How much weight lifting or other exercise is optimal for fitness? What is the right amount of carbohydrate restriction or fasting for sustained weight loss and health? What level of exposure to allergens will reduce allergies? How many hours of sun tanning is healthy? How frequently should plus lenses be worn to reduce myopia? Do I need to take cold showers every day to get their benefit? How much stress is enough — and how much is too much?
Many of the questions I get on this website and the forums are of this type. People understand the general concept of hormesis, namely that exposure to controlled amounts of stress can be beneficial, because it elicits beneficial adaptive responses in the organism. They understand that weight lifting builds muscles, and that intermittent fasting and calorie reduction can be healthful. But too much of any stressor — weight lifting, caloric restriction, sunlight, allergens — can have adverse consequences. With hormesis, it seems, the Goldilocks principle applies: to get a benefit, the level of stress must be “just right”. And because it’s so easy to veer into overload, many people seek to avoid even mild stress: Avoid allergens. Cover up with sunscreen. Eat frequent small meals. Don’t exert yourself. But if you choose this path, you forgo the possible hormetic benefits.
So how do you determine the optimum level and frequency of exposure to a stress? And how much rest or recovery between exposures is optimal?
These are important questions, difficult to answer with certainty. Of course, all over the Internet you will find those who tell you exactly how many days each week is optimal for lifting weights, how much sun tanning is safe or dangerous, what level of dietary carbohydrate or food restriction is optimal or unhealthy. In some cases, they will cite studies to support their position. But there is one big problem with all this advice, even the advice based upon careful scientific studies:
Individual responses to hormetic stressors can vary significantly.
Just as responses differ between individuals, a given individual’s ability to tolerate and benefit from hormesis changes over time, and as a function of previous exposures to stressors. This makes it virtually impossible to reduce hormesis to a simple formula. And yet, the situation may not be so hopeless. There are actually some tools and metrics we can use to quantifiably determine whether hormesis is helping or hurting us, and thus to “adjust” the dose.
Allostasis. There is a general biological principle that can help us dial in the right level of hormesis. The principle is called “allostasis”. Most people are familiar with the related concept of homeostasis, the tendency of a system to maintain a constant internal state, such as the pH, temperature, or oxygen concentration of the blood, within a fairly narrow range. This concept was developed by the famous nineteenth century biologist, Claude Bernard, who observed that organisms strive to control their internal environment, or milieu interieur, within tight physiological constraints, through physiological processes that resist disturbances from the external environment and quickly restore normal operating conditions. This notion was later formalized by Walter Canon as “homeostasis”, the tendency of a biological system to regulate its internal environment within a stable range.
While the concept of homeostasis has some validity, in actuality it is of fairly limited application. In fact, most biological systems do not self-regulate physiological variables within a narrow range, but tolerate a fairly wide range of variation. During the course of a typical day, blood glucose and insulin levels rise and fall by as much as 50% or more. Blood pressure, heart rate, and adrenaline surge upon waking and standing in the morning, and increase to further heights when engaging in vigorous exercise, or responding to threatening or emotional situations.
Bernard and Cannon developed the concept of homeostasis to apply only to regulation of the internal environment, particularly that of the cell or circulatory system. It was not intended to describe the external condition of organs or whole organisms. Yet others have extrapolated this concept and applied it to the misleading notion of “set points”. For example, some have advanced the idea that each of us is born with a body weight set point from which we can only deviate transiently and in a futile manner through diet and exercise, but which we are doomed to return to. But body weight or body fat is not an “internally” regulated physiological variable, despite the efforts of some to tie this to the hormone leptin. Rather, it is the result of a number of interacting systems, which frequently lead to a relatively stable output. I’ve provided a more detailed critique of the set point concept in my post, Change your receptors, change your set point.
On the contrary, when you consider the whole organism, you are struck more by its variability over time than by its constancy. Sterling, Eyer and McEwen have contrasted the stability of homeostasis with what they call “allostasis” or “stability through change”. “Stability” here does not mean a static state, but rather a dynamic physiological process which allows the organism to sustain itself in the face of external challenges. For example, hormones like cortisol, adrenalin and catecholamines, and mediators like cytokines, allow us to adapt to changes in activity level. Digestive hormones like insulin and glucagon, and secreted digestive enzymes like proteases, amylases and lipases, allow us to effectively respond to the sudden ingestion of food, otherwise known as “meals”. On longer time scales, major morphological changes in the overall shape and and size of the body allow animals to handle episodic changes like pregnancy, migration, or hibernation. While organisms and physiology are stable enough to survive, they do not maintain or even strive for a state of constancy.
