The scientific study of obesity has been dominated throughout the twentieth century by the concept of energy balance. This conceptual approach, based on fundamental thermodynamic principles, states that energy cannot be destroyed, and can only be gained, lost or stored by an organism. Its application in obesity research has emphasised excessive appetite (gluttony), or insufficient physical activity (sloth), as the primary determinants of excess weight gain, reflected in current guidelines for obesity prevention and treatment. This model cannot explain why weight accumulates persistently rather than reaching a plateau, and underplays the effect of variability in dietary constituents on energy and intermediary metabolism. An alternative model emphasises the capacity of fructose and fructose-derived sweeteners (sucrose, high-fructose corn syrup) to perturb cellular metabolism via modification of the adenosine monophosphate (AMP)/adenosine triphosphate (ATP) ratio, activation of AMP kinase and compensatory mechanisms, which favour adipose tissue accretion and increased appetite while depressing physical activity. This conceptual model implicates chronic hyperinsulinaemia in the presence of a paradoxical state of ‘cellular starvation’ as a key driver of the metabolic modifications inducing chronic weight gain. We combine evidence from in vitro and in vivo experiments to formulate a perspective on obesity aetiology that emphasises metabolic flexibility and dietary composition rather than energy balance. Using this model, we question the direction of causation of reported associations between obesity and sleep duration or childhood growth. Our perspective generates new hypotheses, which can be tested to improve our understanding of the current obesity epidemic, and to identify novel strategies for prevention or treatment.
The scientific study of obesity has been dominated throughout the twentieth century by the concept of energy balance. This conceptual approach, based on fundamental thermodynamic principles, states that energy cannot be destroyed, and can only be gained, lost or stored by an organism (Hess, 1838; von Helmholtz, 1847). Its application in obesity research has emphasised excessive appetite, or insufficient physical activity, as the primary determinants of the modification of energy balance leading to excess weight gain (Prentice and Jebb, 1995), reflected in current protocols for obesity prevention and treatment. As we review below, this model cannot explain why weight accumulates persistently in individuals, rather than reaching a plateau when weight gain re-establishes the balance between energy intake and expenditure. The energy balance approach also underplays the effect of particular dietary components (for example, carbohydrates, amino-acids and fatty acids) on energy metabolism and fuel oxidation.
Much obesity research, summarised below, has therefore focused on the link between metabolism and excess weight gain. Such research has focused on the hormonal basis of appetite regulation, digestion, molecular signalling and cellular energetics, approaches that collectively also aid understanding of the effects of the various genetic factors that have been associated with obesity risk. For example, the ‘metabolic flexibility’ model emphasises the capacity to switch efficiently between glucose and fatty acids as the main cellular fuels, in the maintenance of energy homeostasis during perturbations of energy supply/demand or manipulations of dietary macronutrient composition (Storlien et al., 2004). Both a diet rich in refined carbohydrate and sedentary behaviour are proposed to perturb this metabolic flexibility (Corpeleijn et al., 2009). Chronic elevated insulin levels in response to a diet high in refined carbohydrate have been proposed to represent a core mechanistic feature of persistent excess weight gain in this conceptual model (Lustig 2006a, 2008; Taubes, 2008), potentially generating a novel approach to obesity prevention and treatment.
We review and develop this discussion in several ways, by describing the mechanisms relating to fructose/sucrose intake and their effects on insulin and intermediate metabolism. Building on the work of others, we consider how the energy balance model may have overemphasised certain environmental factors relevant to the aetiology of obesity, potentially reducing the likelihood of developing appropriate public health campaigns to address the rising prevalence. We then consider two specific examples (sleep and growth) where the apparent direction of causation with obesity may be the reverse of that which is commonly assumed.
The aim of our review is not to claim that our revised model is correct or definitive, but rather to stimulate researchers to acknowledge that the direction of causation in obesity aetiology is less clear than is commonly assumed. Nothing that we propose in this article contradicts the laws of thermodynamics or the energy balance equation; rather, we follow others in questioning whether simplified interpretation of the energy balance equation can become misleading with regard to obesity causation. We suggest that if empirical work specifically addresses this dilemma, regardless of which model proves most compatible with data, the evidence base for obesity prevention is likely to improve.
The concept of energy balance has had an important role in many aspects of physiological research, and thermodynamic principles underlie diverse methodologies for studying energy metabolism and body composition (Hess, 1838; von Helmholtz, 1847). Its central role in the study of human obesity has remained essentially unquestioned: excess weight gain must co-occur with positive energy balance, when energy intake exceeds energy expenditure:
This simple mathematical equation seemingly directs attention to two essential causes of obesity: chronic overeating or chronic low-energy expenditure, usually considered to derive from sedentary behaviour. These supposed causes are often attributed to individual responsibility by being labelled as ‘gluttony’ or ‘sloth’ (Prentice and Jebb, 1995).
Numerous epidemiological or physiological studies have attempted to identify which of ‘gluttony’ or ‘sloth’ is the primary driver of the current obesity epidemic. Some favour the hypothesis that an energy-dense diet, or poor regulation of appetite, drives over-eating (Blundell et al., 1996; Prentice and Jebb, 2003; Woods and D’Alessio, 2008). Others favour the notion that physical activity levels have declined in recent decades (Ferro-Luzzi and Martino, 1996; Bar-Or et al., 1998; Goran and Treuth, 2001), such that the food consumed is not being burned off by sufficient physical work. In terms of evidence, there is marginally more data in favour of the first than the second of these hypotheses, but neither is compelling. Energy intakes are reported to have declined in some populations (Briefel and Johnson, 2004), but are notoriously difficult to assess, while objective data on physical activity trends are sparse, and the evidence remains unconvincing (Westerterp and Speakman, 2008). One recent study attempted to calculate retrospectively the ‘energy balance gap’ on the basis of isotopic measurements of energy expenditure, and concluded that substantial secular increases in energy intake were responsible (Swinburn et al., 2009). Those working at the mechanistic level increasingly direct their attention to the brain and appetite dysregulation (Schwartz and Morton, 2002). In support of this hypothesis is the fact that rare genetic mutations associated with obesity tend to impact on appetite regulation pathways (O’Rahilly et al., 2003), while commoner genetic variants such as the FTO gene also impact on appetite (Speakman et al., 2008).
The concept of energy balance is likewise fundamental to many current approaches to obesity treatment. Dietary management attempts to reduce energy intake while holding energy expenditure constant, whereas the promotion of exercise attempts to increase energy expenditure while holding energy intake constant, or simultaneously reducing it. Each of these approaches targets voluntary behaviour, indicating a belief that energy stores can readily be manipulated by purposive changes in eating patterns or leisure habits. Behavioural obesity prevention uses the same approach, and simply attempts to avoid the onset of ‘gluttony’ or ‘sloth’. Pharmaceutical management in contrast directly targets metabolism, seeking for example to block fat absorption or signals of appetite or satiety (Padwal and Majumdar, 2007), while surgical management seeks to physically constrain the capacity for energy absorption (Maggard et al., 2005) by modifying the anatomy of the gastro-intestinal system. These latter treatments therefore seek to bypass voluntary behaviour. However, for obvious reasons, neither pharmaceutical nor surgical approaches are used in general obesity prevention, which therefore relies entirely on behavioural approaches.
Despite their theoretical logic, neither behavioural prevention nor behavioural treatment is generally successful in practice, and while some programmes generate modest weight loss, bariatric surgery is currently the most effective treatment. The most obvious indication of the failure of behavioural prevention is that obesity is increasing in most populations, and in most age groups (Popkin, 2007). Confronting this scenario, many clinicians question the motivation of individuals to adhere to their treatment regimen. Alarmed at the burgeoning budget implications, governments meanwhile urge people to eat less and exercise more (for example, the UK government's ‘Change4life’ campaign).
The energy balance model has major limitations for ‘explaining’ obesity (Wells, 1998). In the majority of individuals, excess weight gain occurs at an average rate of a few grams per day, equivalent to 1–2 kg per year. Such weight gain is trivial relative to the accuracy and precision of energy metabolism and body composition methodologies, preventing studies from measuring day-to-day changes and attributing obesity specifically to one of the two candidates, ‘gluttony’ or ‘sloth’ (Wells, 1998). Furthermore, it is not clear why such energy imbalance does not simply result in re-stabilisation of body weight at a slightly higher level. Both lean and fat tissue contributes to total energy expenditure (Garby et al., 1988). If energy intake initially exceeds energy expenditure, the resulting increase in body weight should increase energy expenditure until body weight plateaus at a higher level. Several studies have reported high rates of total energy expenditure in obese adults (Prentice et al., 1986) and children (Bandini et al., 1990; Treuth et al., 1998), while also at low levels of physical activity (Ekelund et al., 2002). Why do those with such high absolute energy expenditure continue both to gain weight and to increase their energy intake? Thus, despite decades of research, the energy balance approach has not actually revealed why some people continuously accumulate excess weight.
Thermodynamic theory cannot be wrong, but it is possible that it has been wrongly applied in the scientific study of obesity. Recently, Lustig (2006a, 2008) challenged the direction of causality in the energy balance model, by arguing that hyperinsulinaemia is the primary metabolic derangement leading to excess adiposity, either via an impairment of insulin signalling in the liver, or via a perturbation of signalling of leptin action in the brain. Taubes (2008) developed a similar hypothesis, and argued that the energy balance model detracts attention from links between dietary composition, cellular metabolism and chronic weight gain. Here, we build on their approach by proposing a more comprehensive mechanism whereby contemporary dietary components can determine metabolic, cellular-specific (hepatic, neuronal) perturbations and more prolonged increases in insulin levels. We then consider the implications of this approach for interpreting the epidemiology of human obesity.
Energy balance or energy imbalance
What is remarkable about the development of human obesity is that, over long time periods and in most individuals, millions of calories of energy intake very nearly balance millions of calories expended (Taubes, 2008). The caloric difference in energy intake versus requirements that permits weight gain is proportionally miniscule—for many, it is equivalent to half a biscuit per day, or a few minutes less walking.
