Is the Risk of Cardiovascular Disease Increased in Living Kidney Donors?

A Danish Population-Based Cohort Study

Philip Munch; Christian Fynbo Christiansen; Henrik Birn; Christian Erikstrup; Mette Nørgaard

Disclosures

American Journal of Transplantation. 2021;21(5):1857-1865. 

In This Article

Material and Methods

Study Design and Setting

This nationwide cohort study included data from the Danish National Patient Registry (DNPR), the Danish Civil Registration System (CRS), the Danish National Prescription Registry (DPR), and the Scandinavian Donations and Transfusions (SCANDAT) database. The Danish National Health service provides tax-supported health care for the entire Danish population (5.8 million inhabitants), ensuring free access to general practitioner and hospitals. All Danish residents are assigned a unique personal identifier (the CPR-number), which permits unambiguous individual linkage between each Danish registry.[19]

Data Sources

DNPR contains both administrative and clinical data on diagnoses, examinations, and surgical procedures on all inpatient visits since 1977 and outpatient clinic and emergency room visits since 1995.[20] Diagnoses are classified according to the International Classification of Diseases, Eighth Revision (ICD-8) through 1993, and Tenth Revision (ICD-10) thereafter. Surgical procedures are classified according to the Nordic Medico-Statistical Committee Classification of Surgical Procedures.

DPR contains data on all prescriptions redeemed by Danish residents at community pharmacies since 1995. It includes information on the drug type according to the Anatomical Therapeutic Chemical (ATC) codes and the dispensing date.[21]

The SCANDAT database contains detailed information on virtually all Danish blood donors and transfusion recipients since 1982 including the type of donations (whole blood, plasma, and other types) and dates of donations/transfusions.[22]

CRS contains data on vital status and migration to and from Denmark.[23]

Unfortunately, none of the abovementioned data sources contained information on actual blood pressure levels, GFR, smoking status, or body mass index.

Study population

Kidney Donors. We used DNPR to identify living kidney donors in Denmark from January 1, 1996 to December 10, 2018. We identified living kidney donors based on the procedure code for living donor nephrectomy combined with either a code indicating a hospital contact related to the donation of a kidney or a code indicating examination of a potential donor registered before the date of nephrectomy (see Table S1 for codes). The date of nephrectomy served as index date for kidney donors.

We excluded kidney donors with an inpatient or outpatient clinic diagnosis of one or more of the following CVDs registered in DNPR before index date: AF, angina pectoris, myocardial infarction (MI), ischemic stroke, intracranial hemorrhage, transient ischemic attack, or heart failure. Furthermore, we excluded kidney donors with a prior diagnosis of chronic liver disease, diabetes, COPD, cancer or autoimmune disease before index date. In addition, we excluded individuals with a redeemed prescription of any antihypertensive or antidiabetic drugs within 1 year prior to index date (see Table S1 for codes).

General Population Cohort. We identified 10 healthy age- and sex-matched comparisons from the general population for each kidney donor through CRS. The date of nephrectomy of the corresponding kidney donor served as index date for the general population cohort. To exclude preexisting conditions that would normally exclude living kidney donation we excluded individuals with inpatient or outpatient clinic diagnoses of the abovementioned CVDs, chronic liver disease, chronic kidney disease, diabetes, chronic obstructive pulmonary disease (COPD), cancer, or autoimmune disease before index date or with redeemed prescriptions of any antihypertensive or antidiabetic drugs within 1 year prior to index date (see Table S1 for codes).

Blood Donor Cohort. The SCANDAT database is hosted at a different national server than our main dataset containing kidney donors and the general population cohort and thus, for data protection reasons, direct linkage between the SCANDAT database and our main dataset was not feasible. Instead, we constructed a secondary external comparison cohort including blood donors registered from 1995 to 2017 in the SCANDAT database. The SCANDAT database was linkable to datasets from DNPR, DPR, and CRS only containing information on blood donors. To ensure that included blood donors were properly evaluated similar to kidney donors, who are evaluated by the health care system for some time before donating, we required blood donors to have donated blood a minimum of four times to be included using the date of the fourth blood donation as index date. We excluded blood donors who had diagnoses of the abovementioned CVDs, chronic liver disease, chronic kidney disease, diabetes, COPD, cancer, or autoimmune disease before index date or redeemed prescriptions of any antihypertensive or antidiabetic drugs within 1 year prior to index date.

