Maximal Stress Ergometry Performance is Associated with Troponin T and Abdominal Aortic Calcication Score in Severe Chronic Kidney Disease

Background Cardiac biomarkers Troponin T (TnT) and N-terminal pro-B-type natriuretic peptide (proBNP) and abdominal aortic calcication score (AAC) are associated with cardiovascular events and mortality in patients with chronic kidney disease (CKD). The effects of cardiac biomarkers and AAC on maximal exercise capacity in CKD are unknown and were studied. Methods 174 CKD 4-5 patients not on maintenance dialysis underwent maximal bicycle ergometry stress testing, lateral lumbar radiograph to study AAC, echocardiography and biochemical assessments. Results The subjects with proportional maximal ergometry workload (WMAX%) less than 50% of the expected values had higher TnT, proBNP, AAC, left ventricular end-diastolic diameter, left ventricular mass index, E/e’ and pulse pressure, and lower global longitudinal strain compared to the better performing patients. TnT (β= -0.09, p=0.02), AAC (β= -1.67, p<0.0001) and diabetes (β= -11.07, p<0.0001) remained signicantly associated with WMAX% in the multivariable model. Maximal ergometry workload (in Watts) was similarly associated with TnT and AAC in addition to age, male gender, hemoglobin and diastolic blood pressure in a respective multivariate model. AAC and TnT showed fair predictive power for WMAX% less than 50% of the expected value with AUCs of 0.70 and 0.75, respectively Conclusions TnT and AAC are independent determinants of maximal ergometry stress test workload in patients with advanced CKD.


Introduction
Cardiovascular disease is the prevailing cause of morbidity and mortality in patients with chronic kidney disease (CKD). Alterations in cardiac and vascular structure and function may lead to exercise intolerance. Impairment of maximal exercise performance is well-established in CKD [1][2][3], but less is known about its determinants. Elevated troponin T has been shown to be associated with increased cardiovascular mortality in end stage kidney disease (ESKD) [4][5][6]. The Food and Drug Administration has approved the use of troponin T (TnT) for identifying high mortality risk in CKD patients [5]. N-terminal pro-B-type natriuretic peptide (proBNP) elevation may occur in CKD independent of heart failure [4], but usually re ects underlying heart disease such as left ventricular hypertrophy or coronary atherosclerosis [7,8]. proBNP has been associated with cardiovascular events and cardiovascular mortality in CKD [9]. However, there are no previous data on the association between cardiac biomarkers and maximal physical performance in advanced CKD. Current guidelines recommend the evaluation of abdominal aortic calci cation (AAC) score in CKD [10].
AAC is independently associated with cardiovascular events in the general population and in dialysis patients [11,12]. Previous studies have shown that AAC is independently associated with coronary artery calci cation [13], left ventricular diastolic dysfunction [14], and increased risk for incident claudication [15], all of which may lead to diminished physical performance in CKD.
Deteriorating exercise capacity has great importance on the quality of life, morbidity and mortality.
Evaluating physical performance in CKD patients not yet on dialysis might provide valuable data on health-related risks, and result in improved care, if the causes of declined exercise capacity are better comprehended.
We aimed to study maximal physical performance and its determinants including cardiac biomarkers, AAC and echocardiography in patients with stage 4-5 CKD not on maintenance dialysis.

Study Protocol
The Chronic Arterial Disease, quality of life and mortality in chronic KIDney injury (CADKID) -study (http://www.ClinicalTrials.gov NCT04223726) is an ongoing prospective follow-up study protocol assessing arterial disease, quality of life and mortality in patients with CKD stage 4-5. 210 consecutive patients referred to the predialysis outpatient clinic of Kidney Center, Turku University Hospital between 2013 and 2017 were recruited to the study. Enrolled patients had CKD stage 4-5 de ned as glomerular ltration rate (eGFR) < 30 ml/min/1.73 m 2 and estimated using the Chronic Kidney Disease Epidemiolgy Collaboration (CKD-EPI) formula. [16] The study design was approved by Medical Ethics Committee of the Hospital District of Southwest Finland. All procedures were in accordance with the Helsinki Declaration. All patients gave written informed consent before entering the study.
All 210 patients were invited to attend stress ergometry. 176 patients agreed to attend, and 174 patients underwent a standard maximal bicycle stress test. For two patients stress ergometry was not possible due to patient related issues. Abdominal aortic calci cation was assessed from plain lateral lumbar radiograph, and echocardiography in addition to biochemical studies were examined at baseline.

