Differences in survival on chronic dialysis treatment ... - Oxford Journals

33. Brown AJ, Dusso AS, Slatopolsky E. Vitamin D analogues for secondary hyperparathyroidism. Nephrol Dial Transplant 2002; 17 (Suppl 10): 10–19 34. Lessard M, Ouimet D, Leblanc M et al. Comparison of oral and intravenous alfacalcidol in chronic hemodialysis patients. BMC Nephrol 2014; 15: 27 35. Duque G, El Abdaimi K, Macoritto M et al. Estrogens (E2) regulate expression and response of 1,25-dihydroxyvitamin D3 receptors in bone cells: changes with aging and hormone deprivation. Biochem Biophys Res Commun 2002; 299: 446–454 36. Karakas M, Thorand B, Zierer A et al. Low levels of serum 25hydroxyvitamin D are associated with increased risk of myocardial infarction, especially in women: results from the MONICA/KORA Augsburg case–cohort study. J Clin Endocrinol Metab 2013; 98: 272–280 37. Thorand B, Zierer A, Huth C et al. Effect of serum 25-hydroxyvitamin D on risk for type 2 diabetes may be partially mediated by subclinical inflammation: results from the MONICA/KORA Augsburg study. Diabetes Care 2011; 34: 2320–2322

38. Gombart AF, Bhan I, Borregaard N et al. Low plasma level of cathelicidin antimicrobial peptide (hCAP18) predicts increased infectious disease mortality in patients undergoing hemodialysis. Clin Infect Dis 2009; 48: 418–424 39. Viaene L, Evenepoel P, Meijers B et al. Uremia suppresses immune signal-induced CYP27B1 expression in human monocytes. Am J Nephrol 2012; 36: 497–508 40. Sturmer T, Joshi M, Glynn RJ et al. A review of the application of propensity score methods yielded increasing use, advantages in specific settings, but not substantially different estimates compared with conventional multivariable methods. J Clin Epidemiol 2006; 59: 437–447 41. Chen Y, Briesacher BA. Use of instrumental variable in prescription drug research with observational data: a systematic review. J Clin Epidemiol 2011; 64: 687–700 Received for publication: 20.12.2015; Accepted in revised form: 13.3.2016

ORIGINAL ARTICLE

Nephrol Dial Transplant (2016) 31: 1160–1167 doi: 10.1093/ndt/gfv359 Advance Access publication 22 October 2015

Differences in survival on chronic dialysis treatment between ethnic groups in Denmark: a population-wide, national cohort study Tessa O. van den Beukel1,2, Kristine Hommel3, Anne-Lise Kamper3, James G. Heaf4, Carl E.H. Siegert2, Adriaan Honig5, Kitty J. Jager6, Friedo W. Dekker1 and Marie Norredam7 1

Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, The Netherlands, 2Department of Nephrology, Sint Lucas

Andreas Hospital, Amsterdam, The Netherlands, 3Department of Nephrology, Rigshospitalet, Copenhagen, Denmark, 4Department of Medicine, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark, 5Department of Psychiatry, Sint Lucas Andreas Hospital and VU University Medical Center, Amsterdam, The Netherlands, 6ERA-EDTA Registry, Department of Medical Informatics, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands and 7Danish Research Centre for Migration, Ethnicity and Health, Section of Health Services Research, Department of Public Health, University of Copenhagen, Copenhagen, Denmark

Correspondence and offprint requests to: Tessa O. van den Beukel; E-mail: [email protected]

A B S T R AC T

Results. In total, 8459 patients were native Danes, 344 originated from other Western countries, 79 from North Africa or West Asia, 173 from South or South-East Asia and 54 from sub-Saharan Africa. Native Danes were more likely to die on dialysis compared with the other groups (crude incidence rates for mortality: 234, 166, 96, 110 and 53 per 1000 personyears, respectively). Native Danes had greater hazard ratios (HRs) for mortality compared with the other groups {HRs for mortality adjusted for sociodemographic and clinical characteristics: 1.32 [95% confidence interval (CI) 1.14–1.54]; 2.22 [95% CI 1.51–3.23]; 1.79 [95% CI 1.41–2.27]; 2.00 [95% CI 1.10– 3.57], respectively}. Compared with native Danes, adjusted

Background. In Western countries, black and Asian dialysis patients experience better survival compared with white patients. The aim of this study is to compare the survival of native Danish dialysis patients with that of dialysis patients originating from other countries and to explore the association between the duration of residence in Denmark before the start of dialysis and the mortality on dialysis. Methods. We performed a population-wide national cohort study of incident chronic dialysis patients in Denmark (≥18 years old) who started dialysis between 1995 and 2010. © The Author 2015. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved.

