Authors' Reply: Reconsidering Biomarker Interpretation and Clinical Applicability in High-Intensity Interval Training and Kidney Function Trials

We thank Dr. Sharma1 for the interesting commentary on our article “Long-Term Physical Exercise for Preventing CKD in Older Adults: A Randomized Clinical Trial,” published in JASN.2 The letter suggests that the effect of exercise on cystatin C could be biased by the off-target effects of exercise on inflammation and thyroid function. Although exercise might have a modest effect on these parameters, their contribution to cystatin C levels is very small so the potential influence on our trial results seems only theoretical. By contrast, the effect of exercise on muscle turnover is very direct and leads to immediate changes in creatinine production. Dr. Sharma's comments regarding the generalizability and population impact of our clinical trial are all important considerations. We chose to evaluate the effects of physical exercise in a low-risk population, so of course we used eGFR-based outcomes as has been endorsed by both Kidney Disease Improving Global Outcomes and the US Food and Drug Administration. Intervention trials in nephrology have previously been very difficult to undertake because of the very long timeline in patients with CKD. We believe strongly that nephrology interventions should focus on the early stages of CKD, before irreversible structural damage with fibrosis replacement occurs. Caution should indeed be taken when generalizing and implementing such data into clinical practice, and individualizing to the specific patient will be important. Adherence to assigned interventions is always an issue in clinical research and practice. In a randomized trial like the Generation 100, this will attenuate the effect of the treatment with a bias toward the null result. Therefore, our results are extraordinary in demonstrating the benefits of randomization to an exercise intervention compared with an active control arm. We do agree that our observational analyses of the achieved activity could be biased by unmeasured confounding. Interpretation of the true intervention effect is therefore complicated, as pointed out by Dr. Sharma, but all analyses point in the same direction: Physical exercise preserves kidney function, and higher levels of activity are more effective. The optimal implementation of this new knowledge into real-world practice is not yet known, but we were at least able to include 20% of the general population in this study. Furthermore, high-intensity training is always on a relative scale, which enables participation even for people at older ages and with some degree of frailty and comorbidity.