Influence of nutrition and food on sleep—is there evidence?

According to molecular research, a key for nutritional influence on sleep, sleep quality, and sleep disorders lies in the production of mediators in adipocytes. Both white and brown fat may play a role in circadian rhythm. Contrary to former beliefs that white and brown adipocytes remain constant for a life span, it is now known that white fat may become brown fat or intermediate beige fat and vice versa [16]. Brown adipocytes are the most productive of exogenous agents. The best known agents secreted by adipocytes are bioactive peptides, the adipokines including leptin, resistin, vaspin, visfatin, hepcidin, adiponectin, and inflammatory cytokines [16]. Many neuroactive peptides are also produced. Machado and colleagues reviewed how active brown adipocytes react to the thermogenesis of warming and how thermogenesis and circadian rhythms are dependent upon each other [16]. There is a bidirectional influence on melatonin through uncoupling protein 1 (UCP1) expressed from adipocyte membranes. Melatonin stimulates browning of adipocytes and production of UCP1. UCP1 mRNA then stimulates melatonin expression and thereby stabilizes and re-enforces circadian rhythm and sleep [17]. If there is a disruption of melatonin stimulation for reasons such as insufficient light or reduced sleep, the browning of white fat is disrupted [18]. This may be one explanation on a molecular basis for weight gain associated with insufficient total sleep time.

Other major players in circadian rhythm and sleep are leptin and adiponectin. White fat and brown fat produce both leptin and adiponectin. These metabolic hormones have a direct influence in the clock nuclei involved in the sleep-wake cycle [19, 20]. Leptin and adiponectin influence the ventrolateral preoptic nuclei (VLPO), a major area for sleep promotion [19]. Adiponectin administration increases wakefulness during the rest phase, reduces delta power, and activates wake-promoting neurons, such as those in the locus coeruleus (LC), tuberomammillary nucleus (TMN), and hypocretin/orexin neurons (OX) within the lateral hypothalamus (LH) and perifornical area (PeF). Conversely, leptin promotes REM and NREM sleep, including an increase of delta power during NREM sleep. Leptin induces c-Fos expression in VLPO and melanin concentrating hormone-expressing neurons (MCH) [19].

Most of this neuronal and mediator research has been performed in animals. It remains to be seen if these findings can be substantiated in human physiology.

There are few nutritional substances that have a direct influence on brown adipocytes in animals and stimulate the production of UCP1. Curcumin is found in the spices saffron, thyme, and cinnamon as well as in quercetin contained in onions. Other substances such as pepper and capsaicin may work only in combination in the presence of other mediators. Research on these spices and chemicals is restricted to knock out mice [21‐23]. The application of these findings to humans remains speculative in view of the large doses of chemicals used in this animal research.

The search for studies on the effect of particular nutritional substances on sleep revealed only two serious publications. The first is a meta-analysis of studies investigating effects of increased protein intake on circadian rhythm and sleep. This study concluded that there is no real effect of increased protein intake on sleep [24]. The only effect in the majority of studies is a subjective improvement of sleep quality demonstrated subjectively by questionnaires.

In a survey of over 2000 people, subjects were asked about their type of breakfast (Japanese style with rice, algae, and proteins, Western style with mainly bread or mainly cereal, and alternating Western and Japanese style) and chronotype (morning person vs evening person). Results showed that a Japanese style breakfast is strongly associated with a morning active chronotype and the other breakfast styles are more associated with evening chronotypes [25]. However, these results are based on subjective answers and one may speculate that persons with a so-called healthy diet and lifestyle consider themselves as morning type compared to people with a less healthy diet and who describe themselves as active evening chronotypes. A direct influence on sleep or sleep-wake cycle by food cannot be concluded from these two reports.

A very recent meta-analysis of articles on food intake analyzed 24 studies on this topic [26]. The meta-analysis concluded that morning chronotype has a more healthy eating behavior than the evening type who eats more fat and carbohydrate-containing food than the morning chronotype. All the analyzed studies were based on subjective questionnaires. There have been no data derived from objective measurement of sleep or activity via actigraphy.

There is a larger number of publications that correlate nutrition in children, adolescents, and sleep parameters. These studies present fairly uniform conclusions.

A review of 24 articles of sleep and nutrition in children revealed that a majority of original studies associated missing nutrients and a high amount of sugar in the diets of obese children with insufficient total sleep time. A minority of studies presented a link between obesity and food with sleep parameters [27].

