Obesity is now a major health problem in many societies worldwide. In this issue Stankowska and Gjedde have provided an excellent overview of studies of dopamine (DA) neurotransmission in the development of obesity, and propose an innovative hypothesis regarding the evolution of changes in DA neurotransmission with increasing body mass index (BMI). This hypothesis integrates studies of sensation seeking, which is both found at higher levels in the obese and is believed to be a risk factor for obesity, studies of reward sensitivity and studies of DA neurotransmission in obesity. While there are some monogenetic polymorphisms that lead to obesity, the overwhelming majority of cases of human obesity appear to result from the interaction of polygenetic predispositions to obesity with an environment in which highly palatable food is readily available (Reference Mutch and Clément1–Reference Blakemore and Froguel4). The excessive consumption of highly palatable foods is mediated by their rewarding properties (Reference Mela5–Reference Cohen7). DA neurotransmission in the brain reward circuit, particularly the ventral striatum/nucleus accumbens, plays an important role in regulating reward behaviours (Reference Haber and Knutson8,Reference Wise9). In addition, DA neurotransmission in the dorsal striatum is required for normal eating behaviours and has a permissive role in eating behaviours (Reference Palmiter10). The importance of DA neurotransmission in the consumption of highly palatable foods has resulted in numerous animal studies of DA neurotransmission in the development of obesity as well as a number of studies of human brain function in obesity, but relatively few studies that directly examine DA neurotransmission in humans with the development of obesity.
There have been two theories regarding the changes in DA neurotransmission that predispose to the development of obesity. The first postulates a reward deficient state that leads to compensatory overeating to remedy an anhedonic state; decreased DA neurotransmission in the brain reward circuit is hypothesised to be a significant factor in the reward deficient state leading to obesity (Reference Blum, Braverman and Holder11,Reference Blum, Liu, Shriner and Gold12). Consistent with this hypothesis are the findings of decreased striatal DA D2 receptor levels in the dorsal striatum in extremely obese human subjects (BMI > 45), and animal models of diet-induced obesity that consistently demonstrate decreased ventral striatal/nucleus accumbens DA release and DA D2 receptor levels (Reference Wang, Volkow and Logan13–Reference Johnson and Kenny18). The development of compulsive food ingestion in animals has been related to the development of decreased striatal DA D2 receptor levels in the setting of decreased DA release and elevated reward thresholds (Reference Johnson and Kenny18). These changes are similar to those seen in alcohol and drug abuse which are characterised by decreased DA release, decreased striatal/ventral DA D2 receptor levels, and anhedonia (Reference Leventhal, Brightman and Ameringer19–Reference Volkow, Chang and Wang25). This has led to the concept of obesity being a food addiction (Reference Gearhardt, Corbin and Brownell26,Reference Volkow, Wang, Tomasi and Baler27). The second hypothesis postulates that in individuals prone to obesity there is an increased behavioural salience of highly palatable foods resulting in increased intake of these foods leading to obesity. The increased behavioural salience of food, that is food wanting, is postulated to be mediated by increased ventral striatal/nucleus accumbens DA release in response to highly palatable food stimuli (Reference Wise9, Reference Berridge, Robinson and Aldridge28, Reference Smith, Berridge and Aldridge29). Consistent with this hypothesis are (a) fMRI studies showing greater food related activation in the brain reward circuit both in subjects at high risk for obesity and obese individuals (Reference Stoeckel, Weller, Cook, Twieg, Knowlton and Cox30–Reference Stice, Yokum, Burger, Epstein and Small36), (b) a strong correlation between food cue activation and individual differences in reward drive (Reference Beaver, Lawrence, van Ditzhuijzen, Davis, Woods and Calder37) (c) the role of brain reward circuit DA neurotransmission in food motivation and food seeking energy expenditure (Reference Salamone, Correa, Farrar and Mingote38), and (d) a recent study suggesting increased DA release in binge eating disorder (Reference Wang, Geliebter and Volkow39).