1160

HRs for mortality for Western immigrants living in Denmark for ≤10 years, >10 to ≤20 years and >20 years were 0.44 (95% CI 0.27–0.71), 0.56 (95% CI 0.39–0.82) and 0.86 (95% CI 0.70– 1.04), respectively. For non-Western immigrants, these HRs were 0.42 (95% CI 0.27–0.67), 0.52 (95% CI 0.33–0.80) and 0.48 (95% CI 0.35–0.66), respectively. Conclusions. Incident chronic dialysis patients in Denmark originating from countries other than Denmark have a better survival compared with native Danes. For Western immigrants, this survival benefit declines among those who have lived in Denmark longer. For non-Western immigrants, the survival benefit largely remains over time. Keywords: Denmark, dialysis, ethnicity, survival analysis

INTRODUCTION

Study cohort We used data from the Danish Nephrology Registry (DNR). Details on the DNR have been previously published [15]. In short, the DNR is a clinical database that contains data on all end-stage renal disease patients treated with renal replacement therapy (RRT) in Denmark since 1 January 1990. All Danish dialysis centres extract data from medical records and transfer the information to the DNR. In Denmark, all citizens have a unique personal identification number assigned from birth or immigration by the Danish Civil Registration System. Researchers and official institutions can access and link individual-level data from Danish registries by using these numbers. Consequently, we could link DNR data with data from other Danish registries, including Statistics Denmark [16], the Danish Civil Registration System [17], the Danish National Patient Register [18], the Danish National Prescription Registry [19] and the Danish Register of Causes of Death [20]. For the present study, patients from the DNR were included if they (i) started haemodialysis (HD) or peritoneal dialysis (PD) in Denmark between 1 January 1995 and 1 November 2010 as the first RRT (1995 was chosen because in 1994 many changes were made in the DNR), (ii) continued dialysis treatment for ≥90 days, (iii) were ≥18 years of age at initiation of dialysis and (iv) had data available on country of origin. Country of origin Individuals were divided into groups as immigrants, descendants and native Danes. In line with Statistics Denmark, immigrants were defined as those born outside Denmark whose parents were both born outside Denmark, descendants as those born in Denmark whose parents were both born outside Denmark and native Danes as those having at least one parent born in Denmark [16]. For all patients, country of origin was determined based on data from Statistics Denmark. For immigrants, country of origin was defined as country of birth; for descendants, country of origin was defined as mothers’ country of birth; and for native Danes, country of origin was defined as Denmark. Subsequently, countries of origin were categorized into regions of origin using the United Nations classification system (http://millenniumindicators.un.org/unsd/methods/ m49/m49regin.htm). These regions of origin were used to define ethnic groups. Beside native Danes, five ethnic groups were defined: immigrants and descendants originating from (i) North America, Europe or Oceania (Western countries); (ii) North Africa or West Asia (Arabic counties); (iii) South Asia or South-East Asia; (iv) sub-Saharan Africa and (v) other regions (i.e. central and East Asia, central and South America, Caribbean). Immigrants and descendants originating from ‘other regions’ were excluded from the analyses because of small numbers. Duration of residence Data on the duration of residence were analysed for immigrants. Duration of residence was defined as the number of years living in Denmark before the start of dialysis and was

