Cardiac Mortality in Kidney Transplant Patients: A Population-based Cohort Study 1988–2013 in Australia and New Zealand

Abstract

Background. 

Transplant recipients experience excess cardiac mortality. We compared circulatory death rates in Australian and New Zealand kidney transplant recipients to the general population and identified risk factors for circulatory death in kidney transplant recipients.

Methods. 

The primary cause of death for kidney transplant recipients aged ≥18 was established through ICD-10-AM codes using data linkage between the Australia and New Zealand dialysis and transplant registry and national death registers. We estimated standardized mortality ratios (SMRs) and developed a Fine–Gray competing risks model to determine risk factors for cardiac mortality.

Results. 

Of 5089 deaths in 16 329 kidney transplant recipients (158 325 person-years), 918 (18%) were cardiac. An increased risk of circulatory death was associated with older age (P < 0.001), male sex (P < 0.001), longer dialysis duration (P = 0.004), earlier era of transplantation (P < 0.001), ever graft failure (P < 0.001), known coronary artery disease (P = 0.002), and kidney failure from diabetes or hypertension (P < 0.001). The cardiac SMR was 5.4 [95% confidence interval (CI): 5.0-5.8], falling from 8.0 (95% CI: 4.9-13.1) in 1988 to 5.3 (95% CI: 4.0-7.0) in 2013 (P < 0.001). Females, particularly young ones, had significantly higher relative cardiac mortality than men. In recipients aged 40 years, the cardiac SMR was 26.5 (95% CI: 15.0-46.6) in females and 7.5 (95% CI: 5.0-11.1) for males.

Conclusions. 

Cardiac risks remain elevated in kidney transplant recipients and may be under-recognized, and prevention and treatment interventions less accessed, less effective or even harmful in female recipients.

INTRODUCTION

Patients with end-stage kidney disease (ESKD) have high rates of cardiac mortality.^1,2 It is well recognized that dialysis patients have cardiac mortality rates 10–20 times higher than the general population.^1,3,4 Although this improves posttransplant, patients carry the burden of their time spent living with chronic kidney disease (CKD) and on dialysis, and cardiac mortality rates remain high.^5-8 This is thought to be due to a combination of both traditional risk factors such as hypertension, coronary artery disease, and diabetes and transplant specific risk factors such as the immunologic effects of previous acute rejection episodes and the chronic effects of calcineurin inhibitors and steroids.⁹ Posttransplant cardiac mortality has not been well examined in contemporary cohorts.

Age-adjusted cardiac mortality has been falling in the general population, in part due to primary and secondary prevention and treatment interventions that are well-supported by randomized controlled trials. Such interventions include statins, thrombolysis, and antihypertensives.^10,11 Kidney transplant recipients, like all ESKD patients, are usually excluded from such trials. It is unknown if transplant recipients receive the same benefits, or suffer more harms, than the general population from these interventions.¹² As such, it remains uncertain whether kidney transplant recipients have benefited from the overall reduction in cardiac mortality seen in the general population.

In the general population, there is emerging evidence that women are less likely than men to receive the best practice management of their cardiac disease, and that cardiac outcomes are poorer in women than in men.^13-15 It is not known if this sex disparity exists, or exists to the same extent, within the kidney transplant population.

The aims of our study were to: (1) identify risk factors for circulatory death in kidney transplant recipients; (2) compare cardiac mortality in kidney transplant recipients and the general population; and (3) compare cardiac mortality in women and men post kidney transplant.

Materials and Methods

Study Design and Setting

This was a population-based cohort study of all incident adult kidney transplant recipients (aged ≥18 years at time of transplant) in Australia and New Zealand who were transplanted between January 1, 1980 and December 31, 2013 (Australia) or January 1, 1988 and December 31, 2012 (New Zealand) (the period for which we have complete cause of death data in the general population). This study used data from the Australian and New Zealand Dialysis and Transplant Registry (ANZDATA) and the National Death Registers of each country.

Participants, Data Linkage, and Death Outcomes

ANZDATA collects prospective demographic, kidney-related treatment, and outcome data annually in all dialysis and transplant patients from all kidney units in Australia and New Zealand. ANZDATA data collection methods have previously been described.¹⁶

Kidney transplant recipients in ANZDATA were linked with national death registers to determine the date and primary cause of death, if any.

