Evaluation of Patient-Rated Stiffness Associated With Fibromyalgia: A Post-Hoc Analysis of 4 Pooled, Randomized Clinical Trials of Duloxetine

Robert Bennett, Jon I. Russell, Ernest Choy, Michael Spaeth, Philip Mease, Daniel Kajdasz, Daniel Walker, Fujun Wang, Amy Chappell

Research output: Contribution to journalArticlepeer-review

19 Scopus citations


Background: Patients with fibromyalgia (FM) rate stiffness as one of the most troublesome symptoms of the disorder. However, there are few published studies that have focused on better understanding the nature of stiffness in FM. Objectives: The primary objectives of these analyses were to characterize the distribution of stiffness severity in patients at baseline, evaluate changes in stiffness after 12 weeks of treatment with duloxetine, and determine which outcomes were correlated with stiffness. Methods: These were post-hoc analyses of 3-month data from 4 randomized, double-blind, placebo-controlled studies that assessed efficacy of duloxetine in adults with FM. Severity of stiffness was assessed by using the Fibromyalgia Impact Questionnaire (FIQ) on a scale from 0 (no stiffness) to 10 (most severe stiffness). The association between changes in stiffness and other measures was evaluated by using Pearson's correlation coefficient. The FIQ total score and items, the Brief Pain Inventory (BPI-modified short form), the Clinical Global Impression-Severity scale, the Multidimensional Fatigue Inventory, the 17-item Hamilton Depression Rating Scale, the Sheehan Disability Scale, the 36-item Short-Form Health Survey, and the EuroQoL Questionnaire-5 Dimensions were evaluated in the correlation analyses. Stepwise linear regression was used to identify the variables that were most highly predictive of the changes in FIQ stiffness. Results: The analysis included 1332 patients (mean age, 50.2 years; 94.7% female; and 87.8% white). The mean (SD) baseline FIQ stiffness score was 7.7 (2.0), and this score correlated with baseline BPI pain score and FIQ function. Duloxetine significantly improved the FIQ stiffness score compared with placebo (. P < 0.001) and provided a moderate effect size (0.23 for the 60-mg dose and 0.38 for the 120-mg dose). Changes in stiffness were best correlated (range, 0.52-0.75; all, . P < 0.001) with changes in BPI/FIQ pain and interference scores, FIQ nonrefreshing sleep, FIQ anxiety, 36-item Short-Form Health Survey bodily pain, and Sheehan Disability Scale total score. Variables related to severity of pain, pain interfering with daily activities, and physical functioning were predictors of change in stiffness. Conclusions: Stiffness scores were high in this population with FM and best correlated at baseline with BPI pain score and FIQ function. Not unexpectedly, improvement in stiffness with duloxetine correlated with many of the other markers of FM severity, presumably a result of amelioration in FM comorbidities.

Original languageEnglish (US)
Pages (from-to)824-837
Number of pages14
JournalClinical therapeutics
Issue number4
StatePublished - Apr 2012
Externally publishedYes


  • Correlation
  • Duloxetine
  • Fibromyalgia
  • Pain
  • Stiffness

ASJC Scopus subject areas

  • Pharmacology
  • Pharmacology (medical)


Dive into the research topics of 'Evaluation of Patient-Rated Stiffness Associated With Fibromyalgia: A Post-Hoc Analysis of 4 Pooled, Randomized Clinical Trials of Duloxetine'. Together they form a unique fingerprint.

Cite this