A Possible Source of Heterogeneity in Nociceptive Flexion Reflex (NFR) Threshold Among Patients with Fibromyalgia: Systematic Review and Meta-Analysis

Abstract

Objectives: We aimed to investigate the possibility of sample size as a possible source of heterogeneity in trials comparing the nociceptive flexion reflex threshold (NFRthr) in fibromyalgia (FM) patients and healthy controls. Methods: A systematic search strategy was designed and conducted in electronic databases such as Ovid MEDLINE, Cochrane Libraries, Embase, and PsycINFO. In total, nine scholarly records were filtered by the inclusion and exclusion criteria, passed for data extraction (e.g. study designs and settings; methodology; total, male, and female sample sizes in FM and healthy control groups, and effect sizes), and statistical analyses (e.g. metaanalysis and meta-regression). Using random-effects model, the effect size (i.e. mean difference between the NFRthr in FM patients and healthy controls) was meta-analyzed. Investigating the sample size as a possible source of heterogeneity and utilizing a restricted maximum likelihood (REML) technique, the trials’ effect sizes were meta-regressed in multiple meta-regression against the total, male, and female sample sizes. Prior to any investigations, the aims of this research were registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number: CRD42019140354). Results: Results indicated the deficit in NFRthr (overall mean difference = −3.16; 95% 𝐶𝐼: −6.82 𝑡𝑜 0.50) was heterogenous (𝐼 2 = 0.91, 𝜏 2 = 25.04, 𝜒2= 91.22, df & < 0.00001) and the overall effect was not statistically significant (Z = 1.69; 𝑃 = 0.09). The multiple meta-regression analyses indicated that one of the main sources of bias to the effect size in the trials was found to be the total (𝑆𝑆𝑡𝑜𝑡𝑎𝑙 = −0.0570; 𝑃 = 0.040) and female (𝑆𝑆𝑓𝑒𝑚𝑎𝑙𝑒 = −0.0569; 𝑃 = 0.047) sample sizes that participated in the studies. Conclusion: The findings suggest that the total and female sample sizes might be possible sources of bias to effect sizes determined in NFRthr in FM patient and healthy controls. A possible solution might be to conduct sex-specific trials to determine the NFRthr among FM patients in future research. Funding Statement: Authors received no source of funding for this research. Declaration of Interests: Authors declare no conflict of interest. Ethics Approval Statement: PROSPERO; registration number: CRD42019140354