Overweight and obesity are defined as a BMI of 25–29·9 and >30 kg/m2, respectively, and are positively associated with several chronic diseases, representing a substantial health and economic burden on society( Reference Bray 1 , Reference Roberts and Barnard 2 ). Obesity is a growing concern with the worldwide prevalence of obesity more than doubling between 1980 and 2008( Reference Finucane, Stevens and Cowan 3 ). It has also been shown that BMI increases continuously throughout adulthood( Reference Østbye, Malhotra and Landerman 4 ) suggesting that long-term behavioural changes are required to curtail this pattern of weight gain. This necessitates the study of both lean and overweight populations, to ensure that dietary interventions are applicable for both reducing BMI in overweight and obese populations, and preventing the increase of BMI in lean populations.
Breakfast can be defined as the first meal of the day, consumed within 2 h of waking, before commencing daily activities, and has been suggested to contain 20–35 % of daily estimated energy requirements( Reference Timlin and Pereira 5 ). Breakfast has long been considered an integral part of a healthy balanced diet( Reference Marangoni, Poli and Agostoni 6 ). This is partly due to associations in the literature that show individuals who regularly omit breakfast have a higher BMI( Reference Cho, Dietrich and Brown 7 , Reference Purslow, Sandhu and Forouhi 8 ) and increased prevalence of obesity-related chronic diseases( Reference Timlin and Pereira 5 ), including type-2 diabetes( Reference Mekary, Giovannucci and Willett 9 ) and CHD( Reference Cahill, Chiuve and Mekary 10 ). Despite this, breakfast omission is becoming more common in Western society( Reference Haines, Guilkey and Popkin 11 ), and it was recently reported that 36 % of the UK population either sometimes or always omit breakfast( Reference Reeves, Halsey and McMeel 12 ). Interestingly, a major reason given for omitting breakfast is weight management, which would appear to contradict a proportion of the scientific evidence( Reference Zullig, Ubbes and Pyle 13 ).
Fundamentally, weight gain occurs when energy intake exceeds energy expenditure over a prolonged period of time. Refraining from eating at a prescribed meal time will inevitably create an energy deficit; however, it is thought that the appetite regulatory system will counter perturbations in energy intake, with behavioural and metabolic modifications that target both energy intake and expenditure( Reference Martin, Normand and Sothier 14 ). It is therefore appropriate to determine whether omitting breakfast in the morning will be compensated for by an increased energy intake in both the short and long terms. However, in order to fully understand the influence of breakfast on energy balance, energy expenditure must also be considered, particularly given the overwhelming evidence that physical activity can reduce the risk of developing numerous chronic diseases( Reference Roberts and Barnard 2 ). Inclusion of exercise alongside energy restriction improves the long-term adherence to, and success of a dietary strategy( Reference Franz, VanWormer and Crain 15 ), and therefore it is important to consider the effect that a given dietary strategy, i.e. breakfast omission, has on the ability and willingness to perform exercise. This may be particularly relevant for exercisers concerned with both weight management and maximising exercise performance.
Whilst the efficacy of controlling energy intake via breakfast omission appears to contradict scientific evidence, individuals who regularly consume breakfast often exhibit other healthy lifestyle factors, such as increased physical activity( Reference Wyatt, Grunwald and Mosca 16 ), improved dietary profiles( Reference Galvin, Kiely and Flynn 17 ) and reduced consumption of snacks( Reference O'Connor, Conner and Jones 18 ). Therefore, it is difficult to determine whether improved weight control is mediated through breakfast consumption per se, or whether this may be the result of other lifestyle factors. A recent study also found that presumptions and beliefs about the importance of breakfast on health may predispose studies to biased reporting, further confounding the matter( Reference Brown, Brown and Allison 19 ). This demonstrates a need for causal data from breakfast intervention trials, and a number of studies have recently been performed, helping to elucidate causal links between breakfast and energy balance.
