Central Nervous System and Peripheral Hormone Responses to a Meal in Children

Abstract

Context: Behavioral studies suggest that responses to food consumption are altered in children with obesity (OB).

Objective: To test central nervous system and peripheral hormone response by functional MRI and satiety-regulating hormone levels before and after a meal.

Design and setting: Cross-sectional study comparing children with OB and children of healthy weight (HW) recruited from across the Puget Sound region of Washington.

Participants: Children (9 to 11 years old; OB, n = 54; HW, n = 22), matched for age and sex.

Intervention and outcome measures: Neural activation to images of high- and low-calorie food and objects was evaluated across a set of a priori appetite-processing regions that included the ventral and dorsal striatum, amygdala, substantia nigra/ventral tegmental area, insula, and medial orbitofrontal cortex. Premeal and postmeal hormones (insulin, peptide YY, glucagon-like peptide-1, active ghrelin) were measured.

Results: In response to a meal, average brain activation by high-calorie food cues vs objects in a priori regions was reduced after meals in children of HW (Z = -3.5, P < 0.0001), but not in children with OB (z = 0.28, P = 0.78) despite appropriate meal responses by gut hormones. Although premeal average brain activation by high-calorie food cues was lower in children with OB vs children of HW, postmeal activation was higher in children with OB (Z = -2.1, P = 0.04 and Z = 2.3, P = 0.02, respectively). An attenuated central response to a meal was associated with greater degree of insulin resistance.

Conclusions: Our data suggest that children with OB exhibit an attenuated central, as opposed to gut hormone, response to a meal, which may predispose them to overconsumption of food or difficulty with weight loss.

Trial registration: ClinicalTrials.gov NCT02484976.

Copyright © 2019 Endocrine Society.

Figures

Figure 1.

Timing of study procedures and appetite and appeal ratings in children of HW and children with OB. (A) Participants arrived fasting and underwent four blood draws [fasting (t−30), prior to the test meal (t210), then 30 (t240) and 60 (t270) min after the meal]. The first MRI scan session occurred 3 h after the standardized breakfast and was immediately followed by a standardized test meal, consumed within 10 min. The second MRI scan session began 30 min after the start of the test meal. The breakfast and test meal were titrated to represent 10% and 33%, respectively, of each participant’s estimated daily caloric intake whereas the buffet meal was ad libitum. Bioelectrical impedance analysis (BIA) was used to measure body composition. Appetite ratings were performed serially using visual analog scales (VAS) for hunger and fullness, whereas appeal ratings were obtained once, immediately prior to the buffet meal. (B and C) Appetite ratings of hunger (B) and fullness (C) were not different between groups [hunger, χ2(1) = 0.06, P = 0.80; fullness, χ2(1) = 0.53, P = 0.47]. All meals (filled arrows mark standardized meals; open arrows indicate ad libitum buffet meal) significantly suppressed hunger and increased fullness. (D) Participants rated high-calorie images as more appealing [χ2(1) = 9.99, P < 0.01]. Data are means ± SEM. P values for appetite and appeal ratings were determined by linear mixed models (unadjusted). Gray bars indicate fMRI sessions. n = 22 children of a HW, 54 children with OB (n = 2 missing for appeal ratings in the OB group). *P < 0.01, **P < 0.0001, premeal vs postmeal appetite ratings for both groups; ‡P < 0.01, vs low-calorie images.

Figure 2.

Effect of a meal on average activation across a priori appetite-processing regions. (A–C) Mean parameter estimates across a priori regions in children of HW and children with OB were measured before (filled/open bars) and after (striped bars) a standardized test meal for the contrasts of (A) high-calorie foods vs objects, (B) low-calorie foods vs objects, and (C) high-calorie vs low-calorie foods. Data are means ± SEM. P values were determined by linear mixed models. *P < 0.05, **P < 0.0001, vs premeal within same group; †P < 0.05, HW group vs OB group.

Figure 3.

Effect of a meal on regional activation. (A–F) Mean parameter estimates in regions considered separately in children of HW and children with OB were measured before (filled/open bars) and after (striped bars) a standardized test meal for high-calorie foods vs objects. (A) mOFC, (B) SN/VTA, (C) DS, (D) VS, (E) amygdala, (F) insula. Data are means ± SEM. P values were determined by linear mixed models. *P < 0.05 vs **P < 0.01, vs premeal within same group; †P < 0.05, ††P < 0.01, HW group vs OB group.

Figure 4.

Relationship between meal-induced change of average activation across satiety-processing regions and HOMA-IR and meal-induced changes of circulating ghrelin and insulin. (A) Across all participants, a higher HOMA-IR score was associated with less reduction in brain activation by high-calorie food cues. (B) Considering groups separately, there was a significant interaction between group and meal-induced change in circulating plasma ghrelin on the change in brain activation by a meal. (C) No interaction was present for the meal-induced change in plasma insulin and brain activation by high-calorie food cues. P values were determined by simple linear regression (A) and linear regression with an interaction term (B and C). A Pearson correlation coefficient was calculated for descriptive purposes.

Figure 5.

Exploratory voxelwise analyses outside of a priori regions. The top panel shows clusters of greater activation in children of HW vs children with OB prior to the standardized meal. The bottom panel shows clusters with greater activation in children of HW vs children with OB after consumption of a standardized meal. Analyses excluded previously tested a priori ROIs, and all clusters depict activation by high-calorie foods vs objects. Z statistic maps were corrected for multiple comparisons and were thresholded at Z > 1.65 and a cluster significance threshold of P = 0.05 (FWE corrected). Premeal, n = 22 children of HW, n = 51 children with OB; postmeal, n = 21 children of HW, N = 46 children with OB. Color scales provide Z values of functional activation. Montreal Neurologic Institute coordinates are indicated. A priori ROIs are depicted in bright gray. Peak Montreal Neurologic Institute coordinates are provided in Table 4. AG, angular gyrus; FP, frontal pole; lOcC, lateral occipital cortex; mTG, middle temporal gyrus; OcP, occipital pole; POpC, parietal operculum cortex; ScC, subcallosal cortex; sTG, superior temporal gyrus; TFsC, temporal fusiform cortex.