Allostasis, not homeostasis, better describes how we deal with changing circumstances.
Changes which are beneficial in the short term to handle an external stress, may be harmful or pathological if maintained chronically. So for example, glucocorticoid and catecholamine hormones such as cortisol and adrenaline are helpful, even essential, for gearing the body up to handle acute stress. Without such hormones would we be unable to get up in the morning, much less deal with emergencies. But these same hormones become harmful or deadly when chronically elevated, causing significant damage to the cardiovascular system and neurodegenerative conditions such as depression and memory loss. The “biphasic” effect of cortisol and other arousal hormones and catecholamines is encapsulated by the Yerkes-Dodson Law, illustrated in the figure at the right, which holds that performance increases with physiological or mental arousal, but only up to a point. When levels of arousal become too high, performance decreases.
Similarly, insulin, which is essential for the short term digestion of carbohydrates and protein, and for facilitating tissue growth, can likewise be harmful if elevated chronically, leading to obesity, cardiovascular disease, inflammatory diseases, and possibly cancer. McEwen refers to the elevation of these stress related hormones and effectors as “allostatic load” and their chronic elevation as “allostatic overload”.
Hormones are neither good nor bad in and of themselves. They are helpful at the right time and for the right length of time.
Alternating states and opponent processes. I’ve written about opponent processes as an explanation for psychological adaption in my post on The opponent-process theory of emotion. Here I would like to go further and generalize the opponent process theory to more broadly characterize our adaptive physiology.
Our natural allostatic variability typically manifests itself in an oscillation between two states or “extremes” which alternate or fluctuate over some characteristic interval of time that can range from seconds, to hours, days, months, or years. These two states are often thought of as “high” and “low” levels of some variable hormone, enzyme or effector. But I think they are better considered merely as opposing conditions. That’s because what appears to be “states” are really the results of underlying processes that move the organism in opposite directions — opponent processes. These processes typically come in pairs and act to balance each other, like yin and yang. It is important not to confuse the states and the opponent processes. These alternating states are the resultant outcomes of the opponent processes; the visible “state” reflects the dominant process, but both processes are always in play to greater or lesser extents.
This concept of may be confusing, so here are a few examples of alternating states and associated opponent processes, with widely varying temporal scales. In each case “State A” exists when “Process a” dominates over “Process b”, and “State B” exists when process b dominates:
State A State B Process a Process b Frequency
Eating Fasting Anabolism Catabolism 3-24 hrs
Waking Sleeping “C” process “S” process 24 hrs
Exercise Rest Sympathetic Parasympathetic varies
Eating and fasting. You could attempt to characterize the A and B states as “active vs. passive”, “stressful vs. restful” or “bad vs. good” but that is not quite right. Take eating and fasting, for example. You might argue that eating is the active or stressful state, because it places a demand on the digestive system, and the fasting period between meals allows the digestive system to recover. However, if the fast is continued beyond a certain point, it becomes the stressor. After about 12 hours, the stress of fasting causes a rise in catabolic “breakdown” processes, upregulates the neuroprotective hormone BDNF, and the process of autophagy activates the breakdown of intracellular materials to fuel the mitochondria. Utilized in moderation, the “stress” of fasting thereby activates beneficial processes that protect and defend us. Once you resume eating, the “stress” of fasting is relieved and the anabolic “building” process kick in with the rise of insulin. This has its own benefits, in repair and growth. It is important to note that the anabolic hormones like insulin and the catabolic ones like glucagon or adrenaline are always present at some level; they never “go to zero”. Yet one or the other is dominant at a given time, depending on the state of digestion.
