Table 2. Summary of the cohort-based studies included in the analysis.

Author	Date	Place	Race	Mean age at diagnosis	Design	Cohort
Brandt6	1961–2006	Sweden	Caucasian	NA, aged younger than 75	Cohort study	3,900,000 men from the Swedish cancer registry
Cerhan78	1987–1995	USA, Iowa	Caucasian	73,6 A	Prospective cohort study	1557 population-based controls from case-control study in Iowa from 1987–1989
Chen49	1990–2004	USA	Caucasian	NA	Prospective cohort study B	43494 men from the Health Professionals Follow-Up Study cohort
Kalish79	1987–1997	USA, Boston area	Caucasian	65,2 A	Retrospective cohort study	1149 men from the Massachusetts Male Aging Cohort
Kerber7	1966–2000	USA, Utah	Caucasian	NA	Nested case-control, density samping	11572 cases and 11572 controlsfrom the Utah Population Database
Park80	1993–2006	USA, California	Mixed	NA	Nested case-control, cumulative sampling	729 cases and 729 controls from the Multiethnic Cohort
Schuurman81	1986–1992	The Netherlands	Caucasian	64,2 A	Prospective cohort study	52879 men from the Municipal population registries
Sun82	1992–2006	USA	Caucasian	NA	Nested case-control, density sampling	1157 cases and 1157 controls from the Prostate Lung and Ovarian Cancer Screening Trial cohort

^ACalculated from the reported distribution of age of the cases at diagnosis.

^BData on exposure to family history was available at baseline, however the updated data from 1996 was used in the analysis.