Ethnic differences in survival on dialysis in Denmark

1161

ORIGINAL ARTICLE

Studies have consistently reported that black and Asian dialysis patients in Western countries experience better survival compared with their white counterparts [1–8]. Several reasons have been proposed for these survival paradoxes. It has been suggested that there is a selection of relatively healthier black patients who ultimately require dialysis, because it has been shown that mortality among blacks with chronic kidney disease (CKD) in the pre-dialysis period is higher than pre-dialysis mortality among whites with CKD [9]. Furthermore, therapy with activated vitamin D might explain survival paradoxes. It has been demonstrated that the survival advantage of black dialysis patients was restricted to those receiving activated vitamin D, and this survival advantage was lost after adjustment for vitamin D dosage [10, 11]. Moreover, it has been suggested that a favourable inflammation and nutritional status, including larger muscle mass, may explain the survival advantage of black dialysis patients [12]. It has been reported that among Israeli chronic dialysis patients, those who are ethnically Arab exhibit greater survival than Jewish Israelis [13]. Furthermore, it has been described that immigrant dialysis patients in The Netherlands have better survival than native Dutch dialysis patients [14]. Apart from this, studies have predominantly focused on differences in survival between white, black and—to a lesser extent—Asian dialysis patients. Consequently, it is currently unknown whether—in Western countries—patients originating from other Western countries have different survival rates compared with native dialysis patients. Furthermore, it is unknown whether the duration of residence in the immigration country before the initiation of dialysis is associated with mortality on dialysis. These data, however, provide further insights into possible explanations for ethnic differences in survival on dialysis. In the present study within the group of patients starting dialysis in Denmark between 1995 and 2010, we compared the survival of native Danish patients with that of patients originating from other countries, including Western countries. Moreover, we explored whether the duration of residence in Denmark before the start of dialysis is associated with the mortality on dialysis.

M AT E R I A L S A N D M E T H O D S

calculated based on data on immigration year obtained from Statistics Denmark (available since 1973).

ORIGINAL ARTICLE

Mortality Patients were followed from Day 90 after the initiation of dialysis therapy (baseline) until death or censoring. Patients were treated as censored when they (i) had a recovery of renal function, (ii) emigrated from Denmark, (iii) underwent a renal transplantation or (iv) reached the end of the study period at 18 February 2011. Dates of dialysis initiation, recovery of renal function and renal transplantation were obtained from the DNR. Data on emigration were ascertained from both the Danish Civil Registration System and the DNR. Date of death was ascertained from the Danish Civil Registration System. In the Danish Civil Registration System, death certificates issued abroad were excluded due to validity problems. Data on cause of death were obtained from the DNR using ERA-EDTA codes [21], and where needed, these were completed with data from the Danish Register of Causes of Death. Covariates Data on age, sex, initial treatment modality and primary kidney disease were ascertained from the DNR. Primary kidney disease was classified according to the codes of the 10th revision of the WHO’s International Classification of Diseases (ICD-10) (http://www.who.int/classifications/icd/en/). Household income, including transfer payments, business profits and pensions (except for private pensions), was obtained from Statistics Denmark and corrected for inflation with 2009 as the index year. Data on hospital admissions were ascertained from the Danish National Patient Registry. This administrative registry contains information on dates and ICD-10 (before 1994 ICD-8) diagnoses of hospital admissions. Hospital admissions due to cardiovascular disease were defined as admissions because of acute myocardial infarction and/or stroke. Hospital admissions for cancer were defined as admissions or outpatient contacts due to any kind of cancer excluding non-melanoma skin cancers. For each patient, the Charlson comorbidity index was calculated based on 19 different diagnoses of hospital admissions within 1 year before the start of dialysis. The Charlson comorbidity index is a validated method of classifying prognostic comorbidity in longitudinal studies [22]. Data on prescribed anti-hypertensive and anti-diabetic medication were obtained from the Danish National Prescription Registry. This registry contains information on drugs prescribed and sold, except for in-hospital drug use. Data on household income, comorbidity and medication were based on a period 1 year prior to the start of dialysis to prevent incorrect classification of immigrants who migrated to Denmark shortly before the start of dialysis. Statistical analysis Differences in patient characteristics between native Danes and other ethnic groups were analysed with t-tests, Wilcoxon signed rank-sum tests and χ² tests where appropriate. Crude and age-standardized incidence rates of death, recovery of renal function, emigration and renal transplantation were calculated for the different ethnic groups. Age-standardized

