Data-Driven Path Analytic Modeling to Understand Underlying Mechanisms in ... Long-Term Post-COVID Pain: 2022 Fernández-de-las-Peñas et al

Full title: Data-Driven Path Analytic Modeling to Understand Underlying Mechanisms in COVID-19 Survivors Suffering from Long-Term Post-COVID Pain: A Spanish Cohort Study

Abstract

Pain can be present in up to 50% of people with post-COVID-19 condition. Understanding the complexity of post-COVID pain can help with better phenotyping of this post-COVID symptom. The aim of this study is to describe the complex associations between sensory-related, psychological, and cognitive variables in previously hospitalized COVID-19 survivors with post-COVID pain, recruited from three hospitals in Madrid (Spain) by using data-driven path analytic modeling.

Demographic (i.e., age, height, and weight), sensory-related (intensity or duration of pain, central sensitization-associated symptoms, and neuropathic pain features), psychological (anxiety and depressive levels, and sleep quality), and cognitive (catastrophizing and kinesiophobia) variables were collected in a sample of 149 subjects with post-COVID pain.

A Bayesian network was used for structural learning, and the structural model was fitted using structural equation modeling (SEM). The SEM model fit was excellent: RMSEA < 0.001, CFI = 1.000, SRMR = 0.063, and NNFI = 1.008. The only significant predictor of post-COVID pain was the level of depressive symptoms. Higher levels of anxiety were associated with greater central sensitization-associated symptoms by a magnitude of >β=0.406 (p = 0.008). Males reported less severe neuropathic pain symptoms (−1.50 SD S-LANSS score, p < 0.001) than females. A higher level of depressive symptoms was associated with worse sleep quality (>β=0.406, p < 0.001), and greater levels of catastrophizing (>β=0.345, p < 0.001).

This study presents a model for post-COVID pain where psychological factors were related to central sensitization-associated symptoms and sleep quality. Further, maladaptive cognitions, such as catastrophizing, were also associated with depression. Finally, females reported more neuropathic pain features than males. Our data-driven model could be leveraged in clinical trials investigating treatment approaches in COVID-19 survivors with post-COVID pain and can represent a first step for the development of a theoretical/conceptual framework for post-COVID pain.

Open access, https://www.mdpi.com/2076-0817/11/11/1336

 

When your "predictor" happens after the fact, is obviously a consequence, and no one finds anything wrong with it, you have so many problems to deal with that even having a predictor that happens after the fact is the least of your problems. Like having overlapping sick-enough-to-enter and recovered-enough-to-be-recorded-as-such thresholds for recovery outcomes, so many things had to be wrong to even get there.

Especially when there is massive overlap between all the concepts, while pretending otherwise. Frankly this is all becoming even worse than eminence-based medicine, this is conclusion-based medicine, where a model of the outcome is made-up and everything simply derives from inevitably leading to it, a fully teleological ideology. The exact same as every iteration before.

 

Thanks for watching today's episode of "assuming the causality is in the direction that supports your hypothesis."