Meta-Epidemiology of Testosterone’s Risks and Benefits—Will We Ever Know the Answer? To the Editor Five-years ago, we published the first Mendelian randomization study1 of the cardiometabolic and mortality risk related to testosterone and concluded that reported observational associations might largely result from residual confounding or reverse causation. To overcome these inherent limitations of observational research, the recently published series of results from the Testosterone Trials (T-Trials)2-4 promised to deliver much-awaited evidence about the risks and benefits of testosterone treatment. But despite being the largest randomized clinical testosterone trial series to date, the T-Trials2-4 results—the low generalizability and the mixed nature of the findings, showing some benefits in one area and some risks in others—are insufficient to support any clinical treatment decisions in symptomatic older men. To address the unanswered question about the health effects of testosterone treatment in aging men, future large long-term trials are still needed. But major trial registries (ie, clinicaltrials.gov, clinicaltrialsregister.eu) do not announce any upcoming trials of this nature, and even if trials of this nature were planned, meta-epidemiological insights tell us that 17 664 men will be needed in each trial group to detect a true difference in cardiovascular risk between treatment groups.5 Thus, we may never know the answer. In the meantime, a quasinatural experiment continues, where people we do not know, use testosterone for reasons we do not know, with long-term effects we do not know.

中文翻译：

---

睾酮风险和益处的元流行病学——我们会知道答案吗？——回复

睾酮风险和益处的元流行病学——我们会知道答案吗？致编辑 五年前，我们发表了第一个关于与睾酮相关的心脏代谢和死亡风险的孟德尔随机研究 1，并得出结论，报告的观察性关联可能主要是由残余混杂或反向因果关系造成的。为了克服观察性研究的这些固有局限性，最近发表的睾酮试验 (T-Trials)2-4 系列结果承诺提供期待已久的关于睾酮治疗风险和益处的证据。但尽管是迄今为止最大的随机临床睾酮试验系列，T-Trials2-4 结果——低普遍性和研究结果的混合性质，在一个领域显示出一些好处而在其他领域显示出一些风险——不足以支持有症状的老年男性的任何临床治疗决定。为了解决睾酮治疗对老年男性健康影响的悬而未决的问题，未来仍需要进行大规模的长期试验。但是主要的试验注册机构（即，clinicaltrials.gov、clinicaltrialsregister.eu）没有宣布任何即将进行的此类试验，即使计划进行此类试验，元流行病学见解告诉我们，每个试验都需要 17664 名男性试验组检测治疗组之间心血管风险的真正差异。5 因此，我们可能永远不会知道答案。与此同时，一项准自然实验仍在继续，我们不认识的人出于我们不知道的原因使用睾酮，其长期影响我们不知道。