Response to comment on 'Interventions to improve exercise behaviour in sedentary people living with and beyond cancer: a systematic review'.

Sir,

We thank Broderick for their interest in our manuscript. We agree there seems to be uncertainty in the terminology around exercise interventions, which has not been helped by the implication of a distinction between the terms ‘physical activity' and ‘exercise' in some publications (Chodzko-Zajko ). A more constructive approach would appear to be that taken by Winter and Fowler, (2009) in defining and quantifying exercise according to the Systeme International d'Unites. As stated in the review, this is the approach we chose to follow, for several reasons:

An important aspect of a systematic review is to summarise the current evidence base with a view to identifying prospective research priorities or furthering practice. In this context, making a distinction between physical activity and exercise is unhelpful, particularly in evaluating clinical effectiveness. As in other clinical trials, most notably drug studies, objective documentation of the amount of the intervention that is delivered is imperative to identify dose–response curves and adverse effects. The review criteria were set to include only studies reporting such objective metrics (i.e., frequency, intensity and duration) so as to facilitate reproducibility of the intervention. Any systematic review of cancer therapies will clearly identify the target population and objectively define the intervention; exercise is no different, if we are to take its use as a therapeutic intervention seriously.

The term ‘sedentary behaviour' is open to uncertainty, as considerable exercise may be taking place in a sedentary (i.e., seated) position—for example, rowing or cycling. We agree this needs defining, but to do so by subdividing exercise into different terms, seems counterproductive. Sedentary behaviour could be defined as anyone not achieving the recommendation to take 150 min per week of moderate-intensity aerobic exercise (Rock ). However, in clinical trials involving cancer cohorts, this has frequently been defined as <90, or even <60 min per week (Pinto ; Daley ; Cadmus ; Bourke ). The rationale being that there should be some scope to induce a clinically meaningful benefit from participation in the intervention. If there is not, why do we need to intervene? Nevertheless, as we concluded in the review, most individuals living with or beyond cancer would currently find current guideline targets unachievable, certainly with current published interventions. More research or a revision of the one size fits all approach is warranted. We conclude in the review that a ‘dose response' might be more appropriate. This would include the suggested ‘potential end-points at the lower end of the physical activity continuum.' Such recommendations might require elucidation by further data collection or an individual patient data meta-analysis for any given health outcome.

The tendency of epidemiological studies to imply a distinction between exercise and physical activity reflects imprecision of measurement rather than any fundamental difference. Subjective metrics of exercise behaviour, for example, metabolic equivalents derived from questionnaires (Ainsworth ), are often used in these reports and we agree that objective measurement of exercise behaviour are preferable in clinical trials. Dedicated accelerometers seem an overly expensive option. Less expensive alternatives such as smartphone applications or simple heart rate monitors would be welcome where the technology is available, affordable and contextually appropriate. For individuals on prescribed medication (e.g., β-blockers) that impact the cardiovascular response to physical exertion, reliable measurement of dose–response remains a challenge.

Finally, as Jan Swammerdam's 17th century experiments demonstrated rather elegantly (Needham, 1971; Winter and Fowler, 2009), skeletal muscles neither ‘contract' (i.e., reduce in volume) nor expand significantly during exercise. Furthermore, movement is neither essential nor necessary (as in isometric activity) for exercise to be taking place. What is fundamental to exercise, in the context being discussed, is skeletal muscular activity exerting force and generating a metabolic response i.e. physical activity by a different name.

We look forward to reaching consensus on the role of defined exercise interventions in the treatment of a number of cancers and agree with Broderick et al that consensus on terminology is an essential first step. We would encourage all practitioners in this area to follow the excellent recommendations of Winter and Fowler, which perpetuate reproducibility rather than confusion.