Investigating change across time in prevalence or association: the challenges of cross-study comparative research and possible solutions