1162

incidence rates were calculated using direct standardization with the total dialysis population as the standard population (age groups 18–39, 40–49, 50–59, 60–69 and 70+ years). Competing risk analysis was used to calculate cumulative incidence curves for mortality for the different ethnic groups, taking into account renal transplantation as a competing end point. Cox proportional hazards analysis was used to estimate the hazard ratio (HR) for mortality with accompanying 95% confidence interval (CI) for native Danes compared with the other ethnic groups separately [23]. HRs were adjusted for demographic, socio-economic and clinical characteristics [24]. These characteristics were considered as a consequence of ethnicity and therefore being in the causal pathway between ethnicity and survival. Consequently, multivariable adjustment was done deliberately within the causal pathway in order to explore mechanisms. Furthermore, Cox proportional hazards analysis was used to explore whether the time living in Denmark before the start of dialysis was associated with the mortality on dialysis. For this analysis, we divided the immigrant group into three a priori defined subgroups based on the number of years living in Denmark before the start of dialysis (≤10 years, >10 to ≤20 years and >20 years). HRs for mortality with accompanying 95% CIs were calculated for each subgroup, with native Danes as the reference group. This analysis was done both for all immigrant patients together and stratified for Western immigrants (i.e. immigrants originating from North America, Europe or Oceania) and non-Western immigrants (i.e. immigrants originating from North Africa, West Asia, South Asia, South-East Asia or sub-Saharan Africa). HRs were adjusted for demographic, socio-economic and clinical characteristics to study underlying mechanisms. Proportional hazards assumptions were tested graphically with cumulative residuals. Covariates had no missing data, except for data on household income (50 years of age (although in some strata 95% CIs were wider and included native Danes) (Supplementary data, Data supplement 2). Duration of residence Table 4 shows that the survival advantage for patients originating from countries other than Denmark compared with native Danes becomes smaller with increased duration of residence in Denmark. After multivariable adjustments, these

Region of origin

Sociodemographic Age (years) Sex (% men) Household income (% lowest tertile)b Immigrants (%)c Clinical Number of days admitted to hospital in 30 days after the start of dialysis Dialysis modalityd (% haemodialysis) Primary kidney disease (%) Diabetes mellitus Glomerulonephritis Renal vascular disease Other Comorbidityb Number of days admitted to hospital Charlson comorbidity index Hospital admissions due to Cardiovascular disease (%)e Cancer (%) Medicationb Anti-hypertensive (%) Anti-diabetic (%)

Denmark (n = 8459)

North America, Europe,a Oceania (n = 344)

North Africa, West Asia (n = 79)

South Asia, South-East Asia (n = 173)

Sub-Saharan Africa (n = 54)

63.2 64 32 –

(14.7)

57.7* 59 41* 96

(16.1)

52.5* 63 39* 100

(15.1)

53.8* 51* 47* 99

(15.5)

46.2* 69 43 100

(14)

3

(0–12)

3

(0–13)

4

(0–13)

2

(0–8)

2

(0–7)

69

69

81*

70

76

23 10 12 54

27 13 14 46

35* 19* 10 35*

35* 16* 9 39*

13 26* 19 43

7 3

(2–19) (2–5)

6 2

(2–15) (2–5)

5 2

(2–14) (2–5)

9 3

(2–16) (2–5)

3.5* 2*

5 6

4 3*

4 1

4 1*

0 2

88 24

85* 27

85 38*

83* 38*

67* 7*

(1–10) (2–3)

Values are presented as mean (SD) or median (interquartile range) or percentage. a Excluding Denmark. b In year prior to the start of dialysis. c Versus descendants. d Modality at the start of dialysis treatment. e Defined as myocardial infarction and/or stroke. *P ≤ 0.05 [compared with patients originating from Denmark (native Danes)].


1163

ORIGINAL ARTICLE

Table 1. Baseline characteristics of 9109 patients who started chronic dialysis treatment in Denmark between 1995 and 2010, according to patients’ region of origin

Table 2. Median follow-up time and incidence rates of different outcomes of patients who started chronic dialysis treatment in Denmark between 1995 and 2010, according to patients’ region of origin Region of origin Denmark (n = 8459)

Follow-up months, median (IQR) 22.6 (9.2–45.0) Crude incidence rates (per 1000 py) Death 234 Renal transplantation 63 Recovery of renal function 11 Emigration 0 Age-standardized incidence rates, per 1000 py (95% CI) Death 231 (224–237) Renal transplantation 66 (62–69) Recovery of renal function 11 (10–13) Emigration 0

