How Sleep Timing Preferences Relate to Eating Habits in Adults: A Broad Overview

A literature search was conducted in March 2019. A preliminary scoping search was conducted to determine the feasibility of this review. Once the search strategy was developed, it was independently verified by two reviewers. The following databases, EBSCO Host, Medline & Ovid, Pubmed, Scopus, and The Cochrane Library, were systematically searched to identify relevant publications. All the identified studies were screened for eligibility based on the information contained in the title and abstract. No restriction was set on the year of publication; however, studies were limited to those published in the English language and conducted among human subjects only.

The search terms used for this review were: Chronotype OR circadian preference OR morningness OR eveningness OR circadian timing OR chronobiological Dietary behaviour OR eating habit OR dietary intake OR food intake OR meal consumption OR diet quality OR macronutrient OR micronutrient OR meal timing OR food preference OR portion size OR appetite OR craving OR night eating syndrome OR binge eating 1 AND 2

All studies that reported an association between chronotype and dietary aspects, irrespective of study design (except for letters to the editor and conference proceedings with abstract only), were eligible for this review. However, this scoping review only included studies that were conducted among adults (more than 18 years old) and assessed the circadian rhythm phase as a chronotype. Studies conducted among animals or among individuals with a sleep disorder, mental illness, depression, bipolar disorder, and blind individuals were all excluded.

All the studies that were identified through database search were screened for eligibility by scrutinising the title and abstract of each listed study. The full text of potential eligible studies was retrieved for further screening of their suitability based on the review question and inclusion criteria. The process of selection is outlined in Figure 1. Unlike other scoping reviews, experts were not consulted for this current review.

Circadian rhythm can be measured through physiological markers such as melatonin, cortisol, and body core temperature [50]. The best method to measure circadian rhythm is by measuring the onset of melatonin secretion, which reflects the central circadian clock timing or suprachiasmatic nucleus [51]. However, the melatonin assay is expensive and labour-intensive with the further requirement of participants’ considerable cooperation [52]. The circadian rhythm phase can also be estimated through behavioural patterns such as the sleep–wake cycle, of which the timing varies between individuals, and is known as chronotype [5]. Chronotype or circadian typology is the expression of an individual’s diurnal preferences in performing various biological and psychological traits including the timing of sleep and wakefulness [53]. Individuals who perform daily activities optimally during morning hours, as well as those who sleep and wake up early are known as morning-type or having a morningness preference. In contrast, those who ideally perform activities in the evening hours, sleep and wake up late are recognised as evening-type or having an eveningness preference. The intermediate type refers to individuals who can be either morning or evening type. Chronotype was shown to be correlated with the physiological marker of circadian rhythm, the dim light melatonin onset (DLMO) [52,54]. Hence, this review only included studies that measured chronotype as circadian rhythm proxy. Table 2 shows the type of chronotype assessment and their distribution among included studies. Most included studies measured chronotype via a morningness–eveningness questionnaire (MEQ) by Horne, & Ostberg (1976) (n = 17) and three studies used the shortened version of MEQ [9]. Four studies determined chronotype by quantifying the midpoint of sleep [4]. Two studies from Japan used the Japanese version of morningness-eveningness questionnaire [55] specific to the shift-worker population, and two studies used the composite scale of morningness (CSM) [56].

Horne, & Ostberg (1976) initially categorised chronotypes according to five categories, namely definitely morning, moderately morning, intermediate, moderately evening, and definitely evening type. However, a majority of the included studies narrowed down the outcomes into three main categories, namely morning (M-type), intermediate/neutral (I-type), and evening type (E-type). Three of the studies designed cut-off MEQ scoring based on the median score of the respective population studied [30,42,43]. Some studies analysed the chronotype as a continuous variable, whereby the lower the MEQ, the higher the degree of eveningness, and vice versa. Studies that determined chronotype from the midpoint of sleep also presented the outcome as a continuous variable whereby the latest midpoint of sleep indicates a preference towards eveningness and vice versa. All measurements of chronotype reported are well validated and were employed in numerous studies of the human biological clock.

A wide variety of dietary assessment methods were reported in the included studies. Only 19 of the included studies used an established/standardised dietary assessment, for example, validated diet history questionnaire (DHQ) or food frequency questionnaire (FFQ), dietary recall, food record, or food diary. The majority of these studies reported alcohol and caffeine consumption solely based on a questionnaire (yes or no intake) and not the amount of the food intake. Food group was reported as gram/1000 kcal, gram/day, serving/day or gram.

