Altered hippocampal effective connectivity predicts BMI and food approach behavior in children with obesity

doi:10.21203/rs.3.rs-4301324/v1

Download PDF

Article

Altered hippocampal effective connectivity predicts BMI and food approach behavior in children with obesity

https://doi.org/10.21203/rs.3.rs-4301324/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective

The vicious circle model of obesity proposes that the hippocampus plays a crucial role in food reward processing and obesity. However, few studies focused on whether and how pediatric obesity influences the potential direction of information exchange between hippocampus and key regions, as well as whether these alterations in neural interaction could predict future BMI and eating behaviors.

Methods

In this longitudinal study, a total of 39 children with overweight/obesity and 51 children with normal weight, aged 8 to 12, underwent resting-state fMRI. One year later, we conducted follow-up assessments of eating behaviors and BMI. Resting-state functional connectivity (FC) and spectral dynamic casual modeling technique were used to examine altered functional and effective connectivity (EC) of the hippocampus in children with overweight/obesity. Linear support vector regression, a machine learning method, was employed to investigate whether hippocampal connections at baseline could predict future BMI and eating behaviors.

Results

Compared to controls, children with overweight/obesity displayed abnormal bidirectional inhibitory effects between the right hippocampus and left postcentral gyrus (PoCG), namely, stronger inhibitory EC from the hippocampus to PoCG but weaker inhibitory EC from the PoCG to hippocampus, which further predicted BMI and food approach behavior one year later.

Conclusion

These findings suggest that imbalanced information exchange in the appetitive reward circuitry between the hippocampus to somatosensory cortex may be a sensitive neurobiomarker for childhood obesity and future food approach behavior, which expands the vicious circle model of obesity by revealing the crucial role of hippocampal undirectional and directional connections in childhood obesity. This study is essential for developing effective intervention strategies and for reducing long-term health-care costs associated with obesity.

Biological sciences/Neuroscience

Health sciences/Endocrinology/Endocrine system and metabolic diseases/Obesity

Childhood obesity

Hippocampus

Functional connectivity

Effective connectivity

Spectral dynamic casual modeling

Overweight and obesity have high morbidity and mortality throughout the life course, which is closely linked to diabetes, cardiovascular diseases, and psychological disorders. It is becoming the greatest threat to public health in the 21st century (Lowe et al., 2020). As of 2022, more than one billion people globally are suffering from obesity, with children and adolescents accounting for 159 million (Phelps et al., 2024). In China, the prevalence of childhood overweight is as high as 11.1% and the prevalence of childhood obesity is 7.9% (Pan et al., 2021). In obesogenic environments, food availability and calorie density are the main determinants of excess weight in school-age children(J. Zhang et al., 2024). Treatments for obesity have been sub-optimal, which is partly because understanding of its neurobiological correlates remains to be limited (Ochner et al., 2015; Wang et al., 2023). More importantly, childhood obesity has long been recognized as an increasingly essential predictor of adult obesity (Lister et al., 2023). Therefore, it is of great importance to examine the underlying neural basis of pediatric obesity and its relationship with future weight and eating behaviors during such a critical developmental period.

Hippocampus, a subcortical reward region, is intimately involved in regulating human food intake and obesity (Stevenson & Francis, 2017; Webber et al., 2009). Obesity and high-caloric diets have also been found to increase hippocampal neuroinflammation (Schmitt & Gaspar, 2023). The vicious circle model of obesity has proposed that Western diet (e.g., saturated fat and carbohydrate food) contributes to obesity, in part, by damaging the appetite regulation function of the hippocampus which ultimately induces further over-consumption of unhealthy food, weight gain, and obesity (Kanoski & Davidson, 2011). The appetite processing needs to integrate external food sensory information and internal energy-relevant signals from somatosensory cortex (e.g., postcentral gyrus [PoCG], frontal operculum, and insular) to evaluate when, what and how much to eat (Clasen et al., 2020; Eichenbaum, 2000; Kanoski & Grill, 2017; Quigley et al., 2021). A review study indicated that hippocampus and somatosensory cortex were activated in response to visual, olfactory and taste food stimuli (Huerta et al., 2014).

An emerging body of research has demonstrated that structural impairments and functional alterations in the hippocampus may be a neural vulnerability factor for obesity and future food approach behaviors (e.g., food responsiveness and overeating). Cross-sectional and two-year follow-up studies have consistently found that children with obesity exhibited hippocampal atrophy (i.e., smaller gray matter volume) (Bauer et al., 2015; Jiang et al., 2023; Mestre et al., 2017). Moreover, disrupted hippocampus activation in response to food stimuli has been shown to underlie childhood obesity (Bruce et al., 2010; Davids et al., 2010; Mestre et al., 2017; Samara et al., 2018). Such abnormal reactivity to food in the hippocampus was related to its changed resting-state activity (G. Li et al., 2021). Regarding resting-state functional synchrony, individuals with obesity displayed disrupted functional connectivity (FC) in the hippocampus and somatosensory cortex (PoCG and frontal operculum), which was linked with 12-month changes of eating behaviors and weight after bariatric surgery (Cerit et al., 2019; Geha et al., 2017; Lee et al., 2022). Notably, birth weight of children with overweight/obesity was associated with the FC between hippocampus and PoCG (Solis-Urra et al., 2023). However, existing resting-state fMRI studies have largely focused on the abnormal neurobiology of obesity in adults, rather than in children (for a review, see Syan et al., 2021). As such, the present study sought to determine whether and how the hippocampus-based FC was associated with childhood obesity.

Furthermore, abnormal effective connectivity (EC) between brain regions may act as the basis for functional synchrony, which could determine the directional causal interaction among different brain regions (Friston et al., 2014; Razi et al., 2015). Several methods have been proposed to examine the potential EC pattern, such as Granger causal analyses (GCA) and dynamic causal modeling (DCM) (Friston et al., 2003; Granger, 1969). Although previous studies using the GCA technology found aberrant strength of hippocampus-related directional connectivity in adults with obesity (Duan et al., 2020; Z. Zhang et al., 2022), this method cannot determine the valence of connections between regions (i.e., excitatory or inhibitory effect). Spectral DCM (spDCM), a DCM technique for resting-state data, exhibits higher computational efficiency and sensitivity when comparing group-between differences (Razi et al., 2015). This approach has been widely applied to reveal directional connections contribute to emotion intelligence, disordered eating, and bipolar disorder (Bajaj & Killgore, 2021; Chen et al., 2023; W. Li et al., 2024; Z. Zhang et al., 2022). However, the potential directional neural links involved in obesity are largely unknown. Only one study used the spDCM to examine adults’ obesity-related changes in the direction and valence of connections between key areas, and revealed that excess weight was associated with persistent top-down excitation of the hypothalamus, regardless of homeostatic state (Voigt et al., 2021).

