Association between circadian clock gene expressions and meal timing in young and older adults

ABSTRACT Ageing is associated with a decline in circadian clock systems, which correlates with the development of ageing-associated diseases. Chrononutrition is a field of chronobiology that examines the relationship between the timing of meal/nutrition and circadian clock systems. Although there is growing evidence regarding the role of chrononutrition in the prevention of lifestyle and ageing-related diseases, the optimal timing of meal intake to regulate the circadian clock in humans remains unknown. In this study, we investigated the relationship between clock gene expression and meal timing in young and older adults. In this cross-sectional study, we enrolled 51 healthy young men and 35 healthy older men (age, mean±standard deviation: 24 ± 4 and 70 ± 4 y, respectively). Under daily living conditions, beard follicle cells were collected at 4-h intervals over a 24-h period to evaluate clock gene expression. Participants were asked to record the timing of habitual sleep and wake-up, breakfast, lunch, and dinner. From these data, we calculated “From bedtime to breakfast time,” “From wake up to first meal time,” and “From dinner to bed time.” NR1D1 and PER3 expressions in older adults at 06:00 h were significantly higher than those in young adults (P = 0.001). There were significant differences in the peak time for NR1D2 (P = 0.003) and PER3 (P = 0.049) expression between young and older adults. “From bedtime to breakfast time” was significantly longer in older adults than in young adults. In contrast, “From dinner to bed time” was significantly shorter in older adults than in young adults. Moreover, higher rhythmicity of NR1D1 correlated with longer “From bedtime to breakfast time” (r =  −0.470, P = 0.002) and shorter “From wake up to first meal time” in young adults (r = 0.302, P = 0.032). Higher rhythmicity of PER3 correlated with longer “From bedtime to breakfast time” in older adults (r =  −0.342, P = 0.045). These results suggest that the peak time of clock gene expression in older adults may be phase-advanced compared to that in young adults. In addition, a longer fasting duration from bedtime to breakfast in both young and older adults and earlier intake of meals after waking up in young adults may correlate with robust clock gene expression rhythms.


Introduction
The circadian clock constitutes an approximately 24-h rhythm in the body that regulates many physiological functions, such as body temperature, sleep -wake cycle, feeding time, patterns of physical activity, metabolism, mood, and cognitive function (Neves et al. 2022;Tahara and Shibata 2016).Abnormal circadian rhythms are associated with the development of lifestyle-and ageingrelated diseases such as obesity and diabetes as well as sarcopenia, osteoporosis, and depression.The circadian clock system consists of a central clock that is located in the suprachiasmatic nucleus (SCN) of the hypothalamus as well as peripheral clocks that are located in peripheral tissues such as the liver, kidney, and muscle.Neural firing rhythms and membrane properties of the SCN decrease with ageing (Hood and Amir 2017).In humans, body temperature amplitude, which is controlled by the central clock, is blunted in older adults (Hofman and Swaab 2006).Though peripheral clocks have been shown to exhibit normal oscillatory patterns in aged mice, the relationship between ageing and peripheral clock gene expression in humans is unclear.To appropriately regulate the circadian clock during ageing, it is important to examine how lifestyle habits influence the clock system at the human level.
The evaluation of clock gene expression using beard follicle cells is a non-invasive and convenient method (Akashi et al. 2010(Akashi et al. , 2020;;Takahashi et al. 2017), compared to other methods, such as those which use blood and oral mucosal tissue, and could be an effective tool to evaluate peripheral clock gene expression (Akashi et al. 2010;Sato et al. 2013).Moreover, it has been shown that circadian fluctuations of NR1D1, NR1D2, and PER3 expression in beard follicle cells, which correlate with the time of waking in individuals, may be used to evaluate circadian rhythms in humans (Akashi et al. 2010).NR1D1 and NR1D2 (also known as Rev-erb alpha and beta) are nuclear receptors that regulate several physiological functions including circadian rhythm, lipid metabolism, immune function, and cellular differentiation (Zhang et al. 2015).Some studies have indicated that a PER3 variant is correlated with circadian phenotypes (Morningness-Eveningness Questionnaire [MEQ] score), sleep phase disorder, and sleep homeostasis (Hida et al. 2014;Viola et al. 2007).
In animal models, age-related declines could be prevented by dietary and nutritional interventions (Barzilai et al. 2018).In fact, dietary interventions, especially diet restriction, have positive effects on anti-ageing mechanisms and extends the lifespan by regulating fastingrelated metabolic changes and improving the circadian system (Huffman et al. 2016;Lopez-Otin et al. 2016).Recently, the establishment of chrononutrition, a novel research area within the field of translational chronobiology, has brought attention to the relationship between dietary factors such as meal timings and nutrients and circadian rhythmicity (Aoyama and Shibata 2020;Flanagan et al. 2021).The timing of habits such as light exposure, exercise, and food intake influences circadian rhythms including clock gene expression in peripheral tissues (Tahara and Shibata 2018;Takahashi et al. 2018;Tanaka et al. 2020).Moreover, chrononutritional interventions such as time-restricted eating within 6-12 h during the day have been confirmed to have many health benefits including anti-ageing, anti-obesity, and antidiabetes effects (Sutton et al. 2018;Wilkinson et al. 2020).An important mechanism of time-restricted eating is daily fasting at night, which correlates with eating habits such as skipping breakfast and eating dinner late at night (Leung et al. 2019;Ogata et al. 2019).Moreover, in recent chrononutritional research, the timing of eating habits such as the duration of time from waking to eating breakfast, influenced several physiological functions (Veronda et al. 2020).However, to our knowledge, no study has investigated the association between the peripheral circadian clock and eating habits with a focus on meal timing in young and older adults.
In this study, we examined the relationship between circadian clock gene expression and meal timing in young and older adults.We hypothesised that the association between circadian clock gene expression and meal timing may be stronger in young adults because older adults have more stable and long-term lifestyle habits than young adults (Roenneberg et al. 2012(Roenneberg et al. , 2022)).The current analysis may provide new evidence for the optimal timing of food intake for anti-ageing effects and the prevention of ageing-related diseases from a chrononutritional perspective.

