Outpatient Cardiac Rehabilitation Suppresses Deterioration of Renal Function in Patients ≥75 Years of Age With Heart Disease

Abstract

Background: This study investigated the effect of outpatient cardiac rehabilitation (OCR) and physical activity on the estimated glomerular filtration rate based on serum cystatin C (eGFRcys) in patients with heart disease (HD) aged ≥75 years.

Methods and Results: This non-randomized prospective intervention study involved 136 patients (non-OCR group, n=66; OCR group, n=70), 55 of whom were aged ≥75 years (non-OCR group, n=29; OCR group, n=26). Renal function (eGFRcys) was evaluated at discharge and 3 months thereafter. A linear mixed model (LMM) was used to assess changes in renal function over time. The hospital readmission rate within 3 months after discharge was also evaluated. LMM analysis showed that the change in eGFRcys was −2.27 and +0.48 mL/min/1.73 m² in the non-OCR and OCR groups, respectively (F=2.960, P=0.022). Further, among patients aged ≥75 years in the non-OCR and OCR groups, the change in eGFRcys was −3.83 and −1.08 mL/min/1.73 m², respectively (F=2.719, P=0.039). The proportion of patients aged ≥75 years who were rehospitalized due to exacerbation of HD was 16.9% (n=10) and 6.7% (n=2) in the non-OCR and OCR groups, respectively.

Conclusions: Among patients with HD aged ≥75 years, participation in OCR reduces the decline in renal function and hospital readmission rates.

Chronic kidney disease (CKD) is a public health epidemic that results in end-stage renal disease and increased risk of morbidity.¹ The number of heart disease patients who develop CKD is increasing.^2,3 Elderly heart disease patients have an increased risk of decreased renal function, which tends to be more severe.⁴ Impaired renal function is an independent risk factor for cardiovascular events, hospitalization, and mortality.⁵

Outpatient cardiac rehabilitation (OCR) involving exercise therapy and disease management education is important to prevent deterioration of renal function in heart disease patients.^6–10 Estimated glomerular filtration rate (eGFR) based on serum creatinine (eGFRcr) is typically used to assess renal function. However, serum creatinine is affected by sex, age, diet, and muscle mass, and is not reliable in elderly patients due to aging-related conditions.¹¹ In contrast, serum cystatin C is unaffected by these factors. Therefore, eGFR based on cystatin C (eGFRcys) is a more reliable method to estimate renal function in older patients.¹¹

The effect of OCR on renal function, determined using eGFRcys, in elderly patients with heart disease has not yet been reported. Previous studies have shown that maintaining a high level of physical activity (PA) is important for renal protection.^12,13 One meta-analysis reported an increase in PA with OCR participation compared with a control group.¹⁴ However, the effect of increased PA with OCR on renal function remains unclear. Japan is the world’s first super-aging society, with 25% of Japanese individuals expected to be >75 years of age by 2050.¹⁵ Thus, the number of elderly patients with heart disease and CKD will likely increase also. Preventive strategies for geriatric health are of utmost importance. The aim of this study was to clarify the effects of OCR and PA on renal function in all patients with heart disease and CKD, as well as in those aged ≥75 years.

Methods

Patients

This non-randomized prospective intervention study included 695 consecutive patients with heart disease hospitalized at the Cardiovascular Center, Ota Nishinouchi Hospital, between April 2016 and September 2017 who subsequently underwent inpatient cardiac rehabilitation (ICR; Phase I to early Phase II cardiac rehabilitation). Screening these 695 patients against the exclusion criteria (Figure 1) left 200 patients for evaluation at discharge and 136 for evaluation at 3 months. Thus, 136 patients were included in the final analysis.

Figure 1.

Patient flowchart. HD, heart disease; ICR, inpatient cardiac rehabilitation; OCR, outpatient cardiac rehabilitation.

Patients were divided into a non-OCR group (n=66; i.e., those who did not participate in OCR after discharge) or an OCR group (n=70; i.e., those who participated in supervised exercise sessions, home exercise sessions, and disease management for 3 months after discharge) based on their preference. Of the 55 patients aged ≥75 years, 29 were in the non-OCR group and 26 were in the OCR group. Both groups participated in ICR until discharge.

This study was approved by the appropriate ethics review boards and followed the principles of the Declaration of Helsinki. All patients provided informed consent.

ICR

ICR consisted of 5 or 6 sessions per week, each lasting 30–60 min, adjusted taking into account patients’ safety and health conditions. Sessions included correction of deconditioning and muscle strengthening exercises, including standing exercises, walking, and performing activities of daily life. The ICR was performed in a stepwise manner, with adjustments to exercise load in preparation for hospital discharge along with disease management education.

OCR

OCR consisted of supervised exercise sessions, home exercise sessions, and disease management and nutritional education.¹⁶ OCR continued for 3 months after discharge. Supervised exercise sessions occurred once or twice a week, and home exercise sessions occurred 2 or 3 times per week.¹⁷

Under supervision, exercise intensity was determined according to the anaerobic threshold (AT) level determined during a cardiopulmonary exercise test or at 40–60% of the heart rate (HR) reserve (Karvonen’s equation, k=0.4–0.6).¹⁸ In patients who underwent the cardiopulmonary exercise test, exercise intensity was recorded using an ergometer, in watts, 1 min before the AT (AT-1) for patients on an exercise bike or walking on a treadmill. To calculate the exercise intensity of other types of exercise, the load was adjusted based on HR at the AT (to a rating of perceived exertion of 11–13)¹⁹ and/or based on metabolic equivalents. In patients who did not undergo the cardiopulmonary exercise test, the exercise intensity, in watts, required to reach the HR calculated by Karvonen’s equation with a rating of perceived exertion of 11–13 was recorded.

Supervised sessions consisted of a 5-min warm-up, cycling on an exercise bike or walking on a treadmill for 30 min, resistance and balance training for 20 min, and a 5-min cool down. Home exercise consisted of a 5-min warm-up, walking for 20–30 min, resistance and balance training for 10–20 min, and a 5-min cool down. Home exercise programs were tailored for individual patients, and feedback was obtained using checklists.

Disease management education consisted of secondary prevention, nutrition, medication, and PA information, and was provided at discharge and as outpatient nutritional guidance as needed. A disease management notebook was used to record daily blood pressure, body weight, and drug administration from admission to 3 months after discharge. Patients presented the notebook during OCR sessions, medical examinations, and treatment.

