SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References

Objective: Neurotransmitter systems participate in the regulation of food intake, and their activities are expected to influence eating behavior.

Design and Methods: We investigated possible associations between body mass index (BMI) and central noradrenaline, serotonin, and dopamine activities, as reflected by the cerebrospinal fluid levels of their main metabolites methoxyhydroxyphenylglycol (MHPG), 5-hydroxyindoleacetic acid (5-HIAA), and homovanillic acid (HVA), respectively. We studied 192 subjects (111 males, 81 females) admitted to neurologic clinic for diagnostic investigations that included CSF analysis, and were found not to suffer from any major neurological disease. Subjects were categorized in three groups, namely in lower, in the two middle, and in upper BMI quartiles, the limits calculated separately for males and females.

Results: No differences were found in MHPG levels between groups, while subjects in the upper BMI quartile showed significantly elevated levels of 5-HIAA and HVA compared to the levels of subjects in lower and middle quartiles.

Conclusions: The results provide evidence that in overweight subjects there are enhanced demands in serotoninergic and dopaminergic signaling for their reward system that may lead to increased motivation for food consumption. The implication of reward centers in eating behavior supports the hypothesis of common mechanisms in obesity and drug addiction.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References

Regulation of food intake is a highly complex process involving many peripheral and central pathways, in an interplay of hormones, neuropeptides, and neurotransmitters [1, 2]. Dopamine, the most studied neurotransmitter, has been shown to interact with peripheral and central factors to regulate both homeostatic and hedonic aspects of food intake [3]. Increases in dopamine signaling promote feeding to some limiting value after which dopamine signaling inhibits feeding [4]. Leptin and insulin also tend to inhibit the activation of dopamine neurons while ghrelin fosters it. Experimental models with dopamine deficient mice have demonstrated that these animals are aphagic and do not survive unless dopamine synthesis is restored [5, 6]. Dopamine seems to energize feeding and to reinforce food-seeking behavior, while food reward in turn elevates dopamine levels [7].

Human obesity was associated with chronic elevation of brain serotonin turnover in the study of Lambert et al. [8]. Turnover was calculated using veno-arterial plasma concentration differences in 5-hydroxyindoleacetic acid (5-HIAA), the main serotonin metabolite. Serotonin turnover increased postprandially both in lean and obese subjects. No associations to central noradrenaline turnover (measured by the differences in its main metabolite methoxyhydroxyphenylglycol (MHPG)) were found.

A strong positive association between body mass index (BMI) and density of serotonin receptors (5-HT4R) in reward circuits that regulate food intake was reported recently by Haahr et al. [9]. The authors speculate that stimulation of the 5-HT4R receptor could be considered in the treatment of human hedonic overeating.

Additional information regarding the involvement of neurotransmitter systems in weight control comes from studies of subjects with eating disorders, usually estimating CSF neurotransmitter metabolite levels or brain receptor binding capacities. Underweight subjects with anorexia nervosa have been found to have reduced levels of CSF 5-HIAA, the main serotonin metabolite, which normalized after weight gain [10]. Reduced CSF levels of homovanillic acid (HVA), the main dopamine metabolite, were also found in malnourished individuals with anorexia nervosa, which persisted after recovery, while the levels of the main noradrenaline metabolite MHPG were normal [11].

The aforementioned studies indicate that underweight subjects may have reduced, and overweight subjects have enhanced serotonin and dopamine turnover compared to subjects with middle BMI values. In this study we searched for associations between BMI and CSF levels of the main metabolites of noradrenaline, serotonin, and dopamine, in apparently healthy subjects. To test the hypothesis that overweight is connected to increased serotonin and dopamine but not to noradrenaline turnover, we compared CSF 5-HIAA, HVA, and MHPG levels of the groups in the low, middle, and upper BMI quartiles, the limits of which were calculated separately for males and females.

