A Chronic Fatigue Syndrome (CFS) severity score based on case designation criteria.
James N Baraniuk, Oluwatoyin Adewuyi, Samantha Jean Merck, Mushtaq Ali, Murugan K Ravindran, Christian R Timbol, Rakib Rayhan, Yin Zheng, Uyenphuong Le, Rania Esteitie, Kristina N Petrie
Author Information
James N Baraniuk: Division of Rheumatology, Immunology and Allergy, Georgetown University Washington, DC, USA.
BACKGROUND: Chronic Fatigue Syndrome case designation criteria are scored as physicians' subjective, nominal interpretations of patient fatigue, pain (headaches, myalgia, arthralgia, sore throat and lymph nodes), cognitive dysfunction, sleep and exertional exhaustion. METHODS: Subjects self-reported symptoms using an anchored ordinal scale of 0 (no symptom), 1 (trivial complaints), 2 (mild), 3 (moderate), and 4 (severe). Fatigue of 3 or 4 distinguished "Fatigued" from "Not Fatigued" subjects. The sum of the 8(Sum8) ancillary criteria was tested as a proxy for fatigue. All subjects had history and physical examinations to exclude medical fatigue, and ensure categorization as healthy or CFS subjects. RESULTS: Fatigued subjects were divided into CFS with ≥4 symptoms or Chronic Idiopathic Fatigue (CIF) with ≤3 symptoms. ROC of Sum8 for CFS and Not Fatigued subjects generated a threshold of 14 (specificity=0.934; sensitivity=0.928). CFS (n=256) and CIF (n=55) criteria were refined to include Sum8≥14 and ≤13, respectively. Not Fatigued subjects had highly skewed Sum8 responses. Healthy Controls (HC; n=269) were defined by fatigue≤2 and Sum8≤13. Those with Sum8≥14 were defined as CFS-Like With Insufficient Fatigue Syndrome (CFSLWIFS; n=20). Sum8 and Fatigue were highly correlated (R(2)=0.977; Cronbach's alpha=0.924) indicating an intimate relationship between symptom constructs. Cluster analysis suggested 4 clades each in CFS and HC. Translational utility was inferred from the clustering of proteomics from cerebrospinal fluid. CONCLUSIONS: Plotting Fatigue severity versus Sum8 produced an internally consistent classifying system. This is a necessary step for translating symptom profiles into fatigue phenotypes and their pathophysiological mechanisms.