North America, Europe,a Oceania (n = 344)

North Africa, West Asia (n = 79)

South Asia, South-East Asia (n = 173)

Sub-Saharan Africa (n = 54)

26.9

29.4

28.6

26.3

(10.6–50.4)

166 83 3 0 195 61 3 0

(15.3–63.4)

96 85 4 0 (165–225) (48–75) (0–6)

113 60 2 0

(13.7–54.5)

110 91 0 0 (67–159) (35–86) (0–6)

142 65 0 0

(11.1–63.4)

53 117 0 0 (103–182) (47–84)

239 87 0 0

(0–496) (40–135)

ORIGINAL ARTICLE

py, person-years. a Excluding Denmark.

F I G U R E 1 : Cumulative mortality curves during 10 years of follow-

up for incident dialysis patients in Denmark with different regions of origin, taking into account renal transplantation as the competing end point: (1) native Danes; (2) patients originating from North America, Europe (excluding Denmark) or Oceania; (3) patients originating from South Asia or South-East Asia; (4) patients originating from North Africa or West Asia; and (5) patients originating from sub-Saharan Africa.

differences become less pronounced but remain present. Analyses stratified for Western immigrants and non-Western immigrants show that this pattern of associations is only present in Western immigrants and not in non-Western immigrants (Table 4).

DISCUSSION This population-wide national cohort study, involving 9109 incident chronic dialysis patients in Denmark, demonstrated that native Danes exhibited a substantially higher mortality rate compared with patients originating from sub-Saharan Africa;

1164

South or South-East Asia; North Africa or West Asia; and North America, Europe (excluding Denmark) or Oceania. For Western immigrants, the survival advantage declines among those who have lived in Denmark longer. For nonWestern immigrants, however, the survival benefit largely remains over time. This study presents several interesting findings. First, as far as we know it has not been publicized before that dialysis patients originating from other Western countries have better survival rates compared with native patients. Second, to our knowledge, it has not been previously studied whether the survival of immigrant dialysis patients depends on the duration of residence in the immigration country. In the general population, however, it has been shown that the duration of residence is associated with mortality [25–27]. Third, the low renal transplantation rate for native Danes compared with the other ethnic groups is quite contrary to the literature. It has been found that in the USA, Canada and the UK, ethnic minorities have less access to transplantation [7, 28, 29]. After age standardization, ethnic differences in renal transplantation rates were no longer present in our study. Nevertheless, the reasons for these intercountry differences are unclear. Strengths of our study include the use of an unselected national cohort, the universal healthcare system, minimal missing data, the linkage of national registries and the availability of detailed information on country of birth and duration of residence. There are also some possible limitations. First, although country of birth renders objective information on region of origin, it does not capture the multidimensional character of ethnicity. Ethnicity has been defined as the social group a person belongs to as a result of a mix of cultural and other factors, including language, diet, religion, ancestry and physical features [30]. However, there is no measurable entity of ethnicity that captures all these aspects. Moreover, country of birth is the most frequently used proxy of ethnicity in register-based studies [31]. Second, we had limited data on clinical status, in particular on comorbidity, nutritional status and inflammation

T. O. van den Beukel et al.

Table 3. HRs for mortality for patients originating from Denmark compared with patients originating from the other regions, adjusted using more complex multivariable models Modelb

1. Unadjusted 2. Demographic 3. Socio-economic 4. Clinical

Region of origina Denmark versus North America, Europe, Oceania

Denmark versus North Africa, West Asia

Denmark versus South Asia, South-East Asia

Denmark versus sub-Saharan Africa

HR

95% CI

HR

95% CI

HR

95% CI

HR

95% CI

1.41 1.21 1.23 1.32

1.22–1.63 1.04–1.40 1.06–1.43 1.14–1.54

2.45 1.96 2.03 2.22

1.69–3.56 1.35–2.84 1.40–2.95 1.51–3.23

1.97 1.59 1.64 1.79

1.56–2.50 1.25–2.02 1.30–2.09 1.41–2.27

4.35 2.48 3.14 2.00

2.40–7.87 1.37–4.48 1.45–4.74 1.10–3.57

a

Number of patients: Denmark n = 8459; North America, Europe, Oceania n = 344; North Africa, West Asia n = 79; South Asia, South-East Asia n = 173; sub-Saharan Africa n = 54. Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: further adjusted for household income. Model 4: further adjusted for dialysis modality, primary kidney disease, number of days admitted to hospital in 30 days after the start of dialysis, number of days admitted to hospital in year prior to start of dialysis, Charlson comorbidity index, whether a patient was admitted to hospital due to cardiovascular disease and/or cancer in year prior to start of dialysis and use of anti-hypertensive medication and/or anti-diabetic medication.