The aim of the current scoping review was to systematically identify existing literature and to determine the association between chronotypes and various aspects of dietary habits among adults. Table 5 summarises the findings on morning and evening chronotypes’ dietary intake based on the majority significant results of the included studies. An evening chronotype was mostly related to an unhealthy dietary behaviour such as delay in meal timing, excessive food intake during night and skipping breakfast while morning type exhibited the opposite practice. Most importantly, the current review has managed to capture the patterns of each chronotype’s feeding time. Almost all of the studies that explored the aspect of mealtimes discovered similar findings with earlier chronotypes having earlier mealtimes and later chronotype having later mealtimes. Thus, these findings reflect the possibility that the norm of eating time follows a person’s own biological clock despite restrictions to the social clock [30,39,57,58]. Meule et al. (2012) proposed that the biochemical rhythm of appetite-regulating peptides (leptin and ghrelin) might be underlying this behaviour, since the evening chronotype has a more delayed release of ghrelin and leptin than morning chronotypes [30].

The findings from recent research have added to the mounting evidence supporting the connection between chronotype and breakfast [59,60]. A genetic study was conducted among 53 pairs of female twins in order to examine the impact of genetic heritability on meal times [59]. They found that the timing of breakfast (56%) and lunch (38%) was mostly influenced by genetic but not dinner time. In addition, the genetic heritability of waking time was also bigger than sleep time (55% vs. 38%). Hence, this might be the possible underlying mechanism behind the relationship between chronotype and breakfast habits because chronotype was also linked to sleep–wake time [11,27].

Another interesting finding from the latest research is that only the morning chronotype benefited from a higher energy intake during earlier times of the day (within two hours after waking up) and were protected against obesity, with reduced odds of being overweight or obese (p = 0.0006). Meanwhile, no significant effect was observed among evening chronotypes [60]. Despite many studies reporting the importance of breakfast [61,62], perhaps not everyone (especially late chronotype) can benefit from its advantages. The composition of meals can also affect or alter the beneficial effects of breakfast [63,64]. Thus, the current review suggests the needs for further research to establish appropriate meal timing and composition that are personalised to an individual’s circadian clock.

The central circadian clock coordinates rhythm in the sleep–wake cycle, thermoregulation, hormonal secretion, metabolism, and feeding behaviour [2,65]. In other words, all of these aspects of our physiology are mapped within a 24 h cycle. Delayed and irregular meal timing can cause circadian misalignment whereby the central circadian clock in SCN is desynchronised from the peripheral clock. The reason for this occurrence is because unlike the central clock where light is the main zeitgeber, feeding is the most potent zeitgeber for the peripheral clock [66,67]. Under normal physiology, feeding–fasting rhythm is triggered by the central clock through sleep-wake cycle [68]. During regular meal timings, circadian clock induces food entrainable oscillator (FEO) to generate food anticipatory activity (FAA) before food ingestion [69]. FAA induce food seeking action by increasing locomotor activity, appetite, body temperature, secretion of digestive enzymes and gastrointestinal motility a few hours before scheduled meal timing, preparing the organism to accept food and thus, optimising the food intake [2]. However, sudden food intake out of the usual meal time still induces peripheral circadian clock (in peripheral tissue and organ) action through acute signalling pathways, but this will not affect the central clock [68].Continuous irregular meal timing might however cause circadian desynchronization between the central and peripheral clock because the central clock in the brain will keep entrained to the day–night cycle while the peripheral clock in organs such as the liver, pancreas, muscles, and white adipose tissue are synchronised to the feeding.

A delay in mealtimes, particularly during dinner, might also cause circadian misalignment because feeding is at risk of shifting towards the resting phase (sleeping time). A human being is a diurnal creature, or in other words, we are meant to do vigorous activities such as working, eating during the day while resting and fasting during night. During the day, the ingestion of food will provide substrates (glucose, amino acid and lipid) to support the metabolic processes in body cells, whereas at night, the body’s storage of energy and substrates will be used to maintain metabolic homeostasis [70]. Hence, food intake during the rest phase is against our internal clock rhythm and is linked to diseases such as obesity, diabetes, and cardiovascular disease. There has been a rise in the number of studies investigating the effects of feeding during the circadian rest phase in animals [71,72,73,74,75]. For instance, shifting a rat’s feeding time to the rest phase (during the day) altered the normal rhythm of resting energy requirement (RER) [71,74,75]. The ad libitum fed rat (control group) showed a decrease in RER during the day, whereas it increased and remained high at night (active phase). In contrast, day-fed rats showed an opposite rhythm whereby RER was high during the day and maintained at the level above 1.0, indicating an increase in de novo lipogenesis, and thus, it may have promoted fat mass deposition [76]. A recent study have proven that even short term feeding during resting phase can induce leptin resistance in mice [73]. These mice were fed ad libitum for two weeks and were then divided into day-fed and night-fed groups for 11 days. Leptin was injected on day six for both groups. The night-fed group demonstrated a reduction in food intake (34%) and body weight (5%), however, leptin injection had no significant effect on day-feed mice implying leptin resistance. Furthermore, the researchers found that day-fed mice had higher triglyceride, free fatty acid, and total cholesterol levels compared to the night-fed mice. Nonetheless, those studies were conducted to simulate extreme shifts of meal timing during the rest phase, which mostly occurs among shift workers. There are a limited number of studies available on the impact of delays in meal timing among individuals with evening chronotypes. The current review noted that the delay in dinner or night meals among evening chronotypes was less severe compared to shift workers (range from 19:19 to 21:31). Nonetheless, evening chronotypes remain exposed to a milder form of chronodisruption (regards irregular and delay meal timing) [77]. Thus, more studies are needed to explore whether this habit has the possibility to cause adverse long-term effects.

