Hostname: page-component-7c8c6479df-ph5wq Total loading time: 0 Render date: 2024-03-27T06:34:33.831Z Has data issue: false hasContentIssue false

Circadian rhythms, food timing and obesity

Published online by Cambridge University Press:  24 June 2016

J. Lopez-Minguez
Affiliation:
Department of Physiology, Faculty of Biology, University of Murcia, Murcia, Spain IMIB-Arrixaca, Murcia, Spain
P. Gómez-Abellán
Affiliation:
Department of Physiology, Faculty of Biology, University of Murcia, Murcia, Spain IMIB-Arrixaca, Murcia, Spain
M. Garaulet*
Affiliation:
Department of Physiology, Faculty of Biology, University of Murcia, Murcia, Spain IMIB-Arrixaca, Murcia, Spain
*
*Corresponding author: M. Garaulet, fax +34 968 36 39 63, email 2garaulet@um.es
Rights & Permissions [Opens in a new window]

Abstract

It is known that our physiology changes throughout the day and that several physiological hormones display circadian rhythmicity. The alteration of this normal pattern is called chronodisruption (CD). In recent years, it has been demonstrated that CD is related to obesity. Although several factors may be causing CD, one important aspect to consider is the failure in our internal clock. Indeed, studies performed in mutant animals have demonstrated that mutations in clock genes are related to obesity. In human subjects, mutations are rare (<1 % of the population). Nevertheless, it is rather common to have genetic variations in one SNP, which underlie differences in our vulnerability to disease. Several SNP in clock genes are related to obesity and weight loss. Taking into account that genetics is behind CD, as has already been demonstrated in twins’ models, the question is: Are we predestinated? We will see along these lines that nutrigenetics and epigenetics answer: ‘No, we are not predestinated’. Through nutrigenetics we know that our behaviours may interact with our genes and may decrease the deleterious effect of one specific risk variant. From epigenetics the message is even more positive: it is demonstrated that by changing our behaviours we can change our genome. Herein, we propose modifying ‘what, how, and when we eat’ as an effective tool to decrease our genetic risk, and as a consequence to diminish CD and decrease obesity. This is a novel and very promising area in obesity prevention and treatment.

Type
Conference on ‘Roles of sleep and circadian rhythms in the origin and nutritional management of obesity and metabolic disease’
Copyright
Copyright © The Authors 2016 

Our physiology changes during the day. It is well known that several hormones that are related to obesity, such as cortisol, leptin and adiponectin, among others, display circadian rhythmicity( Reference Milagro, Gomez-Abellan and Campion 1 Reference Ruiz-Lozano, Vidal and de Hollanda 3 ). The alteration of this normal pattern is called chronodisruption (CD), and it is related to several disturbances in many systems and organs of our body. In recent years, it has been demonstrated that CD is related to obesity.

CD or circadian disruption can be defined as ‘a serious disruption of the internal temporal order of the biochemical, physiological and behavioural circadian rhythms’( Reference Erren and Reiter 4 ). In our modern society, CD may be produced by several environmental factors, or external situations that are relatively common in the current society such as jet lag, shift work, night light pollution or overnight recreational activities (social jet lag)( Reference Garaulet, Ordovas and Madrid 5 ). Other factors are internal, and may produce CD by alteration of the core machinery of the molecular circadian clock. Brain- and muscle ANRT-like protein-1 (BMAL1), Period 2 (PER2) and circadian locomotor output cycles kaput (CLOCK), among others clock proteins, have a specific role in our physiology as well as in the circadian molecular clock. Indeed, several studies performed in mutant animals have demonstrated that mutations in clock genes are related to obesity, ageing and other metabolic alterations implicated in several chronic diseases( Reference Turek, Joshu and Kohsaka 6 , Reference Kennaway, Owens and Voultsios 7 ). For example, studies performed in experimental animals have shown that those animals that have mutations in Clock are more obese and display metabolic disturbances( Reference Turek, Joshu and Kohsaka 6 ).

In human subjects, mutations are very rare. Nevertheless, it is rather common to have genetic variations in one single nucleotide (SNP) that underlie differences in our vulnerability to disease. Regarding the circadian machinery, it is well known that several genetic variants are related to obesity, obesogenic behaviours and also to the effectiveness of the treatment in weight-loss programmes.

It is relevant to consider that clock genes are not only associated with obesity, but they also may interact with several obesogenic behaviours for obesity or weight-control parameters. One clear example is the interaction between CLOCK 3111T/C and emotional eating behaviours( Reference Lopez-Guimera, Dashti and Smith 8 ); Our results demonstrate that during a dietary treatment to lose weight, those subjects who were ‘very emotional eaters’ and also carriers of the risk allele C in CLOCK 3111T/C lost less weight, in comparison with (a) C carriers that were non-emotional eaters, and (b) TT carriers independently of their emotional status. Many other examples of interactions between clock genes and several behaviours will be explained in the following text. Thus, through nutrigenetics we know that our behaviours may interact with our genes and may decrease the deleterious effect of one specific risk variant; in other words, we can change our obesity predisposition and, although we are not able to change our genetic code, we can definitely change our behaviours.

From epigenetics, the message is even more positive: it is demonstrated that by changing our behaviour we can even change our genome. For example, we have demonstrated in CLOCK that DNA methylation levels at different CpG sites of CLOCK are higher in obese than in non-obese women, and these methylation levels were associated with several obesogenic behaviours such as snacking frequently, eating when bored or eating from large packages. Therefore, we presume from these data that, through some small but stable changes in our eating behaviours, we are able to change the DNA structure as a consequence of gene expression and, more importantly, we can change our destiny.

