Design: Multicentre, randomised, open-label trial
Population: 700 men on androgen deprivation therapy for prostate cancer
Primary outcome: Quality of life (FACT-P) and fatigue (FACIT-F) at 12 months
Effect: FACT-P adjusted mean difference 4.5 (p=0.0004); FACIT-F 1.9 (p=0.0068)

Androgen deprivation therapy controls prostate cancer but causes fatigue, metabolic change and loss of quality of life, and although guidelines recommend exercise, it is rarely built into care. STAMINA tested an embedded lifestyle programme across 15 NHS trusts, randomising 700 men on ADT.

The intervention, twelve months of supervised aerobic and resistance exercise with dietary and behavioural support, was compared against an optimised usual-care arm that itself included clinician training and exercise advice. It was superior on both primary outcomes: prostate-specific quality of life (FACT-P adjusted mean difference 4.5) and fatigue (FACIT-F difference 1.9). Serious intervention-related events were rare and all resolved.

The practical point is that a structured, supervised programme beats advice alone, even good advice, for men on ADT. The effect sizes are modest but meaningful, and the trial shows such a programme can be delivered within a health service rather than remaining an aspiration in guidelines.

Multicentre trial of 700 men on androgen deprivation therapy across 15 NHS trusts.
Twelve months of supervised exercise with dietary and behavioural support.
Superior to optimised usual care on quality of life (FACT-P difference 4.5).
Also superior on fatigue (FACIT-F difference 1.9).
Serious intervention-related events were rare and all resolved.

Why it matters

It shows supervised delivery, not just exercise advice, is what improves quality of life on ADT, and that it is deliverable in practice.

The statistics, in plain English

The FACT-P difference of 4.5 points is around the threshold often considered clinically meaningful for this scale, and it held against a demanding comparator that already included exercise advice, making the supervised element the likely driver.