We read with interest the paper entitled ‘Neuromuscular impairment at different stages of human sarcopenia’ by Sarto et al. [1], which has used a variety of assessment approaches to innovatively investigate neuromuscular impairment in older human subjects. This study undoubtedly represents an important contribution to our understanding of the impact of age and sarcopenia on the human neuromuscular system, for which we wish to congratulate and thank the authors.

Although the authors have presented a lot of important new data, they state in their conclusions that ‘an altered innervation profile and NMJ instability are prominent features of the ageing of the neuromuscular system’. This is based on the claim that an increase in neural cell adhesion molecule-positive fibres is evidence of ‘increased muscle denervation’. We would like to raise caution against such a conclusion when based on the type of indirect data presented by Sarto et al. [1].

The term ‘muscle denervation’ refers to a specific process whereby the nerve supply (innervation) of a muscle fibre, derived from a lower motor neuron input, is removed or lost. For this term to be accurate, therefore, it requires evidence showing that lower motor neuron inputs at the NMJ are removed or lost. We would politely suggest that an increase in neural cell adhesion molecule-positive fibres is not such evidence. Rather, anatomical and morphological studies of the NMJ are required to be able to draw such conclusions.

Sarto et al. [1] point out in their paper that morphological evidence for NMJ disruption, and hence muscle denervation, is present in rodent models of ageing [2] and use this to support their claims in humans. However, whilst technically challenging, comparable morphological studies of NMJs in young and old humans have been performed, showing that, in stark contrast to rodent models, NMJs remain structurally intact over the normal human lifespan [3]. Such species-specific differences are perhaps not surprising given the inherent differences that exist in NMJ structure and stability between rodents and humans [4, 5]. As such, the strongest direct evidence available to date suggests that motor neuron inputs at the NMJ are not removed or lost with increasing age in humans. This in no way precludes the possibility that age- and/or sarcopenia-related changes, such as those reported by Sarto et al., [1] are occurring in muscle that mimic denervation-induced changes. Rather, it questions the premise that such changes are occurring due to a loss of innervation resulting from structural breakdown of the NMJ.

The use of indirect measurements to draw conclusions about NMJ status and muscle denervation is not an issue solely restricted to Sarto et al.'s paper [1], and we by no means wish to single out this important study in this regard (indeed, we strongly welcome the newly presented evidence of age-related functional decline in the human neuromuscular system). Understandably, many studies are not in the position to obtain fresh muscle biopsies containing NMJs suitable for morphological assessment, for both technical and/or logistical reasons. In our own experience, obtaining such samples is a frustrating, time-consuming and complex task. However, we would contend that experiments facilitating direct morphological assessment of the NMJ are required for studies to be able to draw robust conclusions concerning muscle denervation and NMJ status.

We hope that these comments will be useful for colleagues considering the design and reporting of future studies in this field.

我们饶有兴趣地阅读了 Sarto 等人 [1] 题为“人类肌肉减少症不同阶段的神经肌肉损伤”的论文，该论文使用了多种评估方法，创新性地研究了老年人类受试者的神经肌肉损伤。这项研究无疑代表了对我们理解年龄和肌肉减少症对人类神经肌肉系统影响的重要贡献，我们对此表示祝贺和感谢。

尽管作者提供了许多重要的新数据，但他们在结论中指出，“神经支配谱改变和 NMJ 不稳定是神经肌肉系统衰老的突出特征”。这是基于神经细胞粘附分子阳性纤维的增加是“肌肉去神经支配增加”的证据的说法。当基于 Sarto 等人 [1] 提供的间接数据类型时，我们想对这样的结论持谨慎态度。

术语“肌肉去神经支配”是指一个特定的过程，其中来自下运动神经元输入的肌肉纤维的神经供应（神经支配）被移除或丢失。因此，要使该术语准确，需要有证据表明 NMJ 处的较低运动神经元输入被移除或丢失。我们会礼貌地建议，神经细胞粘附分子阳性纤维的增加并不是这样的证据。相反，需要对 NMJ 进行解剖学和形态学研究才能得出这样的结论。

Sarto 等人 [1] 在他们的论文中指出，NMJ 破坏以及肌肉去神经支配的形态学证据存在于啮齿动物衰老模型中 [2]，并以此来支持他们在人类中的说法。然而，尽管在技术上具有挑战性，但已经对年轻人和老年人的 NMJ 进行了类似的形态学研究，结果表明，与啮齿动物模型形成鲜明对比的是，NMJ 在正常人类寿命中保持结构完整 [3]。考虑到啮齿动物和人类之间 NMJ 结构和稳定性的固有差异，这种物种特异性差异可能并不奇怪 [4， 5]。因此，迄今为止最有力的直接证据表明，人类 NMJ 的运动神经元输入不会随着年龄的增长而消失或丢失。这绝不排除与年龄和/或肌肉减少症相关的变化的可能性，例如 Sarto 等人报道的那些 [1] 发生在模拟去神经支配诱导的变化的肌肉中。相反，它质疑了这种变化是由于 NMJ 结构分解导致神经支配丧失而发生的前提。

使用间接测量得出关于 NMJ 状态和肌肉去神经支配的结论不仅仅是 Sarto 等人的论文 [1] 的问题，我们绝不希望在这方面单独列出这项重要的研究（事实上，我们强烈欢迎新提出的证据，即人类神经肌肉系统中与年龄相关的功能下降）。可以理解的是，由于技术和/或后勤原因，许多研究无法获得含有适合形态学评估的 NMJ 的新鲜肌肉活检。根据我们自己的经验，获取此类样本是一项令人沮丧、耗时且复杂的任务。然而，我们认为，需要促进 NMJ 直接形态学评估的实验，以便研究能够得出有关肌肉去神经支配和 NMJ 状态的可靠结论。

我们希望这些评论对考虑设计和报告该领域未来研究的同事有用。