Wake and sleep. Similarly, you could say that wakefulness is active and stressful, whereas sleep is passive and restorative. But again, this would be misleading. Wakefulness and sleep are the outcome of a dynamic, alternating balance between two essential processes, the “C process” and the “S process”. The “C process” generates a wakeful state based upon activation of the ascending arousal system, including cholinergic, noradrenergic, serotoninergic, dopaminergic, and histaminergic neurons located in the hypothalamus and other brain nuclei. These neurons release corticotropin-releasing factor (CRF), ACTH, and cortiosol on a regular diurnal cycle. This arousal system interacts with inhibitory “sleep-active” neurons in the ventrolateral preoptic nucleus (VLPO), releasing GABA and other sleep-inducing neurotransmitters. These sleep promoting neutrons and neurotransmitters represent the “S” process. The result is a “flip-flop switch” producing distinct sleep and wake states with abrupt transitions. The “C” and “S” processes each never actually stop, but they continuously wax and wane, with one of the two becoming dominant and leading to either wakefulness and sleep. Even within the states of wakefulness and sleep there are many regular oscillating subcycles; for example REM sleep, deep sleep and light sleep. Disruptions in this process can lead to insomnia, and can be corrected by Sleep Restriction Therapy, as I’ve described in my post A cure for insomnia.
The reality is that for each basic physiological process we need both A and B states and the underlying a and b processes. The opponent processes represent polarities of an indivisible “yin-yang” pair. They balance each other, but not in a constant ratio. The a and b processes cannot be indefinitely sustained, but each have within themselves the seeds of their own demise, by inducing their complementary, inhibitory process. Biological organisms are constructed out of complementary and opposing physiological process, which naturally give rise to an alternation between the A and B states. This is a phenomenon I will refer to as stress oscillation.
Stress oscillation builds dynamic range. So what does allostatis and the opponent processes have to do with hormesis? Sometimes hormesis is thought of unidimensionally: lift weights to build muscle. Fast or reduce carbohydrates to lower insulin and reduce weight.
But in reality, hormesis should be thought of as a binary process of alternating stress and recovery.
Lifting weight builds muscles because it induces “catabolic” microtrauma to the muscles; it is the rest between workouts, in combination with adequate diet, that leads to the “anabolic” rebuilding of the muscle. Both stress and recovery are necessary. For the same reasons, weight loss through insulin lowering should be balanced with sufficient periodic insulin raising to maintain lean body mass, and maintain the healthy function of the insulin producing system, including the pancreatic secretory islets and the insulin receptors in the brain and muscle tissues. One risk of an unremitting “insulin sparing” diet, such as a very low carbohydrate diet without periodic insulinogenesis is the induction of a state of physiological insulin resistance. This is indeed a paradoxical outcome of a diet which many pursue in order to improve their insulin sensitivity!
In the wake-sleep cycle, the ascending arousal system or “C-process” is stimulated by the secretion of CRF (corticopin releasing factor) by the hypothalamic-pituitary-adrenal (HPA) axis. But a state of interminable wakefulness or insomnia results in cognitive deterioration. Both the “C” and “S” processes are necessary, and they must oscillate: An unvarying simultaneous activation of both processes would not lead to cognitive stability, but rather mental deterioration. Stress and renewal must follow one another as night follows day.
For any physiological function like digestion, muscle synthesis, or the wake-sleep cycle, the oscillation between State A and State B produces a dynamic stability that exhibits a certain dynamic range between stress and rest. The cycle of eat-fast-eat leads to a cycling of digestive hormones such as insulin, glucagon, and adrenaline. The cycle of wake-sleep-fast leads to a cycling between the arousal system and the sleep system.
And here is the takeway: By exposing ourselves to alternating A and B states of increasing intensity, we build tolerance and dynamic range for the opponent processes. We should strive to increase the magnitude of contrast between the opponent states. I believe that we can generalize the use of dynamic capacity between allostatic states as a marker of fitness. This can be illustrated by several examples:
Example 1. Aerobic capacity. Exercise phyiologists understand that athletes are able to build aerobic capacity (so-called VO2 max) by exerting themselves at or near maximal heart rate. Their state of fitness is manifest in a reduced resting heart rate or pulse, and a higher ratio between peak VO2 and resting VO2. This ratio or difference is sometimes referred to as VO2 reserve or VO2R, and it represents a good measure of aerobic fitness, a kind of dynamic capacity to oscillate between rest and exertion. Yet another measure of dynamic capacity is the rate at which heart rate or VO2 return to normal, after exertion
What is interesting is that training harder does not necessarily increase VO2R or dynamic capacity. A study by the Navy Seals showed that overtraining can actually decrease VO2R, and can elevate resting heart rate by as much as 10-15 beats per minute. Monitoring your resting heart rate is an excellent way to know if you are overtraining. (Caveat: the heart rate measure must be used with judgement, as severe overtraining can lead to extreme exhaustion and an abnormally low heart rate).
