The categorisation of smoking varied and where possible
currently smoking vs never was used as the contrast of
interest. There was mixed evidence for
smoking with the 18 studies varying between having no effect on future knee
problems and a protective effect
contrast
interest.