Thesis Analysis of Cytokine Data in a Case-Control Study of ME/CFS, 2025, Amaratunga

Analysis of Cytokine Data in a Case-Control Study of ME/CFS

Navya Amaratunga

Background
Myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a complex, chronic, and debilitating multi-system disease characterized by profound fatigue, post-exertional malaise (PEM), unrefreshing sleep, cognitive dysfunction, and a range of other symptoms. The etiology of ME/CFS remains unknown, although research suggests potential involvement of immune system dysfunction, viral infections, genetic predisposition, and environmental factors. Currently, there is no cure for ME/CFS, and management focuses on symptom alleviation through strategies such as pacing, medication, and in some cases, graded exercise therapy or cognitive behavioral therapy. Diagnosis is based on clinical criteria, as there are no definitive biomarkers, and it is a lengthy process. For my thesis project, I worked on data previously collected by Dr. Roxana Moslehi as part of a case-control study of ME/CFS. My project involved conducting an analysis of cytokines to determine if there are statistically significant differences between cases and controls with respect to the levels of 48 cytokines.

Methods
The case-control study conducted by Dr. Moslehi involved a total of 61 controls and 60 cases. The levels of 48 cytokines were determined in each subject using the Human Cytokine 48-Plex Discovery Assay and measuring observed concentration of cytokines in serum. Statistical methods used were the T-test, Wilcoxon Rank Sum test, and logistic regression to calculate p-values, odds ratio (OR) estimates, and 95% confidence intervals (CI). Both unadjusted logistic regression and adjusted logistic regression were used to compare the cases and controls with respect to cytokines, adjusting for medication use during the 24 hours prior to blood draw as a potential confounder. A stratified analysis was done based on gender to compare the cases and controls with respect to the cytokines. Unadjusted logistic regression was carried out to analyze the main effect of medication on case-control status as well as the main effect of gender on case-control status.

Results
For all statistical analyses, we used a significance level of 0.05. In the T-test and Wilcoxon Rank Sum test, there is a statistically significant difference (p = 0.031) between cases of ME/CFS and controls with respect to mean Fractalkine levels. Other cytokines such as IL-6 (p = 0.046) and MIP-1α (p = 0.046) also showed borderline statistically significant differences between cases and controls with respect to their mean levels. In our unadjusted logistic regression analysis, there is evidence of a statistically significant association between lower Fractalkine levels and ME/CFS (OR = 1.422, 95% CI: 1.021 - 1.979; p = 0.037). After adjusting for medication use during the 24 hours prior to blood draw, there is a borderline significant association between Fractalkine (OR = 1.448, 95% CI: 1.000 - 2.097; p=0.0498) and ME/CFS. In our gender-stratified logistic regression analysis, there is evidence of a significant association (OR = 0.376, 95% CI: 0.143 – 0.986; p = 0.047) between RANTES levels and ME/CFS among females. There is no evidence to suggest there is a main effect of medication use during 24 hours prior to blood draw on case-control status. There is no evidence of significant differences between cases and controls with respect to gender.

Conclusion
Our findings show us that in our unadjusted logistic regression analysis there is a statistically significant association between the risk of ME/CFS and lower levels of Fractalkine. Lower levels of Fractalkine were associated with a statistically significant increased risk of ME/CFS after adjusting for medication use. The association of ME/CFS with some cytokines (MIP-1α) remained the same in terms of statistical significance whereas with some other cytokines it became less significant (Fractalkine) after adjusting for medication use during 24 hours prior to blood draw. Higher levels of RANTES were associated with a statistically significant decreased risk of ME/CFS in females only. Future direction of this research is to conduct an interaction analysis to assess whether gender influences the association between cytokines and ME/CFS outcomes as well as adjusting for different medication categories [antihypertensives, glucocorticoids, immunosuppressants, anti-viral, anti-bacterial, NSAIDS, anti-malarial, thyroid hormones, anti-diabetic, anti-depressants, anti-psychotic, pain killers, and others]. Future analyses of multiple cytokines may employ multiple-testing adjustments to minimize false positives.

PDF (Thesis)

 

I haven't read the paper, but I'm wondering how they are determining the risk of developing ME if they didn't measure these cytokines before the subject developed ME. If it's just a correlation between having ME and abnormal levels of these cytokines, it doesn't prove it's causal, or even directly caused by the ME mechanism, rather than the lifestyle changes.