Hormesis and the limbic brain

There is a powerful way to re-program your brain that has been largely overlooked.  A way to change your relationship with eating, sleep, sex and basic emotions like fear, love and aggression.  While cognitive therapies can modify behavior, they are of questionable help in altering these basic drives.

Our drives are largely governed by two small primitive brain structures, the hypothalamus and the amygdala — shown in red in the drawing at right.  Remarkably, these two tiny structures are respectively the size of a pea and an almond — representing less than 1% of the brain’s three pounds of neural matter. Together, they constitute the control center of the paleomammalian brain–the “limbic” brain that governs our basic urges and desires as well as our homeostatic “set points” for temperature, sleep, body fat and behavioral urges like sex drive and aggression.

You can attempt to change your behavior by conscious determination and cognitive therapies.  But most attempts at intentional change are temporary and are doomed to fail in the long term because they are strongly resisted by powerful homeostatic processes encoded in our limbic brain.  Modern medicine recognizes the importance of homeostatic drives, and has developed pharmaceuticals to override them with diet pills, sleeping pills and antidepressants.  In fact, these medications do shift the balance of neurotransmitters and neural activity — at least in the short term.  But such chemical interventions are short-sighted “crutches” that promote dependency and come with side effects.  Often they exhibit  a “tolerance” effect: the brain’s control system fights back and weakens the impact of the medication.  To maintain the benefit, doses are increased, but this strategy may not always work.

This article will explain how the hypothalamus and amygdala contribute to the regulation of basic drives like eating, sleeping and sexuality, and how the amygdala can actually override the hypothalamus by enhancing the reward value of foods and other stimuli. (As I will explain, however, my take on “food reward” is different from that of Stephan Guyenet and other advocates of the Food Reward Hypothesis). This dual-control model can help explain anomalies such as obesity, addiction, and disordered sleep.

Finally,  I will provide suggestions on effective and natural ways to re-program the hypothalamus and amygdala and change your homeostatic set points, using the principle of hormesis.

Hormesis. Readers of this blog are familiar with hormesis:  a biological process whereby a beneficial effect (improved health, stress tolerance, growth or longevity) results from exposure to judicious doses of an agent that is otherwise detrimental at higher doses.  The many examples of homesis we’ve discussed on this blog involve adaptations that roughly fall into three categories.  The first two categories are quite well-known:

Structural adaptations to organs and tissues:

Defensive adaptations against foreign subtances:

The third category is perhaps a less well recognized form of hormesis:

 “Psycho-metabolic” adaptations:

Psycho-metabolic adaptations. Let’s now expand upon this third category of adaptations, focusing on how certain types of stimulus or “stress” can bring about long term changes within the brain’s control system — the hypothalamus and amygdala.  These adaptations can induce broad sets of changes to your metabolism and psychological functioning.   These changes are long term adaptations — to be distinguished from short term or “artificial” changes that can temporarily induce weight loss, boost metabolism, energy level, wakefulness, or sex drive.   A true change in “set point” requires a sustainable physiological change that is reflected in real alterations in neuron density or receptor sensitivity within the brain.  In turn, these changes to the brain result in systemic changes elsewhere in the body.

In previous posts, I’ve touched upon a few topics that relate to the general thesis of psycho-metabolic adaptations that involve changes to the brain:

  1. In “Change your receptors, change your set point“, I presented evidence that individuals suffering from obesity, addiction and depression have in common a down-regulation (reduction in the number or sensitivity) of dopamine receptors. In depression, receptors for other neurotransmitters such as serotonin are also down-regulated, a problem that can actually be made worse by chronic use of SSRI antidepressants.  The article also summarized research indicating that intense exercise, caloric restriction and intermittent fasting can up-regulate dopamine receptors and thereby provide a sustainable treatment for certain types of obesity, addiction and depression.
  2. In  “Obesity starts in the brain“, I outlined the Hypothalamic Hypothesis, a brain-centric analysis of obesity.  I argued that there are two different types of obesity–intra-abdominal and subcutaneous obesity–and that these conditions respectively result from  impairments to the insulin sensitivity or leptin sensitivity of a specific part of the hypothalamus — the arcuate nucleus.  Furthermore, it is the hypothalamic impairments that are primary; for example, insulin resistance starts in the brain and later spreads to the liver and muscles.  The article pointed to specific dietary and inflammatory factors that can improve hypothalamic sensitivity to these hormones and reverse obesity.

I will now build upon the Hypothalamic Hypothesis to account for the influence of the amygdala, to consider how the limbic system governs for drives other than eating, and to propose more generally how we can influence the brain’s control system.

The limbic system. Think about this:  By weight, about 85% of the human brain is the elaborate cerebral cortex, devoted to complex perceptual and conceptual processing and executive function.  In contrast, only a tiny piece of the brain is responsible for the full gamut of motivational drives and emotions, and for maintaining the balance of homeostatic functions like metabolism, body temperature, sleep and energy level.  The simultaneous management of all of these diverse functions is tightly packed into two nut-sized structures–evidently without getting signals crossed! When you think about it, this fact is quite astonishing.  It baffles me that, despite great popular interest in neuroscience, there has been so little commentary about this striking fact.