To comprehensively examine energy balance, several determinants of energy intake and energy expenditure must be considered. Whilst absolute energy consumed can be assessed through ad libitum buffet meals and food records, energy intake data collected by food records should be treated with caution( Reference Livingstone, Prentice and Strain 20 ). In addition, the evaluation of breakfast on appetite control has important implications for energy intake. Appetite regulation can be assessed by subjective appetite sensations, as well as through alterations in peripheral appetite hormone concentrations, and these measures might provide insight into the potential effect of breakfast omission outside of rigid experimental control (i.e. laboratory confinement, defined eating times, etc.). Similarly, energy expenditure can be determined from BMR or RMR, dietary-induced thermogenesis and physical activity energy expenditure (AEE). However, perception of effort and performance during exercise are also important considerations that may influence adherence to exercise training and consequently energy expenditure in the long term. This review therefore aims to summarise scientific evidence from breakfast intervention studies, investigating breakfast omission or consumption and energy balance in adults, and will attempt to identify causal mechanisms between morning eating behaviour and energy balance. For the purpose of this review paper, we defined breakfast as containing ⩾10 % of daily estimated energy requirements.
Effect of breakfast on subjective appetite
It is generally believed that omission of breakfast will increase appetite, causing overeating at subsequent meals, resulting in weight gain( Reference Pereira, Erickson and McKee 21 ). In research, visual analogue scales have been used to assess appetite sensations, such as hunger and fullness, and can provide a reproducible and reliable assessment of subjective appetite( Reference Flint, Raben and Blundell 22 ). As would be expected, a well-established pattern of appetite suppression has been observed during the morning when breakfast is consumed, compared with when breakfast is omitted( Reference Astbury, Taylor and Macdonald 23 – Reference Reeves, Huber and Halsey 31 ). However, it is interesting to note that the subjective appetite response to subsequent meals appears to be unaffected by prior omission of breakfast, suggesting that consumption of breakfast only provides a transient suppression of appetite( Reference Astbury, Taylor and Macdonald 23 , Reference Levitsky and Pacanowski 24 , Reference Chowdhury, Richardson and Tsintzas 26 – Reference Clayton, Stensel and James 29 ). This was recently investigated in two studies that were designed to determine how 24 h subjective appetite profiles responded to breakfast consumption or omission, with ad libitum ( Reference Clayton, Barutcu and Machin 28 ) or standardised( Reference Clayton, Stensel and James 29 ) lunch and dinner meals. In these studies, breakfast (25 % estimated energy requirements) was consumed at 08·00 hours, with lunch and dinner meals consumed at 12·30 and 18·00–19·00 hours. These studies found a reduction in appetite throughout the morning when breakfast was consumed compared with when breakfast was omitted. However, the consumption of an ad libitum lunch meal offset appetite to the same extent, independent of breakfast consumption, and this effect persisted throughout the remainder of the day. The same response was observed when standardised lunch (35 % estimated energy requirements) and dinner (40 % estimated energy requirements) meals were provided, consequently preserving the energy deficit created by breakfast omission. Similar studies have also observed a transient suppression of appetite after breakfast consumption in both lean( Reference Levitsky and Pacanowski 24 , Reference Chowdhury, Richardson and Tsintzas 26 ) and obese( Reference Chowdhury, Richardson and Tsintzas 27 ) subjects, with subjective appetite appearing to be offset independent of breakfast consumption, after an ad libitum lunch meal. Results from these studies demonstrate an imprecise regulation of subjective appetite in response to an energy deficit. However, it should be noted that subjective appetite sensations do not always predict subsequent energy intake( Reference Clayton, Stensel and Watson 32 , Reference James, Funnell and Milner 33 ).