Maximal Stress Ergometry
Maximal Stress Ergometry was performed as an incremental, symptom-limited cycling exercise test in accordance with clinical standards. Each patient started with a 30 s warm-up phase during which the target speed of 60 rpm was reached. Primary workload was determined according to an estimated maximum workload and a targeted test duration of 6 to 10 minutes. An increase in workload per minute (10, 15 or 20 W) was accomplished automatically by the ergometer software until symptom limitation within 6 to 10 minutes. Participants were informed to cycle at a speed of 60 rpm, and were encouraged to continue cycling until exhaustion. Perceived strain was reported as the highest rating on the Borg Scale from 1 to 20. The mean proportional workload of the last 4 minutes of the age, sex and body size predicted value (WMAX%), was used in the analyses in addition to the corresponding workload in watts (WMAX). The values considered normal for expected maximal exercise performance measured as watts are derived and extrapolated from large data sets of the Mini Suomi -study. [17] The algorithm for expected exercise performance is incorporated in cycle ergometer software and used in day-to-day clinical work. The study population was divided into two groups according to WMAX% <50% versus ≥ 50% of the expected normal value. . For E/e' the transmitral early diastolic in ow velocity (E wave) was measured using pulse-wave Doppler in the apical fourchamber view, and the peak (e') diastolic mitral annular velocity was measured using tissue Doppler imaging at the septal mitral annulus. Ultrasound examinations were performed using a commercially available ultrasound system (Vivid E9; GE Vingmed Ultrasound, Horten, Norway) with a 3.5-MHz phasedarray transducer (M5S).

Assessment of AAC
Abdominal aortic calci cation (AAC) score was calculated for each subject. Lateral lumbar radiography with standard equipment was performed in a standing position. A validated 24-point scale, as described previously, was used. [18] Calci c deposits of the anterior and posterior wall of the aorta, adjacent to rst through fourth lumbar vertebrae, were assessed at each vertebral segment, and were graded on a scale 0-3 as follows: 0 = no calci c deposits, 1 = small scattered calci c deposits lling less than one-third of the longitudinal aortic wall, 2 = one-thirds to two-thirds of the longitudinal aortic wall calci ed, 3 = at least two-thirds of the longitudinal aortic wall calci ed. The grades of eight segments were summed ranging from 0 to 24 points. Two independent observers recorded AAC scores of all lateral lumbar X-rays, and mean was used for analysis.

Statistical Analysis
Data are presented only for the 174 patients that underwent stress ergometry. We compared the patients not attending stress ergometry (n = 36) to those who did (n = 174). Those not attending were more often women, had a higher proportion of coronary artery disease (CAD), were older and had higher TnT but no differences were observed in AAC or proportion of diabetics (data not shown).
Results are presented as mean ± standard deviation (SD) for the normally distributed variables and as median [inter-quartile range (IQR)] for skewed variables. Skewed variables were log e -transformed to normalize distributions. Normality in continuous covariates was tested with Kolmogorov-Smirnov and Shapiro-Wilk tests. Student's t-test was used to compare continuous normally distributed covariates and Chi-square test for categorical covariates in the study subgroups. For some skewed variables a suitable transformation was not found and thus the comparisons between groups were done using a nonparametric Kruskal-Wallis test.
Univariable associations between the study variables were analyzed by calculating Spearman's correlation coe cients. Multivariable analysis was done using linear regression technique. Variables with signi cant univariable correlations with WMAX%, as well as diabetes and previous coronary artery disease were included as covariates in a stepwise multivariable linear regression model. The multivariable associations between exposure variables and WMAX were also studied using the same analyses. However, for WMAX (expressed in Watts), age, height and gender were also included as covariates in the multivariable model. To make the multivariable results easily comparable, variables were included in the models without transformation.
Receiver operating characteristics (ROC) curve analyses were conducted to estimate the area under the curve (AUC) as a measure of discriminative capacity of TnT and AAC for WMAX%<50%.
All statistical analyses were performed using statistical analysis system, SAS version 9.3 (SAS Institute Inc., Cary NC). P < 0.05 was considered statistically signi cant.

