To the Editor: There is evidence that certain brain areas are functionally altered in obesity. In obese humans, we and others have reported a reduction in dopamine D2/3 receptor (D2/3R) binding in the striatum, an important component of the brain reward system [1–3]. The neurotransmitter dopamine is important for the reinforcing value of food and it has been shown that food can induce a release of endogenous dopamine in the striatum [4]. Obese individuals are thought to be more sensitive to food reinforcement than those who are non-obese. This may underlie the notion that obese humans experience an increased craving for food. In addition, striatal D2/3R availability has been linked to craving and diet-induced obesity [5]. Therefore, it is plausible that dopamine-related mechanisms linked to craving and impulsiveness play a role in the development and pathophysiology of obesity. At present, it is unclear whether lower D2/3R availability is a cause or an effect of obesity. If the latter is true, one would expect that the reduced D2/3R availability observed in obese humans is reversed following the loss of a clinically significant fat mass or during a hypoenergetic state.

Obesity is associated with insulin resistance. Insulin receptors are widely expressed in the human brain, and a relationship between insulin sensitivity and central dopamine signalling has been suggested [6]. It needs to be determined, therefore, whether the striatal D2/3R binding potential correlates with hepatic or peripheral insulin sensitivity.

To determine whether the previously reported reduction in D2/3R availability in obese humans is reversible, we studied striatal D2/3R availability before and 6 weeks after Roux-en-Y gastric bypass (RYGB) surgery in 19 morbidly obese women. Eighteen of these patients participated in a study on the short-term metabolic effects of RYGB surgery (NTR1548; for one additional patient only single photon emission computed tomography [SPECT] data could be acquired) [3]. Informed consent was obtained from all participants and the study was approved by the local medical ethics committee of the Academic Medical Center in Amsterdam. We assumed that a difference of 15% in D2/3R availability would be of clinical relevance. A power analysis indicated that we would be able to detect this difference in a study group of 18 individuals. In the morbidly obese women, striatal D2/3R availability was assessed using a brain-dedicated SPECT scanner and [123I]iodobenzamide (bolus/constant infusion technique). Acquisition, attenuation correction, reconstruction and analyses of SPECT data were performed as previously described [1]. Apart from a classic region-of-interest (ROI) analysis, data also underwent an MRI-driven analysis. In this additional analysis, SPECT images were co-registered to individual MR images, and ROIs were drawn for the caudate nucleus and putamen separately and for the occipital cortex (representing non-specific binding) on the MR images [7]. Insulin sensitivity was determined before surgery by a two-step hyperinsulinaemic–euglycaemic clamp using a stable glucose isotope tracer [8]. Data for four women were not complete owing to technical failures and were excluded from analysis.

The women had a mean age of 40.4 ± 8 (26–49) years (mean ± SD [range]). Weight loss 6 weeks after RYGB was 14 ± 4.6 (8–24) kg, which resulted in a significant reduction in BMI after surgery (before surgery 45.7 ± 6.3 [38.7–61.3] and after surgery 40.9 ± 6.3 [34.1–57.6] kg/m2; p  <  0.001). The ROI analysis showed no significant change in D2/3R availability before vs 6 weeks after RYGB (Fig. 1a). Also, in the MRI-driven analysis, the D2/3R availability in the striatum as a whole and in subregions of the striatum (caudate nucleus and putamen) did not change significantly after surgery (Fig. 1b). There was no correlation between BMI and D2/3R availability before (p  =  0.35; r 2 = 0.054) or after surgery (p  =  0.51; r 2 = 0.027). This suggests that striatal D2/3R is not regulated by acute changes in energy balance and is not influenced by fat mass per se.

Fig. 1
figure 1

Striatal D2/3R availability of 18 obese women before and 6 weeks after bariatric surgery according to (a) classic ROI analysis (without MRI co-registration), mean BPND before vs after surgery (p  =  0.415); (b) boxplots showing MRI-driven analysis expressed for the whole striatum as well as striatal subregions (caudate, p  =  0.272; putamen, p  =  0.661; striatum p  =  0.842, before vs after surgery). White bars represent BPND before RYGB, grey bars represent BPND after RYGB. (c) Striatal D2/3R availability of 16 obese women before bariatric surgery vs insulin-mediated rate of peripheral glucose disappearance (Rd) (r 2  =  0.156; p  =  0.14)

Full size image

Earlier studies on D2/3R binding after bariatric surgery are contradictory and report either an increase [2] or a decrease [3] in striatal D2/3R availability 6 weeks after surgery. Whether a change in D2/3R occurs after long-term weight loss is unknown. Interestingly, when correlating the pre-surgery level of peripheral insulin sensitivity with pre-surgery striatal D2/3R binding availability, a trend for a positive correlation was observed (Fig. 1c), but there was no correlation with hepatic insulin sensitivity. The former is in line with the general insulin-sensitising effects of dopamine agonists in obese diabetic individuals. In addition, dopamine antagonists are known for their negative effects on insulin sensitivity, and drug-naive schizophrenic patients, known for their disturbed central dopamine metabolism, are characterised by hepatic insulin resistance [9]. Furthermore, a correlation between the insulin sensitivity index (SI) and D2/3R availability in the ventral striatum has been reported [10]. This suggests that peripheral glucose uptake, predominantly occurring in skeletal muscle when under hyperinsulinaemic conditions, might be in part regulated by cerebral dopamine metabolism. Although a clear difference in D2/3R availability was found between lean and obese individuals [2], within our obese group no clear correlation between BMI and D2/3R availability was found. This suggests that fat mass per se is not the main determinant of D2/3R availability in obesity. This is in line with the unchanged D2/3R availability, despite clinically significant weight loss. Our study is limited to women only and a potential bias of the menstrual cycle on our outcome cannot be excluded.

In conclusion, surgery-induced weight loss does not significantly increase striatal D2/3R availability in morbidly obese women. This suggests that short-term changes in energy balance in morbidly obese humans do not induce profound alterations in striatal dopaminergic neurotransmission and might predispose obese individuals to regain weight after a hypoenergetic diet. Moreover, the striatal dopamine receptor binding potential is not significantly correlated to hepatic insulin sensitivity but shows a trend for a positive correlation with peripheral insulin sensitivity. This adds to earlier findings on a potential role for cerebral dopamine in glucose metabolism.