Exercise, manual therapy and rehabilitation dosage

Why rehabilitation is difficult to reduce to a protocol

Frozen shoulder changes over time, but not according to a reliable calendar. Pain, stiffness, sleep disturbance, confidence and functional capacity can improve at different rates. Two people six months after onset may require very different starting doses: one may still have marked resting and night pain, while another is mainly limited by stiffness and deconditioning.

Research adds a second problem. “Physiotherapy” may mean advice and home exercise in one trial, supervised stretching and mobilisation in another, or a package containing electrotherapy, manual therapy and exercise in a third. Programme duration, adherence, co-interventions and diagnostic criteria vary. When pooled results are uncertain, clinicians should resist converting a favoured technique into a universal rule.

The practical objective is therefore not to reproduce a named protocol. It is to create a programme with a clear purpose, a documented starting dose, a response rule and a review point.

Begin with a rehabilitation profile

Stage terminology can aid communication, but irritability and function are usually more actionable. Establish the behaviour of pain at rest and at night, where pain begins during movement, how long symptoms take to settle after activity, the amount of movement available, and which tasks the patient most wants to regain.

A person with frequent night pain, pain before end range and prolonged reactions after simple activity has a different dosage ceiling from someone whose pain is low and whose main limitation is a stable mechanical restriction. This does not prove that they occupy different biological stages; it identifies different tolerances today.

A symptom-response dosage framework

Dosage includes more than repetitions. It includes movement range, external load, leverage, speed, hold time, volume, frequency, rest and the number of new variables introduced at once. A modest exercise can become a high dose when performed at end range, for long holds, many times a day.

A response rule should be agreed before the programme begins. “Do what you can tolerate” is too vague. Define what level of discomfort may be acceptable, how quickly it should settle and which changes require contact or review. There is no validated universal next-day rule for frozen shoulder; thresholds must be individualised.

Exercise categories and their clinical purpose

No exercise is mandatory. A pendulum exercise performed with tension and repeated into severe pain may be less suitable than a comfortable active reach. Conversely, strength work should not be withheld automatically until “full range” returns. If a person can generate tolerable effort in an available range, graded loading may support function and confidence while mobility continues to change.

Stretching: distinguish exposure from provocation

Stretching can be described by direction, intensity, duration and frequency, but research does not define one optimal frozen shoulder prescription. The clinical question is whether a particular dose increases useful movement or tolerance without producing a counterproductive flare.

High-irritability presentations often justify comfortable, shorter-range exposure and fewer simultaneous exercises. When resting and night pain recede and movement resistance becomes more dominant, longer holds or greater end-range exposure may be trialled. “More intense” is not synonymous with “more effective.” Forceful treatment that repeatedly worsens sleep or reduces daily use may increase total disability even if a temporary range change is measured immediately afterwards.

Manual therapy: an optional adjunct, not the programme

Manual therapy can include glenohumeral mobilisation, movement with mobilisation, soft-tissue techniques and treatment of adjacent regions. Some patients report short-term pain relief or easier movement; others find hands-on treatment provocative. The evidence does not establish one essential technique or support a claim that the capsule must be mechanically “released” by routine manual treatment.

A more useful framing is to ask whether the intervention creates a worthwhile window. Does it reduce pain enough to sleep, improve confidence, permit a meaningful task or make active rehabilitation easier? If the only outcome is a short-lived passive range change that does not affect function, the value of repeated treatment should be questioned.

What current evidence can and cannot tell us

The Cochrane review of manual therapy and exercise included 32 trials involving 1,836 participants, but intervention heterogeneity limits direct prescription. Moderate-certainty evidence from a key comparison suggested that manual therapy plus exercise produced less short-term improvement than glucocorticoid injection at around seven weeks. That finding should not be interpreted as exercise being unhelpful across the whole course; it shows that treatment effects depend on comparator and timeframe.

The 2025 clinical practice guideline supports manual therapy and exercise while grading much of the evidence as low or very low certainty. It also indicates that combining interventions may be reasonable, including exercise after injection. The UK FROST trial compared early structured physiotherapy with manipulation under anaesthesia and arthroscopic capsular release in secondary care. At 12 months none of the three strategies was clinically superior; arthroscopic release carried higher risks and costs. This does not establish a universally preferred first-line rehabilitation dose, but it reinforces the importance of shared decision-making and proportional escalation.

Strength, conditioning and return to activity

Long periods of pain and protection can reduce strength, endurance and confidence beyond the glenohumeral restriction itself. Rehabilitation should eventually prepare the person for the loads that matter: carrying shopping, lifting a child, manual work, swimming, racquet sport or overhead training. Range alone does not establish readiness.

Progressive resistance can begin within available tolerable movement and expand as capacity improves. Consider intensity, repetitions, proximity to fatigue, weekly frequency and the demands of the target activity. The shoulder does not need to be symptom-free before all conditioning begins, but a programme should not conceal deterioration behind the instruction to “push through.”

Outcome measurement and progression

Use outcomes that can change independently: pain, sleep, patient-selected function, disability score, active movement, passive movement and load tolerance. A programme may be working when the same task causes less pain, when symptoms settle faster, when movement increases at the same irritability, or when a person performs more load without losing sleep.

Communication can alter the rehabilitation dose

Language such as “tear the adhesions,” “no pain, no gain” or “your shoulder will seize unless you force it” may increase threat and encourage excessive loading. Equally, telling a patient never to move into discomfort may reinforce avoidance. Explain the purpose of each component, the uncertainty in the evidence and the response that should trigger adjustment.

Adherence problems are not always motivation problems. A programme may be too long, too painful, poorly connected to goals or incompatible with work and caregiving. A smaller programme completed consistently and reviewed intelligently can be more useful than an idealised list that is abandoned.

Implications for professional practice

1. Prescribe a purpose, not merely an exercise. Every component should connect to pain, sleep, movement, load capacity or a valued task.
2. Write the response rule. Define acceptable symptoms, expected recovery and triggers for reduction or reassessment.
3. Change one major variable at a time. This makes response easier to interpret.
4. Use manual therapy conditionally. Retain it when it creates a useful window; reconsider it when gains are brief and do not transfer.
5. Progress beyond mobility. Restore strength, endurance and task capacity as irritability allows.
6. Revisit the diagnosis when the trajectory changes. New weakness, neurological features, trauma, systemic symptoms or unexpected deterioration require more than a dosage adjustment.