Table of Contents
The relationship between our diet and sleep is bi-directional; i.e. sleep impacts diet and diet impacts sleep. Therefore, we can examine the impact of sleep timing, duration and other dimensions on our dietary intake. And then also examine the impat of both overall diet and specific nutrients on improving/worsening sleep.
The is clear evidence of distinct, acute effects of restricted sleep time on food preferences, eating behaviour, energy intake, and our underlying metabolic physiology.
When it comes to the ability of certain foods or nutrients to improve sleep, often many claims are based on weak evidence or mechanistic reasoning. But there is evidence showing some impacts of certain compounds to either positively or negatively impact sleep.
So what is the accurate way to look at this bi-directional relationship? In this episode, Greg Potter, PhD discusses the evidence to date. Dr. Potter received his PhD from the University of Leeds, where his research focused on circadian rhythms, sleep, nutrition, and metabolism.
Greg Potter, PhD
Greg Potter, PhD completing his doctoral work at the University of Leeds, where his research focused on circadian rhythms, sleep, nutrition, and metabolism. He has consulted with a variety of athletes, organizations and member of the general population on how to improve their health and performance. He is the co-founder of Resilient Nutrition and has in the past created content for HumanOS.
- 03:15 – Sleep architecture and dimensions of sleep
- 10:29 – Influence of sleep on diet
- 35:11 – Chronotypes
- 53:26 – Impact of diet/meals on sleep
- 59:50 – Supplements like melatonin and tryptophan
- 1:20:27 – Rescuing a poor night’s sleep – caffeine and nootropics
- 1:40:31 – Key Ideas segment (Premium only)
Links & Resources
- Sigma Statement: How Sleep & Nutrition Interact
- Website: gregpotterphd.com
- Instagram: @gregpotterphd
- Glossary of Standardized Terminology for Sleep/Biological Rhythm Research
- The STOP-Bang Questionnaire for Sleep Apnea
- 80000hours.org – Your Job Can Make a Difference
Danny’s Key Ideas from this episode are:
- “Good sleep is not one thing” – Looking at many dimensions of sleep
- Diet/nutrient recommendations to improve sleep are very context specific
#1: “Good sleep is not one thing”
Sometime we can think in binary terms about sleep. Our sleep is good or it’s bad. Or even when we do actually distinguish between sleep duration and sleep quality, sometimes we don’t acknowledge that they influence one another. And quality may be an emergent property of enough duration, or appropriate sleep time may be a result of quality being good.
But even within sleep quality, as Greg highlighted in this discussion, there are many dimensions that all interact, and so need to be viewed collectively, rather than in isolation. Namely, we should consider:
- Duration – There are two slightly different measures:
- Time actually asleep
- Time in bed
- Timing – i.e. the time of going to sleep (and waking up)
- Sleep regularity – of those times of sleep/waking, how consistent and regular is that?
- Irregular sleep times can be problematic, e.g. social jetlag
- Sleep latency – defined as the time (in minutes) from ‘lights out’ until the first epoch of any stage of sleep (Berry, 2012)
- epoch = a short interval of arbitrarily defined length; usually 20-60 seconds
- Sleep efficiency – the percentage of time spent asleep while in bed.
- Sleep architecture – the basic structural organization of normal sleep. There are two types of sleep, non-rapid eye-movement (NREM) sleep and rapid eye-movement (REM) sleep. NREM sleep is divided into stages 1, 2, 3, and 4, representing a continuum of relative depth (Institute of Medicine Committee on Sleep Medicine and Research, 2006)
- Sleep-disordered breathing – a condition of repeated episodes of apnea and hypopnea during sleep
#2: Eating to improve sleep is very context specific
If a nutrient or supplement could possibly improve sleep, then it’s likely it will only do so in a specific context (if at all). There are a number of questions to ask that would help uncover if the end result would indeed be improved sleep, for example:
- What is the baseline level of intake of a micronutrient? Is someone deficient in this nutrient?
- How does consuming this food/nutrient impact overall intake?
- Why does someone currently have poor sleep?
In addition to this it’s crucial to remember that diet/supplementation is a distant second to sleep hygiene when it comes to potentially improving sleep. Some of Greg’s practical tips are listed at the bottom of the detailed study notes.
