Comparison of soy isoflavone-breast cancer association among studies by study design (cohort or nested case-control versus. retrospective case-control studies).

$Pooled OR of cohort or nested case-control studies.&Pooled OR of retrospective case-control studies.Ψ The comparison odds ratio, is the ratio of OR1 versus OR2. ORC>1 implies that studies employed prospective design (cohort or nested case-control study) exhibited a weaker protective association than studies employed retrospective design (case-control study). Conversely, ORC**P values (two-sided) were calculated using the Z-statistic [7].