Subjects and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References

We included in the study subjects who were admitted to the Neurological Clinic of the Athens University Medical School, Eginition Hospital, for diagnostic investigations that included CSF analysis, and who were found not suffering from any major neurologic disease. More specifically, we included 126 subjects (82 males and 44 females), evaluated by brain CT scan and CSF analysis for possible neurosyphilis with negative results, as well as 66 subjects (29 males and 37 females), evaluated for possible multiple sclerosis by brain MRI and CSF analysis, again with negative results. Both males and females were in the age range from 18 to 65 years. Informed consent was obtained from all subjects. Lumbar puncture and metabolite estimations were performed as described elsewhere [12].

Body mass index was calculated as weight in kilograms divided by the square of height in meters. The WHO classification for overweight (BMI ≥ 25) and obesity (BMI > 30) for both genders, used in epidemiological studies, was considered not appropriate for the present study. Differences in BMI among countries may exist [13], and we noticed significantly higher BMI values in males compared to females in the population of the present study. The aim of the study was to search for possible deviant CSF metabolite levels in overweight subjects. We thus calculated the limits of BMI quartiles separately for males and females, and categorized the subjects in three groups for comparison of the metabolite data: (a) the lower quartile, (b) the second and third quartiles merged, and (c) the upper quartile.

Multiple regression analysis was performed to test for associations of BMI to sex, age, and the three metabolites. Analysis of variance with age as covariate, followed by planned comparisons was used to evaluate differences in CSF metabolite levels in the three groups. Pearson correlation coefficients were calculated in searching for correlations between age, BMI, and CSF metabolite levels.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References

Mean values (±SD) and ranges for age and BMI, as well as quartile limits for male and female study subjects are shown in Table 1. Body mass index values were 27.1 ± 4.0 (range 18.3-38.7) for males and 25.2 ± 4.7 (range 16.5-39.1) for females, the difference being significant (P = 0.014). BMI was strongly correlated to age, both in males (r = 0.3721, P < 0.0001) and in females (r = 0.3143, P = 0.004).

Table 1. Mean values (±SD) for age and body mass index (BMI), and quartile limits of male and female subjects
GroupNAgeBMIBMI-quartile limits
LowerUpper
  1. a

    Significantly lower compared to male BMI (ANOVA, covariate age, P = 0.014).

Males11142.9±11.527.1±4.024.6930.45
Range 18−6518.3−38.7  
Females8139.8±11.425.2 ± 4.7a21.9728.47
Range 18−6516.5 − 39.1  
All Subjects19241.6 ± 11.526.3 ± 4.423.1829.01
Range 18−6516.5−39.1  

A multiple regression analysis with dependent variable BMI and independent variables sex, age, MHPG, 5-HIAA, and HVA for the 192 subjects (R = 0.4943, F5, 186 = 12.03, P < 0.0001) indicated significant associations of BMI to sex, age, 5-HIAA, and HVA. Further analysis was performed in searching for differences in metabolite levels in the three groups using ANOVA, where sex was considered by assigning subjects in sex-defined quartiles, and age as covariate. It has to be noticed that there were no differences in CSF metabolite levels between subjects in second and third quartiles that were merged for comparisons.

Age, BMI, and CSF metabolite levels of subjects in lower, two middle, and upper BMI quartiles and the results of ANOVA are presented in Table 2.

Table 2. Levels (ng/ml) of the main metabolites of noradrenaline (MHPG), serotonin (5-HIAA), and dopamine (HVA) in CSF of subjects categorized in sex-defined quartiles according to their body mass index (BMI)
BMI quartileNAgeBMIMHPG5-HIAAHVA
  1. Evaluation of the differences by analysis of variance with age as covariate, followed by planned comparisons.

