Measures of semen quality are used as surrogate measures of male fertility
in clinical andrology, reproductive toxicology, epidemiology, and risk asse
ssment, However, only limited data are available to relate those measures t
o fertility. This prospective study with 210 reproductive-age couples was c
onducted to provide information on the value of semen quality measures for
predicting human male fertility potential and for development of models to
estimate the effects of changes in semen quality on fertility in a given po
pulation for risk assessment. Couples without known risk factors for infert
ility and who had discontinued contraception to have a child were accepted.
The study followed each couple for up to 12 menstrual cycles while they at
tempted to conceive and evaluated semen quality measures from multiple ejac
ulates per man with known abstinence intervals. For each cycle, the day of
ovulation was predicted, and the couple was advised to have intercourse mul
tiple times on that day and on the days around it. Among the demographic va
riables assessed, parity, contraception status prior to entering the study,
male education level, and male smoking were associated significantly with
12-cycle pregnancy rate. Several semen quality measures also were associate
d significantly with pregnancy rate, with percentage morphologically normal
sperm by strict criteria and measures involving total number of sperm show
ing particularly strong associations. Localized regression-smoothing plots
of semen quality data against proportion of couples pregnant suggested leve
ls below which fertility declines for several semen quality measures. These
results have applications in both clinical andrology and in assessment of
risk to male fecundity from environmental or pharmaceutical exposures. In p
articular, they contribute information on behavior of fertility with varyin
g semen quality and can allow development of models to predict effects on f
ertility in populations from decrements in semen quality.