Cardiorenal Function and Secondary Data

Cardiorenal function was assessed using B-type natriuretic peptide (BNP), eGFRcys, eGFRcr, and the urinary protein/creatinine ratio (PCR). Urinalysis was performed using a spot urine sample. eGFRcys (mL/min/1.73 m²) and eGFRcr (mL/min/1.73 m²) were estimated using the equations of the Japanese Society of Nephrology:^20,21

eGFRcys = 104 × CysC^−1.019 × 0.996^Age × 0.929 (if female) − 8

eGFRcr = 194 × Cre^−1.094 × Age^−0.287 × 0.739 (if female)

where CysC is the concentration of cystatin C and Cre is the concentration of serum creatinine (mg/dL).

High-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), LDL-C/HDL-C, albumin, HbA1c, hemoglobin, and blood pressure data were collected at the time of discharge and after 3 months.

Physical Function and Psychological Tests

Physical function tests, including the 6-min walk test (6MWT), handgrip strength test, and isometric knee extension strength (IKES) test, were conducted at discharge and after 3 months. The 6MWT was conducted according to recommendations of the American Thoracic Society.²² Handgrip strength was measured using a digital dynamometer (Digital Hand Dynamometer; SAEHAN, Changwon, Korea). Subjects performed the test while sitting with their arms at their sides, elbow joints at 90° flexion, and forearms and wrists in a neutral position, according to standard measurement procedures. Specifically, subjects gripped the handle of the dynamometer twice with each hand and the highest grip strength value for each hand was recorded.²³ IKES was measured using a digital dynamometer (Hydro Muscletor GT-160; OG Wellness, Okayama, Japan). The subjects sat on top of a force transducer mounted on a chair, with their hips and knees at 90°. The subjects were then asked to push with maximal effort against a cuff positioned just above the talocrural joint. Subjects could hold onto the armrest of the chair or lean backward, but were not allowed to rise from the seat. Two trials were performed for each leg, and the single best performance was used in the analysis.²⁴

Psychological evaluations were conducted using the Patient Health Questionnaire (PHQ)-9.²⁵

PA

PA was evaluated using an accelerometer with a triaxial acceleration sensor (Medi Walk MT-KT02DZ; Terumo, Tokyo, Japan). Patients were given the accelerometer sensor at discharge and were instructed to wear it near the waist, except when bathing or sleeping, for the duration of the study. The number of steps per day for 7 consecutive days was retrieved from the accelerometer sensor at 1, 2, and 3 months after discharge. Each 7-day period (21 days in total) was averaged and recorded as the absolute value for this parameter. In addition, we recorded the season when each patient participated in the study so as to remove confounding factors due to the time of year.^12,26

Patients were divided into 2 groups, low and high PA levels, depending on the number of steps taken each day. Among all patients, the median number of daily steps taken by the OCR group was 5,233. Thus, patients were allocated to the low PA group if they walked <5,233 steps/day; those who walked ≥5,233 steps/day were allocated to the high PA group. For patients aged ≥75 years, the median number of daily steps taken by the OCR group was 3,622; thus, patients aged ≥75 years were allocated to low and high PA groups if they walked <3,622 and ≥3,622 steps/day, respectively.

Definition of Participation in OCR

The effect of the number of sessions of OCR on changes in renal function changes was examined. According to previous studies,^7,17 participation in OCR is required at least once a week; therefore, 13 sessions of CR were required during the present study. Thus, the OCR group was divided into a non-active participant group (<13 sessions) and an active participant group (≥13 sessions).

Definition of CKD Severity

Changes in eGFRcys were examined in patients with CKD because it may differ depending on the severity of CKD. In the present study, CKD was defined as eGFRcys <60 mL/min/1.73 m². Mild to moderate CKD was defined as eGFRcys ≥30 and <60 mL/min/1.73 m² (Grade [G] 3 CKD).^9,13

Statistical Analysis

Analyses were performed using data for all patients, as well as for the subgroup of patients aged ≥75 years. The rate of rehospitalization caused by heart disease within 3 months of discharge was investigated. Descriptive statistics are used to present baseline characteristics and parameters related to the non-OCR and OCR groups. Baseline and clinical data, as well as physical function tests at baseline, were compared between the 2 groups using the Chi-squared test, unpaired t-test, or the Wilcoxon signed-rank test. Clinical data at baseline and after 3 months were compared within each group using a paired t-test or the Wilcoxon signed-rank test.

A linear mixed model (LMM) was used to test the group × time interaction to determine whether changes in renal function over time differed between the groups.^27,28 The model included terms for the group (non-OCR, OCR) and time (discharge [0M] and 3 months after discharge [3M]). Covariate models were created and we adjusted the degree of influence of previously reported related factors, including age,²⁹ cardiac function,⁵ diabetes,³⁰ hypertension,²⁹ dyslipidemia,^31,32 and physical function.³³ We included age, BNP, HbA1c, LDL-C/HDL-C, systolic blood pressure (SBP), and IKES in the model as covariates to assess the association between OCR and renal function. Using the same method, we examined the effects of changes in eGFRcys based on differences in PA (low activity vs. high activity groups), number of OCR sessions (non-active vs. active participants), and the CKD category of the patients with heart disease. In the CKD severity classification analysis, we focused on patients with CKD (i.e., eGFRcys <60 mL/min/1.73 m²). All analyses were performed using SPSS 21.0 (SPSS Inc., Chicago, IL, USA). Two-sided P<0.05 was considered significant.

Results

Rate of Rehospitalization

Thirty-one patients (26 [21.5%] in the non-OCR group; 5 [6.3%] in the OCR group; P=0.02) were rehospitalized within 3 months of discharge, including 18 patients aged ≥75 years (16 [27.1%] and 2 [6.7%] in the non-OCR and OCR groups, respectively; P=0.019). Sixteen patients (13 [10.7%] and 3 [6.3%] in the non-OCR and OCR groups, respectively; P=0.062) were rehospitalized due to exacerbation of heart disease, 12 of whom were age ≥75 years (10 [16.7%] and 2 [6.7%] in the non-OCR and OCR groups, respectively; P=0.155). The exacerbations of heart failure in those aged ≥75 years in the non-OCR group were caused by cold symptoms (n=2), anemia and hyperkalemia (n=4), and possible cardiac insufficiency due to insufficient disease management (n=4). The 2 exacerbations of heart failure in patients aged ≥75 years in the OCR group were caused by cold symptoms.

Patient Demographics

There were no significant differences between the non-OCR and OCR groups for patients aged ≥75 years. The OCR group attended a mean of 10 OCR sessions and achieved significantly higher PA levels than the non-OCR group among all patients and for those aged ≥75 years (Table 1).

