Nuanced! Complicated! No. Just Science!

Two years ago I reported the terrible persuasion, but fabulous science, news about calorie restricted diets.  Monkeys who ate a lot less lived longer than monkeys who ate a normal primate diet.  The zealots and pop press FauxItAlls fell for this, while, of course, the science is a bit more unsettled.  Today consider this interesting update on diet and mortality.

Please try and chase down this perspective piece published in Cell Metabolism.  The authors provide a great overview of this research and demonstrate that the simple Eat Less, Live Longer!!! headline is nowhere near true.  They review experimental studies that manipulate not total calories, but types of calories, carbs and proteins for example, in varying proportions with different species and find very different mortality (and fertility) outcomes.  I’m going to show you a fascinating contour graph that illustrates this complexity.  Take a minute to look through this.  Click to enlarge.

If you’re trained as a behavioral scientist, this graphic is a bit confusing because it displays an enormous amount of information in a highly compressed format.  It depicts two independent variables on the X and Y axes with the dependent variables in the body shown with color and contour lines as an implied Z axis.  Let’s get oriented, big picture first.

Realize that the left column looks at Mortality while the right column looks at Fertility.  As you look at each row, you see that each is for a different species, Drosophila, Crickets, and so on.  Within each Figure, the X and Y axis corresponds to the amount of Protein (on the X) and Carbs (on the Y).  The color contours within each Figure show the average survival in days of the insect given the various combinations of Proteins and Carbs.  Think of the colors as contours of a mountain with the red indicating the longest survival and the blue as the shortest.  Thus, these contour maps are 3 dimensional.  You have to imagine the Z axis as a flagpole coming straight out of the page and those different colors as rising “elevation.”  Can you see the different “mountains” of survival or fertility now and how they vary with different combinations of Carbs and Proteins?

Look at the contours for the last two rows corresponding to female and male field crickets.  See that different combinations of Proteins and Carbs produce greater survival depending on sex.  Females live longer with more Carbs and less Protein.  Males live longer with more Protein and less Carbs.  And realize that this is NOT a calorie restricted diet.  It only varies the proportion of nutrient types.

The authors of this paper also note a variety of new studies, most published since 2009, that have found harmful effects to restricted diets.  These directly contradict the Headline of Eat Less, Live Longer.  Now, add in what these color contour Figures show:  Proportion of nutrients determines mortality.  Simpson and Rabenheimer have proposed a Geometric Framework that goes beyond the earlier and simpler models of Eat Less, Live Longer.  GF can explain why calorie restriction works and why it doesn’t work with this balance or proportion framework.  Arguably the positive effect of calorie restriction arises from the fortuitous impact of calorie type – even with fewer calories the critters were getting enough in proper proportion.  And, when calorie restriction is harmful, it arguably arises from a bad proportion.

Now, the GF framework does not work only with Carbs and Proteins and in fact argues it might be a good idea to look at another kind of nutrient, Fat, and work with even more complicated diet proportions.  Like any good theory, GF stimulates as many interesting questions as it does good answers. Shootfire, if you get this kind of variation in survival with just Carbs and Proteins, imagine if you add Fats? More variety expected. And this is with bugs. Do you think it’s going to be less complicated with monkeys and chimps? Humans?

What are the persuasion implications of this science?

Hey, there are a lot of suckers out there and more are born every minute.  With the Drum and Bugle Corps of pop press and health and safety zealots, you can find a lot of ways to make money on simple Headlines.  As long as the FauxItAlls are pitching Eat Less, you can Make More.  Jeepers, the US spends nearly 20% of GDP on health, plans to spend $1 trillion on Health Care We Can Believe In that will make a 1% impact on mortality, passes national laws requiring calories counts on menus that has no impact on eating, and, well, if you read the Persuasion Blog you know the litany. With news like this you have even more plausible confusion. Ratios, baby, ratios. And it’s different for women than for men. And for fertility compared to life span.  It’s a gold rush, so expect a lot of gold digging and gold diggers.

Of course, if you actually want to Make The World A Better Place, you need to think more carefully about your persuasion.  Clearly all the simple Food Police exhortations are not based on Falling Apples.  Eat Less, Ban Fat, Kill Salt, Count Calories are just plain persuasion applesauce aimed at personal and political agendas.  And, I’m not criticizing anyone who’s getting ahead on Simple Science for Simple Minds.  That’s persuasion, baby, and I’ll tip the persuasion hat in your direction.  If that’s how you get tenure, fame and fortune, film at 11 with Charlie Rose, then you are a maven.

But, you’re also a scientific sheep in the persuasion wolf’s clothing.

Piper MD, Partridge L, Raubenheimer D, Simpson SJ. Dietary restriction and aging: a unifying perspective. Cell Metabolism. 2011 Aug 3;14(2):154-60.

PMID: 21803286

P.S. Okay, Steve where are the effect sizes and tests of statistical significance? Why only graphic information? This is a perspective paper that does not report such information, but rather points the interested reader to the original work. I chased down a couple and found effect sizes that range from Small to Large to Ginormous. Each contour graph here shows the results of huge experiments at least by persuasion standards. There are 24 cells in some of these panels with 10-20 “participants” (insects) per cell. The “effect” depends upon what you are looking for and can be played like the statistics guitar, but these are not those trivial observational exclamations of 116%!!! And, they arise from true experiments with replications across species. The numbers look solid to me. But you can chase the references yourself and correct me as needed.