If the number of pregnancies are grouped so that three or more pregnancies are placed in the same category, as in most previous studies (Byers et al., 1982; McMichael & Potter, 1984; Potter & McMichael, 1983; Weiss et al., 1981), there is a highly significant protective trend with increasing numbers of pregnancies (P = 0.0007) (Table II).
Of the previous studies described in Table VII, three reported increasing risk with age at first livebirth (Kune et al., 1989; Potter & McMichael, 1983) or age at first pregnancy (Howe et al., 1985); but three others (Miller et al., 1980; Weiss et al., 1981; Wu et al., 1987) reported no effects. In the present study, there was a weak trend for increasing protection from proximal colon cancer with increasing age at natural menopause (Table VII).