Table 1. Baseline Characteristics and Physical Activity of the OCR and Non-OCR Groups

	All patients			Age ≥75 years
	Non-OCR (n=66)	OCR (n=70)	P value	Non-OCR (n=29)	OCR (n=26)	P value
Age (years)	71.4±9.4	67.5±11.8	0.36	79.4±4.0	78.7±5.6	0.528
Male sex	45 (68.2)	54 (77.1)	0.254	17 (58.6)	16 (61.5)	0.825
BMI (kg/m2)	22.7±4.2	23.6±3.6	0.183	21.7±3.7	21.6±2.3	0.946
Smoking status (n)
Current/former/never smoker	3/29/34	3/41/26	0.222	1/9/19	0/14/12	0.173
Drinking status (n)
Current/former/never drinker	18/22/26	13/37/20	0.162	6/10/13	2/14/10	0.234
Cardiovascular diseases
Acute myocardial infarction	9 (13.6)	22 (31.4)	0.011	3 (10.3)	5 (19.2)	0.291
Cardiomyopathy	1 (1.5)	2 (2.9)	0.522	0 (0)	1 (3.8)	0.473
Heart failure	22 (33.3)	25 (35.7)	0.456	15 (51.7)	10 (38.5)	0.238
Post cardiac surgery	34 (51.5)	21 (30)	0.005	11 (37.9)	10 (38.5)	0.593
Complications
Hypertension	58 (87.8)	61 (87.1)	0.121	27 (93.1)	23 (88.5)	0.659
Diabetes	24 (36.4)	19 (27.1)	0.273	9 (31)	9 (34.6)	0.778
Dyslipidemia	31 (47.0)	35 (50)	0.735	9 (31)	10 (38.5)	0.584
Anemia	17 (25.8)	22 (31.4)	0.57	9 (31)	13 (50)	0.178
Renal function (n)
G1/G2/G3a/G3b/G4/G5	7/30/13/13/3/0	11/28/12/12 /7/0	0.329	1/9/7/9/3/0	1/7/6/8/4/0	0.982
Medications
β-blocker	33 (50)	41 (58.6)	0.663	13 (44.8)	13 (50)	0.79
ACEI/ARB	11 (16.7)	14 (20.0)	0.663	6 (20.7)	5 (19.2)	0.893
Statin	22 (33.3)	31 (44.3)	0.22	7 (24.1)	8 (30.8)	0.763
Diuretics	12 (18.2)	16 (22.9)	0.531	8 (27.6)	8 (30.8)	0.533
Diabetes drugs	24 (36.4)	19 (27.1)	0.369	6 (20.7)	4 (15.4)	0.733
Echocardiography data
LVEF (%)	51.1±11.1	49.6±13.4	0.507	52.2±10.6	50.6±12.9	0.666
OCR data
Median [IQR] OCR sessions (n/3 months)	–	10 [5–13]		–	10 [6–13]
PA data
Mean no. steps/day	3,889±2,774	5,653±2,799	<0.001	2,269±1,276	3,725±1,846	<0.001
Median [IQR] no. steps/day	3,274 [1,919–4,999]	5,233 [3,637–7,648]		2,184 [1,391–3,109]	3,622 [2,446–4,998]
Season of monitoring
Spring (March, April, May)	11 (16.7)	10 (14.3)	0.559	5 (17.2)	4 (15.4)	0.572
Summer (June, July, August)	21 (31.8)	16 (22.9)	0.163	10 (34.5)	9 (34.6)	0.607
Fall (September, October, November)	20 (30.3)	23 (32.9)	0.446	5 (17.2)	6 (23.1)	0.419
Winter (December, January, February)	14 (21.2)	21 (30)	0.165	9 (31)	7 (26.9)	0.486
Data for 3 months after discharge
Smoking continues	1 (1.5)	0 (0)	0.485	1 (3.4)	0 (0)	0.527
Drinking alcohol continues	16 (24.2)	12 (17.1)	0.397	2 (6.9)	2 (6.9)	0.768
Medication changes	19 (28.8)	23 (32.9)	0.493	8 (27.6)	10 (38.5)	0.556

Unless indicated otherwise, data are presented as the mean±SD or n (%). ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; BMI, body mass index; IQR, interquartile range; LVEF, left ventricular ejection fraction; OCR, outpatient cardiac rehabilitation; PA, physical activity.

Patient Clinical Data

The cardiorenal function and secondary data were not significantly different between the non-OCR and OCR groups among all patients or for those aged ≥75 years. The OCR group had significantly higher 6MWT and IKES than the non-OCR group among all patients and among those aged ≥75 years (Table 2).

Table 2. Clinical Data at 0M and 3M in the OCR and Non-OCR Groups

	All patients				Age ≥75 years
	Non-OCR (n=66)		OCR (n=70)		Non-OCR (n=29)		OCR (n=26)
	0M	3M	0M	3M	0M	3M	0M	3M
Cardiorenal function
BNP (pg/mL)	165.4±150.3	103.5±151.4***	152.4±155.0	91.9±114.1**	211.9±169.6	158.8±211.4*	176.3±162.5	99.1±73.0*
eGFRcr (mL/min/1.73 m2)	66.1±21.3	61.2±21.8**	60.8±20.2	59.8±20.0	59.1±22.7	53.4±22.5*	54.6±20.5	52.5±20.5
eGFRcys (mL/min/1.73 m2)	62.8±20.6	60.9±22.4	60.4±24.8	62.6±25.8*	52.5±21.2	49.7±23.3	46.3±18.5	46.6±17.2
PCR (g/g Cr)	0.16±0.43	0.34±1.10	0.15±0.23	0.13±0.23	0.26±0.63	0.62±1.61	0.21±0.32	0.17±0.34
Secondary data
Albumin (g/dL)	3.4±0.5	4.2±0.4***	3.6±0.5	4.2±0.4***	3.4±0.5	4.1±0.4***	3.4±0.6	4.1±0.4***
Serum Cr (mg/dL)	0.91±0.37	0.98±0.42**	1.03±0.45	1.05±0.46	0.98±0.45	1.09±0.51	1.04±0.46	1.09±0.49
Serum cystatin C (mg/dL)	1.16±0.41	1.25±0.54*	1.34±0.60	1.30±0.58	1.30±0.53	1.51±0.71	1.52±0.61	1.50±0.60
HDL-C (mg/dL)	40.5±12.0	52.8±14.9***	37.8±11.1	49.1±13.8***	43.2±12.0	52.6±15.9***	41.7±10.0	54.1±13.4***
LDL-C (mg/dL)	95.2±30.0	114.8±35.5***	92.2±27.9	106.8±34.1***	93.8±32.3	115.1±34.9**	90.0±23.3	103.9±26.4*
LDL-C/HDL-C ratio	2.49±0.92	2.3±1.0	2.55±1.01	2.27±0.91**	2.28±0.88	2.36±0.98	2.18±0.84	2.03±0.69
HbA1c (%)	6.13±0.84	6.0±0.7	6.13±1.01	6.08±1.22	5.9±0.4	5.8±0.4	6.2±1.0	6.1±0.64
Hb (g/dL)	12.2±1.6	13.4±1.4***	12.3±1.8	13.5±1.7***	11.8±1.7	12.6±1.2**	11.7±1.8	12.7±1.3**
Vital signs
SBP (mmHg)	115.6±15.2	127.0±14.0***	121.2±15.5	121.3±13.6	115.6±16.7	126.0±15.4**	123.1±15.3	116.7±15.4*
DBP (mmHg)	69.5±9.6	74.2±8.8***	71.7±11.5	71.8±13.0	68.2±10.1	72.2±8.2	68.1±11.5	65.0±12.1
Physical function
6MWD (m)	324.6±98.5	347.8±108.0***	369.1±90.4	416.4±94.5***	240.0±85.9	253.0±87.1*	316.5±99.1	343.5±92.1**
IKES (% BW)	50.7±16.4	52.0±17.1*	61.1±16.8	66.3±17.7***	37.7±9.0	37.4±1.0	51.2±11.2	54.8±12.2***
Grip strength (kg)	27.3±9.1	28.1±9.6**	30.9±11.0	32.3±10.6**	21.4±6.4	21.5±6.4	22.6±7.0	24.1±7.1*
Psychological function
PHQ-9	4.5±3.8	2.7±3.3***	3.9±3.3	1.8±2.3***	6.3±4.7	4.8±4.1**	4.0±3.5	2.2±3.1**

