Higher urinary dopamine concentration is associated with greater ad libitum energy intake in humans

Abstract

Objective

We aimed to assess the relationship between dopamine, a neurotransmitter involved in feeding behavior, and ad libitum energy intake in humans.

Methods

Healthy individuals (n=158, 72 Native Americans/50 Whites/18 Blacks/18 Hispanics, BMI: 33±9 kg/m², body fat: 33±9%) were admitted for two inpatient studies investigating the determinants of ad libitum energy intake assessed for 3 days using a highly reproducible, computerized vending machine paradigm. Urine was collected for 24 hours during eucaloric conditions prior to the ad libitum feeding period and urinary dopamine excretion rate was quantified by high-performance liquid chromatography.

Results

Urinary dopamine excretion rate was on average 346±106 mcg/24h and positively correlated with BMI (r=0.28, p<0.0001). Higher dopamine concentrations were associated with lower cognitive restraint (ρ=−0.25, p=0.005) and greater total ad libitum energy intake (r=0.29, p=0.0002). However, after adjustment for anthropometrics, in Black and White cohorts, higher dopamine concentrations were associated with greater total ad libitum energy intake (r=0.70, p=0.001 and r=0.33, p=0.02, respectively), whereas no associations were observed in Native Americans or Hispanics (all p>0.3).

Conclusions

Higher urinary dopamine concentrations are associated with greater ad libitum energy intake, indicating a role for dopamine in the reward pathway regulating human feeding behavior.

Introduction

Obesity results from a prolonged imbalance between energy intake and expenditure, is associated with various metabolic diseases, and is considered a threat for public health (1). Many factors influencing both energy intake and expenditure, such as physiologic, environmental, and genetic characteristics, have been identified as determinants of weight gain (2). Physiological effectors of energy intake include neuronal and hormonal signals from periphery such as those from adipose tissue reflecting the state of energy stores (3) and from pancreas, skeletal muscle, liver and gut, which may regulate satiety and hunger in the central nervous system (CNS) (4). The hypothalamus and the brainstem receive input on energy store from multiple peripheral organs and brain regions with the aim of adjusting food intake to regulate energy balance (5). Several neurotransmitters are involved in the regulation of food intake (6) and, in particular, dopamine is a relevant brain messenger (7) modulating food reward via the meso-limbic circuitry of the brain (8). Dopamine is synthesized by DOPA decarboxylase from L-3,4-dihydroxyphenylalanine (L-DOPA), which in turn is produced from L-tyrosine by tyrosine hydroxylase (9). In mice, previous studies have shown that large DOPA-depleting brain lesions induced by injection of the neurotoxin 6-hydroxydopamine (6-HDA) led to pronounced reduction in food intake (10). In rats, dopamine depletion produced by intrastriatal injections of 6-HDA in the ventrolateral striatum in the hypothalamus decreases 24-h food intake (11). Mice with bilateral lesions of midbrain dopaminergic pathways have aphagia (12), and dopamine deficient mice are hypoactive and aphagic (13). In obese rats, the overconsumption of palatable food triggers addiction-like neuroadaptive responses in brain reward circuitries with downregulation of striatal dopamine D₂ receptor (D2R) that drives the development of compulsive eating (14). Lentiviral-mediated knockdown of D2R expression makes rats more susceptible to weight gain when allowed to consume calorically dense food (14). In rats, aphagia resulting from genetic disruption of dopamine production can be reestablished by refurbishment of dopamine signaling in the dorsal or ventrolateral striatum (15).

The knowledge of the behavioral functions of dopamine in the central nervous system is derived primarily from studies focusing on D2R in both mice and humans. In mice, blocking dopamine receptors (D₁ and D₂) reduces the rewarding actions of food (16). It has been hypothesized that alteration of central dopamine might be the cause of disturbed feeding behavior in human obesity (17). Also, in mice and humans, experimental stimulation or inhibition of dopaminergic signaling is related with alteration in motivation to eat, food reward and feeding behavior (18). Hypofunction of the dopamine-mediated reward system might be altered in obesity through the reduction of striatal D2R availability (19). Also, in overweight subjects opportunistic eating and obesity are positively associated with D₂-like receptor binding protein in the hypothalamic lateral striatum (20). In lean and obese adults, food reinforcement and energy intake is greater in obese individuals with the TaqI A1 allele which is associated with decreased D2R density (21). Women with obesity had greater activation in specific brain areas (insula, the Rolandic region, temporal, frontal, and parietal opercular regions) and increased anticipation of the palatability of the milkshake (versus tasteless solution) compared to lean individuals, suggesting a lower availability of dopamine receptors (22). These results are consistent with the observation that subjects with obesity have decreased D2R availability in proportion to their BMI (19). In lean healthy subjects who underwent two [11C]raclopride PET scans (one after an overnight fast and the other after consumption of subject’s “favorite meal), dopamine release in the dorsal striatum increases the meal and in proportion to the perceived pleasantness of the meal (23).

