Why Do Some Exercises Cause Pain Initially but Help Long-Term?

The Pain-Exercise Paradox

One of the most common clinical conversations in musculoskeletal rehabilitation involves a patient reporting that the prescribed exercises are causing discomfort — and wondering whether they should stop. The intuitive response — that pain is always a signal to cease the offending activity — is understandable, but it is not always the clinically appropriate one. In a number of specific contexts, exercises that produce a modest, short-term pain response are precisely the exercises most likely to produce long-term resolution of the underlying problem.

This apparent paradox has a coherent biological basis. Pain is produced by the brain in response to a threat appraisal — and the brain's threat assessment of a loading activity depends on multiple factors beyond the immediate nociceptive signal. Understanding the difference between pain that represents genuine tissue harm and pain that is the product of a sensitised nervous system responding to a therapeutic but unfamiliar stimulus is one of the most clinically important distinctions in rehabilitation practice.

The Tendon Loading Example

Tendinopathy provides the clearest and most well-researched example of therapeutic loading that produces an initial pain response. The most effective evidence-based treatments for Achilles and patellar tendinopathy — eccentric heel drops and heavy slow resistance training protocols — consistently produce local tendon discomfort during and after the loading session. Research by Alfredson, Cook, Rio, and others demonstrates that protocols performed within a moderate pain envelope (typically 3–5/10) produce superior tendon structural outcomes and pain resolution compared to strictly pain-free training protocols.

The mechanism is specific to the tendon's biology. Isometric and isotonic loading under a moderate pain stimulus stimulates tenocyte mechanotransduction, drives collagen synthesis, reorganises the extracellular matrix, and activates the cortical inhibitory mechanisms that reduce tendon pain sensitivity over time. Loading that is insufficient to generate this stimulus — typically loading below the tissue's current threshold — does not drive the required adaptation. The temporary discomfort is the cost of the mechanical stimulus required for structural and neurological improvement.

DOMS vs Injury — Knowing the Difference

Delayed onset muscle soreness (DOMS) — the diffuse muscular aching and tenderness that develops 24–72 hours after unaccustomed exercise — is a normal, expected consequence of the early stages of strength training or when exercise load is increased. It reflects the sarcomere disruption and inflammatory response associated with the exercise-induced adaptation stimulus. DOMS is not an injury; it is a sign that the tissue has received a meaningful adaptive stimulus. Its presence does not indicate damage, and training through low-to-moderate DOMS is generally safe and does not impair adaptation.

Distinguishing DOMS from genuine tissue injury is clinically important. DOMS is diffuse, bilateral (when both sides are trained), develops hours after activity rather than during it, peaks at 24–72 hours, and resolves within three to five days. Acute muscle injury produces immediate, unilateral, well-localised pain with specific tenderness and palpable defect or bruising, and is provoked by even submaximal contraction during the acute phase. When in doubt, a clinician should assess the presentation — but DOMS alone is not a reason to suspend an appropriately progressive training programme.

Pain as Threat Appraisal, Not Damage Signal

In sensitised pain states — central sensitisation, fear-avoidance, post-injury hypervigilance — the brain's threat appraisal of loading activity is elevated beyond what tissue status alone would warrant. Movements that are structurally safe produce a pain response disproportionate to any actual tissue stress, because the nervous system is operating in a state of heightened alertness rather than responding purely to peripheral nociceptive input. In this context, exercises that produce a pain response are not necessarily causing harm — the pain is a feature of the sensitised nervous system, not a report of ongoing tissue damage.

Carefully selected exercises that challenge this sensitised threat appraisal — within a context of adequate explanation, graded progression, and clinical monitoring — provide the counter-evidence that progressively reduces the nervous system's overestimation of threat. Each session of completing a challenging exercise without harm reduces the prediction of harm for subsequent sessions, gradually recalibrating the threat appraisal system and reducing the pain response to the same loading stimulus.

Graded Exposure and Desensitisation

The deliberate, progressive exposure to feared or avoided movements — with careful monitoring of the pain response and explicit reassurance — is a clinically structured application of this principle known as graded exposure. Originally developed in the psychological management of anxiety disorders, graded exposure has been systematically applied to musculoskeletal rehabilitation as a tool for reducing movement-related fear and pain. Its mechanism is straightforward: repeated experience of a feared movement without catastrophic consequence updates the nervous system's prediction model for that movement, gradually reducing the threat appraisal and the pain it generates.

The clinical success of graded exposure in chronic pain populations — including those with chronic low back pain, complex regional pain syndrome, and fibromyalgia — provides direct evidence that therapeutic loading which produces an initial pain response can, when appropriately managed, produce progressive desensitisation and lasting pain reduction. The key is the management context: graded exposure conducted with clear explanation, collaborative goal-setting, and close monitoring of response is a structured clinical intervention, not an invitation to push through severe or worsening pain.

Practical Guidelines for Training With Discomfort

Several principles guide the appropriate use of loading that produces an initial discomfort response. Pain during exercise should remain at or below 4–5 out of 10. Pain should not increase as the session progresses. Pain should return to baseline within 24 hours of the session. Weekly reassessment should show a trend toward lower baseline pain and higher load tolerance. The exercise must be specifically selected for the tissue and presentation — not all exercises that produce pain are therapeutic; clinical judgement guided by these principles is required. When these conditions are met, continuing to load through modest discomfort is not merely acceptable — it is clinically indicated.