cohorts, 5 nested case-control studies, 1 retrospective cohort and 36 case-control studies
were identified that reported data for prostate cancer (pc) and smoking. The effect of
current smoking on prostate cancer varied from protective – odds ratio 0.38 (Keller and
Howe 1994) to increasing the risk 2.9 times (Daniell 1993). Likewise with past smoking:
from 0.6 (Veierod et al 1997; DeStefani et al 1995) to 1.9 (Hsing et al 1990). The quality
of studies also varied greatly: the hospital-based case-control study by Wynder et al
(1971) considered the poorest while the prospective cohort study of US health
professionals the best (Giovannucci et al 1999).