Abstract
Body fat mass increases when energy intake exceeds energy expenditure. In the long term, a positive energy balance will result in obesity. The worldwide prevalence of obesity has increased dramatically, posing a serious threat to human health. Therefore, insight in the pathogenesis of obesity is important to identify novel prevention and treatment strategies. This review describes the physiology of energy expenditure and energy intake in the context of body weight gain. We focus on the components of energy expenditure and the regulation of energy intake. Finally, we describe rare monogenetic causes leading to an impairment in central regulation of food intake and obesity.
2. Energy expenditure
Energy expenditure (EE) consists of three components: resting metabolic rate (RMR), activity-related energy expenditure (AEE), and (diet-induced) thermogenesis (DIT) [7]. The metabolism of an individual at rest is known as the RMR, i.e., the energy requirements for maintenance of vital body functions, such as temperature, circulation, respiration, and cell growth. In sedentary adults, RMR accounts for approximately 60% of total daily EE [8]. A major determinant of RMR is body composition, specifically that of metabolically active tissues such as free-fat mass (FFM). Daily intra- and inter-individual variability in RMR ranges from 2 – 10% [9-13] and 7.5 – 17.9% [14-16], respectively. In general, women have lower RMR than men [17-21], and young adults have higher RMR than older adults [19, 22, 23], which may be attributed to differences in skeletal muscle mass [24, 25]. Notably, however, there is much variation in metabolic rate: at any given body size, subjects can have low, normal, or high metabolic rates [26]. Several studies identified low adjusted RMR as a risk factor for weight gain [27, 28]. The risk of gaining 10kg of body weight is approximately eight times greater in individuals belonging to the lowest tertile of RMR compared to those in the highest tertile [26].
AEE can be categorized into exercise activity thermogenesis (EAT) and non-exercise activity thermogenesis (NEAT) [2]. In most countries, the contribution of EAT to total daily EE is negligible [29], and NEAT is predominant [8]. NEAT comprises EE during all physical activities other than sport-like exercises, such as occupational EE and leisure-time physical activity [29]. It varies widely across and within individuals on a daily basis and can differ up to 2000 kcal/day between two individuals of similar size [29]. The variability in NEAT is in part genetically determined [30].
Energy intake, ironically, also adds to EE, as energy is necessary for the foraging, digestion, absorption, and storage of food [2]. DIT is the increase in metabolic rate associated with the ingestion of food and increases in post-absorptive heat production [31]. Typically, DIT accounts for 5- 15% of total EE [32]. The magnitude of the thermic effect of food depends on the energy content and composition of the food consumed. Measured thermic effects of nutrients are 5-15% for carbohydrates and fat and 20-30% for proteins [33, 34]. In healthy subjects in energy balance with a mixed diet, DIT represents about 10% of daily total ingested energy [32]. One of the determinants of DIT is insulin sensitivity [35]. Insulin-sensitive individuals have a more pronounced thermic effect of food, whereas the most insulin-resistant individuals have negligible effects [35, 36]. Furthermore, food temperature might influence EE. The intake of food or drinks cooler than core body temperature might elicit an increase in EE due to energy required for heating it up to body temperature [37].
Each of these three components of EE is subject to regulation and can vary considerably from day to day and person to person [2]. However, in line with the first law of thermodynamics, when an individual ingests more energy than it expends, a positive energy balance develops and excess energy is stored as high-energy molecules. Sixty to 80% of excess energy is converted into triglycerides.
Small daily aberrations from a neutral energy balance can, over time, contribute to significant weight gain. When energy intake exceeds EE by merely 20 kcal/day (the equivalent of 1 tsp of sugar) a person would gain approximately 1 kg of fat per year (≈ 20 kg over 2 decades) [2]. Remarkably, many adults maintain constant body weight for long periods of time with little conscious efforts. This is partly explained by adaptations in EE (RMR and NEAT) in conditions of over- and underfeeding [8]. In physiological conditions, complex regulatory systems constantly monitor energy status, and an appropriate response in feeding and EE is orchestrated.
2.1 Energy expenditure in obesity
The contribution of reduced EE in the development of obesity has been controversial. Historically, obesity was associated with “slow metabolism” due to lower RMR, AEE and/or DIT contributing to a positive energy balance and subsequent weight gain [27]. Nonetheless, studies conducted over the past three decades have reported contradictory findings; absolute EE in obese individuals is higher compared to their lean counterparts [38-44].