Effect of breakfast on peripheral appetite hormones
Part of the regulation of appetite may involve several gut peptides, including the appetite stimulatory hormone ghrelin and hormones associated with satiation and satiety, such as peptide YY (PYY), glucagon-like peptide-1 (GLP-1), glucose-dependent insulinotropic polypeptide, cholecystokinin and leptin. Understanding the response of these hormones to energy balance fluctuations may provide valuable information about whether dietary interventions (such as breakfast omission) will be tolerable outside of the laboratory environment. Astbury et al. ( Reference Astbury, Taylor and Macdonald 23 ) found that anorexigenic hormones (GLP-1; PYY) were greater up to 30 min after consumption of a 1050 kJ liquid meal 2·5 h after breakfast consumption, compared with after breakfast omission. However, no differences in the orexigenic hormone ghrelin were observed. Additionally, breakfast omission caused an increase in glucose and insulin response to the liquid meal, compared with breakfast consumption. This dampened glycaemic response to the second meal of the day, is known as the ‘second meal effect’ which may be related to glycogen storage( Reference Jovanovic, Leverton and Solanky 34 ). Gonzalez et al. ( Reference Gonzalez, Veasey and Rumbold 25 ) similarly found a tendency for an increased glucose and insulin response to a 1500 kJ liquid meal consumed 3 h after omission, compared with consumption of breakfast, although active GLP-1 concentrations were not different between trials. The different GLP-1 findings may be due to whether total( Reference Astbury, Taylor and Macdonald 23 ) or active GLP-1( Reference Gonzalez, Veasey and Rumbold 25 ) was measured. Additionally, Thomas et al. ( Reference Thomas, Higgins and Bessesen 30 ) examined whether habitual breakfast patterns influence the hormonal regulation of appetite, in response to a standard lunch consumed 4 h after breakfast consumption/omission. Ghrelin concentrations were not affected by the omission or consumption of breakfast, but elevated concentrations of PYY and GLP-1 were reported. Additionally, this study found that the glycaemic response to a standardised lunch was attenuated in habitual breakfast omitters, suggesting some metabolic adaptation may occur over time. In a similar study, Clayton et al. ( Reference Clayton, Stensel and James 29 ) found no difference in acylated ghrelin or GLP-1 concentrations either 1·5 or 3·5 h after a standardised lunch. Collectively, these studies suggest breakfast minimally affects the orexigenic appetite hormone ghrelin, with some evidence that breakfast may increases anorexigenic hormone profiles, in response to subsequent standardised feeding. However, breakfast omission may affect eating behaviour, and the provision of standardised meals does not allow for appetite hormone profiles to be assessed under these circumstances.
This was investigated as part of the Bath Breakfast Project( Reference Betts, Thompson and Richardson 35 ). In these studies, the glycaemic, orexigenic and anorexigenic hormonal responses 3 h after breakfast consumption/omission and 3 h after an ad libitum lunch were determined in both lean( Reference Chowdhury, Richardson and Tsintzas 26 ) and obese( Reference Chowdhury, Richardson and Tsintzas 27 ) subjects. Consumption of breakfast suppressed acylated ghrelin, with concomitant increases in PYY, GLP-1, insulin and glucose, compared with breakfast omission, in both lean( Reference Chowdhury, Richardson and Tsintzas 26 ) and obese( Reference Chowdhury, Richardson and Tsintzas 27 ). After an ad libitum lunch, elevated concentrations of PYY were maintained although no differences in GLP-1 (measured in lean group only) were observed. Paradoxically, acylated ghrelin concentrations were greater in the breakfast consumption trial after lunch in both the lean and obese groups. Similar responses have been reported in another study, with breakfast consumption in the morning appearing to have no effect on acylated ghrelin or GLP-1, 4 h after an ad libitum lunch( Reference Clayton, Barutcu and Machin 28 ).
