Each study was assessed for each of these methodological issues eg the study design of prospective cohort was given preference to that of case-control; selection issues such as participation rates / loss of follow-up of subjects in cohorts; histological confirmation of the diagnosis of prostate cancer; whether smoking status was clearly defined and updated over time in cohorts; if the five most likely potential confounders were adjusted for (age, race, family history of prostate cancer, diet, history of vasectomy) - determined from a review of the literature; statistical issues included provision of confidence intervals or p values.