Participants
The participants included 51 healthy young men (aged 20-34 years) and 35 healthy older men (aged 63-79 years) who provided written informed consent.All participants in the present study were male.They were recruited from the general populations of the local communities such as Tokorozawa and Shinjuku in Japan via word of mouth and distributing study posters.None of the participants undertook athletic activities, though some participants were recreationally active.The exclusion criteria included yielding an insufficient total amount of RNA; the taking of psychotropic, sleep, steroid, anti-diabetes, or hyperlipidaemia-related medicines; and having a body mass index (BMI) >30 kg/m 2 .This study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the ethics committee of Waseda University (2014-G003).

Measurements of anthropometric parameters, meal timing, bedtime and wake-up time, chronotype, and the pittsburgh sleep quality index
For each participant, body mass was measured to the nearest 0.1 kg using a digital scale.BMI was calculated as weight in kilograms divided by the square of height in metres.
Participants were asked to respond to questions related to sleep and meal timing such as "During the last month, what time did you usually go to sleep and wake up?" and "What time did you eat breakfast, lunch, and dinner?"From these data, we calculated "From bedtime to breakfast time," "From waking up to first meal time," and "From dinner to bed time.""From wake up to first meal time" included participants who habitually skipped breakfast.The abovementioned explanatory values, including meal timing, bedtime and wake-up time, were based on the chrononutrition profile questionnaire, which is a valid and reliable method of assessing chrononutrition in daily life (Veronda et al. 2020).
To assess chronotype, we used the Horne -Ostberg MEQ (Horne and Ostberg 1976), which consists of 19 questions related to preferred sleep time and daily performance (e.g., What would be the best time to perform hard physical work?), with scores ranging from 16 to 86 points.Based on their MEQ scores, participants were divided into the following three chronotype groups: morningness (score 59-86), intermediate (score 42-58), or eveningness (score 16-41).
The Pittsburgh Sleep Quality Index (PSQI) scores range from 0 to 21, and a higher score indicates worse sleep quality.A PSQI cut-off score of > 5 provides 89.6% sensitivity and 86.5% specificity for detecting poor sleep quality (Buysse et al. 1989).

