Timmons · Journal of shoulder and elbow surgery 2016 · Within-subject comparative biomechanical study · n=28

Empty can exercise provokes more pain and has undesirable biomechanics compared with the full can exercise.

Level 3 - non-randomized controlled study

Single-group within-subject repeated-measures biomechanical comparison

PubMed 26577128 · doi:10.1016/j.jse.2015.08.046 · record verified 2026-08-26

What was done

Twenty-eight participants with subacromial impingement syndrome performed 5 consecutive repetitions of full can (FC) and empty can (EC) exercises. Researchers evaluated pain on an 11-point numeric rating scale, measured 3-dimensional scapular and clavicular positions, and recorded scapular muscle electromyographic activation during both exercises.

What was found

Pain was significantly higher during the EC exercise compared to the FC exercise (difference: 1 point, P = .003). For kinematics, the EC exercise resulted in greater scapular upward rotation (difference: 3°, P < .001), greater internal rotation (mean difference: 2°, P = .017), greater clavicular elevation (difference: 3°, P < .001), and less scapular posterior tilt (difference: 2°, P < .001). Muscle activation during ascent was greater with EC for upper trapezius (difference: 4%, P = .002), middle trapezius (difference: 3%, P < .001), and serratus anterior (difference: 0.5%, P = .035). During descent, EC produced higher upper trapezius (difference: 2%, P = .005) and middle trapezius activity (difference: 1%, P = .003), but lower activity of the lower trapezius (difference: 1%, P = .039).

Why it matters

These findings show that empty can elevation induces scapular orientations associated with subacromial space narrowing and provokes more pain than full can elevation in symptomatic patients, supporting the preference for full can exercises in rehabilitation.

Limits

The study had a small sample size (n = 28) and evaluated only immediate biomechanical and pain responses rather than longitudinal clinical outcomes. Additionally, absolute kinematic differences (2°–3°) and muscle activation differences (0.5%–4%) were small.