Data are presented as mean±SD. *P<0.05, **P<0.01, ***P<0.001 compared with 0M. 0M, discharge; 3M, 3 months after discharge; 6MWD, 6-min walk distance; BNP, B-type natriuretic peptide; BW, bodyweight; Cr, creatinine; DBP, diastolic blood pressure; eGFRcr, estimated glomerular filtration rate based on serum Cr; eGFRcys, estimated glomerular filtration rate based on serum cystatin C; Hb, hemoglobin; HDL-C, high-density lipoprotein cholesterol; IKES, isometric knee extension strength; LDL-C, low-density lipoprotein cholesterol; OCR, outpatient cardiac rehabilitation; PCR, protein creatinine ratio; PHQ-9, Patient Health Questionnarie-9; SBP, systolic blood pressure.

Among all patients, plasma BNP decreased significantly in both groups from baseline to 3 months after discharge. eGFRcys increased significantly in the OCR group (P<0.05), whereas eGFRcr decreased significantly in the non-OCR group (P<0.01). The LDL-C/HDL-C ratio decreased significantly in the OCR group, and SBP increased significantly in the non-OCR group. The 6MWT, IKES, and handgrip strength increased significantly in both groups over the study period.

Among patients aged ≥75 years, plasma BNP concentrations decreased significantly in both groups from baseline to 3 months after discharge. The eGFRcys in the OCR group was similar at the 2 time points, whereas eGFRcr decreased significantly in the non-OCR group (P<0.05). There was a tendency for a decrease in the LDL-C/HDL-C ratio and SBP decreased significantly in the OCR group. The 6MWT, IKES, and handgrip strength increased significantly in the OCR group from baseline to 3 months after discharge.

LMM Comparing Changes in Renal Function

The LMM analysis revealed that changes in renal function (eGFRcys) over time differed between the 2 groups (Table 3): eGFRcys increased significantly in the OCR compared with non-OCR group (F=6.159, P=0.014). The covariate (adjusted) LMM showed that eGFRcys increased significantly in the OCR compared with non-OCR group when age, LDL-C/HDL-C ratio, HbA1c, and SBP were used as covariates (F=2.96, P=0.022). The change in eGFRcys (∆eGFRcys; calculated by subtracting eGFRcys at discharge from eGFRcys at 3 months after discharge) was −2.27 and +0.48 mL/min/1.73 m² in the non-OCR and OCR groups, respectively (Figure 2A).

Table 3. Intergroup Comparisons Between the OCR and Non-OCR Groups at 0M and 3M: Linear Mixed Model Analysis (Covariate Model)

	All patients				Age ≥75 years
	ΔeGFRcys (95% CI)		Group×time interaction		ΔeGFRcys (95% CI)		Group×time interaction
	Non-OCR (n=66)	OCR (n=70)	F	P value	Non-OCR (n=29)	OCR (n=26)	F	P value
Crude model	−1.84 (−1.87, −1.81)	+2.2 (2.17, 2.23)	6.159	0.014	−2.76 (−2.8, −2.71)	+0.08 (−0.01, 0.16)	1.354	0.251
Covariate models
Age	−1.68 (−1.85, −1.71)	+1.79 (1.76, 1.82)	15.263	<0.001	−2.8 (−2.85, −2.76)	+0.25 (0.19, 0.32)	3.064	0.026
Age, BNP	−2.01 (−2.09, −1.92)	+0.3 (−0.08, 0.69)	3.789	0.006	−2.93 (−2.97, −2.88)	−0.86 (−2.37, 0.64)	2.399	0.064
Age, LDL-C/HDL-C ratio	−2.69 (−2.78, −2.59)	+0.81 (0.51, 1.11)	4.364	0.002	−3.01 (−3.04, −2.98)	−0.65 (−0.97, −0.32)	2.786	0.035
Age, HbA1c	−2.45 (−2.6, −2.3)	+0.86 (0.78, 0.94)	3.803	0.006	−2.7 (−3.06, −2.34)	−0.17 (−0.22, −0.11)	1.096	0.369
Age, SBP	−1.09 (−1.22, −0.95)	+1.50 (1.42, 1.58)	1.656	0.16	−2.11 (−1.93, −2.30)	−0.14 (−0.43, 0.15)	0.769	0.551
Age, IKES (% BW)	−1.68 (−1.8, −1.56)	+0.58 (0.47, 0.69)	1.966	0.103	−2.68 (−2.64, −2.71)	−0.64 (−0.88, −0.40)	0.791	0.537
Age, LDL-C/HDL-C ratio, HbA1c	−3.1 (−3.3, −2.89)	+0.26 (−0.06, 0.57)	4.112	0.004	−3.83 (−3.95, −3.71)	−1.08 (−1.46, −0.69)	2.719	0.039
Age, LDL-C/HDL-C ratio, HbA1c, SBP	−2.27 (−2.33, −2.21)	+0.48 (0.05, 0.91)	2.96	0.022	−3.06 (−3.10, −3.02)	−1.52 (−2.33, −0.71)	2.285	0.072
Age, LDL-C/HDL-C ratio, HbA1c, BNP	−1.88 (−2.00, −1.77)	−0.12 (−0.59, 0.36)	3.030	0.020	−2.21 (−2.29, −2.13)	−1.28 (−2.42, −0.15)	2.021	0.107

The dependent variable was eGFRcys. The change in eGFRcys (ΔeGFRcys) was calculated by subtracting eGFRcys at 0M from eGFRcys at 3M. CI, confidence interval. Other abbreviations as in Table 2.

