Increased connectivity in several bilateral frontal and fronto‐parietal networks predicts depressive symptoms in mid‐ to late‐life diabetics: Neuroimaging / differential diagnosis. (7th December 2020)

• Record Type: Journal Article
• Authors: Salardini, Arash; Shen, Xilin; Hashemi‐Aghdam, Arsalan; Laltoo, Emily; Savoia, Sarah; Tokoglu, Fuyuze; Constable, Todd
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.043619
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Depressive symptoms may be a feature of vascular cognitive impairment secondary to diabetes. There is significant biotypic diversity amongst individuals with depressive symptoms depending on the causes and symptoms. In this study we determined the brain networks which are best correlated with depressive traits in our cohort of diabetic individuals. Method: 49 diabetic patients were recruited and underwent comprehensive neuropsychological testing as well as MRI imaging including structural MRI and rs‐fMRI. A linear regression model was built on a general functional connectivity matrix, derived from the rs‐fMRI data, to predict depression symptoms as measured by Beck's Depression Inventory (BDI). The leave‐one‐out cross validation method was used to test the model. Results: CPM was applied to imaging from 49 individuals aged 55‐90 (Mean 69, SD 8) with a diagnosis of diabetes (Mean HbA1c = 6.88, SD 1) with Fazekas scores of 0‐3 (mean 1.04, SD 0.78). BDI scores ranged 0‐20 (mean 6.12, SD 4.63) while MOCA scores spanned 13‐30 (mean 25.14, SD 3.51). A positive predictive model was discovered (R2 =0.3514, p=0.0143) which contained increased connectivity in several frontal and frontoparietal networks. Specifically, stronger connections between bilateral ventromedial prefrontal cortices and dorsal prefrontal cortices (L > R) appeared to predict higher BDI scores. No statistically significant negative predictive model (where weaker connections predicted higher BDI scores) were found. Conclusion: In this study we use connectome‐based predictive modelling (CPM) to find connections which may predict the presence of depressive symptoms in diabetic individuals. Unlike most connectome‐based techniques which determine connections which have an association with behavioral measures, CPM finds networks which likely predict behavioral scores. Our findings are consistent with what we know about the neural correlates of depression but represent the first hypothesis‐free connectivity study of depressive symptoms in mid to late‐life diabetes.