Frequency of the stages in the cycle of the seminiferous epithelium in the rat

This study determined the optimum number of tubules to be counted per testis cross section, and the number of animals per treatment group, when changes in stage frequencies in the cycle of the seminiferous epithelium are criteria for assessing effects of treatment on spermatogenesis. A data base of 9,672 observed and staged tubules was collected from testicular cross sections of 15 Sprague-Dawley rats. A significant variation between animals was found for the frequencies of Stages I, II, IV, VI, VIII, and XIII. Computer simulation was used to randomly select different combinations of animal and tubule numbers from the observed data. Stage frequency means from each simulation experiment were compared statistically to observed mean frequencies. A model that used data from all 14 stages was analyzed. The following conclusions were made: a) a minimum of 200 tubule cross sections/testis is recommended for estimating stage frequencies; b) for a fixed number of tubules scored, the number of animals sampled is more important than the number of tubules per animal in reducing variance; c) to detect a difference of 2 standard deviations from the mean with a 2% error rate and examining 200 tubules/testis, at least 12 animals must be used per group when assessing all 14 stages; d) when individual stages are examined using 10 animals per group, only Stage VII has 80% or greater power of test (alpha = 0.05) to detect a frequency difference; e) pooling stages into 3-4 groups is recommended to improve the power of detecting a treatment difference.