Estimation error of Leydig cell numbers in atrophied rat testes due to the assumption of spherical nuclei

In this study, Leydig cell numbers in control and atrophied testes (induced via subcutaneous implants of testosterone plus 17β estradiol for 16 weeks; TE‐implanted) of rats, estimated via the fractionator method (independent of any assumptions) were compared to those estimated via the disector (unbiased, but dependent on shrinkage) and Floderus (assumes spherical particles, dependent on shrinkage) methods. Estimates of Leydig cell numbers in control rats produced by all three stereological methods were similar. In rats with atrophied testes, both the fractionator and the disector methods produced significantly lower (P < 0.01; 47% and 41 % with fractionator and disector, respectively) Leydig cell number estimates per testis than in the controls. By contrast, the estimates of Leydig cell number in atrophied testes derived via the Floderus equation were not significantly different from those of controls, but larger than those obtained via the fractionator and the disector methods. These results suggested that the assumptions of the Floderus method were violated in the atrophied rat testes.