Despite the hippocampus has been found to play a crucial role in food consumption and obesity, few studies focused on the neurofunctional alteration in the hippocampus among children with overweight/obesity. More importantly, it remains unknown whether and how pediatric obesity influences the direction of information exchange between key brain regions. Therefore, the current study represents an important first step in combining the FC and EC measurements to investigate the neural basis of pediatric obesity, with the hippocampus as region of interest (ROI). Based on previous findings of lesser FC of hippocampus with somatosensory cortex in obesity adults (Sun et al., 2018), we used the seed-based FC analysis to test the first hypothesis that children with overweight/obesity might show lower functional synchrony between hippocampus and somatosensory cortex. In addition, the spDCM technique was further employed to reveal the abnormal mechanism of information flow in relation to hippocampus (i.e., EC). Due to limited evidence, we hypothesized that the altered directional connections regarding the hippocampus might contribute to pediatric overweight/obesity. Furthermore, altered activity in appetitive brain regions have been shown to be a predictor for future eating behaviors and weight gain (Boswell & Kober, 2016). Regarding children with obesity, the neural vulnerability factors for future eating behaviors and body mass index (BMI) have been understudied. According to the vicious circle model of obesity, we assumed that abnormal hippocampal connections in children with obesity might be a prominent predictor for future eating behaviors and BMI. The current study will expand and deepen the understanding of neurofunctional substrates of pediatric obesity, which is essential for developing effective intervention strategies and for reducing long-term health-care costs associated with obesity (Lowe et al., 2020).

2.1 Participants

All participants were recruited from two primary school in Chongqing (southwest of China) via posters and flyers. A total of 130 children (8–12 years old) participated in this study, each participant completed a series of self-reported measures and behavioral tasks, and underwent neuroimaging scanning in the first year. One year later, we conducted follow-up assessments of their behavioral measures and BMI. Participants met the following inclusion criteria, which were assessed via parental self-report: right-handed; no history of psychiatric or neurological illnesses; no use of psychoactive medications (binary variables indicating the presence or absence of particular disorders); and Chinese as the primary language. After quality control of structural neuroimaging data, a total of 21 participants were excluded due to excessive head motion, defined as a mean framewise displacement (FD) larger than 0.5 mm. Based on the BMI recorded in the first year, a total of 19 underweight participants were further excluded according to WHO growth reference (Onis et al., 2007). The final sample comprised 51 children with normal weight (i.e., NW group) and 39 children with overweight/obesity (i.e., OW group) (for details, see Table 1 and section 2.2.1). All participants and their parents signed an informed consent document prior to enrollment, and each subject received a set of stationery as a reward at the end of the study. The research protocol was reviewed and approved by the Ethics Committee of the Southwest University (IRB No. H22003). All procedures in this study complied the 6th revision of the Declaratifon of Helsinki.

2.2 Food choice task

The food choice task was utilized to assess the subjective values of food (Hare et al., 2009). A total of 36 different food pictures were included in this task, including Chinese and Western traditional high-calories foods (e.g., Peking Ducks, dumplings, hamburgers, and ice creams) selected from the standardized Chinese foods picture database compiled by our team (Kong, 2012). All pictures were matched for size, resolution, brightness, and background. The trials began with a fixation cross (1 s) followed by food picture stimulus presented, with a five-point ratings scale shown on the screen below food picture. The time window to respond lasted 10 s. Food rating consisted of three phases: taste, healthiness rating and decision making. For each food stimulus, participants were first asked “How tasty is it?” and responded with a key press: very untasty (= 1), untasty (= 2), neutral (= 3), tasty (= 4), or very tasty (= 5). Then, they were asked “How healthy is it?” and responded with a key press: very unhealthy (= 1), unhealthy (= 2), neutral (= 3), healthy (= 4), or very healthy (= 5). Finally, they were asked “How much do you want to eat it?” about preference for each food and responded with a key press: strong no (= 1), no (= 2), neutral (= 3), yes (= 4), or strong yes (= 5). The task consisted of 5 practice trials, followed by one formal block of 36 trials. Rating scores of food taste, healthiness and preference (i.e., decision making) have been considered as food subjective reward values (Hare et al., 2009; Kong, 2012). See supplementary Fig. S1 for the task procedure. Participants performed these tasks on laboratory computers using E-Prime 2.0 software, with seated approximately 60 cm from the screen.

2.3 Measures

2.3.1 BMI and WHO weight status

Information regarding participants’ age, sex, height and weight was measured by trained researchers at Southwest University. BMI was calculated by dividing an individual's weight (kg) by the square of their height (m). According to the WHO growth reference, BMI percentiles were measured and adjusted for age and sex for each child. The WHO reference were used to classify children’s BMI percentiles into normal weight (15th percentile to less than the 85th percentile), overweight (85th percentile to less than the 97th percentile), obesity (97th percentile or greater), and thinness (less than the 15th percentile) (Onis et al., 2007; World Health Organization, 2006). Nineteen children with thinness were excluded. Finally, 51 participants were selected in the NW group, and 39 (overweight, 18; obesity, 21) were selected in the OW group. Since BMI percentiles are non-linear, age- and sex-specific BMI z-scores were calculated for statistical analyses (Bhutani et al., 2021; Wang & Chen, 2012).

2.3.2 Children’s Eating Behaviour Questionnaire

The Children’s Eating Behaviour Questionnaire (CEBQ) is a 35-item multidimensional scale that measures children’s eating behaviors about food responsiveness (e.g., “My child is always asking for food”), food enjoyment (e.g., “My child loves food”), satiety responsiveness (e.g., “My child gets full up easily”), slowness in eating (e.g., “My child eats slowly”), food fussiness (e.g., “My child enjoys a wide variety of foods”), emotional overeating (e.g., “My child eats more when anxious”), emotional undereating (e.g., “My child eats less when upset”), and desire for drinks (e.g., “If given the chance, my child would always be having a drink”). This scale was a 5-point Likert scale ranging from 1 (never) to 5 (always), answered by their parents. The CEBQ can be further divided into two subscales: the “food approach” subscale (i.e., food responsiveness, enjoyment of food, and emotional overeating) and “food avoidant” subscale (i.e., satiety responsiveness, slowness in eating, emotional undereating, and food fussiness) (Wardle et al., 2001).

2.3.3 Child Feeding Questionnaire

The Child Feeding Questionnaire (CFQ) measures feeding practice and beliefs about child feeding and obesity proneness (Birch et al., 2001), including perceived responsibility, parents’ concern about child overweight, perceived parent weight, perceived child weight, monitoring, pressure to eat, and restriction. The Chinese version of CFQ used in the current study added an additional subscale of (i.e., food as a reward) in order to more reasonably evaluate feeding practice and beliefs among parents of children in the Chinese context. The final version is a 26-item questionnaire and has good construct validity and satisfactory internal consistency among children (Zheng et al., 2016). All items are scored on a 5-point Likert-type scale, with higher subscale total scores reflecting more disordered feeding practice/beliefs regarding children’s eating and weight concern.

2.3.4 Eating Disorder Inventory-Child

The Eating Disorder Inventory-Child (EDI-C) is a reliable version of the EDI-2 questionnaire that has been formulated for children and adolescents (Eklund et al., 2005). The 7-item bulimia subscale is used to evaluate children’s levels of bulimic eating (e.g., “I stuff myself with food”). Each item was rated on a 6-point scale ranging from 0 = never to 5 = always, with a higher total score indicating more frequent bulimic eating. This scale has been used with Chinese children and was found to have acceptable internal consistency (Chen et al., 2022).

