Short-term changes in bodily pain and associated baseline factors in patients with fibromyalgia receiving paraprobiotic supplementation: a retrospective observational study

Abstract

Background: Fibromyalgia is characterized by heterogeneous pain trajectories, and short-term clinical improvement may vary across patients. We explored whether routinely collected baseline clinical variables were associated with short-term bodily pain improvement in a real-world cohort of patients with fibromyalgia receiving paraprobiotic supplementation. Methods: In this retrospective observational study, 86 women with fibromyalgia receiving paraprobiotic supplementation were followed for 2 months. Bodily Pain (BP) was assessed using the SF-36 at baseline, 1 month, and 2 months. The primary analysis used generalized estimating equations (GEE) with Gaussian family, identity link, and exchangeable working correlation structure, adjusted for age, years since diagnosis, body mass index (BMI), smoking status, hypertension, and dyslipidemia. Sensitivity analyses included alternative GEE working correlation structures, a linear mixed-effects model with patient-specific random intercept, and an ANCOVA model for BP at T2 adjusted for baseline BP. Results: In the primary analysis, time was significantly associated with SF-36 bodily pain (global Wald p=0.0018). Compared with baseline, BP scores increased non-significantly at T1 (β=1.92, 95% CI-0.39 to 4.23; p=0.103) and significantly at T2 (β=6.95, 95% CI 3.12 to 10.78; p

Background: Fibromyalgia is characterized by heterogeneous pain trajectories, and short-term clinical improvement may vary across patients. We explored whether routinely collected baseline clinical variables were associated with short-term bodily pain improvement in a real-world cohort of patients with fibromyalgia receiving paraprobiotic supplementation. Methods: In this retrospective observational study, 86 women with fibromyalgia receiving paraprobiotic supplementation were followed for 2 months. Bodily Pain (BP) was assessed using the SF-36 at baseline, 1 month, and 2 months. The primary analysis used generalized estimating equations (GEE) with Gaussian family, identity link, and exchangeable working correlation structure, adjusted for age, years since diagnosis, body mass index (BMI), smoking status, hypertension, and dyslipidemia. Sensitivity analyses included alternative GEE working correlation structures, a linear mixed-effects model with patient-specific random intercept, and an ANCOVA model for BP at T2 adjusted for baseline BP. Results: In the primary analysis, time was significantly associated with SF-36 bodily pain (global Wald p=0.0018). Compared with baseline, BP scores increased non-significantly at T1 (β=1.92, 95% CI-0.39 to 4.23; p=0.103) and significantly at T2 (β=6.95, 95% CI 3.12 to 10.78; p