Metabolic and structural connectivity within the default mode network relates to working memory performance in young healthy adults

•Metabolic covariance within midline DMN is greater in subjects with a better working memory.•This link seems to be subserved by structural integrity of the cingulate bundle.•PET-quantified metabolic covariance might be useful in studying cerebral networks.

Studies of functional connectivity suggest that the default mode network (DMN) might be relevant for cognitive functions. Here, we examined metabolic and structural connectivity between major DMN nodes, the posterior cingulate (PCC) and medial prefrontal cortex (MPFC), in relation to normal working memory (WM).DMN was captured using independent component analysis of [18F]fluorodeoxyglucose positron emission tomography (FDG-PET) data from 35 young healthy adults (27.1 ± 5.1 years). Metabolic connectivity, a correlation between FDG uptake in PCC and MPFC, was examined in groups of subjects with (relative to median) low (n = 18) and high (n = 17) performance on digit span backward test as an index of verbal WM. In addition, fiber tractography based on PCC and MPFC nodes as way points was performed in a subset of subjects.FDG uptake in the DMN nodes did not differ between high and low performers. However, significantly (p = 0.01) lower metabolic connectivity was found in the group of low performers. Furthermore, as compared to high performers, low performers showed lower density of the left superior cingulate bundle.Verbal WM performance is related to metabolic and structural connectivity within the DMN in young healthy adults. Metabolic connectivity as quantified with FDG-PET might be a sensitive marker of the normal variability in some cognitive functions.