Table 1 shows numerically the extent of caloric imbalance required to alter body weight by 1 kg over a specific time period, through changes in physical activity. For example, if a subject on a stable energy intake replaced 1 h of sitting with 1 h of sleeping every day, equivalent to a daily caloric imbalance of 18 kcal (corresponding to the energy contained in a quarter of a digestive biscuit), the cumulative addition of 372 sleeping hours (lasting ∼1 year) would increase body weight by 1 kg. However, given that the extra weight would also increase resting and total energy expenditure, it would in fact take longer to gain 1 kg. At some periods in the life course, and in some individuals, weight increases more rapidly. Puberty, for example, involves rapid weight gain, and has been identified as a ‘critical period’ for the development of obesity because such weight gain may involve substantial increases in adiposity (Dietz, 1994). Any theoretical explanation for obesity must therefore be able to account for both ‘slow-trickle’ and ‘rapid-onset’ pathways to obesity, even if different exposures and metabolic pathways contribute in different individuals and at different times.
Paradoxically, even moderate rates of weight loss are remarkably difficult to achieve through voluntary behaviour. Most people find it difficult to diet, because when doing so they tend to feel tired and hungry (Laessle et al., 1996). When people exercise more, their appetite may increase to compensate for the energy depletion, although this response varies between individuals (Hopkins et al., 2010). The difficulty of voluntarily inducing weight loss indicates that our predisposition to energy balance is far more remarkable than the minor energy imbalance that underlies weight gain (Taubes, 2008). This paradox can be resolved if we shift attention away from energy balance itself, and consider instead some important metabolic mechanisms underlying fat deposition and oxidation.
Metabolic approaches to obesity
Research into the metabolic basis of obesity has a long history, with a notable early contribution being the discovery that the hormone insulin regulates metabolism and promotes the deposition of lipid in adipose tissue (Berson and Yalow, 1965; Ganong, 1999). Since then, a number of different avenues have been explored.
An important breakthrough was the discovery of the hormone leptin in the 1990s, subsequently shown to signal energy stores to the hypothalamus, and to constrain appetite while increasing energy expenditure (Woods and Seeley, 2000). Although rare mutations in the leptin gene are associated with morbid obesity (Montague et al., 1997), the vast majority of obese humans have high leptin levels. This is widely interpreted as obese individuals being resistant to leptin's effects, but whether leptin resistance induces obesity or vice versa remains controversial (Myers et al., 2010). Related work includes a number of studies, which have identified other signalling molecules involved in the regulation of energy metabolism (Schwartz and Morton, 2002; Morrison and Berthoud, 2007). This work also allows investigation of the physiological effects of the various genotypes that have been associated with obesity risk. A number of genetic disorders have now been associated with obesity (Polychronakos and Kukuvitis, 2002; Haig and Wharton, 2003; Walter and Paulsen, 2003), as have a range of metabolic disorders (Reinehr, 2010; Repaci et al., 2011). However, the global obesity epidemic is clearly driven primarily by environmental factors rather than changes in the frequencies of these rare conditions, and changes in lifestyle, whatever their biological mechanism, must play the key role.
Other researchers have focused on the variable effects of different macronutrients on metabolism. Variability in the glycaemic load of foods affects subsequent changes in blood glucose levels, and hence the insulin response (Foster-Powell et al., 2002). Understanding of this effect has led to low glycaemic load diets, whereby high intakes of fat and protein appear to promote satiety (Johnstone et al., 2008). Others have focused on the effects of diets containing fructose, or sucrose-containing drinks, on the predisposition for obesity (Ludwig et al., 2001; Bray et al., 2004; Lustig, 2006b; Stanhope et al., 2009; Bray, 2010). The relationship between diet and weight gain remains controversial, however, because similar levels of weight loss were reported in a study which randomised obese individuals to low-calorie diets of different macronutrient composition (Sacks et al., 2009). One possible explanation is that studies of diet and weight loss, where individuals attempt to comply with a specific regime, may be a problematic model for understanding the role of diet in weight gain in the general population. Another possible explanation is that multiple factors, including genotype and developmental experience, may affect the response to any given diet.
The importance of developmental experience is demonstrated in many studies by an inverse association between central obesity in childhood or adulthood and birth weight, although not all studies show this association (Wells et al., 2007). This inconsistency may arise because whether birth weight induces later adiposity appears to be mediated by the effect of infant weight gain (Stanner and Yudkin, 2001). Those born small who grow fast during childhood have a tendency not only to catch-up in size but to ‘overshoot’ (Ibanez et al., 2006), as discussed below, although there is also heterogeneity in such growth patterns. Thus, early hormonal adaptations arising from such early growth patterns may affect metabolic sensitivity to subsequent diet.
Recently, the effect of gut flora on energy metabolism has also been investigated, highlighting that individuals may differ in their capacity to extract energy from a given dietary intake. Two bacterial colonisers of the human gut, the Bacteroidetes and the Firmicutes, differ in their metabolic efficiency (Turnbaugh et al., 2006), such that obese individuals, who have a higher proportion of Firmicutes, have an enhanced capacity to harvest dietary energy (Ley et al., 2006) and hence gain weight more readily.
This brief summary shows that metabolic causes of weight gain have been investigated through a number of avenues, and at several different levels of biology. The proposition that metabolism influences the propensity for weight gain emerges in each of these contexts, and the discussion of fructose metabolism that follows therefore builds on this previous work.
Metabolic flexibility and the regulation of fuel supply
Notwithstanding the role of the brain in regulating energy balance (Schwartz and Morton, 2002), the dynamic processes regulating cellular metabolism are also critical (Lustig, 2006a; Taubes, 2008). Healthy individuals are characterised by a high metabolic flexibility, which refers to the ability to shift readily between glucose and free fatty acid oxidation as a source of cellular fuel supply during perturbations of energy supply (Storlien et al., 2004). There is increasing evidence that chronic elevation of circulating insulin levels may disrupt this normal flexibility of insulin-responsive (hepatic, muscular and fat) cells (Bickel, 2004).
Insulin is an anabolic hormone. When insulin levels increase in response to raised plasma glucose levels, glucose oxidation is activated to restore cellular energy levels, while the surplus energy is stored via the stimulation of glycogenesis (limited storage capacity) and lipogenesis (unlimited storage capacity) (Saltiel and Kahn, 2001). However, Lustig (2006a) and Taubes (2008) hypothesised that chronically high insulin levels, induced by certain diets, may disrupt this cellular metabolic regulation.
Sucrose, or table sugar, is composed of 50% glucose and 50% fructose. Taubes (2008) proposed that diets high in sucrose induce a state of ‘cellular starvation’, by reducing adenosine triphosphate (ATP) levels and increasing lactate production in fructose-disposing cells (enterocytes and hepatocytes), even when there is plenty of fuel stored in adipose tissue. According to his model, this cellular ‘starvation’ arises because the raised insulin levels prevent lipolysis and fatty acid oxidation, making the energy already available in adipose tissue temporarily ‘invisible’ to cellular energy demand (Taubes, 2008; Tappy and Le, 2010). The elevation of lactate concentration, which follows fructose ingestion increases the rate of hepatic gluconeogenesis, and contributes to the cellular depletion of ATP (Johnson et al., 1989; Caton et al., 2011). As a consequence, high levels of insulin in response to the raised plasma glucose impact indirectly on the brain to generate feelings of hunger and tiredness through a rapid decline in fuel availability for motor activities in the peripheral tissues, while at the same time the fructose generates systemic effects on ATP levels, depleting cellular energy availability (Johnson et al., 1989; Cha et al., 2008).
Recent data provide support for this model, and the similar model of Lustig (2006a), and illustrate in greater detail how diets high in refined carbohydrate perturb metabolic flexibility. Fundamental to this deleterious effect is the complex metabolic profile of sucrose and high-fructose corn syrup (HFCS), which contain 50:50 or 55:45 ratios of fructose to glucose, respectively, and which are commonly used as sweeteners in soft drinks and baking products. These substances are important because they are key to the hyperinsulinaemia that can drive chronic weight accumulation (Lustig, 2006b).
Fructose and metabolic flexibility
The metabolic capacity of the human liver allows rapid processing of glucose, but becomes defective when challenged by large amounts of fructose (Basciano et al., 2005). A physiological model of the effects of fructose on intermediate metabolism is described in Figure 1. At low intakes, fructose has beneficial effects on glucose disposal by increasing glucose uptake via an increased translocation of glucokinase from the inactive-nuclear form to the cytosolic-active form (VanSchaftingen et al., 1997). However, an excess of fructose floods the glycolytic pathway with lipogenic precursors (glycerol-3-phosphate, acetyl-co-A) by bypassing the regulated step in the pathway catalysed by the enzyme phosphofructokinase (Underwood and Newsholme, 1965; Mayes, 1993). Fructose increases the production of lactate by an increased conversion of pyruvate by the enzyme lactic dehydrogenase (Sahebjami and Scalettar, 1971).
The increased availability of lipogenic and gluconeogenic molecules (pyruvate, glycerol phosphate, acetyl-co-A, lactate and fatty acids), lipogenesis activators (sterol receptor element binding protein 1C) and gluconeogenesis activators (sirtuin 1) induce an insulin-independent increase in de novo hepatic lipogenesis and gluconeogenesis, a metabolic effect observed during fructose supplementation experiments in animals (Noguchi and Tanaka, 1995; Commerford et al., 2002; Matsuzaka et al., 2004) and humans (Mayes, 1993; Dirlewanger et al., 2000; Stanhope et al., 2009). In addition, the rapid metabolism of fructose and the activation of energy-demanding processes increase the adenosine monophosphate (AMP):ATP ratio (Leclerc et al., 1998; Muoio et al., 1999; Cha et al., 2008). This paradoxical state of cellular starvation activates the catabolic enzyme AMP-kinase which, in trying to replenish the cellular ATP levels, orchestrates a series of different oxidative mechanisms including glucose and lipid oxidation and the activation of glycogenlysis (Dzamko and Steinberg, 2009). This can further contribute to the reduced availability of metabolic substrates determined by the high insulin levels.