Further research is required before definitive conclusions can be made regarding the effects of breakfast on the hormonal regulation of appetite. Current research suggests that hormonal markers of appetite are transiently suppressed by breakfast, but differences between breakfast omission/consumption appear to be diminished following lunch, which is in line with subjective appetite sensations. This results in similar hormone concentrations in the afternoon, independent of breakfast consumption. However, there is some evidence of a prolonged anorexigenic response to breakfast, particularly with PYY. Further research is required to determine the long-term effect of breakfast on the hormonal regulation of appetite. This has been partly addressed, with one study finding impaired post-prandial insulin sensitivity after 2 weeks of breakfast omission( Reference Farshchi, Taylor and Macdonald 36 ) and another finding no change in fasted insulin sensitivity as well as no difference in fasted acylated ghrelin, PYY, GLP-1 or leptin after 6 weeks of either consuming or omitting breakfast( Reference Betts, Richardson and Chowdhury 37 ).
Effect of breakfast on ad libitum energy intake
Single exposure studies
The association of regular breakfast omission with a higher BMI( Reference Cho, Dietrich and Brown 7 , Reference Purslow, Sandhu and Forouhi 8 ) has led to the widespread belief that breakfast omission causes overeating at subsequent meals and greater daily energy intake( Reference Pereira, Erickson and McKee 21 ). However, the weight of evidence from well-controlled laboratory intervention studies (Table 1) does not support this belief( Reference Levitsky and Pacanowski 24 – Reference Clayton, Barutcu and Machin 28 , Reference Hubert, King and Blundell 38 ). The majority of single exposure studies have reported either no difference( Reference Levitsky and Pacanowski 24 , Reference Gonzalez, Veasey and Rumbold 25 , Reference Chowdhury, Richardson and Tsintzas 27 ), or an increase( Reference Astbury, Taylor and Macdonald 23 , Reference Levitsky and Pacanowski 24 , Reference Chowdhury, Richardson and Tsintzas 26 , Reference Clayton, Barutcu and Machin 28 , Reference Hubert, King and Blundell 38 ) in energy intake, at the first meal consumed after breakfast (i.e. lunch). However, with the exception of one study( Reference Astbury, Taylor and Macdonald 23 ) the increase in energy intake at lunch was not sufficient to fully compensate for the energy omitted at breakfast, resulting in a reduced gross energy intake (i.e. breakfast + lunch energy intake)( Reference Levitsky and Pacanowski 24 – Reference Clayton, Barutcu and Machin 28 , Reference Hubert, King and Blundell 38 ). With the exception of Astbury et al. ( Reference Astbury, Taylor and Macdonald 23 ), who reported 78 % compensation at lunch for the energy omitted at breakfast, studies have generally reported compensation in the range 0–35 %( Reference Levitsky and Pacanowski 24 – Reference Clayton, Barutcu and Machin 28 , Reference Hubert, King and Blundell 38 ). The amount of compensation observed at lunch might, in part, be related to the energy content of the breakfast supplied. Consuming a low-energy breakfast has been shown to be more accurately compensated for at subsequent meals( Reference Schusdziarra, Hausmann and Wittke 39 ) and might explain why Astbury et al.( Reference Astbury, Taylor and Macdonald 23 ) observed almost complete compensation, whilst others reported much less compensation( Reference Levitsky and Pacanowski 24 – Reference Clayton, Barutcu and Machin 28 , Reference Hubert, King and Blundell 38 ). Whilst it may be possible to increase food intake to compensate for a small energy deficit, a certain threshold may exist, above which complete energetic compensation at a subsequent meal (or meals) is unlikely.
Where available, energy intake at breakfast is presented as mean (sd). Otherwise, mean or absolute intake is presented, as appropriate.
BO, breakfast omission; BC, breakfast consumption; EI, energy intake; AL, ad libitum; CHO, carbohydrate.
* This study compared a very small with a large breakfast, rather than the complete omission of breakfast.
