main measure of depression of the trial was subjective, this could lead people in the diet arm to report large improvements based on their expectation, and people in the control arm dropping out since they were disappointed they weren’t assigned to what they thought would be an effective diet for their depression. The authors of the letter also provide some evidence that the researchers who rated the participants’ depression independently may have had their blinding broken, also inflating the effect size.

If you thought the effect size that the SMILES trial found looked too good to be true, you may have been right.

Gregory Lopez, MA, PharmD
Editor-in-chief, Nutrition Examination Research Digest

See other articles in Issue #51 (January 2019) of Study Deep Dives.