You can think of the the massive cortex as merely an elaborate pattern recognition system wrapped around the limbic brain.  The cortex’s pattern recognition system has evolved to improve the quality of information being fed to the tiny thermostatic hypothalamus and amygdala.  While the cortex gives us a huge advantage over other animals in analyzing our environment, we seem not to exert much real control over basic drives like eating and sleeping.  Despite the evolutionary achievement of “rationality”, we humans remain to a large extent at the mercy of our basic animal drives and emotions.

Things are not so bleak, however, once we recognize what makes the limbic brain tick.  While we may not have direct volitional control over the limbic system, there are actions we can take to influence the balance of neural forces within the hypothalamus and amygdala. Over time, we can literally reprogram our psycho-metabolic control systems.

But first a little anatomy.   And I’ll try to keep things simple.  The point of this interlude is not to teach anatomy, but rather to highlight a few key parts of the limbic control system and how they function. I’ve borrowed much of the following discussion from the excellent and incisive monograph, The Limbic System, by Rhawn Joseph, much of which is also contained in Chapter 4 of his online Brain e-book.

The figure below provides a “macro” view of the major parts of the limbic system.  Located at the center of the brain, perched atop the brainstem, the limbic system includes not only the hypothalamus and amygdala, but other structures such as the hippocampus, cingulate gyrus, pituitary gland.  But particularly note that the amygdala is connected tightly by numerous nerve bundles to the hypothalamus.  The amygdala acts directly on the hypothalamus to control hypothalamic drives, and conversely, the hypothalamus “uses” the amygdala (and to some extent the septum) as a window on the world to satisfy its drives by selectively searching out appropriate foods, potential mates, and sleep and exercise opportunities.


Furthermore, notice that the amygdala is closely connected to the olfactory bulb, and mediates its connections to the hypothalamus.  As Joseph notes, “The hypothalamus is exceedingly responsive to olfactory (and pheromonal) input. Perhaps reflecting this partial and putative olfactory origin is the fact that this structure utilizes chemical (hormonal, humoral) molecules to communicate with other areas of the brain, and reacts to these same molecules as well as olfactory cues, including those directly related to sexual status.”  We will come back to the under appreciated importance of olfactory cues in the limbic system’s control of basic drives, particularly appetite and sexual/social attraction.

For present purposes, there are four important points to understand about the actions of the hypothalamus and the amygdala:

1. The hypothalamus is purely reactive. The hypothalamus regulates drives, but is almost totally “blind” to the outside world.  It is inwardly focused and responds reflexively.  It has no memory and acts “in the moment”.   According to Joseph, the hypothalamus is the physical embodiment of the Freudian id:

Emotional functioning at the level of the hypothalamus is not only quite limited and primitive, it is also largely reflexive… Emotions elicited by the hypothalamus are largely undirected, short-lived, being triggered reflexively and without concern or understanding regarding consequences; that is, unless chronically stressed or aroused. Nevertheless, direct contact with the real world is quite limited and almost entirely indirect as the hypothalamus is largely concerned with the internal environment of the organism. Although it receives and responds to light, it cannot “see”. It has no sense of morals, danger, values, logic, etc., and cannot feel or express love or hate. Although quite powerful, hypothalamic emotions are largely undifferentiated, consisting of feelings of pleasure, unpleasure, rage, hunger, thirst, etc….it tends to serve what Freud (1911) has described as the pleasure principle. Functionally isolated, the hypothalamus at birth has no way of reducing tension of mobilizing the organism for any form of effective action. It is helpless. When tensions associated with immediate needs (e.g. hunger or thirst) become unpleasant the only response available to the hypothalamus is to cry and make rage-like vocalization. When satiated, the hypothalamus can only respond with a feeling state suggesting pleasure or at least quiescence.

2. The hypothalamus operates through a hierarchy of channels.  The hypothalamus receives information about the state of the organism, and in turn sends “commands”,  through three main channels:

  • The bloodstream. Many signals are exchanged through the relatively porous blood-brain barrier.  For example, as discussed in my previous post on obesity, the hypothalamus receives and integrates a range of signals about short term nutrient status (glucose and fatty acids), gut signals (ghrelin, PYY and CCK) and longer term energy storage  (hormones like insulin, glucagon, leptin and adiponectin).   The blood also carries similar signals regarding body temperature, wakefulness and sleep, and state of readiness for action. And the hypothalamus activates the section of neuroendocrine activators via other glands like the pituitary, thyroid and adrenal glands.
  • Nerve fibers –“afferents” and “efferents”.  Certain communication is done via nerve fibers. For example, appetite cues are provided from the nose via the olfactory bulb and from the gut via the vagus nerve.  Body temperature cues are provided from remote thermoreceptors.  The sleep-wake cycle is calibrated by neural inputs from the suprachiasmatic nucleus (SCN), which responds to dark and light cycles.  And conversely, the hypothalamus uses efferent nerves to remotely regulate adrenal glands and digestive organs.
  • Higher order inputs.  The above chemical and neural inputs can be modulated or overridden by “emotional” interpretation of perceptual and cognitive inputs.  This is is where the amygdala comes in.