Chronodisruption and failures in the central clock

It has been previously described that CD can be the result of alterations in circadian system at different levels. Impairments can be present in the inputs or outputs to the circadian central clock located in the suprachiasmatic nucleous, but also in the central clock. With regard to the inputs, failures appear due to different circumstances, such as: (1) no changes between day/night synchronising agents, such as light/dark, timing of food intake (eating/fasting) or programmed exercise (activity/rest); (2) different periods or unusual phasing of synchronising agents, for example: light at night, nocturnal feeding or physical activity; and (3) shifts in the time provided by zeitgebers (i.e. jet lag, shift work). Outputs of the central pacemaker may also be altered: suppression of melatonin at night or loss of glucocorticoids rhythm like cortisol. Other problems in the central clock can result for example from the desynchronisation between the central pacemaker and peripheral oscillators or alterations in the so-called clock genes.

Nowadays we know that the internal clock regulates our physiological changes throughout the day. This clock works as a result of the expression of several clock genes that may activate and deactivate the clock to display a general pattern of 24 h. These clock genes can be classified into two groups: positive and negative. CLOCK and BMAL1 act as positive genes (they activate this clock) and are responsible for the synthesis of two transcriptional factors which, after dimerisation (CLOCK–BMAL1), induce the expression of negative genes; these are PER and cryptochrome (CRY), together with nuclear receptor subfamily 1, group D, member 2 alpha (a transcription factor). These negative elements, after dimerisation (PER–CRY), undergo a nuclear translocation and act as suppressors of CLOCK and BMAL1 expression, consequently slowing and stopping the clock.

The levels of positive and negative elements oscillate in antiphase generating circadian rhythms (with a period of approximately 24 h) in suprachiasmatic nucleus in vitro ( Reference Garaulet and Madrid 9 ). Although the circadian system is mainly formed by the central pacemaker in suprachiasmatic nucleus, it is known since 2001 that this central clock, in turn, synchronises the activity of several peripheral clocks present in our organs and tissues, such as heart, lung, liver, oral mucosa, pancreas and adipose tissue, among others, by a cyclical secretion of hormones and the activation of the autonomic nervous system( Reference Balsalobre, Damiola and Schibler 10 ).

Failures in the central clock: mutations in experimental animals

There are many factors, which may affect the circadian system at different levels. In relation to clock genes, numerous studies performed in experimental animals with mutations in clock genes have proved that there is a relation between these mutations and further failures in circadian system and illness. In this sense, animals with mutations in clock genes show a higher risk of developing certain diseases, such as CVD, cancer and obesity( Reference Garaulet, Ordovas and Madrid 5 ).

One of the first studies which showed the effect of genetic mutations on chronic illnesses was conducted in 2005 by Turek et al.( Reference Turek, Joshu and Kohsaka 6 ) This study proved that homozygous Clock mutant mice were hyperphagic and obese, displayed adipocyte hypertrophy and developed metabolic syndrome (MetS). However, there is already controversy. Results have not proved to be consistent across different experiments. For example, Kennaway et al. showed that Clock (Delta19) mutation in mice did not cause obesity; by contrast, the authors found reduced plasma-free fatty acids and increased insulin sensitivity, together with increased plasma adiponectin, a protective anti-inflammatory cytokine( Reference Kennaway, Owens and Voultsios 7 ).

Further studies have shown that mutations in different clock genes are related to an increased risk of developing certain diseases, such as premature ageing, as has been demonstrated by Kondratov et al. In this study, Bmal1 (−/−) mice had reduced lifespan (mean lifespan of knockout animals was 37 v. 120 weeks for wild-type animals) and displayed various symptoms of premature ageing, including sarcopenia, alteration of the percentage of lymphocytes and impaired vision among others( Reference Kondratov, Kondratova and Gorbacheva 11 ).

Genetic variations of clock genes in human subjects Clock genes associate with obesity in human subjects

Mutations are very rare in human subjects. However, it is rather common to have genetic variations in one SNP at some specific position of the genome, each of these variations being present in more than 1 % of the general population. These genetic variations underlie differences in our vulnerability to several diseases such as obesity.

In this sense, the work performed by Sookoian et al. in 2008 was the first to demonstrate that several variants at CLOCK were associated with obesity, especially with abdominal obesity( Reference Sookoian, Gemma and Gianotti 12 ). In addition, in the same year, Scott et al. confirmed these results by showing that CLOCK could play a relevant role in the development of MetS, type 2 diabetes and CVD( Reference Scott, Carter and Grant 13 ).

Subsequently, our research group replicated the results of Sookoian and Scott and also demonstrated that several genetic variants in clock genes were related to obesity and related diseases such as MetS. For example, CLOCK SNP (rs3749474, rs4580704 and rs1801260 (3111T > C)) were associated with BMI, energy intake and different variables related to obesity( Reference Garaulet, Lee and Shen 14 ). In fact, our results showed that carriers of the minor alleles consumed more energy( Reference Garaulet, Lee and Shen 15 ) (1046 kJ (250 kcal)/d, ate more fat (5–10 g), and were more obese (3 kg body weight) for rs3749474, rs1801260 (3111T > C)). Conversely, minor allele carriers (A) of CLOCK rs4580704 eat less energy and fat, were less obese and showed a decreased risk of developing diabetes (31 % lower) and hypertension (46 % lower) than non-carriers( Reference Garaulet, Lee and Shen 14 ). Besides, one genetic variation in CLOCK rs1801260 (3111T > C) was associated with increased obesity and with less weight loss( Reference Garaulet, Corbalan and Madrid 16 ). Moreover, the proportion of subjects who slept < 6 h/d was higher among subjects with the C minor variant of this polymorphism than non-carriers (59 v. 41 %)( Reference Garaulet, Sanchez-Moreno and Smith 17 ). Some of these associations are functionally explained by the presence of a polymorphism involving a change in the structure of the mRNA leading to a change in gene expression; this is the case of the CLOCK polymorphism rs3749474( Reference Garaulet, Lee and Shen 15 ). Most of these novel results were replicated in two different populations (Mediterranean and North American) with different genetic backgrounds and dietary habits, which emphasise the robustness of the findings.

