Table 1. Details of the 14 studies included in the meta-analysis.
Study first author Year (Country) | Study design | Period of diagnosis | Time of weight estimate |
Number of |
Adjustments |
||||
---|---|---|---|---|---|---|---|---|---|
Sex | Cases | Controls | Proxies | Smoking | Diabetes | ||||
Bueno de Mesquita 1990 (Netherlands) | C–C | 1984–1988 | 2 years before interview | M | 90 | 232 | Yes | Yes | No |
F | 74 | 248 | |||||||
Howe 1990 (Canada) | C–C | 1983–1986 | 2 years before interview | M | 141 | 270 | Yes | No | No |
F | 108 | 235 | |||||||
Ghadirian 1991 (Canada) | C–C | 1984–1988 | 2 years before interview | M, F | 179 | 239 | Yes | Yes | No |
Zatonski 1991 (Poland) | C–C | 1985–1988 | Not known | M, F | 110 | 195 | Yes | No | No |
Friedman 1993 (USA) | Cohort | 1945–1988 | Prospective (measured) | M, F | 450 | 2687 | NA | Yes | No |
Shibata 1994 (USA) | Cohort | 1981–1990 | Prospective | M, F | 65 | NA | NA | Yes | No |
Silverman 1998 (USA) | C–C | 1986–1989 | Usual in adulthood | M | 218 | 1231 | No | Yes | Yes |
F | 213 | 747 | |||||||
Gapstur 2000 (USA) | Cohort | 1967–1995 | Prospective (measured) | M | 96 | NA | NA | Yes | Yes |
F | 43 | NA | |||||||
Michaud 2001 (USA) | Cohort | 1976–1998 | Prospective | M | 140 | NA | NA | Yes | Yes |
F | 210 | NA | |||||||
Hanley 2001 (Canada) | C–C | 1994–1997 | 2 years before interview | M | 173 | 1074 | No | Yes | No |
F | 139 | 1191 | |||||||
Stolzenberg-Solomon 2002 (Finland) | Cohort | 1985–1997 | Prospective (measured) | M | 172 | NA | NA | NA | Yes |
Lee 2003 (USA) | Cohort | 1962–1995 | Prospective | M, F | 212 | NA | NA | NA | Yes |
Calle 2003 (USA) | Cohort | 1982–1998 | Prospective | M | 1908 | NA | NA | Yes | No |
F | 1650 | NA |
NA=not applicable; C–C=case–control study; M=males; F=females.
Michaud (2001) published the results of two cohort studies in one paper. The male subjects were one cohort study (Health Professionals' Follow-up study) and the female subjects were the other (Nurses' Health Study).