More generally, high intensity interval training (HIIT), whether it be in the form of weight lifting, sprinting, or other metabolic training, is based on the very same premise. Maximal exertion, into the anaerobic range, activates the full range of muscle fibers, including the ever-important fast-twitch muscle fibers, empties muscle glycogen, and activates the glycolytic pathway, resulting in an upregulation of insulin receptors (GLUT4 transporters), and improved insulin sensitivity. But for HIIT to work effectively, it is equally important to allow adequate time for rest and recovery. (I’ve discussed this in more detail on the Fitness page of this blog, with particular emphasis on the physiological analysis of Doug McGuff in his book, Body by Science).
For sports as varied as running and weight lifting, the well known principle of periodization recognizes the importance of variation in intensity and proper rest. In short, both high intensity training and aerobic training, if carried out with adequate rest and recovery, build dynamic range.
Example 2. Digestive or metabolic fitness can be measured by a low basal insulin level in combination with a pattern of sharp, but brief insulin secretion in response to ingested carbohydrates or insulinogenic protein. Low basal insulin level is seen, for example in non-industrialized populations such as the Kitavins, whose average basal insulin levels of about 4 mIU/ml are about half those of Western populations. And yet the Kitavans consume meals with a high percentage of carbohydrates and have good insulin sensitivity. So low basal insulin levels alone are not the whole story. The optimal pattern seems to involve an alternation between feast and fast, allowing the digestive hormones and enzymes to cycle between anabolic (insulin) and catabolic (glucagon, adrenaline, and cortisol).
This is also the premise behind the concept of intermittent fasting. By training yourself to cut out snacks and go for longer periods of time between meals, the metabolic system — which includes not only digestive hormones and enzymes, but neurotransmitters and hypothalamic receptors — adapts to increase its dynamic capacity. The resulting benefits are lower basal levels of anabolic hormones like insulin and catabolic hormones like glucagon and adrenaline. But just as importantly, intermittent fasting develops improved sensitivity and the ability to both ramp up and reduce these hormones quickly and responsively.
The benefits of spending time in the fasting state are numerous, including a natural detoxification and nutrient recycling process known as autophagy, and the upregulation of brain-protective growth factors such as Brain-Derived Neurotrophic Factor (BDNF). Fasting allows for the upregulation of fat-liberating enzymes and hormones and a significant and glucose transporters, thereby improving insulin sensitivity. McEwen has compiled research showing that an appropriate level of “stress” or allostatic load will increase markers of brain plasticity. By contrast, following the conventional wisdom to eat six small meals a day of controlled glycemic foods, in the misguided attempt to “regulate” blood glucose at a constant level, deprives your body of these important restorative and protective processes.
But at the other extreme, extensive fasting or strict low carbohydrate dieting can leave the pancreas underutilized and thereby lead to a reduction in glucose transporters in the cells, since these are no longer “demanded”. Our cells and organs tend to “economize” by synthesizing only the machinery they really need: use it or lose it. People who abstain from or never consume milk will lose the ability to produce the enzyme lactase, so they become lactose intolerant. Similarly, we need to regularly “exercise” our ability to secrete insulin on demand and the ability of cells to utilize glucose. This doesn’t necessarily have to occur every day, but several glucose loadings a week are probably necessary.
So the wise course is to apply “stress oscillation” to diet, and alternate judiciously between fasting and nutritious, balanced meals with a variety of macronutients and micronutrients. Remember that the “stress” is binary: fasting represents recovery from the “stress” of eating; and eating relieves the “stress” of fasting. A dynamic approach of hormesis involves stretching the ability to move between these two poles, increasing “allostatic capacity”.
