To the Editor: In their prospective study,
Dr Leitzmann and colleagues1 replicated
the findings of our case-control study2 in
regard to a protective relationship between frequency of ejaculation and risk
of prostate cancer. Although the study of Leitzmann et al avoids the potential
for recall bias, for both their study and ours the accuracy of ejaculation
assessment was likely to be affected by problems in recall of past practice,
particularly of that in early adult life. We believe that the effect of such
a bias would tend to be conservative and would have thus underestimated any
real effect for ejaculation.