A proportion of the endogenous glucose is then released into the circulation to enrich the already replenished plasma glucose, which prolongs insulin secretion by the pancreatic beta cells. The sustained hyperinsulinaemia elicits a systemic anabolic drive, which stimulates fat accumulation in adipose tissue and skeletal muscle. The paradoxical state of cellular energy depletion may be interpreted by the brain as a state of energy deprivation, increasing appetite and energy intake via an activation of AMP kinase (Cha et al., 2008). In rats, for example, central administration of fructose increases appetite and energy intake (Cha et al., 2008). The rapid decrease in ATP levels associated with fructose metabolism is assumed to activate AMP kinase and inhibit the down-stream production of malonyl-co-A, a known appetite suppressant (Wolfgang et al., 2007). This in turn leads to stimulation of the anorexigenic neurocircuitry (decreased levels of cocaine and amphetamine-regulated transcript and proopiomelanocortin) in the hippocampus, and increased activity of the orexigenic pathway (reduced levels of neuropeptide Y and agouti-related peptide) (Cha et al., 2008).
The long-term metabolic effects of fructose intake are still only partially known. Studies in rats have demonstrated that short-term exposure to a high-fructose diet was fully metabolically compensated, however, the animals became hyperinsulinaemic and insulin resistant after 2–5 weeks exposure (Pagliassotti and Prach, 1995; Pagliassotti et al., 1996). A recent study in overweight/obese human adults over 10 weeks showed that fructose-sweetened, but not glucose-sweetened, beverages increased de novo lipogenesis, promoted dyslipidaemia, decreased insulin sensitivity and increased visceral adiposity (Stanhope et al., 2009). However, while this recent research has focused in particular on fructose, it is dietary sucrose and HFCS that are the likely key culprits in the human diet, because the detrimental effects of fructose are exacerbated in the presence of glucose because of its induction of insulin secretion (Lustig, 2006a; Taubes, 2008). Thus, this approach highlights refined carbohydrates as potentially key to chronic fat deposition.
Reinterpreting the logic of energy imbalance
According to this metabolic perturbation model, ‘gluttony’ and ‘sloth’ need not be causal to excess weight gain. Instead, the causation would work in the opposite direction, with the perturbed cellular metabolism driving fat deposition while simultaneously inducing hunger and tiredness (Lustig, 2006a; Taubes, 2008). This could account for widespread epidemiological associations between apparent ‘obesogenic behaviours’ and obesity, but would make them symptoms of fat deposition, not causes. Low activity levels and hunger would not drive excess weight gain, but represent symptoms of the cellular starvation resulting from an anabolic rather than catabolic metabolism.
The proposition that positive energy imbalance is not causal to weight gain is likely, initially, to strike many obesity researchers as problematic. The problem derives, as discussed above, from the fact that the relationship between positive energy balance and weight gain is a truism, not an explanation, hence no direction of causation can be inferred. We therefore emphasise that reversing the conventional direction of causation does not contradict the energy balance equation, or the laws of thermodynamics. Rather, it re-evaluates the tripartite associations between behaviour, metabolism and weight change, and the key point of contention that emerges is the nature of ‘obesogenic’ behaviour. In Box 1, we illustrate these issues using an analogy of an inanimate vehicle to represent the body. The key issue is whether behaviour drives metabolism, or vice versa.
The contrasting cellular metabolic perturbation and energy balance models of obesity are illustrated in Figure 2. Importantly, the model can also account for varying rates of excess weight gain, according to variability in dietary intake (in particular, refined carbohydrate) and its impact on the degree and persistence of hyperinsulinaemia. Insulin resistance might initially emerge in response to high insulin levels, in order to protect some tissues from glucose/insulin overload (Hoehn et al., 2009). This contrasts with the conventional approach, which assumes compensatory hyperinsulinaemia to develop in response to insulin resistance, but again, the temporal connection between these metabolic processes has yet to be established (Lustig, 2008).
Under the influence of the energy balance model, scientific work on the hormonal basis of weight regulation has yet to reach widespread public understanding, a situation exacerbated by continuing emphasis on calorie counting, or ‘balancing’ dietary intake against physical activity practice. The tendency for treadmills in gyms to calculate calorie use during exercise, and for food labels to report the calorie content, ensures that most dieters still do count their calories. Beyond this, the other main public health message, powerfully supported by the food industry, is that low-fat diets protect against obesity, on the justification that fat is the macronutrient with the highest energy density. This then encourages dieters to consume high-carbohydrate diets. Clinicians are also often sceptical of obese patients who insist ‘it's my glands’, and may privately assume that dietary intakes are grossly under-reported, a hypothesis supported by objective data on energy expenditure obtained using stable isotope probes (Goris et al., 2000). Yet these two facts are not in fact contradictory, if energy intakes excessive to energy requirements are directly induced by the composition of the habitual diet. In this case, the patient is correct in that his or her metabolism does contribute, while the clinician is correct because the patient consumes more calories than are required for weight stability.
Taubes (2008) argued that the notion that excess weight gain is essentially the result of diet-induced hormonal perturbations is consistent with several strands of evidence. First, other states of positive energy balance are strongly hormonally regulated, including pregnancy weight gain (Augustine et al., 2008) and childhood growth (Bogin, 1999). Second, since the 1930s, European clinicians have used insulin and carbohydrate to fatten underweight patients (Taubes, 2008), demonstrating a consistent effect of insulin across a wide range of body weight.
Excessive emphasis on energy balance as the primary ‘explanatory’ approach in obesity research may be a case of the tail wagging the dog. In this case, the dog may have been steadily gaining weight because too much attention has been paid to the tail, and not enough to the dog. We now review two widely reported epidemiological associations, between obesity and (a) sleep duration and (b) childhood growth rate, using the metabolic perturbation model described above to re-examine the direction of causation in these associations.
Obesity and sleep
The increasing number of studies associating sleep patterns with obesity risk has attracted considerable attention from the scientific community. In adults, such associations demonstrate poor consistency between studies, with reports variously of no association between sleep duration and body mass index, U-shaped associations or negative linear associations (Marshall et al., 2008; Patel and Hu, 2008). Longitudinal studies show slightly greater consistency, although again the association may be U-shaped or linear, and the effect appears to indicate increased sleep constraining weight gain, rather than promoting weight loss (Marshall et al., 2008; Patel and Hu, 2008).
In children, however, the evidence is much more consistent, with a wide variety of studies reporting a negative linear association between sleep duration and obesity risk (Marshall et al., 2008; Patel and Hu, 2008). Such studies have been reported for populations in Europe (vonKries et al., 2002; Padez et al., 2009), Canada (Chaput et al., 2006), the United States (Gupta et al., 2002), China (Jiang et al., 2009), Japan (Sekine et al., 2002), Australia (Eisenmann et al., 2006), New Zealand (Nixon et al., 2008) and Brazil (Wells et al., 2008). A meta-analysis of data on over 630 000 children showed that, compared with children sleeping >10 h per night, the odds ratio of obesity was 89% in those sleeping less (Cappuccio et al., 2008). Despite such consistency in the paediatric evidence, these findings are difficult to interpret, and their implications for public health policy remain uncertain.
Inverse associations between sleep duration and obesity risk challenge the ‘sloth’ hypothesis for obesity, because a seemingly higher level of physical activity (longer waking hours) is associated with positive energy balance. One possibility is that those sleeping less compensate by being less physically active during the day, and hence have lower overall energy expenditure per 24-h period. However, in our research in Brazil, we found that those sleeping less were actually more active, according to questionnaire assessments (Wells et al., 2008). The challenge to the ‘sloth’ hypothesis appeared more plausible, following research in adults that linked experimentally imposed short sleep duration with declines in leptin (a marker of satiety) and increases in ghrelin (a marker of appetite) (Spiegel et al., 2004). Similar cross-sectional associations between sleep patterns and hormones emerged in a larger observational cohort (Taheri et al., 2004). Thus, according to this model, low levels of sleep would predispose to weight gain by reducing satiety and increasing hunger, with these hormonal effects outweighing any minor increase in total daily energy expenditure because of longer waking hours. Experimentally imposed sleep debt, for example, has been shown to perturb carbohydrate metabolism, with reduced glucose tolerance and raised evening levels of cortisol (Spiegel et al., 1999).
A causal role for short sleep duration in the population aetiology of childhood obesity might also seem plausible. For example, several studies have reported not only cross-sectional associations of short sleep duration and obesity in adolescents, but also inverse associations between sleep duration and the amount of time spent viewing television (Wells et al., 2008; Padez et al., 2009). These data appear to indicate a replacement of sleep with television viewing, another behaviour widely associated with obesity risk (Robinson, 1999; Ekelund et al., 2006a). In particular, children with a television in their bedroom have been shown to have an increased risk of being overweight (chi-Mejia et al., 2007).
Nevertheless, we suggest caution is still required regarding the direction of causation in the paediatric sleep–obesity association. For some, such caution is advocated purely on account of the small magnitude of the effect. Horne has argued that relatively few individuals are short sleepers, that the magnitude of the effect of short sleep is modest (that is, the majority of short sleepers neither are, nor become, obese), and that interventions to remove the inferred effect of short sleep would make a negligible impact on the epidemic of childhood obesity (Horne, 2008). Against this criticism, Young (2008) has responded that even when the magnitude of an effect is small in individuals, changes in the behaviour of an entire population may nevertheless shift a large number of people away from the ‘danger zone’. He calculated the total proportion of childhood obesity attributable to short sleep to be 5–13%, and suggested that such figures would easily justify a public health policy response.
A biochemical perspective also suggests caution regarding causation is merited. At the cellular level, circadian clocks allow coordination of metabolic activities with diurnal environmental variation (Bray and Young, 2006). The sensitivity of body weight and adiposity to seasonal changes in the duration of the light–dark cycle in animals elegantly illustrates the role of circadian clocks in regulating nutritional status (Bray and Young, 2006). In humans, sleep deprivation induces detrimental effects on the metabolism of carbohydrate (reduced glucose clearance) and lipid (increased triglyceride levels), and night-shift work has been associated with an increased risk of the metabolic syndrome (Karlsson et al., 2001, 2003; Di et al., 2003). But it is equally possible that disruption of metabolism may perturb sleep patterns. For example, adult type 2 diabetics tend to suffer from poor sleep quality and duration (Skomro et al., 2001; Cunha et al., 2008). Thus, although sleep is increasingly associated not only with obesity but also with cardiovascular risk (Knutson, 2010), it remains unclear whether short sleep duration exacerbates these risks, or whether those with perturbed metabolism are prone to each of sleep reduction, weight gain and elevation of metabolic risk. Nor can longitudinal studies necessarily determine which of sleep duration or metabolism, if related cross-sectionally at baseline, is the causal predictor of subsequent weight change.
