Arthritis & Rheumatism, Volume 63,
November 2011 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Chicago, Illinois November 4-9, 2011.

Mechanisms of Improvement in Fibromyalgia Symptoms in a Clinical Trial of Exercise: Increased Fitness or Hawthorne Effect?

Mazzuca1,  Steven A., Kaleth2,  Anthony, Saha3,  Chandan, Slaven3,  James, Ang4,  Dennis C.

Indiana Univ Schl of Medicine, Indianapolis, IN
Indiana University Purdue University Indianapolis
Indiana University
Indiana University, Indianapolis, IN


We have previously shown that initiation of supervised aerobic exercise for fibromyalgia (FM) patients, whether followed by a course of motivational interviewing (MI) to promote maintenance or standard instruction in FM self-care, was associated with an overall increase in self-reported physical activity levels and improvement in FM symptoms. To rule out the possibility that self-reported improvements reflected the generalized effects of attention, we explored whether the association between increased moderate-vigorous physical activity (MVPA) and improved clinical outcomes was mediated by an objective increase in physical fitness.


Subjects were 201 FM patients who completed the 36-week Research to Encourage Exercise for Fibromyalgia (REEF) trial. We measured changes in MVPA (CHAMPS), physical fitness [6-minute walk test (6MWT)], pain [Brief Pain Inventory (BPI)], and severity of FM symptoms [Physical Impairment (PI) subscale of the Fibromyalgia Impact Questionnaire (FIQ) and the FIQ global]. General linear models were used to perform a mediation analysis, following the approach described by Baron and Kenny (J Personal Soc Psych 1986;51:1173–82). All analyses were controlled for REEF treatment group (T).


Subjects were divided into 3 subgroups, based on the achievement and/or maintenance of >=90 min/wk increase in MVPA: Group A, achieved and maintained at week 36 (n=51); Group B, achieved but not maintained (N=80); Group C, not achieved (n=70). Groups A and B exhibited significantly greater improvements in fitness (6MWT) than Group C (P= 0.0069 and 0.0199, respectively). Changes in fitness correlated with changes in symptoms (P<0.05 for all outcomes). The table below characterizes the association between MVPA and clinical outcomes, with and without adjustment for changes in 6MWT.

 Parameter Estimate (P-Value)
OutcomeAdjusted for 6MWTGroup A vs. CGroup B vs. CR2 (MVPA|T)D R2 (%)
BPINo-1.03 (0.004)-0.40 (0.199)0.0434 
 Yes-0.89 (0.017)-0.44 (0.188)0.0312-28%
FIQ-PINo-1.74 (<0.001)-0.85 (0.024)0.0824 
 Yes-1.39 (0.002)-0.76 (0.054)0.0522-37%
FIQ GlobalNo-13.64 (<0.001)-4.94 (0.102)0.0794 
 Yes-10.20 (0.005)-2.69 (0.404)0.0470-41%


The magnitude of the association (parameter estimate) between a maintained increase in self-reported MVPA and improved FM symptoms was diminished 14–25% by adjustment for a concurrent, objective measure of change in physical fitness (6MWT). Adjustment for fitness reduced the % of variance in outcomes associated with MVPA levels by 28–41%. These data suggest that while the Hawthorne Effect cannot be totally discounted, post-intervention changes in FM symptoms in the REEF trial were mediated, in part, by fitness-related changes in MVPA.

To cite this abstract, please use the following information:
Mazzuca, Steven A., Kaleth, Anthony, Saha, Chandan, Slaven, James, Ang, Dennis C.; Mechanisms of Improvement in Fibromyalgia Symptoms in a Clinical Trial of Exercise: Increased Fitness or Hawthorne Effect? [abstract]. Arthritis Rheum 2011;63 Suppl 10 :1607

Abstract Supplement

Meeting Menu