For most patients with rotator cuff tendon pain, physical therapy follows a familiar script of resistance exercise that resolves their pain and improves shoulder use. But for a sizable minority, recovery stalls. Now researchers want to understand why.

“Some patients take a longer time to feel better, or never feel better, and we're trying to understand why, and what treatment they need,” said lead author Lori Michener, PT, ATC, PhD, professor and clinical scholar of physical therapy and director of the Clinical Biomechanics and Orthopedic Outcomes Research Laboratory at the University of Southern California.

In an observational study published in Physical Therapy, Michener and her colleagues followed 47 patients with rotator cuff tendinopathy for eight weeks of resistance exercise and follow-up. Tendinopathy refers to persistent tendon pain and can lead to reduced function in the shoulder. The study tracked patients over time, using ultrasound to measure tendon structure while repeatedly assessing pain and function.

A tendon under stress and repetitive loading may temporarily thicken with fluid and inflammatory substances. The tendon’s structure also becomes less organized. With appropriate rest and exercise, that disorganization and related swelling can subside.

Early on, the tendon recovered in clearly observable ways. It became thinner, more organized, and less inflamed, while patients reported improved symptoms. However, over time, this relationship between tendon health and symptoms were not well related. The change in the tendon slowed, even though many patients continued to feel improvement in pain and shoulder use.

“In those first weeks, we saw the most correlation between the change in the tissue and the symptoms,” said first author Oscar Vila Dieguez, PT, PhD, a postdoctoral fellow at the University of California, San Diego, and former graduate student in Michener’s lab. “After that, people were still having improvements in symptoms, but not much in the tendon tissue.”

“You would think if the tissue heals, you should feel better,” Michener added. “But some people continue to feel better even though the tendon isn’t continuing to change.”

Recovery, it turned out, continued beyond what imaging could show. Tendon thickness and structure tell only part of the story.

This study is part of a broader effort by the research team to rethink how tendinopathy patients recover. Tendon healing is one piece of the answer. The researchers are also examining shoulder movement and muscle function, brain’s pain processing, mental health, and metabolic factors such as diabetes and obesity, as well as patterns of widespread pain sensitivity. Their work, supported by a National Institutes of Health (NIH) R01 grant and earlier Foundation for Physical Therapy Foundation funding, aims to move shoulder care toward a more precise, individualized model.

The eventual goal is to classify patients into subgroups based on tendon structure, neuromuscular function, psychological profile, metabolic status, and pain distribution — and then tailor treatments accordingly. Physical therapists, orthopedists, and primary care clinicians could use this classification system to decide which patients are likely to respond well to standard exercise programs and which might need added attention to psychological, metabolic, or systemic pain factors.

“Precision rehabilitation is where we're trying to go with this, so patients get the right treatment at the right time,” Michener said.

This work depends on sophisticated statistical modeling — analyzing eight weeks of tendon imaging and patient’s symptoms across many patients to understand how multiple variables change together over time. For that, the team has relied on biostatistician Wendy Mack, PhD, director of the Biostatistics team at the Southern California Clinical and Translational Science Institute (SC CTSI).

Mack is a co-investigator on the NIH R01 grant supporting this study and multiple related papers. She shaped the team's analytical approach, selecting time-series methods to track how tendon structure, symptoms, and other factors shifted week by week across the patient group. The time-series approach refers repeatedly measuring the same variables over time to analyze how that measurement changed and related to each other across that timeline.

“We could not do this without Wendy,” Michener added. “She helped write the grant and helped us understand which statistical approaches would be most appropriate.”

By guiding that statistical strategy, Mack is helping transform complex longitudinal data into a precision rehabilitation model that could more accurately match patients with the treatments they need.

“CTSI Biostatistics collaborations with Dr. Michener included intensive assistance with study design and statistical analysis plans for her grant proposals that have funded this work and my ongoing collaborations with her research group,” said Wendy Mack, PhD. “I have been excited that both I and CTSI have contributed to this innovative research that may significantly impact physical therapy practice to improve recovery in this group of patients.”

Another boon for researchers is tablet-based ultrasound devices that are becoming more prevalent in clinics, allowing them to monitor tendon thickness and structure more easily than ever. Using these images alongside psychological screening and pain mapping will promote more individualized treatment. Exercise will likely remain the backbone of therapy. But for the many patients whose recovery stalls, this research may eventually help clinicians understand why and point toward what to do about it.