3. The amygdala is the “handmaiden” of the hypothalamus.  It serves as the emotional eyes and ears for the hypothalamus by translating the input of the senses and the great pattern recognition capability of the higher cortex into emotional responses that feed into the hypothalamus.  Going beyond the undifferentiated, spur-of-the moment emotional drives of the hypothalamus, the amygdala provides a highly selective response to specific and often complex sensory stimuli.  As Joseph explains:

In contrast to the primitive hypothalamus, the more recently developed amygdala (the “almond”) is preeminent in the control and mediation of all higher order emotional and motivational activities. Via it’s rich interconnections with various neocortical and subcortical regions, amygdaloid neurons are able to monitor and abstract from the sensory array stimuli that are of motivational significance to the organism. This includes the ability to discern and express even subtle social-emotional nuances such as friendliness, fear, love, affection, distruct, anger, etc., and at a more basic level, determine if something might be good to eat.  In fact, amygdaloid neurons respond selectively to the flavor of certain preferred foods, as well as to the sight or sound of something that might be especially desirable to eat  including even the sight of drugs that induce extreme pleasure…Belying its involvement in emotion, including the pleasure associated with cocaine usage, is the unique chemical anatomy of the amygdala, which is rich in a variety of neuropetides including enkephalins and beta-endorphins as well as opiate receptors. In fact, of all brain regions, the greatest concentration of opiate receptors is found within the human amygdala.

Beyond appetite, the amygdala also provides a selective filter on sensory cues related to other drives such as sociality and sexual attractiveness.  Of significant note, the amygdala is the arbiter of very specific social cues such as facial recognition:

The amygdala is exceedingly responsive to social and emotional stimuli as conveyed vocally, through touch, sight, and via the expressions of the face . In fact, the amygdala, as well as the overlying (and partly coextensive) temporal lobe, contains neurons which respond selectively to smiles and to the eyes, and which can differentiate between male and female faces and the emotions they convey. For example, the left amygdala acts to discriminate the direction of another person’s gaze, whereas the right amygdala becomes activated while making eye-to-eye contact …Moreover, the normal human amygdala typically responds to frightened faces by altering its activity, whereas injury to the amygdala disrupts the ability to recognize faces. With bilateral destruction, emotional speech production and the capacity to respond appropriately to social emotionally stimuli is abolished.

Maybe this explains why Seth Roberts observation that looking at faces in the morning makes people happy–a simple anti depression therapy!

Joseph also notes that “The relationship between hypothalamus and amygdala is bidirectional.  The amygdala interprets sensory information and emotions and passes these inputs on to the hypothalamus to initiate drives. And when a drive like hunger or sex emerges, the amygdala helps out by surveying the environment for suitable choices of food or potential sexual partners.”

4. The hypothalamus and amygdala  are composed of opposing sets of neural clusters or “nuclei”.   These pairs of neural clusters act in an oscillating ying-and-yang fashion to achieve homeostasis. In both the hypothalamus and amygdala, the external or lateral nuclei activate the parasympathetic nervous system, associated with hunger and digestion, pleasure, relaxation and sexual arousal.  In the case of appetite, stimulation of neurons in the lateral hypothalamus (LH) increases  appetite, releases serotonin and dopamine, and activates anabolic storage of  glucose and fatty acids,  In opposition to the lateral nuclei, internal or “medial” nuclei activate the sympathetic (“fight or flight”) nervous system, which readies the organism for action, increases heart rate, suppresses appetite and sexual desire, stimulates release of acetylcholine and norepinephrine, and activates catabolic mobilization of nutrients such as fat or glycogen.  Stimulation of the medial nuclei are also associated with “aversive” non-pleasurable sensation.

Similar pairings of opposing limbic nuclei exist for neurons that control thirst, body temperature, the sleep/wake cycle, or activate social or sexual arousal.

The amygdala has a parallel structure to that of the hypothalamus, which allows direct two-way communication between them.   As Joseph notes:

Moreover, through the massive interconnections maintained with the lateral and medial (ventromedial) hypothalamus, the amygdala is able to act directly on this structure, driving the hypothalamus, so to speak, and thus tapping into its emotional reserviour so that its ends may be met. Indeed, it is able to modulate hypothalamic activity through inhibitory and excitatory projections to this structure. Direct stimulation of the basolateral amygdala and the ventral amydalofugal pathway excites the principle neurons of the medial hypothalamus. By contrast, stimulation of the medial (ventro-medial) amygdala and the stria terminalis pathway, inhibits these same hypothalamic neurons. Hence, whereas the lateral amydala exerts excitatory influences on the hypothalamus, the medial amygdala exerts inhibitory influences, and can thus control, or at least exert excitatory/inhibitory and thus modulatory influences on hunger, thirst, sexual arousal, rage, etc., as well as hormonal, endocrine, and other functions associated with the hypothalamic nuclues. Indeed, the amygdala can be likened to the chief executive of the limbic system and weilds enormous power via its control over the hypothalamus.

Similar sets of paired hypothalamic and amydaloid nuclei govern the balances that control thirst, body temperature, sleep and sex drive.  For example, osmoreceptors that monitor the concentration of salt ions in blood control thirst, and respond by adjusting the hormone vasopressin to regulate water retention by the kidney. Thermoceptors in the body and hypothalamus activate different nuclei in the hypothalamus.

Generalized versus conditioned desires. By serving as the “interpreter” that provides higher-level emotive “meaning” to raw sensory inputs, the amygdala plays a prominent role in learning and laying down reward circuitry.  In effect, it turns complex sensory inputs into cues that the hypothalamus can act upon by establishing Pavlovian circuits that automate the way your basic drives respond to the external environment and even your thoughts.  This applies to both attractive (stimulatory) and aversive (inhibitory) stimuli. As mentioned above, the reward circuitry utilizes a high concentration of dopaminergic neurons to reinforce powerful learned responses of the hypothalamus to sensory cues and thought patterns.