Probabilistic record linkage (matching on date of birth, sex, and name) was used in Australia to link recipients to the National Death Index, a database that has recorded all deaths since 1980 by mandate, drawing from publicly available data from the Australian Bureu of Statistics. Deterministic record linkage (matching on the national health index number) was used in New Zealand to link recipients to the mortality collection database which has recorded all deaths since 1988, drawing from publicly available Ministry of Health Data. Using best-practice privacy-preserving protocols, the Australian Institute of Health and Welfare (AIHW) performed the data linkage for Australian recipients and the New Zealand Ministry of Health performed the data linkage for the New Zealand recipients. Ethics approval was granted for this study from the University of Sydney and the AIHW.

The cause of death in kidney transplant recipients was determined by the ICD-10-AM diagnosis codes for the primary cause of death. We considered all-cause circulatory deaths as: angina (I20), myocardial infarction (I21–I23), ischemic heart disease (I24–I25), valve disorders (I34–I37), cardiomyopathy (I42–I43), conduction disorders (I44–I55), arrhythmias (I46–I49), and heart failure (I50). All other causes of death were classified as noncardiac deaths.

The time at risk was measured from transplant date until the date of transplant recipient death or December 31, 2013 (Australian recipients only) or December 31, 2012 (New Zealand recipients only), whichever was first. As the probabilistic record linkage may lead to a small proportion of incorrect links,¹⁷ Australian patients were censored at the ANZDATA date of death if they were considered to have died by ANZDATA and no death had been captured by the relevant national death register. Australian patients were considered to be alive and censored if ANZDATA recorded clinic visits after the National Death Index registered date of death.

Statistical Analyses

Kidney transplant cardiac mortality rates were initially estimated by dividing the number of circulatory deaths by follow-up time. This was then stratified by age, sex, and calendar year. To assess the agreement of fact of death between ANZDATA and the national death registries, we used the Kappa statistic.¹⁸

To identify risk factors for circulatory death within the transplant population, we developed a competing risks (Fine–Gray) model where circulatory death and noncardiac death were considered competing risks.¹⁹ Covariates in the model were selected a priori including: age at transplant, sex, year of transplant, country, racial background, previous dialysis duration, comorbidities, smoking status, and cause of ESKD.

To compare cardiac mortality rates with the general population, standardized mortality ratios (SMRs) were estimated via indirect standardization, with matching on 5-year age groups, sex, year, and country. A test for trend for the SMRs based on the Poisson regression was performed for age groups in each sex and calendar year.

Height, weight, and smoking status were not regularly collected until after 1995. Missing values for these covariates were imputed using chained equations with 5 iterations. Multinominal logistic regression was used to impute smoking status into the following categories: never, current. and former smokers.

Data were analyzed using Stata version 15 (Stata Corporation, College Station, TX).

RESULTS

Patient Characteristics

Our study included 16 329 kidney transplant recipients who were transplanted between 1980 and 2013 in Australia and 1988 and 2013 in New Zealand and observed for a total of 158 325 person-years of follow-up. During this period, 5089 deaths occurred in kidney transplant recipients of which 918 (18.0%) had a cardiac cause. The majority of circulatory deaths were from ischemic heart disease (n = 780, 85%). Of those with a circulatory death, 45% had experienced graft failure compared with 43% of those with a noncardiac death (P = 0.42). There was extremely high agreement between ANZDATA and the death registries about whether death had occurred in both Australia [Kappa statistic: 0.89 (95% confidence interval [CI]: 0.88-0.90)] and New Zealand [Kappa statistic: 0.93 (95% CI: 0.91-0.95)].

The median age at transplant was 52 years (IQR 43–59 y) in those who had a circulatory death, compared to 50 years (41–58 y) for those with a noncardiac death. Females comprised 32% of cardiac and 42% of noncardiac deaths. Those who were overweight or obese comprised 58% of cardiac and 49% of noncardiac deaths. Current or former smokers made up 55% and 44% of those with a cardiac or noncardiac death, respectively. The cause of ESKD was either diabetes or hypertension in 21% and19% in those with a cardiac or noncardiac death, respectively. Median dialysis time before transplant was similar between groups at ~19 months (Table 1).