Example 3. Stress, health and cortisol. Of all the hormones, cortisol has acquired a reputation as “the bad guy”. It is well known that elevated cortisol levels are the mark of chronic stress and adrenal fatigue. It has been suggested that higher levels of cortisol are linked to disregulated or high blood glucose levels and predispose one to diabetes. Chronically elevated cortisol also damages neurons in the hippocampus, leading to memory loss and cognitive decline. As a result, some practitioners mistakenly advise trying to minimize stress and even eat frequent meals, in order to keep cortisol at bay and avoid “stressing” the adrenal glands. But this is a one-sided perspetive. Cortisol is necessary to normal alertness and mental function, as well as our ability to respond to sudden demands like exercise or threats. The problem comes when cortisol does not exhibit a normal morning peak level, followed by a steady decline through the day, but instead remains flat or even increases in the evening. Chinook et al. classified four different cortisol patterns, shown below. Pattern 1 (Graph A) is normal; Patterns 2, 3 and 4 show the flattening or later peaks that characterize dysregulation:
Diurnal or event-related elevations in cortisol are not problematic, so long as cortisol levels return to baseline at a decent rate, as in Pattern 1. According to Lovell et al., higher percieved stress levels are reflected not so much in average cortisol levels, but rather as higher basal or evening cortisol levels, and flatter diurnal fluctuations in cortisol levels. Matthews et al found that individuals with the flattest cortisol pattern (slowest rate of decline to baseline) were most at risk of coronary calcification. Sephton et al found that flatter cortisol patterns were predictive of suppressed immunity and lower survival rates in women with metastatic breast cancer.
In short, we should be less concerned with absolute cortisol levels, than with the pattern of cortisol secretion. As with other hormones, increased dynamic range and a robust cyclical pattern are indicative of fitness, stress-hardiness, and health.
The larger lesson. James Loehr (about whom I wrote in my earlier post on Stress management and toughness training) has written eloquently about the use of “stress oscillation” to build athletic capicity and resilience in the corporate world in his book The Power of Full Engagement:
Balancing stress and recovery is critical not just in competitive sports, but also in managing energy in all facets of our lives. When we expend energy, we draw down our reservoir. When we recover energy, we fill it back up. Too much energy expenditure without sufficient recovery eventually leads to burnout and breakdown…Too much recovery without sufficient stress leads to atrophy and weakness….Oscillation occurs even at the most basic levels of our being. Healthy patterns of activity and rest lie at the heart of our capacity for full engagement, maximum performance, and sustained health. Linearity, by contrast, ultimately leads to dysfunction and death. (TPOFE, pp. 29-31).
How to apply stress oscillation to your life. Let’s return to the question at the beginning of this post: How much of any kind of stress is enough, but not too much, to generate a hormetic benefit? The answer is: This is the wrong question! You should not be striving for some magic optimum level of constant stress. Rather, you should strive to oscillate stress, by exposing yourself to intermittent, but intense sources of stress. This builds dynamic capacity or strength. The amount and frequency of the stress are variables you can experiment with, but younow have a way to measure the benefit and know whether you are on track. The key metric is dynamic capacity. The appropriate measures of dynamic capacity depend upon what our goals are:
- For physical fitness: a high VO2 max during exertion combined with a low resting VO2, resting pulse, and blood pressure.
- For dietary or metabolic health: a rapid insulin and blood glucose response to food and low basal insulin and blood glucose levels
- For stress hardiness: peak cortisol levels upon waking, followed by steady decline to low evening (basal) levels.
These may be imperfect measures, and they are subject to exceptions and interpretations based upon special health circumstances. Some of these measures are easy to implement at home; others are less convenient because they require blood or saliva analysis (which can be purchased online). But the general principle is valid: Don’t look for average biometric values, but look for the dynamic range between high and low. And look for an oscillatory pattern that demonstrates periods of testing and building capacity, alternating with periods of rest and recovery. I’ve discussed only three applications here in detail: digestion, exercise, and general stress tolerance. But the principle of stress oscillation can be applied to many other applications of hormesis: suntanning, allergen immunotherapy, cold showers or plus lens therapy. I leave it to the curious reader to think about the physiological processes at work, and the appropriate measures of improved dynamic capacity.
The goal of hormetic stress should be to increase dynamic capacity to handle allostatic load — variable stresses — in a measureable way. The precise level and frequency of stress exposure will vary from person to person. This is not a one-size-fits all path to health, but rather a journey that each of us must take for ourselves. But on this journey, our engine is stress oscillation and our compass is increased dynamic capacity.