Open Access

Heterogeneity of clinical indices among the older dialysis population—a study on Japanese dialysis population

Renal Replacement Therapy20173:1

https://doi.org/10.1186/s41100-016-0083-6

Received: 28 April 2016

Accepted: 8 December 2016

Published: 6 February 2017

Abstract

Background

The older dialysis population is increasing, and several studies have investigated the effect of actual age on the relationship between clinical parameters or clinical practice and prognosis. There is concern that biological age does not necessarily reflect the actual effects of aging, and it is important to take heterogeneity into account when considering the older population. In this study, we investigated whether such heterogeneities exist in a large number of clinical parameters and how any disparities might be affected by the domains of clinical indices.

Methods

We investigated the published, summarized results of the Japanese Society for Dialysis Therapy Renal Data Registry, a nationwide survey of the Japanese dialysis population, for the year 2013.

Results

The potential population comprised 306,925 dialysis patients. We investigated coefficient of variations (CVs) of 25 clinical indices across age groups and compared them with the CVs of the 45–59 years age group as a reference group, using tests for homogeneity of variances. Almost all variables showed some heterogeneity by age. The domain for muscle mass or visceral protein exhibited greater heterogeneity in older patients, but surprisingly, the domain for body mass or physique exhibited less heterogeneity. The mean values of most parameters declined in the older population.

Conclusions

This study demonstrated that variations exist in clinical indices in the older dialysis population. Apart from biological age, we should take these heterogeneities into account when interpreting the findings of clinical studies involving older dialysis patients and in their individualized management.

Keywords

Aging phenotype Biological age Coefficient of variances Heterogeneity Older patients Wasting

Background

The dialysis population has become much older in recent years, and the proportion of elderly patients is increasing in many registries [1, 2]. There are several clinical problems to overcome in the older hemodialysis population. One of the most important involves malnutrition-related syndromes, such as sarcopenia [3], protein energy wasting [4], and frailty [5]. These syndromes are in fact associated with a worse prognosis in the dialysis population as a whole [6, 7]; these phenotypes are considered to be most important in the management of the older population.

Strategies against complications for older dialysis patients sometimes differ from those for their younger counterparts [8]. However, the definition of “older” patients varies in different reports [911]. The reason for this may be due to the heterogeneity among the elderly population in terms of phenotype. Some patients look younger than their actual age, while some look older. Moss et al. reported that the “surprise” question, “Would I be surprised if this patient died in the next year?” correlated well with 1 year survival [12], indicating that the phenotype, presumably wasting or malnutrition, seems important for survival.

We speculated that such phenotype of the “older” population does matter and that it is important to take the heterogeneities of such patients into account, which might affect the older phenotype. However, there have been no reports on the heterogeneity of the elderly population. Therefore, here, we focused on the heterogeneity of the older dialysis population and investigated which clinical indices exhibited large heterogeneity and were potentially related to a poor prognosis, using national registry data compiled by the Japanese Society for Dialysis Therapy (JSDT). It is anticipated that our results will help the clinician to undertake individualized management of older dialysis patients.

Methods

Population

Data from the annual Japanese Society for Dialysis Therapy Renal Data Registry (JRDR) survey run by JSDT were used for this analysis. The details of this registry have been described previously [13]. We analyzed the data collected in 2013. The summarized tables are recorded in CD-ROM or are reported on the internet. In this analysis, we used data that appears on JSDT’s website [14].

The potential population for this study was the entire dialysis population as of December 31, 2013, namely 306,925 patients. The population was divided by 15-year age groups. We excluded patients with missing data on age (n = 12) and patients who were younger than 15 years (n = 116).

Clinical indices

The following clinical indices were investigated: (1) mineral and bone disorder markers (corrected calcium and phosphate); (2) malnutrition-related factors (total cholesterol, high-density lipoprotein (HDL) cholesterol, non-HDL cholesterol, percent creatinine generation rate (%CGR) [15], normalized protein catabolic rate (nPCR) [16], and albumin); (3) body mass and volume status (body height, postdialytic body weight (BW), body mass index (BMI), BW change, and fractional BW change); (4) dialysis prescription and its consequences (dialysis duration, spKt/V [16], eqKt/V [17], blood flow rate, predialytic blood urea, and creatinine); (5) circulatory factors (systolic blood pressure, diastolic blood pressure, and pulse rate); and (6) other factors (hemoglobin (Hb) level, glycoalbumin, and HbA1c). Postdialytic BW, dialysis duration, and blood flow rate were investigated among the patients who were treated by hemodialysis, hemodiafiltration, or hemoperfusion. The values of%CGR, nPCR, BW change, fractional BW change, spKt/V, eqKt/V, and creatinine were investigated only in patients who were treated by hemodialysis, hemodiafiltration, or hemoperfusion three times weekly with a dialysis history of ≥2 years. Glycoalbumin and HbA1c were investigated among those who had history of diabetes or were on glucose-lowering therapies. Parameters other than those described above were studied in the entire dialysis population.

Statistical analysis

This analysis consists of two parts: one is to determine the heterogeneities of age populations, and the other is to compare them with the standard of 45–59 year-old group.

