by Lucy Candib, MD Family Health Center, 26 Queen St. Worcester MA 01610 USA
508-860-7700 fax 508-860-7855 lcandib@massmed.org

Abstract:
In this randomized crossover study we enrolled low income Anglo and Hispanic women with chronic pain in a program of 8 weekly one-hour whole body massage treatments. Patients were randomized to either active treatment or an 8 week waiting period followed by the massage program. In preliminary analyses we found a significant and lasting improvement on physical functioning on the SF-36 but no improvement in mental health functioning. The study was limited by the high dropout rate.

Background:
Chronic pain is a frustrating problem for both patients and physicians. Despite frequent office visits and expensive and sometimes harmful medications, many patients do not improve. Sometimes the pain is part of a more complex mental health problem like depression, anxiety, or somatization. Low income patients with chronic pain may be particularly difficult to treat for a variety of reasons. Patients from developing countries or minority ethnic groups may be more likely to articulate distress in bodily symptoms; yet these same patients are also more likely to suffer from chronic debilitating diseases and have fewer resources available to manage their health care. Whole body massage therapy offers an alternative approach to the treatment of chronic pain patients. Massage therapy is widely used in sports medicine to reduce pain and promote flexibility and has been documented to reduce pain and behavioral and biochemical measures of anxiety and depression in various populations.

Objectives:
Our objectives are to present the design and preliminary results of a study of massage therapy for women patients with chronic pain. The following information will be presented and discussed.

Study Design:
Randomized controlled trial (double cross-over) with four months of follow-up observations.

Setting:
Inner-city community health center serving low-income multi-ethnic families.

Participants:
A total of 70 low-income Anglo and Hispanic women patients age 18-65 referred by their family physician or family nurse practitioner were enrolled of whom 48 completed at least 6 massage treatments and at least two follow-up measures. Eligibility requirements: at least one year of primary care at the health center and chronic pain lasting at least 6 months documented in the medical chart.

Intervention:
After completion of the baseline assessment measures, patients were randomized to either the first or second treatment group by means of a random number table. Within each cohort, the initial treatment group received 8 weekly one hour full-body massage treatments by an experienced licensed female massage therapist in a quiet setting within the health center. Follow-up evaluations were performed at the end of treatment and 2 and 4 months later. The crossover group had a baseline assessment and then underwent an 8 week period without any change in treatment (that coincided with the time that the initial group was receiving treatments). At 8 weeks, after assessments had been repeated, the crossover group then received the 8 week massage intervention and the three post-treatment observations. Patients were assigned to one massage therapist for the duration of the program. Whole body massage was conducted using soft tissue manipulation of the head, neck, arms, legs and trunk, focusing on symptomatic areas. A combination of adjunctive therapies such as trigger point, cross fiber friction, Reiki, and oriental massage techniques were incorporated on an individual basis.

Main Outcome Measures:
The composite SF-36 mental health (MCS) and physical functioning (PCS) scores over time; CES-D and STAI-S scores over time; Dartmouth COOP scores over time.

Statistical Methods:
SF-36 scores were available for the 48 subjects who completed at least 6 weeks of massage therapy and 38 subjects had observations out to the final follow-up. A mixed model analysis with repeated measures was employed. The mixed models approach is similar to Analysis of Variance procedures but without some of the assumptions of ANOVA that are often not satisfied in the data. In this particular situation, it was important to employ an analytic approach that allowed cases with missing data to be included in the analysis and also allowed for selection of the appropriate variance-covariance matrix since the more restrictive assumptions of ANOVA were not met. The SAS PROC MIXED procedure was used for the analysis.

Results:
The group undergoing 8 weeks of observation after enrollment and before the intervention experienced no improvement in symptoms and SF-36 scores did not differ significantly from baseline in this group. Therefore, the two groups were combined and subsequent analyses were carried out using only baseline, post-treatment and the 2 follow-up observations for all subjects. Mental health scores (MCS) differed significantly between Latino and Anglo subjects (lower for Latino subjects) but there was no improvement in MCS following massage therapy. Physical health scores (PCS) were similar for Latino and Anglo patients and differed significantly over time (p<.01). There was no significant interaction between ethnicity and time with respect to either outcome. In particular, PCS scores improved in both groups following massage therapy and this improvement was sustained through the entire follow-up period. Scores on CES-D, STAI-S and Dartmouth COOP scales all improved during massage therapy but returned to pretreatment levels within 4 months of completing treatment. Of note, 86% of the Anglo patients and 53% of Latino patients acknowledged some form of prior physical or sexual victimization in either childhood or adulthood.

Conclusions:
Massage therapy may be an effective treatment for some chronic pain patients. For patients who completed 6 or more massage therapy treatments there was an improvement in physical functioning and furthermore, this improvement was sustained over a 4-month follow-up period. In this setting most Anglo women with chronic pain and more than half of Latino women with chronic pain were survivors of prior violent victimization.

Discussion:
This study suggests that massage therapy was a lasting and effective treatment for improving physical functioning in those women with chronic pain problems who were able to participate in a sustained treatment program. The study was limited by the high number of dropouts from treatment. Dropouts offered two main reasons for not continuing in the program: 1) logistical problems typical for low-income families including moving away, sickness in the family, multiple conflicting appointments, and work and child care obligations; and 2) discomfort with massage therapy itself, or lack of result from the initial massage. To offset the first set of difficulties, future studies might need to offer childcare and transportation and increased flexibility in the scheduling appointments. Massage therapy may not be a suitable modality for the second group.

Improvement in physical functioning rather than mental health functioning after a course of massage therapy is a plausible outcome since patients experience chronic pain problems as physical, and the modality of massage directly addresses the location of the pain in the body. Insofar as chronic pain is sometimes linked with depression and somatization, a long-standing improvement in physical functioning may have the potential to improve mental health functioning. The likelihood of prior violent victimization in women with chronic pain should be addressed in considering the use of massage therapy. We were unable to demonstrate any significant effect on mental health scores over a four month follow-up period. However, this analysis included only SF-36 composite scores; more sensitive measures of mental health symptomatology and pain should be assessed in the future.

Trackback URI | Comments RSS