Beard follicle cell collection and laboratory assays
Each participant completed the collection of beard follicle cells under daily living conditions.The participants were asked to refrain from consuming alcohol and to maintain normal eating and sleeping habits during the day before and on the day of sampling.Beard follicle samples were collected over a 24-h period at 4-h intervals (from 06:00 h to 02:00 h the next day) by firmly holding and pulling the roots of beard.More than 10 beard hairs were collected at each sampling point.The collected beard hair was rapidly soaked in a dissolution buffer (RNeasy Micro Kit; QIAGEN).Participants were asked to collect beard follicle cells within ±15 min at each sampling point and to inform the investigator when they missed a collection.None of the participants missed any sample collection points.The sampling locations included the laboratory, workplace, and home.In addition, each participant was asked to store the samples in the refrigerator immediately after collection, and to transport the samples on an ice pack to the laboratory the next day.
After collection, all samples were stored at − 80°C until RNA purification.An RNeasy Micro Kit (Qiagen, Hilden, Germany) was used with frozen cytolysis solution to purify the total RNA, which was reverse-transcribed using an Advanced cDNA Synthesis Kit for RT-qPCR (Bio-Rad Laboratories, Inc., California, USA), and realtime PCR was performed using a TaqMan MGB probe (Applied Biosystems, California, USA) and 1/20 volume of the reverse-transcription product.Data were obtained using a 7500 Fast Real-Time PCR system (Applied Biosystems, California, USA) and expression was normalised to that of 18S rRNA.The primer and probe sequences are listed in Table S1, and the primers and probes were the same as those used in previous studies (Akashi et al. 2010;Takahashi et al. 2017).To evaluate the circadian rhythm of clock gene expression, the cosinor method was used.The amplitude, peak time, and goodness of fit were determined using the single-cosinor method (Acro.exeversion 3.5) (Refinetti et al. 2007).The goodness of fit is used as an index of strength of rhythmicity in sampling distributions that have been empirically determined.Lower goodness of fit values correspond to cosinor curves and indicate a higher level of robustness of circadian rhythms.

Statistical analysis
Data were analysed using Predictive Analytics Software (PASW) version 28.0 for Windows (SPSS Japan Inc., Tokyo, Japan).The Kolmogorov -Smirnov test was used to check for marker distribution, which did not differ significantly from the normal distribution.Unpaired Student's t-tests were used for normal distribution markers to assess differences in baseline data between young and older adults.The Mann -Whitney U test was used for abnormally distributed markers to assess differences in data at baseline.The chi-square test was used to compare the distribution of PSQI (good/poor sleep quality) and MEQ (morningness/intermediate/eveningness).Twofactor repeated-measures analysis of variance (ANOVA) was used to determine the effects of group (young and older adults) and time (24-h period).When significant main or interaction effects were detected, the Bonferroni method was used for post hoc comparisons.Partial correlation was used to examine the relationship between the MEQ score, clock gene expression parameters, and eating habits including meal timing.Statistical significance was accepted at the 5% level.The results are presented as the mean ± standard deviation (SD).

Physical characteristics, sleep -wake and meal timing, PSQI score, and MEQ score in young and older adults
Data for physical characteristics, sleep -wake and meal timing, PSQI score, and MEQ score in young and older adults are shown in Table 1.Both young and older adults were healthy and not taking medicine related to metabolic disease, sleep, and depression.There were significant differences in age (P = 0.001), height (P = 0.001), bedtime (P = 0.001), wake-up time (P = 0.001), lunch time (P = 0.003), dinner time (P = 0.002), and the MEQ score (P = 0.001) between young and older adults.Moreover, the distribution of chronotype (i.e., Moriningness/Intermediate/Eveningness; P = 0.001) differed significantly between young and older adults.Furthermore, "From bedtime to breakfast time" (P = 0.001) and "From dinner to bed time" (P = 0.017) were significantly longer and significantly shorter, respectively, in older adults than in young adults.

Diurnal expression of clock genes in young and older adults
The patterns of NR1D1, NR1D2, and PER3 expression in young and older adults are shown in Figure 1(a-c).Twofactor ANOVA revealed a significant main effect of time on all clock genes (P = 0.001), and a significant group × time interaction was detected for NR1D1 (P = 0.008) and NR1D2 (P = 0.040).Post-hoc tests revealed that NR1D1 and PER3 expressions in older adults at 06:00 h were significantly higher than those in young adults   (P = 0.001).Data for the amplitude, peak time, and goodness of fit for all clock gene expressions are shown in Table 2.There were significant differences in the peak time for NR1D2 (P = 0.003) and PER3 (P = 0.049) expression between young and older adults.

Young adults
Data for the relationships between clock gene expression parameters and sleep or mealtime in young adults are shown in Table 3.The amplitude of PER3 (r = −0.303,P = 0.031) positively correlated with dinner time.The peak time of NR1D1 (r = 0.379, P = 0.006) negatively correlated with dinner time.The peak times of NR1D1 (r = 0.399, P = 0.040), NR1D2 (r = 0.304, P = 0.030), and PER3 (r = 0.340, P = 0.015) positively correlated with the "From wake up to first meal time."The goodness of fit of NR1D1 expression in young adults positively correlated with bedtime (r = 0.357, P = 0.010) and dinner time (r = 0.283, P = 0.044).In contrast, the goodness of fit of NR1D1 expression negatively correlated with "From bedtime to breakfast time" (r = −0.470,P = 0.002) and "From wake up to first meal time" (r = 0.302, P = 0.032).