Other results from our group also showed that different genetic variants of the clock such as nuclear receptor subfamily 1, group D, member 2 alpha rs2314339 were also associated with obesity. Once more, these results were replicated in two independent populations: Mediterranean and North American. This gene is the major regulator of rhythmic BMAL1 transcription and it is considered to be the molecular link between the positive elements of the clock (CLOCK and BMAL1) and the negative (PER and CRY). One specific aspect in this association was that nuclear receptor subfamily 1, group D, member 2 alpha rs2314339 is associated with obesity through a decrease in physical activity( Reference Garaulet, Smith and Gomez-Abellan 18 ), differently to other clock genetic variants, which were associated with obesity through changes in dietary intake. Indeed, our findings are consistent with those obtained in experimental animals, which demonstrated that mutants RevErbα −/− mice displayed a significantly lower spontaneous locomotor activity than their wild-type littermates( Reference Woldt, Sebti and Solt 19 ).

Our research group has also found significant associations between clock genes from the negative element of our circadian system and several obesogenic behaviours. In this regard, a key clock gene is PER2. A missense mutation in human PER2 has been linked to several psychological alterations, such as seasonal variations in mood and behaviour, and winter depression( Reference Partonen, Treutlein and Alpman 20 ). Our group was the first one to demonstrate that some specific PER2 SNP (rs2304672C > G and rs4663302C > T) were associated with abdominal obesity( Reference Garaulet, Corbalan-Tutau and Madrid 21 ). In particular, PER2 rs2304672C > G minor allele carriers G (6 % of the population) displayed several obesogenic behaviours such as an increased attrition of the weight-loss treatment, increased frequency of snacking, stress while dieting, eating while bored and skipping breakfast, when compared with non-carriers C( Reference Garaulet, Corbalan-Tutau and Madrid 21 ). It is impressive to observe how a small change in our genome (a cytosine for a guanine) may produce such a big variation in the mRNA structure and, as a consequence, changes in our gene expression.

Clock genes interact with several behaviours for obesity in human subjects

Clock genes may also interact with behaviours for obesity. This is the case of those behaviours directly related to emotional eating. For example, seeking refuge in food (especially high-energy foods) is a very common strategy to reduce anxiety, sadness and negative emotions that occur when following a long-term diet or due to difficult circumstances of our daily lives. Our results demonstrated that those participants of our weight-loss programme who were carriers of the risk allele C at CLOCK 3111T > C and displayed emotional eating behaviours more frequently, had more difficulties to lose weight during the treatment. Interestingly, C carriers who did not display emotional behaviours, in spite of being potentially at risk due to their genetic background, showed a similar weight loss as TT carriers (protective allele). These results are encouraging because they show that, by changing ‘how’ we eat, we can reduce or even eliminate the deleterious effect of a genetic variant.

In this sense, our research group has developed a ten-item questionnaire called Emotional-Eater-Questionnaire to be used in clinical practice, which classifies obese individuals as a function of the relation between food intake and emotions. The questionnaire allows classification of individuals as non-emotional eaters (emotions have little or nothing to do with their eating behaviour) and emotional eaters (feelings and emotions affect their eating as a response to negative emotions, such as anxiety, depression, anger and loneliness).

Out of a total population of 1500 subjects, we found that 60 % were emotional eaters, 40 % were minor C allele carriers and 30 % were both C carriers and emotional eaters. Considering these results, it may be useful to develop behavioural and cognitive programmes aiming to reduce the frequency of emotional eating episodes, particularly for the 30 % who carry the risk variant C. These results are encouraging and should be used in clinical practice.

Clock genes are related to weight loss

Genetics and weight loss

One of the main problems in weight-loss treatments is the dramatic inter-individual variability in response to the treatment. It is believed that an elucidation of the genetic component will help to predict weight-loss effectiveness. Some studies carried out in monozygotic twins analysed genetic factors for effectiveness of weight-loss programmes, other studies emphasised the familial aggregation in their ability for losing weight; and the role that parental obesity plays in this matter.

However, most of the prior genetic studies focused on candidate obesity-related genes (NUGENOB) found little association between genetics and weight loss, except for some cases (PPARG)( Reference Garaulet, Smith and Hernandez-Gonzalez 22 ). Notably, no association with weight loss was observed for thirteen obesity risk SNP from the genome-wide association study( Reference McCaffery, Papandonatos and Huggins 23 ), suggesting that the genetic architecture of obesity and weight-loss effectiveness may differ, or that gene–environment effects need to be considered.

Surprisingly, clock genes are related to weight loss. In this sense, our most relevant results show that there is an association between a CLOCK 3111T > C (rs1801260) and weight-loss effectiveness( Reference Garaulet, Corbalan and Madrid 16 ) (Fig. 1). This study demonstrated that C carriers were more resistant to weight lost than TT homozygotes, when these individuals were subjected to an energy-restricted diet( Reference Garaulet, Corbalan and Madrid 16 ). These data suggest that clock SNP may predict weight loss in response to a low-energy diet.

Fig. 1. Represents how several impairments in the central clock are related directly with obesity. In experimental animals, these impairments are due to mutations; however in human subjects, mutations are rare and these impairments are caused by human genetic variants. Thus, these human genetic variants may be associated directly with obesity or through different behaviours. Moreover, these genetic variants may interact with different behaviours to influence obesity and weight loss. We can consider several solutions: (a) changing our behaviours, these changes are directly related to a decrease in obesity and an increase in weight loss, or (b) through genetics by changes in our behaviours that may interact with the SNP to produce a decrease in obesity and weight loss.

