"Scapular Muscle Recruitment Patterns: Trapezius Muscle Latency with and without Impingement Syndrome- Cools et al, AJSM '03".
Last month I got hammered pretty good on a discussion thread on the GAIN forum on this topic. The topic had to do with the use of traditional olympic bar pulling movements (shrugs, high pulls et al) in the training of throwing athletes. I made the point that these types of pulling movements create upper trapezius hypertrophy which can cause impingement syndromes in throwers and swimmers. This is one study, of several, which shows trap shrugging preceding middle/lower trap activity in subjects with impingement syndrome. Of course one could say this is an adaptive pattern that people with impingement syndrome take on subconsciously to avoid the portion of the arc that is painful.
My take on the matter is that this muscle firing sequence CREATES impingement syndromes/tendinitis. I first came up with the idea years ago when I was fortunate enough to take a course with Australia's Lyn Watson, a manual P.T. who specializes in shoulders. Lyn felt it had to do with the traditional shrug being more levator scapulae dominant than trapezius. The upper trapezius came more into play when the arms were pushing overhead.
I don't want anyone to be afraid to shrug or pull! In sports like volleyball, the shrug is an important part of the jumping/striking sequence. The same could be said for the javelin thrower, where the implement creates a long lever arm that the shoulder must contend with. But I still think they must be applied prudently in swimmers, baseball players and cricket bowlers.
These pictures are of one of my favorite thrower/swimmer exercises, the see-saw. It can be done with the bands attached to a fixed object, or better yet in opposition with a partner. It creates joint stability/power where you need it the most- where the arms are at the extremes of motion and switching directions. The "transformational zone" as Gary Gray calls it. The goal is to get straight arms behind the ears without arching the back. Many of your atheltes won't be able to do this with moderate resistance. And if they can't, they're really not getting hip to shoulder.
My point is that it's not only about hip to shoulder; but rather how you get hip to shoulder. My example has resistance applied with a horizontal vector bias. The middle/lower traps are integrated quite nicely. And so are the upper traps, but as opposed to a shrug, through a long lever arm. As the arms would during swimming and throwing, no? Think about it a bit, and let me know if I'm on target or just making some wild metaphysical leap with the research.