Non-shivering thermogenesis is the process by which the body generates extra heat without shivering. Shivering is a way for the body to use muscular contractions to generate heat, but non-shivering thermogenesis uses a completely different mechanism to accomplish the same goal: a specialized fat-burning tissue called brown fat. Brown fat is brown rather than white because it's packed with mitochondria, the power plants of the cell. Under cold conditions, these mitochondria are activated, using a specialized molecular mechanism called uncoupling* to generate heat.
The mechanism of brown fat activation has been worked out fairly well in rodents, which rely heavily on non-shivering thermogenesis due to their small body size. Specialized areas of the hypothalamus in the brain sense body temperature (through sensors in the brain and body), body energy status (by measuring leptin and satiety signals), stress level, and probably other factors, and integrate this information to set brown fat activity. The hypothalamus does this by acting through the sympathetic nervous system, which heavily innervates brown fat. As an aside, this process works basically the same in humans, as far as we currently know. Those who claim that rodent models are irrelevant to humans are completely full of hot air**, as the high degree of conservation of the hypothalamus over 75 million years of evolution demonstrates.
Two new studies concurrently published in the Journal of Clinical Investigation last week demonstrate what I've suspected for a long time: brown fat can be 'trained' by cold exposure to be more active, and its activation by cold can reduce body fatness.
It was previously shown that non-shivering thermogenesis is higher in winter than in summer in the same people (1), suggesting that it may be a modifiable trait. As an aside, obese people tend to have low brown fat activity (2)****. The first new study demonstrates that brown fat can be trained to be more active in as little as 10 days, and that this training causes people to become more cold-tolerant (3). The protocol involved exposing people to 60 F (15-16 C) air for two hours on day 1, four hours on day 2, and six hours on days 3-10. Although I assume they were lightly clothed, this is a pretty mild cold exposure.
The second study went further, using a longer cold exposure protocol to investigate changes in fat mass among people with low brown fat activity at baseline (4). Researchers exposed volunteers to 63 F (17 C) air for two hours a day over a six-week period; again I assume they were lightly clothed. As in the previous study, they observed an increase in brown fat activity with cold training, and they found that calorie expenditure was higher when subjects were in the 'cold' air. After six weeks of training, body fat mass had declined by about 5 percent. This is despite the fact that all subjects were lean to begin with!
I think this is pretty impressive, given how wimpy the cold exposure protocol was. Think about it: two hours a day at 63 degrees F. That would seem downright luxurious to a person 100 years ago in winter in almost any temperate climate. Today, most people keep their central heat closer to 70 F in winter, apparently eliminating a major source of passive calorie burning.
I've used heat in my bedroom about 20 times over the last three years, usually when I have a guest. I use less heat because I'm an environmentalist, and because I feel that humans are meant to experience cold stress sometimes just as we're meant to experience exercise stress. The winters aren't particularly cold in Seattle, and I leach heat from downstairs, but the temperature nevertheless fluctuates between 45 and 55 F for several months of winter, with occasional excursions below 40 F (I've awakened with a crust of ice on the inside of my windows).
This may sound extreme, but it's nothing compared to how many people in temperate climates lived in the winter before central heat. Open wood hearths are inefficient at heating a house, wood can be expensive, and the fire burns out overnight. Have you ever experienced a temperature below freezing in your home? It's difficult to stay warm without moving your body, and I must confess it's hard for me to tolerate for long, despite my relative cold hardiness. Indoor temperatures below freezing would have been common two centuries ago, as supported by many historical accounts of even the wealthy having to break ice to wash their faces in the morning.
Is the rise of central heating a contributor to the obesity epidemic? These recent studies suggest that it may be. This is consistent with what one would predict based on the evolutionary/ancestral health principle.
* As protons are pumped across the mitochondrial inner membrane as part of the normal process of ATP generation, a channel protein called UCP1 (uncoupling protein 1) diffuses the proton gradient by allowing protons to cross back passively. Since heat is nothing more than molecules moving around, this process of pumping protons back and forth generates heat. This process is called uncoupling because the Krebs cycle is uncoupled from ATP generation by the diffusion of the proton gradient.
** It is true that certain things don't carry over that well from rodents to humans, for example drug efficacy and perhaps the effects of certain diet elements on health. But basic anatomy and physiology are remarkably conserved in many areas. If you look at a rat's hypothalamus, it contains all the same structures as a human hypothalamus, though the proportions and placement are different. Same for the basal ganglia and many other major systems in the brain (the basal ganglia, remarkably, are highly conserved down to hagfish, the most primitive vertebrate). It's difficult to argue that these structures do something completely different in rodents than in humans when they look the same down to the cellular level, and we have empirically determined in many cases that they do function similarly in rodents and humans. I sometimes hear the impassioned argument "I am not a rat/mouse"! Well, actually you are. We share roughly 99% of our genes with mice, and 85% of the sequence of those genes is identical down to the nucleotide. So yes, you are a mouse, roughly speaking***.
*** OK, to be fair, small changes in the genome can have major phenotypic impacts, so percent genome identity isn't necessarily the last word. But science has abundantly demonstrated the usefulness of rodent models for understanding how humans work, and these insights are going to keep rolling in whether the skeptics like it or not. Much of this skepticism is inspired by animal rights sentiments, and/or the inherent distaste people have for being compared to an animal they consider disgusting and stupid. Although I sympathize with animal rights sentiments, and I think they should be part of the animal research discussion, they have nothing to do with the question of how useful animal research is for understanding humans. The argument from disgust is obviously irrational.
**** This could simply be because they're better insulated. Overall metabolic rate (kcals burned per unit lean mass per unit time, or total kcals burned per individual per unit time) is actually higher in the obese.