The current review also reports in Table 5 that the following dietary behaviour was limitedly assessed by the included studies: food addiction, hunger, eating duration, watching TV during the meal, binge eating and meal skipping. Hence, more evidence is needed to investigate the association between chronotype and these dietary behaviours.

In terms of macronutrient intake, the majority of the included studies agreed that different chronotypes’ intake of energy, carbohydrate, and fat did not significantly differ (Table 5). However, Maukonen et al. (2017) further analysed the nutrient intake based on feeding timing (during morning and at night) and obtained significant results [39]. Evening chronotypes were found to consume less energy by 4%–5% in the morning, and as much as 6%–7% more energy at night compared to morning chronotypes [39]. This finding highlights the differences in the eating patterns between chronotypes in spite of consuming approximately the same amount of energy and nutrients, and also emphasises the importance of including the timing of food ingestion in dietary assessments. Another interesting discovery was that morning chronotypes were positively associated with a greater protein intake while evening types were associated with a greater sucrose intake. The early chronotypes were also linked to having a higher consumption of protein during early times of the day and this was also related to a decrease in the likelihood of being overweight and obese (p < 0.03) [60]. Nonetheless, evening chronotypes were linked to detrimental eating habits of sucrose which have notable negative effects on the health such as obesity, diabetes, an increase in cholesterol and blood pressure levels [78,79]. However, the current review cannot conclude the actual amount of sucrose intake by evening chronotypes due to the variations in the reporting of nutrient intake by the included studies. Hence, these findings need careful interpretation because the effects from normal consumptions of sucrose on the health is still undefined [79]. For vitamins and minerals, the current review was unable to establish a robust relationship with chronotypes because these variables were limitedly evaluated in the chosen studies.

The current review also emphasises on data concerning specific food intake by each chronotype because a single food group can be a major source of a variety of macro and micro-nutrients. For instance, grains constitute a notable supply of carbohydrates and an assortment of nutrients such as fibre, vitamin B, zinc, iron, and magnesium [80]. Animal sources constitute a high quality source of protein that supplies other essential nutrients. For instance, fish provides omega-3 fatty acids [81] while meat fulfils bodily needs of amino acid, iron, zinc, and vitamin B [82]. Hence, it is also important to capture the type of food intake, as well as to translate the information into nutritional value. The authors summarised that evening chronotype are usually engaged with unhealthy eating practices with lower vegetables consumption and higher intake of sweet food/beverages, caffeine, and alcohol (Table 5). On the other hand, no difference was noted between chronotypes for the consumption of legumes, meat, fish, and dairy products. No sufficient evidence was available to conclude an association between grains, fruits or fats, and oil with chronotypes because there was a similar number of studies that showed both the presence and absence of an association.

Importantly, the current review has detected the connection between nutrients and specific food group intake of chronotypes. Even though evening types are mostly linked to low vegetables intake but quite a number of studies have reported that they also consumed similar amounts of fruits as morning types. The current review proposed that this might be the reason for the similar fibre intake between the two chronotype groups. Similar outcomes reported for grains, fats, and oil intake reflected the similar nutrient intake of carbohydrate and fat respectively. There is a need for more studies to allow for a refined analysis of the grains group (whole grains and total grains intake) in a future systematic review. Perhaps, there is a possible significant difference in terms of carbohydrate quality between the chronotypes.

