Authors' reply

摘要

Philipe de Souto Barreto raises two important points: first, was the Geriatric Depression Scale (GDS)-15 the most appropriate measure for this study? And second, was an adequate amount of exercise delivered? For a large pragmatic trial, the GDS-15 is the most appropriate and practical primary outcome. The assessment is shorter than competing measures, it is usable by research staff without specialised training, it does not require carer interviews, it does not include somatic symptoms, and it has recognised measurement properties.1 While some reports have suggested lower internal consistency for GDS-15 in cognitively impaired care home residents, others suggest strong internal consistency and consistency with external measures of depression even in more cognitively impaired residents.