RMR is positively associated with weight [38-43, 45, 46] with reported differences up to 800 kcal/day when comparing individuals with a BMI >50 (RMR = 2157 kcal/day) to lean individuals (RMR = 1331 kcal/day) [42]. In obesity, increases in fat mass occur concurrently with increases in FFM [40]. After correcting RMR for FFM the higher RMR found in obesity is blunted [40, 43, 45, 47-49]. Notably, most studies adjusted RMR simply by dividing it by FFM [38-40, 43, 45, 47, 49].
In addition, obesity is associated with a higher absolute AEE [38, 41, 43, 48]. Differences between lean and obese individuals diminish after adjusting for FFM [38, 41, 43, 48]; indicating that the obese state itself does not intrinsically alter AEE. Nevertheless, obesity is associated with increased sedentary behavior [50] which negatively influences (adjusted) AEE. Therefore, sedentary lifestyle is contributing to positive energy balance and, consequently, obesity.
Studies examining the relationship between DIT and obesity yield inconsistent results. Some described lower DIT in obesity [50-55], whereas others found no difference in DIT between lean and obese individuals [56-58]. Investigating the influence of obesity on DIT is complex due to many confounding variables that may introduce bias, such as the level of physical activity [59] and insulin-resistance [35, 36, 60]. In addition, obesity is associated with a prolonged absorptive state which confounds the duration of DIT measurement. Altogether the evidence is insufficient to support the theory of an altered DIT in obesity. Figure 1 shows variations of energy expenditure between lean and obese individuals.
In summary, studies do not support the hypothesis of an altered RMR and DIT in obesity but lower NEAT and EAT contribute to body weight gain. Emerging obesity phenotypes such as normal weight obesity [46] and sarcopenic obesity [61] might be accompanied with an altered EE compared to lean individuals or obese with normal body composition.
2.2 Effect of caloric restriction on energy expenditure
Calorie-restriction strategies induce weight loss and are, therefore, commonly applied in the treatment of obesity [62]. With (intentional) weight loss, energy requirements fall and compensatory decreases in all components of EE occur [62-65] to match this lower energy intake. Twenty percent weight loss results in a 325-480 kcal/day reduction of EE [66]. The loss of FFM during weight loss largely contributes to lower EE. However, during caloric restriction, the process of metabolic adaptation induces a disproportional decrease in RMR. This metabolic adaptation can account for an additional reduction in RMR up to 500 kcal daily [67]. The mechanism explaining this phenomenon is still unclear in humans but probably involves the hypothalamus-pituitary-thyroid axis and lower sympathetic activity [68]. This reduction in EE comes along with increases in hunger [65] further challenging adherence to caloric restriction in the absence of (permanent) behavioral change. The lack of success in long-term weight loss maintenance [69] suggests that most individuals are not able to match the lower EE with lower food intake mandatory to sustain weight loss.
2.3 Metabolic efficiency
The inter-individual energy required to maintain body weight in individuals with similar physical characteristics can vary widely [27]. Differences in metabolic efficiency might explain this variability and play a role in the susceptibility to weight gain.
The transformation of (energy embedded) in nutrients into actual task performances requires two consecutive metabolic processes [2]. First, nutrients are oxidized to yield ATP that serves as metabolic currency [2]. Second, ATP is utilized into actual task performances (e.g., vital body functions and physical activity) [2]. In line with the second law of thermodynamics; both processes involve heat production [2]. Metabolic efficiency refers to the proportion of ATP vs heat production derived from a given task performance [70].
Ability to dispose part of excess energy as heat decreases the ability to store excess energy as fat and, thus, prevents weight gain [70]. Low metabolic efficiency implies an increase in heat production at the expense of ATP production (and conversion into triglycerides) [71]. Enhanced metabolic efficiency has been reported to contribute to obesity [71]. In rodents, metabolic efficiency is also increased during caloric restriction (23918688).The variability in metabolic efficiency is in part genetically determined [72].
3. Energy intake
Animals, unlike plants, obtain all energy requirements from ingested food and drink. To balance energy intake and expenditure, there are complex systems that regulate feeding behavior. These systems operate at the crossroads of several brain circuits and receive input from peripheral signals that relay information on nutritional state [73].