The heterogeneity was measured by coefficient of variations (CV) in this analysis, because standard deviation (SD) does not necessarily express the degree of heterogeneity when the averages vary. The tables of the JRDR summarized data give the number of patients and mean values and SD for each parameter in all age groups. CVs were calculated by dividing the SD by the mean for each age category. The larger CV is more heterogenic or heteroscedastic the population can be considered.

Relative CVs were calculated by dividing the CVs of each age group by that of the 45–59 years age group. If the value is larger than one, the index is more heterogenic in older population than the younger counterpart. Relative means were also calculated as the ratio of the mean of each age group to that of the 45–59 years age group, similarly to relative CVs.

Finally, we statistically compared the heterogeneities across age groups by investigating the homogeneity of variances. In details, we calculated variances by squaring standard deviations. The tests for homogeneity of variances were performed on the squared relative CVs using Bartlett’s test among overall groups and F tests for each pair of age groups. The former investigated the homogeneity of the variances by age groups as a whole. When the test indicated statistical significance, it shows that the clinical index is heterogenic across age groups. On the other hand, the latter tested statistical significance by comparing the CV of each age group with that of the reference group of ages 45–59. When a test was significant, it indicates that the clinical index is more heterogenic in that age group compared to that in the 45–59-year-old age group. Post hoc adjustment for multiple comparisons was done by Bonferroni correction. The p values less than 0.05 were considered significant. For multiple comparisons, p values less than 0.01 were considered significant because five pairs of comparisons were performed with the 45–59 years age group as the reference group. All statistical calculations and analyses were performed using Microsoft Excel 2013 (Microsoft Corporation, Redmond, WA).

Ethical issues

This study is the secondary analyses of the summarized and published data and does not deal with any personal information. Therefore, we consider that this study is deemed exempt from the Institutional Review Board or Ethics Committee approval. This study adhered to the Declaration of Helsinki.

Results

The demographics and characteristics of the total population who underwent dialysis therapy in 2013 are shown in Table 1 [1]. Mean age was 67.21 ± 12.51 years (mean ± SD). The age groups 15–29, 30–44, 45–59, 60–74, 75–89, and ≥90 years accounted for 0.37, 4.91, 18.56, 45.87, 28.72, and 1.54% of the total population, respectively.
Table 1

Background characteristics of the patients

Number of patients

306,925

Sex (male/female)

194,965/111,960

Age (mean ± SD)

67.21 ± 12.51

Vintage (years, number of patients, %)

 <2

68,475

(22.3)

 2–4

76,589

(25.0)

 5–9

77,197

(25.2)

 10–14

39,490

(12.9)

 15–19

20,874

(6.8)

 20–24

11,421

(3.7)

 25–29

6611

(2.2)

 30+

6155

(2.0)

Primary diagnoses (numbers, %)

 Diabetes

115,484

(37.6)

 Glomerulonephritis

99,492

(32.4)

 Nephrosclerosis

26,569

(8.7)

 Polycystic kidney

10,683

(3.5)

 Others or unknown

54,697

(17.8)

Table 2 shows the number of patients and their proportion of sex, vintage, and primary diagnoses for each age category. In the older age groups, there was a higher proportion of female patients and a higher proportion of patients with the primary diagnosis of nephrosclerosis; less predominant were the proportion of patients with a long history of dialysis and the proportion with diabetes as the primary diagnosis.
Table 2

Patient characteristics by age group

Age groups

15–29

30–44

45–59

60–74

75–89

90+

Sex

 Male

749 (66.8)

10,315 (68.4)

38,610 (67.8)

91,189 (64.8)

51,997 (59.0)

2032 (42.9)

 Female

372 (33.2)

4756 (31.6)

18,340 (32.2)

49,583 (35.2)

36,150 (41.0)

2704 (57.1)

Vintage (years, number of patients, %)

 <2

383 (34.2)

3556 (23.6)

11,484 (20.2)

28,358 (20.2)

23,464 (26.6)

1163 (24.6)

 2–4

323 (28.8)

3756 (24.9)

12,996 (22.8)

32,734 (23.3)

25,177 (28.6)

1565 (33.1)

 5–9

263 (23.5)

3564 (23.7)

13,807 (24.3)

35,296 (25.1)

22,800 (25.9)

1447 (30.6)

 10–14

103 (9.2)

2134 (14.2)

7782 (13.7)

19,446 (13.8)

9598 (10.9)

425 (9.0)

 15–19

34 (3.0)

1176 (7.8)

4673 (8.2)

10,882 (7.7)

4007 (4.5)

102 (2.2)

 20–24

11 (1.0)

605 (4.0)

2890 (5.1)

6174 (4.4)

1715 (1.9)

26 (0.5)

 25–29

4 (0.4)

206 (1.4)

1734 (3.0)

3890 (2.8)

774 (0.9)

3 (0.1)

 30+

0 (0.0)

69 (0.5)

1564 (2.7)

3948 (2.8)

572 (0.6)

2 (0.0)

Primary diagnoses (numbers, %)

 Diabetes

61 (5.4)

4268 (28.3)

21,653 (38.0)

57,903 (41.1)

30,672 (34.8)

920 (19.4)

 Glomerulonephritis

389 (34.7)

5635 (37.4)

20,180 (35.4)

46,654 (33.1)

25,273 (28.7)

1344 (28.4)

 Nephrosclerosis

34 (3.0)

614 (4.1)

2411 (4.2)

8917 (6.3)

13,311 (15.1)

1279 (27.0)