TABLE 1. - Characteristics of kidney transplant recipients

Characteristics	Circulatory deaths	Other deaths	Alive	Total
	n (%)	n (%)	n (%)	n (%)
Total (%)	918 (6)	4171 (26)	11 240 (69)	16 329 (100)
Age at transplant (y)
≤29	47 (5)	344 (8)	1760 (16)	2151
30–49	322 (35)	1616 (39)	5130 (46)	7068
50–64	500 (54)	1867 (45)	3672 (33)	6039
65–74	49 (5)	335 (8)	663 (6)	1047
≥75	0 (0)	9 (<1)	15 (<1)	24
Median [IQR]	52 [43, 59]	50 [41, 58]	45 [34, 55]	47 [36, 56]
Sex
Female	292 (32)	1735 (42)	4333 (39)	6360 (39)
Male	626 (68)	2436 (58)	6907 (61)	9969 (61)
BMI category (kg/m2)
Underweight (<18.5)	18 (4)	90 (4)	405 (4)	513 (4)
Normal (≥18.5 to <25.0)	172 (38)	1179 (47)	4767 (46)	6118 (46)
Overweight (≥25.0 to<30.0)	175 (38)	801 (32)	3453 (33)	4429 (33)
Obese (≥30.0)	91 (20)	417 (17)	1853 (18)	2361 (18)
Not available	462 (–)	1684 (–)	762 (–)	2908 (–)
Previous dialysis time (mo)
0 to <6	111 (12)	574 (14)	2535 (23)	3220 (20)
≥6 to <12	189 (21)	725 (17)	1536 (14)	2450 (15)
≥12 to <18	136 (15)	619 (15)	1319 (12)	2074 (13)
≥18 to <36	255 (28)	1172 (28)	2544 (23)	3971 (24)
≥36	227 (25)	1081 (26)	3306 (29)	4614 (28)
Median [IQR]	19 [10, 36]	20 [10, 37]	19 [7, 41]	19 [8, 40]
Year of transplant
≤1990	468 (51)	1665 (40)	931 (8)	3064 (19)
1991–1999	305 (33)	1553 (37)	2143 (19)	4001 (25)
2000–2009	130 (14)	859 (21)	5143 (46)	6132 (38)
2010–2013	15 (2)	94 (2)	3023 (27)	3132 (19)
Country
Australia	821 (89)	3735 (90)	9799 (87)	14 355 (88)
New Zealand	97 (11)	436 (10)	1441 (13)	1974 (12)
Racial background
White	792 (86)	3611 (87)	9261 (82)	13 664 (84)
Indigenous	83 (9)	345 (8)	699 (6)	1127 (7)
Asian	40 (4)	195 (5)	1133 (10)	1368 (8)
Other	3 (<1)	20 (<1)	147 (1)	170 (1)
Comorbidities at ESKD
Cerebrovascular disease	37 (4)	151 (4)	287 (3)	475 (3)
Diabetes	149 (16)	741 (18)	1585 (14)	2475 (15)
Coronary artery disease	129 (14)	435 (10)	908 (8)	1472 (9)
Peripheral artery disease	68 (7)	280 (7)	512 (5)	860 (5)
Previous malignancy	371 (40)	2149 (52)	3395 (30)	5915 (36)
Smoking status
Current/Former	268 (29)	1359 (33)	4424 (39)	6051 (37)
Never/Unknown	650 (71)	2812 (67)	6816 (61)	10 278 (63)
Cause of ESKD
Diabetes	131 (14)	597 (14)	1238 (11)	1966 (12)
Hypertension/renal artery disease	62 (7)	191 (5)	487 (4)	740 (5)
Glomerulonephritis/IgA	362 (39)	1691 (41)	5343 (48)	7396 (45)
Polycystic kidney disease	108 (12)	526 (13)	1556 (14)	2190 (13)
Othera	255 (28)	1166 (28)	2616 (23)	4037 (25)
Graft failure (ever)
No	509 (55)	2374 (57)	9196 (82)	12 812 (73)
Yes	409 (45)	1797 (43)	2044 (18)	4816 (27)

^aThe largest contributors to this are reflux nephropathy/PUV (n = 1508), analgesic nephropathy (n = 522), and interstitial nephritis (n = 210).

ESKD, end-stage kidney disease; PUV, posterior urethral valves.

