On Eating Moderately

Disclaimer: Nothing I write here is a criticism, in the pejorative sense, of Guyenet’s advice; he reports scientific findings faithfully and we can ask no more than that.  Instead, I want to explore why some people find his advice discouraging.  Also, note that Guyenet’s advice involves more than just reducing food reward – he says you can also lose weight by eating a high-protein diet, exercising more, sleeping better, and reducing stress.

Scott Alexander suggested that the implications of The Hungry Brain are “neo-Puritan”, and I more or less repeated that claim in my review of the book.  However, it could be argued that every diet plan is puritanical – after all, every diet plan, from the quackiest to the most scientifically sound, offers the same basic promise: Restrict in some way your consumption of delicious foods, and you will lose weight.  Is Guyenet’s advice any different?

I think it is.

Consider alcohol.  The actual Puritans drank heavily by modern standards, but nevertheless, we might summarize a (metaphorically) “puritanical” position on alcohol as follows:

  • Heavy drinking is unhealthy.
  • Moderate drinking is also unhealthy, but not as unhealthy as heavy drinking.
  • The healthiest practice is to not drink at all.
  • Therefore, while you should not beat yourself up if you are unable to abstain entirely, you should nevertheless strive to drink as little as possible.

But in reality, there is something approaching a scientific consensus that moderate drinking – perhaps one to two drinks a day – is no more dangerous, and perhaps better for you (where by “you” we mean someone not genetically predisposed to alcoholism), than abstaining entirely.  So “sound scientific advice” on alcohol is something more like this:

  • Heavy drinking is unhealthy.
  • Moderate drinking causes no net harm for most people.
  • Therefore, if you are able to restrict your drinking to one or two drinks a day, you have done enough; there is no need and indeed no benefit in restricting it further.

And it seems to me that most dietary advice is structured this way.  Consider the South Beach Diet:

  • Eating saturated fats and refined carbohydrates is unhealthy.
  • Eating unsaturated fats and unrefined carbohydrates causes no harm.
  • Therefore, once you have eliminated saturated fats and refined carbohydrates from your diet, you have done enough; there is no need and no benefit to excluding other delicious foods.

This advice doesn’t seem “puritanical”, does it?

Or consider Michael Pollan’s advice to “Eat food; not too much; mostly plants” (he’s been saying misleading things about GMOs lately, so he doesn’t get a link):

  • Eating commercially-processed foods or too many animal products is unhealthy.
  • Eating whole foods causes no harm, so long as you don’t eat too much or too many animal products.
  • Therefore, once you have eliminated commercially-processed foods from your diet and reduced your consumption of food in general and animal products in particular to a moderate level, you have done enough; there is no need and no benefit to further reducing your consumption of delicious foods.

These examples “moderate” by restricting certain categories of food, but there are other ways of moderating that we could describe in similar ways:

  • Moderate your calorie count; so long as you do not exceed this count, you have done enough, and can eat whatever delicious foods you want, albeit in small serving sizes; there is no benefit to moderating further.
  • Moderate how frequently you eat, or how frequently you eat certain foods; so long as you meet these goals, you have done enough, and can eat whatever delicious foods you want; there is no benefit to moderating further.

What these approaches all have in common is that they dodge the charge of “puritanism” by giving us an “out”; once we have satisficed the requirements of the diet plan, we take comfort that science has our back, because (we believe) science has proven that further moderating our diet will no benefit.  We have achieved the state that is the diet-marketer’s Nirvana: Guilt Free!

The discouraging thing about the “Food Reward Hypothesis of Obesity” is that it does not seem to suggest that we can ever reach that state – there is, as of yet, no scientifically-justified point at which we can say “further moderation does me no good”:

  • Heavy consumption of high reward foods is unhealthy.
  • Moderate consumption of high reward foods, or consumption of foods with moderate reward, is also unhealthy, but less so.
  • The healthiest option is to eat foods with almost no reward at all – the fastest way to lose weight is a bland liquid diet.

This isn’t to say that Stephan Guyenet would recommend a bland liquid diet to most people; he would surely make allowance for the psychological difficulty of adhering to such a diet.  But to the extent that we do not consume a bland liquid diet, we must blame the failure of our willpower; for most people, no realistic diet plan can promise that state of freedom from guilt – and perhaps as importantly, neither can capitalist marketers nor government regulators honestly sell such a promise.

Fortunately, it’s not clear that the Food Reward Hypothesis works the way I just described.  In particular, the diets of our evolutionary ancestors are at least slightly tastier than a bland liquid diet, which suggests that our hypothalamuses (hypothamali?) are not such fragile snowflakes that even the slightest food reward with disrupt our leptin systems.

The obesity epidemic is said to have begun in the 1970s, so will we have “done enough” if we simply roll back our diets until they are no more “rewarding” than those we ate in the 60s?  Unclear:


This figure, reproduced from Guyenet’s book, shows that obesity was much less of a problem on “The 1960s Diet” than it is on the modern American diet.  But that’s not to say obesity was a tiny problem: Even in 1960, nearly 15% of adults were obese, and (not pictured) roughly 30% were overweight.  It may be that genetic variation is at work: Perhaps most peoples’ physiologies are built such that 1960 levels of “food reward” do not harm their hypothalami, but a significant minority must moderate even further (also, it seems likely that people who are already obese and want their hypothalamus to heal should moderate further than people who are not yet obese.)

But suppose for the moment that The 1960s Diet “does enough” – is it still not discouraging to believe that there is no healthy way to improve on the culinary delights of Mad Men?  Maybe this is still too pessimistic a reading of the Food Reward Hypothesis – Guyenet, in his comment on the previous post, said the following:

…(T)he key aspect of reward that drives us to eat is the motivation, not the pleasure per se. Pleasure is just a convenient marker that tends to travel with motivation.

…which suggests a tantalizing possibility for enlightened hedonism – perhaps it is possible to prepare food that produces great pleasure without producing food reward that is damaging to the hypothalamus – a higher form of pleasure, poetry to junk food’s pushpin.  I linked to one intriguing article in my review; here is another – although note that intuition alone informs my speculation.  Unfortunately, science is probably not yet ready to give us a complete answer to the question, “How can we eat guilt-free?”


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 )

Google photo

You are commenting using your Google 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