It has been hypothesized that dopamine, although not freely, crosses the blood-brain barrier through monoamines transporters and is filtered in the glomerulus entering the urine (24). Thereby, it can provide an index of overall dopamine neurotransmitter status in the central and peripheral nervous system (25). We aimed to evaluate the relationships between dopamine concentrations and body composition, eating behavior traits, and ad libitum food intake as objectively measured by a vending machine paradigm in a large, ethnic-diverse cohort of healthy individuals. We hypothesized that dopamine concentration assessed over 24 hours during eucaloric feeding is a determinant of ad libitum food intake.

Methods

Participants and study design

Data were collected from two different inpatient studies conducted in our research unit which were combined due to the similar procedures performed during the baseline period during each study (Figure 1). Ninety-six volunteers were admitted to the clinical trial NCT00342732 from 2009 to 2017 (The Food Intake Phenotype: Assessing Eating Behavior and Food Preferences as Risk Factors for Obesity) and 62 volunteers were enrolled to the clinical trial NCT00856609 from 2009 to 2015 (The Effects of Exenatide on Energy Expenditure and Weight Loss in Nondiabetic Obese Subjects). Thus, the present analysis was performed in 158 participants (age<55 years) recruited from the greater Phoenix area who had valid assessment of food intake measurement over 3-day period, as well as valid measurement of urinary dopamine concentrations. All subjects, non-smokers and free from any type of medication, were considered healthy based on medical history and laboratory testing. Prior to admission, all subjects signed written and informed consent. The protocols were approved by the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK).

Figure 1. Clinical study outline.

Day 1: Screening examination including labs and informed consent.

Day 2: TFEQ (Three-Factor Eating Questionnaire) administration.

Day 3: DXA (dual energy X-ray absorptiometry) scan.

Day 4: OGTT (oral glucose tolerance test).

Day 5: Twenty-four-hour assessment of energy metabolism inside the metabolic chamber and twenty-four-hour urine collection for measurement of dopamine concentrations in eucaloric conditions.

Days 1–5: Subjects were fed a weight maintaining diet (WMD)

Days 6–7-8: Assessment of ad libitum food intake by a computerized vending machine paradigm.

On the day of admission to our metabolic unit, subjects were instructed to restrict their activities to those available on the research unit during the inpatient stay and were placed on a standard weight-maintaining diet (50% carbohydrate, 30% fat, and 20% protein). For each individual, the weight-maintaining energy needs (WMEN) were initially calculated based on body weight and gender using unit-specific equations (26) and then adjusted by the research dietitian daily to ensure a bodyweight within 1% of the weight on the day of admission. Dual-energy x-ray absorptiometry (DPX-1 and DPX-L; Lunar Radiation, Madison, WI) was used to assess body composition (percentage of body fat, PFAT) and, hence, to calculate body fat mass (FM) and fat free mass (FFM). Height-adjusted FM and FFM indexes (FMI and FFMI) were calculated by dividing the respective masses by height squared. To account for the usage of different DXA machines over time, previously validated regression equations were used to make DXA data comparable across different DXA machines (27). Within 48 hours of admission, the Three-Factor Eating Questionnaire (TFEQ) was administered to the volunteers in the morning approximately one hour after breakfast. The TFEQ is a 51-item questionnaire that classifies eating behavior based on three factors (cognitive dietary restraint, dietary inhibition and susceptibility to hunger) as previously described (28).

A 75-g oral glucose tolerance test was performed after three days on the weight-maintaining diet to assess glucose tolerance according to the American Diabetes Association criteria (29) and subjects with diabetes were excluded. Plasma insulin and glucose concentrations were obtained by an automated immunoenzymometric assay (Tosoh Bioscience Inc., Tessenderlo, Belgium) and the glucose oxidase method (Glucose analyzer GM9, Analox Instruments; Lunenbertg, MA), respectively.

