Table 2.
Summary of studies investigating the relationship between FC, disordered eating and diet
| Author | Design | Variable of interest | FC technique | N | Attrition | Mean age (SD) | Mean BMI (SD) | Findings |
|---|---|---|---|---|---|---|---|---|
| Chen et al., 2016 | Cross-sectional | Eating disorder pathology and FC of the dorsolateral prefrontal cortex | Seed-based resting-state FC; voxel-mirrored homotopic resting-state FC | Restrained eaters: n = 23; healthy controls: n = 24 | n/a | Restrained eaters: 20.74 (1.51); healthy controls: 21.04 (1.85) | Restrained eaters: 21.09 (2.38); controls: 21.01 (2.50) | Restrained eaters exhibited reduced interhemispheric connectivity of the dorsolateral prefrontal cortex, which was associated with higher BN symptoms. Increased dorsolateral prefrontal cortex resting-state connectivity was associated with BN pathology but only among restrained eaters. |
| Lee et al., 2014 | Cross-sectional | The relationship between eating disorder diagnosis and resting-state FC of the dorsal anterior cingulate cortex | Seed-based resting-state FC of the dorsal anterior cingulate cortex | AN: n = 18; BN: n = 20; healthy controls: n = 20 | n/a | AN: 25.2 (4.2); BN: 22.9 (3.9); healthy controls: 23.3 (1.8) | AN: 16.0 (1.7); BN: 21.6 (2.3); healthy controls: 19.9 (1.9) | AN group exhibited stronger synchronous activity between the dorsal anterior cingulate cortex and retrosplenial cortex, whereas the BN group showed stronger synchronous activity between the dorsal anterior cingulate cortex and medial orbitofrontal cortex. Both groups demonstrated stronger synchronous activity between the dorsal anterior cingulate cortex and precuneus, which correlated with higher scores of the Body Shape Questionnaire. |
| Lavagnino et al., 2014 | Cross-sectional | The relationship between BN diagnosis and FC of salience, executive control, somatosensory, DMNs | Seed-based resting-state FC using a single node from each network of interest as a seed | BN: n = 16; healthy controls: n = 18 | n/a | BN: 23 (5); healthy controls: 23 (3) | BN: 21 (2); healthy controls: 22 (2) | BN group showed a decreased resting FC both within the somatosensory network and with posterior cingulate cortex and two visual areas (the right middle occipital gyrus and the right cuneus). |
| Wang et al., 2017 | Cross-sectional | Relationship between resting-state connectivity assessed via graph theory and BN diagnosis | Resting-state connectivity assessed via graph theory | BN: n = 44; healthy controls: n = 44 | n/a | BN: 22 (3.4); healthy controls: 23.1 (3.4) | BN: 21.0 (2.6); healthy controls: 20.5 (1.4) | Nodal strength in BN was higher in the sensorimotor and visual regions as well as the precuneus, but lower in several subcortical regions, such as the hippocampus, parahippocampal gyrus and orbitofrontal cortex. BN group also showed hypoconnectivity involving subcortical limbic and paralimbic regions, which correlated significantly with scores of bulimia and drive for thinness. |
| Stopyra et al., 2019 | Cross-sectional | Relationship between resting-state connectivity and eating disorder diagnosis | Seed-based resting-state connectivity | BED: n = 27; BN: n = 29; healthy weight controls n = 29; overweight controls: n = 28 | n/a | BED: 38.39 (13.06); BN: 27.45 (10.55); healthy weight controls: 26.86 (6.59); overweight controls: 39.40 (10.48) | BED: 32.64 (4.13); BN: 21.33 (2.99); healthy weight controls: 21.85 (1.80); overweight controls: 33.58 (4.54) | ED patients exhibited aberrant FC in the dorsal anterior cingulate cortex within the salience network, as well as in the medial prefrontal cortex. Furthermore, BED and BN groups differed from each other in FC within each network. Seed-based analysis revealed stronger synchronous dorsal anterior cingulate cortex-retrosplenial cortex activity in the BN group. |
| Canna et al., 2017 | Cross-sectional | Resting-state VMHC | VMHC | AN: n = 15; BN: n = 13; healthy controls: n = 16 | n/a | AN: 25.3 (1.6); BN: 27.2 (2.0); healthy controls: 26.1 (3.5) | AN: 16.8 (1.6); BN: not reported; healthy controls: 21.1 (1.6) | Compared to HC, AN patients exhibited reduced VMHC in cerebellum, insula and precuneus, while BN patients showed reduced VMHC in dorsolateral prefrontal and orbitofrontal cortices. |