While the hypothalamus activates generalized drives and provides hard-wired low-level responses to universal and fairly general cues, the amygdala provides finely tuned and highly specific learned responses that can modify or override these low level cues:

The hypothalamus gets hungry and anything will do…,but the amygdala is picky about which foods it likes or dislikes, to the point of craving a specific type of chocolate with a certain texture, or rejecting a wine with a slight off-note
The hypothalamus wants sex…but the amygdala is selective about what turns it on — down to very fine preferences regarding appearance, aroma, or even sense of humor.  It may be so selective as to be monogamous!
The hypothalamus wants to sleep… but the amygdala picks up cues about danger that can rally your alertness.

The key point is this:   The generic drives of the hypothalamus are equally powerful whether they are activated by low level chemical and nerve inputs from the blood stream or stomach nerves — or rather by higher level perceptual and emotional inputs from the amygdala.  And if the reward circuitry from the amygdala is strong enough, it can override the low level signals.   A Pavlovian response to the aroma of a juicy steak or the sight of a decadent chocolate cake can activate the hunger response and fat storage program initiated in the lateral hypothalamus, regardless of the nutritional state conveyed by blood glucose or leptin and insulin levels.  Conversely, an unappetizing meal, or an emotional shock can quickly suppress appetite or activate a state of arousal and access to energy.

The hypothalamus doesn’t know or care why it is getting hungry, sleepy or sexed up.   It matters not whether the signals are based on blood chemicals or high level emotional perception — the actions taken by the hypothalamus are identical in either case.

An aside on food reward. This dual model of direct hypothalamic regulation versus conditioned amygdaloid regulation of drives like hunger can shed some light on the recent debate about the Food Reward Hypothesis of obesity.  Stephan Guyenet has cited compelling evidence for the FRH, based on the  observation that rats fed a “cafeteria diet” of highly palatable junk food became fatter than rats fed calorically matched standard bland rat chow.  Merely adding flavor or flavor variety to the chow also resulted in fatter rats.

However, in an earlier post, “Does tasty food make us fat?,  I argued that Guyenet’s version of the FRH suffers from two logical flaws:  First, Guyenet does not take a clear position on whether “reward” is an inherent property of foods, or rather a learned or conditioned property, relative to individual and cultural experience.  Second, while rewarding food is associated with obesity, the causal sequence can be questioned.  I think it is likely food reward is the is the consequence, not the driver of psycho-metabolic dysregulation.  Food becomes rewarding only after primary hypothalamic regulation becomes impaired, for example by the way that the particular fats and sugars in junk food desensitize hypothalamic receptors to insulin or leptin, as I described in “Obesity starts in the brain“.   Of course, once the amygdaloid food reward circuits are established, they can be expected to perpetuate an increased appetite and shift away from fat mobilization to fat storage.  But the amygdaloid reward circuit is not the primary defect — that remains the impairment to the hypothalamus.  The proof is that it is not just appetite that is impaired — it is also the metabolic consequence of a more active lateral hypothalamus and inhibited ventromedial hypothalamus.   If the hypothermic defect is repaired, the food reward circuit should extinguish.


Hormesis and the hypothalamus.   So how do we use this information?  Specifically, how do we “judiciously” apply “stress”s to re-program our limbic control system. What if we are gaining weight due to both a strong appetite and more “efficient” storage. Or what if we have trouble falling and staying asleep?  Or (more speculatively) what if we want to become more or less aggressive, or more or less sexually motivated?

In short, our understanding of the limbic system suggestions two approaches:

1.  Direct reprogramming of the hypothalamus. Every drive is regulated by a balance of stimulatory and inhibitory neurons.  By the logic of hormesis, we can stimulate the growth of one set of neurons or the other by periodically  “starving” them of their normal stimuli, allowing a compensatory up-regulation of receptor neurons.  Often this process is slow, and the compensating adaptations may take weeks or longer — but with sustainable results. This is the reverse logic illustrated in several posts.

