Brain structural networks and connectomes: the brain–obesity interface and its impact on mental health

Figures & data

Table 1 Demographic and clinical characteristics

	Obesity	Control	P-value
Age (years)	33.1±8.9	45.0±9.7	<0.001a
Age range	23–54	22–58
Gender (male/female)	8/12	3/27
Education (years)	14.1±2.3	14.3±3.0	>0.05
BMI	37.9±5.2	22.6±3.4	<0.001a
Anxiety (of HADS)	6.7±3.1	3.7±2.9	<0.001a
Depression (of HADS)	6.5±3.7	2.9±2.8	<0.001a

Notes: Statistical data are shown as the means ± SD.

a Significant difference.

Abbreviations: BMI, body mass index; HADS, Hospital Anxiety and Depression Scale.

Figure 1 The significant differences of DTI indices between obese subjects and non-obese HC.

Notes: Significantly (A) lower AD of the posterior limb of the internal capsule and (B) lower RD of the right superior corona radiata were found in obese subjects compared with non-obese HC with significantly lower MD of (C) the right posterior limb of the internal capsule and (D) right superior longitudinal fasciculus. The color bar represents the t-score (P<0.05, corrected by the FDR).
Abbreviations: AD, axial diffusivity; DTI, diffusion tensor imaging; FDR, false discovery rate; HC, healthy controls; MD, mean diffusivity; RD, radial diffusivity.

Figure 2 The correlation between DTI indices, BMI, and HADS scores.

Notes: A significant negative correlation between BMI and (A) AD in the posterior limb of the internal capsule, (B) RD in the corpus callosum were found. A significant negative correlation between anxiety scores and (C) AD in the corpus callosum, (D) AD in the superior frontal blade, (E) RD in the superior frontal blade were observed. A significant negative correlation between depression scores and (F) AD in the corpus callosum, (G) RD in the right sagittal stratum were identified. A significant positive correlation between depression scores and (H) MD in the right sagittal stratum was also found. The color bar represents the t-score (P<0.05, corrected by FDR).
Abbreviations: AD, axial diffusivity; BMI, body mass index; DTI, diffusion tensor imaging; FDR, false discovery rate; HADS, Hospital Anxiety and Depression Scale; MD, mean diffusivity; RD, radial diffusivity.

Figure 3 The significant differences of GQI indices between obese subjects and non-obese HC.

Notes: Significantly lower GFA of the (A) left posterior limb of the internal capsule and (B) right posterior corona radiata were observed in obese subjects when compared to non-obese HC. (C) Significantly lower NQA of the lentiform nucleus was also observed. The color bar represents the t-score (P<0.05, corrected by FDR).
Abbreviations: AD, axial diffusivity; DTI, diffusion tensor imaging; GQI, generalized q-sampling imaging; HC, healthy controls; MD, mean diffusivity.

Figure 4 The correlation between GQI indices, BMI, and HADS scores.

Notes: (A) A significant negative correlation between BMI and GFA of the posterior limb of the internal capsule was found, as was a significant negative correlation between BMI and NQA values in the (B) corpus callosum and (C) superior frontal blade. A significant negative correlation between anxiety scores and GFA values in the (D) posterior cingulate, (E) lentiform nucleus, and (F) precuneus were found, (G) as well as a significant negative correlation between anxiety scores and NQA values in the left superior corona radiata. (H) A significant negative correlation between depression scores and GFA values in the precuneus was observed in addition to a significant negative correlation between depression scores and NQA values in the (I) the posterior cingulate and (J) lingual gyrus. The color bar represents the t-score (P<0.05, corrected by FDR).
Abbreviations: BMI, body mass index; FDR, false discovery rate; GFA, generalized fractional anisotropy; GQI, generalized q-sampling imaging; HADS, Hospital Anxiety and Depression Scale; NQA, normalized quantitative anisotropy.

Figure 5 The degree of distribution in the structural network of (A) non-obese healthy controls and (B) obese subjects by graph theoretical analysis with GQI. (C–F) The topological measures of structural networks in obese subjects and non-obese HC by graph theoretical analysis with GQI (P<0.05, corrected by FDR).

Abbreviations: FDR, false discovery rate; GQI, generalized q-sampling imaging; HC, healthy controls.

Figure 6 The disrupted sub-networks of obese subjects identified by the NBS with GQI (P<0.001, corrected by FDR).

Abbreviations: FDR, false discovery rate; GQI, generalized q-sampling imaging; NBS, network-based statistical analysis.