Two studies( Reference Levitsky and Pacanowski 24 , Reference Clayton, Barutcu and Machin 28 ) also assessed energy intake beyond a single meal (Table 1). Consistent with other findings, an increase in energy intake was observed at lunch following the omission of breakfast. Both studies also found no additional energetic compensation occurred at subsequent eating occasions, and therefore gross energy intake (including breakfast) was reduced by 1885( Reference Levitsky and Pacanowski 24 ) and 2740 kJ( Reference Clayton, Barutcu and Machin 28 ) following breakfast omission. Similarly, Thomas et al. ( Reference Thomas, Higgins and Bessesen 30 ) also reported no difference in energy intake at an ad libitum dinner, provided 5 h after a standardised lunch, independent of breakfast consumption in the morning. In this study, gross energy intake was reduced by approximately 710 kJ when breakfast was omitted, but this did not reach statistical significance. Collectively, these studies suggest that energy intake is not accurately regulated in the short term( Reference Levitsky 40 ), and that omission of a single breakfast meal is unlikely to lead to compensation later in the day.
Multiple exposure studies
In a descriptive study, Schusdziarra et al. ( Reference Schusdziarra, Hausmann and Wittke 39 ) measured energy intake of 380 subjects over 10 d, finding that daily energy intake was associated with the amount of energy consumed at breakfast. Specifically, lower energy intake at breakfast was indicative of a reduced daily energy intake. A number of intervention studies have investigated breakfast omission over longer periods of time, often using food records to estimate daily energy intake (Table 2). Whilst the results of these studies are slightly more varied, once again the weight of evidence suggests that omission of breakfast in the morning will reduce daily energy intake in the longer term( Reference Martin, Normand and Sothier 14 , Reference Betts, Richardson and Chowdhury 37 , Reference Reeves, Huber and Halsey 41 ). In one of these studies, a reduction in energy intake was observed in a 6 week between groups breakfast intervention study. Subjects were instructed to consume ⩾2930 kJ before 11·00 hours, or abstain from food completely until 12·00 hours. Timing, type and quantity of foods ingested after 12·00 hours were unaffected by consumption or omission of breakfast in the morning, resulting in a reduced energy intake of approximately 2300 kJ/d when breakfast was omitted( Reference Betts, Richardson and Chowdhury 37 ).
Where available, energy intake at breakfast is presented as mean (sd). Otherwise, mean or absolute intake is presented, as appropriate.
BO, breakfast omission; BC, breakfast consumption; EI, energy intake; AL, ad libitum.
* This study compared a very small with a large breakfast, rather than the complete omission of breakfast.
In contrast to this, Halsey et al. ( Reference Halsey, Huber and Low 42 ) found no difference in daily energy intake, independent of consumption or omission of an ad libitum breakfast. In a study designed primarily to investigate glycaemic control, Farshchi et al. ( Reference Farshchi, Taylor and Macdonald 36 ) found that daily energy intake was increased during 2 weeks of breakfast omission, compared to breakfast consumption. In this study, the authors balanced energy intake in both conditions by providing cereal and milk at a traditional breakfast time (07·00–08·00 hours; breakfast consumption) or later in the day (12·30 hours; breakfast omission). A chocolate-covered cookie was also consumed at 10·30 hours on both trials, and therefore subjects only fasted about 2·5 h longer during the breakfast omission period. The study was designed this way to determine whether the timing of food intake influenced glycaemic control and energy intake, independent of the amount of energy consumed. The experimental design may at least partially explain why the results of this study differ from the majority of the literature.
Surprisingly, there is a sparsity of studies that have investigated breakfast omission in overweight or obese individuals. One study investigated whether daily meal pattern would affect energy intake in obese subjects. Meals were provided as either six meals daily (constituting 4200 kJ) or, four meals daily (constituting 2800 kJ), with the two remaining meals omitted during the morning requiring subjects to fast until 12·00 hours. In addition to the provided meals, subjects were permitted to eat ad libitum after 13·00 hours. This study found a non-significant reduction (approximately 960 kJ) in daily energy intake when meals were provided as four meals daily( Reference Taylor and Garrow 43 ). Reeves et al. ( Reference Reeves, Huber and Halsey 41 ) reported that during 1 week of breakfast omission, energy intake was increased between 12·00 and 18·00 hours in lean subjects and between 12·00 and 21·00 hours in overweight subjects, compared with during 1 week of breakfast consumption. Furthermore, habitual breakfast omitters consumed more after 21·00 hours than habitual breakfast consumers. Despite differing eating patterns, absolute energy intake was reduced by approximately 670 kJ/d during breakfast omission compared with breakfast consumption.