We also reported that SIRT1 and CLOCK 3111T > C combined genotype was associated with a particular chronotype and with weight-loss resistance in a behavioural therapy treatment for obesity( Reference Garaulet, Esteban Tardido and Lee 24 ). In this work, subjects carrying minor alleles at SIRT1 and CLOCK displayed a significantly lower total weight loss and lower weekly weight-loss rate as compared with homozygotes for both major alleles.

Main barriers to lose weight (environmental factors)

It is classically known that weight-loss effectiveness is highly related to environmental factors( Reference Corbalan, Morales and Canteras 25 ). In a study performed in 1400 obese subjects from a Mediterranean area in southeast Spain who attended a weight-reduction programme, those subjects who lost less weight during the treatment had more obstacles and higher ‘barriers to lose weight’ score (double) than those who achieved their goal. One of the main barriers was the lack of motivation. In fact, those patients who did not achieve their weight goal (loss of at least 10 % of their initial weight) displayed 83 % more predisposition to lose their motivation than those who achieved it. These patients were also predisposed to suffer other barriers related to emotional eating, such as stress-related diet, thinking in ‘black or white terms’ or eating while bored( Reference Corbalan, Morales and Canteras 25 ).

Genetics associates with obesogenic behaviours to modulate total weight loss

Furthermore, we have discovered how certain polymorphisms associate with different environmental factors to modulate total weight loss( Reference Lopez-Guimera, Dashti and Smith 8 ). This is the case of CLOCK 3111T > C (rs1801260) SNP. The present author group have shown that minor allele C carriers of CLOCK 3111T > C were more evening type, had less healthy dietary habits and tended to sleep less than T carriers( Reference Lopez-Guimera, Dashti and Smith 8 ). Moreover, the combination of SIRT1 and CLOCK 3111T > C was also associated with evening preference, with less adherence to Mediterranean food intake patterns and with higher plasma levels of ghrelin (hunger hormone)( Reference Garaulet, Esteban Tardido and Lee 24 ).

Currently, one of the greatest challenges in weight-loss treatments is to detect good predictors of success. The analysis of genes, dietary intake and behavioural factors could be used as a tool to increase the success of such treatments.

Genetics variants in clock associate with alterations in circadian rythmicity (chronodisruption), and these alterations may predict a lower effectiveness of the treatment

In order to achieve a better understanding of the relationship between circadian rhythms and obesity, we evaluated changes in circadian rhythmicity with CLOCK 3111T > C SNP in overweight women, by recording continuously body peripheral temperature (wrist temperature (WT)), actimetry and position during 1 week. These variables have been described as good markers of CD( Reference Bandin, Martinez-Nicolas and Ordovas 26 ). Our results showed that risk carriers (C) displayed significant circadian abnormalities such as: (a) lower amplitude, (b) greater fragmentation of the rhythm, (c) less stable circadian pattern and (d) significantly decreased circadian function, as was shown after assessing by the circadian function index. C subjects were also less active, started their activities later in the morning and were sleepier during the day, showing a delayed acrophase that characterises the evening-type subject( Reference Bandin, Martinez-Nicolas and Ordovas 26 ).

In a subsequent study, recording the same markers of CD (WT, actimetry and position), we also demonstrated that alterations in circadian rhythms (CD) could be good predictors of a lower effectiveness of weight-loss treatments. Our results showed that those subjects who were low responders in the treatment displayed a more flattened pattern of WT characterised by lower amplitude, higher intradaily variability and deteriorated circadian functionality index when compared with the individuals who had a greater response to the treatment( Reference Bandin, Martinez-Nicolas and Ordovas 27 ). This work supports the hypothesis that analysing the circadian rhythms of the subjects at the beginning of the treatment could be useful to predict future weight loss( Reference Bandin, Martinez-Nicolas and Ordovas 27 ).

Genetics and chronobiology (twins study)

After having proved that CD and several genetic variants in clock are related to obesity( Reference Garaulet, Lee and Shen 14 ), the next step was to determine to what extent CD is determined by our genetics or if it is mainly driven by environmental factors. For this purpose, in 2013 we determined the circadian system heritability as assessed by WT, using classical twin models( Reference Lopez-Minguez, Ordonana and Sanchez-Romera 28 ). In twin studies, we can distinguish between genetic and environmental variations because monozygotic twins share 100 % of their genes and dizygotic twins share, on average, 50 % of their genetic makeup( Reference Vink, Groot and Kerkhof 29 ).

The sample selected for the study comprised fifty-three pairs of female twins: twenty-eight monozygotic and twenty-five dizygotic, who carried a sensor to measure WT for one week( Reference Lopez-Minguez, Ordonana and Sanchez-Romera 28 , Reference Lavie 30 ).When we analysed the average weekly temperature waves, we realised that in monozygotic twins the patterns were very similar between sisters, not only in those sisters who displayed a regular circadian pattern, but also in those who had an irregular pattern with more flattened curves. However, among dizygotic twins, patterns were rather different between sisters. These first observations demonstrated that genetics play an important role in the circadian rhythm of peripheral body temperature and also in CD.

Further results showed that the individual chronotype, i.e. to be morning or evening type, was mainly driven by genetic factors (about 70 %). Nevertheless, when analysing other characteristics of the WT rhythm as the amplitude of the rhythm and interdaily stability, which play an important role in the CD and obesity; the environmental component was stronger than the genetic. Environmental factors are those external influences that can be controlled and modified by the intervention; for example, changing the subject's behaviours such as the timing of food intake and of physical activity or bedtime. Therefore, we concluded that in the clinical practice, it could be useful to focus efforts in modifying these environmental parameters in order to improve the circadian patient's health.