Patient Characteristics
A total of 174 patients with a mean age of 60.9±13.7 years and median eGFR 12.9 ± 3.4 ml/min/1.73 m 2 underwent a maximal bicycle stress test, echocardiography and lateral lumbar radiography for AAC assessment. Nearly half (43%) had diabetes and 21 (12%) had coronary artery disease. All, but one were on antihypertensive medication. The mean workload of the last four minutes of maximal stress (WMAX) was 83.7 ± 36.5 W and the proportional maximal workload (WMAX%) 55.7 ± 21.5% of the age predicted normal value. The baseline clinical and laboratory characteristics of the study groups (WMAX% <50% vs. ≥50%) are shown in Table 1.

Determinants of Exercise Performance
The subjects with WMAX% of less than 50% of the expected values were older, more likely men, and had a higher prevalence of diabetes and coronary artery disease. TnT, proBNP, AAC, left ventricular end-diastolic diameter, LVMI, E/e', and pulse pressure were higher compared to the better performing group and GLS was lower. There were no signi cant differences between the groups in terms of body mass index, eGFR, hemoglobin, albumin, total cholesterol, LVEF or the use of beta blockers or calcium channel blockers. Only 5 patients in the whole cohort had below normal (< 50%) LVEF.

Discussion
The present study shows for the rst time that maximal stress ergometry performance is associated with TnT and AAC but not with proBNP in CKD stage 4-5 patients without maintenance dialysis.
Previous studies have demonstrated that elevated TnT is associated with an increased risk for incident cardiovascular events and mortality in CKD [4][5][6]. The Food and Drug Administration and the National Kidney Foundation Disease Outcomes Quality Initiative Work Group (KDOQI) have recommended that TnT levels can be considered for risk strati cation of CKD patients to identify those at high mortality risk [5,19]. Our current results suggest that increased TnT in CKD is a determinant of cardiovascular burden that affects the day to day life of patients with advanced CKD by limiting their maximal physical performance in addition to increasing their risk for incident cardiovascular death. In accordance with this nding we have previously shown that the physical composite score of the Short Form 36 Items Health Survey (SF-36 QOL) is inversely associated with TnT in the CADKID study population [20].
A recent large study showed that increased TnT is associated with left ventricular hypertrophy (LVH) and diastolic dysfunction in CKD but not with systolic dysfunction [21]. In line with these results TnT was associated with left ventricular mass index, E/e' and GLS but not with LVEF in present study. Only 5 patients in our study cohort had below normal LVEF (< 50%) which may explain why proBNP was not independently associated with WMAX% in the multivariate model. Increased E/e'-ratio, a marker of cardiac diastolic dysfunction, caused mainly by uid and sodium retention and ventricular stiffness in CKD, showed signi cant univariable correlations with both TnT and WMAX% as did GLS, a more subtle indicator of left ventricular function compared to LVEF. These alterations in cardiac function probably have a signi cant role in the attenuated physical performance in CKD. However, none of the associations between WMAX% and echocardiographic indices remained signi cant in the multivariate model in comparison to TnT which remained highly signi cant. Therefore, increased TnT may at least to some extent be a marker of silent myocardial ischemia in our patients with advanced CKD which could explain the reduced ergometry performance in affected patients. In line with this assumption a previous single photon emission tomography study with mostly maintenance dialysis dependent ESKD patients showed an association between TnT and perfusion defects indicative of myocardial ischemia after pharmacologic and/or exercise stress [22]. The association between TnT and WMAX% also remained signi cant after controlling for previously diagnosed CAD in the multivariable model.