Three decades ago, animal studies clearly demonstrated that patterns of sleep are strongly associated with the utilisation of circulating metabolites. ‘Sleep is essentially a cyclic behaviour … (and) … the metabolic rate appears to be the primary cue in the onset of feeding’ (Danguir and Nicolaidis, 1980a). Rather than habitual sleep duration representing a behavioural ‘choice’, readily alterable on the basis of education and advice, we should therefore consider whether sleep patterns in the general population, especially children who have yet to develop the metabolic syndrome, might be the product of metabolic regulation.
Comparative studies of a variety of animal species have demonstrated strong correlations between sleep duration and metabolic rate, with small animals having higher metabolic rate also sleeping longer (Zepelin and Rechtschaffen, 1974). Figure 3 shows that sleep duration in mammals decreases with increasing weight, according to the regression equation:
This effect of body size is most probably primarily due to the effect of metabolism, with sleep duration increasing in relation to the rate of oxygen consumption (Figure 4). Such an association between metabolism and sleep duration has been interpreted as sleep functioning to enforce rest, and thereby to constrain metabolic requirements (Zepelin and Rechtschaffen, 1974), although the scatter of points around this line shows that individual species vary significantly in the strength of this constraint. Thus, broadly, larger mammals have relatively slower metabolism and require relatively less sleep.
The greater size of obese individuals, who have relatively lower resting energy requirements per unit body weight because of their high adiposity (Hoffmans et al., 1979), might itself predict less sleep. Equation 1 above would predict sleep times of 8.0 and 7.4 h per day for human body weights of 70 and 130 kg. Although an inter-species equation is unlikely to perform accurately when applied within a single species, this illustrates that changes in body size might feasibly alter sleep requirements, both in children and adults.
Aside from any direct effects of body size variability, which reflect allometric metabolic phenomena, the metabolic characteristics of weight flux represent another possible cause of sleep variability. A drastic reduction in sleep duration was a notable consequence of semi-starvation during the classic Minnesota starvation study in the mid-twentieth century (Keys et al., 1950). Animal studies have demonstrated this decrease to derive from nutrient availability at the cellular level. Experimental food deprivation decreased the duration of sleep in rats, while restoration of energy supply through glucose infusions restored normal sleep duration (Danguir and Nicolaidis, 1979). Subsequent studies demonstrated the importance of lipid metabolism in this regulatory mechanism, by using infusions of insulin and adrenaline to reverse the normal association of sleep with the light–dark diurnal cycle (Danguir and Nicolaidis, 1980a) Other studies demonstrated that it was the uptake of nutrients, rather than their mere circulation, that impacted on sleep regulation (Danguir and Nicolaidis, 1980b).
These rat studies demonstrated a clear link between the hormonal regulation of metabolism and sleep patterns, but their significance for humans may initially appear questionable if obesity is considered to arise simply from excessive energy intake, that is, positive energy balance. However, the model that we describe above, of cellular semi-starvation brought about by the perturbing effects of high circulating insulin levels on lipid metabolism, suggests a highly plausible hypothesis for metabolism disrupting sleep patterns. Lustig (2006b) has specifically suggested that a high-carbohydrate diet ‘starves children's brains’. According to this logic, we hypothesise that dietary secular trends could be driving trends in children's sleep duration, rather than vice versa.
Thus, rather than short sleep duration causing obesity, those characterised by higher levels of ATP turnover (lipogenesis, gluconeogenesis) may exhibit as symptoms both weight gain and shorter sleeping hours (Figure 5). Lower levels of sleep in obese children may represent merely a marker of the underlying metabolism that drives their weight gain, and if this is the case, there may be little basis for assuming that increased sleep duration (whatever other health benefits it might have) could reduce weight status. Indeed, if sleep is the product of metabolism rather than vice versa, how easily could such interventions be implemented in children? The causality of the sleep–obesity association therefore remains open to question, and merits further investigation.
Obesity, insulin and growth
Rapid growth in infants and children has likewise been implicated in many studies as a risk factor for childhood or adult obesity (Singhal and Lucas, 2004; Baird et al., 2005; Monteiro and Victora, 2005; Stettler et al., 2005; Dennison et al., 2006). These studies have generated two findings of specific interest in the context of our discussion.
The first finding is that while higher levels of childhood weight gain are ubiquitously associated with increased risk of subsequent overweight (Sachdev et al., 2005; Wells et al., 2005; Ekelund et al., 2006b; Metcalf et al., 2011), obese children are also taller (Haroun et al., 2005; Wells et al., 2006; Metcalf et al., 2011) and have accelerated rate of maturation (Ong et al., 2009a; Wronka, 2010). Thus, childhood obesity involves not just excess weight gain, but is also associated with faster linear growth and maturation.
The second finding is that rapid rates of weight gain in infancy may also be associated with the same outcomes, but with such associations widely reported in industrialised populations (Stettler et al., 2003; Ekelund et al., 2006b; Chomtho et al., 2008; Goodell et al., 2009), but not in populations from developing countries (Li et al., 2003; Sachdev et al., 2005; Wells et al., 2005). Thus, there appear to be two sensitive periods where growth rate, weight gain and later obesity are linked (Botton et al., 2008), one of which is more variable between populations than the other. How growth and obesity are connected has not, however, been clarified.
Rapid growth in infancy is often associated with smaller maternal size and lower birth weight, which in turn is associated with intra-uterine growth retardation. For example, in the UK ALSPAC cohort, those growing rapidly in the first 2 years were small at birth and were more likely to be first-borns, and to have been gestated by mothers who smoked during pregnancy (Ong et al., 2000). However, rapid infant weight gain in western populations is also associated with feeding mode, with formula-fed infants tending to gain weight faster than breast-fed infants (Dewey et al., 1993; Ong et al., 2009b). Rapid weight gain in small versus normal weight babies may therefore represent two contrasting phenomena, and this may account for differences in the obesogenic consequences of rapid infant growth between industrialised and developing countries.
Although low birth weight has been widely associated with later insulin resistance, at the time of birth small-for-gestational-age (SGA) neonates are insulin-sensitive (Soto et al., 2003; Mericq et al., 2005; Ibanez et al., 2009). In absolute terms, SGA neonates have low insulin-like growth factor I (IGF-1) levels at birth (Randhawa and Cohen, 2005; Ibanez et al., 2009), however, they also have higher levels of placental IGF and IGF-1 receptor levels (Iniguez et al., 2010). Cord blood IGF-1 levels predict subsequent catch-up growth (Gohlke et al., 2010), and furthermore IGF-1 levels increase rapidly after birth in SGA infants, but not in normal weight infants (Iniguez et al., 2006). Insulin secretion at birth is predictive of subsequent length gains in SGA infants, whereas insulin sensitivity is more closely associated with weight gain and body mass index at 1 year (Soto et al., 2003).
These early hormonal adaptations in insulin and IGF-1 metabolism appear to induce important effects later in childhood. Ibanez et al. (2006) found that initial catch-up growth benefitted length and lean tissue in SGA infants, and that excess weight gain, insulin resistance and central adiposity only emerged from 2 years of age. In other words, catch-up growth was initially beneficial, but could then undergo an overshoot, at which time excess fatness would accumulate. The notion of such an overshoot is supported by the fact that the magnitude of excess weight gain after 2 years was associated with the degree of catch-up in length in early infancy. We suggest that this overshoot indicates a key role for interactions between (a) early adaptations in the hormonal regulation of growth and (b) subsequent dietary composition, thereby elevating the risk of excess weight gain and childhood obesity.
Our model can potentially account for why both catch-up growth and formula-feeding may increase later obesity risk. On the one-hand, IGF-1 is upregulated during catch-up growth as discussed above, while on the other, early protein intake appears significant in promoting faster growth in formula-fed compared with breast-fed infants (Ziegler, 2006), and may likewise do so through upregulating IGF-1 (Axelsson, 2006; Chellakooty et al., 2006; Larnkjaer et al., 2009). In turn, levels of IGF-1 in infancy are predictive of both linear growth rate and subsequent adiposity (Ong et al., 2009b). Hence, both catch-up growth and infant feeding mode are capable of upregulating IGF-1, and both of these early exposures may then interact with subsequent childhood diet. Importantly, such an interaction may be enabled by the structural similarity of insulin and IGF molecules, as noted by Taubes (2008).
Insulin and IGFs are all members of the insulin protein ‘superfamily’, reflecting their evolution from an ancestral insulin gene early in chordate evolution (Chan et al., 1990). They share a high degree of structural conservation, with 45% amino acid homology and substantial similarity in receptor structure and function (Kasik et al., 2000; Laron, 2004). Insulin can bind with the IGF-1 receptor, although with a much lower affinity than that of IGFs (De Meyts et al., 1994), while IGFs can likewise bind with the insulin receptor (White and Kahn, 1994). We therefore hypothesise that a diet with high refined sugar content may increase the cross-binding activity of insulin and IGF-1 on the insulin receptors, thereby stimulating growth.
Although our model could operate simply by an interaction between upregulated IGF-1 and elevated insulin levels following a diet high in refined carbohydrate, more subtle effects could also occur, through modifications to gene expression. For example, a sub-chronic hyperinsulinaemic state could modify the expression of the insulin receptor gene, and increase the levels of expression of variant A compared with variant B. These two variants differ in their affinity for insulin and IGF-1, and in their effects on cellular function in terms of selectivity and specificity of action. The different levels of insulin receptors are given by a regulation of the splicing of the insulin receptor gene, which determines whether a small region (exon 11) will be translated and included in the insulin receptor molecule (variant B) or not (variant A) (Sesti et al., 2001). Hyperinsulinaemia increases the expression of the variant A, which could explain increased efficiency of excess calorie disposal by canalising energy into tissue accretion and linear growth.