Figure 2.

Linear mixed model analysis comparing the changes in renal function, evaluated as changes in the estimated glomerular filtration rate based on serum cystatin C (∆eGFRcys), between the outpatient cardiac rehabilitation (OCR) and non-OCR groups for (A) all patients and (B) patients aged ≥75 years. Data are the mean±95% confidence intervals. For analyses of data for all patients (A), the covariates in the model were age, the low-density lipoprotein cholesterol (LDL-C) to high-density lipoprotein cholesterol (HDL-C) ratio, systolic blood pressure (SBP), and HbA1c. For analyses of data for patients aged ≥75 years (B), the covariates in the model were age, the LDL-C/HDL-C ratio, and HbA1c.

Among patients aged ≥75 years, the covariate (adjusted) LMM showed a decreasing trend in the negative change in eGFRcys in the OCR compared with non-OCR group when age, LDL-C/HDL-C ratio, SBP, and HbA1c were used as covariates (F=2.285, P=0.072). The decline in eGFRcys decreased in the OCR compared with non-OCR group when age, LDL-C/HDL-C ratio, HbA1c, and SBP were used as covariates (F=2.719, P=0.039). The change in eGFRcys was −3.83 and −1.08 mL/min/1.73 m² in the non-OCR and OCR groups, respectively (Figure 2B).

LMM Comparing Changes in Renal Function Based on PA

The second LMM analysis revealed that the change in renal function (eGFRcys) over time for all patients was different between the low (<5,233 steps/day) and high (≥5,233 steps/day) PA groups (Table 4). The covariate (adjusted) LMM showed that eGFRcys increased significantly in the high compared with low PA group when age, LDL-C/HDL-C ratio, HbA1c, SBP, and participation in OCR were used as covariates (F=6.169, P=0.003). The change in eGFRcys was −0.56 and +1.84 mL/min/1.73 m² in the low and high PA groups, respectively.

Table 4. Results of Linear Mixed-Model Analyses Comparing Changes in Renal Function According to Low or High Physical Activity in the OCR and Non-OCR Groups

	All patients				Age ≥75 years
	Low PA	High PA	Group×time interaction		Low PA	High PA	Group×time interaction
			F	P value			F	P value
OCR group
No. subjects	35	35			13	13
Steps/day	<5,233	≥5,233			<3,622	≥3,622
Covariate models	ΔeGFRcys (95% CI)				ΔeGFRcys (95% CI)
Age, LDL-C/HDL-C ratio, HbA1c, SBP	−0.79 (−1.25, −0.33)	+3.34 (2.76, 3.92)	2.832	0.032	−0.75 (−1.33, −0.16)	+1.33 (0.59, 2.07)	0.661	0.626
Age, LDL-C/HDL-C ratio, HbA1c, SBP, no. OCR sessions	−0.56 (−1.07, −0.04)	+1.84 (1.60, 2.09)	6.169	0.003	−1.80 (−2.78, −0.81)	−0.09 (−0.48, 0.31)	0.338	0.715
Non-OCR group
No. subjects	33	33			14	15
Steps/day	<3,274	≥3,274			<2,184	≥2,184
Covariate model	ΔeGFRcys (95% CI)				ΔeGFRcys (95% CI)
Age, LDL-C/HDL-C ratio, HbA1c, SBP	−1.35 (−1.84, −0.85)	−1.13 (−1.47, −0.79)	2.266	0.07	−5.28 (−5.65, −4.91)	−2.42 (−3.88, −0.96)	0.491	0.743

PA was divided into low and high groups based on the median number of steps, as indicated in the table. The dependent variable was eGFRcys. The change in eGFRcys (ΔeGFRcys) was calculated by subtracting eGFRcys at discharge from eGFRcys at 3 months after discharge. Abbreviations as in Tables 1–3.

Patients aged ≥75 years were also divided into low and high PA groups based on the median number of daily steps taken by the OCR (3,622 steps/day). The covariate (adjusted) LMM showed a decreasing trend in the negative change in eGFRcys in the high compared with low PA group when age, LDL-C/HDL-C ratio, HbA1c, SBP, and participation in OCR were used as covariates (F=0.338, P=0.715). The change in eGFRcys was −1.80 and −0.09 mL/min/1.73 m² in the low and high PA groups, respectively.

LMM Comparing Changes in Renal Function Based on the Number of OCR Sessions

In all patients, the covariate (adjusted) LMM showed that eGFRcys increased significantly in the active compared with non-active participant groups when age and PA were used as covariates (Table 5). The change in eGFRcys was +1.58 and +2.15 mL/min/1.73 m² among the non-active and active participants, respectively (F=2.829, P=0.033).

Table 5. Linear Mixed Model Comparing Changes in Renal Function Based on the Number of Sessions of OCR

Covariate model	All patients				Age ≥75 years
	ΔeGFRcys (95% CI)		Group×time interaction		ΔeGFRcys (95% CI)		Group×time interaction
	Non-active participants (n=46)	Active participants (n=24)			Non-active participants (n=17)	Active participants (n=9)
			F	P value			F	P value
Age	+1.25 (1.23, 1.27)	+1.93 (1.87, 1.99)	8.670	<0.001	+1.53 (1.49, 1.56)	−2.47 (−2.54, −2.41)	1.948	0.142
Age, PA	+1.58 (1.56, 1.60)	+2.15 (2.11, 2.19)	2.829	0.033	+0.81 (0.68, 0.94)	−2.31 (−2.84, −1.79)	0.574	0.684
Age, LDL-C/HDL-C ratio	+0.63 (0.43, 0.83)	+2.21 (1.72, 2.70)	0.592	0.67	+1.73 (1.73, 1.73)	−1.76 (−2.84, −1.79)	1.698	0.189
Age, LDL-C/HDL-C ratio, HbA1c, SBP	+0.48 (0.13, 0.83)	+1.80 (1.27, 2.32)	1.134	0.348	+0.66 (0.06, 1.27)	−2.30 (−3.37, −1.22)	0.760	0.563

The OCR group was divided into non-active participants (<13 sessions) and active participants (≥13 sessions) based on the number of sessions attended during the 3-month period after discharge. The dependent variable was eGFRcys. The change in eGFRcys (ΔeGFRcys) was calculated by subtracting eGFRcys at the time of discharge from eGFRcys 3 months after discharge. Abbreviations as in Tables 1–3.