Higher levels of sensation/novelty seeking behaviour are seen in obese as well as in substance abusing subjects. Sensation/novelty seeking as well as impulsivity appear to be personality traits that predispose to both obesity and substance abuse (Reference Sullivan, Cloninger, Przybeck and Klein40,Reference Jupp and Dalley41). In humans both sensation/novelty seeking and impulsivity are mediated by DA neurotransmission (Reference Leyton, Boileau, Benkelfat, Diksic, Baker and Dagher42–Reference Buckholtz, Treadway and Cowan45). Gjedde has recently reported an inverted ‘U’ shaped relationship between sensation seeking and striatal DA D2 receptor levels in healthy subjects (Reference Gjedde, Kumakura, Cumming, Linnet and Møller46). This inverted ‘U’ shaped relationship is believed to reflect higher occupancies of DA D2 receptors by endogenous DA at either end of this curve. A similar inverted parabolic relationship has been reported between reward sensitivity and BMI (Reference Davis and Fox47,Reference Verbeken, Braet, Lammertyn, Goossens and Moens48). Stankowska suggests that both the relationships of BMI to reward sensitivity and sensation seeking to available striatal DA D2 receptor levels reflect a common aetiology that is related to the evolution of DA neurotransmission with the development of obesity. Such an evolution moves beyond the static concepts of reward deficiency or increased behavioural salience that have previously been proposed as risk factors for obesity. The postulated evolution proposes that at low BMI's, that is BMI's of 18 or less, there are decreased extracellular DA levels along with relatively larger decreases in total levels of DA D2 receptors leading to fewer available DA D2 receptors consistent with animal studies (Reference Pothos, Creese and Hoebel49). The decreased extracellular DA levels are believed to mediate the lower levels of sensation seeking on the left hand side of the curve of sensation seeking versus DA D2 receptors while the decreased available DA D2 receptor levels are hypothesised to mediate the decreased reward sensitivity seen at underweight BMI's. Normal weight to perhaps mildly obese subjects, BMI's of 19 up to the low 30's, in comparison to underweight subjects are postulated to have the highest levels of available DA D2 receptors due to relatively greater increases in total DA D2 receptor levels than extracellular DA; the greater reward sensitivity in these subjects is hypothesised to be due to the higher levels of available DA D2 receptors while the increased DA release is associated with increased sensation seeking. As BMI increases above 35, further increases in both extracellular DA and total DA D2 receptors are postulated to occur. However, a greater increase is believed to occur in extracellular DA levels than in DA D2 receptor levels producing decreased available DA D2 receptor levels; decreased available striatal DA D2 receptor levels have been reported in very obese humans consistent with this hypothesis (Reference Wang, Volkow and Logan13,Reference de Weijer, van de Giessen and van Amelsvoort14). The decreased available DA D2 receptor levels in moderately to severely obese subjects is believed to account for the decreased reward sensitivity while the increased DA release is hypothesised to mediate the high levels of sensation seeking seen in these subjects.
However, Stankowska acknowledges that, given the paucity of human studies of DA neurotransmission in obesity, alternative explanations are possible. Many of the proposed changes in DA neurotransmission in obesity in Stankowska's review are based on Gjedde's study of sensation seeking and DA D2 receptor levels in healthy subjects (Reference Gjedde, Kumakura, Cumming, Linnet and Møller46). The BMI of the subjects studied in Gjedde's study is not stated. While Gjedde's findings in healthy subjects may be accurate in describing the relationship between sensation seeking and DA D2 receptor levels in healthy, nonobese subjects, there may be changes in the relationship of DA neurotransmission to reward-related behaviours in individuals who are very obese or significantly underweight.
As discussed above, Stankowska proposes that total striatal DA D2 levels and extracellular DA levels are increased in very obese human subjects. While no studies of baseline extracellular DA levels and total DA D2 receptor levels, which would require PET studies of DA D2 receptors before and following DA depletion, have been reported in obese humans, animal studies have shown decreased baseline extracellular DA levels, decreased amphetamine induced DA release, and decreased absolute levels of DA D2 receptor levels in obese animals, particularly in those who compulsively consume food (Reference Geiger, Behr and Frank16–Reference Johnson and Kenny18). Both Stankowska and other investigators have postulated that extreme obesity may be due to a food addiction and that extremely obese humans may have changes in DA neurotransmission similar to those seen in substance abuse (Reference Gearhardt, Corbin and Brownell26,Reference Volkow, Wang, Tomasi and Baler27,Reference Everitt and Robbins50). Studies in human substance abusers have shown both decreased extracellular DA levels, and total DA D2 receptor levels as well as decreased psychostimulant induced DA release and available DA D2 receptor levels (Reference Martinez, Gil and Slifstein21–Reference Volkow, Chang and Wang25). The limited available data would suggest decreased rather than increased extracellular DA levels and total DA D2 receptor levels in the extremely obese.