The two major systems that control food intake are often referred to as the homeostatic and hedonic pathways [74]. The homeostatic pathway stimulates feeding behavior when energy stores are low. The hypothalamus and brainstem have been identified as its centers. Here, central and peripheral signals, including circulating concentrations of nutrients, gastrointestinal hormones, and vagal afferents, are integrated to mediate feelings of hunger vs satiety and adjust food intake. The hedonic, or reward-based, pathway adds another layer of control and may override the homeostatic system [74].
By mediating the rewarding and motivational aspects of food intake, the hedonic system can support energy homeostasis during periods of relative energy deficiency, but also increase the intake of highly palatable food during periods of relative energy sufficiency. During evolution, high sensitivity for food cues was probably beneficial, because this increased the chance of successful food foraging and survival. However, in the current obesogenic environment, increased motivation for feeding behavior may be less advantageous. Obesity likely develops when the hedonic and homeostatic regulatory systems are out of balance. Several neurotransmitter and brain circuit-related hypotheses have been postulated to explain the relative abundance of energy intake that causes obesity [65, 75-78].
3.1 Impaired homeostatic inhibition of food intake
The homeostatic hypothesis of obesity development states that decreased serotonin signaling in the hypothalamus impairs negative feedback from energy intake on feeding behavior, thus promoting overconsumption [79]. Serotonin is an important neurotransmitter in the homeostatic pathway, and studies in animals and humans have shown that manipulation of serotonin changes feeding behavior [80, 81].
Serotonin is produced in the raphe nuclei of the brainstem and involved in the regulation of food intake via projections to multiple brain regions of the homeostatic and hedonic regulatory systems [80]. Most of the available evidence points to a model where increased serotonergic signaling is associated with decreased food intake, whereas decreased serotonergic signaling induces hyperphagia and weight gain [82]. The effect of serotonin on food intake is thought to be twofold. Serotonin is able to activate the anorexigenic α-melanocyte-stimulating hormone (α-MSH), a product of pro-opiomelanocortin (POMC) neurons, and inhibit the orexigenic neuropeptide Y (NPY) and Agouti-related peptide (AgRP) neurons located in the arcuate nucleus of the hypothalamus [83]. However, we note that this is a simplification of serotonergic reality and not all recent findings match this straightforward model [84]. Nevertheless, the importance of serotonin is also supported by in vivo human studies that use molecular neuroimaging, including positron emission tomography (PET) or single photon emission tomography (SPECT). Most of these indicate that human obesity is associated with decreased serotonergic signaling (for a review, see [85]), thereby supporting the human relevance of the hyposerotenergic theory of obesity development. In line, the serotonin 2c receptor agonist lorcarserin induces weight loss in obese humans [86].
Functional magnetic resonance imaging (fMRI), another form of neuroimaging, is frequently applied to (indirectly) visualize neuronal activation and functional connectivity of brain regions during fasting or in response to food cues/ingestion [87]. Unfortunately, imaging of the brainstem and hypothalamus is complicated due to physiological noise, their small overall volume, and the numerous smaller, yet functionally distinct nuclei within these brain regions [88]. Therefore, current fMRI data on the homeostatic system are hard to interpret, and few studies have investigated the effect of obesity on these regions. Nevertheless, studies do report differences in the activation and connectivity of the brainstem and hypothalamus in obese adults, consistent with impaired homeostatic regulation in human obesity [89-92].
3.2 Increased hedonic drive for food intake
Another explanation for the imbalance between homeostatic and hedonic regulation is postulated in the reward deficiency theory. This hypothesis states that decreased dopaminergic signaling, which normally relays the rewarding aspects of (food-related) stimuli, promotes the overconsumption of palatable food beyond homeostatic needs in order to compensate for lower reward sensations [93].
