Functional connectivity predictors and mechanisms of symptom change in functional neurological disorder 2026 Westlin et al

[Andy]

Abstract​

Clinical trajectories in patients with functional neurological disorder (FND) are variable, and the neural mechanisms underlying this heterogeneity remain poorly understood. This longitudinal brain imaging study examined resting-state functional connectivity predictors and mechanisms of symptom change in FND.

Thirty-two adults with FND (motor and/or seizure phenotypes) completed baseline questionnaires and functional MRI (fMRI), followed by naturalistic treatment for 6.8 ± 0.8 months. All participants completed follow-up questionnaires; 28 completed follow-up fMRI. At each timepoint, three graph-theory network metrics of resting-state functional connectivity were computed: whole-brain weighted-degree (centrality), cortical integration (between-network connectivity), and cortical segregation (within-network connectivity). All analyses adjusted for age, sex, antidepressants, head motion, time between sessions and baseline score of interest, with cluster-wise correction. Results were contextualized against 50 age-, sex-, and head motion-matched healthy controls (HCs).

Based on patient-reported Clinical Global Impression of Improvement ratings, 59.4% improved, 31.3% were unchanged, and 9.3% worsened. Core FND symptom (i.e. Screening for Somatoform Symptoms-7 Subscale for Conversion Disorder) and non-core physical symptom (Patient Health Questionnaire-15) scores showed variable trajectories, with no group-level changes.

For whole-brain weighted-degree analyses, baseline centrality in right middle frontal, precentral, and left cerebellar regions was positively associated with core FND symptom change; longitudinally, centrality decreases in right precentral, superior parietal, lateral occipital, and cerebellar regions were associated with symptom improvement.

For cortical integration analyses, baseline between-network connectivity in ventral attention, frontoparietal, and default mode network regions was positively associated with core FND symptom change; longitudinally, decreases in between-network connectivity for regions of these same networks were associated with symptom improvement.

For cortical segregation analyses, baseline within-network connectivity in frontoparietal network regions was positively associated with core FND symptom change; no regions showed longitudinal segregation changes associated with symptom change.

The right anterior insula emerged as a convergent site across baseline and longitudinal integration analyses, with the most improved participants showing elevated baseline between-network connectivity relative to HCs that normalized at follow-up. More modest functional connectivity associations were observed with non-core physical symptom change, spanning baseline within-network connectivity in dorsal attention network regions and longitudinal between-network connectivity increases in visual network regions. Findings remained significant adjusting for FND phenotype, although several attenuated when accounting for baseline affective symptoms or trauma burden.

In conclusion, this study identified baseline and longitudinal resting-state functional connectivity features linked to symptom change in FND, highlighting the potential of large-scale network interactions as prognostic markers and providing mechanistic insights that set the stage for novel, biologically informed interventions.

Open access

 

[Sean]

So where is the 'functional' bit among all these objective brain features?

Clinical trajectories in patients with functional neurological disorder (FND) are variable, and the neural mechanisms underlying this heterogeneity remain poorly understood.​

Or it remains so poorly conceptualised and characterised that it is not at all clear the proposed FND entity even exists in any meaningful useful sense, beyond being a convenient god of the gaps and a source of employment and status for FND advocates.

What is clear is that the FND advocates are simply going to label whatever findings they get, positive or negative, as evidence for the FND concept, regardless of what the findings actually show or don't show.

 

[rvallee]

Based on CGI-I scores, 5 participants (15.6%) reported being ‘much improved’, 14 (43.8%) ‘improved’, 10 (31.3%) ‘unchanged’, and 3 (9.3%) ‘worse’

They don't include any numbers, only those categories, so it's pretty much guaranteed that the 'improvements' were trivial, and most likely unrelated, as they would otherwise have emphasized them. CGI-I is very generic and about the most susceptible to suggestion ("if you think of yourself as better, you will become better, now how well do you think you are?") you can imagine.

No baseline demographic or psychometric variables were associated with symptom improvement

Ah, well, that's the entire basis of their models. They keep insisting about those multidimensional holistic factors, and yet they never actually support anything. It's always only ever "humans, probably fee-male (said with a spitting Ferengi accent), between the age of 0 and 99, alive at the moment, experiencing symptoms which we decided are generated by tiny ghosts/beliefs/trauma in their brains".

Compared to longitudinal cohort participants (n = 32), those lost to follow-up (n = 20)

That's a whole lot of dropouts. This is a big tell, considering how loose the whole thing was, participants seem to have simply self-reported participation in some sort of generic psychological program ("Among the 32 participants, 30 (94%) received psychotherapy, with 21 (65.6%) specifically reporting receipt of CBT. Seventeen participants (53%) reported receiving physiotherapy, 16 (50%) endorsed receiving occupational therapy, and 10 (31%) reported receiving speech and language therapy. Eleven participants (34.4%) had all FND care within our healthcare system; the remainder either received a combination of treatment within our healthcare system and the community (n = 16, 50.0%) or entirely in the community (n = 5, 15.6%), limiting our ability to objectively capture treatment-related frequency and duration information").

As is tradition, they made no effort to gather information on reasons for dropping out. None ever does. Somehow the abstract seems to suggest that there were no dropouts:

All participants completed follow-up questionnaires

Core FND symptom (i.e. Screening for Somatoform Symptoms-7 Subscale for Conversion Disorder) and non-core physical symptom (Patient Health Questionnaire-15) scores showed variable trajectories, with no group-level changes.

So, the "treatment" is associated with participants reporting improvement, despite not improving, given that the entire concept is defined as those symptoms. Fascinating.

Bottom-tier quality slop, as always.