Why Do We Eat? A Neurobiological Perspective. Part III

In the first post, I explained that all voluntary actions are driven by a central action selection system in the mesolimbic area (the reward system).  This is the part of you that makes the decision to act, or not to act.  This system determines your overall motivation to obtain food, based on a variety of internal and external factors, for example hunger, the effort required to obtain food, and the sensory qualities of food/drink.  These factors are recognized and processed by a number of specialized 'modules' in the brain, and forwarded to the reward system where the decision to eat, or not to eat, is made.  Researchers divide food intake into two categories: 1) eating from a true energy need by the body (homeostatic eating), e.g. hunger, and 2) eating for other reasons (non-homeostatic eating), e.g. eating for social reasons or because the food tastes really good.

In the second post of the series, we explored how the brain regulates food intake on a meal-to meal basis based on feedback from the digestive system, and how food properties can influence this process.  The integrated gut-brain system that accomplishes this can be called the satiety system.

In this post, we'll explore the energy homeostasis system, which regulates energy balance (energy in vs. energy out) and body fatness on a long term basis.

The Energy Homeostasis System

As we saw in the last post, food intake can vary considerably from day to day based on a variety of factors. If we compare the energy intake to the energy expenditure of an individual person on a particular day, the two often match up poorly (1).  Some days, a person eats hundreds of calories more than energy needs and sits on the couch, while other days a person exercises a lot and eats less than she needs to break even.  Yet if we look on a longer time scale, say a week, energy intake and expenditure match up fairly well.  If we look at an even longer time scale, say two weeks, the two usually line up precisely (2).  Even a modest persistent mismatch between intake and expenditure would lead to rapid changes in fat mass over time*, but in most people this doesn't occur.  This implies the existence of a system that matches intake with expenditure over a long time scale to maintain the stability of fat stores.

The first evidence for such a system came in 1840, when a German doctor named B. Mohr determined that a subset of his obese patients carried tumors in a part of the brain called the hypothalamus (B. Mohr. Wschr Heilkd, 6:565–574. 1840), which we now know cause obesity.  In the 172 years since then, an enormous body of research has accumulated in support of the idea that body fatness is regulated by the brain, and that the seat of this regulation lies primarily in the hypothalamus (4).  This makes sense, since the hypothalamus specializes in homeostatically regulating a variety of processes in the body, including body temperature, blood sugar, blood pressure, and electrolyte balance.  There is no other organ in the body that has been shown to regulate body fatness**.

Similar to the satiety system, the energy homeostasis system stabilizes body fat stores by measuring body fatness and adjusting food intake and energy expenditure accordingly***.  Only this time, the regulation happens over days, weeks and months rather than minutes and hours.  As with the satiety system, the "control center" of the energy homeostasis system resides in the brain, but not the brainstem.  The brain is able to modify calorie intake much more easily than calorie expenditure in humans, and so changes in food intake are the main way this system acts to 'defend' the size of fat stores.  It does this at least in part by influencing the satiety system, changing the degree of hunger and the desire for food at individual meals (5).  For example, if you haven't eaten for a week, you'll be very hungry and you'll be able to eat much more than usual before reaching fullness for a few days.  This reflects the energy homeostasis system influencing the satiety system to restore long-term energy balance.

As with a thermostat that measures temperature in order to adjust it, the hypothalamus measures body fatness in order to adjust it.  It does this by monitoring the circulating concentration of hormones that reflect body fat levels (6).  Chief among these is leptin****.  Leptin is secreted by fat tissue in proportion to its size.  The more fat, the more leptin in the circulation. Overeating and fat gain increase leptin levels, signaling to the brain to decrease subsequent food intake, and in some cases increase calorie use, to restore fat mass to its original level.  Conversely, undereating and fat loss lower leptin, which signals to the brain to increase hunger, increase interest in in food, and decrease calorie use in an attempt to recover the lost fat (7).  Replacing leptin to the pre-weight loss level attenuates these responses by tricking the brain into thinking that no fat was lost (8, 9).  This is the crux of why it's hard to lose fat: the brain 'defends' current fat stores against changes in the short/medium term, whether you're lean or obese.

Well Then How Does Anyone Gain Fat?

The energy homeostasis system wasn't designed for super sized french fries, pizza, or 32 oz Slurpees that you can get in 5 minutes by jumping into a car or simply making a phone call.  It wasn't even really designed for a roasted leg of lamb with rosemary, garlic and salt that you can make easily in your own kitchen.  It was really designed for simple foods like raw fruits, plain meat cooked on a fire, and plain roasted nuts, with no added salt and few added fats, sweeteners or other flavorings-- foods that also required work to obtain and prepare.  It was designed for daily physical activity, sleep and sunlight.  In that environment, the energy homeostasis system doesn't have to work very hard to constrain body fatness because food intake isn't constantly being driven to excess by non-homeostatic factors such as easy availability, high palatability, liquid calories, and constant advertising.  Its main worry historically was preventing starvation, which it evolved to be very good at.  The energy homeostasis system can only do so much to constrain body fatness in the face of excessive eating pressure, and our ancient brain is simply no match for the modern food environment in which non-homeostatic factors strongly drive food intake beyond energy needs.

This energy homeostasis system is more robust in some people than in others.  In a fascinating experiment published in the journal Science in 1999, researchers overfed 16 non-obese young adults with an excess of 1,000 calories per day (beyond energy needs) for 8 weeks (10).  They found an amazing 10-fold difference in fat gain between individuals, ranging from 0.4 to 4.2 kg of fat mass.  Some people were able to burn off the excess calories without any increase in voluntary exercise*****, while others gained most of the excess calories as body fat.  This suggests that the energy homeostasis system is more robust in some people than in others, and those who inherited a weak system are much more prone to fat gain.  Keep in mind that this system doesn't have to defend against fat gain unless a person is overeating, so these individual differences are only relevant in the context of excess food intake.  In 2012, I believe most people in affluent nations frequently overeat due to our unnatural food environment, and those who have a less robust homeostatic system gain fat over time as a result.