2.4 Neuroimaging data acquisition and preprocessing

For each participant, an 8 min resting-state fMRI and 5 min structural fMRI scans were performed in a 3T Trio scanner (Siemens Medical, Erlangen, Germany). We used a gradient echo planar imaging sequence to obtain resting-state functional images. Scanning parameters were as follows: repetition time (TR) = 2000 ms, echo time (TE) = 30 ms, slices = 33, slice thickness = 3.5 mm, field of view (FOV) = 224 × 224 mm², flip angle = 90^°, resolution matrix = 64 × 64, voxel size = 3.5 × 3.5 × 3.5 mm³, phase encoding direction = PC > > AC. Each section contained 180 volumes. High-resolution T1-weighted structural images were acquired for coregistration purposes. The magnetization prepared rapid acquisition gradient echo sequence employed the following parameters: TR = 2530 ms, TE = 3.48 ms, FOV = 256 × 256 mm², flip angle = 7^°, base resolution = 256 × 256, voxel size = 1 × 1 × 1 mm³, phase encoding direction = AC > > PC. Prior to the brain scanning, each participant underwent a 5-min simulated scanning training to adapt the real scanning environment. During formal scanning, participants were instructed to remain still and relaxed, keep their eyes closed, and avoid thinking of anything deliberately. Foam pads and earplugs were employed to reduce head motion and scanning noise.

The publicly available Data Processing Assistant for Resting-State fMRI (DPARSF) toolbox (http://www.restfmri.net) was used to preprocess neuroimaging data. Preprocessing was conducted as follows: the first 10 time points were removed to minimize the instability of initial scanning. The remaining 170 volumes were slice timing-corrected, realigned. Functional volumes were then spatially normalized to the Montreal Neurological Institute (MNI) 152-brain template with a resolution voxel size of 3 × 3 × 3 mm³ by Diffeomorphic Anatomical Registration through Exponentiated Lie Algebra (DARTEL). All functional images were smoothed using a Gaussian kernel of 6 mm full width at half-maximum and further denoised by regressing out several nuisance signals, including the Friston-24 head motion parameters and signals from white matter and cerebrospinal fluid, before temporal band-pass filtering (0.01–0.08 Hz). After preprocessing, a total of 21 participants were deleted due to the head movements (i.e., mean FD values) greater than 0.50 mm (Jenkinson et al., 2002; Yan et al., 2013).

2.5 Seed-based functional connectivity

The seed regions of bilateral hippocampus were derived from the Automated Anatomical Labelling Atlas 3 (AAL3) (https://www.gin.cnrs.fr/en/tools/aal/) (Rolls et al., 2020). We generated spherical seeds with 5 mm radius in the right and left hemisphere (MNI coordinate: left hippocampus: x = − 24.76, y = − 22.46, z = − 11.88; right hippocampus: x = 29.45, y = − 21.49, z = − 12.08). The mean BOLD signal time series from each ROI was extracted and correlated with the time series of all other voxels within the brain to create whole-brain Pearson’s correlation coefficient maps. Fisher’s r-to-z transformation was then applied in these maps to improve the normality of the correlation coefficients.

2.6 Effective connectivity

Based on the FC results, we further calculated the EC and examined the group differences using the spDCM method. We briefly explain each step as follows. For a detailed description of the spDCM technique, refer to Razi et al., (2015).

ROI time series extraction

Two nodes of FC results were considered as ROI in EC analysis. To extract ROI time series, a general linear model was built for preprocessed (slice timing, realigned, normalized, and smoothed) rs-fMRI data with six head-motion parameters, cerebrospinal fluid, and white matter as nuisance regressors. Low-frequency signal drifts were filtered using a 128-s high-pass filter. ROI times series were extracted by performing a principal components analysis to retain the first principal eigenvariate across a sphere of 8 mm radius centered on the peak coordinates, as in previous studies (Bajaj & Killgore, 2021; Preller et al., 2019).

First level analysis

Specification and estimation of DCM. The spDCM analyses were conducted using DCM12 implemented in the SPM12 (revision 7771, www.fil.ion.ucl.ac.uk/spm). We build a fully-connected DCM for significant FC at subject level, which assumed each node to be connected to all other nodes (22 connectivity parameters, including 2 intrinsic self-connections) and fits the complex cross spectral density using a power-law model of endogenous neuronal fluctuations. Then we estimated each participant's DCM with Bayesian model, finding the posterior density over parameters and offering the best trade-off between explaining the data and minimizing complexity of model. For more details, see Friston et al., (2014).

Second level analysis

Parametric empirical Bayes. Second-level analysis with the Parametric Empirical Bayes (PEB) framework examined the between-group differences in EC. The PEB framework specifies a hierarchical statistical model of connectivity parameters. We built a PEB model to partition the variability in connectivity parameters across subjects into hypothesized group-level effects and uninteresting between-subject variability. Two covariates were specified (overall group mean and between-group difference) in the design matrix. The parameters of the PEB model are estimated using the standard variational Laplace procedures at second-level. Then, Bayesian model averaging was performed on the second-level PEB model to investigate which the direction of connection best described the between-group differences. Under the PEB framework, parameters best describing between-group effects were reported not in terms of p value but instead in terms of posterior probability (PP). PP values > 0.95 were considered “strong evidence” for significant effects (Blain et al., 2023; Preller et al., 2019; Randeniya et al., 2023; Zeidman et al., 2019).

2.7 Longitudinal predictive analysis

The machine learning approach allows for the prediction of unseen participants, offering information at the individual level rather than group level (Cui et al., 2018; Shen et al., 2017; Yarkoni & Westfall, 2017). Thus, the current study performed a machine learning method named linear support vector regression (SVR) and leave-one-out cross-validation (LOOCV) procedure (Beaty et al., 2018; Supekar et al., 2013) to examine the predictive performance of the brain connectivity indices that differed in the two groups (baseline) on future individual-level behavioral variables (one year later), that is, eating behaviors and BMI z-score. Behavioral variables were taken as the dependent variables, and connectivity strength indices (e.g., the hippocampus→PoCG connection) were taken as independent variables in the linear regression algorithm. The predictive model was trained on n − 1 participants’ neural indexes and eating behaviors, and subsequently tested on the left-out participant. After all the LOOCV folds were completed, we obtained the predicted scores of each participant. The predictive power of the model was assessed by the magnitude and statistical significance of Pearson’s correlation between observed and predicted values. If the observed and predicted eating behavior scores were significantly positively related, this would suggest that the model was successful in its prediction. Permutation testing was then used to assess the significance of the prediction results. We randomly shuffled the label between observed eating behaviors scores and neural indexes each time and reran the above LOOCV prediction procedure. The 1,000 Pearson’s correlations between observed and predicted scores composed null distributions of r values. The number of null r values was greater than or equal to the r_{(observed, predicted)} value and was then divided by 1,000, providing an estimated p_perm value.

2.8 Statistical analyses

A χ² test was conducted to test group differences in sex. A t test was applied to test group differences in demographic characteristics. An analysis of covariance (ANCOVA) model was implemented to examine group differences in eating behaviors and food choice task, with age and sex as covariates.

In neuroimaging data analyses, the ANCOVA was used to examine the differences in hippocampus-based FC values based on their z-maps between groups, which was implemented in the CONN toolbox (https://web.conn-toolbox.org). The model used random field theory parametric analyses, with statistical significance set at a voxel height threshold of p < 0.001, and the false discovery rate (FDR)-corrected cluster-size threshold of p < 0.05. To assess the effects of possible confounding factors, we controlled for the effects of age, sex, and head motion. Based on the FC results, we further used the spDCM approach to examine the directionality of connections, with statistical significance set at PP values > 0.95 (see 2.6. Effective connectivity for more details).