Recent advances in neurobiology have greatly enhanced our understanding of the neuronal and peripheral factors involved in feeding behavior; for an overview of the functional organization of feeding circuits, we refer to the available literature [78, 83, 94]. Briefly, the hedonic system is headquartered in the striatum, with close connections to the hypothalamus and homeostatic system [95]. In animal models with reduced dopamine signaling through striatal D2 receptor knockdown, a phenotype of compulsive-like food seeking and obesity is observed [96]. In addition, in humans the presence of the A1 allele of the DRD2/ANKK1 Taq1A polymorphism, which is associated with lower dopamine D2 receptor (D2/3R) availability, increases the risk for the development of obesity [97]. There is also increasing evidence from human neuroimaging trials in support of reward deficiency. Recent fMRI data demonstrate reduced striatal activity in response to food consumption in obese compared to lean subjects [89, 98]. In fact, reduced neuronal activation in the striatum upon food intake may be predictive of future weight gain, and this finding was particularly strong in individuals with the Taq1A A1 allele [98, 99]. In accordance, PET and SPECT data show decreased striatal availability of the dopamine D2 and D3 receptor (D2/3R) in obese vs lean individuals, although the correlations between BMI and D2/3R availability have not been consistent and the relationship is not necessarily linear [85]. Few molecular neuroimaging trials have assessed changes in D2/3R availability, which reflect dopamine release and receptor binding, in response to food-related stimuli, but the available data suggest that striatal dopamine release is impaired in obese humans [100, 101]. Overall, these data are consistent with the hypothesis that reduced striatal dopamine signaling may drive people to overeat in compensation [98].
Another theory, which may in fact complement the reward deficiency hypothesis, postulates that overeating in obesity is caused by higher expectations of reward for food. This is supported by fMRI studies, where obese subjects had increased neuronal activation in brain regions of the cerebral reward system upon exposure to visual food cues or taste [102-107]. A combination of increased expectations of reward and reduced sensations of reward would surely prompt people to overeat.
3.3 Nutritional feedback
Both the homeostatic and hedonic regulatory systems receive input from multiple signals that convey information on energy intake, current energy status, and long-term energy stores. These signals close the energy expenditure-intake feedback loop and are therefore essential for the maintenance of energy balance. Obesity is associated with several changes in nutritional feedback.
3.3.1 Taste and smell
Taste is one of the first food intake-related signals to the homeostatic system [108]. Taste receptor cells are located in taste buds on the tongue, activated by molecules in food and drink, and linked to the brainstem and hypothalamus via gustatory sensory afferent neurons in the facial and glossopharyngeal nerves. Most of us have also experienced that the taste of palatable food produces reward, indicating a link to the hedonic system.
Studies on taste perception in obesity have produced mixed results [109]; overall, higher BMI seems to be associated with lower intensity of sweet, salt, and umami perception [110-112]. In addition, more recent data from animals [113] and humans [114, 115] suggest that dietary fatty acids may produce their own gustatory sensations, i.e., the taste of fat [116]. In this light, decreased fatty acid chemoreception has been implicated in the development of obesity [117], although a recent meta-analysis of 7 trials did not reveal any differences in fatty acid taste detection between lean and obese subjects [115].
One mechanism for taste dysfunction, which was recently demonstrated in mice [118], implicates obesity-induced chronic low-grade inflammation in the reduction of taste bud abundance. This finding suggests that obesity precedes taste dysfunction, but it is likely not the only mechanism. Nevertheless, taste intensity does seem to be important. When healthy volunteers were randomly assigned to pharmacological inhibition of sweet taste perception (vs control), they developed a preference for food with higher sucrose content, i.e., more intense stimuli, suggesting that impaired sweet taste may contribute to increased sugar consumption [119]. Obesity may also be associated with changes in smell [111], but the directionality of this relationship is, as of yet, unclear. Interestingly, an acquired loss of olfactory function protects mice from becoming obese and promotes weight loss and insulin sensitivity in diet-induced obese (DIO) mice [120]. These effects were, surprisingly, not caused by a reduction in food intake, but were found to be mediated by sympathetic nervous activity and adipose tissue thermogenesis, indicating a previously unknown link between the olfactory sense and metabolism.
3.3.2 Gastrointestinal hormones
Gut hormones and peptides are well-known regulators of gastrointestinal motility and digestive function. In addition, it has become increasingly clear that hormones and peptides secreted by endocrine cells in the stomach, gut, and pancreas are important modulators of whole-body metabolism and food intake, thereby contributing to the control of energy balance [121]. Several gut hormones, secreted in response to fasting or exposure to ingested nutrients, affect feeding behavior by promoting satiety or appetite, and their circulating concentrations have been shown to differ between lean and obese individuals [122].
