The current analyses are part of a larger, prospective, single-centre cohort study, PROGRESS 2002, investigating factors impacting the progress of renal insufficiency. The details of this study have been described previously [8].
Study protocol
We included 52 patients with non-dialysis CKD stage 2–3 (CKD; aged 18–65 years) and 54 healthy controls who had completed an exercise test at baseline. The cohort has been monitored over a follow-up of 5 years. GFR was measured by iohexol plasma clearance [19]. During follow-up, GFR measurement by iohexol was repeated in the CKD group. In both CKD and controls, estimated GFR (eGFR) was calculated using the Chronic Kidney Disease–Epidemiology Collaboration (CKD–EPI) equation. This equation was chosen because it has been shown to provide a the most precise estimate of filtration capacity in the range of mild CKD [20]. The methodological considerations have been described in detail elsewhere [21].
Patients were recruited consecutively from the outpatient clinic at the Department of Renal Medicine at the Karolinska University Hospital during 2002–2009 if they had renal function corresponding to CKD stage 2–3a as defined by the National Kidney Foundation [22]. The controls, matched for age and sex with the CKD 2–3 group, were randomly selected from the Swedish Total Population Register or recruited through the website of the regional University Hospital. The exclusion criteria for all participants were current malignancy, kidney transplantation or kidney donation, or blood-transmitted disease. The inclusion criteria for the controls were absence of kidney disease, cardiovascular disease, diabetes or any chronic medication. After inclusion, all participants underwent clinical investigation, anthropometric measurement, an exercise stress test, a handgrip strength test, laboratory testing, carotid ultrasound, transthoracic echocardiography and a body composition scan. The participants with CKD were closely monitored and treated aggressively for hypertension, hyperlipidaemia and proteinuria at the Department of Renal Medicine.
The study population and the excluded individuals are presented in Figure 1. Two participants in the CKD group died from cancer before the end of the follow-up period. No patient in the CKD group progressed to renal replacement therapy within the follow-up period. However, several patients were lost to follow-up. At year 5, 49 participants in the CKD group and 43 in the control group participated in the examinations. Forty-six participants in the CKD group and 40 in the control group performed an exercise test; however, one participant with CKD was excluded from the follow-up exercise test analysis because the exercise test was terminated early owing to the occurrence of atrial fibrillation with a high HR. Two other participants with CKD had atrial fibrillation before and during the exercise test; therefore, these individuals were not included in the peak HR analyses.
Fig. 1. Study population at baseline and 5-year-follow up.
The study protocol was reviewed and approved by the Local Ethics Committee and Institutional Review Board of the Karolinska Institutet at the Karolinska University Hospital. All participants gave their written informed consent.
Measurements
Aerobic exercise capacity. A symptom-limited incremental cycle ergometer test on an electronically braked cycle ergometer (RE990; Rodby Innovation AB, Uppsala, Sweden) was administered according to clinical standards. The initial workload and workload increase/minute (10, 15 or 20 W) were individualized with the goal of achieving symptom limitation within 6 to 10 minutes. Participants were instructed to cycle at a speed of 60 rpm and were encouraged to continue cycling until exhaustion. Perceived exertion was reported as the highest rating on the Borg CR10 scale for either leg fatigue, dyspnoea or general exhaustion as a limiting symptom [23]. Aerobic ExCap was defined as the peak workload in W. Continuous 12-lead electrocardiography was used to measure HR and for ST–T-segment monitoring and safety purposes. Resting HR was measured in the supine position before the exercise test. Predicted peak HR was calculated as 220 minus age. HR reserve was defined as the difference between peak and resting HR.
Muscular function. A handheld dynamometer (Grip-A; Takei Scientific Instruments Co., Ltd, Tokyo, Japan) was used to measure maximum voluntary isometric contraction as a measure of handgrip strength. The test was performed with the participant in a standing position. The measurement was repeated three times using the dominant arm, and the highest value was recorded as handgrip strength.