Rapid childhood growth could therefore simply represent a marker of dietary-induced hyperinsulinaemia, in children with upregulated IGF-1 exposed to high insulinogenic diets (Figure 5). Although the first opportunity for such exposure could be the introduction of complementary foods in infancy, a significant increase in refined dietary carbohydrate is more likely to occur slightly later, when commercial sucrose- and HCFS-rich foods first enter the diet. This is consistent with recent reports suggesting that the obesity epidemic starts after infancy (Tate et al., 2006), and may help explain why SGA infants only develop central fat and insulin resistance from around 2 years of age (Ibanez et al., 2006).
In populations that catch up in infancy but remain unexposed to a sucrose-rich diet in childhood, our model suggests that the metabolic vulnerability to excess weight gain would not be activated, and no association between early growth and subsequent obesity would manifest. This is consistent both with literature from populations unexposed to the nutrition transition, and may also account for the proliferation of obesity in countries where low birth weight is common, and where the nutrition transition has also rapidly increased exposure to high-sucrose diets (Popkin, 2003). Indeed, our model may help account for the ‘double-burden’ of growth faltering and obesity (Doak et al., 2005), as such growth-retarded individuals consuming a diet high in refined carbohydrate would expose their upregulated IGF-1 levels (induced by early catch-up growth) to the influence of chronically raised insulin levels, and hence express their potential risk of obesity. However, such a mechanism might be complemented by other mechanisms, for example a reduction in fat oxidation has also been proposed to contribute to obesity in stunted children from Brazil (Hoffman et al., 2000).
Dietary trends in sucrose
To account for the emerging obesity epidemic, our model assumes rapid secular increases in the consumption of a high-sucrose diet. Such increases have indeed been documented, driven largely by trends in products containing conventional table sugar through the twentieth century, but also since the 1970s through rapid increases in HFCS (Duffey and Popkin, 2008). HCFS consumption mirrors the rise of obesity in the United States (Bray et al., 2004), although clearly such a correlation not sufficient to demonstrate causation.
Overall, sweetener consumption has increased by ∼90% in the last century and recent estimates suggest that HFCS now represents over 20% of total daily carbohydrate intake in the United States (Gross et al., 2004). In addition, intakes in US children are high (Bleich et al., 2009), with ∼25% of American children consuming more than the 25% recommended of total energy intake from sweeteners (American Dietetic Association, 2004). As in adults, these beverages have also been implicated in excess weight gain (Bray et al., 2004; Malik et al., 2006; Bray, 2010). Broadly, increases in refined carbohydrate consumption constitute a significant role in the nutrition transition (Popkin, 2009), however, it remains unclear how important they are for the global obesity epidemic as the transition involves a number of related changes in dietary intake and physical activity behaviour. Nevertheless, this area strongly merits further research.
Significance for public health
Our approach builds on recent insights by Lustig and Taubes to develop a model of obesity based on metabolic perturbation, rather than energy imbalance. What does this perspective imply for efforts to prevent and treat obesity?
The first implication is that the composition, rather than the quantity, of the food that we eat may be critical. The food industry is dominated by products rich in refined carbohydrate, for numerous reasons: sucrose and HFCS lend themselves readily to palatable foods that are easily stored, displayed and branded—all the processes whereby industry ‘adds value’ to food and thereby generates its profits (Nestle, 2007) as well as stabilising consumer demand. These products are central to the high glycaemic load that may generate chronic high insulin levels. At a secondary level, current dietary advice does not help, because although a low intake of refined carbohydrate is recommended, much attention is still given to dietary energy density and hence dietary fat intake (Taubes, 2008). Products labelled, for example, 99% fat-free are typically rich in carbohydrate and therefore contribute to glycaemic load.
Recent evidence on the relative efficacy of different diets might appear to contradict such a focus on the importance of refined carbohydrates. Several studies have concluded that different diets may generate similar levels of weight loss (Astrup, 2001; Sacks et al., 2009). However, evidence of associations between dietary composition and weight loss may have little relevance for understanding the role of diet in population weight gain, and may simply indicate a similar level of compliance across different treatment regimes, each of which may omit some of the key culprits for population weight gain. Furthermore, fructose-rich soft drinks are one of the product types most strongly implicated in excess weight gain (Bell and Sears, 2003; Sacks et al., 2009; Stanhope et al., 2009; Bray, 2010).
The second implication is that an entire generation of obesity researchers may have been studying the symptoms, not the causes, of weight gain (Lustig, 2006b; Taubes, 2008). Total energy intake and total physical activity level, seemingly so obviously the determinants of weight gain, might actually be somewhat peripheral to persistent population increases in body mass index, although enforced changes in diet and exercise would undoubtedly impact on energy balance. Diet and activity are undoubtedly critical for obesity, but persistent weight gain requires persistent metabolic dysregulation, whereas behavioural trends would be predicted to cause weight to increase and then plateau. The benefits of physical activity may, for example, interact strongly with those of reducing dietary refined carbohydrate. At the very least, we suggest that more attention is now paid to the overweight dog and less to the tail, where preventive and treatment effects based on the theory of energy balance remain unsatisfactory.
The third implication is that the current emphasis on the early origins of adult disease risk may be focusing attention unduly on the first component of a two-stage mechanism. There is now compelling evidence that birth weight is inversely associated both with faster infant growth (Ong et al., 2000), and with the risk of many chronic degenerative diseases (Barker et al., 2009). Some now specifically advocate slower infant growth to benefit cardiovascular health in later life (Singhal and Lucas, 2004; Singhal et al., 2004). Yet faster infant growth appears beneficial in developing countries, and benefits human capital (Victora et al., 2008). We speculate that many of the adverse consequences of rapid infant growth may be dependent on exposure to the hyperinsulinaemia that results from a diet high in refined carbohydrate. It may be more appropriate to target this diet, than infant growth rate (ideally regulated by breast-milk intake), in order to reduce long-term disease risk. In other words, we need to re-evaluate what proportion of public health initiatives should target on the one hand early development, and on the other hand the broader environment.
Our perspective offers many opportunities to test between two models of obesity causation, potentially improving understanding of its aetiology while also generating new opportunities for prevention, management and treatment. Indeed, it is the apparent ‘truism’ of the energy balance approach that has stifled attempts to challenge it as a hypothesis for obesity aetiology—contradictory results are rejected because the model ‘must be true’. While interventions to promote sleep, reduce growth and constrain energy intake may all appear warranted on the basis of existing evidence, further analyses are required to identify which factors are actually causal. A key area for childhood obesity research would involve investigating possible alterations in sleep and growth outcomes according to dietary glycaemic load, to see if metabolism does indeed drive growth and sleep patterns, or vice versa.
It is highly unlikely, in the case of a complex condition such as obesity, sensitive to multiple genomic and environmental effects, that any one hypothesis can explain the entire epidemiology. Obesity is not a homogeneous phenotype, and there are multiple pathways to its development (Wells, 2009). But if supported by further experimental evidence, the metabolic perturbation hypothesis can potentially explain a great deal about escalating national and global obesity epidemics, as well as epidemics of associated diseases, and could therefore offer new approaches to public health policies. At the very least, weighing up two hypotheses is a more scientific approach to data analysis than testing one only, and a more sceptical approach is likely to bring valuable returns. We re-emphasise that nothing proposed here contradicts the energy balance equation; rather we, following others, have questioned how a mathematical truism has been used to interpret obesity data. As shown in Figure 2, reversing the direction of causation from the conventional approach still leaves diet and physical activity as central components of obesity prevention and treatment, and voluntary behaviour as a potential target for intervention, but through a different scientific logic, and therefore potentially through new public health policies. We argue that more rigorous tests of the direction of causation in obesity are likely to prove extremely valuable for improving efforts at prevention and treatment.
Ainsworth BE, Haskell WL, Leon AS, Jacobs Jr DR, Montoye HJ, Sallis JF et al. (1993). Compendium of physical activities: classification of energy costs of human physical activities. Med Sci Sports Exerc 25, 71–80.
American Dietetic Association (2004). Position of the American Dietetic Association: use of nutritive and nonnutritive sweeteners. J Am Dietet Assoc 104, 255–275.
Astrup A (2001). The role of dietary fat in the prevention and treatment of obesity. Efficacy and safety of low-fat diets. Int J Obes Relat Metab Disord 25 (Suppl 1), S46–S50.
Augustine RA, Ladyman SR, Grattan DR (2008). From feeding one to feeding many: hormone-induced changes in bodyweight homeostasis during pregnancy. J Physiol 586, 387–397.
Axelsson I (2006). Effects of high protein intakes. Nestle Nutr Workshop Ser Pediatr Program 58, 121–129; discussion 129–131.
Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, Law C (2005). Being big or growing fast: systematic review of size and growth in infancy and later obesity. BMJ 331, 929–92.
Bandini LG, Schoeller DA, Dietz WH (1990). Energy expenditure in obese and nonobese adolescents. Pediatr Res 27, 198–203.
Bar-Or O, Foreyt J, Bouchard C, Brownell KD, Dietz WH, Ravussin E et al. (1998). Physical activity, genetic, and nutritional considerations in childhood weight management. Med Sci Sports Exerc 30, 2–10.
Barker DJ, Osmond C, Kajantie E, Eriksson JG (2009). Growth and chronic disease: findings in the Helsinki Birth Cohort. Ann Hum Biol 36, 445–458.
Basciano H, Federico L, Adeli K (2005). Fructose, insulin resistance, and metabolic dyslipidemia. Nutr Metab (Lond) 2, 5.
Bell SJ, Sears B (2003). Low-glycemic diets: impact on obesity and chronic diseases. Crit Rev Food Sci Nutr 43, 357–377.
Berson SA, Yalow RS (1965). Some current controversies in diabetes research. Diabetes 14, 549–572.
Bickel PE (2004). Metabolic fuel selection: the importance of being flexible. J Clin Invest 114, 1547–1549.