  • Change your receptors, change your set point”  demonstrates how exposure to uncomfortable stresses such as intermittent fasting, strenuous exercise, cold showers and the like can up-regulate dopaminergic neurons and thereby counteract conditions such as obesity, addiction and depression.  While the research cited in that article doesn’t specifically locate the dopamine neurons, , we know they have a high density in the hypothalamus, amygdala and other limbic structures, and the PET scans indicate a brain location consistent with the hypothalamus and amygdala.
  • A cure for insomnia?” describes the use of Sleep Restriction Therapy (SRT).  By forcing extended wake cycles, there is an apparent rebalancing of hypothalamic neurons in the ascending arousal system, thereby activating sleep-active neurons in the ventrolateral preoptic nucleus (VLPO) associated with the  “flip-flop switch” that produces distinct sleep-wake states.  As a result, SRT reduces the  excessive production of corticotropin-releasing factor (CRF) that is associated with many cases of insomnia.
Several other articles suggest the possibility of re-adjusting the homeostatic set points of our hypothalamic drives:
  • Flores et al have found that extended exercise can directly improve insulin and leptin sensitivity in the hypothalamus, based upon IL-6 signaling.
  • Marnia Robinson and her husband Gary Wilson have developed a therapeutic method to “reboot” sex drive and romantic interest, based upon deliberate restriction of sexual stimulation for several weeks, combined with alternate forms of intimacy.  Their rebooting method can even reverse problems such as erectile dysfunction and has been found useful in combatting addiction to pornography. They cite evidence that dopamine and prolactin circuitry is at work with both the problem and the solution.  Both the hypothalamus and amygdala regulate sex drive, so it would be interesting to know exactly how “rebooting” affects the relevant neural nuclei.
2.  Reprogramming the amygdala. This is the indirect way to re-program the hypothalamus, by altering the amygdaloid reward circuitry that feeds it.  There are a number approaches to achieving this, some of which I’ve outlined in previous articles, but all of them fall generally under the umbrella of classical or Pavlovian conditioning.  There are a few basic strategies:
  • Extinction.  An addictive response becomes weaker and eventually dies out when you stop responding to a triggering cue.   This approach works, but can take a long time and requires patience and discipline.
  • Cue exposure or deconditioning.  This involves deliberate, repeated and provocative exposure to the triggering cue, withholding the response.  After some initial discomfort, this approach proceeds rapidly and can be quite effective.  Success is improved the more realistic and varied the presentation of the cue.
  • Putting on cue.  A new cue is developed and the behavior is only allowed in the presence of this cue.  It could be a special sound, or a location.  Then the special cue is withheld and the behavior disappears.
  • Counter conditioning.  This involves the substitution of an alternative behavior to actively displace the old reward circuitry.  It can be very effective.
I’ve written several posts that illustrate the use of classical conditioning to alter reward circuitry.  These were written before my research into the limbic system, so they are lacking or wrong in the details regarding the role of the hypothalamus and amydala in the re-programming process.  (I hope to flesh out those details in future posts):

The anatomy of the limbic system offers one other strong leverage point into reprogramming the amygdala-hypothalamus axis: namely, the prominence of the olfactory bulb.  The olfactory bulb directly innervates the amygdala, and there is ample support that smell and taste are powerful triggering cues for the appetitive and sex drives.

Several diets are based on control of this powerful trigger, as I have argued in my post on Flavor control diets.  Flavor and flavor variety tend to stoke appetite, due to direct classical conditioning of the amygdala (and without the hypothesized intermediation of a preprandial insulin and blood glucose mechanism, as I erroneously speculated in my original article, which I intend to re-write based on my current understanding). While some diets work by either suppressing flavor (Shangri-la Diet) or limit flavor variety to induce sensory-specific satiety (Flavor Point Diet), these approaches don’t reprogram the amygdaloid flavor-appetite reward circuit. They merely avoid appetitive triggers, which remain intact until re-activated.  I think the most effective way to change your appetite is via the above-mentioned Deconditioning Diet, which directly modifies reward circuitry, presumably within the amygdala.

The use of olfactory cue conditioning to modulate other hypothermic drives is worthy of exploration.

A final speculation.  Admittedly, this is one of my more speculative articles.  While I have started out in the known physiology of the limbic system, I am to some extent going beyond proven data in my judgements and recommendations.  So I’ll continue one step further down the path with a parting thought.  At the beginning of this article, I expressed my astonishment that the control of so many apparently distinct drives — eating, sleeping, body temperature, aggression, sex drive and sociality — are all packed into two structures the size of a pea and an almond.  It seems quite remarkable the the neurons and circuitry for these different drives remain distinct and do not interfere with one another.  But perhaps they are not so distinct.  In fact there is some evidence that they interact.  For example, many have reported that fasting makes them feel colder and may depress thyroid function, at least in the short term.  Fasting also may result in reduced sex drive and changes to the sleep cycle.   So the hypothalamic control of feeding,  body temperature, sleep and sex drive may interact.  To some extent, these effects may be compensated for by actively exercising, which appears to increase body temperature.   In addition, these short term interactions may or may not persist during longer term adaptations.

I take cold showers every day and have found they raise my energy level and mood and help me to stay lean.  In his book, The Four Hour Body, Tim Ferriss correctly rejects the thermodynamic argument of Ray Cronise that cold showers and baths promote weight loss based because  shivering involves significant energy expenditure.  Not only do the energy calculations fail to add up, but this explanation would defy the principle of homeostasis:  If we lose weight by shivering, and nothing else changed, our hypothalamus and leptin accounting system should compensate by driving us to increase appetite to restore the lost weight.  Ferriss proposes what I think is a more plausible explanation, namely that cold exposure induces metabolic changes that cause a replacement of white adipose tissue (WAT) with more metabolically active brown adipose tissue (BAT). Interestingly, work by Cao et al at Ohio State recently found that the conversion of WAT to thermogenic BAT is triggered by the the action of BDNF in the hypothalamus.  Interestingly, BDNF is a stress response hormone that is also up-regulated by intermittent fasting or calorie restriction.  Furthermore, it is  is known that the hypothalamus responds to cold exposure by up regulating the production of thyroid stimulating hormone (TSH) which directs the thyroid gland to output thyroid hormones T3 and T4, increasing basal metabolic rate.  That alone could explain increased energy levels and weight loss, which may be sustained so long as the cold stimulus is provided at a certain frequency.   There are likely many other examples cross talk” between temperature regulation, eating behavior and hypothalamic regulation of other drives.

