Premium
Semen quality and waiting time to pregnancy explored using association mining
Author(s) -
Romero Herrera Jose Alejandro,
Bang Anne Kirstine,
Priskorn Lærke,
Izarzugaza Jose M. G.,
Brunak Søren,
Jørgensen Niels
Publication year - 2021
Publication title -
andrology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.947
H-Index - 43
eISSN - 2047-2927
pISSN - 2047-2919
DOI - 10.1111/andr.12924
Subject(s) - semen , semen quality , sperm , andrology , sperm motility , biology , sperm quality , gynecology , medicine
Background Assessment of semen quality is a key pillar in the evaluation of men from infertile couples. Usually, semen parameters are interpreted individually because the interactions between parameters are difficult to account for. Objectives To determine how combinations of classical semen parameters and female partner age were associated with waiting time to pregnancy (TTP). Materials and methods Semen results of 500 fertile men, information of TTP, and partner age were used for regressions and to detect breaking points. For a modified Association Rule Mining algorithm, semen parameters were categorized as High, Medium, and Low. Results Men ≤32.1 years and women ≤32.9 years had shorter TTP than older. Decreasing TTP was associated with increasing level of individual semen parameters up to threshold values: sperm concentration 46 mill/mL, total sperm count 179 mill, progressive motility 63%, and normal morphology 11.5%. Using association mining, approximately 100 combinations of semen parameters and partner age were associated with TTP. TTP ≤ 1 month often co‐occurred with high percentages of progressive motility (≥62%) and morphologically normal spermatozoa (≥10.5%). Furthermore, TTP ≤ 1 did not tend to appear with lower percentages of these two semen parameters or high partner age (≥32 years). However, high percentages of motile or normal spermatozoa could not compensate for sperm concentration ≤42 mill/mL or total sperm count ≤158 mill. The prolonging effect of high partner age could not be compensated for by the man's semen quality. Discussion and Conclusion Using association mining, we observed that TTP was best predicted when combinations of semen parameters were accounted for. Sperm counts, motility, and morphology were all important, and no single semen parameter was inferior. Additionally, female age above 32 years had a negative impact on TTP that could not be compensated for by high semen parameters of the man.