The Shoulder Problem Everyone Keeps Getting Wrong

How modern biases in healthcare shape what patients believe, what clinicians say, and why outcomes often disappoint

Walk into almost any clinic with shoulder pain and you will hear a familiar script. Your rotator cuff is weak, your shoulder is impinged, you need to strengthen your shoulder…The explanation feels tidy, mechanical, and reassuring. Pain has a cause, the cause has a fix, and the fix is exercise.

The problem is that this story, while convenient, is increasingly at odds with what the research actually shows.

Over the past two decades, evidence has steadily suggested many of the assumptions that dominate shoulder rehabilitation are wrong, especially for rotator cuff related shoulder pain, now commonly referred to as RCRSP. What has replaced those assumptions is something far more uncomfortable for both clinicians and patients, complexity. Shoulder pain is not a simple strength deficit problem, and physical therapy is not a linear equation where more muscle equals less pain.

Modern shoulder care is biased, not out of malice or ignorance, but out of habit. The biases are baked into how we diagnose, how we talk, how we measure success, and how we justify treatment. And those biases matter, because they shape patient expectations, clinician confidence, and ultimately outcomes.

Bias 1: The belief that shoulder pain is primarily a strength problem

The most dominant bias in modern shoulder physical therapy is the idea that pain improves because strength increases. This belief is intuitive, easy to communicate, and deeply ingrained in both professional education and public health messaging.

But intuition is not evidence.

Clinical trials repeatedly show that while exercise improves pain and function in people with rotator cuff related shoulder pain, the associated strength gains are often modest and frequently clinically unimportant, even when pain improves substantially, as documented in controlled trials reviewed by Powell and Lewis in the Journal of Orthopaedic & Sports Physical Therapy. In some studies, external rotation (rotating out) and abduction (out to the side) strength deficits of up to 43 percent are observed at baseline, yet improvements in pain do not correlate strongly with changes in those deficits.

This disconnect is not unique to the shoulder. A 2025 editorial in British Journal of Sports Medicine highlights similar findings across musculoskeletal conditions. In knee osteoarthritis, knee extension strength mediates only about 2 percent of the treatment effect of exercise, and in Achilles and patellar tendinopathy, improvements in pain occur without meaningful changes in muscle or tendon structure or strength.

Yet clinicians continue to default to strength-based explanations, even when the mechanism does not hold up under scrutiny. This bias persists because strength is measurable, visible, and culturally associated with resilience and recovery. Pain, on the other hand, is subjective and messy.

Bias 2: Overvaluing biomechanics and undervaluing biology

Another modern bias is the assumption that improving shoulder mechanics is the primary pathway to recovery. Scapular dyskinesis (which is actually now essentially viewed as a nonsense term by modern research), muscle timing, and movement optimization are frequently targeted, despite inconsistent evidence that normalizing these factors is necessary for pain relief.

Research summarized by Powell and colleagues shows that scapular movements do not need to be normalized for pain and function to improve, and that scapula-focused exercise is not clearly superior to general shoulder strengthening.

Even when high-load resistance programs are used, strength gains remain underwhelming and do not reliably predict outcomes.

What this exposes is a deeper bias toward visible mechanical explanations. Muscles, joints, and movement patterns feel concrete. Systemic inflammation, neuroendocrine changes, and central pain modulation sound complicated, confusing, and harder to pitch to patients.

A 2022 scoping review in Musculoskeletal Science and Practice identified 32 proposed mechanisms for how exercise might help RCRSP, grouped into neuromuscular, tissue, neuro-endocrine-immune, and psychological themes. Despite this breadth, 95 percent of proposed mechanisms in clinical trials were biomedical, and neuromuscular explanations dominated, even though direct causal evidence supporting them was weak.

This imbalance reflects not what is most true, but what is most comfortable to explain.

Bias 3: The misleading certainty of diagnostic labels

Labels shape beliefs. Few areas illustrate this better than shoulder pain diagnosis.