More needs to be explored on how control of our apparently distinct drives interact with each other.  This can be helpful in designing strategies for effective diet and exercise, and for addressing sleep and sexual problems.

Given the speculative nature of this article, I would be more than interested in feedback and suggestions for further investigation or development of the ideas presented here.


  1. Brian

    The suggestion made in this post is provocative – that cold showers affect both metabolism and dopamine receptors. This sounds right to me, based on my own experience.

    Also, intense exercise seems to work best alongside, or “within,” aerobic exercise. Much of the research is done on interval training within a longer treadmill session, for example.

    With regard to your last question, the one suggestion is that you look into citicholine:


    This is one of the few studies to show a substance having this sort of affect on the dopamine receptors themselves, instead of dopamine levels.

    (citicholine is a white powder, taken like a supplement)

  2. Todd

    Thanks, Brian. I will take a look at the citicholine study you linked.

  3. I would love to see this material applied to getting rid of visceral fat! I keep reading your articles, and they make sense, but I don’t get how to apply them.

    • Todd

      Hi Ted,

      The key fat loss recommendation in this latest article, as well as the previous one (“Obesity starts in the brain”) to apply hormetic stress. Intermittent fasting, high intensity exercise (not slow aerobics), and cold showers will lower basal insulin, increase BDNF in the brain, stimulate norepinephrine and alter fat metabolism. A secondary recommendation is to minimize (not necessarily avoid) foods that contain compounds known to inflame the hypothalamus — principally high levels of fructose and sucrose, or palmitic acid (found in meat and dairy that is grain-fed rather than grass fed). You can also add anti-inflammatory oils (fish oil, coconut oil) and anti-inflammatory vitamins and minerals (vitamin D, magnesium, zinc) to your diet.

      The above recommendations concern direct modification of hypothalamic function. The second main point I made in the article is that regulation of body fat (and other drives) can be changed via “deconditioning” of behavioral responses to cues that are coded in the amygdala. The most specific advice I can provide is given in my article on the Deconditioning Diet (http://bit.ly/x2EvOh). In short: (1) cut back on carbohydrates and cut out snacks (2) use cue-exposure therapy to extinguish your conditioned cravings; (3) cut out occasional meals and attempt intermittent fasts of 12-20 hours.

      Changes to the hypothalamus typically take weeks to months, so don’t expect an immediate benefit. On the positive side, the resulting adaptations are typically quite robust and sustainable.

      Hope that helps give you a more specific idea of how to apply these ideas.

      • Thank you Todd. Hopefully I’ll have good results to report in a year. 🙂

  4. Rob R

    In the realm of the amygdala you may be interested in the work of Joesph LeDoux on the disruption of memory reconsolidation.

  5. Todd wrote: “However, in an earlier post, ‘Does tasty food make us fat?’, I argued that Guyenet’s version of the FRH suffers from two logical flaws: First, Guyenet does not take a clear position on whether “reward” is an inherent property of foods, or rather a learned or conditioned property, relative to individual and cultural experience. Second, while rewarding food is associated with obesity, the causal sequence can be questioned. I think it is likely food reward is the is the consequence, not the driver of psycho-metabolic dysregulation.”

    Neither would be a logical flaw (if either is a flaw at all).

    The first is merely a feature of his view. He allows that both are factors in food reward. The second is an empirical question, not a question of logic.

    It is unlikely that Guyenet, as careful as he is, will make logical errors by the time his views hit print. Not impossible, but very unlikely. His views invite many empirical questions, though. That’s the nature of Science.

    • Todd


      On the strict deductive sense of the term “logic”, you are of course correct. But there is another, broader sense of the term “logic”, namely inductive logic, which is well recognized by philosophers of science. “Logic” also has the informal meaning of “argument” or “reasoning” that applies in many fields including science and the law.

      The point I was trying to make is that Stephan is probably (as usual) correct regarding his reporting of empirical facts, but I think his reasoning is not airtight. First of all, he uses the term “reward” equivocally — sometimes to mean an inherent property of food; other times to mean a property that is variable or can be conditioned. But logically, it can’t be both at the same time, and this equivocation has implications for how he interprets experiments and the conclusions he draws.

      Sorting out causation from correlation may seem like a simple empirical matter. But it is often a complex question that relies not only on the immediate experimentation, but a wider body of science, and conceptual or “logical” issues of interpretation. But I will grant you that sorting out causation involves a significant empirical effort.

      Stephan is usually careful not to mistake correlation for causation, but I think he does just that when he argues that rats that become obese while eating a “rewarding” cafeteria diet do so because of the rewarding character of the diet. My point is that “reward” likely starts out as a consequence, not a cause, of the “obesity” (which appears to have its genesis in what the food does to the brain).

      Whether or not I’m correct in my specific objections to the Food Reward Hypothesis, my objections regarding equivocation and causal logic are not about empirical mistakes, but rather about flaws or fallacies in argumentation, or “logic”.

      Hope that clarifies.


  6. Am I using ‘food’ equivocally if I use it sometimes to mean ‘fruit’ and other times ‘meat’?

  7. Tim


    How much can an improvement in dopamine sensitivity for a given reward circuit(s) relating to one type of motivation (i.e. food-based) effect the dopamine sensitivity on another reward circuit (i.e. sexuality-based)? Are dopamine receptors shared between the circuits? If so, to what extent?

