Session Information
Date: Sunday, November 8, 2015
Title: Fibromyalgia, Soft Tissue Disorders, Regional and Specific Clinical Pain Syndromes Poster I
Session Type: ACR Poster Session A
Session Time: 9:00AM-11:00AM
Background/Purpose:
Attempts to individuate subsets of fibromyalgia patients based on the aggregation of different symptoms have so far produced conflicting results. The 2010 preliminary diagnostic criteria for fibromyalgia (1), with the suggestion of the symptoms to be considered in the diagnosis, offers a frame for further investigation on this topic.
Methods:
338 patients (314 F; 24 M, mean age 45.47±11.37 yrs) diagnosed with fibromyalgia (FM) according to 2010 criteria were studied. FM was evaluated by tender points (TP), Widespread Pain Index (WPI), and Symptom Severity Scale (SS). The clustering in a two-dimentional space of the 41 somatic symptoms to be considered according to 210 criteria was assessed by Multidimentional Scaling analysis (MDS) by the ALSCAL routine of IBM SSPS 20.0 package, and the existence of an undelying classificative structure was explored by Latent Class Analysis (LCA) by the R poLCA package.
Results:
In the whole population, fibromyalgia scores were: TP 12.92±4.46, WPI 11.43±4.1 SS 9.04±2.10. Among somatic symptoms, muscle pain, muscle weakness, fatigue and thinking difficulties showed the highest similarity (Dice ≥ 0.8). Multidimensional scaling analysis was performed on the 41 somatic symptoms reported by the diagnostic criteria: the model was found to account for a substantial proportion of the variability of data (RSQ = 0.893); Kruskal stress was = 0.167. The Euclidean distance model (Fig. 1, left panel) showed at least 3 individualized grouping of symptoms: Group A symptoms, comprising muscle symptoms, fatigue and thinking difficulties, were prevalent in the majority (N = 225, 66.6%) of patients, while Group B (mainly digestive and urinary symptoms), and Group C (miscellaneous symptoms, including depression) symptoms were prevalent in 17.2% (N=58) and 16.3% (N=55) respectively. Group A patients reported the lowest total number of symptoms (16.8±5.4 vs 21.5±6.6 in group B and 23.0±7.9 in group C, p <0.001), and the lowest WPI (11.0±3.8 vs 11.5±4.5 and 14.1±4.6, p = 0.01); Group C showed also the highest TP count (15.1±2.9 vs 12.9±3.9 in group A and 12.0±5.0 in group B). Fatigue, waking unrefreshed, and cognitive symptoms intensity were not different in the 3 groups (p > 0.05).Latent Class Analysis by poLCA showed that modeling the data from 1 to 3 latent classes produced only a small improvement in the model estimate (AIC coefficient from 15538.65 to 14578.15).
Conclusion:
The results of our study are consistent with the hypothesis of the existence of separate subsets FM patients, characterized by different clusters of symptoms stratified on the common core of altered pain sensation. The definite characterization of these subsets requires further studies, and more complex modeling of data with consideration of covariates.
References: 1) Wolfe F, et al. Arthritis Care Res. 2010;62:600-10
To cite this abstract in AMA style:
Antivalle M, Battellino M, Batticciotto A, Ditto M, Mutti A, Varisco V, Rigamonti F, Atzeni F, Sarzi-Puttini P. Exploratory Analysis of Somatic Symptoms in Fibromyalgia By Multidimensional Scaling and Latent Class Analysis [abstract]. Arthritis Rheumatol. 2015; 67 (suppl 10). https://acrabstracts.org/abstract/exploratory-analysis-of-somatic-symptoms-in-fibromyalgia-by-multidimensional-scaling-and-latent-class-analysis/. Accessed .« Back to 2015 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/exploratory-analysis-of-somatic-symptoms-in-fibromyalgia-by-multidimensional-scaling-and-latent-class-analysis/