Food as zeitgeber

This article by Dr Piotr Wozniak is part of SuperMemo Guru series on memory, learning, creativity, and problem solving.

Introduction

Food and the expectation of food are both zeitgebers that may be used in the stabilization of the circadian cycle or in phase shifts in chronotherapy. However, the understanding of food as zeitgeber belongs to a relatively new area of science, and there are many things that need to be elucidated before feeding can be efficiently employed in support of healthy sleep.

Advice from experts is often contradictory. This text attempts to answer the following questions: What is the circadian impact of breakfast? Is it about food as zeitgeber, or more about food anticipation, or the ancestral food hunt, or food reward? What is the circadian impact of evening fasting? Or late meals? What timing is optimum?

Suggestions compiled for Curing DSPS and insomnia and Optimum diet indicate that there are many competing forces that need to be well understood and reconciled in order to optimize feeding and to prevent sleep phase shift.

What makes things worse, the advice will differ depending on the goals, and depending on the source. Is it about weight loss, weight gains, overall health, morning productivity, high quality of sleep, etc. Those suggestions must be filtered out to understand the value of food as zeitgeber.

Meal timing

The main circadian oscillator of the suprachiasmatic nucleus (SCN) is primarily driven by light-and-dark cycles. However, food can impact peripheral oscillators, incl. liver clock genes. The main oscillator is considered a master clock, but there are indications that resetting the food entrainable oscillator (FEO) may have some impact on the phase shift in the master clock.

The secondary and transient nature of the FEO can be shown in experiments in which food anticipatory activity is abolished within days upon switching to ad libitum feeding. This implies that a form of restricted feeding is necessary to achieve an entrainment. It is unclear if time-restricted feeding will have the same effect if it follows natural circadian feeding cycle. It is even less clear, if the entrainment can have a significant impact on the main oscillator of the circadian cycle.

Decoupling of the master clock and the FEO occurs in shift workers, which is known to have negative health effects. This suggests that the FEO should ideally be in synchrony with the master clock.

The phase response curve needs to be considered to understand the impact of meals on the sleep phase. By the way circadian oscillators operate, we can conclude that only morning and evening meals or fasting will have a significant impact on the phase. However, it is conceivable that meals may also stabilize the cycle by means other than cycle phase shifts (e.g. by a mere effect on the amplitude of the oscillator).

Additionally, mid-day meals may indirectly affect the phase by their impact on the siesta. If a meal increases the depth of sleep in a late siesta (e.g. Phase 10 or later), it is likely to result in a lesser homeostatic sleepiness in the evening. This in turn might result in phase delays.

The problem of food as zeitgeber primarily boils down to the following questions:

How does breakfast affect the sleep phase?
How does evening fasting affect the sleep phase?
How do late meals affect the sleep phase?

In 1996, Challet observed, and in 2005, Mendoza confirmed that when rats are fed on a regular schedule in constant darkness, locomotor entrainment was achieved. This indicates that in the absence of strong light zeitgeber (e.g. in winter for humans), meals or fasting may assume a larger role. As indicated elsewhere in this text, wakefulness itself may be a zeitgeber that relies on neurotransmitters that promote alertness. Similarly, Mendoza hypothesized that motivation and arousal may form the basis of entrainment.

Synchronization dilemmas

If the master clock is hard to manipulate with peripheral clocks, a major question arises: should we use peripheral clocks to induce phase shifts, or should we rather let peripheral clocks synchronize spontaneously. Spontaneous synchronization seems like a good way to achieve a robust circadian cycle, which may show more desirable properties in response to all healthy zeitgebers.

Moreover, some experiments indicate that clock synchronization may additionally be a subject of mutual stabilization with mechanisms other than phase shift. This in theory might mean that a well-timed meal may have a synchronizing power even if taken in the middle of a subjective day (beyond the range of a PRC).

Neuroscientists with a good understanding of control theory noted that the SCN does not comply with the predictions of limit cycle theory (as it is the case in primitive animals such as Neurospora). The explanation seems to be that in a poorly synchronized neural network, population phase shifts will be a resultant of individual cycle shifts for separate neurons. As a result, well-synchronized networks will respond better to zeitgebers. The power of the synchronized network can be demonstrated in computer simulations (source). This reasoning emphasizes the utmost importance of non-intervention. In modern lifestyle, not only do we need to protect the brain from artificial scheduling. We also need to shelter it from fake zeitgebers (e.g. electronic devices in the evening).