    The point of my question – if DA receptors are highly shared between different types of motivations/addictions, a change in one behavior would affect others, i.e. short-term/medium-term sexual abstinence may be an effective strategy to combat obesity.

    If there is any truth to this, I cannot find any research making the link.

    • Todd


      While dopamine receptors are ubiquitously distributed throughout the brain, reward circuits are mostly separate. However, to answer your question, “starving” the brain of pleasure will tend to result in a generalized upregulation of dopamine receptors. (See my article on Overcoming addiction for hard evidence of this based on PET scans of the brain). As Rhawn Joseph put it, the hypothalamus strives to maintain a stable “pleasure budget”, regardless of the source of pleasure. So witholding food, sex, or good music, should improve our general senstivity to any sort of pleasure.

      In this sense, abstinence or curtailing of pleasures may be helpful in dampening our appetite for food. Keep in mind, however, that upregulation and sensitization of dopamine receptors takes weeks to months. So this is a long term project. It can be frustrating for those bent upon quick results, but the results can be dramatic and worth the wait for those who have the discipline.

      One big caveat: Besides a generalized pleasure level, most pleasures also contain a specific component: our preferences for food, sex or music can be incredibly specific. That specific component cannot be readily substituted by differente pleasures. Listening to our favorite music may not do much to quell our cravings for our favorite dessert.

  8. Shadowfoot

    I would be interested in your thoughts on the importance of stress oscillation in neurological conditioning compared with physical adaptation. For example, does the necessity of a “rest” period apply when training food cues?

    • Todd

      Good question, Shadowfoot. I think that psychological and physical conditioning are not too dissimilar, although perhaps less “rest” is required for effective psychological conditioning.

      Some clues come from Pavlov’s original studies in his opus “Conditioned Reflexes” which report detailed experiments with various schedules of stimulus frequency. Other clues come from Richard Solomon’s studies of “opponent process” conditioning, which I summarized in my post on the Opponent process theory of emotion. In particular, Solomon highlighted the importance of 3 factors: the intensity of the stimulus, the duration of the stimulus and the interstimulus interval. He found that the interstimulus interval had a major effect and he identified the “critical decay duration” that characterizes the the rate at which adaptive conditioning fades or extinguishes. There is also a residual or “savings” component which represents a long term behavioral memory, one that re-activates even after a long time has passed.

      While the studies of both Pavlov and Solomon support the idea that increased frequency of cue exposure generally improves reinforcement, there is also a well-known saturation effect to conditioning or learning. My post on Overcoming addictions cites the work of Conklin and Tiffany on cue exposure therapy. They found that deconditioning of addictive behavior is most effective when exposures are separated by sufficient time, both within and between cue exposure sessions. Some time needs to pass for short-term memory to consolidate into long-term memory. This is intuitive: you can only study for a test or practice piano so much and then it becomes counterproductive, so you are better off sleeping on it and resuming training the next day.

      I’ve not seen studies that specifically measure the time constants of reinforcement, savings, and saturation for cue exposure to food cues, but I see no reason why these could not be measured. There would undoubtedly be some variation among individuals and by type of cue, but the results might be instructive and helpful in developing a training protocol.

  9. Todd, have you seen this? http://addiction-dirkh.blogspot.com/2012/01/reward-and-punish-say-hello-to.html

    I thought the concept of “reward prediction error” interesting.

    • Todd

      Thanks for the link, Beth. Bozarth’s concept of “motivational toxicity” is indeed interesting, particularly the idea that this arises not from excess dopamine per se, but rather from impairment to the inhibitory GABA circuitry that counterbalances it. This results in the “reward prediction error” you refer to. The normal checks and balances are replaced by an open-ended circuit.

      The question then becomes how to restore the normal balance and thereby recover a homeostatic ability to “predict” reward. This can be a very difficult hole to dig out of, for serious addicts. I think that sustained extinction is the only sure route, assisted by active cue exposure reconditioning and reinforcement of alternate pathways. Some of the empirical support for this approach, as explored in the study by Conklin and Tiffany, is discussed in my post on Overcoming Addictios.

  10. Sue

    Todd, you are a fabulous writer – you write like one should – imagine your audience is a complete novice on the subject. some writers just like to use big words while not making much sense.

    • Todd

      Thanks for the kind words, Sue. Actually, I often start out as a novice in researching many of my articles, so I need to distill the science down to its essence to make sure I understand it myself.

      It also helps that I have a central organizing principle — hormesis — that provides a compelling theme and lens through which to focus my analysis. I try to provide a perspective that is both thought provoking and useful to my readers.


  11. Bill Rowles

    Since we are speculating here, and adding my n=1, I have been practising IF for over 2 years, and cold showers for 2 months or so.

    If I were to apply the pareto principle to my experience, and admittedly, the cold showers are relatively new, I would say that any fat loss achieved is 80% mediated via IF, and only 20% by CS..

    Also, in my experience, while it less important than IF, low – moderate CHO is helpful, and if you pause for a moment, completely harmonious with fasting metabolism.

    Having said all that – as a combination – IF/CS (let’s include strength/anaerobic exercise) should provide a powerful synergistic hormetic stimulus. Low CHO just makes sense – ie why derail a good thing you’ve got going?

  12. anna

    What if a person with low thyroid (hypothyroidism) makes the cold showers? High TSH in such cases is not welcomed.

