Speaker 1:
Speaker 2:
Good day, it's Jared. Today we talk about a question that every single musculoskeletal clinician faces, often multiple times a day that we still don't have a definitive answer for, and that question is, should we allow pain during exercise or should we avoid it if we do allow it? How much? At what intensity does it matter what condition the patient has or does it matter who the patient is?
Speaker 2:
Now the answer to these questions shape how we prescribe exercise. Also, how we counsel and reassure patients, how we judge progress and what we say. When a patient comes back to us with desperation in their eyes and they report that the exercise that we prescribed them made them sore. Now, full disclosure, I'm a co-author on a recent paper led by Ivy Tran in the Journal of Orthopedic and Sport Physical Therapy, where we conducted a meta-analysis on this very question. Now, the findings have made me think differently about how I talk about pain during rehab with my own patients, and there's a bunch more evidence coming. A trial called Pays, PASE run out of Copenhagen is underway and is specifically designed to test whether allowing pain during shoulder exercise produces better outcomes than avoiding it. Now, the results aren't out yet, but the question is gathering steam.
Speaker 2:
So today I want to walk through where the evidence currently sits, what I think the clinical message should be, and the nuances that complicate the picture. Let's go. If your practice for any length of time in musculoskeletal medicine or rehab, you've had some version of this conversation. A patient is doing their exercises at home, they come back and say, I did them, but they hurt. Should I keep going? And you have to decide in that moment what to tell them. Now for most of history in musculoskeletal rehabilitation, the implicit answer was avoid pain. Pain is the body's signal that something is wrong. It's a threat signal. If an exercise hurts, you are probably doing damage or at least irritating tissue that needs to settle down, back off, modify, reduce the load, wait for the pain to subside before progressing. Now that advice made intuitive sense.
Speaker 2:
It still does in some contexts, and there are clinical scenarios where it's clearly correct. The early postoperative period, an acute inflammatory flare, a stress fracture of the navicular bone, for example, a patient who's already in significant distress. Pain in those settings is meaningful and I think pushing through it is probably ill advised. But for persistent or chronic musculoskeletal pain, the population we spend most of our time with and who prove frustratingly the most resistant to our care, that framing has been increasingly challenged over the last decade or so. The argument is as follows, in persistent pain, the relationship between pain and tissue damage is loose. Imaging findings don't often correlate with symptoms. Movements that hurt are not necessarily movements that harm and a strict pain avoidance approach can reinforce the very fear of movement that's driving the disability in the first place. So if you took that argument seriously, you ended up at a different position.
Speaker 2:
Pain during exercise might not be a red flag, quite the opposite. It might even be desirable on the theory that exposing the patient to tolerable amounts of their own pain in a controlled setting could help recondition or desensitize the system that's keeping them stuck in pain. That was the position that started to crystallize around the year 2017 when Ben Smith and colleagues published a systematic review in the British Journal of Sports Medicine that became influential. Now what they did was pull the available trials comparing painful and non-painful exercise across various persistent musculoskeletal pain conditions, and they found a small short-term benefit, yes, benefit for painful exercise on pain intensity with no difference for disability. The finding was modest, but it seemed revelatory at the time. The clinical translation that emerged from it was that exercising into pain was not just permissible, but may even be marginally preferable.
Speaker 2:
Now that shaped a lot of clinical practice over the last seven or eight years, including mine. The Smith Review was due an update because it was eight or nine years ago now many new trials have been published since 2017, and the question deserved a fresh look. So led by Ivy Tran and a team including me, Mitch Gibbs, Ben Smith himself, who published the original paper and Matt Jones down at the University of Sydney, and now we, we redid the review with additional data and the paper came out in the Journal of Orthopedic Sport Physical Therapy in 2025. The updated picture is in some ways less transformative than the original. With additional trials included the short-term advantage for painful exercise on pain intensity dissolved. It went away, it collapsed. The pooled estimate moved closer to no difference for disability. The picture remained as it was no difference between painful and non-painful exercise.
Speaker 2:
Now you might read that or hear that and think it weakens the case for permitting pain during exercise, but I want to suggest the opposite. The clinical message coming out of this updated picture is indeed clearer and more useful than the message that came out of the 2017 review. If painful exercise had been clearly better, we'd have to recommend it with all the difficulty that involves telling patients their exercise should hurt, navigating the inevitable resistance, taking responsibility for when things flare up. If painful exercise had been clearly worse, we'd have to avoid it, which would constrain what we prescribe and reinforce the avoidance behaviors that often drive chronic pain. But what the updated evidence shows is that painful exercise and non-painful exercise produce roughly equivalent outcomes. On average, pain during exercise is not a determinant of recovery in either direction. It's neither a feature nor a bug.