b

Table 4. Years living in Denmark before starting dialysis and associated risk of mortality for immigrants compared with native Danes who started chronic dialysis treatment in Denmark between 1995 and 2010 Modela

1. Unadjusted

3. Socio-economic

4. Clinical

≤10 years >10 to ≤20 years >20 years Native Danes ≤10 years >10 to ≤20 years >20 years Native Danes ≤10 years >10 to ≤20 years >20 years Native Danes ≤10 years >10 to ≤20 years >20 years Native Danes

All immigrants (n = 633)

Western immigrantsc (n = 329)

Non-Western immigrantsd (n = 304)

HR

95% CI

HR

95% CI

HR

95% CI

0.29 0.49 0.71 Reference 0.46 0.64 0.78 Reference 0.43 0.62 0.77 Reference 0.43 0.54 0.71 Reference

0.21–0.39 0.37–0.65 0.60–0.84


0.18–0.45 0.45–0.95 0.73–1.08


0.24–0.60 0.24–0.58 0.36–0.67

0.34–0.64 0.49–0.85 0.67–0.92 0.31–0.59 0.47–0.82 0.65–0.91 0.31–0.59 0.41–0.72 0.61–0.84

0.31–0.81 0.53–1.11 0.74–1.09 0.29–0.76 0.51–1.07 0.74–1.09 0.27–0.71 0.39–0.82 0.70–1.04

0.31–0.77 0.34–0.82 0.43–0.80 0.28–0.71 0.33–0.80 0.41–0.78 0.27–0.67 0.33–0.80 0.35–0.66

Patient numbers: ≤10 years: all immigrants, n = 158; Western immigrants, n = 67; non-Western immigrants, n = 91; >10 to ≤20 years: all immigrants, n = 154; Western immigrants, n = 70; non-Western immigrants, n = 84; >20 years: all immigrants, n = 321; Western immigrants, n = 192; non-Western immigrants, n = 129; Native Danes, n = 8459. a Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: further adjusted for household income. Model 4: further adjusted for dialysis modality, primary kidney disease, number of days admitted to hospital in 30 days after the start of dialysis, number of days admitted to hospital in year prior to start dialysis, Charlson comorbidity index, whether a patient was admitted to hospital due to cardiovascular disease and/or cancer in the year prior to the start of dialysis and use of anti-hypertensive medication and/or anti-diabetic medication. b Before starting dialysis. c Immigrants from North America, Europe and Oceania. d Immigrants from North Africa, West Asia, South Asia, South-East Asia and sub-Saharan Africa.

status. We partly addressed this limitation by using diagnoses of hospital admissions and outpatient contacts in the year prior to the start of dialysis. In doing so, we may have slightly underestimated comorbidity because patients may have had comorbid diseases without contacting a hospital in the year before starting dialysis. Furthermore, in the DNR, data on the number of patients who died or came off dialysis in the first 90 days were incomplete. Therefore we only included patients who continued dialysis treatment for ≥90 days. This could be a source of selection bias. However, a nationwide study from the UK demonstrated that both the survival from the start of RRT to

Day 90 and the survival after Day 90 were significantly better for ethnic minority groups compared with Caucasians [7]. This implies that if any selection bias exists, this could not fully explain the better survival for ethnic minority groups. Finally, although registered remigrations have been documented among immigrants and descendants, unregistered remigration followed by death in the home country was not documented. This may have led to an underestimation of the number of deaths in immigrants and descendants. Several explanations for the better survival of immigrants and descendants could be proposed. First, selection effects may play a


