Five-week Outcomes From a Dosing Trial of Therapeutic Massage for Chronic Neck Pain

Karen J. Sherman, PhD, MPH; Andrea J. Cook, PhD; Robert D. Wellman, MS; Rene J. Hawkes, BS; Janet R. Kahn, PhD; Richard A. Deyo, MD, MPH; Daniel C. Cherkin, PhD


Ann Fam Med. 2014;12(2):112-120. 

In This Article

Abstract and Introduction


Purpose. This trial was designed to evaluate the optimal dose of massage for individuals with chronic neck pain.

Methods. We recruited 228 individuals with chronic nonspecific neck pain from an integrated health care system and the general population, and randomized them to 5 groups receiving various doses of massage (a 4-week course consisting of 30-minute visits 2 or 3 times weekly or 60-minute visits 1, 2, or 3 times weekly) or to a single control group (a 4-week period on a wait list). We assessed neck-related dysfunction with the Neck Disability Index (range, 0–50 points) and pain intensity with a numerical rating scale (range, 0–10 points) at baseline and 5 weeks. We used log-linear regression to assess the likelihood of clinically meaningful improvement in neck-related dysfunction (≥5 points on Neck Disability Index) or pain intensity (≥30% improvement) by treatment group.

Results. After adjustment for baseline age, outcome measures, and imbalanced covariates, 30-minute treatments were not significantly better than the wait list control condition in terms of achieving a clinically meaningful improvement in neck dysfunction or pain, regardless of the frequency of treatments. In contrast, 60-minute treatments 2 and 3 times weekly significantly increased the likelihood of such improvement compared with the control condition in terms of both neck dysfunction (relative risk = 3.41 and 4.98, P = .04 and .005, respectively) and pain intensity (relative risk = 2.30 and 2.73; P = .007 and .001, respectively).

Conclusions. After 4 weeks of treatment, we found multiple 60-minute massages per week more effective than fewer or shorter sessions for individuals with chronic neck pain. Clinicians recommending massage and researchers studying this therapy should ensure that patients receive a likely effective dose of treatment.


Neck pain is a common condition, with a 12-month prevalence of 30% to 50% and rates of activity-limiting pain of 1.7% to 11.5%,[1] and it accounts for more than 10 million ambulatory medical care visits per year in the United States.[2] At least one-half of persons with neck pain report persistent or recurrent neck problems at 1 to 5 years of follow-up.[3,4] Neck pain is the eighth leading cause of disability in the United States[5] and fourth worldwide.[6] It is the second leading reason for use of complementary and alternative medicine (CAM),[7] with chiropractic and massage most commonly used.[8] In a national survey, 61% of persons with neck pain who used both CAM and conventional therapies perceived CAM therapies to be more helpful for this condition, whereas just 6% perceived conventional treatments to be better.[9]

Massage is the second most commonly used CAM therapy for neck pain.[10,11] Although it is often used as a stand-alone treatment for chronic neck pain in the United States, reviews of research on massage for neck pain draw inconsistent conclusions. Furlan et al[12,13] found massage superior to various controls, Brosseau et al[14] found it effective immediately posttreatment with further follow-up data lacking, and Patel et al[15] were unable to draw conclusions because of the poor quality of mostly clinically irrelevant studies. An earlier Cochrane review[16] noted that studies have used such different types and doses of massage that the optimum dose for practice and clinical trials is unknown. We therefore designed a study to evaluate the optimal combination of frequency and length of sessions with therapeutic massage in persons with chronic neck pain. This article describes outcomes 1 week after the end of a 4-week treatment period, our primary endpoint.