LMM Comparing Changes in Renal Function Based on CKD Severity

Among patients with CKD aged ≥75 years, the covariate (adjusted) LMM showed that eGFRcys increased significantly in the OCR compared with non-OCR group when age, LDL-C/HDL-C ratio, HbA1c, and SBP were used as covariates (F=2.857, P=0.037; Table 6). The change in eGFRcys was −4.05 and +0.48 mL/min/1.73 m² in the non-OCR and OCR groups, respectively. Further, among patients with G3 CKD, eGFRcys increased significantly in the OCR compared with non-OCR group (F=4.114, P=0.009). Similarly, among all patients with G3 CKD, eGFRcys increased significantly in the OCR compared with non-OCR group (F=3.065, P=0.024).

Table 6. Result of Linear Mixed-Model Analysis Comparing the Changes in Renal Function Between the OCR and Non-OCR Groups in Patients With CKD

	All patients				Age ≥75 years
	Non-OCR	OCR	Group×time interaction		Non-OCR	OCR	Group×time interaction
			F	P value			F	P value
CKD G3–G4
No. subjects	29	31			19	18
Covariate models	ΔeGFRcys (95% CI)				ΔeGFRcys (95% CI)
Age, LDL-C/HDL-C ratio	−3.00 (−3.11, −2.88)	+1.86 (1.50, 2.21)	2.855	0.03	−3.90 (−3.93, −3.88)	+0.85 (0.45, 1.25)	3.380	0.018
Age, LDL-C/HDL-C ratio, HbA1c, SBP	−2.88 (−3.02, −2.75)	+1.42 (0.93, 1.90)	1.901	0.12	−4.05 (−4.08, −4.01)	+0.48 (−0.28, 1.24)	2.857	0.037
CKD G3
No. subjects	26	24			16	14
Covariate models	ΔeGFRcys (95% CI)				ΔeGFRcys (95% CI)
Age, LDL-C/HDL-C ratio	−3.39 (−3.58, −3.20)	+1.06 (0.71, 1.42)	5.163	0.031	−4.84 (−4.90, −4.78)	−0.52 (−0.70, 0.19)	5.501	0.002
Age, LDL-C/HDL-C ratio, HbA1c, SBP	−3.07 (−3.25, −2.89)	+0.69 (0.14, 1.24)	3.065	0.024	−4.58 (−4.62, −4.54)	−0.46 (−1.27, 0.34)	4.114	0.009

The dependent variable was eGFRcys. The change in eGFRcys (ΔeGFRcys) was calculated by subtracting eGFRcys at discharge from eGFRcys at 3 months after discharge. CKD, chronic kidney disease; G3, mild to moderate CKD. Abbreviations as in Tables 2,3.

Relationship Between Changes in eGFRcys, Age, and the LDL-C/HDL-C Ratio

The LMM showed that eGFRcys increased significantly in the OCR compared with non-OCR group when age and the LDL-C/HDL-C ratio were used as covariates (Tables 3,5).

Discussion

In this study, the change in eGFRcys was significantly higher in the OCR than non-OCR group, and the decline in eGFRcys among patients aged ≥75 years was suppressed in the OCR compared with non-OCR group. Further, this effect was verified by focusing on patients with CKD. The OCR group had a higher mean PA levels than the non-OCR group, and the change in eGFRcys was significantly higher in the high than low PA group. In addition, the changes in eGFRcys and the LDL-C/HDL-C ratio may be associated. Finally, the hospital readmission rate for exacerbation of heart disease was lower for the OCR than non-OCR group.

The present study is the first to elucidate the protective effects of OCR on renal function assessed by eGFRcys in ≥75-year-old patients with heart disease. Previous studies found that eGFRcr improved significantly in patients with heart disease who participated in OCR focused on aerobic exercise for 3–6 months.^6–9 Further, more frequent participation resulted in more favorable effects.^7,8 Age stratification showed a significant improvement in eGFRcr in patients aged ≤70 years and no improvement in patients aged >70 years.⁹ Analyses based on the severity of renal dysfunction showed improvement in the G3 CKD group, but not in the severe CKD group.^7,9

Hama et al¹⁰ used serum cystatin C (eGFRcys) to assess renal function and found that it increased significantly with 3 months of OCR, indicating that OCR had a favorable effect on renal function; however, there was no control group and PA was not included in that study. In the present study we examined changes in eGFRcys from discharge to 3 months after discharge in the non-OCR and OCR groups and measured PA using an accelerometer. The change in eGFRcys was significantly higher in the OCR group, and there was a significant difference in the change in eGFRcys between the 2 groups. The results for the OCR group were similar to those reported by Hama et al.¹⁰ In addition, in the CKD group verification, eGFRcys increased significantly in the OCR compared with non-OCR group among patients with G3 CKD, similar to previous studies.^7,10 Takaya et al⁷ demonstrated a favorable effect on renal function when patients with G3 CKD performed moderate intensity exercise (50–60% of HR reserve or Borg’s score 12–13). The present study consisted of similar exercise intensity and showed suppressed deterioration of renal function, indicating that “functional renal reserve capacity” may be reversibly improved in patients with G3 CKD.³⁴

This study is also the first to elucidate the effects of increased PA on participation in OCR and the change in renal function in heart disease patients. The OCR group had a higher level of PA than the non-OCR group, which is consistent with previous studies.¹⁴ The change in eGFRcys was significantly higher in the high than low PA group among OCR patients. Earlier studies reported an association between PA and changes in renal function. For example, Robinson-Cohen et al¹³ suggested that increased PA was associated with a lower risk of rapid kidney function decline among older adults when using eGFRcys to evaluate renal function and a questionnaire to evaluate PA. Sato et al¹² concluded that high PA may suppress the decline in renal function in patients after acute myocardial infarction (AMI) using an accelerometer to evaluate PA and eGFRcys to evaluate renal function.

A previous study reported the cardioprotective mechanisms of PA, proving that continued exercise reduces the risk of cardiovascular events.³⁵ The benefits of PA and exercise have also been proven at the cellular level, because aerobic exercise has been shown to have several beneficial antiatherogenic effects, including increasing HDL-C and decreasing LDL-C.^30,31 A moderate increase in PA and exercise improves endothelial function, autonomic function, and systemic blood pressure.^36,37 In the present study, renal function significantly improved in the OCR compared with non-OCR group. It is likely that the cardioprotective effects of PA overcame the negative relationship between heart disease and renal function and may have contributed to the protection of cardiac and renal function.