Nutrigenetics says: our genome may interact with our behaviours

Over the course of many years, researchers have been focused on the search for potential interactions between genetic and environmental behaviours, such as nutrition, for various physiological and pathophysiological aspects. In fact, over the last few years it has been demonstrated that different SNP in our circadian machinery interact with dietary intake and several behaviours for obesity and for other metabolic risk (MetS)-related variables( Reference Lopez-Guimera, Dashti and Smith 8 , Reference Garaulet, Lee and Shen 14 , Reference Dashti, Smith and Lee 31 ).

In this sense, in 2009 we revealed several interactions between CLOCK variants and dietary intakes for different MetS traits. We showed a significant interaction between CLOCK SNP rs 4580704 and MUFA intake for glucose plasma values and homeostatic model assessment: the protective effect of the minor allele on insulin sensitivity was only present when MUFA intake was >13·2 % of energy. We also found different effects across CLOCK 3111T > C genotypes for SFA intake (% of energy). The deleterious effect of C risk variant on waist circumference was only found with high SFA intakes (>11·8 %)( Reference Garaulet, Lee and Shen 14 ).

More recently, one study developed in 2014 by the present author group showed a novel CRY × diet interaction for insulin resistance in a Mediterranean population, that was also replicated in a North American population( Reference Dashti, Smith and Lee 31 ). Findings from the meta-analyses indicated that an increase in carbohydrate intake was associated with an increase in HOMA-IR, in fasting insulin and a decrease in QUICKI, only among individuals homozygous for the minor C allele in CRY1 polymorphism, rs2287161( Reference Dashti, Smith and Lee 31 ). CRY (negatives elements of clock) are implicated in the regulation of glucose metabolism. Studies in experimental animals have demonstrated that gluconeogenesis can be modulated by circadian changes in the hepatic expression of Cry ( Reference Zhang, Liu and Dentin 32 ). Therefore, factors which affect the expression of CRY could impair hepatic regulation of glucose homeostasis and thus increase risk for diabetes in human subjects( Reference Garaulet and Madrid 9 ).

These results can help us design more effective dietary programmes: for those CC carriers at CRY1 rs2287161, our advice would be focused on helping them reduce their carbohydrate intake so HOMA risk could be even lower than G carriers (see Fig. 1).

In another example, but this time for gene × drug interaction (i.e. exogenous melatonin), is related to the melatonin receptor 1B rs10830963 variant. Hence, melatonin receptor 1B rs10830963 risk variant G worsens the effect of exogenous melatonin on glucose tolerance, thus G carriers should avoid having food together with exogenous melatonin administration( Reference Garaulet, Gomez-Abellan and Rubio-Sastre 33 ). These results could affect those subjects taking melatonin (i.e. for sleep problems), but also G carriers who usually have food (dinner) together with high levels of endogenous melatonin, as happens in natural late eaters or in shift workers. Further studies should answer these specific questions about timing of food intake, genetic risk and insulin resistance.

In Table 1, we include advice to prevent genetic and behavioural association/interactions to obesity that may be crucial for genetically informed personalised nutrition in obesity treatment.

Table 1. Advices to prevent genetic and behavioural association/interactions for obesity that may be crucial for genetically informed personalised nutrition in obesity treatment.

HOMA, homeostatic model assessment.

All these examples and many others show us how our behaviours may interact with our genes and might decrease the deleterious effect of one specific risk variant. Consequently, to the question of whether or not we are predestinated nutrigenetics answers: no, we are not predestinated; even though we cannot change our genome, we can change our behaviours in order to improve our health.

Epigenomics says: we can change our genome with our behaviour

Presently, it is evident that genes are modulated by a set of regulators, which switch on or switch off altering the person's phenotype. The field focused on the study of these genetic expression regulators is called epigenetics( Reference Campion, Milagro and Martinez 34 ).

The main idea underlying this field is the following: epigenetics does not change our DNA; however, it decides how much or whether some genes are expressed in different cells of our body. These changes can occur as a result of three processes: (1) DNA methylation; (2) chromatin modifications; (3) post-transcriptional gene silencing mediated by RNA. The best characterised epigenetic mechanism is methylation, which consists of the addition of one methyl group to a cytosine of DNA when it stands next to guanine nucleotide (dinucleotide CpG, shorthand for ‘–C–phosphate–G–’) or cytosine and guanine separated by one only phosphate.

In this sense, it has been shown that circadian clock genes expressions can be regulated by epigenetic mechanisms( Reference Sahar and Sassone-Corsi 35 ). Moreover, methylation levels in clock genes have been associated with obesity and metabolic disturbances( Reference Milagro, Gomez-Abellan and Campion 1 ).

In 2012, the present group published a study, which demonstrated significant associations between the methylation levels of several CpG located in CLOCK with MetS, weight loss and obesity( Reference Milagro, Gomez-Abellan and Campion 1 ). This study, performed on sixty women, demonstrated that the methylation degree in the CpG sites of CLOCK, such as CpG 1, 5, 6 and 8, increased with obesity. More importantly, the ‘how’ we eat was related to methylation levels at CLOCK. We observed that those patients who tended to snack frequently had twelve times higher methylation levels in CLOCK CpG1, whereas those who tended to eat when bored or eat from large packages had nine to nineteen times higher methylation levels. These increases in methylation levels of blood mononuclear cells suggest a suppression of CLOCK expression, which has also been related to obesity. In Table 2, we have represented the effect of different behaviours in methylation levels of CLOCK CpG1, which shows the importance of ‘how we eat’.

Table 2. Effect of different behaviours in methylation levels of CLOCK CpG1, which shows the importance of ‘how we eat’

Milagro et al. Chronobiology International. 2012( Reference Milagro, Gomez-Abellan and Campion 1 ).

