Advanced neuroimaging in obesity

Abstract

The Obesity Society, founded in 1982, conducts the most comprehensive annual scientific meeting on obesity in North America. Annually, the meeting brings together over 2000 scientists and healthcare professionals of various specialties to allow a multidisciplinary forum on this burdensome disease. Numerous basic science presentations addressed the importance of the brain’s role in obesity. Investigating the role of the human brain in obesity has been limited until more recent advances in biomedical imaging. This review details the symposium that was held at this year’s annual meeting of The Obesity Society on ‘Neuroimaging in obesity’, bringing together current leader’s in this emerging field.

Keywords::

• dopamine
• functional MRI
• neuroimaging
• obesity
• PET
• receptor

Figure 1. Dopamine type 2 receptor binding (B_max/Kd) versus BMI in ten obese patients.

Redrawn with permission from [16].

Figure 2. Activation in the caudate nucleus of the dorsal striatum in response to food receipt versus BMI.

Redrawn with permission from [18].

Innovations in biomedical imaging have greatly impacted biological research and the treatment and diagnosis of disease [1]. Standard imaging was developed to define structure; however, a growing number of techniques allow for measures of function. Imaging to measure function is especially important since neural activity is coupled to glucose metabolism and regional cerebral blood flow (rCBF), hence enabling a better estimate of neural activity. There are various techniques that allow the measurement of the cerebral glucose metabolism and rCBF [2], many of which have been applied to the study of the brain’s role in obesity [3,4]. Of particular note is functional MRI (fMRI) and blood oxygen-level-dependent (BOLD) signals. BOLD signals are based on the magnetic properties of hemoglobin that change with the state of oxygenation, allowing a measure of rCBF that reflects neural activation [5].

Dopaminergic neurotransmission is pivotal for sensing reward, which is necessary for survival, but it also susceptible to pathologic disruption. Inability to synthesize dopamine leads to death by starvation that can be prevented by restoration of dopamine in the dorsal striatum [6]. On the contrary, excess stimulation of the dopaminergic system occurs with intake of various substances of abuse and contributes to the development of addiction [7], a disease that is often refractory to therapy. Various imaging techniques have been utilized to investigate the role of dopaminergic neurotransmission in addiction. PET imaging utilizing radioligands specific for dopamine type 2/3 (D₂) receptors allows for measurement of D₂ receptor levels, the most abundant dopamine receptor and essential for sensing of reward [7]. Currently employed D₂ radioligands compete with the endogenous dopamine; therefore, performing PET imaging with a D₂ radioligand both with and without exposure to a stimulus that increases endogenous dopamine in the synapse allows estimation of dopamine levels [8]. Since growing evidence has revealed the importance of dopaminergic neurotransmission in control of eating behaviors, these techniques have been recently applied to human obesity.

Nora Volkow, director of the National Institute on Drug Abuse (NIDA; MD, USA), opened the symposium on ‘Neuroimaging in obesity’. Volkow has spent a great deal of her career utilizing neuroimaging to discern the importance of dopaminergic neurotransmission in various disease states but most notably addiction [7]. Volkow and others have shown that decreased D₂ receptor binding occurs in the striatum in cocaine [9], alcohol [10], methamphetamine [11] and opiate [12] addiction. The striatum, which includes the dorsal and ventral striatum, has an important influence over the cerebral cortex and is especially involved in decision making involving reward processing and motivation [13]. Overall, it is considered that reduced dopaminergic neurotransmission in addiction, partially resultant of low D₂ receptors, contributes to a decreased sense of reward and compensatory behavior of ongoing drug intake.

As proof of this principle, Volkow and colleagues used a D₂ receptor adenoviral vector to temporarily increase expression in the nucleus accumbens (ventral striatum) in rodent models of alcohol [14] and cocaine [15] addiction. Their work revealed that directly increasing D₂ receptor levels dramatically reduced the intake of the drug and then as receptor levels returned to baseline low levels, excessive drug intake resumed.

Volkow and colleagues further investigated how decreased striatal D₂ receptors in addictive states could influence the cerebral cortex by using PET imaging with ¹⁸F-fluorodeoxyglucose (FDG) to measure cerebral glucose metabolism. In cocaine [9] and methamphetamine [11] addiction, decreased striatal D₂ receptor binding is associated with reduced prefrontal cortex metabolism and, therefore, reduced neuronal activation. Since the prefrontal cortex is pivotal for inhibitory control, Volkow hypothesized that this contributes to an impaired ability to inhibit unwanted behaviors, such as intake of substances of abuse.

In 2001, Wang and Volkow published a landmark study in human obesity. They utilized PET imaging to demonstrate that low levels of striatal D₂ receptor binding occur in human obesity [16]. In the obese, D₂ receptor binding was negatively correlated with BMI (Figure 1); hence, the more severe the obesity, the lower the D₂ receptor levels. PET imaging with a D₂ radioligand with and without exposure to food reveals that the pleasure of actual food consumption is directly correlated with dopamine release in the dorsal striatum [8]. Therefore, these studies extend the concept of reduced dopaminergic neurotransmission in drug addiction to obesity, as low D₂ receptors found in obese individuals would contribute to reduced reward with eating and, ultimately, excessive food intake to compensate.