 Polycystic kidney

21 (1.9)

433 (2.9)

2738 (4.8)

5373 (3.8)

2050 (2.3)

63 (1.3)

 Others or unknown

615 (54.9)

4119 (27.3)

9967 (17.5)

21,917 (15.6)

16,832 (19.1)

1130 (23.9)

The number of patients (percentage) in each age group is shown. Patients with missing or unknown values for age or background characteristics were excluded. Therefore, the sum of the numbers is not equal to the total population shown in Table 1

Table 3 shows the SDs and means of the variables investigated as well as the number of the patients. Table 4 shows the CVs relative to those of the 45–59 years age group, which is used as the reference group. The overall differences of variances, including chi-square values and p values calculated with the Bartlett test, are shown, as well as p values for the differences of variances of the age groups, taking the variance of the 45–59 years age group as the references. For almost all variables, the variances differed by age group. However, the trends themselves differed by index. We can categorize these indices according to the relative CV for the 75–89 years age group into four categories: (1) relative CVs of more than 1.1 (indicating the most heteroscedastic indices), namely for %CGR, albumin, creatinine, fractional BW reduction, blood urea nitrogen, and diastolic blood pressure; (2) relative CVs of 1.0–1.1 (mildly heterogeneous), namely for phosphate, total cholesterol, nPCR, body height, BW reduction, blood flow rate, systolic blood pressure, pulse rate, and Hb; (3) relative CVs of 0.9–1.0 (less heterogeneous), namely for corrected calcium, non-HDL, HDL, session duration, spKt/V, and eqKt/V; and (4) relative CVs of less than 0.9 (least heterogeneous), namely for BW, BMI, glycoalbumin, and HbA1c.
Table 3

Details of investigated clinical indices by age group

Age groups

15–29

30–44

45–59

60–74

75–89

90+

 

n

Mean

SD

n

Mean

SD

n

Mean

SD

n

Mean

SD

n

Mean

SD

n

Mean

SD

Corrected calcium (mg/dl)

982

9.23

1.01

13,372

9.21

0.95

50,910

9.28

0.95

125,857

9.29

0.92

78,390

9.28

0.89

4159

9.32

0.88

Phosphate (mg/dl)

982

6.20

1.75

13,395

5.94

1.70

51,003

5.66

1.51

126,082

5.24

1.38

78,480

4.86

1.35

4163

4.58

1.35

Total cholesterol (mg/dl)

733

151.42

34.05

10,072

161.17

36.00

38,259

161.91

36.68

94,475

156.90

35.81

58,543

152.40

34.76

3074

152.01

34.89

Non-HDL (mg/dl)

572

98.53

30.52

8163

110.54

34.88

30,913

111.28

35.04

75,683

107.96

33.74

46,507

105.29

32.34

2386

105.10

31.91

HDL cholesterol (mg/dl)

649

51.96

15.40

9291

50.68

17.81

35,085

50.48

18.04

85,526

48.81

16.71

52,604

47.17

15.33

2707

47.06

14.59

%Creatinine generation rate (%)

576

97.84

22.30

9279

102.51

22.47

37,051

101.84

23.50

92,017

100.60

25.51

52,501

96.04

27.78

2763

89.88

30.00

Normalized protein catabolic rate (g/kg/day)

576

0.96

0.19

9283

0.93

0.18

37,071

0.90

0.18

92,080

0.88

0.18

52,547

0.83

0.17

2768

0.81

0.18

Albumin (g/dl)

957

3.94

0.46

13,123

3.86

0.40

49,868

3.76

0.39

123,713

3.62

0.41

77,239

3.44

0.44

4082

3.26

0.45

Body height (cm)

951

160.35

12.28

12,879

165.28

9.76

49,316

164.41

9.00

121,116

160.36

8.77

74,553

155.74

8.97

3895

150.46

9.31

Postdialytic body weight (kg)

904

55.07

15.10

12,921

64.44

17.75

49,320

62.10

14.59

123,611

55.35

11.02

77,661

49.86

9.67

4130

43.68

8.29

Body mass index (kg/m2)

907

20.96

4.33

12,468

23.40

5.26

47,739

22.89

4.49

117,550

21.51

3.55

72,249

20.56

3.24

3751

19.36

3.22

Body weight reduction (kg)

598

2.84

1.20

9692

3.19

1.22

38,748

2.95

1.10

96,978

2.52

0.91

55,482

2.17

0.83

2954

1.84

0.78

Fractional body weight reduction (%)

588

5.23

1.75

9627

5.09

1.67

38,616

4.83

1.57

96,725

4.61

1.56

55,325

4.41

1.61

2945

4.26

1.70

Session duration (h)

928

4.11

0.60

13,153

4.15

0.61

50,237

4.11

0.56

125,686

4.00

0.50

79,179

3.84

0.51

4228

3.68

0.54

spKt/V

576

1.55

0.33

9260

1.47

0.32

36,997

1.47

0.31

91,943

1.50

0.29

52,461

1.48

0.30

2759

1.45

0.31

eqKt/V

576

1.35

0.30

9270

1.29

0.30

37,018

1.29

0.28

91,982

1.31

0.27

52,478

1.29

0.27

2762

1.25

0.29

Blood flow rate (ml/min)