More recently, research demonstrated that weight-loss nutritional intervention modifies the methylation pattern of BMAL1, CLOCK and nuclear receptor subfamily 1, group D, member 1 in the whole blood. More importantly, these changes in methylation levels of BMAL1 associated with a reduction in metabolic risk parameters, i.e. serum lipids( Reference Samblas, Milagro and Gomez-Abellan 36 ). Interestingly, evening types presented more epigenetic modifications caused by the weight-loss intervention than morning types. Moreover, baseline methylation of BMAL1 positively correlated with energy intake and carbohydrate intake, suggesting that interventions designed to decrease total energy intake and carbohydrates may help decrease the vulnerability to CD and obesity, especially in evening type subjects (see Table 1).

What can we do? We can change what, how and when we eat

In this review, we have shown so far that we can interact with our genome, even change it, and improve our health through changes in how we eat, what we eat and other daily behaviour. One novel aspect to consider in dietary interventions may be when we eat: the timing of food intake. If we take into account that eating is an external synchroniser of our peripheral clock, and that an unusual eating time may cause a disruption of our circadian system( Reference Arble, Bass and Laposky 37 , Reference Sherman, Genzer and Cohen 38 ), the ‘when’ we eat may have a significant role in obesity treatment. In this line, a novel observational study performed in 2013 revealed that eating late may influence the success of weight-loss therapy, leading to a decrease of its effectiveness. This study was performed on 411 overweight and obese subjects who underwent a dietary weight-loss treatment: 199 subjects were early eaters (had their main meal of the day, lunch, before 15·00 hours) and 212 subjects were late eaters (had lunch after 15·00 hours). Late eaters lost significantly less weight than early eaters although having similar age, appetite hormones values, energy intake, sleep duration or macronutrients distribution. It is remarkable that late eaters were more evening types( Reference Garaulet, Lee and Shen 14 , Reference Garaulet, Lee and Shen 15 ). The conclusion of this study was that, if we could change our behaviour in Spain towards an earlier lunch, perhaps we could lose more weight during a dietary treatment.

More recently, another observational study developed in Barcelona showed that the timing of food intake might also influence the effectiveness of bariatric surgery in severely obese subjects. It is true that many subjects were good responders to the treatment (68 %) losing 80 % of their initial excess weight during the first year after surgery, and maintaining this weight loss during 6 years follow-up. However, about 11 % of the population is defined as ‘primarily weight loss poor responders’; these subjects lose only about 40 % of their initial excess weight during the first year after surgery. From a clinical point of view, it is important to detect this group of poor responders before the surgery, in order to make a decision about their further treatment. The results of this study demonstrated that the percentage of late eaters was significantly higher among the primarily poor weight-loss responders (about 70 %) than in good weight-loss responders (about 37 %). Moreover, poor weight-loss responders had lunch later than good weight-loss responders. Surprisingly, obesity-related variables, biochemical parameters, pre-surgical total energy expenditure, sleep duration, chronotype, energy intake and macronutrient distribution, were similar among groups( Reference Ruiz-Lozano, Vidal and de Hollanda 3 ).

To discover why the timing of food intake could influence weight loss, we carried out a randomised study in which thirty-two lean and young women completed two protocols: one including assessments of resting energy expenditure (indirect calorimetry) and glucose tolerance, and the other including circadian-related measurements, such as profiles in salivary cortisol and WT (Fig. 2)( Reference Bandin, Scheer and Luque 2 ). Participants received standardised meals during both meal interventions weeks and were studied under two timings: early eating (lunch at 13·00 hours) and late eating (lunch 16·30 hours). The results of this study showed that these lean women, after 1 week of eating late, suffered from metabolic alterations that usually characterise obese women, such as decreased glucose tolerance, decreased resting energy expenditure and decreased carbohydrate oxidation. Moreover, late eaters had a flattened pattern of daily cortisol and alterations in daily rhythms of body temperature similar to those that characterise obese women as previously described( Reference Corbalan-Tutau, Madrid and Ordovas 39 ).

Fig. 2. Two typical examples of how behaviours may influence genetic variants. (a) Shows an example of genetic association between the genetic variant of CLOCK 3111T/C and weight loss, where the C carriers exhibit a greater difficulty in losing weight during 28 weeks treatment in a clinical nutrition centre than TT carriers. (b) Shows an example of gene × behaviour interaction for insulin resistance; it is remarkable that when subjects are CC carriers of CRY1 rs2287161 and also eat high values of carbohydrates, their insulin resistance homeostatic model assessment (HOMA) is higher. However, among G carriers (GG + GC) values of HOMA are independent of carbohydrates intake and they remain constant.

In Table 3, we summarise the specific recommendations for the prevention of obesity and MetS by improving the circadian system health, based on available scientific evidence proved in epidemiological, clinical and experimental animals (A);  on interventional studies in human subjects or in animal models (B); and on preliminary studies or low number of studies (C).

Table 3. Summarises the specific recommendations for the prevention of obesity and metabolic syndrome by improving the circadian system health, based on available scientific evidence proved in epidemiological, clinical and experimental animals (A); on interventional studies in human subjects or animal models (B); and on preliminary studies or low number of studies (C)

These recommendations can be very useful for the general population and also in the clinical practice to prevent or treat obesity alterations.

Summary

In Fig. 3, we represent a summary of the current revision. As it is shown, several impairments in the central clock are related directly with obesity. In experimental animals, these impairments are due to mutations; however in human subjects, mutations are rare and these impairments are caused by human genetic variants. Thus, these human genetic variants may be associated directly with obesity or through different behaviours. Moreover, these genetic variants may interact with different behaviours to influence obesity and weight loss. We can consider several solutions: (a) changing our behaviours as these changes are directly related to a decrease in obesity and an increase in weight loss; or (b) through genetics by changes in our behaviour that may interact with the SNP to produce a decrease in obesity and weight loss.

Fig. 3. Represents how eating late affects the metabolic characteristics of lean women; showing that, after 1 week of late eating, their metabolic pattern was more similar to the metabolic pattern of obese women.