Older Adults
Data for the relationships between parameters of clock gene expression and sleep time or mealtime in older adults are shown in Table 4.The amplitude of NR1D1  expression in older adults positively correlated with wake-up time (r = 0.427, P = 0.010).The goodness of fit of PER3 negatively correlated with the "From bedtime to breakfast time" (r = −0.342,P = 0.045).

Discussion
To the best of our knowledge, this study is the first to examine and compare the relationship between circadian clock gene expression and meal timing in young and older adults.The main findings were that "From bedtime to breakfast time" in both young and older adults and "From wake up to first meal time" in young adults were associated with the rhythmic expression of clock genes.These findings indicate that a longer fasting duration until breakfast in both young and older adults and an earlier intake of meals after waking up in young adults may correlate with the robustness of clock gene expression rhythms.
It is well known that ageing is influenced by circadian clock systems that are related to the morningness chronotype (Acosta-Rodriguez et al. 2021;Holler et al. 2021).A comparison between young and older adults showed different chronotype distributions, with a lower rate of morningness in young adults.Several studies have shown that chronotype is influenced by age and reflects an individual's circadian preferences in sleep timing (Roenneberg et al. 2007(Roenneberg et al. , 2012)).Compared with morningness individuals, eveningness individuals prefer a later bedtime and wake-up time, which may delay breakfast and dinner times (Gill and Panda 2015;Roenneberg et al. 2012).In the present study, the MEQ score negatively correlated with bedtime, wakeup time, breakfast time, and dinner time in both young and older adults.Thus, we infer that chronotype distribution differs between young and older adults, whilst sleep -wake cycles and meal timings are strongly correlated with chronotype in both young and older adults.
Some studies have reported that evaluating peripheral clock gene expression in beard follicle cells could be an effective method for assessing circadian rhythms in humans (Akashi et al. 2010;Hattammaru et al. 2019;Takahashi et al. 2017;Watanabe et al. 2012).Our previous study showed that chronotype is closely correlated with the rhythm of clock gene expression in young adults (Takahashi et al. 2018).Another study reported that, compared to the morningness chronotype, the eveningness chronotype confers a delay in the phase of circadian clock gene expression in beard follicle cells (Ferrante et al. 2015).In the present study, we compared the expression of clock genes in beard follicle cells in young and older adults.Previous studies have shown that the neural firing rhythms and membrane properties of the SCN in mice decrease with aging (Hood and Amir 2017).In contrast, peripheral clocks have been reported to exhibit normal oscillatory patterns in aged mice (Hofman and Swaab 2006).Our study showed that NR1D1 and PER3 expression in older adults at 06:00 h was significantly higher than that in young adults.Moreover, there were significant differences in the peak time for NR1D2 and PER3 expression between young and older adults, which showed that the peak time of clock gene expression in older adults was phaseadvanced compared with that in young adults.Furthermore, the above-discussed findings provide evidence of changes in peripheral clock gene expression with ageing in relatively large samples from human participants.
Food/nutritional intake directly changes the phases of peripheral clocks (Tahara andShibata 2013, 2018).Based on the phase-response characteristic of foodinduced peripheral clock entrainment, breakfast may advance the phase of the peripheral clock, but late dinners may delay it.In fact, delayed eating in humans delays the plasma glucose rhythm and adipose clock gene expression rhythms (Wehrens et al. 2017).However, a previous study reported that skipping breakfast for 6 d did not change clock gene expression under the sleep -wake cycle in healthy young men (Ogata et al. 2020).Other studies have reported that a longer fasting duration before food intake may ensure more phase entrainment of the peripheral clock in mice, suggesting that breakfast (i.e., the first meal after overnight fasting) may play an important role in peripheral phase entrainment during the day (Flanagan et al. 2021;Hara et al. 2001;Tahara et al. 2011).In addition, breakfast skipping and late dinner are associated with the eveningness chronotype, which causes obesity and diabetes (Lopez-Minguez et al. 2019;Xiao et al. 2019).However, no information is available regarding the association between peripheral circadian clock and daily meal timing in young and older adults.In the present study, the peak time of NR1D1 expression in young adults negatively correlated with dinner time.Moreover, the goodness of fit of NR1D1 expression in young adults positively correlated with dinner time.Thus, late dinner was phase-delayed and correlated with decreased robustness of the clock gene expression rhythm in young adults.In contrast, the amplitude of PER3 expression in young adults positively correlated with dinner time.A previous study reported that the amplitude of clock gene expression is elevated by high meal volumes or distributions of meal intake in mice (Itokawa et al. 2013).Individuals who eat dinner