910

215.90

40.58

12,992

223.89

39.60

49,599

218.83

37.45

124,081

206.67

34.32

78,016

192.68

33.07

4149

179.45

32.49

Predialytic blood urea (mg/dl)

601

69.22

16.40

9729

67.31

15.90

38,895

64.89

15.54

97,289

62.47

15.51

55,647

58.45

15.43

2956

56.73

15.77

Creatinine (mg/dl)

601

13.83

3.12

9727

13.38

2.79

38,905

11.99

2.62

97,292

10.48

2.46

55,646

8.88

2.30

2955

7.42

2.15

Systolic blood pressure (torr)

844

147.58

22.58

11,584

152.31

24.84

43,661

153.49

24.07

108,150

152.08

23.81

67,379

148.93

24.16

3586

145.85

25.04

Diastolic blood pressure (torr)

843

87.65

16.31

11,557

88.44

15.51

43,545

84.86

14.03

107,857

77.84

13.33

67,193

72.10

13.14

3576

68.01

13.34

Pulse rate (/min)

785

80.69

13.35

11,009

81.29

12.92

41,376

78.50

12.58

102,757

74.40

12.46

64,244

72.38

12.63

3396

71.16

12.49

Hemoglobin (g/dl)

971

10.83

1.31

13,204

10.90

1.27

50,256

10.85

1.26

124,487

10.70

1.24

77,553

10.52

1.27

4096

10.37

1.25

Glycoalbumin (%)

25

21.38

7.09

1692

21.41

7.01

8742

20.82

5.82

23,723

21.24

5.21

13,001

21.35

4.93

420

21.41

5.31

A1c (%)

39

6.34

1.51

2363

6.64

1.51

12,166

6.40

1.29

32,540

6.18

1.12

17,481

6.00

1.05

544

5.83

0.99

HDL high-density lipoproteins, spKt/V single pool Kt/V, eqKt/V equilibrated Kt/V, HbA1c hemoglobin A1c, n number of patients, SD standard deviations

Table 4

Trends in relative and standardized standard deviations of all clinical indices by age group

Age groups

15–29

30–44

45–59

60–74

75–89

90+

Trend analysis

 

Relative CV

p value

Relative CV

p value

Relative CV

p value

Relative CV

p value

Relative CV

p value

Relative CV

p value

Chi-square values

p value

Corrected calcium (mg/dl)

1.069

0.001

1.008

0.134

1

Ref

0.967

<0.001

0.937

<0.001

0.922

<0.001

356.9

<0.001

Phosphate (mg/dl)

1.058

0.006

1.073

<0.001

1

Ref

0.987

0.000

1.041

<0.001

1.105

<0.001

463.8

<0.001

Total cholesterol (mg/dl)

0.993

0.396

0.986

0.038

1

Ref

1.007

0.042

1.007

0.073

1.013

0.160

11.3

0.080

Non-HDL (mg/dl)

0.984

0.299

1.002

0.405

1

Ref

0.993

0.057

0.975

<0.001

0.964

0.008

32.9

<0.001

HDL cholesterol (mg/dl)

0.829

<0.001

0.983

0.021

1

Ref

0.958

<0.001

0.909

<0.001

0.868

<0.001

482.4

<0.001

%creatinine generation rate (%)

0.988

0.347

0.950

<0.001

1

Ref

1.099

<0.001

1.254

<0.001

1.446

<0.001

3282.4

<0.001

Normalized protein catabolic rate (g/kg/day)

0.990

0.370

0.968

<0.001

1

Ref

1.023

<0.001

1.024

<0.001

1.111

<0.001

122.8

<0.001

Albumin (g/dl)

1.126

<0.001

0.999

0.448

1

Ref

1.092

<0.001

1.233

<0.001

1.331

<0.001

3463.1

<0.001

Body height (cm)

1.399

<0.001

1.079

<0.001

1

Ref

0.999

0.401

1.052

<0.001

1.130

<0.001

693.7

<0.001

Postdialytic body weight (kg)

1.167

<0.001

1.172

<0.001

1

Ref

0.847

<0.001

0.825

<0.001

0.808

<0.001

5473.5

<0.001

Body mass index (kg/m2)

1.053

0.013

1.146

<0.001

1

Ref

0.841

<0.001

0.803

<0.001

0.848

<0.001

5456.5

<0.001

Body weight reduction (kg)

1.133

<0.001

1.026

0.001

1

Ref

0.968

<0.001

1.026

<0.001

1.137

<0.001

401.1

<0.001

Fractional body weight reduction (%)

1.029

0.155

1.009

0.123

1

Ref

1.041

<0.001

1.123

<0.001

1.228

<0.001

884.0

<0.001

Session duration (h)

1.071

0.001

1.079

<0.001

1

Ref

0.917

<0.001

0.975

<0.001

1.077

<0.001

1243.2

<0.001

spKt/V

1.010

0.367

1.032

<0.001

1

Ref

0.917

<0.001

0.961

<0.001

1.014

0.160

604.6

<0.001

eqKt/V

1.024

0.207

1.071

<0.001

1

Ref

0.950

<0.001

0.964

<0.001

1.069

<0.001

410.3

<0.001

Blood flow rate (ml/min)

1.098

<0.001

1.034

<0.001

1

Ref

0.970

<0.001

1.003

0.239

1.058

<0.001

243.6

<0.001

Predialytic blood urea (mg/dl)

