Table 1.
Summary of studies investigating the relationship between functional connectivity and obesity
| Author | Design | Variable of interest | FC technique | N | Attrition | Mean age (SD) | Mean BMI (SD) | Findings |
|---|---|---|---|---|---|---|---|---|
| Chao et al., 2018 | Cross-sectional | Resting-state FC among individuals with obesity prior to bariatric surgery | Seed-based resting-state FC plus graph theory; the authors did not specify which seed(s) they used. | Obese: n = 20 | n/a | Obese: 33.11 (8.86) | Obese: 37.66 (5.07) | Obesity was associated with increased FC in the bilateral anterior cingulate cortex and precuneus and with decreased FC in the medial prefrontal cortex. Obesity predicted lower cluster coefficients and modularity and higher global efficiency, suggesting a shift toward more random networks. |
| Healthy weight: 45.03 (9.65) | Healthy weight: 22.64 (3.45) | |||||||
| Healthy weight: n = 30 | *Significantly different | |||||||
| Dietrich et al., 2016 | Cross-sectional | Relationship between obesity and FC during the regulation of food cravings; BMI was analyzed continuously. | PPI analysis during regulation of cravings | N = 43 | n/a | 26.7 (3.5) | 27.5 (5.3) | BMI was positively associated with FC between the putamen and dorsal lateral and medial prefrontal cortex during regulation of craving. |
| García-García et al., 2013 | Cross-sectional | Functional activation and FC in response to high- and low-calorie food cues | ICA | Obese: n = 18 | n/a | Obese: 34.78 (4.45) | Obese: 34.89 (4.78) | During the presentation of high-calorie food, obesity was associated with reduced FC of lateral and medial prefrontal regions and precuneus. |
| Healthy weight: n = 19 | Healthy weight: 32.00 (5.87) | Healthy weight: 22.44 (1.93) | ||||||
| Kube et al., 2018 | Cross-sectional | Relationship of obesity with FC during a probabilistic learning task | PPI using the ventral striatum as the seed | Obese: n = 19 | n/a | Obese: 29.5 (5.6) | Obese: 35.4 (4.5) | Individuals with obesity exhibited increased FC between the ventral striatum, insula and superior temporal gyrus during prediction error processing. |
| Healthy weight: n = 23 | Healthy weight: 30.0 (5.0) | Healthy weight: 22.5 (1.7) | ||||||
| Sadler et al., 2018 | Cross-sectional | Effect of weight on resting-state FC in weight discordant identical twins | ICA | Weight discordant twins: n = 94 | n/a | Weight discordant twins: 29.4 (3.5) | Weight discordant twins: 28.1 (5.6) | In BMI-discordant twins, twins with lower BMI had stronger FC between striatal/thalamic and prefrontal networks. FC patterns observed in the BMI-discordant twin sample were not seen in a BMI-similar sample, providing evidence that the results are specific to BMI discordance. |
| Weight concordant twins: n = 94 | Weight concordant twins: 28.5 (3.6) | Weight concordant twins: 25.6 (4.1) | ||||||
| Weight discordant unrelated individuals: n = 94 | Weight discordant unrelated individuals: 29.3 (3.5) | Weight discordant unrelated individuals: 28.1 (5.7) | ||||||
| Stoeckel et al., 2009 | Cross-sectional | Functional activation and FC in response to high- and low-calorie food cues | PPI using nucleus accumbens, amygdala and orbitofrontal cortex as seeds | Obese: n = 12 | n/a | Obese: 27.8 (6.2) | Obese: 30.8–41.2 | In response to high-calorie foods, individuals with obesity exhibited reduced amygdala FC with the orbitofrontal cortex and nucleus accumbens, but increased FC between orbitofrontal cortex and nucleus accumbens. |
| Healthy weight: n = 12 | Healthy weight: 28 (4.4) | Healthy weight: 19.7–24.5 | ||||||
