What if my exercises hurt?
Finding the right balance between challenge and comfort
Tim Cook - Physiotherapist, lecturer and researcher
Tim Cook - Physiotherapist, lecturer and researcher
“My physiotherapist has given me exercises to help—but what if they hurt?”
It is an understandable question. When your shoulder already aches, your knee complains on the stairs or your back makes everyday tasks difficult, deliberately doing more can feel counterintuitive.
Exercise can help with many musculoskeletal problems, supporting function, confidence and participation in activities that matter. But recommending exercise is only the beginning. Finding a manageable starting point—and deciding what to do when it hurts—is part of the treatment.
Ideally, exercise should feel comfortable but challenging. That might involve working against varied resistance, for multiple repetitions over time, or simply learning a new movement.
Challenging does not have to mean painful. Pain is not proof that exercise is working.
Three questions often become muddled: does exercise help; can exercise that allows pain help; and is painful exercise better than pain-free exercise? These are different questions.
Babatunde and colleagues’ overview of treatments for common musculoskeletal problems supported exercise therapy, while recognising uncertainty about the best dose and delivery. It also identified recommendations for self-management advice and education, although supporting evidence was limited.[1] Self-management should mean having useful skills and support, rather than being left to manage alone.
For persistent low back pain, Rathnayake and colleagues reviewed nine trials of self-management programmes incorporating exercise. They found small average improvements in pain and disability compared with usual-care.[2] This supports the combined approach; it does not establish that exercising into pain produces those benefits.
An influential 2017 review by Smith and colleagues found a small short-term pain benefit from programmes allowing or provoking pain compared with pain-free programmes. There was no clear overall superiority over the medium or long term.[3] However, an updated 2025 review by Tran and colleagues included 16 trials and found no clear difference in pain or disability between painful and nonpainful exercise across follow-up periods. All trials were judged at high risk of bias, and evidence certainty was low or very low.[4]
The practical message is that pain need not always prevent improvement, but deliberately producing pain has no established general advantage. Uncertainty also means we should avoid claiming the approaches are definitively equivalent for everyone.
In a small Achilles tendinopathy trial, Silbernagel and colleagues compared continuing running and jumping under a pain-monitoring model with stopping these activities for six weeks. Both groups followed the same rehabilitation programme. Both improved, with no significant between-group difference in improvement.[5]
The model allowed pain up to 5/10 during and immediately after activity, with restrictions on next-morning symptoms and no progressive weekly increase in pain or stiffness. These were study-specific guidelines, not a universal safety threshold.
This supports monitored continuation of activity in selected patients. It was not a direct trial proving painful strengthening superior to pain-free strengthening, nor a guarantee against injury.[5]
For osteoarthritis, NICE recommends exercise tailored to the person and explicitly advises that joint pain may initially increase. Regular exercise can nevertheless improve pain, function and quality of life.[6] This applies to managing conditions such as knee and hip osteoarthritis; pain-free movement is not a prerequisite for starting exercise.
Research on longer-term knee osteoarthritis exercise programmes found no detectable worsening of radiographic disease severity or cartilage morphology compared with no exercise.[7] That is reassuring about appropriately prescribed exercise. It does not show that every painful activity, at any intensity, is harmless.
Pain is real, but its intensity is not a direct reading of how much tissue damage is present. The International Association for the Study of Pain recognises that biological, psychological and social factors influence pain.[8]
Tissue loading and the condition itself matter - So can sensitivity, previous experiences and the wider circumstances in which someone is moving. Persistent pain should not be dismissed as imaginary or “just sensitivity”. Equally, familiar discomfort does not automatically tell us that an exercise is causing further damage.
Context remains essential. A rehabilitation plan for persistent pain may differ considerably from one following a fracture, recent operation or acute injury with specific restrictions.
The useful question is: what does this response mean for you, with this condition, at this stage?
Deciding whether to continue with discomfort or ease back can be one of the trickier parts of rehabilitation. The following are practical options to discuss, not three independently validated treatments or steps everyone must follow.
Within an agreed plan, you might choose to continue with familiar, manageable discomfort while concentrating on what you are doing:
“Despite your familiar discomfort, focus on the movement or task.”
For example, during an agreed sit-to-stand exercise, your attention might be on standing smoothly and confidently rather than repeatedly checking how a person's back and knee pain feel.
This does not mean ignoring symptoms. Notice the overall response, and avoid forcing through escalating pain or dismissing new symptoms. The aim is to give movement a chance without making every sensation the sole judge of success. Over time the aim is to encourage the body to adapt to this stimulus and for the pain to naturally begin to fade.
Some people prefer a more structured check during and after exercise:
Does discomfort ease after stopping or modifying the exercise?
Are everyday activities about as manageable later that day?
Is sleep no more disturbed than usual?
Are symptoms returning towards their usual level by the following day, without significant worsening over the next 24–48 hours?
These are questions to personalise with your physiotherapist, not universal safety rules or an exact reproduction of Silbernagel’s protocol.
For example, a shoulder exercise might cause manageable discomfort during the session, but leave dressing and sleep unchanged afterwards. That provides useful information for the next session. If symptoms remain substantially increased, fewer repetitions, less resistance, a smaller movement or more recovery time may make the exercise more manageable.
A pain score can help describe the experience, but everyone's number will be different. Sometimes it can be more helpful to classify pain during exercise as either tolerable or intolerable.
You can also choose an amount of movement that does not bring on your familiar symptoms.
If you already know that discomfort usually starts after a particular walking duration, begin with a shorter, manageable walk. You do not need to deliberately trigger pain to establish a starting point.