On the 5^th day, subjects spent 24 hours inside a whole-room indirect calorimeter (metabolic chamber) to measure energy expenditure and substrate oxidation in eucaloric conditions. During these 24 hours, urine was collected and immediately frozen at −70 degrees for later measurements of urinary dopamine concentration. This measurement was performed by Mayo Clinic Laboratories (Rochester, MN, USA) using High-Performance Liquid Chromatography (HPLC) which has been shown to be a reliable measurement technique (30) chromatography. Urinary dopamine excretion rate over 24 hours was obtained by multiplying dopamine concentration (mcg/L) by total urinary volume (L) and extrapolating the actual urine collection time (range: lower quartile 23.0 to upper quartile 24.3 hours) to 24 hours. Throughout the manuscript, the terms “urinary dopamine concentrations” refer to 24-h urinary excretion rate. According to the Mayo Clinic laboratory reference ranges for catecholamines, urinary dopamine concentrations of 42 subjects were above the normal range (65–400 μg/24 hours). Thus, a sensitivity analysis was performed by excluding these subjects with above-normal dopamine values, which led to similar results (data not shown).

Ad libitum food intake measures

Upon exiting the chamber on the 6^th day, a highly reproducible, validated and computerized vending machine paradigm (31) was used for the assessment of ad libitum food intake over 3 days. On the day of admission, the food preferences of each subject had been assessed via a food selection questionnaire in which a 9-point Likert scale (1=dislike extremely, 5=neutral, 9=like extremely) was given to rate each food item. The 40 different food items who were rated by the volunteer as intermediate (scores: 4–8) were stocked in an individual-specific, computer-operated vending machines during the 3-day period when the volunteers had free access to food for 23.5 hours.

Prior to placing the food items in the vending machines, all food was weighed and any uneaten food (leftovers) at the end of the vending day was also weighed by the metabolic kitchen to precisely estimate actual intake. The daily total and individual macronutrient kilocalories consumed was calculated using the CBORD Professional Diet Analyzer Program (CBORD, Inc., Ithaca, NY, USA) and the Food Processor database (ESHA version 10.0.0, ESHA Research, Salem, OR, USA).

The average total ad libitum food intake over 3 days was calculated and expressed as total kcal eaten daily. Similar calculations were performed for each macronutrient intake (carbohydrate, fat, and protein). The total energy intake was also expressed as percentage of the WMEN determined prior to the 3-day vending period.

Macronutrients categories

Each food item on the vending machine questionnaire was categorized in six groups of food based on the macronutrient content as a percentage of the total energy intake. Food categories were identified as low in fat (<20% kcal) or high in fat (≥45% kcal). Also, food groups were categorized as high in complex carbohydrates (≥30% kcal), high in protein (≥13% kcal) and high in simple sugars (≥30% kcal). Based on this classification, six different groups were identified: high-fat/high-complex carbohydrate (HF/HCC), high-fat/high-protein (HF/HP), high-fat/high-simple sugar (HF/HSS), low-fat/high-complex carbohydrate (LF/HCC), low-fat/high-simple sugar (LF/HSS), low-fat/high-protein (LF/HP) (32).

Statistical analysis

Statistical analyses were performed using SAS software (SAS 9.3, Enterprise guide version 5.1; SAS Institute, Cary, NC). Data are expressed as mean±SD or mean with 95% confidence interval (CI).

Student’s t-test and ANOVA were used to assess differences according to gender and ethnicity, respectively. Pairwise differences between ethnicities were assessed by ANOVA post-hoc tests using the Tukey-Kramer adjustment. Associations between normally distributed data were assessed by the Pearson’s correlation coefficient. Spearman’s correlation was used to assess the correlation between dopamine concentrations and storage time and TFEQ scores. To assess determinants of urinary dopamine concentrations, we performed multivariate regression analysis including age, sex, ethnicity, and body fat as predictors. Similar analyses adjusting for confounders were performed to assess the associations between urinary dopamine concentrations and ad libitum food intake and macronutrient intake. Specifically, linear regression analysis was first used to calculate adjusted values (i.e., residuals) of total food intake and macronutrient intakes after adjustment for age, sex, FFMI and FMI. Linear regression analysis were then used to evaluate the relationships between 24-h dopamine concentrations and residual values of food intake measures (Figure 5).

Figure 5. Relationships between urinary dopamine concentrations and adjusted measures of ad libitum food intake in each ethnic group.

