DOS 2017

DOS Kongressen 2017 ·

Reproducibility of dual x-ray absorptiometry and as-

sessment of changes in regional body composition fol-

lowing shoulder arthroplasty for osteoarthritis

Mustafa Al-Hamdani, Bo S. Olsen, Bo Zerahn, Jeppe V. Rasmussen

Herlev Hospital og Københavns universitet; Ortopædkirurgisk afdeling,

Background:

Measurements of body composition using a DXA may objectively

reflect the outcome of various rehabilitation programs and surgical procedures

in the upper extremity.

Purpose / Aim of Study:

To assess the reproducibility of dual x-ray absorpti-

ometry (DXA) measurements of regional body composition in the shoulder with

or without the presence of a shoulder arthroplasty. Furthermore, we used the

DXA to assess changes in body composition after shoulder arthroplasty.

Materials and Methods:

Body composition was measured in one region of

interest (ROI) corresponding to the deltoid muscle. Each patient had two dupli-

cate scans for both the affected and the contralateral arm. Data on functional

outcome score (e.g., the Constant Score) and muscle strength were collected

preoperatively whereas the body composition scans were collected the day af-

ter surgery. The patients were re- examined at 3 and 12 months.

Findings / Results:

Intraclass correlation values (ICC) between duplicate scans

were 0.991 and 0.996 with and without a shoulder arthroplasty respectively.

The Bland-Altman plots showed narrow limits of agreement. Friedman’s and

Wilcoxon test showed highly significant declines in muscle mass 3 months after

surgery, p <0.001 and the muscle mass was not regained at the 12-months

follow-up. There was a statistically significant correlation between muscle mass

and muscle strength, r=0.23, P=0.03.

Conclusions:

DXA is an excellent method for measuring body composition in

the upper extremity. The presence of a shoulder arthroplasty did not affect re-

producibility. Muscle mass decreased during the first 3 months and was not fully

regained a year after surgery. The reason for this is unknown, but the results of

the present study underline the importance of a better understanding of factors

that influence postoperative recovery programs after shoulder arthroplasty.

No conflicts of interest reported

23.