Ghrelin, the only known orexigenic or hunger hormone, is primarily produced in the gastric fundus. Its levels go up during fasting and spike just before eating; its secretion may be controlled by the sympathetic nervous system [123] and is suppressed postprandially, i.e., in response to nutrient intake [124]. Ghrelin stimulates appetite, promotes meal initiation, and has been implicated in the regulation of long-term energy balance [125]. Fasting ghrelin levels and BMI are negatively correlated, and human obesity is associated with reduced postprandial ghrelin suppression [126]. Anorexigenic intestinal hormones, including glucagon-like peptide 1 (GLP1), peptide YY (PYY), and cholecystokinin (CCK), are secreted in response to food intake or nutrient exposure and involved in, among others, digestion, insulin secretion and post-digestive metabolism, and satiety [121]. Meal consumption in obese humans is associated with delayed, reduced, or otherwise attenuated activity of anorexigenic hormones [127, 128].
Perturbations in the suppression of ghrelin and/or rise of anorexigenic hormones may contribute to impaired homeostatic inhibition and reward deficiency; consequently, the gut hormones have been intensively investigated over the past few decades, but their precise role in the complex regulation of feeding behavior and obesity development in humans– and their potential as therapeutic targets [129] – is not fully elucidated yet. Recently, clinical trials with GLP-1 analogues showed promising results on body weight regulation [130-132].
3.3.3 Insulin
In addition to orchestrating postprandial metabolism, insulin also contributes to nutritional feedback: its effects on the hypothalamus promote satiety [133], whereas it enhances dopamine release in the striatum, thereby signaling reward [134]. Obesity is characterized by insulin resistance, a condition that may also develop in the brain [135], suggesting that insulin-mediated nutritional feedback may be impaired in obesity. Indeed, insulin’s effect on striatal dopamine dynamics was disturbed in rats fed a high fat high sucrose diet [136]. Whether insulin contributes to the development of overconsumption and thus obesity or facilitates reward-driven food intake in the setting of obesity is unknown. Interestingly, we have recently shown that striatal dopamine release promotes whole-body insulin sensitivity [137], indicating a bidirectional link between insulin action and the reward system.
3.3.4 Cerebral nutrient sensing
Mice without sweet taste receptors still develop a preference for sucrose over non-caloric sweeteners [138], indicating that the reinforcing effects of sugar are not only attributed to taste. Circulating nutrients, now called post-absorptive nutrient signals [139], also provide nutritional feedback to the CNS. In mice, intragastric nutrient infusion, bypassing naso-oral chemoreception, induced striatal dopamine release depending on caloric value of the infusion [140]; in humans, it altered brainstem and hypothalamic neuronal activity [141, 142]. Mechanisms underlying post-absorptive nutrient signaling remain to be elucidated, but emerging rodent data suggest that separate dopaminergic circuits are involved in chemoreceptive (olfactory, gustatory) vs post-absorptive (gut hormones, circulating nutrients) signals [139]. It is currently unknown how these steps in nutritional feedback are altered in human obesity, but one study, comparing obese to lean subjects, showed that the increase in cerebral ATP upon intravenous glucose infusion is diminished [143], indicating either impaired brain energy delivery or reduced ATP production, which may be one mechanism contributing to the “hungry brain”.
3.3.5 Leptin
Leptin is predominantly secreted by white adipose tissue. Circulating levels correlate with fat mass and represent a hormonal signal of body energy stores [144-147]. In individuals with more body fat, serum, plasma and cerebrospinal fluid leptin levels are elevated [147]. Leptin binds to its receptors, primarily relaying information on energy status [147-150], but also on acute energy availability [151-153]. In the hypothalamus, leptin mediates most of its actions [154-157]. Here, the activity of several hypothalamic neurons and expression of various orexigenic and anorexigenic neuropeptides are under its influence [150, 158-162]. [163]. Briefly, leptin activates neurons that synthesize anorexigenic peptides, including pro-opiomelanocortin (POMC), and suppresses the activity of orexigenic neurons [148, 156, 157, 164-167]. In addition, leptin counterbalances the effects of ghrelin [168, 169]. Overall, in leptin-sensitive individuals, leptin signaling results in a decrease in food intake and an increase in energy expenditure to maintain the size of energy stores [170-175]. Conversely, low leptin levels augment food intake and suppress energy expenditure [170-175]. Despite elevated leptin levels, obesity is characterized by impaired leptin signaling, i.e., leptin resistance, which explains why leptin administration to most obese individuals is not effective [176]. Leptin resistance is thought to result from hypothalamic inflammation and gliosis, but many other mechanisms are involved [177-179]. Mutations in genes encoding components of the leptin-melanocortin pathway result in early onset obesity (further described below) and subjects with these rare conditions benefit from therapy with leptin or melanocortin receptor 4 agonists [180, 181].