When comparing kidney donors with blood donors we further restricted the population to kidney and blood donors aged 25–65 years at index date because of a very low number of kidney donors below age 25 and an upper age limit for blood donation.

In the blood donor cohort, diagnoses recorded in relation to emergency room visits could not be separated from diagnoses recorded in relation to outpatient visits from 2014 and onwards. We therefore included emergency room diagnosis codes in our definition of diagnoses within the last 4 years of the study period. In order to minimize bias, we included emergency room diagnosis codes within the last 4 years for the living kidney donor cohort as well when comparing with blood donors.

In Denmark, neither kidney donors nor blood donors receive any financial compensation for their donation.

Outcomes

We examined the following outcomes: (1) hypertension; (2) AF; (3) angina pectoris; (4) MI; (5) ischemic stroke; (6) major adverse cardiovascular events (MACE) defined as the composite of MI, ischemic stroke and all-cause mortality; and (7) all-cause mortality. Hypertension was defined by the redemption of prescriptions of minimum two different antihypertensive drug classes with less than 180 days between the two prescriptions, and with the date of the last prescription serving as the date of outcome. The other cardiovascular outcomes were based on inpatient and outpatient clinic diagnosis codes (see Table S1 for codes). Vital status was ascertained trough CRS.

Covariates

To address potential confounding from selected, preexisting diseases, we identified any inpatient or outpatient clinic diagnoses of alcohol-related disorders recorded at least 2 years before the index date using the DNPR. Furthermore, we identified prescriptions of lipid-lowering agents and antidepressants recorded in DPR within 1 year before the index date (see Table S1 for codes).

Statistical analysis

For each outcome, we followed all persons from index date until the first diagnosis of the specific outcome, death, emigration or end of follow-up whichever came first (see Figure 1 for illustration). For kidney donors and the general population cohort, follow-up ended September 30, 2018 when examining hypertension and December 10, 2018 when examining the remaining outcomes. For blood donors, follow-up ended September 30, 2017 when examining hypertension and December 10, 2017 in the remaining analyses.

Figure 1.

Illustration of observation periods in kidney donors, the general population cohort, and the blood donor cohort

For kidney donors and the blood donor cohort the study period was divided into four calendar periods. Calendar periods were used when comparing kidney donors with blood donors. In order to compare kidney donors to blood donors with approximately same potential follow-up time, calendar periods for blood donor cohort were shifted 1 year backwards compared with kidney donors (see Figure S1 for illustration).

Patients' characteristics are presented as median with interquartile range (IQR) or proportion in percentage. In addition, we calculated weighted medians and weighted proportions for blood donors based on the kidney donors' distribution of sex, age (25–39, 40–49, 50–59, and 60–65 years) and calendar periods. Age-, sex-, and calendar period-weighted estimates were provided to enable comparison of kidney donors and blood donors while controlling baseline differences in age and sex.

First, we compared the risks of the outcomes in kidney donors with those in the general population cohort. For each endpoint we calculated the incidence rates (IRs) with 95% confidence intervals (CIs) in both cohorts and the hazard ratio (HR) with 95% CI using Cox proportional-hazards regression analysis. Assumptions of proportional hazards were checked using log-minus-log plots and found acceptable. We also calculated the cumulative 10-year risks of the outcomes with 95% CIs, treating death as a competing risk (except for MACE and all-cause mortality). Furthermore, we plotted the cumulative incidence of MACE and all-cause mortality, as well as the cumulative incidence of hypertension and AF treating death as a competing risk.

Next, we compared the risks of the outcomes in kidney donors with those in the blood donor cohort. For each endpoint we calculated the IRs with 95% CIs in both cohorts as well as standardized incidence ratio (SIR) with 95% CI as a measure of the relative risk. SIRs were calculated as the number of observed cases among kidney donors divided by the expected number of cases, which is computed as the time at risk multiplied by the blood donors' incidence rates according to sex, age (25–39, 40–49, 50–59, and 60–65 years) and calendar periods and summing the products. In addition, we repeated the analysis while restricting to the period of 1996–2017 and comparing kidney donors and blood donors from the same period.

As a sensitivity analysis to address the impact of including emergency room diagnoses in the last part of our study period when comparing kidney donors with blood donors, we repeated the analysis comparing kidney donors with the general population cohort while allowing emergency room diagnosis codes when excluding cohort members with previous diagnoses.

Statistical analyses were performed using STATA 14.0 and 16.1 (StataCorp, College Station, TX).

processing....