Navigating Unsettled Science: Lessons from the Macronutrient Wars

By Max Prokopy

There is no shortage of diet advice or money to be made therein.  A quick Amazon search using the terms “diet and health” shows about 75,000 publications.  Several of these are national bestsellers and “The Biggest Loser” is wildly popular.  The point to be made is that losing weight is a common goal, yet as a whole Americans are not particularly good at it.  Buried within this cultural obsession is an interesting observation: people pay a lot of money for diet advice the government hands out for free.  The question is: with all the money and effort spent on diets, have we actually found one that works?

Innumerable scientific studies try to answer the above question.  However, obtaining accurate data is difficult.  Participants are not always tracked for many years and don’t always eat what they were told to eat.  Other confounders like smoking status, pregnancy, physical activity, and emotional stress underscore the difficulties with generating reliable information.  This article highlights two effective but divergent approaches of diet evaluation.  The primary intention is to help health-conscious people read and interpret research more effectively.  A secondary objective is to examine the merits of several popular diets.

A Randomized Long-term Intervention: the A to Z study

Chris Gardner of Stanford valiantly tried to produce a randomized prospective trial, published in JAMA in 2007.  The study was loosely termed “A to Z” as the following four diets (all bestsellers) were examined:

“A”tkins (low carbohydrate [CHO], high fat, modest exercise);

“T”he LEARN approach (USDA recommended intakes plus lifestyle modification – what you can expect from many registered dietitians);

“O”rnish approach (very low-fat plus exercise and stress management); and

“Z”one (slightly more protein and less CHO: 40% CHO, 30% fat, 30% protein, modest exercise).

The above plans run the gamut of CHO recommendations, from very low (Atkins) to very high (Ornish).  Numerous studies show that dieters tend to switch back and forth between fat and CHO proportions, while protein intake remains relatively constant.

The study participants (311 total) were randomly assigned to one of the above groups.  They were supplied with the diet book/program plus 8 weeks of material discussion with an enthusiastic aid.  Participants were told the diet’s efficacy and resumed their lifestyles outside of the study.  Investigators followed subjects for 12 months, periodically conducting weigh-ins and health screens.  In Dr. Gardner’s publication the Atkins diet is credited with achieving significantly better mean weight loss (about 10 lbs.) than any of the other three groups (about 4 pounds each).  Furthermore, the Atkins diet group experienced significantly greater reductions in blood pressure and triglycerides (fat in the blood).  HDL (“good”) cholesterol significantly increased in the Atkins diet.  Such results seem to seal the deal for Atkins.  However, a closer look is warranted.

The study has some advantages but is saddled with several significant flaws.  On the positive side this is a large group of people for a randomized diet trial and the drop-out rate was low.  The study was designed as “intention to treat” so dropouts were included in the final analyses.  This gives the study a “real-world” hue: this is what happens if people buy the diet materials and try to follow it for a year.

Yet, a significant concern is how researchers assessed participants’ diets.  These participants completed three days of food recall surveys every 2 months.  Using food recalls to collect diet information is unreliable, although one might argue it’s equally unreliable across the four diets.  However, the fact remains that only three days of every 60 was examined by a survey.  Neither the Atkins nor the Ornish participants came close to the plan’s recommendations.  After one year, Ornish participants were getting 29% of their calories from fat (instead of 10%) and Atkins eaters ate approximately 32% CHO (instead of 10-15%).  Such uncomfortable facts make it difficult to extrapolate results with any certainty.

Perhaps the most interesting data from the study went unpublished (it can be viewed here: http://www.youtube.com/watch?feature=player_detailpage&v=eREuZEdMAVo#t=2811s).  Overall averages can be misleading as each of the four diets had similar individual results.  Several participants in each group were able to lose 30+ pounds.  Likewise, several in each group actually gained weight.  The Atkins group significantly reduced its reported caloric intake relative to the other plans. Atkins had the lowest drop-out rate of all four plans.  It might just come down to which diet best controls your hunger and is easy to stick with, which varies from one person to another.

An Acute Intervention: the effect of fat on arterial function

In 2006, Brock and Gaesser from the University of Virginia highlighted a different look at the macronutrient debate and a different way to conduct research.  Since the effects of diet on body weight and cholesterol (as measured in the “A to Z” study) are relatively slow to change,   Brock and Gaesser used flow-mediated dilation (FMD) to examine the more immediate effect of diet on physiological function.  FMD measures the ability of a blood vessel to expand (dilate).  This is important as vascular dysfunction plays a role in the development of both atherosclerosis and adult-onset diabetes.

12 healthy study participants stayed overnight in a research facility and fasted.  The following morning they were supplied either a high-CHO/high-fiber/low-fat breakfast or a high-fat/low-fiber/low-CHO breakfast.  The low-fat breakfast was similar in composition to the Ornish diet while the low-CHO breakfast resembled Atkins.  The volunteers had FMD measured before, right after, and four hours after the breakfast.  Each participant stayed on two separate occasions to evaluate each breakfast for each subject (termed a “cross-over” study).

The high-fiber breakfast produced an acute improvement in FMD while the high-fat breakfast acutely impaired FMD.  Over time, these small effects might turn into chronic impairments.  Again, it is the individual results that tell an interesting tale.  Such figures should be included in any good diet study.  Three of 12 participants (25%) showed constant or increased FMD (better vascular function) with the high-fat diet.  The high-fiber diet showed a virtually unanimous improvement in FMD.  This study is an example of a well-controlled intervention.  However, the sample size was small and numerous environmental factors were not involved.  The FMD study appears in stark contrast to “A to Z” which suggested Atkins is a better diet; had a large study group; allowed for environmental factors; but was poorly controlled.


So, which study is better?  Does either one provide better insight into how people can eat healthier?  What can we make of the apparent contradictory results?  Several kernels of truth can be culled out of both data sets:

  1. Individual results vary.  Anyone can lose weight and improve health if they find what’s right for them.
  2. Fiber is a good thing and can be obtained through fresh produce (lower CHO) or cereal grains (higher CHO).  Atkins counts fiber as valuable by subtracting it from total CHO recommendations.
  3. It’s presumptuous to label fat or CHO as inherently better or worse for an entire population.

Final Thoughts

The process of reading science is worthwhile but only if you read carefully.  Keep the following in mind when reading diet studies:

  1. Look for individual results.  Group averages often hide the true responses.
  2. Look at what subjects actually ate and how diet was assessed.
  3. Look for how well controlled the diets were.
  4. Consider the population being studied.  Were participants already in poor health? Were they diverse?  How were factors like smoking and exercise accounted for?

By paying close attention you can obtain valuable information.  However, no one study has all the answers.  The fine print is very important.

*Image source http://www.thelast3reps.com/what-are-macronutrients/

Max Prokopy is a first-year Biochemical and Molecular Nutrition PhD student who has certifications to and experience with training prep, collegiate and professional athletes with a particular focus on ice hockey.

The Friedman Sprout is a monthly student run newspaper that aims to serve the student population at the Friedman School of Nutrition Science and Policy, prospective students, and alumni. Our mission is to report on newsworthy information that affects the Friedman community including nutrition research, food policy, internship and volunteer opportunities, as well as school events. Our editorial slant is that of sustainability in food and nutrition.

0 comments on “Navigating Unsettled Science: Lessons from the Macronutrient Wars

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: