We thank Drs White and Regan for their comments [1] regarding our article [2]. Unfortunately, most of the markers of insulin resistance mentioned in their correspondence, namely HOMA-IR, HOMA2 and QUICKI, are not available in the UK Biobank. Therefore, this exposure, even if we had wanted to include it, would require another dataset to address this research question. In terms of addressing upstream causes of elevated HbA1c, we did adjust for BMI and comorbidity in our model to isolate the causal effect of elevated HbA1c. Concerning their point about our agreement to ‘pre-operative testing of surgical patients for hyperinsulinaemia as well as hyperglycaemia and HbA1c glycation’, we have two main concerns. First, we wonder whether using surrogate measures of insulin resistance in everyday practice is feasible. We must consider the added cost to the NHS, the availability of the tests in hospital laboratories, the turnaround times of these investigations in the short pre-operative assessment window and, of course, the ability of all clinicians to interpret the results. Indeed, interpretation of results relies on clinically relevant thresholds, which brings us to our second concern. While there is a large amount of observational data on glucose and HbA1c and postoperative outcomes, there is a relative paucity of evidence for markers of hyperinsulinaemia and postoperative complications. This is an area that is no doubt important, but requires further assessment to elucidate its clinical relevance.

我们感谢 White 博士和 Regan 博士对我们的文章 [ 2 ] 发表的评论 [ 1 ]。不幸的是，在他们的通信中提到的大多数胰岛素抵抗标志物，即 HOMA-IR、HOMA2 和 QUICKI，在英国生物银行中都没有。因此，即使我们想包含这种曝光，也需要另一个数据集来解决这个研究问题。在解决 HbA1c 升高的上游原因方面，我们确实在我们的模型中调整了 BMI 和合并症，以隔离 HbA1c 升高的因果效应。关于他们关于我们同意对手术患者进行高胰岛素血症以及高血糖和 HbA1c 糖化的术前检测的观点'，我们有两个主要担忧。首先，我们想知道在日常实践中使用胰岛素抵抗的替代措施是否可行。我们必须考虑到 NHS 的额外成本、医院实验室检测的可用性、这些调查在较短的术前评估窗口中的周转时间，当然还有所有临床医生解释结果的能力。事实上，结果的解释依赖于临床相关的阈值，这让我们想到了第二个问题。虽然有大量关于血糖和 HbA1c 以及术后结果的观察数据，但关于高胰岛素血症和术后并发症标志物的证据相对缺乏。这是一个毫无疑问很重要的领域，但需要进一步评估以阐明其临床相关性。