Saturday, February 2, 2013

Why Do We Eat? A Neurobiological Perspective. Part V

In previous posts, I explained that food intake is determined by a variety of factors that are detected by the brain, and integrated by circuits in the mesolimbic system to determine the overall motivation to eat.  These factors include 'homeostatic factors' that reflect a true energy need by the body, and 'non-homeostatic factors' that are independent of the body's energy needs (e.g. palatability, habit, and the social environment).

In this post, we'll explore the hedonic system, which governs pleasure.  This includes the pleasure associated with food, called palatability.  The palatability of food is one of the factors that determines food intake.

The Hedonic System

The pleasure of food (and everything else) is determined by the activity of circuits mostly in the mesolimbic system and brainstem, which in turn respond to other brain areas that perceive the sensory qualities of food and compare them to pre-existing food preferences.  Basically, the hedonic system gives you a good feeling when you eat food it considers valuable, and it does this in part by releasing a class of compounds called endogenous opioids or 'endorphins' (a contraction of endogenous + morphine) (1, 2).

It has been known for a long time that injected opioids can increase food palatability and food intake. Conversely, blocking opioid signaling reduces food palatability and specifically reduces the excess intake of highly palatable foods, without reducing the naturally lower intake of simple foods.  This has been shown both in animals and humans.  Here's Ann E. Kelley* and colleagues (3):
Simply stated, it is argued that opioid receptor agonism generally enhances food intake by increasing the positive hedonic valence of food, while opioid receptor antagonism reduces or blocks this affective response.  Such a position gains support from human studies as well as the animal literature. Specifically, systemic infusions of the opioid antagonist naltrexone in humans reduce reported affective or pleasantness ratings of sweet and fatty foods while not affecting subjective reports of hunger or the ability to detect sweet or salty tastes. Similar data has been reported in rats from a variety of paradigms.
The opiate receptor blocking drug naltrexone is half of the anti-obesity drug Contrave, which is effective at reducing body fat but failed to secure FDA approval due to the possibility of cardiovascular side effects.

The cannabinoid system is also involved in hedonic and reward signaling in the brain.  Cannabinoids were named after cannabis, because they are the body's own chemicals that act on the same receptors that cannabis hijacks.  Most people know that cannabis increases food palatability and food intake, a phenomenon called 'the munchies'.  The striking effects of frequent cannabis use on food intake and body weight were demonstrated in a 13-day experiment published in 1988 (3b):
Two cigarettes containing active marijuana (2.3% delta 9 THC) or placebo were smoked during both the private work period and the period of access to social activities. Smoked active marijuana significantly increased total daily caloric intake by 40%. Increased food intake was evident during both private and social periods. The increase in caloric intake was due to an increased consumption of snack foods as a consequence of an increase in the number of snacking occasions. There was no significant change in caloric consumption during meals. The principal increase within the category of snack foods was in the intake of sweet solid items, e.g., candy bars, compared to sweet fluid, e.g., soda, or savory solid items, e.g., potato chips. Increases in body weight during periods of active marijuana smoking were greater than predicted by caloric intake alone.
With an understanding of the neurocircuits that underlie food palatability, and a sprinkling of common sense, it's not difficult to imagine that the pleasure associated with eating food is an important determinant of food intake.  Excessive palatability suppresses and/or overrides satiety, leading to continued food intake beyond energy needs (4).  This is easy to demonstrate in rats.  If you leave a rat in a cage with free access to normal rat chow, it will eat to fullness, consuming all the energy its body needs based on hunger.  If you wait until a rat has finished eating, then put a highly palatable food item in the cage such as a piece of candy or salami, it will find its 'second stomach' and continue eating beyond the point of fullness (5).  This is called the 'dessert effect', and it's the same thing that happens to you when a big chocolate brownie finds its way onto your plate after a meal.  It happens at least in part via the action of endogenous opioids in the mesolimbic system, which are a signal to continue eating a preferred food beyond energy needs (6, 7).

As I wrote in a previous post, the work of John de Castro has shown that in people living their normal lives, we eat 44 percent more calories at meals rated as highly palatable, than at meals rated as moderate or low palatability (8):

Controlled trials have repeatedly confirmed this phenomenon: people eat more food if it tastes better (910, 111213141516).  That doesn't mean we need to eat unpleasant food all the time to stay lean, but it does suggest that eating simple food that has a 'natural' level of palatability will help support a healthy calorie intake, and that this is more important for people who have a tendency to gain fat.  Remember, eating ancestrally is not just about the nutrients in food, it's also about the way food is prepared and its sensory qualities-- these are an important part of the human ecological niche.  Modern food frequently overstimulates hedonic pathways that originally evolved to keep us well nourished, and this has almost certainly played a role in today's obesity epidemic.

What Determines Food Palatability?

We can break down palatability into two fundamental categories: 1) inborn, and 2) acquired (learned).