We treated sex, age, and eating behaviors (baseline, e.g., food enjoyment) as control measures and regressed the effect of these confounding factors (baseline) on eating behavior scores (one year later, e.g., food enjoyment) to ensure that the remaining variance was specific to the corresponding variables one year later, as suggested by Daker et al., (2020). We used the resultant residualized eating behavior scores for the longitudinal analyses. To identify the neural vulnerability factors for future eating behaviors and BMI, the current study used the Pearson’s correlation analysis to examine the potential correlational patterns between brain connections (baseline) and eating behaviors (one year later). This study additionally employed a machine learning method named SVR and LOOCV procedure to further examine whether baseline brain connectivity indexes could predict individual-level eating behaviors one year later). Permutation test was used to assess the significance of the prediction models.

3.1 Demographic and behavioral results

Demographic characteristics are presented in Table 1. No group differences were observed in age and head motion. Significant differences were observed in BMI z-score (t₍₈₈₎ = 17.67, p < 0.001) and sex (\({{\chi }}_{\left(1\right)}^{2}\)= 11.40, p < 0.001) between OW and NW groups. We further tested the group differences in different eating behaviors (i.e., the CEBQ sub-scales). We observed that the OW group showed lower scores in satiety responsiveness (F_(1,86) = − 4.70, p = 0.033), and slowness in eating (F_(1,86) = − 14.83, p < 0.001) sub-scales, relative to the NW group. See details in supplementary Table S1.

Table 1

Demographic characteristics of the samples
Variables	OW (n = 39)	NW (n = 51)	Statistics
Variables	OW (n = 39)	NW (n = 51)	χ²	t	df	p-value
Sex, female/male	12/27	34/17	11.40		1	< 0.001^**
Age, years	10.54 (0.90)	10.42 (0.87)		0.67	88	0.51
BMI z-score	1.98 (0.52)	0.03 (0.51)		17.67	88	< 0.001^**
BMI z-score_{_T2}	1.86 (0.65)	0.32 (0.62)		11.47	88	< 0.001^**
Head motion, mm	0.16 (0.09)	0.17 (0.10)		−0.80	88	0.43
Notes: Variables are expressed as mean (standard deviation), or n/n, as appropriate. ^** p < 0.001.

Abbreviation: OW = overweight group; NW = normal weight group; BMI = body mass index; T2 = one year later.

[Insert Table 1 here]

3.2 Group comparison of seed-based functional connectivity

Seed-based FC analyses found that compared to the NW group, the OW group exhibited lower connectivity strength between the right hippocampus and left PoCG (Table 2).

Table 2

Altered resting-state functional connectivity of the hippocampus between two groups
Seeds	Regions	MMI coordinates			Cluster size	t-value
Seeds	Regions	x	y	z	Cluster size	t-value
OW > NW
Hippocampus_R	PoCG_L	−48	−10	20	219	−4.80
Notes: Only significant results are reported. Statistical significance was set at a voxel height threshold of p < 0.001, and the cluster-size threshold of p < 0.05 false discovery rate (FDR) corrected (two-sided). Covariates included sex, age, and head motion.

[Insert Table 2 here]

3.3 Group comparison of effective connectivity

Compared to the NW group, the OW group showed higher inhibitory effect from the right hippocampus to left PoCG (EC_OW = − 0.338, EC_NW = − 0.129, PP > 0.95), and lower inhibitory effect from the left PoCG to right hippocampus (EC_OW = − 0.04, EC_NW = − 0.266, PP > 0.95) (Fig. 1).

[Fig. 1 around here]

3.4 Hippocampus connectivity predicts future BMI and eating behaviors

Machine learning analyses showed that the baseline hippocampus→PoCG EC could successfully predict individual-level BMI z-score (r_{(observed, predicted)} = 0.28, p_perm = 0.012), food approach (r(observed, predicted) = 0.29, p_perm = 0.004), and food as a reward (r_{(observed, predicted)} = 0.30, p_perm = 0.014) one year later. The baseline PoCG→hippocampus EC also effectively predicted individual-level food approach (r_{(observed, predicted)} = 0.16, p_perm = 0.047) one year later (Fig. 2). Interestingly, we further observed the opposite pattern of the relationship of effective connectivity with future eating behaviors and BMI (Fig. 3 and supplementary Fig. S2).

[Fig. 2 around here]

[Fig. 3 around here]

Given the crucial role of the hippocampus in obesity and food intake, the current study used the FC and EC indices, combined machine learning technique, to investigate the abnormal neurofunction in the hippocampus in children with obesity. Relative to the NW group, the OW group showed lower scores in satiety responsiveness and slowness in eating, in line with previous findings that lower slowness in eating and satiety contributed to childhood obesity (Dalrymple et al., 2020; Demir & Bektas, 2017). At the neural level, the OW group displayed weaker strength of FC between the right hippocampus and left PoCG, compared to controls. Regarding the connectivity directions, there was a stronger inhibitory effect from the hippocampus to PoCG but a weaker inhibitory effect from the PoCG to hippocampus in OW group. These ECs successfully predicted BMI and eating behaviors at one-year follow-up. More interestingly, the relationship of baseline ECs with behavioral variables one year later exhibited an opposite pattern. The current study provides novel insights into pathophysiology of obesity and highlights potential neural targets for the preventions and treatments of pediatric obesity.

4.1 Altered functional connectivity in children with obesity

The finding that the OW group exhibited lower FC strength between the hippocampus and PoCG than NW group was in accordance with the results of previous adult studies (Park et al., 2020; Sun et al., 2018). The hippocampus, a part of the limbic reward network, plays a crucial role in regulating food intake and integrates multimodal appetite-related signaling (Kanoski & Grill, 2017; Stevenson & Francis, 2017). The PoCG is an essential hub of the primary somatosensory cortex, mainly located in the oral sensory regions (Avery et al., 2021; Miyamoto et al., 2006). Activation of both the hippocampus and PoCG has been observed during tasting food, particularly unhealthy food (Huerta et al., 2014; Stice et al., 2013). Evidence from task-evoked fMRI studies suggested that higher responsivity in the hippocampus and PoCG when tasting food may reflect sensitivity to palatable food in children with obesity, which needed the somatosensory cortex to convey specific relevant food information (Mestre et al., 2017; Stice et al., 2011). Our finding that lower connectivity strength between the hippocampus and PoCG may suggest the imbalanced information communication of food properties processing in children with overweight/obesity.

4.2 Altered effective connectivity in children with obesity

To reveal the directional effect in the hippocampus–PoCG connectivity, the spDCM approach was further used to examine the group differences in EC. We found the significant alterations in the bidirectional effects between the hippocampus and PoCG in OW group. Specifically, the OW group displayed a stronger inhibitory effect from the hippocampus to PoCG. Although no prior study has explored directional links between key regions at rest in pediatric obesity samples, a recent study has observed a greater strength of the hippocampus→PoCG connection contributing to adults obesity (P. Zhang et al., 2021). The PoCG is involved in unimodal gustatory sensory processing, while the hippocampus serves as an important neural substrate of food reward processing, receiving and integrating multimodal food attributes from the primary somatosensory cortex (Kanoski & Grill, 2017; Miyamoto et al., 2006). Obesity has been shown to be associated with disrupted top-down pathway of food reward processing (Voigt et al., 2021). Therefore, the greater inhibitory influence from the hippocampus to PoCG represents the crucial neural substrates that could explain the abnormal interference of appetitive reward function on food sensory processing in children with obesity.