3.4 Circadian rhythms
Circadian clocks orchestrate all biological functions (from gene expression to apoptosis) in a rhythmic 24-hour periodicity. The central circadian clock is located in the CNS, specifically in the suprachiasmatic nucleus (SCN) of the hypothalamus [182, 183]. Most metabolically active cells also have an internal, peripheral clock that enables tissues to regulate gene expression locally in an autonomic manner [184-187]. The central clock is primarily synchronized by afferent signals from the retina [188, 189], but fine-tuned in response to other signals such as meal timing and food composition [190, 191]. Peripheral clocks, in turn, are synchronized by signals from the SCN and have several mechanisms to influence metabolism, including circulating levels of hormones and metabolites. Circadian misalignment, i.e., disruption of the circadian rhythm, due to altered timing of food intake and diet composition can give rise to the development of metabolic disorders [192-194]. In animals, changes in the circadian rhythm are associated with changes in feeding behavior and weight gain [195]. As a result, (night) shift workers are at greater risk of obesity and obesity-related disorders [196, 197]. Modifying the time of feeding alone can greatly affect body weight [198, 199].
3.5 Cause and effect
Almost all human data on the homeostatic and hedonic system are derived from cross-sectional studies, which do not support conclusions on causality. It is still debated whether the described changes in obesity are the result of (long-term) obesity or predispose to a positive energy balance and weight gain. We have shown that serotonin transporter availability increases within six weeks of hypercaloric high-fat high-sugar snacking in lean adults, suggesting that cerebral effects of a hypercaloric diet arise early and may contribute to the progression of weight gain [200]. This effect was not observed with other hypercaloric diets, indicating a role for diet composition in mediating these effects. Meal timing may be similarly important: obese subjects in a hypocaloric diet intervention study, who consumed most calories at breakfast, had increased thalamic serotonin and striatal dopamine transporter availability, whereas those, who consumed calories at dinner, had decreased availability of both transporters [201]. We have also shown that striatal D2/3R availability is reduced in obese women, but partially reversible after bariatric surgery-induced long-term weight loss [202]; this suggests that obesity may, in fact, cause the observed changes in the dopamine system. Evidently, longitudinal follow-up and controlled intervention trials are required to resolve the issue of causality, but the available evidence suggests it may, in fact, go both ways.
4. Genetic causes of obesity
4.1 Leptin-melanocortin pathway
As outlined above, the leptin-melanocortin pathway plays a pivotal role in food intake and energy balance. Leptin stimulates POMC neurons in the arcuate nucleus to produce a series of melanocortin peptides. The melanocortin α-MSH binds with high affinity to melanocortin receptor-3 (MCR3) and MCR4 in the paraventricular nucleus [203, 204]. These signaling pathways subsequently coordinate energy intake and expenditure [205, 206]. Mutations in genes involved in the leptin-melanocortin tract have, to a greater or lesser extent, been associated with (childhood-onset) obesity.
First, patients with homozygous or compound heterozygous mutations in leptin have reduced leptin levels- and activity, or in case of homozygous/ compound heterozygous mutations in the leptin receptor (LEPR) an impaired leptin receptor signaling ability, leading to obesity [207]. A leptin or LEPR mutation occurs in approximately 1 – 5% of the morbidly obese population [206, 207]. Second, homozygous/ compound heterozygous mutations of the POMC gene are associated with severe early-onset obesity, combined with features of adrenocorticotropin hormone (ACTH) deficiency, pale skin and red hair. The last two features are due to the role POMC plays in the determination of melanocytes [208]. Third, MC4R mutations in humans result in a phenotype of early onset obesity and hyperphagia [209]. Approximately 2 – 5% of the childhood-onset obesity cases are due to a heterozygous mutation of MC4R; making a MC4R mutation the most common cause of monogenic obesity [209-211]. Finally, heterozygous mutations in MC3R have been identified in humans with obesity, but the underlying mechanism is unclear. However, mice lacking the melanocortin 3 receptor show increased body fat and reciprocal decreased lean mass due to increased food efficiency and relative inactive behavior. This phenotype was more prominent on a high-fat diet [212, 213].