Cardiac and vascular function. Echocardiography was performed with the participant in the supine left decubitus position and according to current guidelines [24]. The images were acquired using an ultrasound scanner (Sequoia 512; Siemens Medical Solutions, Mountain View, CA, USA) with an appropriate transducer. Cardiac function was evaluated and expressed for systolic left-ventricular (LV) function as ejection fraction (EF), for right-ventricular systolic function as tricuspid annular plane systolic excursion (TAPSE) and for diastolic LV function as an E/é ratio (mitral flow velocity E, divided by tissue LV velocity é). Carotid artery ultrasound was performed to assess arterial stiffness. Measurements of the diameter of the right common carotid artery and calculations of the pressure strain elastic modulus (Ep) were performed according to a standardized protocol [8].
Ambulatory blood pressure monitoring was performed over 24 hours from morning to morning using Spacelab 90217 (Spacelab Healthcare, Issaquah, Washington, USA) and a cuff of appropriate size on the non-dominant arm. Blood pressure was measured three times per hour day and night.
Body composition. Body composition was measured with a whole-body dual-energy X-ray absorptiometry scan to determine LBM (QDR 4500 Discovery A, software version 12.3; Hologic, Bedford, MA, USA). Only the CKD group were examined at follow-up.
Physical activity level. Physical activity level was rated by the participants using a four-point scale modified from the Saltin–Grimby Physical Activity Level Scale as follows [25]:
Level 1 = Regular exercise: running, swimming, tennis, badminton, gymnastics or similar activity on three or more occasions per week; every session should last at least 30 min and cause sweating.
Level 2 = Moderate amount of regular exercise: running, swimming, tennis, badminton, gymnastics or similar activity on 1–2 occasions per week; every session should last at least 30 minutes and cause sweating.
Level 3 = Light exercise: walking or cycling or other physical activity during at least 2 h per week, usually without sweating; this includes walking or cycling to/from work, Sunday walks, gardening, fishing, table tennis and bowling or similar activity.
Level 4 = Sedentary: mostly reading, watching television, movies or other sedentary activities, or walking, cycling or light exercise for less than 2 h per week.
Blood samples. Haemoglobin, creatinine and high-sensitivity C-reactive protein (hs-CRP) concentration were measured in venous blood obtained using routine laboratory methods at the Karolinska University Laboratory at the inclusion visit before the exercise test.
Statistical analysis
Continuous variables are presented as mean and standard deviation, and categorical variables as number and percentage of the study sample. Variables with a skewed distribution are presented as median and interquartile range. The normality assumption was assessed graphically using histograms. Potential outliers were examined graphically by box plots and their validity were assessed. Logarithmic transformation was performed for variables with a skewed distribution when they were analysed in mixed models. Group comparisons at baseline were performed using Student’s t-test or Mann–Whitney U-test when the assumptions of normality and homogeneity were not fulfilled. For categorical variables, a chi-square test was used. Wilcoxon’s signed-rank test for related samples was used to assess the change over time for ordinal data. Statistical significance was defined as a p-value < 0.05 for a two-tailed test.
Linear mixed models were considered to be the most suitable method for the analysis of longitudinal data because this method provides effect estimates from the whole study sample provided that the values missing at follow-up are random. Changes over time were analysed in CKD and controls and the changes were compared between the groups by the interaction factor group ´ time in the mixed model. The analyses were adjusted for different background factors (depending on the chosen outcome variable), including age at baseline, sex, height and beta-blocker use.
To analyse longitudinal data where the outcome variables were ordered categoricals, generalized estimating equations (GEE) model were used. The model was set up with the same factors as in the mixed linear model mentioned above. The parameter estimates from the GEE model are presented as odds ratios (OR) and 95% confidence intervals (CI).
Pearson (r) or Spearman (rs) correlation coefficients were used to analyse the relationship between two variables.
Statistical analyses were performed using IBM SPSS Statistics (version 23.0; IBM, Armonk, NY, USA) and SAS (version 9.3; SAS Institute, Cary, NC, USA) software.