Machine Learning Predictive Models for Evaluating Risk Factors Affecting Sperm Count: Predictions Based on Health...

Area of Science:

• Reproductive Health
• Biostatistics
• Machine Learning

Background:

• Global decline in fertility rates, particularly in developed nations.
• Taiwan faces a record low fertility rate, prompting government interventions.
• Older age at marriage is linked to declining physical status and potential impacts on gamete quality.

Purpose of the Study:

• To identify major risk factors affecting male sperm count using machine learning.
• To analyze health screening data from a general male population in Taiwan.
• To explore novel risk factors beyond traditional associations with male fertility.

Main Methods:

• Utilized five machine learning algorithms: random forest, stochastic gradient boosting, least absolute shrinkage and selection operator regression, ridge regression, and extreme gradient boosting.
• Analyzed annual health screening data of 1375 males from 2010 to 2017.
• Employed metrics such as symmetric mean absolute percentage error and root mean squared error for performance evaluation.

Main Results:

• Identified sleep time (ST), alpha-fetoprotein (AFP), body fat (BF), systolic blood pressure (SBP), and blood urea nitrogen (BUN) as top five risk factors for sperm count.
• Confirmed negative impacts of metabolic syndrome indicators (BF, SBP) on sperm count.
• Highlighted alpha-fetoprotein (AFP) and blood urea nitrogen (BUN) as novel risk factors for male fertility.

Conclusions:

• Machine learning models effectively identify complex relationships influencing sperm count.
• Findings support the role of metabolic syndrome and sleep duration in male reproductive health.
• Identified novel risk factors (AFP, BUN) provide new avenues for research and public health strategies to address declining fertility rates.