Panels A-C-E-G. Relationships urinary dopamine excretion rate (mcg/24h) and adjusted total food intake (panel A), carbohydrate (panel C), fat (panel E), and protein (panel G) intake in Native Americans and Hispanics.

Panels B-D-F-H. Relationships between urinary dopamine excretion rate (mcg/24h) and adjusted total food intake (panel B), carbohydrates (panel D), fat (panel F), and protein panel (H) intake in Blacks and Whites.

In each panel, the Pearson’s correlation coefficient (r) is reported along with its significance (p).

Adjusted values of total food and macronutrients intake (i.e., residual values) were calculated via linear regression analysis including age, sex, FMI and FFMI as covariates.

Results

Demographic and anthropometric characteristics are reported in Table 1. On average, volunteers were young (36.8 ± 10.9 yrs) with obesity (33.2±9.3 kg/m²) and we observed the expected differences in body composition between males and females. No differences in urinary dopamine concentration and ad libitum food intake measurements were observed between the two study protocols (all p>0.2). Urinary dopamine concentrations were not affected by storage time (p=0.3).

Table 1.

Demographic and anthropometric characteristics of the study cohort.

	Whole study group (n=158)	Men (n=92)	Women (n=66)
Ethnicity	18 BLK, 50 WHT, 18 HIS, 72 NAM	9 BLK, 33 WHT, 12 HIS, 38 NAM	9 BLK, 17 WHT, 6 HIS, 34 NAM
Age (years)	36.8 ± 10.9	38.7 ± 10.8	34.2 ± 10.5*
Body weight (kg)	93.7 ± 23.2	94.9 ± 22.7	91.9 ± 23.9
Height (cm)	169.1 ± 9.7	174.9 ± 7.5	161.1 ± 5.8*
BMI (kg/m2)	33.2 ± 9.3	31.7 ± 9.7	35.3 ± 8.3
Fat free mass (kg)	61.5 ± 12.9	67.1 ± 11	53.7 ± 11.4*
Fat mass (kg)	32.2 ± 14.6	27.9 ± 13.7	38.2 ± 13.9*
Body fat (%)	33 ± 9	27.8 ± 7.2	40.3 ± 5.5*
Fasting glucose (mg/dL)	95.1 ± 8.1	95.2 ± 8.2	94.9 ± 8
2-h OGTT glucose (mg/dL)	130.3 ± 27.1	127.9 ± 27.9	133.5 ± 25.8
Glucose AUC (mg/dL×180min)	24055.4 ± 3702.6	23944.9 ± 3716	24209.1 ± 3707.7
Fasting insulin (μIU/mL)	16.6 ± 17.9	16.6 ± 21	16.7 ± 12.2
2-h OGTT insulin (μIU/mL)	117.5 ± 125.1	115.4 ± 140.1	120.4 ± 101.1
Insulin AUC (μIU/mL ×180min)	20020.8 ± 16238.9	19051.8 ± 17181.9	21377.5 ± 14866.6
Urinary dopamine (mcg/24h)1	345.5 ± 106.4	363.7 ± 114.7	320.2 ± 88.3*
TFEQ cognitive restrain (score)	7.3 ± 4.2	6.8 ± 4.2	7.9 ± 4.2
TFEQ dietary disinhibition (score)	4.9 ± 3.3	4.1 ± 2.6	5.9 ± 3.9 *
TFEQ hunger cues (score)	4.4 ± 3.2	4.0 ± 3.5	4.9 ± 3.5

Data are presented as mean ± SD.

Abbreviations: AUC: area under the curve, BLK: Black, HIS: Hispanic, OGTT: oral glucose tolerance test, NAM: Native American, TFEQ: three-factor eating questionnaire, WHT: White.

^*: p<0.05 between males and females as determined by Student’s t-test.

¹: Urinary dopamine excretion rate was measured over 24 hours period while in the metabolic chamber during eucaloric conditions.

A higher urinary dopamine concentration was associated with lower cognitive restraint assessed by the TFEQ (Spearman ρ=−0.25, p=0.005, Supplementary Fig. 1), whereas no associations were found with dietary disinhibition (p=0.06) and hunger scores (p=0.22). This inverse relationship between restraint and intake was still observed after adjustment for dopamine concentration as partial variable (partial ρ= −0.23, p=0.008) but was only observed in Whites (ρ= −0.45, p=0.0042) but not in any other ethnicities (all p>0.08).