1165

ORIGINAL ARTICLE

2. Demographic

Years living in Denmarkb

ORIGINAL ARTICLE

role. Selection can play a role in the trajectory prior to the start of dialysis. Mehrotra et al. [9] found that among patients with early stages of CKD, black individuals have a higher risk of mortality compared with whites. This may result in the selection of relatively healthy black patients who ultimately need dialysis treatment. In addition, selection can play a role in the migration process. In general, only relatively healthy patients are able to migrate. This may result in the selection of healthy individuals into migration in the home country [32]. Our finding that Western immigrants have a better survival compared with native Danes, while both groups are likely to be white, may support this hypothesis. Furthermore, our finding that the survival advantage for Western immigrants declines among those who have lived in Denmark longer might imply that in Western immigrants selection effects disappear over time. The finding that for non-Western immigrants the survival benefit largely remains over time may imply that in non-Western immigrants other mechanisms must be present as well. Second, differences in lifestyles between natives and immigrants could explain ethnic differences in survival on dialysis. A Danish survey on health behaviour shows that immigrants in Denmark experience healthier lifestyles than native Danes [33]. Unfortunately, in the present study we did not have data on individual lifestyle factors, such as dietary habits and smoking status. Third, genetic variation may account for ethnic differences in survival on dialysis. It could be that particular genetic profiles induce additional mortality risks once these persons are treated with dialysis therapy. In the literature, a number of other hypotheses have been presented. Therapy with activated vitamin D might explain ethnic differences in survival on dialysis [10, 11]. Moreover, it has been suggested that a favourable inflammation and nutritional status, including larger muscle mass, may explain the survival advantage of black dialysis patients [12]. In conclusion, we demonstrated ethnic differences in survival on dialysis in Denmark that were beneficial to immigrants and their descendants. Challenging new findings were that patients originating from other Western counties have a better survival compared with native Danes and that this survival advantage for Western immigrants declines among those who have lived in Denmark longer. For non-Western immigrants, however, the survival benefit largely remains over time. Future studies should focus on the relative importance of selection effects, lifestyle differences and genetic factors as explanations for ethnic differences in survival on dialysis.

S U P P L E M E N TA R Y D ATA Supplementary data are available online at http://ndt.oxfordjournals.org.

AC K N O W L E D G M E N T S T.O.B. is grateful to the European Renal Association–European Dialysis and Transplant Association for support with a research fellowship (53-2009).

1166

C O N F L I C T O F I N T E R E S T S TAT E M E N T The results presented in this article have not been published previously in whole or part, except in abstract format. (See related article by Rhee et al. Why minorities live longer on dialysis: an in-depth examination of the Danish nephrology registry. Nephrol Dial Transplant 2016; 31: 1027–1030)