Under a similar scientific premise as in their addiction studies, Volkow went on to compare cerebral glucose metabolism measured with FDG scans, and estimate D₂ receptor levels in obese individuals [17]. A similar relationship was demonstrated in obesity as in drug addiction, where D₂ receptor binding was associated with reduced prefrontal metabolism, extending the idea that reduced D₂ receptor contributes to an impairment to inhibit unwanted behaviors – in this case, overeating in obese individuals. Volkow stressed that this work does not indicate that all of obesity is a state of ‘addiction’, rather that similar pathways, especially dopaminergic pathways that are involved in sensing pleasure from natural rewards (food), are also involved in sensing pleasure from unnatural rewards(drugs of abuse). Just as altered neurotransmission can perpetuate maladaptive behaviors in drug addiction, similar processes may occur in obesity in reference to eating behaviors. There is likely an underlying predisposition in individuals who develop altered neurotransmission, as not everyone who uses drugs becomes addicted nor everyone who occasionally overeats develops obesity. Investigations into susceptibility to obesity would be valuable.

Eric Stice from the Oregon Research Institute (OR, USA) presented his laboratory’s work utilizing fMRI with BOLD signals to provide further insight into traits that increase susceptibility to weight gain. Stice and colleagues imaged young females with a paradigm that utilized a combination of cues and receipt of a taste of chocolate milkshake or tasteless solution to distinguish between the brain regions involved in consummatory and anticipatory food reward – two different and relevant aspects of eating behavior [18]. Imaging revealed decreased neural activation in the dorsal striatum in obese compared with lean participants in response to taste of the chocolate milkshake (Figure 2). Even though fMRI does not allow a direct measure of dopaminergic signaling, it does support the finding of decreased D₂ receptors in this region [16] and the relevance of the dorsal striatum to consummatory food reward [8]. They also found obese participants to have increased activation in the primary gustatory cortex and the somatosensory cortex in response to anticipatory food reward, suggesting that obese individuals have increased sensitivity to the anticipation of food. Therefore, their work indicated separate neural circuits for consummatory and anticipatory aspects of food reward that could contribute to ongoing vulnerabilities to obesity, yet it is uncertain whether the differential activations are consequences of obesity or predate the disease [18]. To begin to distinguish between the two, Stice and colleagues investigated how the genetic predisposition to decreased striatal D₂ receptors levels could predispose to obesity. They employed fMRI with BOLD signals during food receipt in young females with and without the TaqIA polymorphism; the presence of the A1 allele is associated with an approximately 30% decrease in striatal D₂ receptor levels [19]. They found that in those with the A1 allele, a blunted dorsal striatal response to food intake predicted increased weight gain over 1 year, providing evidence that low levels of D₂ receptors actually predispose to weight gain and are not just a consequence of obesity [19].

Much of the morbidity associated with obesity is resultant of the metabolic changes due to excess bodyweight. Insulin and leptin are peripherally secreted hormones involved in metabolic diseases, such as diabetes mellitus, but are also pivotal central signals that promote appetite suppression. Both insulin and leptin action are suppressed in obesity [20]. Joy Hirsch, the Director of the Program in Imaging and Cognitive Sciences at Columbia University (NY, USA), spoke on her work using fMRI with BOLD signals to evaluate leptin’s role in cerebral response for food cues during maintenance of weight loss. Hirsch and colleagues studied individuals with obesity at three time points: weight stable at baseline, maintenance of a 10% weight loss, and maintenance of a 10% weight loss with leptin therapy. Leptin decreases during weight loss and replacing leptin during weight maintenance improves certain negative metabolic effects of weight loss, such as reduced energy expenditure and suppressed thyroxine [21]. This study aimed to determine the effects of replacing leptin to pre-weight-loss levels on the central response to food cues.

Leptin replacement restored activation in the hypothalamus, the center of the control of homeostatic eating behaviors, which was diminished by weight loss. Other areas with similar activity in response to food cues at baseline while maintaining weight loss with leptin therapy were the cingulate gyrus and middle frontal gyrus, which are involved in dietary restraint and aspects of executive function, respectively. During maintenance of weight loss (without leptin therapy), increased neural activity occurred in key limbic areas, including the brainstem, parahippocampus and globus pallidus. While leptin therapy had little effect on activity in the brainstem and parahippocampus, it did decrease activity in the globus pallidus, which mediates reward responses to highly palatable foods. Various studies have revealed that weight loss and obesity lead to altered neural activity and it is considered that these changes contribute to the challenge of weight loss and the high rate of weight recidivism [3,22]. Hirsch and colleagues revealed that replacing a peripherally secreted hormone can reverse certain alterations in neural activity. Their work brings to the forefront a highly relevant observation regarding obesity. Alterations of peripherally secreted hormones during weight-loss efforts not only impair the metabolism’s ability to maintain lower weight but may also impair an individual’s ‘central’ ability to lose and maintain weight loss.

The work of these investigators has contributed greatly to the understanding of obesity. Their work reveals obesity is not simply a failure of ‘will power’ but a complex cascade that occurs in those with underlying susceptibility and functions to prevent weight loss, even when weight is detrimentally excessive. Further investigations are needed to build a more comprehensive knowledge of the central control of obesity and how it is influenced by peripheral hormones. Such work will open avenues for interventions to prevent and treat obesity.

Acknowledgements

We thank Robyn Tamboli for editing support.

Financial & competing interests disclosure

Julia P Dunn received support from the Vanderbilt Environmental Health Science Scholars Program (NIEHS K12 ESO15855). This work was supported by NIH grants RO1-DK070860, National Institute of Diabetes and Digestive and Kidney Diseases to NNA, and DK20593, to the Vanderbilt Diabetes Research and Training Center. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.