Further comment by the author
Some concern have been raised about the generalizability of our
results due to the fact that women had discontinued their use of any birth
control just prior to enrollment. To the extent that prior birth control
methods were hormonal, there could be residual effects on patterns of
ovulation during the period of observation.
In fact, only 5% of women were using oral contraceptives just prior
to entering the study, and none had been using injectables. (These data
have been provided in detail in earlier papers, e.g. NEJM 319:189-94,
1988.) The three most common methods of birth control just prior to
enrollment were the diaphragm (40%), condom (29%) and NFP (13%).
Thus, we find no reason to believe that the variable ovulation seen
in our study could be an artifact of prior birth control use.
Regarding the exclusion of couples with known fertility problems,
this was to avoid the over-representation of such couples. This is not to
say that all women in our study were fertile. To the contrary, about ten
percent had not conceived within their first year of trying -- similar to
estimates from the general population.
Competing interests: No competing interests