This study suggests that there is a strong relationship between changes in eGFRcys and the LDL-C/HDL-C ratio, indicating that factors related to arteriosclerosis may have a renoprotective effect. A previous study reported a high correlation between the severity of renal function decline and an increase in the LDL-C/HDL-C ratio. If elevated blood pressure and exacerbated arteriosclerosis are both present, the decline in kidney function is faster.³¹ A previous meta-analysis demonstrated that aerobic exercise significantly improved lipid abnormalities and arteriosclerosis in patients with heart disease.³⁸ Another study demonstrated that continuous aerobic exercise improved renal function via improved lipid metabolism in heart disease patients with CKD,⁶ similar to the findings in the present study. The progression of renal disease causes body fluid regulation disorders, leading to hemodynamic abnormalities in the intrarenal blood vessels, glomerulosclerosis, and other types of endothelial dysfunction that are worsened by the progression of arteriosclerosis. Similarly, cardiovascular diseases activate the renin-angiotensin-aldosterone system, increasing insulin resistance. Cardiovascular diseases also maintain an activated state of the sympathetic nervous system, which is worsened by the progression of arteriosclerotic vascular lesions. The responses of the heart and kidneys are interconnected physiologically and pathologically. Arteriosclerosis is an important risk factor for the deterioration of cardiorenal function;³⁹ therefore, improvements in factors related to arteriosclerosis may have favorable effects on renal function.

It has been suggested that the degree of renal function recovery is less in ≥75-year-old patients with heart disease and that there is a high probability of rapid renal function deterioration if appropriate exercise therapy and disease management are not continued. Therefore, continuous follow-up is essential for heart disease patients aged ≥75 years; however, this is currently insufficient in Japan.^40–42 Endo et al⁴³ reported that elderly patients have difficulty continuing OCR, although it provides significant prophylaxis against cardiac-related readmission compared with non-participation. Therefore, a reliable long-term follow-up and education system for heart disease patients aged ≥75 years needs to be established.

This study has several limitations. The study was not a randomized control trial. To minimize selection bias, we enrolled consecutive patients who underwent ICR within the study period. However, the patients decided whether to participate in OCR after discharge. Therefore, the OCR group likely had a higher rate of compliance than the non-OCR group. In addition, the study was conducted at a single center with a small sample size. Future multicenter studies should include a large sample size. In addition, the patients who had various comorbidities and complications were not considered during analysis. The prevalence of AMI and post cardiac surgery (PCS) were significantly different between the non-OCR and OCR groups. The AMI group was significantly younger with higher PA than the PCS group, which may have affected the results. However, in the covariate (adjusted) LMM analysis, renal protection was more effective in the OCR than non-OCR group for both the AMI and PCS groups. Finally, a short intervention period is another limitation of this study, and long-term studies are needed.

Conclusions

OCR is safe for heart disease patients aged ≥75 years and lowers hospital readmission rates. Participation in OCR and the maintenance of high PA levels may suppress the deterioration of renal function and prevent rehospitalization in all patients with heart disease, including those aged ≥75 years. Participation in OCR increased PA, suppressing the decline in renal function. Therefore, improved long-term follow-up of OCR for heart disease patients aged ≥75 years is necessary.

Acknowledgments

The authors acknowledge all the patients included in this study and the Ohta Nishinouchi Hospital Cardiac Rehabilitation Team. The authors thank Editage for English language editing.

Sources of Funding

This study did not receive any specific funding.

Disclosures

The authors declare that they have no conflicts of interest.

IRB Information

This study was approved by the Ethics Committee of Tohoku University Graduate School of Medicine (Approval no. 2015-1-859) and the Ethics Committee of Ohta General Hospital Foundation, Ohta Nishinouchi Hospital (Approval no. 2015-40).