Mutations in POMC-derived transcripts such as in the proprotein convertase subtilisin/kexin type 1 (PCSK1) inhibitor have also been associated with an obese phenotype. Patients with homozygous/compound heterozygous mutations in PCSK1 suffer from obesity. This could be the result of impaired POMC processing, since similar phenotypic aspects, such as glucocorticoid deficiency, are also seen inpatients POMC mutations [206]. Other mutations in the leptin-melanocortin pathway, for example in the accessory proteins interacting with the melanocortin receptors, could also result in early onset obesity. Mice with disruption of the melanocortin-2 receptor accessory protein 2 (Mrap2), a modulator of Mc4r, show an obese phenotype [214, 215]. In human obesity, rare MRAP2 variants have been identified, but their exact role in obesity has to be further explored.
In addition to these causes of monogenic obesity without intellectual deficit, where a variant in one gene is causing the phenotype, obesity can also be part of a syndrome, where obesity is associated with congenital malformations, dysmorphic features and/or intellectual deficits. An overview of syndromic causes of obesity is provided in Table 1, but will not be further elaborated in this review [216].
4.2 Polygenic obesity
Monogenic causes of obesity are rare and account for approximately 8-12% of (severe) childhood-onset obesity (ref) and for approximately 4% of severe obesity in the general population [231]. In most individuals, however, genetic predisposition to obesity is expected to be polygenic, in which the phenotype is caused by the additional effect of variants in multiple genes. In 2007, common variants in specific parts of fat mass and obesity associated (FTO) gene were associated with a higher BMI in humans [232-235]. Subsequently, other polygenic variants were identified after large meta-analysis of genome wide association studies (GWAS) for BMI. Altogether, these studies made it possible to develop and apply genetic risk scores for the determination of the role of polygenic variants in obesity [236]. Domingue and coworkers showed that these genetic risk scores are positively correlated with BMI [235]. Notably, due to the small effect size of each single gene variant, most GWA studies are underpowered. Therefore, definite conclusions cannot be drawn from these studies. Hence, future research needs to focus on larger patient cohorts to further elucidate genetic variances in obesity prone genes.
4.3 Genetic obesity: implementation in clinical care
Knowledge of genetic obesity syndromes and the molecular mechanisms underlying these syndromes is crucial for reproductive decision making, reducing obesity stigma, and the discovery of novel mechanism-based pharmacologic treatments. Diagnosing genetic obesity has already led to personalized therapies for obesity. The most illustrative example is the successful treatment with subcutaneous leptin injections in patients with congenital leptin deficiency. In these patients, leptin therapy has also shown to be effective for other associated symptoms, such as preventing recurrent infections and inducing puberty [180]. Notably, leptin therapy in patients without (congenital) leptin deficiency has no effect [176]. The newest therapeutic agent in genetic obesity is the MC4R-agonist setmelanotide. This drug showed promising results in two patients with proopiomelanocortin deficiency with a weight reduction of 20.5 kg after 12 weeks in the first and 51 kg after 42 weeks in the second patient [181]. A follow-up study, using setmelanotide in patients with leptin receptor deficiency described similar results [217]. Finally, clustered regularly interspaced short palindromic repeats (CRISPR) and its associated protein (Cas) genome editing techniques have potential to treat genetic (obesity) syndromes. Developments in this technique are therefore closely monitored by all experts in the field of genetics.
5. Conclusion
Significant advances in understanding the pathophysiological mechanisms in the development and maintenance of obesity have been made. Obesity seems to be the result of impaired brain circuits and neuroendocrine feedback associated with pathological overeating and physical inactivity. A small proportion of the obese population is affected by a monogenetic mutation causative of obesity. Additionally, many obesity susceptible genes have been identified in GWA studies. Despite our increased knowledge on the development and progress of obesity, our understanding of its etiology and pathophysiology is still incomplete. In particular, longitudinal follow-up and controlled intervention trials are required to resolve issues of causality. Nevertheless, due to the accelerating effects of obesity on metabolic outcomes and cancer, it has the potential to be pernicious to mankind if preventive measures and/or effective therapies are not realized.