There has been an increase in the number of studies on lifestyle and sleep behavior in adolescents recently. Late consumption of carbohydrates was associated with shorter sleep in Chinese adolescents though this association was complicated by sleep restriction linked to depression and other mental health problems in young people [28]. However, young Chinese athletes with a healthy lifestyle found that a healthier diet gave them the best results in competition. The sleep of these athletes was less fragmented and more stable with an overall higher sleep quality and less signs of depressive mood [29]. A problem with these two studies is the unclear use of questionnaires and the lack of objective data collection (e.g., polysomnography, polygraphy, actigraphy) and the possible bias towards a push for a so-called healthy diet.

In a sample of over 800 students, data collection was based on validated questionnaires such as the Pittsburg Sleep Quality Index (PSQI) [30]. There was a high odds ratio for low sleep quality in the so-called unhealthy eaters (i.e., those who ate more sweets and meat and less vegetables etc.). However, a majority of young women volunteered for the survey (67%) and they considered themselves to be healthy eaters, leaving a small minority of unhealthy eaters as the focus of the study. There were a number of issues that biased the outcome of the investigation.

A large number of investigations [31‐33] were based upon epidemiologic studies in children, adolescents, and young adults concerning sleep duration, time online. or time on computer gaming and the consumption of fast food or ultra-processed foods. A systematic review listed a total of 61 such studies [34]. Almost all studies concluded that there is a correlation of short sleep due to working late, gaming into the night, increased consumption of ultra-processed food, and a risk for obesity. However, none of these investigations showed direct disruption of sleep or shortening of sleep by intake of ultra-processed food. A direct connection between length of sleep and particular food intake has not been established.

One study enrolled 850 students equally divided between the genders and showed that morning chronotype tended to increase BMI and change eating behavior [35]. However, the study clearly showed that the questionnaire survey results for sleep, chronotype, and eating behavior differ at different time points in adolescents or young adults.

In another study [36], eating behavior and late sleep onset time do not differ much between adolescents and adults.

The effect of parental education on changing consumption in teenagers is moderate according to a recent review of randomized controlled studies on lifestyle behavioral change [37]. Parents were able to achieve a slight increase in physical exercise, shortened screen time, earlier bedtime with longer sleep time, and reduced junk food consumption.

Changing sleep patterns in nursing home residents with dementia via food intake was described in a survey of nursing staff [38]. Nursing staff often encourages demented residents to eat sufficiently during day time and have only a snack for dinner in the belief that this keeps the residents asleep at night. If demented residents nevertheless have a chaotic sleep-wake cycle and get up at night, the nursing staff keeps them awake with coffee or tea in addition to activities in order to get them tired. While this seems a practical approach in elderly demented patients, the authors concluded that the caffeine delivery negatively influences the sleep-wake cycle. There has been no controlled trial to determine if certain food, certain amounts of food, or certain meal times may help persons with dementia to sleep at night.

In a randomized controlled trial [39] with 224 participants, 198 subjects were sleep restricted to 4 h time in bed (TIB) and 27 controls subjects received 10 h TIB over a period of 5 nights. With a control group of 10 subjects, 35 participants spent after two baseline nights ten nights with restricted sleep (4 h) in two exposures of five nights each. Daily calorie intake increased by an impressive 527 kcal per day in the sleep-restricted subjects versus control subjects and sleep-restricted subjects gained an average of 0.86 kg over the period. The study results strongly suggested a connection between sleep restriction and weight gain. A possible mechanism of this effect of sleep restriction may be the influence of the sleep-wake cycle on hormones affecting adipocytes.

Another study approached the issue of timing of food intake [40] by measuring the influence of circadian rhythm on hormone-sensitive lipase, a fat dissolving enzyme. The results showed that the later that dinner was eaten after 10:00 pm, the lower the effect of lipase. With lipase most active around midnight, longer duration of fasting at night was associated with higher fat mobilization and fat burning. The findings suggest that the sleep-wake cycle affects the role of fat dissolving enzymes on metabolism.

Unfortunately there are no human studies that investigate how nutrition may influence sleep. A large trial in mildly obese women who were followed for several weeks found no influence of the MIND diet (Mediterranean-DASH Intervention for Neurodegenerative Delay Diet) on sleep quality assessed with the Pittsburgh Sleep Quality Index questionnaire [41]. The MIND diet combined two forms of so-called healthy nutrition with low amounts of both carbohydrates and saturated fat and increased amounts of unsaturated fat and vegetables.

Other weight loss programs and diets have been unable to show any influence on sleep parameters [42‐48] though there may be gender differences concerning success with weight loss and the effects of behavioral changes on bedtimes and nutrition [43].