Risk Factors for Cardiac Mortality

In our multivariate model, an increased risk of circulatory death was associated with older age at transplant, male sex, longer duration of previous dialysis, earlier era of transplant, graft failure, no previous history of malignancy, a history of coronary artery disease, and diabetes or hypertension as the cause of ESKD (univariate model Appendix 1, multivariate model Table 2). Noncardiac mortality was also associated with each of these risk factors except for history of coronary artery disease and hypertension as the cause of ESKD. Additional factors associated with noncardiac mortality include a history of peripheral artery disease and being a current or former smoker (Table 2).

TABLE 2. - Summary of the competing risks multivariable model, with circulatory death (n = 918) and noncirculatory death (n = 4171) deaths as the events of interest

	Circulatory death			Noncirculatory death
	Subhazard ratio	(95% CI)	P	Subhazard ratio	(95% CI)	P
Age at transplant (y)			<0.001			<0.001
≤29	0.42	(0.31, 0.58)		0.58	(0.51, 0.65)
30–44	Ref			Ref
45–54	2.08	(1.73, 2.50)	–	1.68	(1.54, 1.84)	–
≥55	3.07	(2.55, 3.69)	–	2.81	(2.56, 3.08)	–
Sex			<0.001			0.04
Female	Ref			Ref
Male	1.54	(1.33, 1.879)		0.93	(0.87, 1.00)
Body mass index (kg/m2)			0.05			0.53
Underweight (≤18.4)	1.37	(0.95, 1.96)		1.00	(0.82, 1.22)
Normal (18.5–24.9)	Ref			Ref
Overweight (25.0–29.9)	1.21	(1.00, 1.47)	–	1.00	(0.90, 1.10)	–
Obese (≥30.0)	1.37	(1.02, 1.85)	–	1.09	(0.97, 1.21)	–
Previous dialysis time (mo)			0.004			<0.001
0 to <6	Ref			Ref
≥6 to <12	1.43	(1.12, 1.81)	–	1.14	(1.02, 1.27)	–
≥12 to <18	1.29	(1.00, 1.66)	–	1.27	(1.13, 1.43)	–
≥18 to <36	1.38	(1.10, 1.74)	–	1.37	(1.24, 1.53)	–
≥36	1.58	(1.24, 2.00)	–	1.50	(1.34, 1.67)	–
Year of transplant			<0.001			<0.001
≤1995	Ref			Ref
1996–2005	0.37	(0.31, 0.44)	–	0.65	(0.60, 0.70)	–
2006–2013	0.16	(0.12, 0.22)	–	0.34	(0.30, 0.39)	–
Graft failure (ever)			<0.001			<0.001
No	Ref			Ref
Yes	3.13	(2.72, 3.60)		3.37	(3.14, 3.61)
Racial background			0.35
Nonwhite	Ref			Ref
White	1.11	(0.90, 1.36)		1.05	(0.94, 1.16)	0.38
Country			0.34			0.11
Australia	Ref			Ref
New Zealand	1.12	(0.89, 1.40)		1.09	(0.98, 1.21)
Comorbidities
Cerebrovascular disease	1.05	(0.71, 1.56)	0.79	1.21	1.00, 1.46)	0.05
Coronary artery disease	1.41	(1.14, 1.76)	0.002	1.02	(0.91, 1.15)	0.76
Peripheral artery disease	1.07	(0.81, 1.43)	0.63	1.17	(1.01, 1.35)	0.04
Previous malignancy	0.64	(0.55, 0.74)	<0.001	1.10	(1.02, 1.18)	0.01
Smoking status			0.18			<0.001
Never	Ref			Ref
Former	1.15	(0.97, 1.36)	–	1.14	(1.05, 1.24)	–
Current	1.26	(0.94, 1.70)		1.39	(1.26, 1.52)	–
Cause of ESKD			<0.001			<0.001
Diabetes	1.73	(1.38, 2.16)	–	1.94	(1.74, 2.16)	–
Hypertension/renal artery disease	1.43	(1.06, 1.93)	–	1.06	(0.90, 1.25)	–
Glomerulonephritis/IgAN	Ref			Ref
Polycystic kidney disease	0.95	(0.76, 1.18)	–	0.94	(0.85, 1.04)	–
Other	1.21	(1.02, 1.42)	–	1.09	(1.01, 1.19)	–

Global P value were Wald tests for heterogeneity. Bold = P < 0.05.

