Gene Expression Factor Analysis to Differentiate Pathways Linked to Fibromyalgia, Chronic Fatigue Syndrome, and Depression in a Diverse Patient Sample

Abstract

Objective To determine if independent candidate genes can be grouped into meaningful biological factors and if these factors are associated with the diagnosis of chronic fatigue syndrome (CFS) and fibromyalgia (FMS) while controlling for co-morbid depression, sex, and age. Methods We included leukocyte mRNA gene expression from a total of 261 individuals including healthy controls (n=61), patients with FMS only (n=15), CFS only (n=33), co-morbid CFS and FMS (n=79), and medication-resistant (n=42) or medication-responsive (n=31) depression. We used Exploratory Factor Analysis (EFA) on 34 candidate genes to determine factor scores and regression analysis to examine if these factors were associated with specific diagnoses. Results EFA resulted in four independent factors with minimal overlap of genes between factors explaining 51% of the variance. We labeled these factors by function as: 1) Purinergic and cellular modulators; 2) Neuronal growth and immune function; 3) Nociception and stress mediators; 4) Energy and mitochondrial function. Regression analysis predicting these biological factors using FMS, CFS, depression severity, age, and sex revealed that greater expression in Factors 1 and 3 was positively associated with CFS and negatively associated with depression severity (QIDS score), but not associated with FMS. Conclusion Expression of candidate genes can be grouped into meaningful clusters, and CFS and depression are associated with the same 2 clusters but in opposite directions when controlling for co-morbid FMS. Given high co-morbid disease and interrelationships between biomarkers, EFA may help determine patient subgroups in this population based on gene expression.

Percolation transition of cooperative mutational effects in colorectal tumorigenesis

“…By propagating all of the mutations of each colorectal cancer patient, we obtained a gene list included in the corresponding GC and estimated the enrichment of the hallmark gene sets in each patient through the hypergeometric test. To reduce the dimensionality of the resulting matrix of the hallmark gene set enrichment test (191 patients × 50 hallmark gene sets), we did factor analysis, which is often used in gene expression data for patient clustering as a robust feature selection method 26 , 27 using standardized z -scores of −log( p -value) for each hallmark gene set because the hallmark gene set variables have different scales. For an optimal predefined factor number ( k = 5) (Supplementary Fig.…”

Using Clinical Characteristics and Patient‐Reported Outcome Measures to Categorize Systemic Lupus Erythematosus Subtypes

“…In trying to understand the etiology of fibromyalgia, some investigators have suggested that this condition results from a maladaptive reaction to chronic illness, possibly in patients with a genetic predisposition exacerbated by environmental or social stressors (37,38). Other researchers propose that fibromyalgia may have a unique inflammatory or neuroinflammatory etiology based on heightened levels of substance P, interleukin (IL)‐8, IL‐6, and tumor necrosis factor, and based on the identification of unique gene signatures, including those associated with immune system regulation and with glutamine, purinergic, nociception, and mitochondrial domains (39–42). Further studies are needed, specifically in patients with SLE with fibromyalgia to better understand the underlying pathobiology in this subgroup.…”

Chronic fatigue syndrome in the emergency department

“…The criteria for CFS have similarities with the ancillary criteria of depression and somatoform disorders including fatigue, lack of energy, poor concentration, insomnia, and reduced activity (Table 3). 53,54 This may be a superficial subjective oversimplification because major depressive disorder and CFS are distinct entities when their objective findings such as leukocyte mRNA expression patterns are contrasted 55,56. CFS is associated with plasma metabolomics changes indicating mitochondrial dysfunction and reduced utilization of amino acids for ATP production 32,57.…”