Background: The Cochrane Database of Systematic Reviews potentially contains valuable evidence to support health care of older people. However its utility depends on a consistent definition of what constitutes an 'older' person.

Objectives: To identify Cochrane reviews and protocols that focus specifically on 'older people' and to describe the variations in criteria used to identify this population.

Methods: We searched the Cochrane Database of Systematic Reviews (CDSR) 2006, Issue 1, using: Aged explode [MeSH] and key words: old$, elderly, adult$, people, veteran. Reviews and protocols were eligible if they included a minimum age limit greater than 40 or stated that their focus was 'older' people. The descriptions of types of participants were extracted from the text.

Results: The search identified 311 reviews and 67 protocols, of which 42 reviews and nine protocols (published between 1998 and 2006) met the criteria, representing 1.6% of reviews and 0.6% of protocols in the CDSR 2006, Issue 1. Twenty-six reviews and all nine protocols stated a minimum chronological age, mode 60 years (n=17), range 40 (n=1) to 65 (n=12). The descriptors that authors used included: 'elderly' (n=12), 'geriatric' (n=8) and 'older adults' (n=8). In addition 18 reviews focused on reduced cognitive function in their descriptors.

Conclusions: There are very few Cochrane reviews or protocols indexed for terms equivalent to 'older people'. There is variation in the descriptive terms used to indicate 'older', and in the age restrictions applied. To optimize the utility of future reviews the development of a consensus of taxonomy for this population is recommended.