Why Do We Eat? A Neurobiological Perspective. Part I

As with all voluntary movements, eating food is an expression of activity in the brain.  The brain integrates various inputs from around the body, and outside the body, and decides whether or not to execute the goal-directed behaviors of food seeking and consumption.  Research has uncovered a lot about how this process works, and in this series I'll give a simplified overview of what scientists have learned about how, and why, the brain decides to eat.

The Gatekeeper of Voluntary Behaviors

Let's start at the broadest level.  The brain has the capacity to drive feeding behavior at any time of day.  Why do we only seek food and eat it at some times and not others?  Why do we perform any behavior at specific times and not others?  The brain contains a sort of 'gatekeeper' function that selects among all possible behaviors at any given moment, executing those that are currently the most relevant. 

This gatekeeper function is integrated with the reward system*, centered in the basal ganglia.  This system is responsible for selecting/motivating all voluntary behaviors, including the seeking and consumption of food.  Stimulating this system in the right way can strongly influence feeding behaviors (1, 2).

The neurotransmitter dopamine is a critical element of the reward system.  Mice lacking dopamine are a fascinating case study in the function of this system-- they have no motivation to do anything.  Their behaviors almost all remain latent and they just sit in their cages, not eating or drinking until their dopamine is chemically replaced (3)**.  Dopamine-deficient mice can still react to things-- if you place food in their mouths, they'll chew and swallow; if you startle them, they'll jump; if you place them in water, they'll swim.  However, when they swim the path they take through the water is random, compared to the goal-oriented path a mouse would usually take to try to escape.  Dopamine-deficient mice still have the ability to move their bodies, and they can still react to certain things using hard-wired behaviors, but they are totally unable to execute voluntary, goal-directed behaviors.  They're constitutionally apathetic toward everything, including food.

Conversely, increasing dopamine levels in mice increases motivation to eat, increases food intake, and increases body weight (3b)**.

The reward system is a gatekeeper through which all voluntary behaviors must pass to be expressed.  Whether or not the reward system selects/motivates a behavior depends on a variety of inputs.  Keep this in mind, because it's central to the process of food intake regulation. 

What is Hunger?

In addition to being a sensation, hunger is a motivational state.  When a person or animal is hungry, he is motivated to seek and consume food.  That motivation can make people run many miles in pursuit of a gazelle with nothing more than a spear and a water bottle (in the case of hunter-gatherers), or on a smaller scale, it can make you move a fork from your plate to your mouth.  Seeking food and eating it represents the reward system motivating those behaviors rather than other potential behaviors, because the brain has decided based on various sources of input that eating is a high priority at that moment.

Hunger is Only One of the Reasons We Eat

Researchers have divided eating into two categories, which are important to understand: 1) 'homeostatic eating', in which food intake is driven by a true need for energy, and 2) 'non-homeostatic eating', in which food intake is driven by other factors.  Eating in response to hunger is mostly homeostatic, while eating for pleasure, emotional/stress reasons, social reasons, or just because it's mealtime, is non-homeostatic.  As I'll explain in more detail later in this series, non-homeostatic factors determine food intake at least as much as homeostatic factors in the modern food environment.  Here's Dr. Hans-Rudolf Berthoud, a researcher who has written extensively on this topic (4):

The initiation and maintenance of ingestive behavior is co-determined by metabolic and non-metabolic factors. Among the latter, environmental cues, as well as reward, cognitive, and emotional factors, play an important role, particularly in human food intake in the modern world.

A common sense example is all we need to begin to understand this.  The holiday season is the scenario in which Americans are most likely to overeat and gain fat.  That's not because we're suddenly hungrier on Thanksgiving-- holiday weight gain is driven almost exclusively by non-homeostatic overeating: the presence of readily accessible, delicious, energy-dense, diverse food, and social eating and drinking.  The average American overeats during the holidays, gains fat, and hangs on to most of it indefinitely (5):

In subjects who completed one year of observation, the weight increased by an average of 0.32 kg during the holiday period and 0.62 kg over the entire year, suggesting that the period contributing most to yearly weight change is the six-week holiday period.

Holiday weight gain accounts for about half of total annual weight gain in American adults, and is therefore an excellent example of non-homeostatic overeating leading to weight gain (5).

Another example of non-homeostatic eating is soda consumption.  People don't choose calorie-dense soda over plain water because they're hungry or thirsty-- they choose it because they like soda.  Most people only weakly compensate for the extra calories they drink by eating less later.

As obvious as it sounds, we eat because we're motivated to eat, and there are many factors that can motivate us to eat.  These factors are recognized and processed by many specialized hardware 'modules' in the brain, and forwarded to the reward system to determine if they are sufficient cause for action.

When we're trying to understand what causes obesity (and how to reverse it), the most important question is not "what makes us eat?"-- it's "what makes us eat more than what we require for leanness?"  To answer that question, we need to understand both homeostatic and non-homeostatic eating-- the motivations driven by hunger and factors other than hunger.  That's one of the reasons why models that focus exclusively on hunger and satiety fail to explain human eating behavior in the real world.

The other reason such models fail is they don't take into account the system that regulates long-term energy balance and body fat mass (the energy homeostasis system).  This system influences the circuits that govern reward and satiety, adjusting motivation and hunger in accordance with current energy stores.  For example, if you've under-eaten for a day, or weeks, your energy homeostasis system detects the energy deficit and responds by making you hungry and motivated to seek food.  If you've overeaten, it detects the excess energy and responds by reducing hunger and the attractiveness of food in subsequent meals (6)   In some people, it can also compensate by cranking up energy expenditure (7).  The degree of compensation for excess energy intake varies between individuals, and is one of the factors that determines a person's susceptibility to fat gain (7).

A Neurobiological Model of Food Intake Regulation

Let's start with this diagram, illustrating the key systems that govern food intake.  It will be highly simplified, but sufficient to illustrate the central points.  Each colored shape loosely represents a functional and anatomical brain 'module' specialized for a particular task (or set of tasks).  As we proceed, I'll fill in the other modules that converge on the reward system (and thus determine our motivation to eat), and the external factors they respond to:

As you can see, the reward system receives a number of inputs from other brain regions, makes a decision about what behaviors to motivate, and allows those behaviors to be expressed through regions of the brain that control muscular movements (motor systems).  Every time you buy food at the grocery store, every time you eat out of hunger, every time you eat out of boredom, every time you drink an alcoholic beverage, every time a fork moves from your plate to your lips, it's happening because of coordinated activity in these brain systems.

In the next post, I'll continue to unveil this model by exploring homeostatic eating.


* I ran this post by two researchers who work in reward-related fields and are familiar with these circuits, to verify its accuracy. One of them commented that although it's commonly called the 'reward' system in the scientific literature, he thinks a more intuitive name for the function I'm describing is the 'action selection' system. I'm going to keep calling it the reward system just for consistency (with previous blog posts and the literature), but I do think his point is worth noting.

** Dopamine replacement does not quite bring them back to the level of a normal mouse, and consequently dopamine-deficient mice eat less and are leaner than normal mice.  Conversely, mice with more dopamine eat more and weigh more than normal mice.  These are obvious predictions of the food reward hypothesis.  Lowering food reward reduces the motivation to eat, whether that is accomplished by genetic manipulation or by eating food that is inherently low in reward value; increasing food reward increases the motivation to eat.