Detailed Study Notes
Overview of Sleep
There are two primary processes that regulate sleep (Scheer & Shea, 2009, pg. 199):
- Sleep pressure (homeostatic sleep drive)
- Wake drive (circadian waking drive)
As shown in the diagram below, sleep pressure accumulates across the day and drops throughout a bout of sleep. Greg mentioned in the episode that more sleep pressure tends to lead to more slow-wave (deep) processes during sleep.
The second process in this ‘two-process’ model is often referred to as ‘wake drive’ and is, as the name suggests, wake-promoting (and therefore the opposing effect to sleep pressure). This waking drive is regulated by the circadian system.
- Circadian rhythms are internally driven cycles of biochemical, physiological, and behavioral processes that fluctuate across a 24-hour repeating cycle. [For more discussion on circadian biology, check out episode 296]
This wake drive begins to decline at night, helping to promote sleep onset. An appropriate amount of sleep will reduce sleep pressure as shown above, in addition to the circadian wake drive increasing, and the cycle starts over.
Sleep Architecture: the basic structural organization of normal sleep.
The time spent in bed (see ‘bedrest’ in diagram below) is made up of:
- Time before falling asleep
- The sleep episode
- The “final waking” between the end of the sleep episode and full waking
Sleep latency: the amount of time (in minutes) from ‘lights out’ until the first epoch of any stage of sleep (Berry, 2012)
- Note: epoch = a short interval of arbitrarily defined length; usually 20-60 seconds.
A sleep episode is made up of several individual sleep cycles, which are roughly 90 minutes in duration. Within each individual sleep cycle we move through a number of sleep stages (and back again).
There are two types of sleep, non-rapid eye-movement (NREM) sleep and rapid eye-movement (REM) sleep. NREM sleep is divided into stages 1, 2, 3, and 4, representing a continuum of relative depth (Institute of Medicine Committee on Sleep Medicine and Research, 2006)
- Stage 1 is the lightest phase of sleep.
- Stage 2 is deeper and accounts for the majority of total sleep time.
- Stages 3 and 4 are known as “slow-wave sleep” (SWS), which is a deep phase of sleep associated with memory consolidation.
- REM sleep is a temporary state of bodily paralysis, which is believed to reflect the dream state.
As noted earlier, one sleep cycle sees us progress throught the stages (and back again) in around 90 minutes. And this cycle then gets repeated multiple times over the course of the night (sleep episode).
The Sleep-Diet Bidirectional Relationship
Sleep and our diet can impact one another. Sleep (and the different dimensions of sleep) can have impacts on our eating behaviour and metabolic physiology. While our dietary intake may potentially impact sleep.
How Sleep Can Impact Nutrition
Impact of restricted sleep on food choices and intake
University of Chicago RCT (Tasali et al., 2022)
- RCT with 80 adults young, overweight adults who habitually slept fewer than 6.5 hours a night
- After a personalized sleep hygiene counseling session, participants were able to increase their sleep duration by an average of 1.2 hours per night.
- The increased sleep duration reduced participants’ average caloric intake by ~270 kcal/d, compared to controls
Work by Dr. Marie-Pierre St-Onge has suggested men may respond worse to sleep loss (in terms of subsequent energy intake) than women:
Greg discussed how the brain is the important player in regulating intake:
- Whilst there are studies showing impacts of sleep on leptin (a hormone secreted by adipose tissue) and ghrelin (an appetite hormone), the evidence doesn’t reliably and consistently show such changes to a significant degree
- However, brain imaging of changes after sleep loss this seems to predict intake
Indeed neuroimaging studies have revealed that sleep curtailment results in:
- exaggerated food-reward
- increased appetite for high-calorie foods
- compromised inhibitory control of food intake
In addition to an increased drive to consume food, sleep curtailment results in reduced dietary restraint and increased disinhibited eating.
Sleep Fragmentation Impacts BMI Sawamoto et al., (2014):
- 90 people underwent a 7-month weight-loss phase
- On average BMI was reduced by 13.6%, but differences emerge when looking at the number of wake episodes (WEs) per night
- Participants with 5 or more WEs per night had a significantly lower reduction in BMI compared with those with fewer than five, after adjusting for confounding variables.
- Chaput & Tremblay, 2012 – A pooled analysis of three studies examining hypocaloric diets over 15-24 weeks found that increasing sleep duration by one hour, from a baseline of seven hours, was associated with a 0.7kg decrease in fat mass.
