Article Text

other Versions


Are baseline characteristics related to responses to acupuncture treatment for musculoskeletal conditions? Analysis of a dataset from a teaching centre
  1. Kai-Yin Hsu1,
  2. Kenneth Kwan Ho Chui2,
  3. Lisa Conboy3
  1. 1Research Department, New England School of Acupuncture at MCPHS University, Newton, Massachusetts, USA
  2. 2Department of Public Health and Community Medicine, Tufts School of Medicine, Tufts University, Boston, Massachusetts, USA
  3. 3Director of Research, New England School of Acupuncture at MCPHS University, Newton, Massachusetts, USA
  1. Correspondence to Kai-Yin Hsu, Research Department, New England School of Acupuncture at MCPHS University, 150 California Street, Newton, MA 02458, USA; Kai-Yin.Hsu{at}

Statistics from


Acupuncture treatment outcomes can differ according to patient characteristics. However, whether these factors are merely correlational or involved in the causal framework is largely inconclusive. A deeper understanding of clinical, psychosocial, and treatment factors would allow us to form more accurate prognostic predictions, enhancing follow-ups and treatment outcomes. We performed an observational study in an acupuncture teaching centre of patients treated for musculoskeletal conditions. Our objectives were to: (1) follow patients who received acupuncture treatment; and (2) identify and further describe patients who were especially likely to adhere to and benefit from acupuncture.


This is a secondary analysis of data1 collected from October 2009 to July 2010 from 421 patients who filled out our Measure Yourself Medical Outcome Profile (MYMOP)2 forms at the New England School of Acupuncture (NESA) teaching centre. The study was approved and overseen by the New England Institutional Review Board (reference no. NEIRB#: 13-121).

Through a literature review, we identified eight independent variables potentially related to outcomes,3 ,4 namely five sociodemographic measures (sex, age, marital status, educational level, employment status); MYMOP health status; prior exposure to acupuncture or oriental medicine; and confidence level in the acupuncture treatments. Baseline study variable values were from a health history form that all patients complete. MYMOPs were used to measure clinical outcomes.

For objective 1, patients were chosen only if they provided complete study variable data and listed diseases of the musculoskeletal system and/or connective tissue as the main or secondary complaints on their health history form. For objective 2, patients also needed to have completed a 6th week follow-up MYMOP.

For objective 1, we compared study variables and baseline MYMOP score for the first chief complaint using the t-test, Wilcoxon rank sum and χ2 tests. For objective 2, we examined the direction and magnitude of association between the change in the MYMOP score and each factor. Due to the limited sample size, we performed bivariate instead of multivariate regression using Stata V.14 (Stata Corp, College Station, Texas, USA). Cases with any missing study variable values were removed from the analysis. Statistical significance was set at p<0.05.


The baseline variables of patients with and without 6th week MYMOP (n=45 and n=256, respectively) are shown in table 1.

Table 1

Baseline variables of patients with and without 6th week MYMOP scores

The initial sample size was 421. Removing missing data yielded 301 cases (256 non-returning, 45 returning). Returning and non-returning patients were largely similar in terms of sex, race, education, employment status, experience and confidence in Traditional Chinese Medicine (TCM), and the baseline MYMOP score of the first chief complaint. However, returning patients were, on average, 9.5 years older (p<0.001) and more likely to be married or previously married rather than never married (p=0.009). Returning patients had an average reduction in MYMOP score of 1.3 points (p<0.001). However, only two factors showed significant associations with change in MYMOP score: employment status (p=0.022) and baseline MYMOP (p=0.005) using linear regression.


We found that age and marital status were unrelated to baseline MYMOP scores but may influence return for follow-up treatments. Previous research has found a relationship between baseline age/pain and subsequent follow-up in the management of musculoskeletal conditions.5 The MYMOP baseline was significantly associated with change in MYMOP scores. Additionally, regardless of MYMOP baseline, returning patients tended to have large improvements in MYMOP score at 6th week follow-up.

The small sample and observational design of our study both limit its power to detect significant differences and prevent us from deriving causal inference of the results. Future studies should strive for a higher sample size. However, we did find potentially important baseline predictors for successful acupuncture treatments. The generalisability of our findings is limited due to our low follow-up rate of 18%. The 45 patients we have post-treatment data on may possibly have favoured acupuncture treatment more than those that discontinued potentially due to dissatisfaction with the treatment. However, our findings are still applicable to those patients engaged in long-term acupuncture therapy.


We found some variables potentially affecting acupuncture treatment outcomes for musculoskeletal conditions. Larger research studies are needed for stronger conclusions.


We would like to thank NESA faculty, staff, students and patients, and Thomas Yan.


View Abstract


  • Contributors All contributors meet the criteria for authorship.

  • Competing interests None declared.

  • Ethics approval New England Institutional Review Board (NEIRB#: 13-121).

  • Provenance and peer review Not commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.