0.989

0.364

0.986

0.044

1

Ref

1.037

<0.001

1.102

<0.001

1.161

<0.001

620.9

<0.001

Creatinine (mg/dl)

1.032

0.130

0.954

<0.001

1

Ref

1.074

<0.001

1.185

<0.001

1.326

<0.001

1992.9

<0.001

Systolic blood pressure (torr)

0.976

0.164

1.040

<0.001

1

Ref

0.998

0.342

1.034

<0.001

1.095

<0.001

184.0

<0.001

Diastolic blood pressure (torr)

1.126

<0.001

1.061

<0.001

1

Ref

1.036

<0.001

1.102

<0.001

1.186

<0.001

692.0

<0.001

Pulse rate (/min)

1.032

0.100

0.992

0.139

1

Ref

1.045

<0.001

1.089

<0.001

1.095

<0.001

449.3

<0.001

Hemoglobin (g/dl)

1.042

0.035

1.003

0.315

1

Ref

0.998

0.289

1.040

<0.001

1.038

0.001

185.3

<0.001

Glycoalbumin (%)

1.186

0.089

1.171

<0.001

1

Ref

0.877

<0.001

0.826

<0.001

0.887

<0.001

703.6

<0.001

A1c (%)

1.182

0.054

1.128

<0.001

1

Ref

0.899

<0.001

0.868

<0.001

0.842

<0.001

564.6

<0.001

Indicated CVs were relative (i.e., SD of the 45–59 years age group was set as 1) CV. The p values for each age group were calculated from F tests compared with those of the 45 and 59 years age group. Trend analysis was performed using Bartlett’s test; chi-square and p values are shown. HDL high-density lipoproteins, spKt/V single pool Kt/V, eqKt/V equilibrated Kt/V, HbA1c hemoglobin A1c, n number of patients, SD standard deviation, ref reference

Figure 1 indicates that the disparities in representative indices between several domains differ significantly across the age groups. The relative CVs of indicators relating to the muscle mass or visceral proteins increased steadily with age, becoming more heteroscedastic. The parameters related to dietary intake showed a modest relationship across the age groups. On the other hand, indices of body mass demonstrated that the dispersion progressively decreased with the age of the population.
Fig. 1

Heterogeneity of representative clinical parameters by age group. The figure shows the trends in relative coefficient of variations (CV) (i.e., the CV of the 45–59 years age group was set as 1) for the clinical parameters related to a muscle mass or visceral proteins, b dietary intake, and c body mass or physique. The CVs for indices of muscle mass or visceral proteins became larger among the older population, whereas the CVs for those of body mass became smaller. Indices for dietary intake tended to be dispersed in the very old population. CV coefficient of variance, %CGR percent creatinine generation ratio, a surrogate marker of muscle mass, Cr creatinine, Alb albumin, nPCR normalized protein catabolic rate, BW body weight, BMI body mass index, ref reference

Figure 1 was designated to demonstrate the CVs by themselves so that the heteroscedasticity can be recognized easily. On the other hand, the CVs were calculated from the SDs and means. The means are also subjected to change by age groups, and the relative means indicate the trend of the distributions of the absolute values as a whole by age groups. Therefore, in Fig. 2, we added the information about the relative means to the results shown in Fig. 1 so that information about whether the absolute values tended to become smaller among the older population as well as the information about the heteroscedasticity. This figure indicates that the mean values of all indices included in Fig. 1 were lower in the older age groups, irrespective of relative CVs. This fact might indicate that the normal aging process accompanies universal declines in clinical indices.
Fig. 2

Evolution of the relationship between mean and SD for representative indices among age groups ≥45 years. The relationship between the set of relative coefficients of variations (CV) and relative mean values are shown. The 45–59 years age group was set as a standard for both CV and means. The mean of all the indices decreased to a large degree in the older population. Each point indicates the age groups of 45–59, 60–74, 75–89, and ≥90 years. Open circles, %CGR; open squares, Alb; open triangles, Cr; closed circles, nPCR; closed squares, phosphate; closed triangles, body weight reduction; gray circles, body weight; and gray squares, body mass index

Discussion

Heterogeneity as determined by CV increased with age in some domains, especially that of muscle mass or visceral proteins. However, in some other domains such as body mass, the disparities seem to be diminished in the older population. This is the first study investigating the heteroscedasticity of clinical indices, using nationwide registry data.

There were differences among the domains of indices; factors related to sarcopenia or wasting were most prominently dispersed. Moreover, the mean values of these parameters declined with age. Therefore, muscle mass and albumin level generally decreases with age, but their speed of decline varies among individuals. The decrease in these parameters, as discussed above, reportedly relates to poor survival relating to wasting [6, 7]. Therefore, the heterogeneity of these clinical indices might reflect the diversity of the older population in terms of “old” phenotypes and relate to a worse prognosis. Moreover, the mean values of CGR and albumin decreased less in older population, compared to other indices as shown in Fig. 2. This fact might reflect that CGR and albumin are less affected by aging and they could be good markers of aging phenotype. In this analysis, we investigated the disparity in muscle mass in terms of CGR. However, muscle strength is reported to relate to survival to a greater degree [6]. The concept of dynapenia has been proposed [18, 19]. Future studies investigating the heterogeneity of muscular strength and its relation to survival are required.

