Risk of ESRD in Prior Living Kidney Donors

Jennifer L. Wainright; Amanda M. Robinson; Amber R. Wilk; David K. Klassen; Wida S. Cherikh; Darren E. Stewart


American Journal of Transplantation. 2018;18(5):1129-1139. 

In This Article


Our analysis of 123 526 LKDs who donated in the United States between April 1, 1994 and September 30, 2016 identified 218 cases of ESRD in the first 2 decades after donation. Statistically significant risk factors for ESRD in multivariable analyses included being male, black, having lower eGFR, and having higher BMI at donation. We found a significant interaction between age and race, with older age at donation conferring higher risk for white donors, but younger age at donation conferring higher risk for black donors. After adjusting for race and other factors, among lower–income donors (median ZIP code income <$65.1k), LKDs in less wealthy neighborhoods had higher ESRD risk than those from higher income neighborhoods; the effect was not evident above $65.1k. Parents and full siblings of their recipient had a higher risk than unrelated donors, with identical twins of recipients having an exceedingly high relative risk. Because of the small number of identical twins and events in the cohort, however, this finding should be interpreted with caution. There was great variation in absolute risk among donors, with age, race, and sex playing important roles in predicting risk, and risk increasing exponentially over time after donation.

Several of our results support findings of past research conducted with other cohorts, including concern about high BMI at donation, a modifiable risk factor,[12,14] and donor sex, with male donors being at greater risk.[4,12–15,53] We have also expanded on others' findings, including the interaction between age at donation and race found in prior research.[14] We confirmed that older age was a significant predictor of 20–year ESRD risk for white donors, but also found a significant association between younger age at donation and greater 20–year risk of ESRD for black donors. Steiner asserts that some younger donors, and black donors in particular, have a higher lifetime risk of ESRD than donors who donate at older ages; and this is the first study to show such a trend exists within the first 2 decades after donation.[29,30,32] The lack of higher risk for older black donors may be because development of clinically identifiable renal disease in susceptible black individuals is likely to have occurred by age 50 or 60, precluding donation.[14,54] Our findings suggest additional screening and consent procedures may be warranted for younger black potential donors,[55] keeping in mind that more stringent standards could increase disparities in access to transplantation or cause harm to rejected potential donors.[56–58]

Although previous research has documented familial clustering of ESRD in the population,[28,59] and donors' relationship to their recipient has predicted ESRD in past research, no research on LKDs to date has studied donor risk by type of relative (eg, sibling vs parent).[1,4,12,14] Our study examines such differences in risk, and as expected, we found significant associations between being the parent or sibling of the recipient and risk of ESRD, compared with unrelated donors. Our study expands this area of research by finding a highly significant association for identical twins of recipients. Clinicians should consider this high relative risk in the context of the small cohort of identical twin donors and resulting wide confidence intervals (eg, absolute 20–year risk for a 20–year–old white female LKD who is the identical twin of her recipient and has median clinical values and income is 150 per 10 000 donors [95%CI: 0–302]).

We found a significant association between lower eGFR at donation and higher 20–year risk of ESRD that is consistent with previous research[12] and the relationship between premorbid eGFR and lifetime risk of ESRD in population studies.[32] Our finding for systolic blood pressure at donation did not reach statistical significance, but warrants further investigation because of the direction and magnitude of the effect. Clinicians should consider both eGFR and systolic blood pressure in their decisions and conversations with potential donors.

This study investigated SES and ESRD outcomes in the general LKD population. Data limitations likely affected Lentine and colleagues' ability to find significant results for SES in their study of LKDs with private insurance,[16,17] while Garg found that SES was associated with poor outcomes for donors.[11] Our study found greater risk of ESRD among donors from poorer neighborhoods, and this effect was seen for white, black, and other race donors, but not for Hispanic donors. While familial trends in ESRD may account for some of the risk seen for donors, environmental factors such as access to health care may also be important factors.[54] Previous research has found many donors are uninsured at donation, and younger, male, and racial minority LKDs are more likely to be uninsured,[60] suggesting those at greatest risk of ESRD are also least likely to have adequate healthcare resources. Between June 1, 2004 and September 30, 2016, 15.6% of our cohort did not have insurance at donation and data were missing for another 28.1%. The fact that we assessed neighborhood income with ZIP code at donation addresses concerns that donor illness could be causing lower SES instead of lower SES causing higher rates of illness, but our results suggest all LKDs need regular follow–up.

One of the most important strengths of our study is excellent ascertainment of postdonation ESRD and death. Aside from rare cases that developed in another country or resulted in death before a CMS Form 2728 was submitted, we have complete and unbiased ESRD data for our cohort. Our decision to limit our cohort to donors who donated on or after April 1, 1994, when the OPTN began collecting SSN for donors, had the disadvantage of shortening the follow–up time. We judged it more important, however, to eliminate known biases in ascertainment of post–April 1, 1994 onset of ESRD that result if earlier donations are included. These biases are based on factors known to be important in predicting ESRD in LKDs: sex and age (ie, young female donors missed in the linkage if they marry and change their names between donation and ESRD onset), and ethnicity (ie, false positives for common names, a problem that varies by ethnic group). Restriction of our cohort to donations after April 1, 1994 also reduced these biases in assessment of postdonation death for death–censored analyses. Because we knew these biases occur if older donations are included, we sacrificed longer follow–up time for unbiased ascertainment of ESRD and death.

An important limitation is our inability to assess ESRD occurring more than 20 years after donation. Prior research[1] found median time from donation to onset of ESRD to be greater than 20 years. Diabetes—the most common cause of ESRD[32]—often develops later in life and requires several decades before causing onset of ESRD. As a result, there is an exponential increase in risk of ESRD with age,[61] and we see evidence of this pattern—especially for black LKDs, first–degree relatives, and lowest SES LKDs—in the increasing slopes in Figure 1. Our knowledge of this pattern, relatively short follow–up time, and the small percentage of LKDs with ESRD caused by diabetes force us to assume our study does not give a complete view of donors' lifetime risk of ESRD.[15,61]

Another limitation is the use of ZIP code–level median income data as a proxy for donor SES. Donor–level income data would be a better measure of SES. A final limitation is the assessment of donor race, which was reported by transplant programs to the OPTN and may be based on self–report by donors or staff assessment of a donor's background.

Our results confirm that while 20–year risk of ESRD is low for LKDs overall, risk is not uniform among donors.[14,29,30,32,62] Relative risks among donors are useful to clinicians, but potential donors must also understand their absolute risk when making the decision to donate. Potential LKDs should receive the best available information—including limitations of existing research—about their own personal risk of ESRD after donation.[17,56,62–66]