In conclusion, herein, we propose modifying what, how and when we eat as an effective tool to decrease our genetic risk, and as a consequence to diminish CD and to decrease obesity. The research addressed in this revision presents a novel and very promising area in obesity prevention and treatment.

Financial Disclosure

This study was supported by grants from Spanish Government of Economy and Competitiveness (SAF2014-52480-R) and European Regional Development Fund (ERDF) to Marta Garaulet.

Conflicts of Interest

None.

Authorship

All authors contributed equally to the preparation of this paper.

References

1. Milagro, FI, Gomez-Abellan, P, Campion, J et al. (2012) CLOCK, PER2 and BMAL1 DNA methylation: association with obesity and metabolic syndrome characteristics and monounsaturated fat intake. Chronobiol Int 29, 11801194.CrossRefGoogle ScholarPubMed
2. Bandin, C, Scheer, FA, Luque, AJ et al. (2015) Meal timing affects glucose tolerance, substrate oxidation and circadian-related variables: a randomized, crossover trial. Int J Obes (Lond) 39, 828833.Google Scholar
3. Ruiz-Lozano, T, Vidal, J, de Hollanda, A et al. (2016) Timing of food intake is associated with weight loss evolution in severe obese patients after bariatric surgery. Clin Nutr (Epublication ahead of print version).Google Scholar
4. Erren, TC & Reiter, RJ (2009) Defining chronodisruption. J Pineal Res 46, 245247.Google Scholar
5. Garaulet, M, Ordovas, JM & Madrid, JA (2010) The chronobiology, etiology and pathophysiology of obesity. Int J Obes (Lond) 34, 16671683.CrossRefGoogle ScholarPubMed
6. Turek, FW, Joshu, C, Kohsaka, A et al. (2005) Obesity and metabolic syndrome in circadian Clock mutant mice. Science 308, 10431045.Google Scholar
7. Kennaway, DJ, Owens, JA, Voultsios, A et al. (2007) Metabolic homeostasis in mice with disrupted Clock gene expression in peripheral tissues. Am J Physiol Regul Integr Comp Physiol 293, R1528R1537.Google Scholar
8. Lopez-Guimera, G, Dashti, HS, Smith, CE et al. (2014) CLOCK 3111T/C SNP interacts with emotional eating behavior for weight-loss in a Mediterranean population. PLoS ONE 9, e99152.Google Scholar
9. Garaulet, M & Madrid, JA (2009) Chronobiology, genetics and metabolic syndrome. Curr Opin Lipidol 20, 127134.Google Scholar
10. Balsalobre, A, Damiola, F & Schibler, U (1998) A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93, 929937.CrossRefGoogle ScholarPubMed
11. Kondratov, RV, Kondratova, AA, Gorbacheva, VY et al. (2006) Early aging and age-related pathologies in mice deficient in BMAL1, the core component of the circadian clock. Genes Dev 20, 18681873.Google Scholar
12. Sookoian, S, Gemma, C, Gianotti, TF et al. (2008) Genetic variants of Clock transcription factor are associated with individual susceptibility to obesity. Am J Clin Nutr 87, 16061615.Google Scholar
13. Scott, EM, Carter, AM & Grant, PJ (2008) Association between polymorphisms in the Clock gene, obesity and the metabolic syndrome in man. Int J Obes (Lond) 32, 658662.Google Scholar
14. Garaulet, M, Lee, YC, Shen, J et al. (2009) CLOCK genetic variation and metabolic syndrome risk: modulation by monounsaturated fatty acids. Am J Clin Nutr 90, 14661475.CrossRefGoogle ScholarPubMed
15. Garaulet, M, Lee, YC, Shen, J et al. (2010) Genetic variants in human CLOCK associate with total energy intake and cytokine sleep factors in overweight subjects (GOLDN population). Eur J Hum Genet 18, 364369.CrossRefGoogle ScholarPubMed
16. Garaulet, M, Corbalan, MD, Madrid, JA et al. (2010) CLOCK gene is implicated in weight reduction in obese patients participating in a dietary programme based on the Mediterranean diet. Int J Obes (Lond) 34, 516523.CrossRefGoogle Scholar
17. Garaulet, M, Sanchez-Moreno, C, Smith, CE et al. (2011) Ghrelin, sleep reduction and evening preference: relationships to CLOCK 3111T/C SNP and weight loss. PLoS ONE 6, e17435.Google Scholar
18. Garaulet, M, Smith, CE, Gomez-Abellan, P et al. (2014) REV-ERB-ALPHA circadian gene variant associates with obesity in two independent populations: mediterranean and North American. Mol Nutr Food Res 58, 821829.Google Scholar
19. Woldt, E, Sebti, Y, Solt, LA et al. (2013) Rev-erb-alpha modulates skeletal muscle oxidative capacity by regulating mitochondrial biogenesis and autophagy. Nat Med 19, 10391046.CrossRefGoogle ScholarPubMed
20. Partonen, T, Treutlein, J, Alpman, A et al. (2007) Three circadian clock genes Per2, Arntl, and Npas2 contribute to winter depression. Ann Med 39, 229238.Google Scholar
21. Garaulet, M, Corbalan-Tutau, MD, Madrid, JA et al. (2010) PERIOD2 variants are associated with abdominal obesity, psycho-behavioral factors, and attrition in the dietary treatment of obesity. J Am Diet Assoc 110, 917921.Google Scholar
22. Garaulet, M, Smith, CE, Hernandez-Gonzalez, T et al. (2011) PPARgamma Pro12Ala interacts with fat intake for obesity and weight loss in a behavioural treatment based on the Mediterranean diet. Mol Nutr Food Res 55, 17711779.CrossRefGoogle Scholar
23. McCaffery, JM, Papandonatos, GD, Huggins, GS et al. (2013) FTO predicts weight regain in the Look AHEAD clinical trial. Int J Obes (Lond) 37, 15451552.Google Scholar
24. Garaulet, M, Esteban Tardido, A, Lee, YC et al. (2012) SIRT1 and CLOCK 3111T> C combined genotype is associated with evening preference and weight loss resistance in a behavioral therapy treatment for obesity. Int J Obes (Lond) 36, 14361441.CrossRefGoogle Scholar
25. Corbalan, MD, Morales, EM, Canteras, M et al. (2009) Effectiveness of cognitive-behavioral therapy based on the Mediterranean diet for the treatment of obesity. Nutrition 25, 861869.CrossRefGoogle ScholarPubMed
26. Bandin, C, Martinez-Nicolas, A, Ordovas, JM et al. (2013) Differences in circadian rhythmicity in CLOCK 3111T/C genetic variants in moderate obese women as assessed by thermometry, actimetry and body position. Int J Obes (Lond) 37, 10441050.CrossRefGoogle ScholarPubMed
27. Bandin, C, Martinez-Nicolas, A, Ordovas, JM et al. (2014) Circadian rhythmicity as a predictor of weight-loss effectiveness. Int J Obes (Lond) 38, 10831088.CrossRefGoogle ScholarPubMed
28. Lopez-Minguez, J, Ordonana, JR, Sanchez-Romera, JF et al. (2015) Circadian system heritability as assessed by wrist temperature: a twin study. Chronobiol Int 32, 7180.CrossRefGoogle ScholarPubMed
29. Vink, JM, Groot, AS, Kerkhof, GA et al. (2001) Genetic analysis of morningness and eveningness. Chronobiol Int 18, 809822.Google Scholar
30. Lavie, P (1985) Ultradian rhythms: gates of sleep and wakefulness. Exp Brain Res 12, 148164.Google Scholar
31. Dashti, HS, Smith, CE, Lee, YC et al. (2014) CRY1 circadian gene variant interacts with carbohydrate intake for insulin resistance in two independent populations: Mediterranean and North American. Chronobiol Int 31, 660667.Google Scholar
32. Zhang, EE, Liu, Y, Dentin, R et al. (2010) Cryptochrome mediates circadian regulation of cAMP signaling and hepatic gluconeogenesis. Nat Med 16, 11521156.Google Scholar
33. Garaulet, M, Gomez-Abellan, P, Rubio-Sastre, P et al. (2015) Common type 2 diabetes risk variant in MTNR1B worsens the deleterious effect of melatonin on glucose tolerance in humans. Metabolism 64, 16501657.Google Scholar
34. Campion, J, Milagro, FI & Martinez, JA (2009) Individuality and epigenetics in obesity. Obes Rev 10, 383392.Google Scholar
35. Sahar, S & Sassone-Corsi, P (2013) The epigenetic language of circadian clocks. In Circadian clocks pp. 2944. Heidelberg: Springer Berlin.Google Scholar
36. Samblas, M, Milagro, FI, Gomez-Abellan, P et al. (2016) Methylation on the circadian gene BMAL1 is associated with the effects of a weight loss intervention on Serum lipid levels. J Biol Rhythms 31, 308317.CrossRefGoogle ScholarPubMed
37. Arble, DM, Bass, J, Laposky, AD et al. (2009) Circadian timing of food intake contributes to weight gain. Obesity (Silver Spring) 17, 21002102.Google Scholar
38. Sherman, H, Genzer, Y, Cohen, R et al. (2012) Timed high-fat diet resets circadian metabolism and prevents obesity. FASEB J 26, 34933502.Google Scholar
39. Corbalan-Tutau, MD, Madrid, JA, Ordovas, JM et al. (2011) Differences in daily rhythms of wrist temperature between obese and normal-weight women: associations with metabolic syndrome features. Chronobiol Int 28, 425433.Google Scholar
Figure 0