Bleich SN, Wang YC, Wang Y, Gortmaker SL (2009). Increasing consumption of sugar-sweetened beverages among US adults: 1988–1994 to 1999–2004. Am J Clin Nutr 89, 372–381.
Blundell JE, Lawton CL, Cotton JR, Macdiarmid JI (1996). Control of human appetite: implications for the intake of dietary fat. Annu Rev Nutr 16, 285–319.
Bogin B (1999). Patterns of Human Growth 2nd edn. Cambridge University Press: Cambridge.
Botton J, Heude B, Maccario J, Ducimetiere P, Charles MA (2008). Postnatal weight and height growth velocities at different ages between birth and 5 y and body composition in adolescent boys and girls. Am J Clin Nutr 87, 1760–1768.
Bray GA (2010). Soft drink consumption and obesity: it is all about fructose. Curr Opin Lipidol 21, 51–57.
Bray GA, Nielsen SJ, Popkin BM (2004). Consumption of high-fructose corn syrup in beverages may play a role in the epidemic of obesity. Am J Clin Nutr 79, 537–543.
Bray MS, Young ME (2006). Circadian rhythms in the development of obesity: potential role for the circadian clock within the adipocyte. Obes Rev 8, 169–181.
Briefel RR, Johnson CL (2004). Secular trends in dietary intake in the United States. Annu Rev Nutr 24, 401–431.
Cappuccio FP, Taggart FM, Kandala NB, Currie A, Peile E, Stranges S . et al. (2008). Meta-analysis of short sleep duration and obesity in children and adults. Sleep 31, 619–626.
Caton PW, Nayuni NK, Khan N, Wood EG, Corder R (2011). Fructose induces gluconeogenesis and lipogenesis through a Sirt1-dependent mechanism. J Endocrinol 208, 273–283.
Cha SH, Wolfgang M, Tokutake Y, Chohnan S, Lane MD (2008). Differential effects of central fructose and glucose on hypothalamic malonyl-CoA and food intake. Proc Natl Acad Sci USA 105, 16871–16875.
Chan SJ, Cao QP, Steiner DF (1990). Evolution of the insulin superfamily: cloning of a hybrid insulin/insulin-like growth factor cDNA from amphioxus. Proc Natl Acad Sci USA 87, 9319–9323.
Chaput JP, Brunet M, Tremblay A (2006). Relationship between short sleeping hours and childhood overweight/obesity: results from the ‘Quebec en Forme’ Project. Int J Obes (Lond) 30, 1080–1085.
Chellakooty M, Juul A, Boisen KA, Damgaard IN, Kai CM, Schmidt IM et al. (2006). A prospective study of serum insulin-like growth factor I (IGF-I) and IGF-binding protein-3 in 942 healthy infants: associations with birth weight, gender, growth velocity, and breastfeeding. J Clin Endocrinol Metab 91, 820–826.
chi-Mejia AM, Longacre MR, Gibson JJ, Beach ML, Titus-Ernstoff LT, Dalton MA (2007). Children with a TV in their bedroom at higher risk for being overweight. Int J Obes (Lond) 31, 644–651.
Chomtho S, Wells JC, Williams JE, Davies PS, Lucas A, Fewtrell MS (2008). Infant growth and later body composition: evidence from the 4-component model. Am J Clin Nutr 87, 1776–1784.
Commerford SR, Ferniza JB, Bizeau ME, Thresher JS, Willis WT, Pagliassotti MJ (2002). Diets enriched in sucrose or fat increase gluconeogenesis and G-6-Pase but not basal glucose production in rats. Am J Physiol Endocrinol Metab 283, E545–E555.
Corpeleijn E, Saris WH, Blaak EE (2009). Metabolic flexibility in the development of insulin resistance and type 2 diabetes: effects of lifestyle. Obes Rev 10, 178–193.
Cunha MC, Zanetti ML, Hass VJ (2008). Sleep quality in type 2 diabetics. Rev Lat Am Enfermagem 16, 850–855.
Danguir J, Nicolaidis S (1979). Dependence of sleep on nutrients’ availability. Physiol Behav 22, 735–740.
Danguir J, Nicolaidis S (1980a). Circadian sleep and feeding patterns in the rat: possible dependence on lipogenesis and lipolysis. Am J Physiol 238, E223–E230.
Danguir J, Nicolaidis S (1980b). Intravenous infusions of nutrients and sleep in the rat: an ischymetric sleep regulation hypothesis. Am J Physiol 238, E307–E312.
De Meyts P, Wallach B, Christoffersen CT, Urso B, Gronskov K, Latus LJ et al. (1994). The insulin-like growth factor-I receptor. Structure, ligand-binding mechanism and signal transduction. Horm Res 42, 152–169.
Dennison BA, Edmunds LS, Stratton HH, Pruzek RM (2006). Rapid infant weight gain predicts childhood overweight. Obesity (Silver Spring) 14, 491–499.
Dewey KG, Heinig MJ, Nommsen LA, Peerson JM, Lonnerdal B (1993). Breast-fed infants are leaner than formula-fed infants at 1 y of age: the DARLING study. Am J Clin Nutr 57, 140–145.
Di LL, De PG, Zocchetti C, L’Abbate N, Basso A, Pannacciulli N et al. (2003). Effect of shift work on body mass index: results of a study performed in 319 glucose-tolerant men working in a Southern Italian industry. Int J Obes Relat Metab Disord 27, 1353–1358.
Dietz WH (1994). Critical periods in childhood for the development of obesity. Am J Clin Nutr 59, 955–959.
Dirlewanger M, Schneiter P, Jequier E, Tappy L (2000). Effects of fructose on hepatic glucose metabolism in humans. Am J Physiol Endocrinol Metab 279, E907–E911.
Doak CM, Adair LS, Bentley M, Monteiro C, Popkin BM (2005). The dual burden household and the nutrition transition paradox. Int J Obes (Lond) 29, 129–136.
Duffey KJ, Popkin BM (2008). High-fructose corn syrup: is this what's for dinner? Am J Clin Nutr 88, 1722S–1732S.
Dzamko NL, Steinberg GR (2009). AMPK-dependent hormonal regulation of whole-body energy metabolism. Acta Physiol (Oxf) 196, 115–127.
Eisenmann JC, Ekkekakis P, Holmes M (2006). Sleep duration and overweight among Australian children and adolescents. Acta Paediatr 95, 956–963.
Ekelund U, Aman J, Yngve A, Renman C, Westerterp K, Sjostrom M (2002). Physical activity but not energy expenditure is reduced in obese adolescents: a case-control study. Am J Clin Nutr 76, 935–941.
Ekelund U, Brage S, Froberg K, Harro M, Anderssen SA, Sardinha LB et al. (2006a). TV viewing and physical activity are independently associated with metabolic risk in children: the European Youth Heart Study. PLoS Med 3, e488–e48.
Ekelund U, Ong K, Linne Y, Neovius M, Brage S, Dunger DB et al. (2006b). Upward weight percentile crossing in infancy and early childhood independently predicts fat mass in young adults: the Stockholm Weight Development Study (SWEDES). Am J Clin Nutr 83, 324–330.
Ferro-Luzzi A, Martino L (1996). Obesity and physical activity. Ciba Found Symp 201, 207–221; discussion 221–207.
Foster-Powell K, Holt SH, Brand-Miller JC (2002). International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 76, 5–56.
Ganong WF (1999). Review of Medical Physiology, 19th edn. Appleton and Lange: Stamford, CT.
Garby L, Garrow JS, Jorgensen B, Lammert O, Madsen K, Sorensen P . et al. (1988). Relation between energy expenditure and body composition in man: specific energy expenditure in vivo of fat and fat-free tissue. Eur J Clin Nutr 42, 301–305.
Gohlke BC, Schreiner F, Fimmers R, Bartmann P, Woelfle J (2010). Insulin-like growth factor-I in cord blood is predictive of catch-up growth in monozygotic twins with discordant growth. J Clin Endocrinol Metab 95, 5375–5381.
Goodell LS, Wakefield DB, Ferris AM (2009). Rapid weight gain during the first year of life predicts obesity in 2–3 year olds from a low-income, minority population. J Community Health 34, 370–375.
Goran MI, Treuth MS (2001). Energy expenditure, physical activity, and obesity in children. Pediatr Clin North Am 48, 931–953.
Goris AH, Westerterp-Plantenga MS, Westerterp KR (2000). Undereating and underrecording of habitual food intake in obese men: selective underreporting of fat intake. Am J Clin Nutr 71, 130–134.
Gross LS, Li L, Ford ES, Liu S (2004). Increased consumption of refined carbohydrates and the epidemic of type 2 diabetes in the United States: an ecologic assessment. Am J Clin Nutr 79, 774–779.
Gupta NK, Mueller WH, Chan W, Meininger JC (2002). Is obesity associated with poor sleep quality in adolescents? Am J Hum Biol 14, 762–768.
Haig D, Wharton R (2003). Prader-Willi syndrome and the evolution of human childhood. Am J Hum Biol 15, 320–329.
Haroun D, Wells JC, Williams JE, Fuller NJ, Fewtrell MS, Lawson MS (2005). Composition of the fat-free mass in obese and nonobese children: matched case-control analyses. Int J Obes (Lond) 29, 29–36.
Hess GH (1838). The evolution of heat in multiple proportions. Poggendorf Ann Chemie Physik 47, 210–212.
Hoehn KL, Salmon AB, Hohnen-Behrens C, Turner N, Hoy AJ, Maghzal GJ et al. (2009). Insulin resistance is a cellular antioxidant defense mechanism. Proc Natl Acad Sci USA 106, 17787–17792.
Hoffman DJ, Sawaya AL, Verreschi I, Tucker KL, Roberts SB (2000). Why are nutritionally stunted children at increased risk of obesity? Studies of metabolic rate and fat oxidation in shantytown children from Sao Paulo, Brazil. Am J Clin Nutr 72, 702–707.
Hoffmans M, Pfeifer WA, Gundlach BL, Nijkrake HG, Oude Ophuis AJ, Hautvast JG (1979). Resting metabolic rate in obese and normal weight women. Int J Obes 3, 111–118.