Obesity in animals and humans is associated with a resistance to the actions of leptin (and other related signals) in the brain.  In other words, similar to insulin resistance, the brain can't 'hear' the leptin signal very well.  That means it takes a lot more leptin, and therefore a lot more fat mass, for the hypothalamus to be satisfied that a person isn't starving.  Consequently, the brain of an obese person 'defends' current fat stores as if the person were lean, initiating a starvation program if fat mass declines, even if the person still has twice as much total fat mass as a lean person (11).  As a result, a weight-stable obese person has a higher energy 'flux' than a lean person-- more energy leaves the body because obese people have a larger tissue mass to sustain, and therefore more calories must enter the body to maintain weight (12).  The brain maintains this situation by keeping food intake high-- an obese person has to eat more food on average to feel satisfied than a lean person, but the difference isn't very large-- only about 20 percent higher than the lean level for someone who is very obese (13).  The difference is even smaller for someone who is only moderately overweight. This difference is therefore neither immediately obvious nor easy to measure.

What Causes Leptin Resistance and an Increased Setpoint?

To understand why we hold on to the fat we gain, and why it's hard to lose fat, we must understand what causes resistance to leptin and other signals of body energy status.  In earlier posts, I speculated about a variety of factors that may contribute to leptin resistance and an increased 'setpoint' (14).  I believe that physical inactivity can increase the setpoint over time, and in some overweight people exercise can lower it, although for most people exercise alone isn't a very effective fat loss strategy.  The ability of exercise to prevent fat gain is well supported by animal studies, which suggest that exercise maintains leptin sensitivity (15, 16).  Certain drugs, such as atypical antipsychotics, can increase the setpoint, presumably by acting on brain circuits that regulate body fatness.  There is some evidence that the sensory qualities of food, in addition to promoting non-homeostatic eating, can also increase the setpoint (17, 18).  One of the most compelling hypotheses is that leptin resistance is caused by low-grade inflammation in the hypothalamus, since animals that can't develop this inflammation are partially protected against fat gain (19, 20, 21).  Yet we still don't know for sure what causes the inflammation, even though many people including myself are working on it.  There are other likely contributors to an increased setpoint that I won't get into here.

I think all of these things contribute, but I recently had a breakthrough, and I now believe there's something else at play, something bigger.  It's actually quite simple.

Something I've noticed over the years is that when you make an animal gain fat by giving it a fattening diet, and then return it to the original healthy diet, it will lose most of the excess fat but frequently retain a portion of it (22, 23).  Similarly, in human overfeeding studies, after subjects return to a normal diet, they'll spontaneously undereat for a while, lose most of the excess fat, but if you read the studies carefully you find that they often hang on to a fraction of the excess fat indefinitely (24, 25).  The piece of evidence that pushed me over the edge was a study showing that half of annual weight gain in the US occurs during the 6-week holiday period (26).  People gain weight during the holidays by voluntary (non-homeostatic) overeating, lose a little bit of it in January, but hang on to most of it indefinitely.

I believe this is the missing piece to the puzzle of how the body fat setpoint gradually increases over time, leading to the 'defense' of a higher fat mass.  What increases the setpoint, in a susceptible person, is fat gain itself.  Each time a susceptible person eats more calories than necessary to meet calorie needs, the excess is stored as body fat.  Afterward, a portion of the excess fat is lost as the energy homeostasis system kicks in, but a portion of it remains and is 'defended' in addition to what was already there.  Over time, the setpoint gradually ratchets up in response to temporary periods of overeating.  How this happens on a cellular/molecular level, I don't know, but it must coincide with the incremental development of leptin resistance in the brain since that's what supports the elevated setpoint (along with resistance to other circulating factors).  This process may also tie in with inflammatory signals.  I want to be clear that there are probably multiple contributors to an increased body fat setpoint, and this particular one may not apply to everyone, but I've come to see it as the most likely primary driver in most people.

This is both exciting and sobering at the same time.  On one hand, it gives us a simple explanation for this frustrating phenomenon that has undermined the fat loss efforts of millions of people.  On the other hand, it suggests that there's no easy fix for an increased setpoint.  I wish it was just a nutrient deficiency, a lack of sleep, or some such 'secret', and fixing it would melt away the fat effortlessly, but that's simply not the case.  That being said, there are scientifically supported strategies that can cause durable fat loss without having to bean count calories, and some probably act by changing the setpoint.  I've researched these strategies and used them to design a fat loss program in collaboration with sleep and food intake researcher Dan Pardi and other scientists.  This program will be available through the Dan's Plan website in late December 2012.

The Updated Model

Here's the new version of the simplified model of food intake regulation.  As before, the colored shapes are brain 'modules', and the words outside them are the external and internal factors they respond to.  We're most of the way through.

* This doesn't mean that eating 50 extra calories per day more than you do now will lead to rapid fat gain.  Actually, fat gain will plateau at a modestly higher level because larger bodies expend more calories.  But if you continually eat 50 calories more than your body expends, thus continually increasing your calorie intake, you will gain fat rapidly.

** I'm using the word "regulate" in the true sense of the word here, as in something that controls a variable in a purposeful manner.  To influence a process is not the same as to regulate it.

*** It also regulates a number of other variables related to body fatness, including the release of fat from fat tissue.

**** There are others as well, including but not limited to insulin, ghrelin, and amylin.

***** Involuntary increases in non-exercise activity thermogenesis accounted for the differences in energy expenditure.