Interestingly, the bottom-up pathway of food processing have also changed in children with obesity, namely, the OW group displayed weaker inhibitory effect from the PoCG to hippocampus. There is evidence that the EC of the somatosensory areas toward hippocampus may act as a potential pathway for providing primary sensory information to higher cognitive processing (Rolls et al., 2022). Recently, P. Zhang et al., (2021) revealed abnormal top-down information flow (the hippocampus→PoCG connection) contributed to adult obesity, rather than the bottom-up information exchange (the PoCG→hippocampus connection). This inconsistent result could in part be due to the fact that, as individuals age, they tend to rely more on the top-down cognitive processes than the bottom-up sensory processes in response to external stimuli (Açık et al., 2010; Fuchs et al., 2024). Therefore, the inhibitory effect from the PoCG to hippocampus might reflect the potential information exchange, in which food sensory information in the PoCG converted into food reward responsivity information in the hippocampus, thus promoting healthier diets. However, the reduced PoCG→hippocampus connection strength of the OW group might suggest a decrease in such information exchange ability. This finding may reflect the specificity of the neural correlates of childhood obesity.

Taken together, the current findings expanded previous EC studies involving adults with obesity (Voigt et al., 2021; P. Zhang et al., 2021) and deepened our understanding of neurobiomarker of childhood obesity by revealing that the dysfunctional food reward processing system includes both the top-down (hippocampus→PoCG) and bottom-up (PoCG→hippocampus) pathways during such a critical developmental period. Since comparative researches on the relationship between directional interactions within food consumption regions and childhood obesity is lacking, this finding should be interpreted with caution and more research is clearly warranted.

4.3 Hippocampus connectivity predicts future BMI and eating behaviors

Using machine learning method, there were two major findings in the OW group: (i) the baseline hippocampus→PoCG connection effectively predicted one-year follow-up BMI z-score, food approach, and parental use of food as a reward; (ii) the baseline PoCG→hippocampus connection successfully predicted food approach one year later. The hippocampus has been identified as an appetitive region and involved in higher cognitive processing, such as food intake, reward, and memory (Stevenson & Francis, 2017). The PoCG is the primary somatosensory cortex and conveys sensory information to the hippocampus, contributing to food intake regulation (Kanoski & Grill, 2017). Evidence from recent longitudinal studies have demonstrated that activities in the hippocampus and somatosensory were linked with future BMI of adolescents and older adults (Burdette et al., 2022; Park et al., 2020). Based on larger neuroimaging datasets, our recent multivariate resting-state functional connectome-based prediction study also suggested that dysfunctions in the sensory and reward regions that involved in encoding food cues may act as reproducible biomarkers of obesity (Wang et al., 2023). In addition, adults with obesity showed that presurgery resting-state functional synchrony between the hippocampus and PoCG could predict changes of food approach one year later after bariatric surgery (Cerit et al., 2019). It should be noted that several patterns of EC may cause the same observations in FC (Z. Zhang et al., 2022; Zhou et al., 2018). The EC provides the directional communication between brain regions, further informing our understanding of FC patterns (Razi et al., 2015). Using the perspective of directional interactions, our findings support and expand the vicious circle model of obesity and suggest that impaired appetitive regulation between the hippocampus and somatosensory cortex may drive future consumption of harmful foods, thereby facilitating weight gain and obesity for children (Rapuano et al., 2016). Notably, we found that the EC from the hippocampus to PoCG was associated with parental use of food as a reward for their children with overweight/obesity one year later. Indeed, parental behaviors (e.g., parental use of food as a reward) have been found to have the potential impact on child food preference and overeating behavior (Benton, 2004; Powell et al., 2017). Our finding highlights that parental feeding behaviors play a vital role in children’s eating and the development of obesity (Anzman et al., 2010).

Furthermore, this study found the opposite pattern of the relationship of ECs with OW children’s eating behaviors and BMI one year later. In general, future BMI and eating behaviors tended to be positively correlated with the top-down, hippocampus→PoCG connection, while negatively correlated with the bottom-up, PoCG→hippocampus connection. These findings suggested that the dysfunctional top-down mechanism of food reward processing may be a more sensitive neural predictor of children’s future BMI and eating behaviors, relative to the bottom-up pathway (Kaisari et al., 2019). Meanwhile, machine learning analyses also indicated that the top-down EC showed significantly stronger predictive efficacy for eating behaviors compared to the bottom-up EC, which may again support the inference above to some extent. Future interventions of children’s obesity and problematic eating should be improved in particular the capacity for food regulatory processing by targeted training (Kemps et al., 2020).

4.4 Limitations and future directions

This study has several potential limitations. First, the current study was a longitudinal design and found the sensitive neural predictors of children BMI and eating behaviors, but it cannot infer the causal relationship between them. Second, although all participants underwent brain scanning at a fixed time period between 11:00 and 13:30 after lunch, the level of feelings of hunger were not collected, which should be considered in future studies. Additionally, the follow-up time between two visits was rather short (1 year). Future studies could follow up a longer interval (more than 10 years) and use more sophisticated analyses, such as growth curve modeling (Rogers et al., 2023), to investigate how brain changes in children with obesity over time, and how these changes contribute to future eating behaviors and weight gain during childhood. Another limitation of this study was that our sample had a fairly restricted age range (i.e., 8–12 years old) and had relatively small sample size, which may affect the generalizability of our results. Future studies are needed to investigate and ascertain the neural specificity and comorbidity across different age periods, which could contribute to the identification of neurological risk factors for childhood obesity and further determine the key age points for early and more targeted prevention and intervention.

This is the first attempt to investigate the critical role of hippocampal undirectional and directional connections in childhood obesity from cross-sectional and longitudinal perspectives. Imbalanced information exchange in the appetitive reward circuitry between the hippocampus to somatosensory cortex, as reflected by bidirectional inhibitory effects, may be the crucial neurofunctional substrates of pediatric obesity and future food approach behavior. This study expands the vicious circle model of obesity by revealing the crucial role of hippocampus-based undirectional and directional integration in childhood obesity, which is essential for developing effective intervention strategies and reducing long-term health-care costs associated with obesity.

Compliance with Ethical Standards：

The research protocol was reviewed for compliance with the standards for the ethical treatment of human participants and approved by the Ethical Committee for Scientific Research at the university with which the authors are affiliated.

Conflict of Interest:

The authors declare that they have no conflict of interest.

Research Involving Human Participants:

All ethical guidelines for human subjects’ research were followed.

Informed Consent:

All participants provided written informed consents to participate in the study.

Funding:

This study was funded by National Natural Science Foundation of China (No. 32271087).

Acknowledgments

This study was funded by National Natural Science Foundation of China (No. 32271087). We would like to thank all the members of the Self and Health Research Lab, Southwest University for their invaluable help in data acquisition.

