Hemispheric divergence of interoceptive processing across psychiatric disorders

Reviewer #1 (Public Review):

Summary:
The aim of this study is to explore the neurocircuitry of top-down and bottom-up interception, and how this differs in psychiatric disorders. Using functional neuroimaging, the research focuses on individuals with anxiety, depression, and/or eating disorders compared to healthy individuals. The findings highlight the dysgranular mid-insula as a key cortical area where attention and real-time bodily inputs converge, potentially serving as a disruption point in psychiatric disorders.

Strengths:
The authors used robust and validated methods to answer their research question efficiently. They illustrate a complete picture of the theoretical impact of the study and their own strengths and weaknesses.

Weaknesses:
One concern is regarding the experimental task design. Currently, only subjective reports of interoceptive intensity are taken into account, the addition of objective behavioural measures would have given additional value to the study and its impact.

This brings me to my second concern. The authors mostly refer to their own previous work, without highlighting other methods used in the field. Some tasks measure interoceptive accuracy or other behavioural outcomes, instead of merely subjective intensity. Expanding the scientific context would aid the understanding and integration of this study with the rest of the field.

Lastly, the suggestions for future research lack substance compared to the richness of the discussion.

Reviewer #2 (Public Review):

Summary:
The authors have conducted an exceptionally informative series of studies investigating the neural basis of interoception in transdiagnostic psychiatric symptoms. By comparing differential and overlapping neural activation during 'top-down' and 'bottom-up' interoceptive tasks, they reveal convergent activation largely localised to the ventral dysgranular subregion ('mid-insula'), which differs in extent between patients and controls, replicating and extending previous suggestions of this region as a central locus of disruption in psychiatric disorders. Their work also reveals different extents of divergent activation in the anterior insula during anticipation of interoceptive disruption. This substantially advances our previous knowledge of the anatomy of interoception and confirms theoretical predictions of the roles of different cytoarchitectural subregions of the insula in interoceptive dysfunction in mental health conditions.

Strengths:
The work is exceptional in terms of breadth and depth, making use of multiple imaging and analysis techniques which are non-standard and go well beyond what is known today. The study is statistically well-powered and the tasks are well-validated in the literature. To my knowledge, these functions of the insula in interoception and mental health have never been compared directly before, so the results are novel and informative for both basic science and psychiatry. The work is strongly theory-driven, building on and directly testing results from influential theories and previous studies. It is likely that the results will strengthen our theoretical models of interoception and advance psychiatric studies of the insula.

Weaknesses:
The study has three current limitations. (1) The interpretation of the resting-state data is not quite as clear-cut as the task-based data - as presented currently, changes could potentially represent fluctuations over time rather than following interoception specifically. In contrast, much stronger conclusions can be drawn from the authors' task-based data. (2) The transdiagnostic sample could be better characterised in terms of diagnostic information, and was almost entirely female; it is also unclear what the effect of psychotropic medications may have been on the results given the effects of (e.g.) serotonergic medication on the BOLD signal. (3) As the authors point out, there may have been task-specific preprocessing/analysis differences that influenced results, for example, due to physiological correction in one but not both tasks.

Reviewer #3 (Public Review):

Summary:
Adamic and colleagues present fMRI data from ADE patients and a healthy control group acquired during two interoceptive tasks (attention and perturbation) from the same session. They report convergent activity within the granular and dysgranular insular cortex during both tasks, with a patient group-specific lateralisation effect. Furthermore, insular functional connectivity was found to be linked to disease severity.

Strengths:
The study is well-designed and - despite some limitations noted by the authors - provides much-needed insight into the functional pathways of interoceptive processing in health and disease. The manuscript is clear, concise, and well-written so that I only have a few comments I would mostly regard as minor points.

Weaknesses:
There are a few instances where it is not entirely clear whether the authors' claims are fully supported by the underlying statistics.

Author Response

We appreciate the thoughtful comments from the reviewers. All reviewers express common support for the study’s meaningful contribution to understanding interoceptive neurocircuitry in health and in psychiatric disorders. Specifically, the reviewers highlight the strong theoretical backing and the novel combination of tasks and analytical methods. In turn, the reviewers identify several areas for improvement that we plan to address in our resubmission. These include a more detailed demographic characterization of the study participants, increased clarity when describing the statistics that support each conclusion, and additional discussion when interpreting the resting state findings, as we did not include a separate control condition for the effect of time. One reviewer commented that we largely cite our previous work with the isoproterenol paradigm; while we will provide an updated and broader view of the literature in our resubmission, there remains a limited number of comparable interoceptive perturbation studies. Finally, one comment referred to our reliance on ratings of interoceptive intensity without included additional behavioral measures. While our measures of interest were chosen for their relevance to our hypotheses, we will consider adding additional measures such as interoceptive accuracy (correspondence between heart rate and dial ratings) that were collected during the perturbation task, should they provide additional insight into the insular responses of the participants.