Albumin has been used as a marker of malnutrition, but many other factors also affect albumin levels [20]. Dietary restriction only does not decrease the albumin level and it is now regarded as a factor of wasting [20]. The 2013 annual JRDR survey, the source of the data analyzed in this study, also investigated C-reactive protein (CRP). However, the distribution of CRP does not follow a normal distribution, and its heterogeneity cannot be assessed by SD or CV as we performed on other variables because the variables that do not follow normal distribution cannot be summarized by SDs or means. Therefore, we could not investigate the heterogeneity of CPR, and we cannot draw the conclusion that the disparity of albumin relates to inflammation status.

Dietary intake is reduced in the elderly population because of reduced appetite [21]. Moreover, the degree of appetite itself relates to prognosis [21]. In this study, we used nPCR as a surrogate of dietary intake. On the other hand, phosphate is contained in foods that contain much protein [22]. Phosphate binders and dialysis dosage, especially session length, also affect the level of phosphate [22]. It is interesting to note then that nPCR and phosphate exhibited similar trends. Intradialytic BW reduction can be related to sodium intake during interdialytic periods [23], which also followed a similar trend to that of nPCR and phosphate levels. Therefore, these clinical indices can be considered as surrogate markers of dietary intake in patients.

Surprisingly, the distribution of body mass tended to be smaller in the older population even though the disparities of muscle mass become wide among the elderly population. An inverse relationship can be observed in the dialysis population [24]; patients with larger BMI experience better survival [25]. However, in the Asian population, such relationship might be weaker than in the white or black populations [25]. Female patients tend to have better survival among the Japanese population, and the average age of female patients is higher than their male counterparts [1]. Actually, Table 2 shows that the proportion of female patients was larger among the older age groups. Obviously, female patients have a smaller physique than male patients. Therefore, the present results might be attributed to the predominance of female patients in the older age groups. On the other hand, it is possible that the findings of the smaller and relatively homogeneous physique in older patients merely reflect the trend in the general population because this study did not examine variances in the general population.

Moreover, there was a discrepancy in terms of heteroscedasticity between the indices regarding body mass such as BMI or BW and the indices regarding muscle mass such as %CGR or creatinine. The former showed to be less heterogenic, while the latter showed to be more heterogenic in older population. BMI is one of the simplest indices to describe the physiques. However, it cannot distinguish fat mass from muscle mass. The dialysis patients are shown to reduce muscle mass, while they often lose less amount of fat or even gain it during the course of dialysis therapy [26]. Therefore, BMI can underestimate the changes in muscle mass [27]. This point might cause the discrepancy of heteroscedasticity between body mass and muscle mass. On the other hand, body height can contribute to BMI. However, as Table 4 shows, the heterogeneity of height remained small even in the older population. Moreover, the mean values of body height were less in older population. Several recent studies demonstrated that taller dialysis patients might experience worse survival [28, 29]. Thus, the smaller and less heterogenic older population in terms of their body height might be due to the survival effects partly.

Nonetheless, we should pay attention to the clinical indices which exhibited large heterogeneity, such as creatinine generation rate, creatinine, and albumin, when we perform clinical practice on the older population. The deterioration of these indices might reflect the degree of the morbid conditions among older population. Future study will be required to investigate the relationship to the clinical outcomes, such as mortality, quality of life (QOL), and activity of daily living (ADL). After that, the interventions to the indices are warranted to improve the outcomes of the patients.

This study has several limitations. The first is its cross-sectional and observational nature. Therefore, we cannot draw any conclusions about the time course or cause-effect relationship. The second is that this study was based on results that appeared in summarized tables, and therefore, detailed investigations, such as further stratification by sex or primary diagnoses, could not be performed. The third is that the data that we used did not contain the data for clinical outcomes. Therefore, we could not investigate the effect of heterogeneity observed on mortality, QOL, or ADL. The fourth is that the JRDR data were almost exclusively derived from Japanese people. Therefore, the applicability of our findings to other ethnicities is unclear. Lastly, data was missing on some parameters in some patients. Although the number of the patients was sufficiently large, which is the most important advantage of this study, it is possible that errors exist.

Although there are several limitations in this study, the strengths of this study are the large number of patients and parameters that were examined. These results may contribute to the individualized management of the older dialysis population. Further studies on the cause of disparities and the relationship to the prognosis of the patients are required. We anticipate that the findings of such studies will provide a better understanding of phenotypes in the older dialysis population.

Conclusions

This is the first study to investigate heterogeneity in the older dialysis population. There were some disparities in heterogeneity across the parameters. Clinical parameters related to muscle mass or wasting exhibited larger heterogeneity, while those related to body mass or BW exhibited less heterogeneity. Individualized management for older dialysis patients with these heterogeneities might contribute to better outcomes in this population.

Abbreviations

%CGR: 

Percent creatinine generation rate

ADL: 

Activity of daily living

BMI: 

Body mass index

BW: 

Body weight

CD-ROM: 

Compact disk read-only memory

CRP: 

C-reactive protein

CV: 

Coefficient of variation

eqKt/V: 

Equilibrated Kt/V

Hb: 

Hemoglobin

HbA1c: 

Hemoglobin A1c

HDL: 

High-density lipoprotein

JRDR: 

The Japanese Society for Dialysis Therapy Renal Data Registry

JSDT: 

The Japanese Society for Dialysis Therapy

nPCR: 

Normalized protein catabolic rate

QOL: 

Quality of life

SD: 

Standard deviation

spKt/V: 

Single pool Kt/V

Declarations

Acknowledgements

The data reported here were provided by the JSDT. The interpretation and reporting of these data are the responsibility of the authors and do not reflect the official interpretation or views of the JSDT.