Açık, A., Sarwary, A., Schultze-Kraft, R., Onat, S., & König, P. (2010). Developmental changes in natural viewing behavior: Bottom-up and top-down differences between children, young adults and older adults. Frontiers in Psychology, 1, 7198.
Anzman, S. L., Rollins, B. Y., & Birch, L. L. (2010). Parental influence on children’s early eating environments and obesity risk: Implications for prevention. International Journal of Obesity, 34(7), 1116–1124.
Avery, J. A., Liu, A. G., Ingeholm, J. E., Gotts, S. J., & Martin, A. (2021). Viewing images of foods evokes taste quality-specific activity in gustatory insular cortex. Proceedings of the National Academy of Sciences, 118(2), e2010932118.
Bajaj, S., & Killgore, W. D. S. (2021). Association between emotional intelligence and effective brain connectome: A large-scale spectral DCM study. NeuroImage, 229, 117750. https://doi.org/10.1016/j.neuroimage.2021.117750
Bauer, C., Moreno, B., González-Santos, L., Concha, L., Barquera, S., & Barrios, F. (2015). Child overweight and obesity are associated with reduced executive cognitive performance and brain alterations: A magnetic resonance imaging study in M exican children. Pediatric Obesity, 10(3), 196–204.
Beaty, R. E., Kenett, Y. N., Christensen, A. P., Rosenberg, M. D., Benedek, M., Chen, Q., Fink, A., Qiu, J., Kwapil, T. R., & Kane, M. J. (2018). Robust prediction of individual creative ability from brain functional connectivity. Proceedings of the National Academy of Sciences, 115(5), 1087–1092.
Benton, D. (2004). Role of parents in the determination of the food preferences of children and the development of obesity. International Journal of Obesity, 28(7), 858–869.
Bhutani, S., Christian, I. R., Palumbo, D., & Wiggins, J. L. (2021). Reward-related neural correlates in adolescents with excess body weight. NeuroImage: Clinical, 30, 102618. https://doi.org/10.1016/j.nicl.2021.102618
Birch, L. L., Fisher, J. O., Grimm-Thomas, K., Markey, C. N., Sawyer, R., & Johnson, S. L. (2001). Confirmatory factor analysis of the Child Feeding Questionnaire: A measure of parental attitudes, beliefs and practices about child feeding and obesity proneness. Appetite, 36(3), 201–210.
Blain, S. D., Taylor, S. F., Lasagna, C. A., Angstadt, M., Rutherford, S. E., Peltier, S., Diwadkar, V. A., & Tso, I. F. (2023). Aberrant effective connectivity during eye gaze processing is linked to social functioning and symptoms in schizophrenia. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 8(12), 1228–1239.
Boswell, R. G., & Kober, H. (2016). Food cue reactivity and craving predict eating and weight gain: A meta-analytic review. Obesity Reviews, 17(2), 159–177.
Bruce, A., Holsen, L., Chambers, R., Martin, L., Brooks, W., Zarcone, J., Butler, M., & Savage, C. R. (2010). Obese children show hyperactivation to food pictures in brain networks linked to motivation, reward and cognitive control. International Journal of Obesity, 34(10), 1494–1500.
Burdette, J. H., Bahrami, M., Laurienti, P. J., Simpson, S. L., Nicklas, B. J., Fanning, J., & Rejeski, W. J. (2022). Longitudinal relationship of baseline functional brain networks with intentional weight loss in older adults. Obesity, 30(4), 902–910.
Cerit, H., Davidson, P., Hye, T., Moondra, P., Haimovici, F., Sogg, S., Shikora, S., Goldstein, J. M., Evins, A. E., & Whitfield-Gabrieli, S. (2019). Resting‐state brain connectivity predicts weight loss and cognitive control of eating behavior after vertical sleeve gastrectomy. Obesity, 27(11), 1846–1855.
Chen, X., Li, W., Liu, Y., Xiao, M., & Chen, H. (2023). Altered effective connectivity between reward and inhibitory control networks in people with binge eating episodes: A spectral dynamic causal modeling study. Appetite, 188, 106763. https://doi.org/10.1016/j.appet.2023.106763
Chen, X., Li, W., Qin, J., Gao, X., Liu, Y., Song, S., Huang, Y., & Chen, H. (2022). Gray matter volume and functional connectivity underlying binge eating in healthy children. Eating and Weight Disorders-Studies on Anorexia, Bulimia and Obesity, 27(8), 3469–3478.
Clasen, M. M., Riley, A. L., & Davidson, T. L. (2020). Hippocampal-dependent inhibitory learning and memory processes in the control of eating and drug taking. Current Pharmaceutical Design, 26(20), 2334–2352.
Cui, Z., Su, M., Li, L., Shu, H., & Gong, G. (2018). Individualized prediction of reading comprehension ability using gray matter volume. Cerebral Cortex, 28(5), 1656–1672.
Daker, R. J., Cortes, R. A., Lyons, I. M., & Green, A. E. (2020). Creativity anxiety: Evidence for anxiety that is specific to creative thinking, from STEM to the arts. Journal of Experimental Psychology: General, 149(1), 42.
Dalrymple, K. V., Flynn, A. C., Seed, P. T., Briley, A. L., O’Keeffe, M., Godfrey, K. M., & Poston, L. (2020). Associations between dietary patterns, eating behaviours, and body composition and adiposity in 3-year‐old children of mothers with obesity. Pediatric Obesity, 15(5), e12608.
Davids, S., Lauffer, H., Thoms, K., Jagdhuhn, M., Hirschfeld, H., Domin, M., Hamm, A., & Lotze, M. (2010). Increased dorsolateral prefrontal cortex activation in obese children during observation of food stimuli. International Journal of Obesity, 34(1), 94–104.
Demir, D., & Bektas, M. (2017). The effect of childrens’ eating behaviors and parental feeding style on childhood obesity. Eating Behaviors, 26, 137–142.
Duan, S., Ji, G., Li, G., Hu, Y., Zhang, W., Wang, J., Tomasi, D., Volkow, N. D., Nie, Y., Cui, G., Wang, G.-J., & Zhang, Y. (2020). Bariatric surgery induces alterations in effective connectivity between the orbitofrontal cortex and limbic regions in obese patients. Science China Information Sciences, 63(7), 170104. https://doi.org/10.1007/s11432-019-2817-x
Eichenbaum, H. (2000). A cortical–hippocampal system for declarative memory. Nature Reviews Neuroscience, 1(1), 41–50.
Eklund, K., Paavonen, E. J., & Almqvist, F. (2005). Factor structure of the eating disorder inventory-C. International Journal of Eating Disorders, 37(4), 330–341.
Friston, K. J., Harrison, L., & Penny, W. (2003). Dynamic causal modelling. Neuroimage, 19(4), 1273–1302.
Friston, K. J., Kahan, J., Biswal, B., & Razi, A. (2014). A DCM for resting state fMRI. NeuroImage, 94, 396–407. https://doi.org/10.1016/j.neuroimage.2013.12.009
Fuchs, B. A., Pearce, A. L., Rolls, B. J., Wilson, S. J., Rose, E. J., Geier, C. F., & Keller, K. L. (2024). Does ‘portion size’matter? Brain responses to food and non-food cues presented in varying amounts. Appetite, 107289.
Geha, P., Cecchi, G., Todd Constable, R., Abdallah, C., & Small, D. M. (2017). Reorganization of brain connectivity in obesity. Human Brain Mapping, 38(3), 1403–1420.