Speaker 2:
That's a useful finding because it removes the false binary that was never that helpful in the first place. So the question stops being should this exercise hurt and actually becomes something like what exercise does this person need to do and how much pain can they tolerate while doing it? They're different questions and I think the second question is more clinically tractable and appropriate. So the headline message I take from all of this is pretty simple.
Speaker 2:
Pain during rehab is not a red flag. It is not to be conflated with harm. It doesn't predict worse outcomes. It's a clinical phenomenon that we must navigate and I don't think it's a danger to avoid. Again, as I stated in the introduction at the top of this episode, there's more evidence coming. The pays trial run by Baji Kia and colleagues in Copenhagen is a randomized trial specifically designed to test pain allowance versus pain avoidance in chronic rotator cuff related shoulder pain. The results are pending. They may well refine the picture further, but we don't have all the data to hand at the moment. Let me translate this into how it plays out in the room with a patient. You've got someone with chronic rotator cuff pain. They come in for their second appointment. You ask how the exercises you prescribed went. They tell you the upright row brought on their familiar pain somewhere around four out of 10 in intensity lasted for maybe an hour afterwards after they cease the exercise and then settled by the next morning.
Speaker 2:
They want to know directly if they should stop doing the exercise that you gave them the old answer and even the current answer, depending on the clinician giving the answer would have been some version of if it hurts back off, reduce the weight, reduce the reps, modify the position, wait for the pain to fully settle before progressing. The new answer I'd argue looks very different. So firstly, you sit with them and you ask a few questions, was the pain familiar? Did it stay within a tolerable range, say under five out of 10, whatever that means for the patient. Now, I'm not a huge fan of numbering pain, but sometimes it's easy to do or you could just say, did it feel tolerable to you? Now, did it settle within a reasonable window after the session, say within 24 hours, did the next day's exercise session feel similar or did it feel worse now?
Speaker 2:
Did it affect your sleep at night? Are they functionally normal outside of that exercise window or is it affecting their ability to participate in activities of daily living? How did it make them feel psychologically and emotionally? Now, if the answers to all those questions were directionally yes, yes, it was tolerable, yes, it settled pretty quick. No, it didn't keep me up and no, it's not worrying me, then keep going. The pain during the exercise is not signaling any damage. It might be indicating that the tissue is being challenged in a way that the system is currently sensitive to and that it may even be important to recalibrate the system or change erroneous beliefs. Conversely, if the answers are the pain is unfamiliar, it's escalating, it's persisting well beyond the session, it's keeping me up at night, it's affecting my daily function, that's different. That's a different category.
Speaker 2:
That's something you probably need to look to adjust. The point is that the decision isn't being driven by a dogmatic position on whether pain is allowed or whether it's not. It's being driven by what the pain is doing. It's behavior, it's familiarity, it's intensity, it's duration, and its effect on the person's life. This is a more flexible and clinically useful framework than pain is bad, avoided, or pain is fine. Push into it. Both of those positions impose an outside answer on what should be a context sensitive judgment, and the updated evidence supports that more flexible framework because if it doesn't make a meaningful difference whether the exercise hurts or not, then we can take the question off the pedestal and put it back where it belongs as one variable among many in the clinical conversation. So as I like to do in these episodes, I want to be careful not to overgeneralize.
Speaker 2:
There's enough of that in the industry as it is because I think there's a nuance in the literature that complicates the picture in an interesting way. So let's get into some of the nuance. A few years ago, Mikel Beck Klausen and colleagues over in Denmark, again published a secondary analysis of a trial they'd done on subacromial pain syndrome. The original trial had tested whether adding a large dose of resistance exercise to usual care produced better outcomes. The trial is called the SEXY trial, which is one of the great trial names of all time. Now, the overall answer was no. The extra dose of resistance exercise did not help. But when they looked at who responded and who didn't, they found something interesting. The patient's baseline level of pain catastrophizing, modified the effect of the extra dose of resistance exercise. So what this means is that patients with low pain catastrophizing scores benefited from the extra dose of resistance exercise.