CI, confidence interval; ESKD, end-stage kidney disease.

Kidney transplant recipients who were older at age of transplant had increased risk of circulatory death (P < 0.001). Those aged ≥55 years at time of transplant were 3 times more likely to have a circulatory death compared to those aged 30–44 years at transplant [subhazard ratio (SHR): 3.07, 95% CI: 2.55-3.69]. Men were 54% more likely to have a circulatory death than women (SHR: 1.55, 95% CI: 1.33-1.79). Recipients who were obese were 37% more likely to die from a cardiac cause compared to those in the normal weight range (SHR: 1.37, 95% CI: 1.02-1.85). The previous dialysis time of >6 months was associated with an increased risk of circulatory death. Those transplanted in the most recent era of 2006–2013 were less likely to have a circulatory death than those transplanted before 1995 (SHR: 0.16, 95% CI: 0.12-0.22). Recipients who experienced graft failure were 3 times more likely to have a circulatory death than those who did not (SHR: 3.13, 95% CI: 2.72-3.60). A prior history of coronary artery disease was associated with a 40% increased risk of circulatory death (SHR: 1.41, 95% CI: 1.14-1.76), while a prior history of malignancy was associated with a reduced risk of circulatory death (SHR: 0.64, 95% CI: 0.55-0.74). The cause of ESKD was also a risk factor for a circulatory death, with those having diabetic (SHR: 1.73, 95% CI: 1.38-2.16) or hypertensive nephropathy (SHR: 1.43, 95% CI: 1.06-1.93) at the highest risk (P < 0.001).

Mortality Rates

The overall crude cardiac mortality rate for kidney transplant recipients was 581.1 (95% CI: 544.7-619.9)/100 000 person-years. Annual crude cardiac mortality rates fell from 915.0 (95% CI: 560.6-1493.6)/100 000 person-years in 1988 to 511.2 (95% CI: 386.4-676.4)/100 000 person-years in 2013 (Figure 1A).

FIGURE 1.: Cardiac mortality rates. A, Absolute cardiac mortality rates per 100 000 person-years, by year. B, Absolute cardiac mortality rates per 100 000 person-years, by age and sex.

There was a trend for absolute cardiac mortality rates to increase with older age in both women and men and to be lower for women than men in each age group (Figure 1B). In recipients aged 40 years, crude cardiac mortality rates were 168.4 (95% CI: 95.6-296.5)/100 000 person-years in women and 214.1 (95% CI: 143.5-319.5)/100 000 person-years in men. These rates increased to 1505.2 (95% CI: 935.8-2421.5)/100 000 person-years in women and 2301.1 (95% CI: 1618.3-3272.0)/100 000 person-years in men by age 75 years.

Cumulative Incidence

The cumulative incidence of cardiac mortality was 0.4% (95% CI: 0.3%-0.5%) at 3 months posttransplant; 0.7% (95% CI: 0.5%-0.8%) at 1 year posttransplant; 0.9% (95% CI: 0.7%-1.0%) at 2 years posttransplant; 2.0% (95% CI: 1.7%-2.2%) at 5 years; and 4.5% (95% CI: 4.1%-4.8%) at 10 years posttransplant.

In those with a history of coronary artery disease, at 3 months posttransplant 1.0% (95% CI: 0.6%-1.7%) had died, increasing to 1.3% (95% CI: 0.8%-2.0%) at 1 year posttransplant, 1.7% (95% CI: 1.1%-2.5%) at 2 years posttransplant, 4.7% (95% CI: 3.6%-6.0%) at 5 years posttransplant, and 9.5% (95% CI: 7.7%-11.5%) at 10 years posttransplant. The cumulative incidence of cardiac mortality was lower in those without a history of coronary artery disease, but also increased over time with cumulative incidence rates of 0.3% (95% CI: 0.2%-0.4%), 0.6% (95% CI: 0.4%-0.7%), 0.7% (95% CI: 0.6%-0.9%), 1.6% (95% CI: 1.4%-1.8%), and 3.8% (95% CI: 3.4%-4.1%) at 3 months, 1, 2, 5, and 10 years posttransplant, respectively (Figure 2).

FIGURE 2.: The estimates of cumulative incidence in transplant recipients by pre-existing coronary artery disease for: (A) circulatory deaths and (B) noncirculatory deaths.