Lower (1st)4835.3 ± 10.321.6 ± 1.97.00 ± 2.5916.2 ± 8.526.1 ± 11.0
Middle (2nd + 3rd)9642.7 ± 10.825.8 ± 2.07.09 ± 2.3316.9 ± 7.029.2 ± 12.0
Upper (4th)4845.7 ± 211.832.2 ± 2.77.21 ± 2.4822.6 ± 8.136.3 ± 12.1
F (2, 189)   0.189.408.67
P   0.830.00010.0002
LQ vs. MQ, P    0.590.46
LQ vs. UQ, P    0.00010.0001
MQ vs. UQ, P    0.00010.0002

No differences were found for any of the three metabolites between lower and middle quartiles. Subjects in the upper quartile though, i.e. with BMI more than 30.4 for males and more than 28.5 for females, had significantly higher CSF 5-HIAA and HVA levels compared to subjects in the lower or in the middle BMI quartiles. No differences regarding the noradrenaline metabolite MHPG were found. In previous studies, the expected significant correlation between 5-HIAA and HVA levels was present for the subjects in lower (r = 0.6198), middle (r = 0.4991), and upper (r = 0.6444) quartiles.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References

The role of dopamine in energizing feeding and food-seeking behavior through its participation in brain reward circuits has been recognized [7, 14]. Moreover, Kranz et al. [15], reviewing electrophysiological, pharmacological, genetic, and imaging studies, postulated a role for serotonin in emotional, motivational, and cognitive aspects of reward representation, possibly as important as for dopamine. In the study of Erritzoe et al. [16], BMI was negatively correlated to cortical and subcortical serotonin transporter binding. The higher levels of HVA and 5-HIAA found in the present study in subjects in the upper BMI quartile indicate that dopamine and serotonin may be implicated in food-seeking behavior. Food reward elevates dopamine levels [7], so that motivation to eat and food reward capacity is expected to be elevated in overweight subjects.

It is suggested that humoral factors from the periphery influence the dopaminergic system and affect food intake, either stimulating, like ghrelin, or inhibiting dopamine signaling, like leptin or insulin [4]. Serum leptin levels correlate positively to percentage of body fat and to BMI, and are higher in women than in men, both in normal-weight and obese subjects [17]. Leptin signals nutritional status to other physiological systems and modulates their function [18]. When administered to animals it reduces the activation of dopaminergic neurons and decreases food intake [19].

The question of whether increased dopamine turnover in overweight subjects is a cause or effect of increased food intake has been addressed in several studies [14, 20]. Volkow et al. [20] proposed that enhanced dopaminergic neurotransmission contributes, as the authors write, to improved eating behavior by strengthening cognitive control mechanisms. An overactivity of the mesolimbic dopamine system that has been connected to the risk for addiction may increase the sensitivity to reward from food but also from addictive drugs, and Kenny [21] proposes common mechanisms in obesity and drug addiction. In the study by Davis et al. [22], sensitivity to reward correlated positively with measures of emotional overeating, and overweight women were more sensitive to reward than those of normal weight. If enhanced dopaminergic transmission characterizes reward dependent subjects, food intake is one way to increase dopaminergic signaling, and the results of Small et al. [23] show that the amount of dopamine release after meal consumption is associated with the degree of experienced pleasure.

The present study provides evidence that overweight subjects may have increased motivation to eat in order to maintain enhanced serotonin and dopamine stimulation that is required by their reward system. If so, drugs that cause increases in serotonin and dopamine signaling should be expected to have a reducing effect on food consumption. Haahr et al. [9] propose that pharmacological stimulation of serotonin receptors should be considered in the treatment of human hedonic overeating. Serotoninergic drugs have been used for the treatment of obesity [24], dopaminergic treatment of patients with Parkinson's disease decreases body weight [25], and sibutramine, a drug that increases both serotonin and dopamine release [26], has been shown to cause weight reduction in obese subjects [27].

The results indicate that in overweight subjects, the need for enhanced serotoninergic and dopaminergic signaling may lead to increased motivation for food consumption, while the implication of dopaminergic reward centers in eating behavior supports the hypothesis of common mechanisms in obesity and drug addiction [28].

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Subjects and Methods
  5. Results
  6. Discussion
  7. References