| Verdejo-Román et al., 2017 | Cross-sectional | Characterize FC during processing of food and monetary rewards | Graph theory | Excess weight: n = 39 | n/a | Excess weight: 33.59 (6.23) | Excess weight: 30.41 (3.69) | Excess weight was associated with decreased FC during the processing of food rewards in a network involving primarily frontal and striatal areas and increased FC during the processing of monetary rewards in a network involving principally frontal and parietal areas. |
| Healthy weight: n = 37 | Healthy weight: 33 (6.53) | Healthy weight: 22.28 (1.77) | ||||||
| Lips et al., 2014 | Cross-sectional | Associations between fasting and resting-state FC among individuals with obesity with and without type II diabetes (T2DM) | Seed-based resting-state FC using the posterior cingulate cortex, hypothalamus and amygdala as seeds | T2DM: n = 19 | n/a | T2DM: 47.7 (6.4) | T2DM: 43.8 (3.2) | No significant differences between normal-glucose-tolerant and T2DM subjects were observed. In the fasting state, obesity was associated with stronger hypothalamic FC with the medial prefrontal cortex and the dorsal striatum. The amygdala was differentially connected to the right insula in those with obesity. Food intake dampened hypothalamic FC with the frontal regions in lean subjects, whereas these connections were not affected in those with obesity. |
| Non-T2DM: n = 27 | Non-T2DM: 51 (7.1) | Non-T2DM: 42 (5.5) | ||||||
| Healthy weight: n = 12 | Healthy weight: 49.2 (6.22) | Healthy weight: 21.7 (1.6) | ||||||
| Wijngaarden et al., 2015a | Cross-sectional | Associations between fasting and resting-state FC | Seed-based resting-state FC using the posterior cingulate cortex, hypothalamus and amygdala as seeds | Obese n = 13 | n/a | Obese: 31 (3) | Obese: 35.4 (1.2) | At baseline, obesity was associated with stronger FC between hypothalamus and left insula. This effect diminished upon the prolonged fast. |
| Healthy weight: n = 11 | Healthy weight: 28 (3) | Healthy weight: 23.2 (0.5) | ||||||
| Healthy weight: n = 31 | Healthy weight: 27.05 (7.03) | Healthy Weight: 22.32 (19.52 to 25.09 kg m2) | ||||||
| SD not reported | ||||||||
| Coveleskie et al., 2015 | Cross-sectional | Associations between obesity and resting FC of the nucleus accumbens | Seed-based resting-state FC using the nucleus accumbens as a seed | Obese: n = 19 | n/a | Obese: 25.42 (5.86); | Obese: 31.83 (25.88 to 37.56 kg m2) | Subjects with high BMI had greater FC of the left nucleus accumbens with bilateral anterior cingulate cortex and right ventromedial prefrontal cortex. |
| Doucet et al., 2017 | Cross-sectional | Relationship between BMI and resting-state FC of canonical networks; BMI was analyzed continuously. | Graph theory | N = 496 | n/a | 29 (Range of 22–37 years) | 26.6 (Range of 16.8–47.8 years) | Elevated BMI was associated with reduced functional cohesion and increased integration of sensory-driven networks (sensorimotor and visual) and internally guided networks (default mode and central executive). |
| Beyer et al., 2017 | Cross-sectional | Relationship between BMI and resting-state FC of canonical networks; BMI was analyzed continuously. | ICA | N = 521 | n/a | 70.1 (3.8) | 27.5 (4.1) | Higher BMI was significantly associated with lower default mode FC in the posterior cingulate cortex and precuneus. |
| Baek et al., 2017 | Cross-sectional | Relationship between obesity and resting-state FC | Resting-state FC between 90 regions defined using the AAL atlas | Obese: n = 20 | n/a | Obese: 42.7 (11.1) | Obese: 33.4 (3.9) | Obesity was associated with global and local network efficiency as well as decreased modularity. In regional metrics, the putamen, pallidum and thalamus exhibited significantly decreased nodal degree and efficiency among individuals with obesity. Obesity was also associated with decreased FC of cortico-striatal/cortico-thalamic networks. |