Standardized Mortality Rates

The overall SMR for circulatory death was 5.4 (95% CI: 5.0-5.7). Over the follow-up period, the annual SMR fell from 8.0 (95% CI: 4.9-13.1) in 1988 to 5.3 (95% CI: 4.0-7.0) in 2013 (test for trend P < 0.001) (Figure 3A).

FIGURE 3.: Relative (standardized) cardiac mortality rates. A, Relative cardiac mortality by year (standardized mortality ratios). B, Relative cardiac mortality rates by year and sex (standardized mortality ratios). C, Relative cardiac mortality rates by age and sex (standardized mortality ratios).

The cardiac SMR fell with age in both women and men over time (test for trend P < 0.001) (Figure 3B). Unlike absolute mortality, women had a higher relative cardiac mortality compared with the general population across all age groups (Figure 3C). In recipients aged 40 years, the cardiac SMR was 26.5 (95% CI: 15.0-46.6) in women and 7.5 (95% CI: 5.0-11.1) for men. This fell to 3.3 (95% CI: 2.1-5.4) for women and 2.7 (95% CI: 1.9-3.8) for men when aged 75 years (Table 3).

TABLE 3. - Standardized mortality ratio estimates for all-cause circulatory death by age group and sex

Age at circulatory death (y)	Female				Male
	Observed	Expected	SMR	95% CI	Observed	Expected	SMR	95% CI
30–49	44	1.7	25.4	19.0-34.0	11	11.6	9.5	7.8-11.4
50–64	134	10.9	12.3	10.4-14.5	311	54.9	5.7	5.1-6.3
65–74	92	14.9	6.2	5.0-7.6	166	46.4	3.6	3.1-4.2
≥75	20	10.6	1.9	1.2-2.9	38	19.7	1.9	1.4-2.7

CI, confidence interval; SMR, standardized mortality ratio.

DISCUSSION

In our comprehensive binational cohort study of kidney transplant recipients, we have shown that kidney transplant recipients have >5 times the burden of circulatory deaths compared with the general population, and that female recipients bear the greatest excess burden of circulatory death. While it is well recognized that transplant recipients have higher cardiac mortality rates than the general population, our work provides a contemporary analysis spanning 33 years with national-linked data sets which has enabled us to update and extend what is known about cardiac mortality following kidney transplant. We have 3 key findings. First, circulatory death rates in transplant recipients, while remaining high compared with the general population, have fallen significantly over time. Notably, these improvements have outpaced the well-recognized improvement in cardiac mortality in the general population.²⁰ Second, while men have higher absolute rates of circulatory death, relative to the general population, women disproportionately bear the burden of excess cardiac mortality following transplantation. Third, despite the widespread use of cardiac screening both before transplant listing and while on the waitlist, those with preexisting coronary artery disease have significantly higher rates of circulatory death than those without coronary artery disease before transplant.

The importance of traditional cardiac risk factors in the kidney transplant population was confirmed by our multivariate model, with older age, male sex, a history of coronary artery disease, and diabetes or hypertension as the cause of ESKD being significantly associated with circulatory death. Transplant factors including longer dialysis vintage before transplantation, earlier era of transplantation, and graft failure were also important risk factors. Longer duration of dialysis has previously been found to be an important risk factor for circulatory death following transplantation. Meier-Kriesche et al⁸ used data from the US Renal Data System to analyze 58 900 recipients transplanted between 1988 and 1998 and found that those with longer dialysis durations pretransplant were at an increased risk of cardiovascular death with a relative risk of 1.3 (95% CI: 1.1-1.6) for those on dialysis for 6–12 months before transplant compared to preemptive transplant, increasing in a dose-response fashion to a relative risk of 1.5 for those on dialysis for >24 months.²¹ It is likely that the high rates of circulatory death seen in transplant recipients are due not just to traditional risk factors but also transplant-related factors including the effects of immunosuppression and graft function.⁹

Transplant recipients are at the greatest risk for circulatory death in the first 3 months posttransplant. In our study, the 3-month cumulative incidence of circulatory death was 0.4%, half of the total 2-year cumulative incidence of 0.8%. In an effort to reduce posttransplant circulatory death, contemporary transplant programs have extensive cardiac risk screening programs in place both before listing on the deceased donor transplant waitlist as well as annually or biannually while on the waitlist. Despite such screening programs, transplant recipients with preexisting coronary artery disease had higher rates of circulatory death compared with those without coronary artery disease. The excess mortality for those with a history of coronary artery disease was highest in the early posttransplant period with a 3-month cumulative incidence of circulatory death >3 times higher in those with, compared to without, a history of coronary artery disease (1% versus 0.3%). However, those with a history of coronary artery disease remain at significantly higher risk at each time point with ~2.5 times the cumulative incidence of circulatory death up to 10 years.