Although not directly assessing energy intake, three further studies assessed the impact of breakfast on weight loss in overweight and obese subjects( Reference Geliebter, Astbury and Aviram-Friedman 44 , Reference Schlundt, Hill and Sbrocco 45 ). Schlundt et al. ( Reference Schlundt, Hill and Sbrocco 45 ) investigated a prescribed energy restricted diet in two groups, with equal energy provisions provided in either two (breakfast omission) or three (breakfast consumption) meals daily. Whilst subjects in both groups lost weight, no difference in weight loss was observed between groups after 12 weeks. The authors also stratified subjects according to their habitual breakfast habits, and found that subjects who changed their breakfast habits lost more weight than those who maintained their breakfast habits. This suggests that the success of a dietary regime might be governed, in part, by the degree in which that regime differs to an individual's normal dietary behaviour. However, this study involved a degree of dietary restriction beyond the consumption or omission of breakfast in the morning, and as such may not reflect true alterations in eating behaviour. Dhurandhar et al. ( Reference Dhurandhar, Dawson and Alcorn 46 ) investigated the effect of recommendations to consume or omit breakfast, in free-living adults attempting to lose weight. Subjects (n 283) were randomly assigned to either consume or omit breakfast for 16 weeks, and results were compared with a control group. Although subjects in this study were attempting to lose weight, in contrast to Schlundt et al. ( Reference Schlundt, Hill and Sbrocco 45 ) this study did not impose any dietary restraint on subjects after 11·00 hours. Results found that either consuming or omitting breakfast did NS affect weight change over a 16 week period( Reference Dhurandhar, Dawson and Alcorn 46 ). In another study, Geliebter et al. ( Reference Geliebter, Astbury and Aviram-Friedman 44 ) found that 4 weeks consuming water in the morning (i.e. breakfast omission) reduced body weight to a greater extent than when 1470 kJ high- or low-fibre breakfasts were consumed.
Overall, these findings do not support the notion that omission of breakfast causes overeating at subsequent meals. Indeed several studies have found that energy intake is not sufficiently increased to compensate for omission of breakfast in the morning, therefore at least partially preserving the energy deficit achieved by breakfast omission.
Effect of breakfast on energy expenditure
It is interesting to note that some of the aforementioned longer term breakfast intervention studies have failed to observe a reduction in body weight( Reference Betts, Richardson and Chowdhury 37 , Reference Schlundt, Hill and Sbrocco 45 ), despite observing reductions in energy intake when breakfast is omitted. This may be due to underreporting of energy intake as opposed to a genuine reduction( Reference Livingstone, Prentice and Strain 20 ), but also could suggest an interaction between breakfast and energy expenditure. The intake of food in the morning will inevitably increase morning energy expenditure due to an increase in dietary-induced thermogenesis( Reference Westerterp 47 ). Consumption of breakfast has been shown to increase RMR during the morning, compared with when no breakfast was provided( Reference Martin, Normand and Sothier 14 , Reference Clayton, Stensel and James 29 , Reference Thomas, Higgins and Bessesen 30 ). Beyond lunch, breakfast does not appear to affect RMR, whether the energy deficit from breakfast omission is recovered at subsequent meals( Reference Kobayashi, Ogata and Omi 48 ) or if the energy deficit is sustained by matching energy intake at lunch across trials( Reference Clayton, Stensel and James 29 , Reference Thomas, Higgins and Bessesen 30 ). Typically, daily energy expenditure has been assessed using a calorimetry chamber( Reference Taylor and Garrow 43 , Reference Kobayashi, Ogata and Omi 48 ) or indirect calorimetry( Reference Martin, Normand and Sothier 14 , Reference Clayton, Stensel and James 29 , Reference Thomas, Higgins and Bessesen 30 ). However, the most malleable component of energy expenditure, physical activity, may be underestimated from these studies, as confined testing spaces and experimental control is likely to restrict free-living physical activity.