| BED: n = 20 | BED: 43.7 (9.6) | BED: 33.0 (2.4) | ||||||
| Healthy weight: n = 40 | Healthy weight: 41.8 (11.7) | Healthy weight: 22.5 (2.0) | ||||||
| Garcia-Garcia et al., 2015 | Cross-sectional | Relationship between obesity and resting-state FC | Graph theory | Obese: n = 20 | n/a | Obese: 33.55 (5.61) | Obese: 35.90 (5.83) | Individuals with obesity exhibited less degree centrality in the left middle frontal gyrus and the lateral occipital cortex. |
| Healthy weight: n = 21 | Healthy weight: 31.33 (5.96) | Healthy weight: 22.33 (1.87) | ||||||
| Frank et al., 2014 | Cross-sectional | Effect of gastric bypass surgery on resting-state and task-evoked (food cue) FC | ICA | Obese: n = 11 | n/a | Obese: 42.6 (4.0) | Obese: 40.2 (0.8) | Individuals with obesity who had not undergone surgery exhibited stronger FC in frontal regions of the DMN during resting state compared to healthy weight individuals and those who underwent surgery. |
| Bypass surgery: n = 9 | Bypass surgery: 42.0 (2.8) | Bypass surgery: 27.1 (0.9) | ||||||
| Healthy weight: n = 11 | Healthy weight: 36.6 (3.8) | Healthy weight: 21.4 (0.5) | ||||||
| Atalayer et al., 2014 | Cross-sectional | Sex differences in neural responsitivity to food cues | PPI analysis; seed-based FC on amygdala in response to high- vs low-calorie food cues in fed vs fasted state | Female: n = 14 | n/a | Female: 35 (6.9) | Female: 36.9 (5.6) | In response to high- vs low-calorie food cues, obese men (vs women) had greater FC between amygdala and right subgenual anterior cingulate, whereas obese women had greater functional connectivity with amygdala in left angular gyrus and right primary motor areas. |
| Male: n = 17 | Male: 35 (9.0) | Male: 36.2 (5.5) | ||||||
| Carnell et al., 2014 | Cross-sectional | Functional activation and FC in response to high and low calorie visual and auditory food cues | PPI analysis of a visual food cue task | Obese: n = 10 | n/a | Obese: 22.4 (2) | Obese: 32.9 (5.3) | Individuals with obesity exhibited greater response to high-calorie cues, as well as relatively greater FC between the ventral tegmental area and cerebellum. |
| Healthy weight: n = 10 | Healthy weight: 21 (1.2) | Healthy weight: 22.1 (1.2) | ||||||
| Geha et al., 2017 | Cross-sectional | Global brain FC during anticipation and tasting of a milkshake | Seed-based FC during the tasting of a milkshake | Healthy weight: n = 15 | n/a | Obese: 27.7 (1.7) | Obese: 35.3 (.9) | At rest and during milkshake consumption, global FC was consistently decreased in the ventromedial and ventrolateral prefrontal cortex, insula and caudate nucleus and was increased in brain regions belonging to the dorsal attention network including premotor areas, superior parietal lobule and visual cortex. |
| Obese: n = 15 | Healthy weight: 27.4 (1.7) | Healthy weight: 21.9 (.5) | ||||||
| Kullmann et al., 2013 | Cross-sectional | Relationship between obesity and FC during a visual food cue task | ICA | Obese/overweight: n = 12 | n/a | Obese/overweight: 24.66 (2.42) | Obese/overweight: 30.46 (1.77) | Overweight/obesity was associated with increased FC of the salience network and networks involved in object recognition, motivational salience. |
| Healthy weight: n = 12 | Healthy weight: 22.91 (2.1) | Healthy weight: 21.16 (1.13) | ||||||
| Nummenmaa et al., 2012 | Cross-sectional | Functional activation and FC in response to high- and low-calorie food cues | PPI analysis | Obese: n = 19 | n/a | Obese: 45.72 (9.60) | Obese: 43.97 (3.74) | Obesity was associated with increased FC between the caudate nucleus and the amygdala, posterior insula and somatosensory cortex during the presentation of high calorie food. |
