Statistical Methods for Assessing the Health of Sperm

Environmental Health Sciences

Investigators

Linked publications & trials

Abstract

We are developing improved methods for assessing the health of sperm, both for application to reproductive toxicology studies and for development of improved biomarkers for human male fertility. Work has progressed in two areas: (1) developing new methods for characterizing the short-term movement of individual sperm and (2) developing methods for comparing samples of populations of sperm, i.e. sets of specimensProper characterization of the motion of sperm is needed to detect and interpret effects of male exposures to reproductive toxicants. Large numbers of sperm can be individually tracked using computer-assisted semen analysis (CASA), which records X-Y coordinates of the planar projection of the sperm's path over approximately one second, and outputs seven summary measures describing that track. Our method reduces the required number of measures to three, which includes assessment of the predictability of the path. Applied to a toxicologic data set, our three measures discriminated between normal and abnormal sperm as well as did the seven CASA measures. Moreover, our predictability measure significantly improved discrimination between treated and untreated sperm when added to a model containing all the CASA measures.Many individual sperm are studied for each animal/specimen in an study. We have developed a method for jointly analyzing the sperm-specific kinematic measures and sperm survival. The method can be used to characterize effects of spermatotoxicants and to derive a summary measure that scores the progressive motility of an individual's sperm.