Granger, C. W. (1969). Investigating causal relations by econometric models and cross-spectral methods. Econometrica: Journal of the Econometric Society, 424–438.
Hare, T. A., Camerer, C. F., & Rangel, A. (2009). Self-control in decision-making involves modulation of the vmPFC valuation system. Science, 324(5927), 646–648.
Huerta, C. I., Sarkar, P. R., Duong, T. Q., Laird, A. R., & Fox, P. T. (2014). Neural bases of food perception: Coordinate-based meta‐analyses of neuroimaging studies in multiple modalities. Obesity, 22(6), 1439–1446.
Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage, 17(2), 825–841.
Jiang, F., Li, G., Ji, W., Zhang, Y., Wu, F., Hu, Y., Zhang, W., Manza, P., Tomasi, D., & Volkow, N. D. (2023). Obesity is associated with decreased gray matter volume in children: A longitudinal study. Cerebral Cortex, 33(7), 3674–3682.
Kaisari, P., Kumar, S., Hattersley, J., Dourish, C. T., Rotshtein, P., & Higgs, S. (2019). Top-down guidance of attention to food cues is enhanced in individuals with overweight/obesity and predicts change in weight at one-year follow up. International Journal of Obesity, 43(9), 1849–1858.
Kanoski, S. E., & Davidson, T. L. (2011). Western diet consumption and cognitive impairment: Links to hippocampal dysfunction and obesity. Physiology & Behavior, 103(1), 59–68.
Kanoski, S. E., & Grill, H. J. (2017). Hippocampus contributions to food intake control: Mnemonic, neuroanatomical, and endocrine mechanisms. Biological Psychiatry, 81(9), 748–756.
Kemps, E., Goossens, L., Petersen, J., Verbeken, S., Vervoort, L., & Braet, C. (2020). Evidence for enhancing childhood obesity treatment from a dual-process perspective: A systematic literature review. Clinical Psychology Review, 77, 101840.
Kong, F. (2012). The neural correlates of food cues processing in restrained eaters. Southwest University: Doctor Thesis.
Lee, H., Kwon, J., Lee, J., Park, B., & Park, H. (2022). Disrupted stepwise functional brain organization in overweight individuals. Communications Biology, 5(1), 11.
Li, G., Hu, Y., Zhang, W., Ding, Y., Wang, Y., Wang, J., He, Y., Lv, G., von Deneen, K. M., & Zhao, Y. (2021). Resting activity of the hippocampus and amygdala in obese individuals predicts their response to food cues. Addiction Biology, 26(3), e12974.
Li, W., Chen, X., Luo, Y., Xiao, M., Liu, Y., & Chen, H. (2024). Altered connectivity patterns of medial and lateral orbitofrontal cortex underlie the severity of bulimic symptoms. International Journal of Clinical and Health Psychology, 24(1), 100439.
Lister, N. B., Baur, L. A., Felix, J. F., Hill, A. J., Marcus, C., Reinehr, T., Summerbell, C., & Wabitsch, M. (2023). Child and adolescent obesity. Nature Reviews Disease Primers, 9(1), 24.
Lowe, C. J., Morton, J. B., & Reichelt, A. C. (2020). Adolescent obesity and dietary decision making—A brain-health perspective. The Lancet Child & Adolescent Health, 4(5), 388–396. https://doi.org/10.1016/S2352-4642(19)30404-3
Mestre, Z. L., Bischoff-Grethe, A., Eichen, D. M., Wierenga, C. E., Strong, D., & Boutelle, K. N. (2017). Hippocampal atrophy and altered brain responses to pleasant tastes among obese compared with healthy weight children. International Journal of Obesity, 41(10), 1496–1502.
Miyamoto, J. J., Honda, M., Saito, D. N., Okada, T., Ono, T., Ohyama, K., & Sadato, N. (2006). The representation of the human oral area in the somatosensory cortex: A functional MRI study. Cerebral Cortex, 16(5), 669–675.
Ochner, C. N., Tsai, A. G., Kushner, R. F., & Wadden, T. A. (2015). Treating obesity seriously: When recommendations for lifestyle change confront biological adaptations. The Lancet Diabetes & Endocrinology, 3(4), 232–234.
Onis, M. de, Onyango, A. W., Borghi, E., Siyam, A., Nishida, C., & Siekmann, J. (2007). Development of a WHO growth reference for school-aged children and adolescents. Bulletin of the World Health Organization, 85(9), 660–667.
Pan, X.-F., Wang, L., & Pan, A. (2021). Epidemiology and determinants of obesity in China. The Lancet Diabetes & Endocrinology, 9(6), 373–392. https://doi.org/10.1016/S2213-8587(21)00045-0
Park, B., Byeon, K., Lee, M. J., Chung, C., Kim, S., Morys, F., Bernhardt, B., Dagher, A., & Park, H. (2020). Whole-brain functional connectivity correlates of obesity phenotypes. Human Brain Mapping, 41(17), 4912–4924.
Phelps, N. H., Singleton, R. K., Zhou, B., Heap, R. A., Mishra, A., Bennett, J. E., Paciorek, C. J., Lhoste, V. P., Carrillo-Larco, R. M., & Stevens, G. A. (2024). Worldwide trends in underweight and obesity from 1990 to 2022: A pooled analysis of 3663 population-representative studies with 222 million children, adolescents, and adults. The Lancet.
Powell, E. M., Frankel, L. A., & Hernandez, D. C. (2017). The mediating role of child self-regulation of eating in the relationship between parental use of food as a reward and child emotional overeating. Appetite, 113, 78–83.
Preller, K. H., Razi, A., Zeidman, P., Stämpfli, P., Friston, K. J., & Vollenweider, F. X. (2019). Effective connectivity changes in LSD-induced altered states of consciousness in humans. Proceedings of the National Academy of Sciences, 116(7), 2743–2748. https://doi.org/10.1073/pnas.1815129116
Quigley, K. S., Kanoski, S., Grill, W. M., Barrett, L. F., & Tsakiris, M. (2021). Functions of interoception: From energy regulation to experience of the self. Trends in Neurosciences, 44(1), 29–38.
Randeniya, R., Vilares, I., Mattingley, J., & Garrido, M. (2023). Increased functional activity, bottom-up and intrinsic effective connectivity in autism. NeuroImage: Clinical, 37, 103293.
Rapuano, K. M., Huckins, J. F., Sargent, J. D., Heatherton, T. F., & Kelley, W. M. (2016). Individual differences in reward and somatosensory-motor brain regions correlate with adiposity in adolescents. Cerebral Cortex, 26(6), 2602–2611.
Razi, A., Kahan, J., Rees, G., & Friston, K. J. (2015). Construct validation of a DCM for resting state fMRI. Neuroimage, 106, 1–14.
Rogers, C. R., Jimenez, V., Benjamin, A., Rudolph, K. D., & Telzer, E. H. (2023). The effect of parents and peers on the neural correlates of risk taking and antisocial behavior during adolescence. Journal of Youth and Adolescence, 52(8), 1674–1684.
Rolls, E. T., Deco, G., Huang, C.-C., & Feng, J. (2022). The effective connectivity of the human hippocampal memory system. Cerebral Cortex, 32(17), 3706–3725.
Rolls, E. T., Huang, C.-C., Lin, C.-P., Feng, J., & Joliot, M. (2020). Automated anatomical labelling atlas 3. Neuroimage, 206, 116189.
Samara, A., Li, X., Pivik, R. T., Badger, T. M., & Ou, X. (2018). Brain activation to high-calorie food images in healthy normal weight and obese children: A fMRI study. BMC Obesity, 5(1), 31. https://doi.org/10.1186/s40608-018-0209-1
