What holding the bar asks of the shoulders, and what no study has countedEmerging evidence
In a high-bar squat the bar rests on the muscle of the upper back just below the neck; in a low-bar squat it sits a few centimetres lower, across the back of the shoulders. A review of lab studies found the low bar leans the trunk further forward and works the hips and back harder, while the high bar keeps you more upright with more knee bend; one small study of 12 trained men found the opposite pattern of knee and hip bend, so the details are mixed. The low bar usually lets people lift the most.
To keep the bar on that shelf, the shoulders turn out and pull back with the elbows behind the bar, much like a milder version of the arm-up-and-back "high-five" position that went with signs of a shoulder that might slip in 159 weight trainers. In our reading of coaching sources and our squat guide, a narrow grip, elbows driven up or wrists bent back make the hands, wrists and elbows hold load the back should carry. Nobody has measured any of this: no study has counted shoulder, elbow or wrist pain from squats, or measured the shoulders, elbows or wrists in a back squat.
Sources 1, 2, 3, 4, 5
Who gets itEmerging evidence
Often the squat shows up a shoulder that pressing has already irritated. In 104 Swedish powerlifters, a current shoulder injury was strongly linked with starting during bench press training (9 of 16 current shoulder injuries started there). In powerlifting, the commonest injury sites were the lower back (30.8%), the shoulder (19.6%) and the elbow and upper arm (8.0%), at 1.0 to 4.4 injuries per 1,000 hours of training, which is low next to team sports.
Up to 36% of resistance training injuries were at the shoulder in one review; shoulder, elbow and hand or wrist injuries were more common in powerlifters over 40; and a scoping review found shoulder instability and overuse the commonest shoulder injuries in weightlifting. In US emergency departments, the shoulder was a third of upper limb weightlifting injuries, mostly from drops and crushes rather than holding a bar. No study gives the share of each cause in people who squat.
Sources 6, 7, 8, 9, 10, 11, 12
The elbows: golfer's elbow and a torn biceps tendonEmerging evidence
Golfer's elbow (medial epicondylitis) is pain where the forearm muscles attach to the inner elbow. It makes up about 10% of elbow tendon diagnoses, is commonest from 40, is diagnosed by examination, and is treated without surgery first; surgery is considered only after 6 to 12 months of other care. Other causes of inner elbow pain include the ulnar nerve and the inner elbow ligament.
A torn biceps tendon at the elbow is different: often a pop, then bruising, a bulge in the upper arm and weakness turning the palm up. It's commonest in men over 30, and in people who smoke or use steroids. Surgery to reattach the tendon is needed for full strength, and in 183 strength athletes who had it repaired, all returned to sport, while a longer delay before surgery went with worse results. Deadlifts and curls, not squats, were the commonest lifting causes.
Sources 13, 14, 15, 16
The neck under the barEmerging evidence
The bar should sit on muscle, never on the bones of the neck. A review of case reports describes rare fractures of the bony points at the base of the neck (C6 to T1) in weightlifters, from a blow or a hard muscle pull, felt as a crack and sharp pain between the shoulder blades that usually settles over weeks. They are often missed at first, so a crack with sharp pain is worth checking.
Source 17
When a scan helps, and what it can't tell youStrong evidence
Not at first for rotator cuff, AC joint or elbow tendon pain. In 96 shoulders without symptoms, a magnetic resonance imaging (MRI) scan showed a rotator cuff tear in 34%, rising to 54% over 60, and golfer's elbow is a clinical diagnosis with an ultrasound or MRI only when the diagnosis is unsure.
Scans matter for a suspected biceps tear at the elbow (an ultrasound or MRI can confirm it, though an obvious tear may not need one), sudden weakness after an injury, a shoulder that has slipped, a crack between the shoulder blades and worsening nerve signs. The NHS says a health professional may send you for tests such as an X-ray to check the cause.
Sources 18, 13, 15, 19
How it usually goesEmerging evidence
No study follows shoulder pain from squats. In the largest rotator cuff trial (708 adults), shoulder pain and function improved substantially over a year in every group. The NHS says shoulder pain usually needs 2 weeks of looking after before it starts to ease, and can take 6 months or longer. Most injured powerlifters kept training with changes: only 16% of those currently injured had stopped completely. No one can promise a date.
Sources 20, 19, 13, 6