Evening chronotypes were reported to consume more alcohol and caffeine, which is widely recognised for its stimulant properties that can boost wakefulness and concentration by increasing brain activity [83,84,85]. However, many other studies have demonstrated how an excessive intake of alcohol and caffeine can result in reduced sleep duration and quality, increased fatigue and daytime sleepiness, and delayed sleep onset [85,86,87,88,89,90]. According to Galli et al. (2013), each alcoholic drink (contain 14g of alcohol) can reduce sleep duration by 30 min. In addition, the history of binge alcoholic drinking during the adolescent age was found to be linked to a diminishment of sleep quality during the adult stage among Mexican Americans and American Indians [87]. Fucito et al. (2018) also discovered that reduction in sleep duration influences a greater number of alcoholic drinks consumed on the following day among young adults. The mechanism underlying this association is still unclear, but it is proposed that both alcohol and caffeine intakes can delay the onset on nocturnal melatonin release, thus suppressing sleepiness [83,91]. Interestingly, some studies revealed how the consumption of alcohol and caffeine can increase the odds of later chronotype [90,92]. It was postulated that evening chronotypes have more chances to engage with ‘evening habits’ (drinking) due to a delay in sleep time compared to morning chronotypes, who sleep earlier [93]. Complementing that statement, the included studies also revealed that evening chronotypes significantly consumed the most number of alcohol beverages after 20:00 than morning and intermediate chronotypes [39]. Taking these reported findings into consideration, the present review emphasises the importance of appropriate timing for specific food intake.

Unhealthy eating practices is one of the risk factors linked to the development of obesity. A systematic review concluded that a diet consisting of high dietary fibre intake predicts less weight gain, whereas high consumption of sweets and desserts predicts greater weight gain [94]. Besides, another study also reported that excessive intake of sugary beverages, and a low intake of fruit and vegetables are all linked to the increase of adiposity [95]. Aforementioned in the review, the majority of the selected studies agreed that evening chronotypes displayed unhealthy eating practices which can raise risks of obesity. Most studies included in this review also reported information on weight or BMI, and most of the cross-sectional studies have shown no difference in weight markers between morning and evening chronotypes (Table 1). This might be due to many of these studies being conducted among university students, and this age group population is young and mostly healthy; making them more resistant towards the effects of misalignment in sleep/wake timing [96]. Studies that were conducted among the same population would usually report the same outcome where the majority of their respondents had a healthy weight range [97,98]. Furthermore, a cross-sectional study captures the weight parameter at one time only.

In contrast, a follow-up study [32] discovered that although there were no association detected between chronotypes and weight during baseline, the researcher found that later chronotypes were significantly linked to higher weight and BMI after eight weeks. Moreover, a prospective cohort study among six-years post-bariatric surgery patients demonstrated that evening types had a greater weight pre-surgery and experienced less excess weight loss compared to the morning types [35]. Although both chronotypes had almost the same energy intake, evening types had later lunch and dinner timing compared to the morning types, aligning with the finding of the current review. Consistent with this finding, Garaulet et al. (2013) also found that those who had lunch later were exhibited slower weight loss progress in their intervention despite having the same calorie intake and dietary composition as early lunch eaters; with majority of late lunch eaters being evening chronotypes [21]. Other intervention studies also found that morning types had better weight loss outcome compared to evening types [34,99]. In fact, recent findings discovered that circadian misalignment can also increase obesity risk among morning types [96]. A short sleep duration (less than 6 h), reduced sleep efficiency (less than 85%) or staying awake after sleep timing (more than 60 min) can increase body fat percentage, waist:hip ratio and waist:height ratio. Lastly, the impact of night meals differed according to the chronotypes in that late chronotypes acquired an increased risk of being overweight or obese yet no significant effects were found affecting early chronotypes [60].

With this in consideration, chronotypes might play an important role in the mechanism of weight gain and obesity in relation to dietary habits. Disturbance in active (feeding) and resting (sleeping) phase activities have become risk factors for the development of obesity and can hamper weight loss effectiveness. More studies are needed to determine the relationship between chronotype and dietary habits especially in relation to meal timing and obesity. The strength and limitations of this review should be carefully considered. This review has compiled a database of studies on the association between chronotype and three major dietary aspects, namely dietary behaviour, nutrient intake, and specific food group intake. This step is crucial in identifying each chronotype’s dietary aspects to develop a specific chrono-nutrition intervention for weight management or any dietary related health interventions. There are several limitations in the current scoping review. First, most of the included studies were cross-sectional design due to insufficient evidence on clinical trial in this area. Secondly, the current review generally discusses individuals without a sleep disorder because chronotype is affected by sleep and wake timing. However, it is important to highlight that there are other factors that influence chronotype, such as age, gender, latitude, work schedule, and social engagement [100]. Hence, future systematic reviews could more specific control for these modifiable and non-modifiable factors that affect chronotype. Lastly, the chronotype and the dietary assessment applied in the included studies varied widely, and thus no statistical analysis was conducted to conclude this scoping review. The current review highlights the need for more randomised clinical trial studies and thorough dietary assessments and diet quality to further clarify the link between chronotype and dietary aspects to enhance the effectiveness of weight management intervention.