| Healthy weight: n = 16 | Healthy weight: 47.75 (10.44) | Healthy weight: 24.10 (2.07) | ||||||
| Tuulari et al., 2015 | Cross-sectional | FC during craving regulation | PPI analysis | Obese: n = 27 | n/a | Obese: 42.1 (9.3) | Obese: 41.4 (3.9) | Obesity was associated with higher FC in the executive control network during craving regulation. |
| Healthy weight: n = 14 | Healthy weight: 44.9 (11.9) | Healthy weight: 22.6 (2.7) | ||||||
| Legget et al., 2016 | Longitudinal | Effects of a 6 month exercise program on resting-state FC | BNC with ICA | N = 11 | 9% (N = 1 did not complete post testing) | 33.6 (1.4) | 38.2 (3.2) | BNC in the posterior cingulate cortex was reduced following chronic exercise; change in BNC was related to changes in aerobic fitness level and perceived hunger. |
| Lepping et al., 2015 | Longitudinal | Resting-state FC after surgical and behavioral weight loss | Seed-based resting-state FC using mPFC, precuneus and inferior temporal gyrus | Bariatric surgery: n = 15 | 13% (N = 2 did not have useable follow-up data or failed to complete follow-up testing) | Bariatric surgery: 42 (10.35) | Bariatric surgery: 41.35 (1.97) | Following weight loss, behavioral dieters exhibited increased FC between left precuneus/superior parietal lobule and bilateral insula pre- to post-meal and bariatric patients exhibited decreased FC between these regions pre- to post-meal. |
| Behavioral diet: n = 13 | Behavioral diet: 40.23 (8.01) | Behavioral diet: 40.10 (1.8) | ||||||
| Li et al., 2018 | Longitudinal | Resting-state FC among individuals with obesity who were awaiting bariatric surgery and a subset of that group who completed surgery | Resting-state FC between 90 regions defined using the AAL atlas | Preoperative: n = 29 | 24% (N = 7 surgery patients dropped out do to travel restrictions) | Preoperative: 27.8 (6.9) | Preoperative: 40.0 (6.5) | Pre-operative individuals exhibited increased FC in orbitofrontal cortex, MFG and superior frontal gyrus. Post-surgery, these differences were no longer present. |
| Postoperative: n = 22 (subset of the preoperative group who underwent surgery) | Postoperative: Not described | Postoperative: 34.4 (5.9) | ||||||
| Healthy weight: n = 19 | Healthy weight: 26.7 (6.8) | Healthy weight: 21.8 (1.8) | ||||||
| Weygandt et al., 2013 | Longitudinal | FC during a dietary impulse healthy weight task as a predictor of weight loss | PPI using ventromedial prefrontal cortex as the seed | N = 16 | NR | 43.0 (12.2) | Pre-diet BMI: 34.5 (3.2) | Stronger FC between the dorsolateral and ventromedial prefrontal cortex was associated with better dietary success and impulse control. |
| Post-diet BMI: 30.2 (1.8) | ||||||||
| Contreras-Rodríguez et al., 2017 | Longitudinal | Relationship between obesity and resting-state FC | Seed-based resting-state FC using the nucleus accumbens as a seed | Obese: n = 42 | 28% (N = 11 did not return to follow-up in the obese group) | Obese: 33.59 (6.16) | Obese: 30.51 (3.63) | Participants with excess weight displayed increased FC between the ventral striatum and the medial prefrontal and parietal cortices and between the dorsal striatum and the somatosensory cortex. Dorsal striatum FC correlated with food craving and predicted BMI gains. |
| Healthy weight: n = 39 | Healthy weight: 33.07 (6.73) | Healthy weight: 22.09 (1.74) | ||||||