| Frank et al., 2016 | Cross-sectional | Effective connectivity between regions involved in energy homeostasis and reward during the ingestion of a sucrose solution among individuals with an ED diagnosis | Effective connectivity | AN: n = 26; BN: n = 25; healthy controls: n = 26 | n/a | AN: 23.23 (5.26); BN: 24.64 (3.22); healthy controls: 24.39 (3.49) | AN: 16.23 (1.09); BN: 23.56 (5.89); healthy controls: 21.61 (1.21) | Only the controls had an effective connectivity pattern from the hypothalamus to ventral striatum bilaterally. On the right side both eating-disorder groups showed effective connectivity from the anterior cingulate to ventral striatum and from there to the hypothalamus. Both AN and BN showed effective connectivity on the left from ventral anterior insula to inferior orbitofrontal cortex, middle to inferior orbitofrontal cortex and dorsal to ventral anterior insula. |
| Geliebter et al., 2016 | Cross-sectional | Task-evoked FC of the dorsal anterior cingulate cortex in response to food cues among individuals who binge eat | Seed-based connectivity, PPI | Binge eating: n = 10; healthy controls: n = 10 | n/a | Binge eating: 22.1 (2.3); healthy controls: 21.3 (0.6) | Binge eating: 27.4 (5.8); healthy controls: 27.7 (7.2) | In response to high-calorie (vs low-calorie) food cues, binge eaters exhibited greater connectivity between dorsal anterior cingulate cortex and insula, cerebellum and supramarginal gyrus. None of these effects were significant at the whole-brain level. No effects of body weight were observed. |
| Kim et al., 2012 | Cross-sectional | Task-evoked functional and effective connectivity of the anterior insula during the processing of food cues | Seed-based connectivity, PPI with anterior insula as the region on interest | AN: n = 18; BN: n = 20; healthy controls: n = 20 | n/a | AN: 25.2 (4.2); BN: 22.9 (3.9); healthy controls: 23.3 (1.8) | AN: 16.0 (1.7); BN: 21.6 (2.3); healthy controls: 19.9 (1.9) | In response to food images compared to non-food images, both the AN group and BN group demonstrated increased activity in the left anterior insula. In the AN group, the left anterior insula demonstrated significant interactions with the right insula and right IFG. In the BN group, the left anterior insula demonstrated significant interactions with the medial orbitofrontal cortex. |
| Bohon and Stice, 2012 | Cross-sectional | Task-evoked connectivity during anticipation and receipt of a chocolate milkshake | Seed-based connectivity, PPI with amygdala as the ROI | Subthreshold BN: n = 13; healthy controls: n = 13 | n/a | 20.3 (1.87); subgroup means not reported | BN: 23.93 (2.82); healthy controls: 23.19 (2.42) | Connectivity analyses revealed a greater relation of amygdala activity to activation in the left putamen and insula during anticipated receipt of milkshake in the BN group relative to the control group. The opposite pattern was found for the taste of milkshake. |
| Neveu et al., 2018 | Cross-sectional | FC of dorsolateral prefrontal cortex during a food choice task | Seed-based PPI with dorsolateral prefrontal cortex as the ROI | BN: n = 35; healthy controls: n = 26 | n/a | BN: 24 (3.87); healthy controls: 23 (2.70) | BN: 19.9 (2.15); healthy controls: 21.3 (2.36) | BN patients chose unhealthy food more often. FC analysis showed that the activity in the dorsolateral prefrontal cortex was coupled with ventromedial prefrontal cortex activity in uncontrolled food choices. |
| Diet and FC studies (humans) | ||||||||
| Garcia-Casares et al., 2017 | Longitudinal intervention (no control group) | Resting-state seed-based connectivity | Seed-based | N = 19 | 15.8% (N = 3 did not complete follow-up) | 46.31 (4.07) | 38.15 (4.7) | After the intervention, there was decreased FC between the left inferior parietal cortex and the right temporal cortex, and bilateral posterior cingulate; decreased connectivity between the left superior frontal gyrus and the right temporal cortex; decreased connectivity between the prefrontal cortex and the somatosensory cortex; and decreased connectivity between the left and right posterior cingulate. |
| Talukdar et al., 2019 | Cross-sectional | Whole-brain connectivity | Multivariate distance-based matrix regression | N = 96 | n/a | 69.0 (3.0) | 26.0 (4.0) | Omega 3 levels were associated with individual differences in FC within regions that support executive function (prefrontal cortex), memory (hippocampus) and emotion (amygdala). |
Note. ED, eating disorder; PPI, psychophysiological interaction; FC, functional connectivity; VMHC, voxel-mirrored homotopic connectivity.