Comfortable pain-free exercise is a valid choice. It can make a plan easier to repeat, particularly when repeated flare-ups have undermined confidence. Experiencing pain-free movement helps to de-threaten that exercise. It need not become a permanent limit: review and build the activity as tolerance and confidence change.
Your body can adapt to new challenges, but it often helps to introduce them gradually. Doing much more than usual—particularly after a quieter period—may leave you more sore than you need to be.
Progressive loading is central to rehabilitation, while unfamiliar exercise is a recognised trigger for muscle soreness.[9,10] Neither provides a formula that fits every person or condition.
“More” can mean heavier resistance, extra repetitions, longer sessions or exercising more often. It can also mean adding rehabilitation on top of gardening, work and a busy weekend.
A practical approach is to start with an amount you can repeat, then build from there. When returning to walking, that might mean several manageable outings before attempting a much longer walk on the first good day.
Changing one aspect at a time can make the response easier to interpret. Allow recovery between demanding sessions, and agree progression with your physiotherapist. These are practical strategies, not proven universal rules: there is no single percentage increase or timetable that everyone needs to follow.
Gradual progression does not mean avoiding effort or waiting until every symptom disappears. It means finding a challenge you can sustain and develop.
After unfamiliar exercise, you may notice a new ache or tenderness in the muscles you have used. This is often called delayed-onset muscle soreness, or DOMS.
It typically develops after the session, becomes most noticeable over the next one to three days and then eases, commonly within about a week.[10,11] It involves a complex response to unfamiliar loading, including local tissue and inflammatory changes.
This may feel different from your usual knee, shoulder or back symptoms. Think about where the ache is, whether those muscles did unfamiliar work and whether it is settling. Sometimes the distinction is unclear, and both can occur together.
Soreness is not required for progress and does not prove that the exercise dose was right. Muscle adaptation cannot be judged simply by how sore you feel afterwards.[12]
If soreness makes the next session difficult, adjust the amount or allow more recovery. Seek advice about unusually severe, unexpected or non-settling symptoms rather than automatically labelling them DOMS.
An increase in familiar symptoms may mean the exercise amount needs adjusting; it does not automatically establish new damage. Nevertheless, a flare-up can be distressing and disrupt sleep, work or family life. It deserves attention.
Temporarily easing back, changing the exercise or reviewing its place among your other activities may help you find a more manageable plan. Repeated or prolonged flare-ups are a reason to review together. New, severe or distinctly different symptoms warrant assessment.
Individualisation means taking account of your goals, preferences, confidence, condition and response. It does not require claiming that every individually tailored programme outperforms every standard programme.
Notice progress beyond pain scores: can you walk further, manage stairs more confidently, return to gardening or exercise more consistently?
There is no single correct amount of pain for everyone. Comfortable exercise and exercise allowing manageable discomfort can both have a place.
Start small. Find the right challenge. Notice your response. Adjust together.
This article supports, rather than replaces, individual assessment and advice.
Babatunde OO, et al. Effective treatment options for musculoskeletal pain in primary care: A systematic overview of current evidence. PLOS ONE. 2017;12:e0178621. doi:10.1371/journal.pone.0178621
Rathnayake APS, Sparkes V, Sheeran L. What is the effect of low back pain self-management interventions with exercise components added? A systematic review with meta-analysis. Musculoskeletal Science and Practice. 2021;56:102469. doi:10.1016/j.msksp.2021.102469
Smith BE, et al. Should exercises be painful in the management of chronic musculoskeletal pain? A systematic review and meta-analysis. British Journal of Sports Medicine. 2017;51:1679–1687. doi:10.1136/bjsports-2016-097383
Tran I, et al. Effectiveness of painful versus nonpainful exercise on pain intensity, disability, and other patient-reported outcomes in adults with chronic musculoskeletal pain: An updated systematic review with meta-analysis. JOSPT. 2025. doi:10.2519/jospt.2025.13253
Silbernagel KG, Thomeé R, Eriksson BI, Karlsson J. Continued sports activity, using a pain-monitoring model, during rehabilitation in patients with Achilles tendinopathy: A randomized controlled study. American Journal of Sports Medicine. 2007;35:897–906. doi:10.1177/0363546506298279
NICE. Osteoarthritis in over 16s: diagnosis and management. NG226. 2022. Recommendations
Effects of long-term exercise therapy on knee joint structure in people with knee osteoarthritis: A systematic review and meta-analysis. Seminars in Arthritis and Rheumatism. 2019;48:941–949. PubMed
International Association for the Study of Pain. IASP announces revised definition of pain. 2020. Definition and explanatory notes
Silbernagel KG, et al. Current clinical concepts: Conservative management of Achilles tendinopathy. Journal of Athletic Training. 2020;55:438–447. doi:10.4085/1062-6050-356-19
Hotfiel T, et al. Advances in delayed-onset muscle soreness (DOMS): Part I: Pathogenesis and diagnostics. Sportverletzung Sportschaden. 2018;32:243–250. doi:10.1055/a-0753-1884
Cheung K, Hume P, Maxwell L. Delayed onset muscle soreness: Treatment strategies and performance factors. Sports Medicine. 2003;33:145–164. PubMed
Damas F, et al. Resistance training-induced changes in integrated myofibrillar protein synthesis are related to hypertrophy only after attenuation of muscle damage. Journal of Physiology. 2016. PubMed
This article provides general education based on evidence and professional interpretation. It does not replace individual assessment or treatment. For personal advice, consult an HCPC-registered physiotherapist or another qualified healthcare professional. Clinicians should apply their own clinical judgement.
Tim Cook © 2026 Physio Principles