An Application of the Multivariate Linear Mixed Model to the Analysis of Shoulder Complexity in Breast Cancer Patients

Int J Environ Res Public Health. 2016 Mar 2;13(3):274. doi: 10.3390/ijerph13030274.

Abstract

In this study, four major muscles acting on the scapula were investigated in patients who had been treated in the last six years for unilateral carcinoma of the breast. Muscle activity was assessed by electromyography during abduction and adduction of the affected and unaffected arms. The main principal aim of the study was to compare shoulder muscle activity in the affected and unaffected shoulder during elevation of the arm. A multivariate linear mixed model was introduced and applied to address the principal aims. The result of fitting this model to the data shows a huge improvement as compared to the alternatives.

Keywords: autoregressive of order one; correlated random effects; multivariate linear mixed model.

MeSH terms

  • Adult
  • Aged
  • Arm
  • Breast Neoplasms / physiopathology
  • Breast Neoplasms / therapy*
  • Electromyography*
  • Female
  • Humans
  • Linear Models
  • Lymph Node Excision*
  • Mastectomy*
  • Middle Aged
  • Muscle, Skeletal / physiopathology*
  • Radiotherapy, Adjuvant
  • Range of Motion, Articular
  • Scapula / physiopathology
  • Shoulder Joint / physiopathology*
  • Shoulder Pain / physiopathology
  • Treatment Outcome