| McFadden et al., 2013 | Longitudinal | Effects of exercise on resting-state default mode and salience network activity in overweight/obese adults | ICA | N = 12 | 0% | 32.8 (9.5) | 33.3 (4.3) | The intervention was associated with a reduction in DMN activity in the precuneus, which was associated with greater fat mass loss as well as reduced perceived hunger and hunger ratings in response to a meal. No changes were observed in the salience network in response to the exercise intervention. |
| Tregellas et al., 2011 | Randomized cross-over design comparing fed and fasted states | Effect of weight loss on FC of the DMN during the presentation of high calorie food cues | ICA | Reduced-obese: n = 18 | NR | Reduced-obese: 35.2 (5.7) | Reduced-obese: 27.5 (2.6) | In the eucaloric state, greater activity among individuals who lost weight, compared to healthy weight individuals, was observed in the lateral inferior parietal and posterior cingulate cortices. Lateral parietal activity correlated positively with appetite. Overfeeding was associated with increased posterior cingulate default network activity in healthy weight individuals compared to those who lost weight. |
| Healthy weight: n = 24 | Healthy weight: 34.7 (5.4) | Healthy weight: 21.6 (1.7) | ||||||
| Hinkle et al., 2013 | Randomized cross-over design comparing placebo vs leptin | FC of the hypothalamus during reduced-weight maintenance and leptin repletion | PPI analysis using the right hypothalamus as the seed | N = 10 | NR | 36.8 (6.5) | Initial: 39.9 (8.2) | During reduced-weight maintenance with placebo injections, the FC of the hypothalamus increased with visual areas and the dorsal anterior cingulate in response to food cues. During reduced-weight maintenance with leptin injections, FC of the right hypothalamus increased with the mid-insula and the central and parietal operculae, suggesting increased coupling with the interoceptive system and decreased with the orbital frontal cortex, frontal pole and the dorsal anterior cingulate. |
| Following weight loss: 34.8 (7.0) | ||||||||
| Kahathuduwa et al., 2018 | RCT | Effect of two dietary manipulations on task-evoked FC | PPI analysis | Meal replacement (TMR) group: n = 16 | 12.5% (N = 4 did not complete study) | TMR: 31.27 (11.85) | TMR: 35.14 (3.75) | Compared to TD, TMR was also associated with negative modulation of FC of the nucleus accumbens, orbitofrontal cortex and amygdala by dorsolateral prefrontal cortex. |
| Typical diet (TD) group: n = 16 | TD: 32.15 (14.67) | TD: 34.82 (2.63) | ||||||
| Filbey and Yezhuvath, 2017 | RCT | Examine the relationship between BMI and successful inhibition during an inhibitory control task; BMI was analyzed continuously. | PPI analysis with IFG as the seed | N = 34 | 24% (N = 10 due to data quality; N = 1 due to BMI) | 32.6 (10.6) | 30.7 (6.3) | Positive correlations were found between BMI and impulsivity. Further, there was a positive association between BMI and FC between the right IFG and right middle frontal gyrus during successful response inhibition. |
| Mehl et al., 2019 | RCT | Task-evoked FC during cognitive bias modification training and resting-state FC after training | PPI analysis using the angular gyrus as a seed. Resting-state FC using dorsolateral and medial prefrontal cortex, amygdala, nucleus accumbens and MFG as seeds | Cognitive bias modification: n = 17 | NR | Cognitive bias modification: 28(5) | Cognitive bias modification: 35.57 (4.63) | Analysis of brain FC during rest revealed training-related FC changes of the inferior frontal gyrus and bilateral middle frontal gyri. |
| Healthy weight: n = 16 | Healthy weight: 31 (4) | Healthy weight: 6.95 (7.63) | ||||||
| *Significantly different |
Note. BNC, between-network FC; PPI, psychophysiological interaction; FC, functional connectivity; ICA, independent component analysis; RCT, randomized controlled trial; NR, not reported.