Hopkins M, King NA, Blundell JE (2010). Acute and long-term effects of exercise on appetite control: is there any benefit for weight control? Curr Opin Clin Nutr Metab Care 13, 635–640.
Horne J (2008). Too weighty a link between short sleep and obesity? Sleep 31, 595–596.
Ibanez L, Lopez-Bermejo A, Diaz M, Marcos MV, Casano P, de Zegher F (2009). Abdominal fat partitioning and high-molecular-weight adiponectin in short children born small for gestational age. J Clin Endocrinol Metab 94, 1049–1052.
Ibanez L, Ong K, Dunger DB, de ZF (2006). Early development of adiposity and insulin resistance after catch-up weight gain in small-for-gestational-age children. J Clin Endocrinol Metab 91, 2153–2158.
Iniguez G, Gonzalez CA, Argandona F, Kakarieka E, Johnson MC, Cassorla F (2010). Expression and protein content of IGF-I and IGF-I receptor in placentas from small, adequate and large for gestational age newborns. Horm Res Paediatr 73, 320–327.
Iniguez G, Ong K, Bazaes R, Avila A, Salazar T, Dunger D et al. (2006). Longitudinal changes in insulin-like growth factor-I, insulin sensitivity, and secretion from birth to age three years in small-for-gestational-age children. J Clin Endocrinol Metab 91, 4645–4649.
Jiang F, Zhu S, Yan C, Jin X, Bandla H, Shen X (2009). Sleep and obesity in preschool children. J Pediatr 154, 814–818.
Johnson MA, Tekkanat K, Schmaltz SP, Fox IH (1989). Adenosine triphosphate turnover in humans. Decreased degradation during relative hyperphosphatemia. J Clin Invest 84, 990–995.
Johnstone AM, Horgan GW, Murison SD, Bremner DM, Lobley GE (2008). Effects of a high-protein ketogenic diet on hunger, appetite, and weight loss in obese men feeding ad libitum. Am J Clin Nutr 87, 44–55.
Karlsson B, Knutsson A, Lindahl B (2001). Is there an association between shift work and having a metabolic syndrome? Results from a population based study of 27,485 people. Occup Environ Med 58, 747–752.
Karlsson BH, Knutsson AK, Lindahl BO, Alfredsson LS (2003). Metabolic disturbances in male workers with rotating three-shift work. Results of the WOLF study. Int Arch Occup Environ Health 76, 424–430.
Kasik JW, Lu C, Menon RK (2000). The expanding insulin family: structural, genomic, and functional considerations. Pediatr Diabetes 1, 169–177.
Keys A, Brozek J, Henschel A, Mickelsen O, Taylor HL (1950). The Biology of Human Starvation.. University of Minnesota Press: Minneapolis.
Knutson KL (2010). Sleep duration and cardiometabolic risk: a review of the epidemiologic evidence. Best Pract Res Clin Endocrinol Metab 24, 731–743.
Laessle RG, Platte P, Schweiger U, Pirke KM (1996). Biological and psychological correlates of intermittent dieting behavior in young women. A model for bulimia nervosa. Physiol Behav 60, 1–5.
Larnkjaer A, Hoppe C, Molgaard C, Michaelsen KF (2009). The effects of whole milk and infant formula on growth and IGF-I in late infancy. Eur J Clin Nutr 63, 956–963.
Laron Z (2004). IGF-1 and insulin as growth hormones. Novartis Found Symp 262, 56–77; discussion 77–83, 265–268.
Leclerc I, Kahn A, Doiron B (1998). The 5′-AMP-activated protein kinase inhibits the transcriptional stimulation by glucose in liver cells, acting through the glucose response complex. FEBS Lett 431, 180–184.
Ley RE, Turnbaugh PJ, Klein S, Gordon JI (2006). Microbial ecology: human gut microbes associated with obesity. Nature 444, 1022–1023.
Li H, Stein AD, Barnhart HX, Ramakrishnan U, Martorell R (2003). Associations between prenatal and postnatal growth and adult body size and composition. Am J Clin Nutr 77, 1498–1505.
Ludwig DS, Peterson KE, Gortmaker SL (2001). Relation between consumption of sugar-sweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 357, 505–508.
Lustig RH (2006a). Childhood obesity: behavioral aberration or biochemical drive? Reinterpreting the First Law of Thermodynamics. Nat Clin Pract Endocrinol Metab 2, 447–458.
Lustig RH (2006b). The ‘skinny’ on childhood obesity: how our western environment starves kids’ brains. Pediatr Ann 35, 898–897.
Lustig RH (2008). Which comes first? The obesity or the insulin? The behavior or the biochemistry? J Pediatr 152, 601–602.
Maggard MA, Shugarman LR, Suttorp M, Maglione M, Sugerman HJ, Livingston EH et al. (2005). Meta-analysis: surgical treatment of obesity. Ann Intern Med 142, 547–559.
Malik VS, Schulze MB, Hu FB (2006). Intake of sugar-sweetened beverages and weight gain: a systematic review. Am J Clin Nutr 84, 274–288.
Marshall NS, Glozier N, Grunstein RR (2008). Is sleep duration related to obesity? A critical review of the epidemiological evidence. Sleep Med Rev 12, 289–298.
Matsuzaka T, Shimano H, Yahagi N, memiya-Kudo M, Okazaki H, Tamura Y et al. (2004). Insulin-independent induction of sterol regulatory element-binding protein-1c expression in the livers of streptozotocin-treated mice. Diabetes 53, 560–569.
Mayes PA (1993). Intermediary metabolism of fructose. Am J Clin Nutr 58 (5 Suppl), 754S–765S.
Mericq V, Ong KK, Bazaes R, Pena V, Avila A, Salazar T et al. (2005). Longitudinal changes in insulin sensitivity and secretion from birth to age three years in small- and appropriate-for-gestational-age children. Diabetologia 48, 2609–2614.
Metcalf B, Hosking J, Frémeaux AE, Jeffery AN, Voss LD, Wilkin TJ (2011). BMI was right all along—taller children really are fatter 3 (implications of making childhood BMI independent of height) EarlyBird 48. Int J Obes (Lond) 35, 541–547.
Montague CT, Farooqi IS, Whitehead JP, Soos MA, Rau H, Wareham NJ et al. (1997). Congenital leptin deficiency is associated with severe early-onset obesity in humans. Nature 387, 903–908.
Monteiro PO, Victora CG (2005). Rapid growth in infancy and childhood and obesity in later life—a systematic review. Obes Rev 6, 143–154.
Morrison CD, Berthoud HR (2007). Neurobiology of nutrition and obesity. Nutr Rev 65 (12 Part 1), 517–534.
Muoio DM, Seefeld K, Witters LA, Coleman RA (1999). AMP-activated kinase reciprocally regulates triacylglycerol synthesis and fatty acid oxidation in liver and muscle: evidence that sn-glycerol-3-phosphate acyltransferase is a novel target. Biochem J 338 (Part 3), 783–791.
Myers Jr MG, Leibel RL, Seeley RJ, Schwartz MW (2010). Obesity and leptin resistance: distinguishing cause from effect. Trends Endocrinol Metab 21, 643–651.
Nestle M (2007). Food Politics: How the Food Industry Influences Nutrition and Health. University of California Press: Berkeley.
Nixon GM, Thompson JM, Han DY, Becroft DM, Clark PM, Robinson E et al. (2008). Short sleep duration in middle childhood: risk factors and consequences. Sleep 31, 71–78.
Noguchi T, Tanaka T (1995). Insulin resistance in obesity and its molecular control. Obes Res 3 (Suppl 2), 195S–198S.
O’Rahilly S, Farooqi IS, Yeo GS, Challis BG (2003). Minireview: human obesity-lessons from monogenic disorders. Endocrinology 144, 3757–3764.
Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger DB (2000). Association between postnatal catch-up growth and obesity in childhood: prospective cohort study. BMJ 320, 967–971.
Ong KK, Emmett P, Northstone K, Golding J, Rogers I, Ness AR et al. (2009a). Infancy weight gain predicts childhood body fat and age at menarche in girls. J Clin Endocrinol Metab 94, 1527–1532.
Ong KK, Langkamp M, Ranke MB, Whitehead K, Hughes IA, Acerini CL et al. (2009b). Insulin-like growth factor I concentrations in infancy predict differential gains in body length and adiposity: the Cambridge Baby Growth Study. Am J Clin Nutr 90, 156–161.
Padez C, Mourao I, Moreira P, Rosado V (2009). Long sleep duration and childhood overweight/obesity and body fat. Am J Hum Biol 21, 371–376.
Padwal RS, Majumdar SR (2007). Drug treatments for obesity: orlistat, sibutramine, and rimonabant. Lancet 369, 71–77.
Pagliassotti MJ, Prach PA (1995). Quantity of sucrose alters the tissue pattern and time course of insulin resistance in young rats. Am J Physiol 269 (3 Part 2), R641–R646.
Pagliassotti MJ, Prach PA, Koppenhafer TA, Pan DA (1996). Changes in insulin action, triglycerides, and lipid composition during sucrose feeding in rats. Am J Physiol 271 (5 Part 2), R1319–R1326.
Patel SR, Hu FB (2008). Short sleep duration and weight gain: a systematic review. Obesity (Silver Spring) 16, 643–653.
Polychronakos C, Kukuvitis A (2002). Parental genomic imprinting in endocrinopathies. Eur J Endocrinol 147, 561–569.
Popkin BM (2003). The nutrition transition in the developing world. Dev Policy Rev 21, 581–597.
Popkin BM (2007). The world is fat. Sci Am 297, 88–95.
Popkin BM (2009). Global changes in diet and activity patterns as drivers of the nutrition transition. Nestle Nutr Workshop Ser Pediatr Program 63, 1–10; discussion 10–14, 259–268.
Prentice AM, Black AE, Coward WA, Davies HL, Goldberg GR, Murgatroyd PR et al. (1986). High levels of energy expenditure in obese women. BMJ (Clin Res Ed) 292, 983–987.