Funding

The authors did not have specific funding source to be disclosed.

Availability of data and materials

The data we used for the analysis have been already publicized by the JSDT on the web pages.

Authors’ contributions

NH conceived of the study design, analyzed the data, and wrote the manuscript. SS and MN made the critical revisions to the manuscript and helped to finalize the article. All authors read and approved the final manuscript.

Competing interests

NH has received research consulting fees from Kyowa Hakko Kirin Co., Ltd.; lecture fees from Bayer Yakuhin, Ltd., Chugai Pharmaceutical Co., Ltd., Kyowa Hakko Kirin Co., Ltd., and Nikkiso Co., Ltd.; and grant support from Chugai Pharmaceutical Co., Ltd. and Kyowa Hakko Kirin Co., Ltd. NH was also a member of a division that is funded by Terumo Corporation. The other authors declare that they have no competing interests.

Consent for publication

This study does not include any individual person’s data. Therefore, this point is not applicable.

Ethics approval and consent to participate

Although this study employs the data originally derived from the patients, the data is widely available from the JSDT web pages and the data have already been anonymized. We the authors recognized that this study was not within the scope of the Ethical Guidelines for Medical and Health Research Involving Human Subject issued by the Ministry of Health, Labour and Welfare of Japan in December 2014, pursuant to the chapter 1 of part 3 of the guideline. Therefore, the ethics approval by the institutional review boards is exempted.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
Department of Blood Purification, Kidney Center, Tokyo Women’s Medical University
(2)
Division of Nephrology and Endocrinology, The University of Tokyo School of Medicine