References

• 1.   Astor BC, Matsushita K, Gansevoort RT, Velde M, Woodward M, Levey AS, et al. Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease: A collaborative meta-analysis of kidney disease population cohorts. Kidney Int 2011; 79: 1331–1340.
• 2.   Hamaguchi S, Tsuchihashi-Makaya M, Kinugawa S, Yokota T, Ide T, Takeshita A, et al. Chronic kidney disease as an independent risk for long-term adverse outcome in patients hospitalized with heart failure in Japan: Report from the Japanese Cardiac Registry of Heart Failure in Cardiology (JCARE-CARD). Circ J 2009; 73: 1442–1447.
• 3.   Shiba N, Matsuki M, Takahashi J, Tada T, Watanabe J, Shimokawa H. Prognostic importance of chronic kidney disease in Japanese patients with chronic heart failure. Circ J 2008; 72: 173–178.
• 4.   Imai E, Horio M, Watanabe T, Iseki K, Yamagata K, Hara S, et al. Prevalence of chronic kidney disease in the Japanese general population. Clin Exp Nephrol 2009; 13: 621–630.
• 5.   Go AS, Chertow GM, Fan D, McCulloch CE, Hsu CY. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296–1305.
• 6.   Toyama K, Sugiyama S, Oka H, Sumida H, Ogawa H. Exercise therapy correlates with improving renal function through modifying lipid metabolism in patients with cardiovascular disease and chronic kidney disease. J Cardiol 2010; 56: 142–146.
• 7.   Takaya Y, Kumasaka R, Arakawa T, Ohara T, Nakanishi M, Noguchi T, et al. Impact of cardiac rehabilitation on renal function in patients with and without chronic kidney disease after acute myocardial infarction. Circ J 2014; 78: 377–384.
• 8.   Kimura S, Ueda Y, Ise T, Yagi S, Iwase T, Nishikawa K, et al. Impact of supervised cardiac rehabilitation on urinary albumin excretion in patients with cardiovascular disease. Int Heart J 2015; 55: 105–109.
• 9.   Iso Y, Kitai H, Kowaita H, Kyuno E, Maezawa H, Hashimoto T, et al. Association of aging with glomerular filtration changes in cardiac rehabilitation participants with chronic kidney disease. Int J Cardiol 2015; 187: 283–285.
• 10.   Hama T, Oikawa K, Ushijima A, Morita N, Matsukage T, Ikari Y, et al. Effect of cardiac rehabilitation on the renal function in chronic kidney disease: Analysis using serum cystatin-C based glomerular filtration rate. Int J Cardiol Heart Vasc 2018; 19: 27–33.
• 11.   Gastro S, Fusco S, Corica F, Rosignuolo M, Marino A, Montesanto A, et al. Estimating glomerular filtration rate in older people. Biomed Res Int 2014; 2014: 916542.
• 12.   Sato T, Kohzuki M, Ono M, Muto M, Osugi T, Kawamura K, et al. Association between physical activity and change in renal function in patients after acute myocardial infarction. PLoS One 2019; 14: e0212100.
• 13.   Robinson-Cohen C, Katz R, Dariush M, Dalrymple LS, Boer L, Sarnak M, et al. Physical activity and rapid decline in kidney function among older adults. Arch Intern Med 2009; 169: 2116–2123.
• 14.   Dibben G, Dala HM, Taylor RS, Doherty P, Tang LH, Hillsdon M. Cardiac rehabilitation and physical activity: Systematic review and meta-analysis. Heart 2018; 104: 1394–1402.
• 15.   Ouchi Y, Rakugi H, Arai H, Akishita M, Ito H, Tobe K, et al. Redefining the elderly as aged 75 years and older: Proposal from the Joint Committee of Japan Gerontological Society and the Japan Geriatrics Society. Geriatr Gerontol Int 2017; 17: 1045–1047.
• 16.   JCS Joint Working Group. Guidelines for rehabilitation in patients with cardiovascular disease (JCS 2012): Digest version. Circ J 2014; 78: 2022–2093.
• 17.   Davidson PM, Cockburn J, Newton PJ, Webster JK, Betihavas V, Howes L, et al. Can a heart failure-specific cardiac rehabilitation program decrease hospitalizations and improve outcomes in high-risk patients? Eur J Cardiovasc Prev Rehabil 2010; 17: 393–402.
• 18.   Fletcher GF, Balady GJ, Amsterdam EA, Chaitman B, Eckel R, Fleg J, et al. Exercise standards for testing and training: A statement for healthcare professionals from the American Heart Association. Circulation 2001; 104: 1694–1740.
• 19.   American College of Sports Medicine Position Stand. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in healthy adults. Med Sci Sports Exerc 1998; 30: 975–991.
• 20.   Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S, Collaborators Developing the Japanese Equation for Estimated GFR. GFR estimation using standardized serum cystatin C in Japan. Am J Kidney Dis 2013; 61: 197–203.
• 21.   Matsuo S, Imai E, Horio M, Yasuda Y, Tomita K, Nitta K, et al. Revised equation for estimated GFR from serum creatinine in Japan. Am J Kidney Dis 2009; 53: 982–992.
• 22.   ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories. ATS statement: Guidelines for the six-minute walk test. Am J Respir Crit Care Med 2002; 166: 1111–1117.
• 23.   Wieczorek M, Wiliński W, Struzik A, Rokita A. Hand grip strength vs. sprint effectiveness in amputee soccer players. J Hum Kinet 2015; 48: 133–139.
• 24.   Bijlsma AY, Pasma JH, Lambers D, Stijntjes M, Blauw GJ, Meskers C, et al. Muscle strength rather than muscle mass is associated with standing balance in elderly outpatients. J Am Med Dir Assoc 2013; 14: 493–498.
• 25.   Inagaki M, Ohtsuki T, Yonemoto N, Kawashima Y, Saitoh A, Oikawa Y, et al. Validity of the Patient Health Questionnaire (PHQ)-9 and PHQ-2 in general internal medicine primary care at a Japanese rural hospital: A cross-sectional study. Gen Hosp Psychiatry 2013; 35: 592–597.
• 26.   Aoyagi Y, Park H, Watanabe E, Park S, Shephard R. Habitual physical activity and physical fitness in older Japanese adults: The Nakanojo Study. Gerontology 2009; 55: 523–531.
• 27.   Heber S, Sallaberger-Lehner M, Hausharter M, Volf I, Ocenasek H, Gabriel H, et al. Exercise-based cardiac rehabilitation is associated with a normalization of the heart rate performance curve deflection. Scand J Med Sci Sports 2019; 29: 1364–1374.
• 28.   Ennis S, Lobley G, Worrall S, Powell R, Kimani PK, Khan AJ, et al. Early initiation of post-sternotomy cardiac rehabilitation exercise training (SCAR): Study protocol for a randomized controlled trial and economic evaluation. BMJ Open 2018; 8: e019748.
• 29.   Imai E, Horio M, Yamagata K, Iseki K, Hara S, Ura N. Slower decline of glomerular filtration rate in the Japanese general population: A longitudinal 10-year follow-up study. Hypertens Res 2008; 31: 433–441.
• 30.   Iseki K, Oshiro S, Tozawa M, Ikemiya Y, Fukiyama K, Takishita S. Prevalence and correlates of diabetes mellitus in a screened cohort in Okinawa, Japan. Hypertens Res 2002; 25: 185–190.
• 31.   Muntner P, Coresh J, Smith JC, Eckfeldt J, Klag MJ. Plasma lipids and risk of developing renal dysfunction: The atherosclerosis risk in communities study. Kidney Int 2000; 58: 293–301.
• 32.   Manttari M, Tinla E, Alikoski T, Manninen V. Effects of hypertension and dyslipidemia on the decline in renal function. Hypertension 1995; 26: 670–675.
• 33.   Roshanravan B, Robinson-Cohen C, Patel K, Ayers E, Littman AJ, Boer LH. Association between physical performance and all-cause mortality in CKD. J Am Soc Nephrol 2013; 24: 822–830.
• 34.   Barai S, Gambhir S, Prasad N, Sharma RK, Ora M. Functional renal reserve capacity in different stages of chronic kidney disease. Nephrology 2010; 15: 350–353.
• 35.   Kachur S, Chongthammakun V, Lavie CJ, De Schutter A, Arena R, Milani RV, et al. Impact of cardiac rehabilitation and exercise training programs in coronary heart disease. Prog Cardiovasc Dis 2017; 60: 103–114.
• 36.   Fletcher GF, Landolfo C, Niebauer J, Ozemek C, Arena R, Lavie CJ. Promoting physical activity and exercise: JACC health promotion series. J Am Coll Cardiol 2018; 72: 1622–1639.
• 37.   Amiya E, Taya M, Watanabe M. Physical activity a useful marker for cardiac rehabilitation? Int Heart J 2015; 56: 583–584.
• 38.   Kelley GA, Kelley KS, Franklin B. Aerobic exercise and lipids and lipoproteins in patients with cardiovascular disease: A meta-analysis of randomized controlled trials. J Cardiopulm Rehabil 2006; 26: 131–144.
• 39.   Ronco C, Happio M, House AA, Anavekar N, Bellomo R. Cardiorenal syndrome. J Am Coll Cardiol 2008; 52: 1527–1539.
• 40.   Goto Y. Current state of cardiac rehabilitation in Japan. Prog Cardiovasc Dis 2014; 56: 557–562.
• 41.   Kohzuki M. Late Phase II cardiac rehabilitation, which is adding life to years and years to life, is still underused in Japan. Circ J 2016; 80: 1697–1699.
• 42.   Kamiya K, Yamamoto T, Makaya MT, Ikegame T, Takahashi T, Sato Y, et al. Nationwide survey of multidisciplinary care and cardiac rehabilitation for patients with heart failure in Japan: An analysis of the AMED-CHF Study. Circ J 2019; 83: 1546–1552.
• 43.   Endo N, Goto A, Suzuki T, Matsuda S, Yasumura S. Exploring the effect of outpatient cardiac rehabilitation on prevention of short-term re-admission: A hospital-based study in Fukushima, Japan. J Gen Fam Med 2016; 17: 170–173.