Wyatt et al. ( Reference Wyatt, Grunwald and Mosca 16 ) administered physical activity questionnaires during a cross-sectional study, and reported an association between breakfast consumption and greater physical activity. However, there are very few studies that have directly investigated the effect of breakfast on physical activity, particularly in adults. Two studies used pedometers to estimate free-living physical activity and found no difference after 1 week of breakfast consumption or omission( Reference Reeves, Huber and Halsey 31 , Reference Halsey, Huber and Low 42 ). Stote et al. ( Reference Stote, Baer and Spears 49 ) used accelerometers to estimate physical activity, and similarly found no difference when food was provided as one evening meal or three (breakfast/lunch/dinner) meals daily. Verboeket-van der Venne et al. ( Reference Verboeket-van de Venne, Westerterp and Kester 50 ) used doubly labelled water to determine energy expenditure, and also found no difference in AEE when energy was provided in two or seven meals daily. Whilst these studies provide some information about free-living physical activity, the methodology employed in the studies limits their interpretation or makes it difficult to apply the findings directly to breakfast habits. The measurement tools used in some of these studies( Reference Reeves, Huber and Halsey 31 , Reference Halsey, Huber and Low 42 , Reference Stote, Baer and Spears 49 ) may lack reliability and sensitivity when applied to free-living environments( Reference Corder, Ekelund and Steele 51 ) or these studies have assessed daily meal patterns( Reference Stote, Baer and Spears 49 , Reference Verboeket-van de Venne, Westerterp and Kester 50 ), as opposed to the consumption or omission of breakfast.
Recently, the first data suggesting that breakfast consumption in the morning may have a causal effect on AEE in adults was reported( Reference Betts, Richardson and Chowdhury 37 ). Using a combined heart rate and accelerometer device, this study found a reduction in AEE, attributable primarily to a decline in light intensity physical activity, during 6 weeks of breakfast omission compared with breakfast consumption( Reference Betts, Richardson and Chowdhury 37 ). Whilst this reduction in AEE (1885 kJ/d) was not sufficient to fully offset the decrease in energy intake (2300 kJ/d), this study does demonstrate that appreciation of both energy intake and energy expenditure is essential to fully understand the implications of breakfast omission on energy balance. No detectable changes in body weight were observed between groups over the 6 week intervention period, and therefore future studies should look to extend this monitoring period, to determine whether any further behavioural changes occur.
Effect of breakfast on exercise performance
It has been reported that weight loss interventions that combine both dietary restriction and exercise are more successful in the long term( Reference Franz, VanWormer and Crain 15 ). Therefore, it is important to consider the effect that a given dietary intervention, such as breakfast omission, has on the capability and desire to perform exercise. Traditional western breakfasts tend to be high in carbohydrate, and previous studies have observed that omission of breakfast alters dietary profiles, primarily through a reduction in daily carbohydrate intake( Reference Deshmukh-Taskar, Nicklas and O'Neil 52 , Reference Shriver, Betts and Wollenberg 53 ). Therefore, it appears that breakfast could play a crucial role in meeting daily carbohydrate requirements and thus maximising carbohydrate availability( Reference Williams and Lamb 54 ). Whilst individuals concerned purely with weight management may not be overly concerned about carbohydrate availability, consuming adequate carbohydrate is of primary importance to individuals wanting to maximise athletic performance( Reference Cermak and Van Loon 55 , Reference Ormsbee, Bach and Baur 56 ).