While acknowledging that ‘the changes in dopaminergic neurotransmission in mild to moderate obesity are unclear’, Stankowska proposes that extracellular DA levels, available and total DA D2 receptor levels increase as BMI's increase from an underweight range, that is 18 or under, up to at least a normal to overweight level and possibly extending into the mildly obese range, that is BMI's of 30–35. While no studies of baseline receptor occupancy by extracellular DA versus BMI have been published, we have presented preliminary data regarding d-amphetamine induced DA release in 16 healthy subjects with BMI's ranging from 19 to 35 – low normal to mildly obese (Reference Kessler, Zald, Ansari, Li and Cowan51). Positive correlations of BMI with DA release were seen in all regions except the right temporal cortex (r = −0.17) with significant positive correlations seen in the right putamen (r = 0.581, p = 0.023) and in the left substantia nigra (r = 0.568, p = 0.027). These results are consistent with increased extracellular DA levels with increasing BMI's from 19 to 35 and are consistent with Stankowska's model. In regard to available DA D2 receptor levels, Haltia reported no change in striatal DA D2/3 receptor levels in overweight and mildly obese subjects (mean BMI = 33, BMI > 27) compared with normal weight subjects (mean BMI = 22, BMI < 24) (Reference Haltia, Rinne and Merisaari52). We have recently presented data consistent with the Haltia study showing nonsignificant decreases, rather than increases, in available DA D2 levels in striatal and most extrastriatal regions over a BMI range of 19–35 in 34 healthy subjects (Reference Kessler, Zald, Ansari, Li and Cowan51). These results, while preliminary, suggest increasing extracellular DA levels as BMI's rise from underweight to mildly obese levels, but with little change in available DA D2 receptor levels.
Although the evolution of changes in DA neurotransmission occurring with the development of obesity are not completely understood at present, the limited available studies in humans do suggest that significant changes in DA neurotransmission occur (Reference Wang, Volkow and Logan13,Reference de Weijer, van de Giessen and van Amelsvoort14,Reference Kessler, Zald, Ansari, Li and Cowan51,Reference Haltia, Rinne and Merisaari52). There are very few studies of the factor(s) mediating changes in DA neurotransmission with increasing BMI in humans. While the DA transporter is an important regulator of extracellular DA levels, a recent study has shown no correlation of BMI with DA transporter levels in humans (Reference van de Giessen, Hesse, Caan, Zientek, Dickson, Tossici-Bolt, Sera, Asenbaum, Guignard, Akdemir, Knudsen, Nobili, Pagani, Vander Borght, Van Laere, Varrone, Tatsch, Booij and Sabri53). Altered enteric hormone regulation of DA neuronal function in the brain reward circuit with the development of obesity has been demonstrated in animals and recently a single study reporting significant correlations of cerebral DA D2 receptor levels with enteric hormone levels has in humans has been published (Reference Figlewicz and Benoit54–Reference Dunn, Kessler and Feurer57). Obesity induced changes in enteric hormone regulation of cerebral DA neurotransmission appear to be a potentially important factors in mediating altered DA neurotransmission and altered reward behaviours with the development of obesity.
In summary, the excellent review of Stankowska and Gjedde puts forth an important hypothesis regarding the evolution of DA neurotransmission in the development of obesity and the role that such changes in DA neurotransmission produce in reward-related behaviours, particularly sensation seeking, in predisposing to the development of obesity. As noted by the authors, alternative hypotheses are possible given the paucity studies of DA neurotransmsion in humans with the development of obesity. There are numerous animal studies which demonstrate that DA neurotransmission plays a critical role in the development of obesity related to increased highly palatable food ingestion. However, we are at a relatively early stage in delineating the nature of such changes in humans and the factor(s) that drive these changes.