Determinants of urinary dopamine concentrations

Urinary dopamine excretion rate assessed during eucaloric conditions was on average 346±106 mcg/24h and differed by sex (p=0.008, Figure 2A) and ethnicity (p<0.0001, Figure 2B), such that dopamine concentrations were on average lower in women compared to men by ~13% (Δ=−44 mcg/24h, CI: −77 to −10) and in Whites compared to Blacks (Δ=−74 mcg/24h, CI: −145 to −3), Hispanics (Δ=−99 mcg/24h, CI: −170 to −28) and Native Americans (Δ=−86 mcg/24h, CI: −134 to −39).

Figure 2. Urinary dopamine concentrations by sex and ethnicity.

Average differences (Δ) in urinary dopamine concentrations between males and females (panel A) by Student’s t-test and across ethnicities (panel B) by ANOVA with Tukey-Kramer adjustment of the least square means for multiple comparisons. Error bars represent mean ± 95% CI.

Abbreviations: BLK, Black, WHT, Caucasian; HIS, Hispanic; NAM, Native American.

Dopamine concentration was positively correlated with FMI (r=0.24, p=0.003, Figure 3A), FFMI (r=0.40, p<0.0001, Figure 3B), BMI (r=0.28, p<0.0001, Figure 3C), while there was a negative correlation with age (r=−0.20, p=0.01, Figure 3D). In a multivariable model, age (p=0.006), ethnicity (p=0.007), sex (p=0.03), and body fat (p= 0.0017), were all independent determinants of urinary dopamine concentrations (total R2=0.28) (Supplementary Table S1).

Figure 3. Relationships between urinary dopamine concentrations and body composition and age.

Relationships between urinary dopamine excretion rate (mcg/24h) and fat mass index (panel A), fat free mass index (panel B), body mass index (panel C), and age (panel D).

In each panel, the Pearson’s correlation coefficient (r) is reported along with its significance (p).

After adjustment for age, sex, ethnicity, and body fat, dopamine concentrations were not associated with fasting glucose, glucose AUC, 2-h glucose, fasting insulin concentrations, insulin AUC, 2-h insulin during the OGTT and insulinogenic index (all p>0.09).

Relationship between urinary dopamine concentrations and ad libitum food intake measures

Energy intake measurements on the vending machines are reported in Table 2. The average, daily, total ad libitum energy intake was 3307±1041 kcal/day (or 120±36% when expressed as percentage of WMEN) and was higher in males compared to females by nearly 28% (Δ=1074.5 kcal/day, 95% CI: 788.4 to 1360.6, p<0.0001) with no difference according to ethnicity (p=0.7). Dopamine concentration was positively associated with total energy intake (r=0.29, p=0.0002, Figure 4A), reflecting greater carbohydrate (r=0.27, p=0.0005, Figure 4B), fat (r=0.26, p=0.0008, Figure 4C) and protein (r=0.27, p=0.0007, Figure 4D) intake. However, after adjustment for known determinants of energy intake (age, sex, ethnicity, FMI and FFMI), these associations were highly attenuated (all p>0.1). When testing for differences according to ethnicity given that Whites had on average lower urinary dopamine concentrations compared to other ethnic groups, we found that the relationship between dopamine concentration and total energy intake (adjusted for age, sex, FMI and FFMI) was very different across the four ethnic groups (dopamine×ethnicity interaction term p=0.018). Specifically, in Blacks and Whites dopamine concentrations were positively associated with adjusted total food intake (β=635.0 kcal/day, p=0.001 and β= 325.2 kcal/day per 100 mcg/24h difference in dopamine, p=0.02, respectively, Figure 5B), carbohydrate (β=263.1 kcal/day, p=0.03 and β=167.8 kcal/day per 100 mcg/24h, p=0.02, respectively, Figure 5D), fat (β=248.7 kcal/day, p=0.009 and β=160.9 kcal/day per 100 mcg/24h, p=0.03, respectively, Figure 5F) and protein intake (β=116.6 kcal/day per 100 mcg/24h, p=0.003 in Blacks). Conversely, no associations were found in Hispanics and Native Americans (all p>0.3, Figure 5A–5C–5E). No association was found when dopamine concentrations were correlated with residuals of protein intake neither in Native American and Hispanic (all p>0.3, Figure 5G) nor in Whites (p>0.9, Figure 5H). Similar results were obtained after adjustment for individual energy requirement (data not shown).

Table 2.