Prentice AM, Jebb SA (1995). Obesity in Britain: gluttony or sloth? BMJ 311, 437–439.
Prentice AM, Jebb SA (2003). Fast foods, energy density and obesity: a possible mechanistic link. Obes Rev 4, 187–194.
Randhawa R, Cohen P (2005). The role of the insulin-like growth factor system in prenatal growth. Mol Genet Metab 86, 84–90.
Reinehr T (2010). Obesity and thyroid function. Mol Cell Endocrinol 316, 165–171.
Repaci A, Gambineri A, Pasquali R (2011). The role of low-grade inflammation in the polycystic ovary syndrome. Mol Cell Endocrinol 335, 30–41.
Robinson TN (1999). Reducing children's television viewing to prevent obesity: a randomized controlled trial. JAMA 282, 1561–1567.
Sachdev HS, Fall CH, Osmond C, Lakshmy R, Dey Biswas SK, Leary SD et al. (2005). Anthropometric indicators of body composition in young adults: relation to size at birth and serial measurements of body mass index in childhood in the New Delhi birth cohort. Am J Clin Nutr 82, 456–466.
Sacks FM, Bray GA, Carey VJ, Smith SR, Ryan DH, Anton SD et al. (2009). Comparison of weight-loss diets with different compositions of fat, protein, and carbohydrates. N Engl J Med 360, 859–873.
Sahebjami H, Scalettar R (1971). Effects of fructose infusion on lactate and uric acid metabolism. Lancet 1, 366–369.
Saltiel AR, Kahn CR (2001). Insulin signalling and the regulation of glucose and lipid metabolism. Nature 414, 799–806.
Schwartz MW, Morton GJ (2002). Obesity: keeping hunger at bay. Nature 418, 595–597.
Sekine M, Yamagami T, Handa K, Saito T, Nanri S, Kawaminami K et al. (2002). A dose-response relationship between short sleeping hours and childhood obesity: results of the Toyama Birth Cohort Study. Child Care Health Dev 28, 163–170.
Sesti G, Federici M, Lauro D, Sbraccia P, Lauro R (2001). Molecular mechanism of insulin resistance in type 2 diabetes mellitus: role of the insulin receptor variant forms. Diabetes Metab Res Rev 17, 363–373.
Singhal A, Cole TJ, Fewtrell M, Deanfield J, Lucas A (2004). Is slower early growth beneficial for long-term cardiovascular health? Circulation 109, 1108–1113.
Singhal A, Lucas A (2004). Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 363, 1642–1645.
Skomro RP, Ludwig S, Salamon E, Kryger MH (2001). Sleep complaints and restless legs syndrome in adult type 2 diabetics. Sleep Med 2, 417–422.
Soto N, Bazaes RA, Pena V, Salazar T, Avila A, Iniguez G et al. (2003). Insulin sensitivity and secretion are related to catch-up growth in small-for-gestational-age infants at age 1 year: results from a prospective cohort. J Clin Endocrinol Metab 88, 3645–3650.
Speakman JR, Rance KA, Johnstone AM (2008). Polymorphisms of the FTO gene are associated with variation in energy intake, but not energy expenditure. Obesity (Silver Spring) 16, 1961–1965.
Spiegel K, Leproult R, Van CE (1999). Impact of sleep debt on metabolic and endocrine function. Lancet 354, 1435–1439.
Spiegel K, Tasali E, Penev P, Van CE (2004). Brief communication: sleep curtailment in healthy young men is associated with decreased leptin levels, elevated ghrelin levels, and increased hunger and appetite. Ann Intern Med 141, 846–850.
Stanhope KL, Schwarz JM, Keim NL, Griffen SC, Bremer AA, Graham JL et al. (2009). Consuming fructose-sweetened, not glucose-sweetened, beverages increases visceral adiposity and lipids and decreases insulin sensitivity in overweight/obese humans. J Clin Invest 119, 1322–1334.
Stanner SA, Yudkin JS (2001). Fetal programming and the Leningrad Siege study. Twin Res 4, 287–292.
Stettler N, Kumanyika SK, Katz SH, Zemel BS, Stallings VA (2003). Rapid weight gain during infancy and obesity in young adulthood in a cohort of African Americans. Am J Clin Nutr 77, 1374–1378.
Stettler N, Stallings VA, Troxel AB, Zhao J, Schinnar R, Nelson SE et al. (2005). Weight gain in the first week of life and overweight in adulthood: a cohort study of European American subjects fed infant formula. Circulation 111, 1897–1903.
Storlien L, Oakes ND, Kelley DE (2004). Metabolic flexibility. Proc Nutr Soc 63, 363–368.
Swinburn BA, Sacks G, Lo SK, Westerterp KR, Rush EC, Rosenbaum M et al. (2009). Estimating the changes in energy flux that characterize the rise in obesity prevalence. Am J Clin Nutr 89, 1723–1728.
Taheri S, Lin L, Austin D, Young T, Mignot E (2004). Short sleep duration is associated with reduced leptin, elevated ghrelin, and increased body mass index. PLoS Med 1, e62.
Tappy L, Le KA (2010). Metabolic effects of fructose and the worldwide increase in obesity. Physiol Rev 90, 23–46.
Tate AR, Dezateux C, Cole TJ (2006). Is infant growth changing? Int J Obes (Lond) 30, 1094–1096.
Taubes G (2008). The Diet Delusion. Vermillion: London.
Treuth MS, Figueroa-Colon R, Hunter GR, Weinsier RL, Butte NF, Goran MI (1998). Energy expenditure and physical fitness in overweight vs non-overweight prepubertal girls. Int J Obes Relat Metab Disord 22, 440–447.
Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, Gordon JI (2006). An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 444, 1027–1031.
Underwood AH, Newsholme EA (1965). Properties of phosphofructokinase from rat liver and their relation to the control of glycolysis and gluconeogenesis. Biochem J 95, 868–875.
VanSchaftingen E, Veiga-da-Cunha M, Niculescu L (1997). The regulatory protein of glucokinase. Biochem Soc Trans 25, 136–140.
Victora CG, Adair L, Fall C, Hallal PC, Martorell R, Richter L et al. (2008). Maternal and child undernutrition: consequences for adult health and human capital. Lancet 371, 340–357.
von Helmholtz H (1847). Uber Die Warmeenwicklung Bei Der Muskelaction. G Reiner: Berlin.
vonKries R, Toschke AM, Wurmser H, Sauerwald T, Koletzko B (2002). Reduced risk for overweight and obesity in 5- and 6-y-old children by duration of sleep—a cross-sectional study. Int J Obes Relat Metab Disord 26, 710–716.
Walter J, Paulsen M (2003). Imprinting and disease. Semin Cell Dev Biol 14, 101–110.
Wells JC (1998). Is obesity really due to high energy intake or low energy expenditure? Int J Obes Relat Metab Disord 22, 1139–1140.
Wells JC (2009). Thrift: a guide to thrifty genes, thrifty phenotypes and thrifty norms. Int J Obes (Lond) 33, 1331–1338.
Wells JC, Chomtho S, Fewtrell MS (2007). Programming of body composition by early growth and nutrition. Proc Nutr Soc 66, 423–434.
Wells JC, Fewtrell MS, Williams JE, Haroun D, Lawson MS, Cole TJ (2006). Body composition in normal weight, overweight and obese children: matched case-control analyses of total and regional tissue masses, and body composition trends in relation to relative weight. Int J Obes 30, 1506–1513.
Wells JC, Hallal PC, Reichert FF, Menezes AM, Araujo CL, Victora CG (2008). Sleep patterns and television viewing in relation to obesity and blood pressure: evidence from an adolescent Brazilian birth cohort. Int J Obes (Lond).
Wells JC, Hallal PC, Wright A, Singhal A, Victora CG (2005). Fetal, infant and childhood growth: relationships with body composition in Brazilian boys aged 9 years. Int J Obes (Lond) 29, 1192–1198.
Westerterp KR, Speakman JR (2008). Physical activity energy expenditure has not declined since the 1980s and matches energy expenditures of wild mammals. Int J Obes (Lond) 32, 1256–1263.
White MF, Kahn CR (1994). The insulin signaling system. J Biol Chem 269, 1–4.
Wishnofsky M (1958). Caloric equivalent of gained or lost weight. Am J Clin Nutr 6, 542–546.
Wolfgang MJ, Cha SH, Sidhaye A, Chohnan S, Cline G, Shulman GI et al. (2007). Regulation of hypothalamic malonyl-CoA by central glucose and leptin. Proc Natl Acad Sci USA 104, 19285–19290.
Woods SC, D’Alessio DA (2008). Central control of body weight and appetite. J Clin Endocrinol Metab 93 (11 Suppl 1), S37–S50.
Woods SC, Seeley RJ (2000). Adiposity signals and the control of energy homeostasis. Nutrition 16, 894–902.
Wronka I (2010). Association between BMI and age at menarche in girls from different socio-economic groups. Anthropol Anz 68, 43–52.
Young T (2008). Increasing sleep duration for a healthier (and less obese?) population tomorrow. Sleep 31, 593–594.
Zepelin H, Rechtschaffen A (1974). Mammalian sleep, longevity, and energy metabolism. Brain Behav Evol 10, 425–470.
Ziegler EE (2006). Growth of breast-fed and formula-fed infants. Nestle Nutr Workshop Ser Pediatr Program 58, 51–59; discussion 59–63.
The authors declare no conflict of interest.
About this article
Cite this article
Wells, J., Siervo, M. Obesity and energy balance: is the tail wagging the dog?. Eur J Clin Nutr 65, 1173–1189 (2011). https://doi.org/10.1038/ejcn.2011.132
- energy balance
- metabolic flexibility
American Journal of Physiology-Endocrinology and Metabolism (2020)
The carbohydrate-insulin model does not explain the impact of varying dietary macronutrients on the body weight and adiposity of mice
Molecular Metabolism (2020)
Proceedings of the Nutrition Society (2020)
Convenience Behavior and Being Overweight in Adults: Development and Validation of the Convenience Behavior Questionnaire
Frontiers in Public Health (2019)
European Journal of Nutrition (2019)