Despite the ongoing high risk of cardiac mortality after transplant, it is important to note that cardiac mortality in kidney transplant recipients has fallen >40% from 915 to 511/100 000 person years between 1988 and 2013. Cardiac mortality rates have been improving in the general populations of most industrialized countries.^20,22-26 However, the improvement seen in kidney transplant recipients has been even more sizable as evidenced by the SMR falling from 8.0 to 5.3 between 1988 and 2013. While the rate of cardiac mortality in transplant recipients remains elevated compared with the respective general population, the narrowing of the gap likely reflects both that transplant recipients are at higher risk, enabling successful prevention efforts to deliver larger impact, as well as the introduction of management guidelines such as KDIGO and the consequent improved cardiac risk factor management in this population. It is possible that selection bias is a contributing factor, but this is unlikely to be a major factor as contemporary recipients are older and have more comorbid cardiac risk factors than historical cohorts.³

As in the general population, men with kidney transplants have the highest absolute rates of circulatory death, and these rates increase with age. However, the relative rates (SMR) reveal that following kidney transplantation women carry the heaviest burden of excess cardiac mortality. Similarly, excess mortality is highest at younger ages. After transplant, young women have almost 27 times the circulatory deaths than similarly aged women in the general population. This is compared to young men with a transplant who have 7.5 times the circulatory deaths seen in the general population. In an analysis of a European cohort of hemodialysis patients, there was no significant difference in excess cardiac mortality between men and women, where men had 9.1 times the standardized rate and women had 8.3 times the standardized rate.²⁷ It is unclear why women are so disproportionately burdened following transplantation. Possible reasons include cardiac risks are under-recognized, and prevention and treatment interventions are not applied or less effective in women with ESKD. A recent paper looking at a community cohort of Americans aged 35–54 found that hospitalizations for acute myocardial infarction were increasing for young women, but decreasing for young men, and that young women had a lower probability of receiving guideline recommended treatment.²⁸ Similar findings exist from Australia, Canada, the United Kingdom, and the Netherlands.^15,29-32 Clinicians have been found to underestimate the cardiovascular risk in women in the general population.^33,34 Further work is needed to elucidate the reasons why women with kidney transplants have such high rates of cardiac mortality.

There were limitations to our study. First, as with all population-based studies, the extent to which risk factors were, or were not, optimally managed was not measured (eg, whether blood pressure targets were met or prescription of and adherence to statin therapy), and so such information could not be included in our analysis. Second, we could not measure additional elements that may affect the risk of posttransplant circulatory death including new onset diabetes after transplant and the effects of immunosuppression. Third, we used probabilistic data linkage for the Australian transplant recipients. Inherent in probabilistic data linkage is the risk of incorrect linkage, although this has previously been shown to be low in this setting at 0.5%.³⁵ Finally, reflecting the underlying demographics of Australia and New Zealand, our study population was predominantly Caucasian. This may have limited our ability to detect race-related differences in cardiac mortality.

In conclusion, although there have been sizable improvements in cardiac mortality for kidney transplant recipients over time, likely due to improved cardiac risk factor management, recipients continue to have significantly higher rates of circulatory death than the general population. Women, particularly young women, have the greatest excess of circulatory deaths, suggesting cardiac risks may be under-recognized in this population, and prevention efforts are less effective or more harmful. This suggests that targeted cardiac risk reduction programs, particularly those focused on younger women, could substantially improve graft and patient outcomes.

ACKNOWLEDGMENTS

The data in this article was supplied by the ANZDATA, the AIHW, and the New Zealand Ministry of Health. We are most grateful to AIHW who assisted with the data linkage process, as well as Kidney Health Australia for their support. The analysis, interpretation, and presentation of these data are the responsibility of the authors alone.