Measures of ad libitum food intake by the computerized vending machine system.

	Whole study group (n=158)	Men (n=92)	Women (n=66)
Total energy intake (kcal/day)1	3306.8 ± 1041.1	3755.7 ± 950	2681.2 ± 819.7*
WMEN (kcal/day)2	2772.1 ± 257.3	2867.2 ± 228.2	2642.6 ± 239.3*
Total energy intake (% WMEN)	120.2 ± 36.5	136 ± 34.8	99.0 ± 26.6*
Carbohydrate intake (kcal/day)	1710.9 ± 535.7	1927.6 ± 511.3	352.2 ± 102.6*
Fat intake (kcal/day)	1212.0 ± 474.5	1386.5 ± 462.2	968.8 ± 376.5*
Protein intake (kcal/day)	416.8 ± 142.4	478.9 ± 131.0	330.4 ± 109.1*

Data are presented as mean±SD.

^*: p<0.05 between males and females as determined by Student’s t-test.

¹Ad libitum food intake measures are reported as the average of three days.

²WMEN: weight maintaining energy needs.

Figure 4. Relationships between urinary dopamine concentrations and ad libitum food intake measures.

Relationships between urinary dopamine excretion rate (mcg/24h) and total ad libitum food intake (panel A), carbohydrate intake (panel B), fat intake (panel C), and protein intake (panel D). The total ad libitum food intake and each macronutrient intake during the 3-day vending period are expressed as the average over 3 days.

In each panel, the Pearson’s correlation coefficient (r) is reported along with its significance (p).

Furthermore, urinary dopamine concentrations were not associated with daily energy intake from any food group nor in the whole cohort neither in any ethnicity separately (all p >0.07) after adjustment for age, sex, FFMI and FMI.

Discussion

In the current analysis, we tested whether dopamine concentration is a determinant of ad libitum food intake and eating behavior constructs in a large cohort of 158 healthy subjects without diabetes in carefully controlled settings where food intake was directly and objectively assessed by a validated, computerized vending machine paradigm. Our results show that higher urinary dopamine excretion rate assessed during 24 hours of eucaloric diet and energy balance was associated with lower cognitive dietary restraint and greater ad libitum energy intake. However, after adjustment for body size and composition, these dopamine-intake associations were only observed in White and Black cohorts but not in Native Americans or Hispanics.

It is well known that several limbic and cortical brain regions play a major role in the control of food intake both in rats and humans (33). Several neurotransmitters are implicated in the regulation of food intake (6) and especially dopamine seems to be an essential messenger for feeding regulation in rodents (7). The lack of dopamine production in dopamine deficient mice, created by inactivating the tyrosine hydroxylase, leads to an inability to initiate feeding although this ability is restored after injection of L-DOPA into the striatum (7, 13). In mice, extensive depletion of dopamine results in food intake reduction (10). Reversely, increased release of dopamine concentrations via mesulargine in the medial hypothalamic nuclei is associated with stimulation of food intake in rats (34). It has been hypothesized that there is an “afferent-efferent neurotransmitter unit” (35) where the afferent information (such as metabolic status) is conveyed to hypothalamus and stored in the presynaptic afferent neurons which, in turn, stimulate the postsynaptic cells to release dopamine. Dopamine concentrations might stimulates food intake via neuropeptides such as NPY, orexin, and melanin-concentrating hormone (35). In line with these findings, we observed that higher dopamine concentrations were associated with lower cognitive dietary restraint and greater daily energy intake when food was available with no constraints. However, dopamine did not fully mediate the association between restraint and intake, suggesting that other factors from the periphery such as signals from adipose tissue, skeletal muscle or gut might explain this association.

It has been shown that dopamine has an essential influence on motivation and reward circuits by modulating food reward via meso-limbic circuitry of the brain trough D2R (19). Mesolimbic dopamine neurons are an important link in the neural circuitry of reward (16). In both mice and humans, it has been shown that subjects with obesity might have altered D2R expression in the hypothalamic area that are activated by food related cues (36). Rats fed with highly palatable and energy dense food have reduced D2R expression in the striatum (37). In individuals with obesity, striatal D2R availability was lower than the lean control group, indicating that dopamine deficiency may lead to pathological eating to compensate the decreased activation of the motivation and reward circuits (19). In subjects with obesity, dopamine neurocircuitry alterations raises the susceptibility for opportunistic overfeeding making food intake less rewarding and more habitual (20). Taken together, our current results support and confirm the role of dopamine in the regulation of food intake in humans, although the positive association between urinary dopamine concentration and ad libitum food intake was found only in races (White and Black) other than Hispanics and Native Americans after adjustment for the known determinants of energy intake (mainly, FFM (38)).

