In their letter to the editor, Cheng et al.¹ performed additional meta-regression analyses using age stratification, based on our meta-analysis of the effects of vitamin D supplementation monotherapy on indices of sarcopenia.² We did not perform age-stratified subgroup analyses, because it is important for systematic reviews and meta-analyses to follow a priori analysis plans registered in the international prospective register of systematic reviews (PROSPERO).

The authors categorized older adults into ‘young-old’ (60–69 years old), ‘middle-old’ (70–79 years old) and ‘old-old’ (≥80 years old) age strata and showed that vitamin D supplementation decreased general physical performance in the middle-old population (standardized mean difference [SMD]: −0.15; 95% CI: −0.27 to −0.02), but not young-old or old-old populations. The classification of such age strata is not well established,³ and it should be noted that significant heterogeneity exists in the health and physical function of older adults of similar chronological ages.⁴ Regardless, the meta-regression by Cheng et al. essentially confirms our findings.

In relation to the authors' statement highlighting the need for larger scale vitamin D trials in different groups across old age, several studies,^5-10 including ours, have observed unfavourable effects on muscle strength, physical performance and risk of falls, particularly with higher doses of vitamin D. We included in our discussion that ‘… mechanistic studies may be preferable to investigate this relationship and any randomized controlled trials of high-dose vitamin D supplementation should potentially be restricted to those at low risk of falling’. Moreover, given our own and other recent meta-analyses that have demonstrated no effect of vitamin D on muscle strength/physical performance,^{2, 5} there is limited benefit to conducting further trials in the general population of older adults. Indeed, in a trial sequential analysis that determined the effects of vitamin D supplementation on falls and fractures, Bolland et al.¹¹ concluded that ‘… vitamin D supplementation does not have meaningful clinical benefits …’ and that ‘Further similar trials are unlikely to alter the conclusions …’. These findings suggest that strong rationale is required to support the need for further studies on vitamin D supplementation for improving musculoskeletal health. Hence, we reiterate our main conclusion that any future trials of vitamin D supplementation for prevention or treatment of sarcopenia should be targeted at specific populations, such as those at greatest risk for low vitamin D status and its associated complications.

在给编辑的信中，Cheng 等人。¹根据我们对维生素 D 补充剂单一疗法对肌肉减少症指标影响的荟萃分析，使用年龄分层进行了额外的荟萃回归分析。²我们没有进行按年龄分层的亚组分析，因为系统评价和荟萃分析遵循国际前瞻性系统评价登记册 (PROSPERO) 中注册的先验分析计划很重要。

作者将老年人分为“青年”（60-69 岁）、“中年”（70-79 岁）和“老年”（≥80 岁）年龄层，并表明维生素 D补充剂会降低中老年人群的一般身体机能（标准化均值差 [SMD]：-0.15；95% CI：-0.27 至 -0.02），但不会降低年轻或老年人群。此类年龄层的分类尚不明确³，应该注意的是，年龄相近的老年人的健康和身体机能存在显着异质性。⁴无论如何，Cheng 等人的元回归。基本上证实了我们的发现。

关于作者强调需要在不同年龄组进行更大规模维生素 D 试验的声明，包括我们在内的^5-10项研究观察到对肌肉力量、身体机能和跌倒风险的不利影响，尤其是较高的剂量的维生素 D。我们在讨论中提到“……机制研究可能更适合研究这种关系，任何高剂量维生素 D 补充剂的随机对照试验都可能仅限于跌倒风险较低的人群”。此外，鉴于我们自己和其他近期的荟萃分析表明维生素 D 对肌肉力量/身体表现没有影响，^{2, 5}在一般老年人群中进行进一步试验的益处有限。事实上，在确定补充维生素 D 对跌倒和骨折影响的试验序贯分析中，Bolland 等人。¹¹得出的结论是“……补充维生素 D 没有有意义的临床益处……”并且“进一步的类似试验不太可能改变结论……”。这些发现表明需要强有力的理由来支持进一步研究补充维生素 D 以改善肌肉骨骼健康的必要性。因此，我们重申我们的主要结论，即未来任何补充维生素 D 以预防或治疗肌肉减少症的试验都应针对特定人群，例如维生素 D 水平低下及其相关并发症风险最高的人群。