We are born with certain hard-wired food preferences that are (for the most part) common to everyone.  The most deeply wired, and the easiest to identify, is the taste for sweetness.  This is present from infancy in humans and many other animals, and arises from the oldest parts of the brain: even infants born without a forebrain (anencephaly) respond to the taste of sweetness (17).  This suggests that we have been under consistent evolutionary pressure to consume sweet foods, and therefore that sweet foods would have been healthy for our ancestors in a natural environment.  Sweetness would have come mostly from fresh fruit and sometimes honey, which are generally safe and nutritious foods.

It's hard to know exactly which food preferences are inborn and which are acquired, but it's likely that the most basic preferences are inborn, such as:
  • Sweetness
  • Fat
  • Starch
  • Calorie density
  • Salt
  • Free glutamate ('umami')
  • Absence of bitterness
Humans are adaptable animals.  We have a high capacity to learn, and food is no exception.  Since we're able to subsist on so many different types of foods, many of which are new to us from an evolutionary perspective, there's no way we can be hard wired to enjoy all things that are healthy and avoid all things that aren't.  This is where the reward system kicks in to determine the palatability of foods over time based on experience.  I'll explain this in more detail in the next post, but basically what happens is that we can acquire a taste for something when a food contains an inherently preferred property repeatedly paired with a new set of properties** (18, 19, 20).

For example, if you smother your brussels sprouts in fat every time you eat them, you will probably acquire a taste for the brussels sprouts themselves over time (21).  The brain comes to enjoy everything associated with the inherently preferred property (fat), in this case the aroma, flavor, texture, and appearance of brussels sprouts.  Brussels sprouts themselves become more palatable.  Conversely, if you always eat brussels sprouts plain, you will never acquire a taste for them because they're somewhat bitter and low in calories.  This is how the brain favors the consumption of foods that it has determined, based on prior experience, are desirable.  The brain can also form taste aversions; for example, you may stop enjoying the flavor of oysters if they make you sick (22, 23).

If you don't like brussels sprouts, a more compelling example may be alcohol, which most people dislike the first time they try it, but nearly everyone ends up enjoying.  In this case, the inborn quality that causes us to acquire a taste for alcohol is the drug itself, which acts directly on reward circuits.  Like highly palatable/rewarding food, alcohol can overstimulate hedonic/reward circuits, causing excessive and uncontrolled consumption, a phenomenon we call alcoholism.

The Updated Model

Here is the simplified neurobiological model of food intake regulation, including everything we've covered so far.  As always, the colored shapes represent functional brain 'modules', and the words on the exterior represent the factors that influence them.

* An outstanding behavioral neuroscience researcher who unfortunately is no longer with us.

** This is a form of classical (Pavlovian) conditioning: pairing an unconditioned stimulus with a conditioned stimulus until the animal learns to respond to the (previously irrelevant) conditioned stimulus.


G said...

This is an excellent series!

I was wondering whether smoking cannabis may also contribute to fat gain. You are likely aware that cannabis causes the "munchies" - excess appetite and increased pleasure from eating.

It seems that it temporarily increases the palatability of ALL food. Not sure if this is technically correct. May just be my amateur description of what's going on.

Tomas said...

I heard someone say that when you like certain food - oysters in this case - it means or could mean that you have specific enzymes to process oysters. Is such a mechanism possible, or confirmed by studies?

Ivan Nikolov said...

How does a 4-oz glass of wine every evening cause me to develop desire to this drink, I wonder? If it is the alcohol in it as a drug that makes me like wine - I've never been drunk from wine and the last time I overdid the alcohol was probably a half a century ago when I was a teen. Any way, what is the process that makes one enjoy wine when 1) Wine doesn't taste all that good the fist time one tries is (just like every other alcoholic beverage), and 2) When I never drink it for the drug (alcohol) influence in it?

Travis Culp said...

The drug would be affecting even if it's not perceptible. The same would be true of coffee even if it doesn't make you noticeably jittery or wired.

Unknown said...

Learn about how you can reach your New Year resolutions at

Unknown said...

Excellent series on why we eat! Of course, emotionally-based cognitive learned responses are a major factor too - the process whereby we "train" ourselves to respond to certain feelings by eating. For example, our mother buys us ice-cream after the dentist, after an exam, when we're not feeling well, etc ... and over time our brain begins to SUBCONSCIOUSLY associate eating ice-cream with making us feel better in uncomfortable situations. Net result: as soon as something uncomfortable happens, we find ourselves craving ice-cream. It is possible to "retrain your brain" and thereby change your subconscious thought patterns, but it takes time and effort.

Stephan Guyenet said...

Hi G,

Absolutely. THC hits the cannabinoid CB1 receptor, which is part of the hedonic system. There was a controlled trial a while back showing that smoking cannabis 3x daily causes rapid weight gain.

Hi Ivan,

It may be acting even at lower concentrations. I can't speak for you, but in my case a glass of wine will have noticeable effects on me even if I'm not drunk.

Hi Bryan,

I do think that's important. I'll be covering it briefly later in the series.

Unknown said...

This site was... how do you say it? Relevant!! Finally I have found something which helped me.
Thank you!

G said...

Thanks, Stephan. This is valuable information. :) Do you happen to have the reference handy? If not, then don't go searching for it.

Stephan Guyenet said...

Hi G,

Here it is:

e-mail me if you want the full text.

sertac kaya said...

nice job for peace.!