Native Americans, the largest ethnic group in our cohort, had higher urinary dopamine concentrations compared to White cohorts, but these were not associated with energy or macronutrient intake. This finding might suggest decreased dopamine responsiveness in Native Americans with obesity and uncoupling of the dopaminergic reward system with food intake in this ethnic group. The underpinning of this decrease responsiveness to dopamine is not clear but reduced activity of the D2R in the brain is one possibility. The Ser311Cys mutation (rs1801028) of the human D₂-dopamine receptor gene (DRD2) is associated with higher BMI and reduced energy expenditure in this Native American population (39). In this latter cohort, the DRD2 mutation is more common (15%) (39) whereas it is much less common in other ethnicities (3%, 1000 Genome Project). We might speculate that, in overweight Native Americans, urinary dopamine might not play a substantial role in food intake regulation partly because of the “resistance” of D2R to dopamine (due to genetics and/or greater adiposity), which might lead to higher-than-normal dopamine concentrations in Native Americans but with no effect on food intake. While the observed ethnic differences for the effect of dopamine on energy intake may reflect its central effect, differences in peripheral metabolism of dopamine and/or glomerular filtration by ethnicity might further explain the observed differences between Whites and Native Americans. Further, dopamine plays an important role in physiological regulation (production/excretion) outside the central nervous system (40). It has been shown in vivo that the kidney deconjugates dopamine compounds, leading to free dopamine excretion (41). In rats, the action of dopa decarboxylase on L-dopa in the renal tubular cells leads to the dopamine production which, in turn, is secreted into the kidney tubular lumen (42).

We also evaluated the anthropometric determinants of urinary dopamine concentration. Many studies in literature have investigated the relationship between basal striatal dopamine functioning and BMI. It has been shown that subjects with obesity had lower D₂-D₃ binding potentials (19) and a lower striatal dopamine tone (43) compared to normal weight controls. However, to our knowledge, no studies have shown the association between urinary dopamine concentrations and body composition. We found that that urinary dopamine concentrations were positively associated with BMI, FMI, FFMI, and body fat, indicating higher dopamine excretion with increased adiposity. The reason why urinary dopamine concentration was lower in Whites is unclear but might be due to differences in adipose tissue mass and functions (e.g., expression and function of dopamine receptors in adipocytes) compared to other races/ethnicities.

Our study has several limitations. We did not have any measurement of D2R availability with positron emission tomography (PET) (19, 20), which might have been helpful to understand the inter-individual difference in urinary dopamine concentrations, especially between ethnicities. Also, we measured 24-h dopamine excretion rate in urine, which might not be an exact measurement of dopamine concentration in the brain, although it has been shown that dopamine crosses the blood brain barrier and urinary dopamine concentrations are a good indicator of the peripheral and nervous system neurotransmitter functional status (44, 45).

Conclusion

In conclusion, in a large cohort including 158 healthy subjects from four different ethnicities, higher dopamine concentrations were associated with lower cognitive restraint of eating and greater ad libitum energy intake, although the associations between dopamine and food intake were observed only in Whites and Blacks after controlling for body size and composition. Our results suggest that, in humans, dopamine has a role in the regulation of the food reward processing by modulating impulsivity related to eating behavior that increases the drive to eat.

What is already known about this subject?

• Dopamine represents an essential messenger for feeding regulation in rodent models.

• Dopamine has an essential influence on motivation and reward circuits by modulating food reward.

• Lack of results on the role of physiologic dopamine concentrations on eating behavior and energy intake in humans.

What are the new findings in your manuscript?

• Higher dopamine concentration associates with lower cognitive dietary restraint.

• Higher dopamine concentration determines greater ad libitum energy intake.

• However, the dopamine-intake relationship highly depends upon ethnicity.

How might your results change the direction of research or the focus of clinical practice?

• Therapies acting on dopamine to decrease intake may be beneficial for the treatment of obesity.

Abbreviations

DXA

dual energy X-ray absorptiometry

FFM

fat free mass

FFMI

fat free mass index

FM

fat mass

FMI

fat mass index

OGTT

oral glucose tolerance test

PFAT

percent of body fat

TFEQ

three-factor eating questionnaire