Mapping the network of persistent somatic symptoms across diseases: Cross-sectional & longitudinal analyses from the SOMACROSS research 2026 Strahl+

[Andy]

Highlights​

• Symptom clusters: gastrointestinal, musculoskeletal, cardiopulmonary, and general
• Fatigue and functional impairment are central symptom nodes across time points
• Global impairment and symptom severity are connected to individual somatic symptoms
• Symptom networks remained stable over a time period of six months
• Shared network structure in functional somatic/mental disorders and somatic disease

Abstract​

Background​

Persistent somatic symptoms (PSS) are common, functionally disruptive, and stable over time. Understanding symptom interactions may clarify transdiagnostic patterns and inform treatment. This study investigated symptom interrelations in a transdiagnostic sample to identify central symptoms and assess temporal stability.

Methods​

Patients (n = 1134; 63.5% female; mean age 50.6 ± 16.3 years) from the transdiagnostic SOMACROSS research unit were analysed regarding the Patient Health Questionnaire-15 (PHQ-15), the modified Pain Disability Index (PDI), and two items on symptom severity and impairment (EURONET-SOMA) at baseline and 6-month follow-up. Regularised partial correlation networks were compared between patients with functional somatic/mental disorders (n=316) and patients with somatic diseases (n=818), as well as across different measurements. Stability and invariance was tested with the Network Comparison Test (NCT).

Results​

Four symptom clusters emerged: gastrointestinal, musculoskeletal, cardiopulmonary, and general symptoms. Fatigue/low energy emerged as a central node linking domains. Global symptom severity and impairment (EURONET-SOMA) showed the highest connectivity. Disability in home responsibilities and recreation (PDI) bridged somatic symptoms and daily functioning. The NCT confirmed temporal stability (global strength: p = .815; structure invariance: p = .180), indicating that symptom clusters and central nodes remained consistent over time. The somatic diseases network showed significantly higher connectivity than the functional somatic/mental disorders network (global strength: p = .001), while structural invariance was not significant (p = .197).

Paywall

 

[rvallee]

So, sick people presenting with symptoms have symptoms, and those look similar to other people presenting with symptoms. Damn. Who knew?

It's almost like building models based entirely on the knowledge and subjective perception of professionals, rather than having anything to do with the patients, was a really absurdly inept and destructive to do. Almost.

Plus they created a machine that spits out as generic output the generic input they put in. Damn, that is suuuuper useful and genius and stuff. This is almost an improvement on the old reading tea leaves and stuff. Almost as smart as the very first version of ChatGPT, but not quite.

Also:

Four symptom clusters emerged: gastrointestinal, musculoskeletal, cardiopulmonary, and general symptoms

Since "the conversion disorder" is mainly focused on neurological symptoms, and so are functional models because they are essentially the same thing wearing a different hat, I guess they decided to arbitrarily relabel neurological as general? ¯\_(ツ)_/¯

Disability in home responsibilities and recreation (PDI) bridged somatic symptoms and daily functioning

Is there an actual legitimate meaning to their use of the word "bridged" here? Or it's just something that sounds sciencey enough but is just an arbitrary bullshit label?

Minor annoying is noticing how they keep using using transdiagnostic when they actually mean transdisciplinary. Again defined not based on patient properties but on properties relating entirely to them and their daily work.

They call this patient-centred medicine, because damn does someone here have a dark sense of humor, I guess.