Can intervention modify adverse lifestyle variables in a rheumatoid population? Results of a pilot study