Supplementary Material for: The impact of age and body composition on the agreement between estimated and measured GFR in heart transplant recipients

Background
Estimated GFR (eGFR) has shown poor agreement with measured GFR (mGFR) in several populations. We investigated the impact of age and body composition on the accuracy and precision of eGFR in heart transplant (HTx) recipients.

Methods
In a longitudinal, observational, retrospective study-design, patients receiving first-time HTx with at least one registered mGFR value within 15 months after HTx and a corresponding plasma creatinine were included. GFR was measured by 51Cr-EDTA and eGFR calculated by creatinine-based CKD-EPI formula.

Results
150 patients with a total of 723 mGFR measurements were included. During the first year after HTx, mean weight increased by 4.2 kg (CI:3.2–5.1) followed by an annual decrease of 0.35 kg/year (Cl:-0.05–0.74). mGFR increased by 7.5 ml/min (Cl:3.2–11.8) the first year but was stable hereafter (0.0 ml/min/year CI:-1.0–1.0). The initial weigh gain and increase in mGFR were most pronounced in patients <45 years. Neither eGFR adjusted nor unadjusted for BSA detected the initial increase in mGFR.
At one year after HTx, limits of agreement on the Bland-Altman plot were -37.2–33.1 ml/min with a bias of -2.1 ml/min (Cl:-5.0–0.9). In patients <45 years, eGFR significantly overestimated mGFR by 7.1 ml/min (Cl:1.0–13.2) and showed a significant lower precision than patients >45 years. There was no effect of BMI class, weight, BSA or change in BMI class on the difference between eGFR and mGFR.

Conclusion
eGFR is, on average, accurate but imprecise in HTx patients. The agreement is affected by age, but not body composition.

### 背景
多项研究显示，估算肾小球滤过率（eGFR）与实测肾小球滤过率（mGFR）在多个人群中一致性欠佳。本研究旨在探讨年龄与身体成分对心脏移植（HTx）受者eGFR准确性与精密度的影响。

### 方法
本研究采用纵向观察性回顾性研究设计，纳入首次接受心脏移植、且在术后15个月内至少有1次记录的实测肾小球滤过率值及对应血浆肌酐检测结果的患者。采用51铬-乙二胺四乙酸（51Cr-EDTA）测定肾小球滤过率，并通过基于肌酐的慢性肾脏病流行病学合作研究（CKD-EPI）公式计算eGFR。

### 结果
本研究共纳入150例患者，累计获取723次mGFR检测数据。心脏移植术后1年内，患者平均体重增加4.2kg（置信区间CI：3.2~5.1），此后每年平均下降0.35kg/年（置信区间CI：-0.05~0.74）；术后1年内mGFR平均升高7.5ml/min（置信区间CI：3.2~11.8），此后维持稳定（0.0ml/min/年，置信区间CI：-1.0~1.0）。上述初始体重增加与mGFR升高在年龄小于45岁的患者中最为显著。无论是否以体表面积（BSA）校正eGFR，均未检测到mGFR的初始升高。
心脏移植术后1年时，布兰德-奥特曼图（Bland-Altman plot）的一致性界限为-37.2~33.1ml/min，偏倚为-2.1ml/min（置信区间CI：-5.0~0.9）。在年龄小于45岁的患者中，eGFR较mGFR平均高估7.1ml/min（置信区间CI：1.0~13.2），且精密度显著低于年龄大于45岁的患者。身体质量指数（BMI）分级、体重、体表面积或BMI分级变化对eGFR与mGFR的差值无显著影响。

### 结论
心脏移植受者的eGFR整体平均水平准确，但精密度欠佳；其与mGFR的一致性受年龄影响，但与身体成分无关。