In my own life, for decades, I would try to protect the circadian cycle from unhealthy external interference. I paid far less attention to food as zeitgeber. Meal timing would be largely spontaneous with pretty regular natural timing. Meal composition would be dictated by the rules of healthy eating. In other words, I rarely tried and never observed easily detectable impact of healthy and naturally timed meals on the sleep phase. All major influences could be summarized as "negative impact of badly timed meals on the quality of sleep" (e.g. when being late with meals for one reason or another). I may then be biased to think that food should rather not be used as a phase-shifting tool in chronotherapy. This may quarrel with some expert advice and some research. This text attempts to show all components of the equation.

Food may play a role of either (1) a zeitgeber or (2) a phase stabilizer. The two may imply different meal timing strategies

Breakfast

Dr Clifford Saper postulates that breakfast can be a powerful zeitgeber (source). However, his reasoning on the role of the DMH in circadian control was undermined by the findings of Mistlberger who noticed the persistence of FAA in DMH-lesioned animals (source).

The old dogma says: "Eating breakfast is healthy". However, there is little research that would explain or prove the value of breakfast. Most of the reasoning about breakfast comes from epidemiological studies based on correlations that fail to account for reasons people skip or delay breakfast. Those who do not eat breakfast are more likely to be in high stress in the morning (e.g. due to school or employment), unmarried, waking up with an alarm clock, smokers, etc. They are also less physically active, and may consume more alcohol (source). Dr Jason Fung is one of those who disagree with the dogma (see more: Optimum diet). Self-experimenting insomniac, Seth Roberts would also recommend skipping breakfast as a form of circadian control.

If breakfast is not taken as dogmatically good, we can possibly use it as a tool in chronotherapy.

The impact of breakfast has two prime components: (1) the impact of the meal itself, and (2) the impact of the meal on the levels of cortisol and alertness.

There is very little research on how sleep and/or wakefulness are zeitgebers on their own independent of light, food, social cues, etc. There is one artificial zeitgeber that capitalizes on the power of wakefulness: the alarm clock. In a similar fashion, insomnia can contribute to phase delays. This is why perfect timing of bedtime is so essential in DSPS. A great deal of people destroy their morning sleep with the alarm clock. They are then exposed to artificial light or social interaction that make the analysis a bit complex. However, insomnia is a form of suffering in darkness and isolation. If there are any social cues involved, these must be the problems of the day circulating in the mind. The ability to adapt the cycle to the customary period of wakefulness seems like an important physiological adaptation to life on earth when dark-and-light cycles cannot provide a full set of cues on the optimum rest-vs-activity allocations. The concept of wakefulness as zeitgeber is the basis of the recursive phase response curve idea:

Figure: Recursive phase response curve (rPRC): In free running sleep, sleep episodes are good markers of circadian sleep propensity. Conversely, all forms of artificial intervention that affect the position of sleep block can affect the circadian cycle and the sleep phase. We can iteratively compute the position of the subjective night and the impact of wakefulness on the phase shift. The graph demonstrates the PRC for wakefulness as zeitgeber. In this graph, delaying sleep by four hours results in a shift of sleep phase equal to 1.4 hours (which seems to be close the maximum shift possible). Phase advance would require a natural onset of sleep that preceded the optimum retirement time by as much as 6 hours. Going to sleep at the optimum hour results in the natural daily delay, in this particular case 1.0 hour, typical of DSPS disorders or conditions of isolation from natural zeitgebers (e.g. due to constant lighting)

Using the above reasoning, and my own experience (I do not eat breakfast these days), I may be tempted to believe that skipping breakfast may also be a form of zeitgeber based on the impact of cortisol and glucagon on the level of alertness. We know that animals adapt their sleep to the timing of feeding. However, they should also be able to adapt their sleep to the period in which they are most fit to search for food. A major problem with that picture is that cortisol itself is a product of the circadian cycle. It is released as an natural alarm clock. Along the principles of the control theory, the hormone itself cannot therefore be efficiently used as a zeitgeber even though we know it is able to affect the phase of peripheral clocks. Many neurohormonal responses in the brain are also a byproduct of waking up. Without more data and more research, it is hard to settle the question of the mechanism by which breakfast or morning fast affect the circadian cycle.