- Spiegel et al., 2004 – Extending sleep duration from 6.5 hours to 8.5 hours for two weeks led to significant improvements in appetite regulation, and decreased desire for hyperpalatable foods.
Potential problem with too much time in bed?: If someone is experiencing difficultly with sleeping when giving themselves a long period of time in bed (e.g. 10 hours), they may paradoxically improve sleep quality by reducing time in bed. Greg termed this “matching your time in bed to your sleep capacity”. So only spending the amount of time in bed as you can actually sleep.
Impacts on body composition
- Participants consumed a calorie intake of 90% of their resting metabolic rate
- Each participant spent 14 days in each of two conditions (i.e. crossover study):
- 5.5 hours or 8.5 hours of time in bed per night
- In both conditions, the participants lost the same absolute amount of weight (3kg)
- However in the 5.5 hours sleep condition, particpants lost 55% less fat mass and lost 60% of lean mass.
- Using measures of carbohydrate and fat oxidation (respiratory quotient), the investigators showed that in the 5.5 hour condition, there was impaired fat oxidation.
Sleep restriction study by Nicholas Saner and colleagues (2020):
- Participants underwent a sleep restriction protocol: 5 nights of 4 hours in bed each night.
- These participants had lower rates of muscle protein synthesis (MPS).
- However, by performing high-intensity interval exercise during the period of sleep restriciton, MPS rates could be maintained at the same levels as controls.
Greg also mentioned the work of Pål Jåbekk and colleagues (2020), who published a pilot trial looking at the interaction of sleep and resistance training:
- 23 untrained healthy men randomly assigned to:
- exercise and sleep optimization (n=10)
- exercise only ( =12).
- Both groups performed a whole-body resistance exercise program twice a week for 10 weeks.
- The “sleep optimization” group received education on how to improve both sleep quantity and quality.
- After 10 weeks both groups had increased lean body mass by a similar amount. The ExS group experienced an increase of 1.7±1.1 kg while the Ex group experienced an increase of 1.3±0.8 kg (P=0.29 for difference between groups).
- However, the men in the exercise and sleep optimization group lost body fat (-1.8±0.8 kg). While the exercise-only group did not (non-significant increase of 0.8±1.0 kg).
Most of the studies on sleep’s impact on diet and/or body mass tend to focus on sleep duration. However, differences in chronotype seem to be associated with a variety of outcomes, independent of sleep duration.
Chronotype: the behavioural preference for going to bed (and rising) at a particular time (earlier vs. later) There is a spectrum of chronotype with the two ends of the spectrum relating to preference for ‘morningness’ or ‘eveningness’, which reflects the timing of their internal circadian rhythms.
- Eveningness = delayed sleep period; most active and alert in the evening)
- Morningness = advanced sleep period; most active and alert in the morning
The below graph shows the typical distribution of chronotype across a population. The time on the x-axis relates to the mid-point of sleep. The height of each bar (y-axis) indicates percentage of people with that chronotype. So while there is wide variation and extreme chronotypes, most people have a relatively “normal” chronotype, with mid-point of sleep falling between 3am and 6am.
A later mid-point of sleep is used as an indication of the relationship between chronotype and sleep timing. And a later mid-point is associated with increased energy intake in persons with obesity, and impaired long-term blood glucose regulation in persons with T2DM.
So independent of sleep duration, a late/evening chronotype is associated with:
- increased risk for type-2 diabetes
- lower overall diet quality
- redistribution of energy and macronutrient intake to later in the day
- “social jetlag” (associated with increased metabolic disease risk)
External factors can cause “chronodisruption”, which can lead to jet-lag or “social jet lag”. Such factors include:
- Work schedules
- Lack of getting outdoors (in the US people may spend 87% of their time in enclosed buildings)
- Daylight Savings
- Time-zone travel
How Nutrient Intake Impacts Sleep
Gill & Panda (2015) showed that when overweight individuals with an eating window of more than 14 hours each day, were restricted to an eating window of 10-11 hours daily for 16 weeks, it improved their improved sleep.
However, a review of human trials by McStay et al., (2021) found that, on the totality of the human trials using an intermittent fasting or time-restricted eating protocol, there is little to no effect on sleep quality or other dimensions of sleep.
Do high-carb meals help?
A common recommendation that gets echoed is that a high-carb meal at the end of the day will enhance sleep. But is this actually rooted in strong evidence?