Schmitt, L. O., & Gaspar, J. M. (2023). Obesity-induced brain neuroinflammatory and mitochondrial changes. Metabolites, 13(1), 86.
Shen, X., Finn, E. S., Scheinost, D., Rosenberg, M. D., Chun, M. M., Papademetris, X., & Constable, R. T. (2017). Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nature Protocols, 12(3), 506–518.
Solis-Urra, P., Esteban‐Cornejo, I., Mora‐Gonzalez, J., Stillman, C., Contreras‐Rodriguez, O., Erickson, K. I., Catena, A., & Ortega, F. B. (2023). Early life factors and hippocampal functional connectivity in children with overweight/obesity. Pediatric Obesity, 18(3), e12998.
Stevenson, R. J., & Francis, H. M. (2017). The hippocampus and the regulation of human food intake. Psychological Bulletin, 143(10), 1011.
Stice, E., Burger, K. S., & Yokum, S. (2013). Relative ability of fat and sugar tastes to activate reward, gustatory, and somatosensory regions. The American Journal of Clinical Nutrition, 98(6), 1377–1384.
Stice, E., Yokum, S., Burger, K. S., Epstein, L. H., & Small, D. M. (2011). Youth at risk for obesity show greater activation of striatal and somatosensory regions to food. Journal of Neuroscience, 31(12), 4360–4366.
Sun, Q., Chen, G.-Q., Wang, X.-B., Yu, Y., Hu, Y.-C., Yan, L.-F., Zhang, X., Yang, Y., Zhang, J., & Liu, B. (2018). Alterations of white matter integrity and hippocampal functional connectivity in type 2 diabetes without mild cognitive impairment. Frontiers in Neuroanatomy, 12, 21.
Supekar, K., Swigart, A. G., Tenison, C., Jolles, D. D., Rosenberg-Lee, M., Fuchs, L., & Menon, V. (2013). Neural predictors of individual differences in response to math tutoring in primary-grade school children. Proceedings of the National Academy of Sciences, 110(20), 8230–8235.
Syan, S. K., McIntyre-Wood, C., Minuzzi, L., Hall, G., McCabe, R. E., & MacKillop, J. (2021). Dysregulated resting state functional connectivity and obesity: A systematic review. Neuroscience & Biobehavioral Reviews, 131, 270–292.
Voigt, K., Razi, A., Harding, I. H., Andrews, Z. B., & Verdejo-Garcia, A. (2021). Neural network modelling reveals changes in directional connectivity between cortical and hypothalamic regions with increased BMI. International Journal of Obesity, 45(11), Article 11. https://doi.org/10.1038/s41366-021-00918-y
Wang, Y., & Chen, H.-J. (2012). Use of percentiles and z-scores in anthropometry. In Handbook of anthropometry: Physical measures of human form in health and disease (pp. 29–48). Springer.
Wang, Y., Dong, D., Chen, X., Gao, X., Liu, Y., Xiao, M., Guo, C., & Chen, H. (2023). Individualized morphometric similarity predicts body mass index and food approach behavior in school-age children. Cerebral Cortex, 33(8), 4794–4805.
Wardle, J., Guthrie, C. A., Sanderson, S., & Rapoport, L. (2001). Development of the children’s eating behaviour questionnaire. The Journal of Child Psychology and Psychiatry and Allied Disciplines, 42(7), 963–970.
Webber, L., Hill, C., Saxton, J., Van Jaarsveld, C., & Wardle, J. (2009). Eating behaviour and weight in children. International Journal of Obesity, 33(1), 21–28.
World Health Organization. (2006). WHO child growth standards: Length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: Methods and development. World Health Organization.
Yan, C.-G., Cheung, B., Kelly, C., Colcombe, S., Craddock, R. C., Di Martino, A., Li, Q., Zuo, X.-N., Castellanos, F. X., & Milham, M. P. (2013). A comprehensive assessment of regional variation in the impact of head micromovements on functional connectomics. Neuroimage, 76, 183–201.
Yarkoni, T., & Westfall, J. (2017). Choosing prediction over explanation in psychology: Lessons from machine learning. Perspectives on Psychological Science, 12(6), 1100–1122.
Zeidman, P., Jafarian, A., Corbin, N., Seghier, M. L., Razi, A., Price, C. J., & Friston, K. J. (2019). A guide to group effective connectivity analysis, part 1: First level analysis with DCM for fMRI. NeuroImage, 200, 174–190. https://doi.org/10.1016/j.neuroimage.2019.06.031
Zhang, J., Li, Y., Li, F., He, M., Li, J., Zhang, S., Zhao, W., Tang, Y., Li, Y., & Xiong, J. (2024). Association between sugar-free beverage intake and childhood obesity among Chinese children and adolescents. Pediatric Obesity, 19(3), e13096.
Zhang, P., Liu, Y., Yu, F., Wu, G., Li, M., Wang, Z., Ding, H., Wang, L., Zhao, K., Zhang, Z., Zhao, P., Li, J., Yang, Z., Lv, H., Zhang, Z., & Wang, Z. (2021). Hierarchical integrated processing of reward-related regions in obese males: A graph-theoretical-based study. Appetite, 159, 105055. https://doi.org/10.1016/j.appet.2020.105055
Zhang, Z., Bo, Q., Li, F., Zhao, L., Wang, Y., Liu, R., Chen, X., Wang, C., & Zhou, Y. (2022). Altered effective connectivity among core brain networks in patients with bipolar disorder. Journal of Psychiatric Research, 152, 296–304. https://doi.org/10.1016/j.jpsychires.2022.06.031
Zheng, L., Song, D., Chen, C., Li, F., & Zhu, D. (2016). Reliability and validity of a Chinese version of Child Feeding Questionnaire among parents of preschoolers. Chin. J. Child Health Care, 24, 1019–1023.
Zhou, Y., Friston, K. J., Zeidman, P., Chen, J., Li, S., & Razi, A. (2018). The Hierarchical Organization of the Default, Dorsal Attention and Salience Networks in Adolescents and Young Adults. Cerebral Cortex, 28(2), 726–737. https://doi.org/10.1093/cercor/bhx307

There is NO conflict of interest to disclose

Supplementarymatieral.pdf

Download PDF

Version 1

posted

You are reading this latest preprint version

Altered hippocampal effective connectivity predicts BMI and food approach behavior in children with obesity

Status:

Version 1

Abstract

Objective

Methods

Results

Conclusion

Figures

1 Introduction

2 Methods

2.1 Participants

2.2 Food choice task

2.3 Measures

2.3.1 BMI and WHO weight status

2.3.2 Children’s Eating Behaviour Questionnaire

2.3.3 Child Feeding Questionnaire

2.3.4 Eating Disorder Inventory-Child

2.4 Neuroimaging data acquisition and preprocessing

2.5 Seed-based functional connectivity

2.6 Effective connectivity

2.7 Longitudinal predictive analysis

2.8 Statistical analyses

3 Results

3.1 Demographic and behavioral results

3.2 Group comparison of seed-based functional connectivity

3.3 Group comparison of effective connectivity

3.4 Hippocampus connectivity predicts future BMI and eating behaviors

4 Discussion

4.1 Altered functional connectivity in children with obesity

4.2 Altered effective connectivity in children with obesity

4.3 Hippocampus connectivity predicts future BMI and eating behaviors

4.4 Limitations and future directions

5 Conclusions

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1