AAC is independently associated with cardiovascular events in the general population and in dialysis patients and the KDOQI Work Group recommends the assessment of AAC for risk strati cation in CKD [10][11][12]. There are no previous data available on the association between AAC and maximal ergometry performance or other physical stress tests in patients with CKD. AAC has previously been shown to be associated with left ventricular mass, left atrial volume and left ventricular diastolic dysfunction including decreased E/e' in CKD [14,23]. Furthermore, former studies have shown that AAC is independently associated with coronary artery calci cation [13], and increased risk for incident claudication [15], both of which may lead to diminished physical performance in CKD. In line with previous ndings AAC was associated with E/e' in addition to attenuated exercise capacity in the present study. As AAC is closely associated with coronary artery calci cation the association with poor maximal exercise tolerance may be at least partly attributed to silent myocardial ischaemia during the maximal ergometry test. Other potential mechanisms responsible for the observed attenuation of maximal stress test workload with increasing AAC in our study include universal atherosclerosis and peripheral arterial disease and increased ventricular stiffness and diastolic dysfunction that are known to be associated with aortic calci cation. The nding that AAC has an impact not only for risk strati cation of patients with advanced CKD, but also on their ability to endure exercise, may be of value in clinical work to target treatments and further diagnostics accordingly.
The present study has limitations. The data included in this current report are cross-sectional. The study sample was somewhat limited. There was a degree of expected selection bias in attending stress ergometry as the patients not attending were more often women and had CAD, were older and had higher TnT but no differences were observed in AAC or proportion of diabetics. Nevertheless, in our opinion the studied cohort with a high degree of comorbidities represents the overall CKD stage 4-5 population at our center well. It is not likely that our current results on the association between TnT, AAC and WMAX were signi cantly affected due to this selection bias. As we did not perform spiroergometry to de ne the peak oxygen uptake in this study, the maximal ergometry performance was therefore limited by subjective exhaustion. However, we believe the results give a reliable estimate of maximal aerobic performance as the observed relative value of 55.7 ± 21.5% of the age predicted performance was similar to the reduction in maximal oxygen uptake reported in patients on maintenance dialysis [24]. A high proportion of patients were on beta blockers and the ergometry was performed without medication pauses, which has probably in uenced the workloads achieved in affected patients. However, beta blocker use was similar between patients with WMAX% <50% compared to others. CKD is associated with inferior exercise capacity and the decrease in physical function appears to emerge in the predialysis period [25]. Maintaining functional independence is of most importance in patient-reported questionnaires lled by CKD patients [26]. The recent prospective randomized multicenter EXCITE trial showed that a personalized walking exercise program at home improves the physical performance of dialysis patients and reduces their rate of hospitalization [27]. The nding that TnT and AAC are independently associated with maximal physical performance in advanced CKD may have clinical implications in recognizing patients at risk and targeting treatment not only for decreasing mortality risk but also to increase the functionality and quality of life of affected patients.
Declarations FUNDING This work was supported by the Finska Läkaresällskapet, the Perklén Foundation, Helsinki, Finland and the Aarne Koskelo Foundation

CONFLICTS OF INTEREST
The authors declare that they have no con ict of interest.

ETHICS APPROVAL
The study design was approved by Medical Ethics Committee of the Hospital District of Southwest Finland. All procedures were in accordance with the Helsinki Declaration.

CONSENT TO PARTICIPATE
All patients gave written informed consent before entering the study.

CONSENT FOR PUBLICATION
All patients gave written informed consent before entering the study. All authors have approved submission of the manuscript, and the manuscript has not been published and is not being considered for publication elsewhere.

AVAILABILITY OF DATA AND MATERIAL
Data that support the ndings of this study are available from the datasets of the Kidney Center of Turku University Hospital upon reasonable request and after permission of the Ethics Committee from Hospital District of Southwest Finland.

CODE AVAILABILITY
All statistical analyses were performed using statistical analysis system, SAS version 9.3 (SAS Institute Inc., Cary NC).