- A trial by Afaghi et al. (2007) found that consumption of a high glycemic meal (GI = 109) of Jasmine rice 4 hours before bed resulted in a shorter sleep latency (time to fall asleep), compared to either a low GI (50) meal or having the high GI meal 1 hour before bed.
- A study of 2 crossover trials from Christoforos Giannaki’s lab (Vlahoyiannis et al., 2018) had 10 recreationally trained male volunteers perform sprint interval training in the evening. After training, the men consumed either a high-GI or low-GI meal. Total sleep time and sleep efficiency were greater in the high-GI trial compared to the low-GI trial, while sleep latency was reduced four-fold.
- A review by St-Onge et al. (2016) found associations between high-carbohydrate diets and: increased REM sleep, reduced sleep latency, and less slow-wave/light sleep. However, they note that longer-term effects have not been examined in RCTs.
Role of Specific Nutrients and/or Supplementation
Greg suggested that commonly recommended supplements such as magnesium, glycine and valerian root have very weak evidence for improving sleep. Those with better evidence are:
Melatonin supplementation in the range of 1-5mg has been shown to effectively shift circadian phase, correct in cases where there is circadian misalignment (e.g. due to jet-lag or drastic change in sleep/wake times). In these specific instances is can reduce sleep onset latency and increase total sleep time. However, in situations where this is no correction of circadian phase needed, melatonin is unlikely to reliably improve sleep.
Certain supplements may be able to “rescue” physical and/or cognitive performance that has been lost due to acute sleep curtailment (e.g. 1-2 nights of poor sleep). The most common compound considered in this situation is caffeine. A systematic review by Pomeroy et al. (2020) suggested that tyrosine or caffeine could be used in a military context to enhance cognitive performance when personnel are sleep-deprived. Similarly, a meta-analysis by Irwin et al. (2020) found that caffeine supplementation can be beneficial across a whole range of physical and cognitive performance outcomes, following sleep restriction or deprivation.
Greg highlighted that a less obvious compound that could be beneficial is creatine. In an RCT of sleep deprivation for 18, 24, and 36 hours (McMorris et al., 2007), consuming 20 g/d of creatine for a week before the trial was able to improve central executive functioning, relative to those who were not supplemented. Cook et al. (2011) examined the impact of creatine supplementation on a physical skill (rugby passing) following sleep restriction (3-5 hours). While the sleep restriction alone negatively impacted performance, when creatine was supplemented at 50 or 100 mg/kg, no deficit was observed.
In relation to tart cherry juice concentrate, Gao & Chilibeck (2020) published a meta-analysis showing potential benefits for endurance perfomance. Glyn Howatson has done a lot of work on both exercise and sleep, with one of his papers (Hill et al., 2021) collating all the findings on recovery from exercise, then there were quite large effects of tart cherry juice intake on reducing muscle soreness, and improving the recovery of muscle strength and power.
Endotypes and Phenotypes
Consider why there is a sleep issue. Sleep disorders can be considered from both the standpoint of underlying mechanisms (endotypes) and in its clinical expression (phenotypes). Therefore “bad sleep” can have many different phenotypes (e.g. difficulty falling asleep, unable to get back to sleep, breathing issues during sleep, etc.), and any one of these phenotypes can have different root causes (endotypes). For example, if the underlying mechanism of a particular sleep issue is anxiety, then is sleep likely to be improved through supplementation?
Pragmatic Tips for Better Sleep
Greg suggested the following:
- Match time in bed to your sleep capacity
- If you had no obligations or external factors impacting wake/sleep time, when and for how long would you likely sleep? Try to set time in bed to match that.
- Spend as much time outdoors as possible/practical, within the first 2 hours of waking
- Acts as a circadian regulator and will help you fall asleep quicker and more easily that night
- Avoid using screens/devices in the 30 minutes before bed
- Stimulus control of behaviour: Our brain is good at creating associations. If you spend lots of time in bed not sleeping, then your brain can associate those. So only go to bed when actually sleepy, if you can’t sleep at all then get out of bed and go somewhere else and do something non-stimulating until sleepy again.
- Any stress reduction interventions are important. Even simply writing a to-do for tomorrow before going to bed.
- Have set times of the day when you do zero work (especially remote workers who can check email and other work tasks throughout the day.
- Obstructive sleep apnoea