REFERENCES 1. Bleyer AJ, Tell GS, Evans GW et al. Survival of patients undergoing renal replacement therapy in one center with special emphasis on racial differences. Am J Kidney Dis 1996; 28: 72–81 2. Bloembergen WE, Port FK, Mauger EA et al. Causes of death in dialysis patients: racial and gender differences. J Am Soc Nephrol 1994; 5: 1231–1242 3. Kucirka LM, Grams ME, Lessler J et al. Association of race and age with survival among patients undergoing dialysis. JAMA 2011; 306: 620–626 4. Medina RA, Pugh JA, Monterrosa A et al. Minority advantage in diabetic end-stage renal disease survival on hemodialysis: due to different proportions of diabetic type? Am J Kidney Dis 1996; 28: 226–234 5. Pei YP, Greenwood CM, Chery AL et al. Racial differences in survival of patients on dialysis. Kidney Int 2000; 58: 1293–1299 6. Robinson BM, Joffe MM, Pisoni RL et al. Revisiting survival differences by race and ethnicity among hemodialysis patients: the Dialysis Outcomes and Practice Patterns Study. J Am Soc Nephrol 2006; 17: 2910–2918 7. Roderick P, Byrne C, Casula A et al. Survival of patients from South Asian and Black populations starting renal replacement therapy in England and Wales. Nephrol Dial Transplant 2009; 24: 3774–3782 8. Tanna MM, Vonesh EF, Korbet SM. Patient survival among incident peritoneal dialysis and hemodialysis patients in an urban setting. Am J Kidney Dis 2000; 36: 1175–1182 9. Mehrotra R, Kermah D, Fried L et al. Racial differences in mortality among those with CKD. J Am Soc Nephrol 2008; 19: 1403–1410 10. Kalantar-Zadeh K, Miller JE, Kovesdy CP et al. Impact of race on hyperparathyroidism, mineral disarrays, administered vitamin D mimetic, and survival in hemodialysis patients. J Bone Miner Res 2010; 25: 2448–2458 11. Wolf M, Betancourt J, Chang Y et al. Impact of activated vitamin D and race on survival among hemodialysis patients. J Am Soc Nephrol 2008; 19: 1379–1388 12. Streja E, Kovesdy CP, Molnar MZ et al. Role of nutritional status and inflammation in higher survival of African American and Hispanic hemodialysis patients. Am J Kidney Dis 2011; 57: 883–893 13. Kalantar-Zadeh K, Golan E, Shohat T et al. Survival disparities within American and Israeli dialysis populations: learning from similarities and distinctions across race and ethnicity. Semin Dial 2010; 23: 586–594 14. van den Beukel TO, Dekker FW, Siegert CE. Increased survival of immigrant compared to native dialysis patients in an urban setting in the Netherlands. Nephrol Dial Transplant 2008; 23: 3571–3577 15. Hommel K, Rasmussen S, Madsen M et al. The Danish Registry on Regular Dialysis and Transplantation: completeness and validity of incident patient registration. Nephrol Dial Transplant 2010; 25: 947–951 16. Danmarks Statistik. http://www.statistikbanken.dk (February 2012, date last accessed) 17. Pedersen CB, Gotzsche H, Moller JO et al. The Danish Civil Registration System. A cohort of eight million persons. Dan Med Bull 2006; 53: 441–449 18. Mosbech J, Jorgensen J, Madsen M et al. The national patient registry. Evaluation of data quality. Ugeskr Laeger 1995; 157: 3741–3745 19. Kildemoes HW, Sorensen HT, Hallas J. The Danish National Prescription Registry. Scand J Public Health 2011; 39: 38–41 20. Helweg-Larsen K. The Danish Register of Causes of Death. Scand J Public Health 2011; 39: 26–29

T. O. van den Beukel et al.

21. Appendix 2—Grouping of causes of death. In: ERA-EDTA Registry 2008 Annual Report. Amsterdam, The Netherlands: Academic Medical Center, Department of Medical Informatics, 2010, p. 129 22. Charlson ME, Pompei P, Ales KL et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40: 373–383 23. Noordzij M, Leffondre K, van Stralen KJ et al. When do we need competing risks methods for survival analysis in nephrology? Nephrol Dial Transplant 2013; 28: 2670–2677 24. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996; 15: 361–387 25. Harding S. Mortality of migrants from the Indian subcontinent to England and Wales: effect of duration of residence. Epidemiology 2003; 14: 287–292 26. Kliewer EV, Smith KR. Ovarian cancer mortality among immigrants in Australia and Canada. Cancer Epidemiol Biomarkers Prev 1995; 4: 453–458 27. Kliewer EV, Smith KR. Breast cancer mortality among immigrants in Australia and Canada. J Natl Cancer Inst 1995; 87: 1154–1161

28. Malek SK, Keys BJ, Kumar S et al. Racial and ethnic disparities in kidney transplantation. Transpl Int 2011; 24: 419–424 29. Yeates KE, Schaubel DE, Cass A et al. Access to renal transplantation for minority patients with ESRD in Canada. Am J Kidney Dis 2004; 44: 1083–1089 30. Bhopal R. Glossary of terms relating to ethnicity and race: for reflection and debate. J Epidemiol Community Health 2004; 58: 441–445 31. Norredam M, Kastrup M, Helweg-Larsen K. Register-based studies on migration, ethnicity, and health. Scand J Public Health 2011; 39: 201–205 32. Razum O, Twardella D. Time travel with Oliver Twist—towards an explanation for a paradoxically low mortality among recent immigrants. Trop Med Int Health 2002; 7: 4–10 33. Singhammer J. Ethnic Minority Health—A Survey Based Study. Århus, Denmark: Region MidJutland, 2008

Received for publication: 4.3.2015; Accepted in revised form: 15.9.2015

ORIGINAL ARTICLE


1167