  13. Tom

    Hi Todd , tnx for the interesting articles. I have a few questions for u:

    For me , the most important mood booster is intense natural workout (pull ups , dips and push ups) for 50 minutes duration. as a result of a car accident I broke my hand and leg and I must avoid training for few months. As a former soccer player , and 3-4 times a week trainer it is really hard for me not to do sport. my past concludes ocd and depression, which both reduced a lot after doing intense workout. to find solution I found your articles , and for few months I do cold showers and for 1 month I do 16-8 Intermittent fasting. I do both activities in order to up regulate my dopamine receptors. I must say that it helps but much less than intense workout. so my question is how much the fasting process take for full results? and if you have another recommendations for me to up regulate the “mood”

    • Todd


      I sympathize with your being unable to do your intense workout routine while you heal. I think the fasting and cold showers should help keep those dopamine receptors sprouting while you are out of action, but I agree that the “intensity” is less than what you get with physical workouts. I found that the full adaptation to intermittent fasting takes about 3-4 weeks. For cold showers, try to go longer (more than 5 minutes) and colder (less than 6o degrees F) or even try a cold bath with ice cubes or an ocean or lake swim — for at least 10 minutes.

      If that’s not enough, read my post on the Opponent-Process Theory of Emotion. Consider what really exciting activities you could do without hurting your leg or hand. Sky diving or bungie jumping might not be advisable, but can you think of other activities that will get your heart pumping? One of my personal favorites is rock climbing. It’s both a physical and emotional challenge that never ceases to produce a long lasting “high”.

      Let me know what you come up with!


  14. tom

    Hey todd, 6 months ago I posted here. I asked you for alternatives to sport due to injury. I am not fully heald as I need another few months to recover. right now i can do cycling. I remember that you recommended on interval training. I am very intersted on this topic. what duration of training do u recommens? how much intervals?. I know that is depends on many variables, but I am looking for energy and emotional benefits. tnx

    • Todd


      Cycling sounds good, and it’s an ideal sport for intervals. If you are well enough to ride outside, try mixing hills or sprints of at least one minute, with more relaxed cycling. Or you can use a stationary bike for “spinning” type workouts. I’m not sure what your injury is, but you could also consider rowing, swimming or weight lifting. You can certainly do some long endurance workouts, but it is better to mix it up, add at least 1 or 2 days each week with intense intervals, several minutes long, that push you to your limits, combined with low intensity recover to catch your breath between sprints.


  15. tom

    Hey todd thanks a lot for the replay, and the caring. I am very glad discovering this blog. As for my injury – I have tricep tendinosis that never goes away, and also overuse wrist pain which prevents me from lifting weights. the wrist pain developed 3 months ago as I did negative push ups which push a lot of stress on the wrist and not resolved to this day. the pain does not bother me daily basis in my daily activities, only when I am trying to workout. do u know any good herb/suplement or certain food which helps with tendon and ligaments repair? a freind of mine recommended on indian herb called “Cissus”, are you familiar with that?. the injuries are quite depreesing as the doctor did not find any problems in my wrist. as for another hormetic stressors I use cold showers, Intertimment fasting, and moderate exposure to the sun. I also took your advice and did sky-dyving which was quite rewarding. but for me nothing is like weight training.I know i need to rest but it is hard for me, I will happy to hear if you have ideas how to help the body recover from those overuse injuries. also i have an off-topic question- I know that the fasted state and also the cold showers are increasing noradrenaline in the brain. noradrenaline are made from dopamine, so basically when my noradrenaline levels increase do the dopamine levels are increasing also ? or the dopamine increase only for a short time and then they converts to noradrenaline, do we have direct relasionship between high/low dopamine and high/low noradrenaline?

    thanx a lot, much appreciated, Tom.

    • Todd


      Thanks for the update. Skydiving and weight lifting are both great, enjoyable sources of hormesis!

      Tendonitis, such as lateral or medial epicondylitis, is inflammation of the extensor or flexor forearm muscles due to overuse. For overuse injuries you need allow adequate rest and recovery. But it’s also useful to use “opposing” exercises. If you are doing a lot of “pulling”, compensate with “pushing motions”, and vice versa. One exercise that is surprisingly effective is the Thera-band Flexbar:

      I don’t know about Cissus. For muscle soreness, you might consider curcumin (in a bioavailable form), omega-3 fatty acids, and magnesium to strengthen your endogenous anti-inflammatory capacity.

      But long term, and better than supplementation, is building your strength gradually through judicious weight training with care not to overtrain.

      As you mention, dopamine gets converted to norepinephrine and then to epinephrine – all 3 are catecholamine neurotransmitters. However, unless you are really stressed out and overtraining, your metabolism is perfectly capable of replenishing these neurotransmitters. Don’t worry about “depletion”. For your general sense of well being and pleasure in life, you should focus more on the level and sensitivity of your dopamine receptors, not the transient levels of neurotransmitters, as I discussed in this post:



  16. Joelyn Cada

    Would you say that the Hypothalamus is why vegans and vegetarians have lost their sense of humor regarding diets? I think there may be a possibility that their lymbic system is in constant ready to attack mode because they subconsciously are hungry for and in need of meat.

    • Todd

      I never thought of that — maybe you’re right! By the way, did you hear the joke about where the term “vegan” came from? It’s an old Inuit word for “bad hunter”.


Add Your Comment