Where alcohol use disorder meets interoception: A meta-analytic view on structural and functional neuroimaging data

Introduction: A growing body of research suggests that alcohol use disorder (AUD) involves a disruption in the processing of internal bodily signals, known as interoception. Changes in brain structure, particularly in the insula, are thought to underlie altered interoception. As studies specifically investigating this relationship are largely lacking, this analysis takes an approach that compares meta-analytic results on structural changes in AUD with meta-analytic results on neural representations of interoception. Methods: A systematic literature review was conducted to identify studies on functional representations of interoceptive processes. Activation likelihood estimation (ALE) was then used to test where the outcome clusters overlapped with the clusters of the recent ALE on grey matter reductions in AUD. Coactivation (MACM) and task-independent resting-state functional connectivity (rsFC) analyses were used to examine the functional network of interoception, and again to test where this network overlapped with the clusters of the AUD meta-analysis. The overlapping regions were functionally characterized using behavioral domain analysis. Results: 25 studies with a total of 647 healthy participants were included in the ALE analysis of interoception through a systematic literature search. The results showed a convergence cluster in the left middle insula. There was no overlap with clusters of reduced grey matter in AUD. Analysis of the co-activation patterns (MACM) of the cluster resulting from the interoception analysis revealed a large network located mainly in the insulae, thalami, basal nuclei, cingulate and medial frontal cortices, and pre- and postcentral gyri. Task-independent resting-state analysis of the interoception cluster confirmed this finding, showing the strongest connections to nodes of the salience and sensorimotor networks. Five of the eight clusters showing structural reductions in AUD were located in these regions. The behavioral profiles of these clusters also point to higher level processes such as salience control and sensorimotor functions and skin sensations. Conclusion: The results suggest altered salience mapping of interoceptive signals in AUD, consistent with recent models. Connections to the sensorimotor network may relate to action control and integration of skin sensations. Mindfulness-based interventions, pleasant touch and (deep) transcranial magnetic stimulation may be targeted interventions that can reduce interoceptive deficits in AUD and thus contribute to reduced use and relapse prevention.