As we know that peripheral and central oscillators can be uncoupled (source), playing with natural feeding rhythms may have an effect on peripheral clocks (e.g. liver clock), but leave the SCN rhythm behind. If we induce a phase shift between the two, we may achieve two effects that differ wildly on desirability: (1) peripheral clocks will have a minor and positive impact on stabilizing the circadian cycle or (2) minor asynchrony will lead to minor and yet negative metabolic side effects.

Due to the complexity of the mechanism and scarcity of consistent models, it is difficult to recommend the use of breakfast as a zeitgeber. In 2020, the verdict seems to be:

Timing and composition of breakfast should best be dictated by natural appetites

Late meals

As discussed in Optimum diet, meals that come very late, interfere with the circadian gut rest. As such they might simply be unhealthy, and unwelcome. Their impact on the sleep phase will probably depend on the negative digestive side effect. Some meals might delay sleep, others might cause premature waking. This will change their phase shifting power and direction. The effect may be hard to predict, but it is unlikely that the undesired impact on health could be compensated by minor changes in the sleep phase.

If hunger pangs were to cause insomnia, late meals make sense if they are easily tolerated. A proverbial cup of warm milk is fine. Most people have their own preferences and may just consult appropriate sources to see if their preference may have negative side effects.

From the point of view of sleep and phase shift, the optimum timing of the last meal must find the balance between: (1) being early enough to prevent interference with sleep, and (2) being late enough to prevent increased alertness or discomfort and bedtime. The composition of the optimum last meal is one more complex subject to study. However, a simple rule of thumb says that the last meal should: (1) follow basic healthy eating guidelines, and (2) be based on natural appetites (assuming they have not been distorted by years of dieting).

Late meals, within the last two hours before bedtime, are unlikely to be beneficial in chronotherapy

Fasting

Ghrelin have been found to advance sleep phase in vitro in the SCN culture in mice (source). In vivo, it failed to show a similar effect in ad libitum fed animals. However, it seemed to work in time-restricted feeding condition. This may indicate that fasting may facilitate circadian response to feeding-related neuropeptides.

Caldelas and Mendoza found that caloric restriction also affects the SCN making it more sensitive to light signals. Restricted feeding was shown to increase the amplitude of the circadian clock (source).

Intermittent fasting may be beneficial for the amplitude of the circadian clock, and its sensitivity to zeitgebers

Exemplary, time-restricted schedule anchored in the circadian cycle is presented below:

Figure: Example of optimized meal timing inspired by Mediterranean habits. One major meal comes in Phase 8, and directly precedes the siesta. Evening snack comes in Phase 16 (or earlier). In terms of intermittent fasting, in the presented example, the period without meals spans 16.5 hours. Important: each person's circadian cycle may differ in terms of phase, period and amplitude. The timing in the graph is exemplary and cannot be followed without a study of individual sleep data, health, metabolic profile, lifestyle, latitude, climate, exercise regimen, and many more. For example, for people who need more hours for night sleep, circadian phase may be lower. It is more important to make sure that last meal precedes sleep by no less than 2-4 hours. Sleep data derived from SleepChart. The chart is based on: Optimizing the timing of brainwork

Snacking

The effect of snacking may be the opposite of the one described for fasting. If food is to play a role of a zeitgeber and a cycle stabilizer, snacking would likely weaken the signal. We should aim at a high amplitude of the circadian wave, intermittent metabolic switching, and sharp boundaries between creativity, exercise, stress and meals. Snacking is equivalent to flattening of the metabolic cycle. In areas of sleep or creativity, it can be compared to the chaos of polyphasic sleeping, or multitasking based on social media alerts.

For cycle stabilization, well-timed meal routine is superior to a routine with occasional snacks

Conclusions

Food is a zeitgeber, but it is hard to employ food as a significant tool of chronotherapy
Time-restricted feeding may improve the stability, amplitude, and the responsiveness of the circadian clock (see: Optimum diet)
Peripheral clocks seem to play a secondary role to the brain's master clock responsive to light
Natural appetites are likely to lead to natural feeding habits that can stabilize the cycle, e.g. through intermittent metabolic switching
In habit reinforcement, it is important to follow observations from good days (e.g. sleep deprivation may skew the appestat)
Natural habitual circadian meal timing may assist in synchronizing body clocks via food zeitgebers

For healthy sleep, naturally time-restricted feeding schedule seems to work best