Speaker 2:
Conversely, patients with high pain catastrophizing scores actually trended towards worse outcomes when given more loading. It wasn't just that the average effect was null, it was that the average effect was hiding two opposite effects in different subgroups. Now that finding hasn't been replicated robustly yet, and I don't want to overweight a single secondary analysis. I'm not that naive, but it's interesting and it points to something that I think is important. The decision about how much pain to allow during exercise isn't just about the condition, it's also about the person. So for a patient who's psychologically robust, motivated, not particularly fearful of their pain, and able to interpret bodily sensations without alarm, exposing that group to some pain during exercise is unlikely to do harm and may help them rebuild their relationship with movement and with their body. So the pain becomes information rather than an out and out threat.
Speaker 2:
However, for a patient who is highly catastrophizing, fearful, hypervigilant, interpreting every twinge as a sign of irreparable damage, pushing that person or that group into pain might be counterproductive. It may confirm the fears that they actually came in with. It may erode their trust in you, the clinician, and it may flare them in ways that take longer to settle and reinforce the avoidance pattern you were trying to break in the first place. The clinical implication isn't a hard rule or a universal law that must hold in all conditions. Before you decide how to handle pain during a person's rehab, you need to know who they are, what they believe about their pain, how they interpret their pain, how much fear they're carrying into the exercise program itself. And that's not always easy to assess in a first appointment. It often takes a few sessions of careful listening and reflection, but it shapes what kind of program you build and how much pain you're going to be comfortable allowing for that person.
Speaker 2:
The evidence-based average tells you that pain during exercise doesn't determine the outcome. The clinical reality tells you that for some patients it might. Both things can be true at once. The clinician's job is to figure out which patient is in front of them and tailor their advice to that person. That's the art of being a good clinician. Now, I want to declare a few things we still don't know. We don't know whether the answer differs across conditions. Most of the trials in our latest review pulled across various chronic persistent musculoskeletal conditions that were tendinopathy, low back pain, neo A and neck pain. The theme across these conditions was reasonably consistent, but the underlying mechanisms may be quite different. The PACES trial will help on the shoulder front. Specifically, we need similar trials in other conditions. We don't know the optimal level of pain to permit is three out of 10 different from seven out of 10.
Speaker 2:
Probably most of the trials have used some version of a five out of 10 threshold. That number is essentially a convention rather than a finding. It doesn't hold in every clinical scenario, one person's four out of 10 may be another person's eight out of 10. I understand the limitations there. It may turn out that the precise level matters less than the patient's interpretation of it, but at the moment, we don't have a really robust, reliable, and consistent way of measuring a person's pain experience. It's a real challenge. We don't know whether pairing painful exercise with pain education changes the picture. There's some early work suggesting that the cognitive framing around the pain is important, almost as important as the pain itself, that helping a patient understand what their pain means before exposing them to it may shift how they experience it and how they respond to it.
Speaker 2:
That's an interesting area for further research. We also don't have good adverse event reporting in clinical trials. We rarely know with any precision who flared up, who got worse, who dropped out of the trial because the exercise was too painful. We need to know who is the patient most likely to be harmed by a pain permissive approach to exercise, and how do we identify them in advance. We don't have great answers to that yet. So let me begin to conclude. Now, pain during musculoskeletal rehabilitation is not a red flag. The updated evidence does not support the view that exercising into pain produces worse outcomes than avoiding it. It also does not support the view that exercising into pain produces meaningfully better outcomes. The two approaches appear on average to produce roughly equivalent results. So it means that the decision about whether to permit pain during exercise can be made based on what works for the specific person in front of you rather than on a dogmatic position.
Speaker 2:
Some patients will tolerate pain well and benefit from being exposed to it, others won't. The clinician's job is to read the situation, not to apply a rule. This is the art of being a good clinician. The questions that seem important to me are about what the pain is doing. It's familiarity, it's intensity, it's duration, it's impact on function and life, and about who the patient is, their beliefs, their fears, their interpretive framework. Those are the conversations you need to have, and they're not deficits that you can easily measure, I'm afraid. So pain during exercise is a clinical phenomenon for every musculoskeletal clinicians to navigate. It's a footnote in a larger story, and that story is about the relationship between a person and their body and between a person and a clinician. They trust enough to work through this with. Thanks for listening. I'm Dr. Jerry Powell, and this has been another episode of The Shoulder Physio Podcast. I'll chat to you soon. The Shoulder Physio podcast would like to acknowledge that this episode was recorded from the lands of the Ang people. I also acknowledge the traditional custodians of the lands on which each of you are living, learning, and working from every day. I pay my respects to elders past, present, and emerging, and celebrate the diversity of Aboriginal and Torres Strait Islander peoples and their ongoing cultures and connections to the lands and waters of Australia.
