I’ve written many times that losing weight is the second hardest thing I ask my patients to do. (Breaking an addiction like smoking or alcoholism is the hardest.) The frustrating thing is how little we know about how to lose weight successfully. But we are learning more all the time about why losing weight is so difficult.

Much about dieting and weight loss is poorly understood, but let’s first lay out some facts that are well established.

Weight loss and weight gain are caused by an imbalance between calories ingested and calories burned. That’s not controversial. If you eat fewer calories than you use in exercising, you will lose weight. If you eat more, you will gain. How many calories it takes to simply maintain one’s weight varies between individuals and the mechanisms behind that variation are still being explored, but for every person there is a number of ingested calories below which weight loss will happen. That means that if someone else is in control of what you eat (for example in a prison in a totalitarian country) and doesn’t provide you enough food, you will lose weight.

That makes it sound fairly simple, right? If you eat less, you lose weight. Since eating is a volitional behavior, overweight people should simply choose to eat less, and their failure to do so simply reflects poor judgment or weak willpower.

Wrong.

Permit me a brief digression about control systems. I think about them a lot because of my engineering background. Our body has many mechanisms that very tightly regulate certain biological parameters, like the sodium concentration in our blood, or the amount of light that is shining on our retinas. Many of these mechanisms are entirely out of our conscious control. For example, if we walk into a brighter environment our pupils automatically constrict, letting less light hit our retinas. That happens without our attention or knowledge.

The control of our breathing is a very interesting example. Our breathing is usually not under our conscious control. Our brain monitors the amount of carbon dioxide (CO2) in our blood from moment to moment. When the level of CO2 increases we take a breath, lowering the CO2 level. The cycle repeats continuously even in our sleep. Without our attention or intention the CO2 level in our blood is kept within a fairly narrow range. But anyone who plays a wind instrument or sings can tell you that breathing is also volitional. You can take a breath purposefully between sentences and blow through a horn exactly when you want to. So which is it? Is breathing voluntary or not?

The answer depends on the time scale. From second to second you can control your breathing. You can hold your breath for a few seconds or you can hyperventilate for a few seconds. But over minutes you will not be able to override the drive to keep your CO2 at a certain level. That is, if you try to hold your breath or slow down your breathing over minutes, your CO2 will slowly climb and your urge to breathe faster will eventually prove to be irresistible. Similarly if you try to hyperventilate over minutes, your CO2 will fall and your urge to slow your breathing will eventually overwhelm your conscious control. So breathing is voluntary over seconds but entirely involuntary over minutes or longer.

Are you getting a sense of how this may relate to control of weight?

Long ago researchers began suspecting that there were control mechanisms responsible for maintaining weight within some range. Just as there is an internal set point for our blood sodium concentration that the kidneys maintain, and a set point for our CO2 concentration maintained by our breathing, researchers argued that there must be an internal set point for our weight. A set point simply means a normal level of some measure that a control mechanism tries to achieve – the temperature that the thermostat is set to, for example.

I first discovered the idea of a possible weight set point in a fascinating paper by Seth Roberts, a psychologist. He cites much evidence that weight must be controlled by an internal set point. For instance, many people occasionally fast for a day. This results in a small weight loss. Without an internal set point for weight, that weight loss would be permanent or would fade very slowly. But weight loss after a fast usually disappears within a few days, suggesting that hunger is increased for the subsequent few days until the weight renormalizes.

The general idea is that the quantity of fat stores in our body is monitored by our brain (perhaps using hormones released by fat cells) and compared to some set point. Whenever our weight (or fat stores) falls below this set point various hormonal mechanisms increase hunger and decrease physical activity. Research is currently attempting to unfold the details of these mechanisms. The current understanding and consequences of this theory is explained in an illuminating article in the New York Times Magazine, The Fat Trap. If you’re trying to lose weight, I urge you to read it.

The article cites several studies including a study published in The New England Journal of Medicine. The study enrolled 50 overweight or obese adults and for 10 weeks put them on a very low calorie diet. They lost an average of 30 lb. Before the study and periodically for a year after, the levels of hormones thought to mediate hunger and satiety were measured. The subjects were also asked for their subjective levels of hunger and appetite.

The results showed that the hormones that cause hunger and weight gain increased after the weight loss and remained increased a full year later, even after most subjects had partially regained their lost weight. More sobering is the fact that the subjects’ self-reported sense of hunger rose after the weight loss and didn’t return to baseline levels throughout the one year study.

The authors state that the result “supports the view that there is an elevated body-weight set point in obese persons and that efforts to reduce weight below this point are vigorously resisted … suggesting that the high rate of relapse among obese people who have lost weight has a strong physiological basis and is not simply the result of the voluntary resumption of old habits.”

For now, this isn’t a particularly helpful discovery, but it helps explain a lot. It explains, for example, why the myriad diets on the market all have approximately the same lousy long-term success rates. It also explains that eating, like breathing, and like refraining from scratching that patch of eczema, is a voluntary behavior only on short time scales. I can choose whether to have a snack now or not, but I can’t choose to fast for three days or to eat much less than my caloric needs for a month.

Overweight people have a “weight thermostat” that is turned up too high. We need researchers to to find a medical solution to reset this set point or to break one of the mechanisms that mediate hunger.

The best we have to offer overweight patients at this point is the advice to diet and exercise, though in the long term this seems to be effective only for a small minority of patients. For the morbidly obese, surgery for weight loss is an increasingly evidence-based option.

Perhaps the best advice we can learn from this is to at least encourage patients not to gain more weight. We now know that losing it will be much more difficult and that maintaining the current weight after weight gain and loss will be harder than never gaining in the first place.

Albert Fuchs is an internal medicine physician who blogs at his self-titled site, Albert Fuchs, MD.

Submit a guest post and be heard on social media’s leading physician voice.