late tend to be of the eveningness chronotype, which causes elevation of the amplitude of clock gene expression due to higher energy intakes at dinner.Furthermore, the amplitude of NR1D1 expression in older adults positively correlated with wake-up time.Though the underlying mechanisms are unclear, the abovementioned correlations may be influenced by the duration of time from breakfast to dinner.In general, individuals who wake up later and eat dinner late have a longer gap from breakfast to dinner, which causes the large amplitude of clock genes in young and older adults.The amplitude of circadian clock gene expression is thought to indicate the strength of the circadian rhythm, which decreases with age (Hofman and Swaab 2006).Our results indicate that wake-up and dinner times are important factors for changing circadian rhythm amplitude.
In recent chrononutritional research, the timing of eating habits such as the time from bedtime to breakfast and waking to breakfast have been reported to influence several physiological functions (Veronda et al. 2020).The importance of these eating habits is attributed to the proper timing of eating/fasting in relation to the body's energy needs, which can help maintain metabolic health (Potter et al. 2016).However, inappropriate meal timings can also impair physical and mental health.In the present study, the peak times of NR1D1, NR1D2, and PER3 expressions positively correlated with "From wake up to first meal time" in young adults.Moreover, the goodness of fit of NR1D1 expression positively correlated with "From wake up to first meal time" in young adults.Thus, earlier meal intake after waking up was phaseadvanced and correlated with the robustness of the clock gene expression rhythm in young adults.Furthermore, the goodness of fit of NR1D1 in young adults and PER3 in older adults negatively correlated with "From bedtime to breakfast time" in older adults.These results indicate that a longer fasting duration until breakfast in both young and older adults correlates with the robustness of clock gene expression rhythms.
The present study had several limitations.First, it is difficult to separate the cause from the effect because of the cross-sectional nature of the study.Therefore, an interventional study is required to examine the effects of meal timing on age-specific changes in clock gene expression.Second, it is unclear whether other external synchronisers, such as light and exercise, had an effect on clock gene expression.Further investigation will be needed to determine these factors by using a device that can accurately measure light exposure and physical activity, including exercise.Third, we did not assess the association between meal timing and clock gene expression by adipose tissue biopsy because the use of beard follicle cells was a non-invasive and more convenient method for humans.As some studies have reported an association between meal timing and the adipose clock gene expression rhythms in mice and humans (Mazzoccoli et al. 2012;Wehrens et al. 2017), it may be appropriate to use adipose tissue biopsy to examine this relationship in future studies.In addition, it was difficult to find significant correlations between MEQ, sleep times or meal times and clock gene expression rhythms using beard follicle cells.One reason for these results is caused by the difference for peak phase and expression patterns of each clock genes.In addition, MEQ, sleep times and meal times were evaluated by the questionnaire.It may be small sample size for analysing the correlations between results of questionnaire-based survey and clock gene expression.Lastly, we did not assess the participants' daily dietary intake or dietary content.Several nutrients/compounds, such as caffeine and amino acids, have regulatory effects on the circadian clock (Burke et al. 2015;Fukuda et al. 2018;Ikeda et al. 2018).Though we did not evaluate dietary intake on the day of the experiment, we asked each participant to maintain normal eating habits during the day before and on the day of sampling.Thus, we assumed that irregular and abnormal dietary intake did not influence clock gene expression patterns.
In conclusions, the peak time of clock gene expression in older adults was phase-advanced compared with that in young adults.Moreover, a longer fasting duration until breakfast in both younger and older adults and earlier meal intake after waking up in young adults were associated with the robustness of clock gene expression rhythms.

Figure 1 .
Figure 1.Diurnal expression of clock genes in young and older adults.The values are presented as means and standard deviation represented by bidirectional bars.The mean values are significantly different from those of the young adults.*P < 0.05

Table 1 .
Physical characteristics, sleep and diet time, PSQI and MEQ in young and older adults.
All data are presented as mean±standard deviation.The Pittsburgh Sleep Quality Index (PSQI): Good/Poor, and Morningness -Eveningness Questionnaire (MEQ) distributions: Chi-squared test, Others: unpaired t-test, or Mann-Whitney Test.*The number of

Table 2 .
The amplitude, peak time, and goodness of fit for clock genes expression in young and older adults.

Table 3 .
Relationships between parameters of clock genes expression and sleep or meal time in young adults.

Table 4 .
Relationships between parameters of clock genes expression and sleep and meal time in older adults.