References

  1. Masakane I, Nakai S, Ogata S, Kimata N, Hanafusa N, Hamano T, Wakai K, Wada A, Nitta K. An overview of regular dialysis treatment in Japan (as of 31 December 2013). Ther Apher Dial. 2015;19:540–74.Google Scholar
  2. Saran R, Li Y, Robinson B, Ayanian J, Balkrishnan R, Bragg-Gresham J, Chen JT, Cope E, Gipson D, He K, et al. US renal data system 2014 annual data report: epidemiology of kidney disease in the United States. Am J Kidney Dis. 2015;66:Svii. S1-305.Google Scholar
  3. Fielding RA, Vellas B, Evans WJ, Bhasin S, Morley JE, Newman AB, Abellan van Kan G, Andrieu S, Bauer J, Breuille D, et al. Sarcopenia: an undiagnosed condition in older adults. Current consensus definition: prevalence, etiology, and consequences. International working group on sarcopenia. J Am Med Dir Assoc. 2011;12:249–56.Google Scholar
  4. Fouque D, Kalantar-Zadeh K, Kopple J, Cano N, Chauveau P, Cuppari L, Franch H, Guarnieri G, Ikizler TA, Kaysen G, et al. A proposed nomenclature and diagnostic criteria for protein-energy wasting in acute and chronic kidney disease. Kidney Int. 2008;73:391–8.View ArticlePubMedGoogle Scholar
  5. Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, Seeman T, Tracy R, Kop WJ, Burke G, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–56.View ArticlePubMedGoogle Scholar
  6. Isoyama N, Qureshi AR, Avesani CM, Lindholm B, Barany P, Heimburger O, Cederholm T, Stenvinkel P, Carrero JJ. Comparative associations of muscle mass and muscle strength with mortality in dialysis patients. Clin J Am Soc Nephrol. 2014;9:1720–8.View ArticlePubMedPubMed CentralGoogle Scholar
  7. Johansen KL, Chertow GM, Jin C, Kutner NG. Significance of frailty among dialysis patients. J Am Soc Nephrol. 2007;18:2960–7.View ArticlePubMedGoogle Scholar
  8. Hanafusa N, Nomura T, Hasegawa T, Nangaku M. Age and anemia management: relationship of hemoglobin levels with mortality might differ between elderly and nonelderly hemodialysis patients. Nephrol Dial Transplant. 2014;29:2316–26.View ArticlePubMedPubMed CentralGoogle Scholar
  9. Locatelli F, Pisoni RL, Combe C, Bommer J, Andreucci VE, Piera L, Greenwood R, Feldman HI, Port FK, Held PJ. Anaemia in haemodialysis patients of five European countries: association with morbidity and mortality in the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant. 2004;19:121–32.View ArticlePubMedGoogle Scholar
  10. Canaud B, Tong L, Tentori F, Akiba T, Karaboyas A, Gillespie B, Akizawa T, Pisoni RL, Bommer J, Port FK. Clinical practices and outcomes in elderly hemodialysis patients: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Clin J Am Soc Nephrol. 2011;6:1651–62.View ArticlePubMedPubMed CentralGoogle Scholar
  11. Foote C, Ninomiya T, Gallagher M, Perkovic V, Cass A, McDonald SP, Jardine M. Survival of elderly dialysis patients is predicted by both patient and practice characteristics. Nephrol Dial Transplant. 2012;27:3581–7.View ArticlePubMedGoogle Scholar
  12. Moss AH, Ganjoo J, Sharma S, Gansor J, Senft S, Weaner B, Dalton C, MacKay K, Pellegrino B, Anantharaman P, Schmidt R. Utility of the “surprise” question to identify dialysis patients with high mortality. Clin J Am Soc Nephrol. 2008;3:1379–84.View ArticlePubMedPubMed CentralGoogle Scholar
  13. Hanafusa N, Nakai S, Iseki K, Tsubakihara Y. Japanese society for dialysis therapy renal data registry—a window through which we can view the details of Japanese dialysis population. Kidney Int Suppl. 2015;5:15–22.View ArticleGoogle Scholar
  14. An Overview of Regular Dialysis Treatment in Japan [http://member.jsdt.or.jp/member/contents/cdrom/2013/main.html]. Accessed 17 Apr 2016.
  15. Shinzato T, Nakai S, Miwa M, Iwayama N, Takai I, Matsumoto Y, Morita H, Maeda K. New method to calculate creatinine generation rate using pre- and postdialysis creatinine concentrations. Artif Organs. 1997;21:864–72.View ArticlePubMedGoogle Scholar
  16. Shinzato T, Nakai S, Fujita Y, Takai I, Morita H, Nakane K, Maeda K. Determination of Kt/V and protein catabolic rate using pre- and postdialysis blood urea nitrogen concentrations. Nephron. 1994;67:280–90.View ArticlePubMedGoogle Scholar
  17. Daugirdas JT. Simplified equations for monitoring Kt/V, PCRn, eKt/V, and ePCRn. Adv Ren Replace Ther. 1995;2:295–304.PubMedGoogle Scholar
  18. Clark BC, Manini TM. Sarcopenia =/= dynapenia. J Gerontol A Biol Sci Med Sci. 2008;63:829–34.View ArticlePubMedGoogle Scholar
  19. Kim JC, Kalantar-Zadeh K, Kopple JD. Frailty and protein-energy wasting in elderly patients with end stage kidney disease. J Am Soc Nephrol. 2013;24:337–51.View ArticlePubMedGoogle Scholar
  20. Friedman AN, Fadem SZ. Reassessment of albumin as a nutritional marker in kidney disease. J Am Soc Nephrol. 2010;21:223–30.View ArticlePubMedGoogle Scholar
  21. Lopes AA, Elder SJ, Ginsberg N, Andreucci VE, Cruz JM, Fukuhara S, Mapes DL, Saito A, Pisoni RL, Saran R, Port FK. Lack of appetite in haemodialysis patients—associations with patient characteristics, indicators of nutritional status and outcomes in the international DOPPS. Nephrol Dial Transplant. 2007;22:3538–46.View ArticlePubMedGoogle Scholar
  22. Kalantar-Zadeh K, Gutekunst L, Mehrotra R, Kovesdy CP, Bross R, Shinaberger CS, Noori N, Hirschberg R, Benner D, Nissenson AR, Kopple JD. Understanding sources of dietary phosphorus in the treatment of patients with chronic kidney disease. Clin J Am Soc Nephrol. 2010;5:519–30.View ArticlePubMedGoogle Scholar
  23. Levin NW, Kotanko P, Eckardt KU, Kasiske BL, Chazot C, Cheung AK, Redon J, Wheeler DC, Zoccali C, London GM. Blood pressure in chronic kidney disease stage 5D-report from a kidney disease: improving global outcomes controversies conference. Kidney Int. 2010;77:273–84.View ArticlePubMedGoogle Scholar
  24. Kalantar-Zadeh K, Block G, Humphreys MH, Kopple JD. Reverse epidemiology of cardiovascular risk factors in maintenance dialysis patients. Kidney Int. 2003;63:793–808.View ArticlePubMedGoogle Scholar
  25. Kalantar-Zadeh K, Kopple JD, Kilpatrick RD, McAllister CJ, Shinaberger CS, Gjertson DW, Greenland S. Association of morbid obesity and weight change over time with cardiovascular survival in hemodialysis population. Am J Kidney Dis. 2005;46:489–500.View ArticlePubMedGoogle Scholar
  26. Marcelli D, Brand K, Ponce P, Milkowski A, Marelli C, Ok E, Merello Godino JI, Gurevich K, Jirka T, Rosenberger J, et al. Longitudinal changes in body composition in patients after initiation of hemodialysis therapy: results from an international cohort. J Ren Nutr. 2016;26:72–80.View ArticlePubMedGoogle Scholar
  27. Abramowitz MK, Sharma D, Folkert VW. Hidden obesity in dialysis patients: clinical implications. Semin Dial. 2016;29:391–5.View ArticlePubMedGoogle Scholar
  28. Shapiro BB, Streja E, Ravel VA, Kalantar-Zadeh K, Kopple JD. Association of height with mortality in patients undergoing maintenance hemodialysis. Clin J Am Soc Nephrol. 2015;10:965–74.View ArticlePubMedPubMed CentralGoogle Scholar
  29. Elsayed ME, Ferguson JP, Stack AG. Association of height with elevated mortality risk in ESRD: variation by race and gender. J Am Soc Nephrol. 2016;27:580–93.View ArticlePubMedGoogle Scholar

Copyright

© The Author(s) 2017

Advertisement