Fig. 1. Represents how several impairments in the central clock are related directly with obesity. In experimental animals, these impairments are due to mutations; however in human subjects, mutations are rare and these impairments are caused by human genetic variants. Thus, these human genetic variants may be associated directly with obesity or through different behaviours. Moreover, these genetic variants may interact with different behaviours to influence obesity and weight loss. We can consider several solutions: (a) changing our behaviours, these changes are directly related to a decrease in obesity and an increase in weight loss, or (b) through genetics by changes in our behaviours that may interact with the SNP to produce a decrease in obesity and weight loss.

Figure 1

Table 1. Advices to prevent genetic and behavioural association/interactions for obesity that may be crucial for genetically informed personalised nutrition in obesity treatment.

Figure 2

Table 2. Effect of different behaviours in methylation levels of CLOCK CpG1, which shows the importance of ‘how we eat’

Figure 3

Fig. 2. Two typical examples of how behaviours may influence genetic variants. (a) Shows an example of genetic association between the genetic variant of CLOCK 3111T/C and weight loss, where the C carriers exhibit a greater difficulty in losing weight during 28 weeks treatment in a clinical nutrition centre than TT carriers. (b) Shows an example of gene × behaviour interaction for insulin resistance; it is remarkable that when subjects are CC carriers of CRY1 rs2287161 and also eat high values of carbohydrates, their insulin resistance homeostatic model assessment (HOMA) is higher. However, among G carriers (GG + GC) values of HOMA are independent of carbohydrates intake and they remain constant.

Figure 4

Table 3. Summarises the specific recommendations for the prevention of obesity and metabolic syndrome by improving the circadian system health, based on available scientific evidence proved in epidemiological, clinical and experimental animals (A); on interventional studies in human subjects or animal models (B); and on preliminary studies or low number of studies (C)

Figure 5

Fig. 3. Represents how eating late affects the metabolic characteristics of lean women; showing that, after 1 week of late eating, their metabolic pattern was more similar to the metabolic pattern of obese women.