Several studies have demonstrated that consumption of carbohydrate in the morning can improve exercise performance compared with performing exercise in the overnight fasted state( Reference Neufer, Costill and Flynn 57 – Reference Chryssanthopoulos, Williams and Nowitz 63 ). However, the majority of these studies provided carbohydrate drinks rather than a typical breakfast meal, and therefore may not accurately represent breakfast consumption and omission per se. Chryssanthopoulos et al. ( Reference Chryssanthopoulos, Williams and Nowitz 63 ) demonstrated that consumption of a high-carbohydrate breakfast meal 3 h before exercise, increased exercise capacity by approximately 9 % compared with when no breakfast was provided. This would likely be due to the effect of an overnight fast on glycogen stores. An overnight fast results in a substantial (approximately 40 %) reduction in liver glycogen( Reference Nilsson and Hultman 64 ), therefore decreasing endogenous glucose availability. Consumption of a high-carbohydrate breakfast will replenish liver glycogen content( Reference Hawley and Burke 65 ), and has also been shown to increase muscle glycogen concentrations by 11–17 %( Reference Chryssanthopoulos, Williams and Nowitz 66 , Reference Wee, Williams and Tsintzas 67 ). Therefore, the omission of breakfast may limit glycogen availability for muscle metabolism, and potentially reduce exercise performance( Reference Coyle and Coggan 68 ).
This evidence would suggest that individuals performing exercise in the morning should aim to consume breakfast between 1 and 4 h before exercise in order to avoid any decrements in exercise performance. However, it has been reported that exercise in the evening may be more acceptable and tolerable than exercise in the morning( Reference Maraki, Tsofliou and Pitsiladis 69 ), suggesting that this may be a more preferable time to exercise for those exercising for weight management. Therefore, it would be of interest to determine whether the detrimental effect of breakfast omission on exercise performance is exclusive to the morning, or whether these effects continue throughout the day. This has been determined in one study, in which breakfast was provided or omitted at 08·00 hours, with an ad libitum lunch provided at 12·30 and a 60 min bout of exercise performed at 17·00 hours. Despite the breakfast intervention occurring 9 h before exercise and consumption of an ad libitum lunch 4·5 h prior to exercise in both trials, a 4·5 % decrement of exercise performance was observed after breakfast omission( Reference Clayton, Barutcu and Machin 28 ). This data suggest that the effects of breakfast may be prolonged and therefore consumption of breakfast appears to be important for individuals aiming to maximise exercise performance.
The weight of evidence from breakfast intervention studies suggests that the omission of breakfast may lead to an increase in energy intake at subsequent meals. However, this increase has rarely been shown to fully compensate for the energy omitted at breakfast, and therefore daily energy intake appears to be reduced when breakfast is omitted from the daily meal pattern. Additionally, whilst the omission of breakfast will lead to an increase in subjective appetite and associated hormonal profiles during the morning, this appears to be transient, and can be offset by the first meal consumed after breakfast. This suggests that food intake is not accurately regulated, at least in the short term, and the observed trend for a higher BMI amongst breakfast omitters does not appear to be augmented by dietary compensation. There is some evidence to suggest an interaction between breakfast and energy expenditure, specifically, reductions in low-intensity physical activity have been observed. However, over a 6 week period, this reduction in energy expenditure did not completely offset the reduction in energy intake produced by breakfast omission, therefore generating a state of negative energy balance. Longer term intervention studies are required to confirm whether this negative energy balance is maintained, and whether this would lead to meaningful weight loss. Finally, omission of breakfast appears to reduce exercise performance throughout the day and therefore individuals aiming to maximise exercise performance should consume breakfast.
This research was supported by the National Institute for Health Research (NIHR) Diet, Lifestyle & Physical Activity Biomedical Research Unit based at University Hospitals of Leicester and Loughborough University. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.
Conflicts of Interest
D. J. C. conceived and wrote the manuscript with assistance from L. J. J. Both authors approved the final manuscript.