Woldemariam Sarah R, Xie Feng, Roldan Alennie, Roger Jacquelyn, Tang Alice S, Oskotsky Tomiko T, Stevenson David K, Lathi Ruth B, Rajkovic Aleksandar, Allen Isabel E, Aghaeepour Nima, Eisenberg Michael, Sirota Marina
Bakar Computational Health Sciences Institute, University of California San Francisco, San Francisco, CA, USA.
Department of Anesthesiology, Perioperative, and Pain Medicine, Stanford University, Stanford, CA, USA.
Commun Med (Lond). 2025 Sep 1;5(1):380. doi: 10.1038/s43856-025-01071-7.
Male infertility (MI) is the sole cause of 20-30% of infertility cases, and it is a contributing factor for an additional 15-20% of cases. However, the full breadth of potential MI risk factors and adverse health outcomes has not been explored.
We used electronic health records (EHRs) from the University of California (UC) and Stanford to identify MI-associated comorbidities. We identified 6531 and 5551 MI patients at UC and Stanford, respectively, and 8353 and 2464 vasectomy control patients at UC and Stanford, respectively. Low-dimensional embeddings of patients' diagnosis profiles based on MI status, demographics, or hospital utilization were compared using either Kruskal-Wallis tests followed by post hoc Dunn's tests or Mann-Whitney U tests. We used logistic regression to identify MI-associated comorbidities prior to or after 6 months of a patient's first MI or vasectomy-related record. Pearson correlation coefficients were used to compare primary versus sensitivity logistic regression analyses as well as UC versus Stanford logistic regression analyses. Cox regression was used to assess whether patients had a higher risk of receiving diagnoses significantly associated with MI after the 6-month cutoff at UC.
Here, we identify 15 diagnoses that are positively associated with MI before the 6-month cutoff across both hospital systems and all analyses, including less expected comorbidities such as hypothyroidism and other anemias. Using Cox regression, we find that patients have a higher risk of receiving 11 out of 13 diagnoses positively associated with MI after the 6-month cutoff at UC.
Our findings can set the groundwork for future studies to clarify the relationship between less expected comorbidities and MI.
男性不育（MI）是20%-30%不育病例的唯一病因，在另外15%-20%的病例中是一个促成因素。然而，尚未对MI潜在风险因素和不良健康结局的全部范围进行探索。
我们使用了来自加利福尼亚大学（UC）和斯坦福大学的电子健康记录（EHR）来识别与MI相关的合并症。我们分别在UC和斯坦福大学识别出6531例和5551例MI患者，以及分别在UC和斯坦福大学的8353例和2464例输精管切除对照患者。基于MI状态、人口统计学或医院利用情况的患者诊断概况的低维嵌入，使用Kruskal-Wallis检验，随后进行事后Dunn检验或Mann-Whitney U检验进行比较。我们使用逻辑回归来识别患者首次MI或输精管切除相关记录6个月之前或之后与MI相关的合并症。Pearson相关系数用于比较主逻辑回归分析与敏感性逻辑回归分析以及UC与斯坦福逻辑回归分析。Cox回归用于评估在UC的6个月截止期后，患者接受与MI显著相关诊断的风险是否更高。
在此，我们识别出在两个医院系统的所有分析中，在6个月截止期之前与MI呈正相关的15种诊断，包括如甲状腺功能减退和其他贫血等较意外的合并症。使用Cox回归，我们发现患者在UC的6个月截止期后，接受与MI呈正相关的13种诊断中的11种的风险更高。
我们的发现可为未来研究奠定基础，以阐明较意外的合并症与MI之间的关系。