Historically, terms like subacromial impingement syndrome implied a structural conflict that needed correction. Yet imaging studies repeatedly show that many so-called pathological findings, including rotator cuff tears, are common in pain-free individuals, especially with age.

The debate over diagnostic labeling is laid out bluntly in a guest editorial in the New Zealand Journal of Physiotherapy, where clinicians are divided between “leavers,” who want to abandon pathoanatomical labels, and “retainers,” who believe labels are essential for communication and validation.

Rotator cuff related shoulder pain was introduced as a middle ground, a hypothesis rather than a definitive diagnosis. It acknowledges structural involvement without claiming structural causation. The problem is that many clinicians and patients still hear it as a tissue failure diagnosis, reinforcing fear, fragility narratives, and the belief that something is torn, damaged, or degenerating.

As the editorial points out, people want explanations, but unexplained pain is worse than imperfect explanations. The bias lies not in labeling itself, but in overselling certainty where none exists.

Bias 4: Assuming exercise works the same way for everyone

Another subtle but damaging bias is the assumption that exercise is a standardized input with a predictable output. Do the program, get better.

Qualitative research published in Physical Therapy & Rehabilitation Journal challenges this assumption by exploring how people with RCRSP actually experience exercise therapy.

Participants consistently identified three perceived reasons exercise helped: improved strength, improved psycho-emotional state, and general health benefits. But crucially, these mechanisms only activated under certain conditions, including a strong therapeutic relationship, a tailored program, and timely progress.

When these conditions were absent, exercise failed, not because the movements were wrong, but because belief, trust, and confidence were undermined.

This exposes a major bias in healthcare research and practice. We often measure exercises, loads, and repetitions, but we under-measure context. The relationship between clinician and patient, the framing of progress, and the narrative surrounding pain all can influence outcomes, yet they are rarely treated as primary variables.

Bias 5: Confusing population averages with individual reality

Exercise therapy shows modest average effects in RCRSP. That statement is both true and misleading.

Roughly half of people continue to experience symptoms up to two years after onset, despite receiving recommended care, according to longitudinal data summarized in the 2022 scoping review.

This does not mean exercise is ineffective; it just means it's not universally effective and our current models do a poor job predicting who will benefit and why.

The bias arises when clinicians interpret average benefits as guarantees. When patients do not improve, the narrative often shifts to compliance, effort, or incorrect technique, reinforcing a sense of personal failure.

Modern evidence suggests that exercise benefits are more plausibly mediated through reductions in fear, improvements in self-efficacy, changes in pain perception, and systemic health effects rather than isolated strength gains, as emphasized in the 2025 British Journal of Sports Medicine editorial.

Reframing shoulder physical therapy without throwing it away

None of this argues against exercise. Exercise remains a first-line recommendation in every clinical practice guideline for RCRSP. What must change is how we explain it, measure it, and emotionally frame it.

Strength still matters for general health, independence, and resilience, but it is not the sole or even primary driver of pain relief in most cases. Movement quality matters, but not because it restores an idealized biomechanical model. Diagnosis matters, but only when communicated as a working hypothesis rather than a verdict.

Modern shoulder rehabilitation works best when it abandons single-cause thinking and embraces a biopsychosocial framework that reflects how pain actually behaves. The shoulder is adaptable, sensitive, and deeply influenced by context. That definitely does not mean it is broken.

The bias is not that clinicians prescribe exercise. The bias is believing that exercise works for the reasons we were taught, rather than the reasons the evidence now supports.

The bottom line

Shoulder physical therapy has outgrown its old stories. Pain is not a simple strength deficit. Recovery is not a mechanical repair job. And progress is not linear.

The sooner clinicians and patients let go of these biases, the sooner shoulder care can become both more honest and more effective. If you want modern physical therapy, where we acknowledge our biases and work extremely hard to stay up to date with the latest shoulder research call Pain & Performance Coach LLC at 971-364-0909 or fill out our contact form so we can reach out to you.