BACKGROUND
The global burden of end-stage kidney disease (ESKD) continues to increase [1]. Although kidney transplantation (KT) confers clear survival and quality-of-life advantages over dialysis [2], the donor supply remains insufficient relative to need. KT accounts for only ~22% of kidney replacement therapy, and the dialysis population is expanding more rapidly [3].
In this context, living donations have become central to KT delivery, shortening wait times, facilitating preemptive transplantation, providing a better graft survival rate, and reducing waitlist mortality [24]. Consequently, living-donor utilization and the pool of eligible donors continue to expand, and ABO-incompatible transplantation and desensitization strategies further broaden access to living donation [5]. In this setting, programs increasingly aim to optimize living donor pathways with careful risk stratification to balance donor safety and recipient benefit [6].
Donor health should always be the primary priority, and favorable long-term outcomes are warranted after donation. However, unilateral nephrectomy reduces renal mass with downstream activation of the renin-angiotensin system and sympathetic tone, and promotes volume-dependent hypertension mechanisms and endothelial dysfunction, which induces increased arterial pressure in living kidney donors [7]. Although many donors tolerate this physiological load without sustained blood pressure (BP) elevation, a subset appears to be more sensitive to BP elevation.
Indeed, observational studies have reported that 3–30% of living kidney donors develop incident hypertension within 2 years [89]. Importantly, several studies have demonstrated a close relationship between donor and recipient outcomes [10]. Paired recipients of donors who later develop ESKD have a higher risk of graft failure than recipients of donors who do not develop ESKD [11].
This suggests that post-donation blood-pressure responses can provide prognostic information in kidney transplant recipients (KTRs). However, the association between donor systolic blood pressure (SBP) and recipient outcomes has not been systematically evaluated.
Accumulating evidence suggests that longitudinal BP trends with age may have greater prognostic importance than baseline BP [12]. An increasing SBP trajectory is reportedly associated with a high risk of developing chronic kidney disease [13]. Additionally, a rapidly increasing SBP over the course of life is associated with a higher risk of cardiovascular diseases (CVDs) [14].
Therefore, we hypothesized that post-donation longitudinal BP changes in kidney donors could serve as markers of donor and paired recipient outcomes. This study aimed to classify donors by characterizing longitudinal SBP patterns using latent class-mixed modeling and to evaluate whether the donor SBP trajectory class was associated with long-term donor and matched recipient outcomes.
METHODS
Study population
Data were obtained from the Korea Organ Transplantation Registry (KOTRY), a prospective multicenter nationwide cohort study of KTRs in South Korea, that collects transplantation-related data from 42 centers [1516]. A total of 4,915 living kidney donors with at least 1 year of follow-up and their matched recipients who underwent transplantation between May 2014 and December 2023 were initially identified. Donors with prior hypertension (n = 1,051) and those with incomplete data (n = 20) were excluded; 3,844 living kidney donors and their matched recipients were included in the final analysis.
The study protocol was approved by the Institutional Review Board (IRB) of the Kyung Hee University Hospital (IRB No. KHUH 2025-12-094), and written informed consent was obtained from all participants before enrollment. This study was conducted in accordance with the principles of the Declaration of Helsinki.
Data collection
The KOTRY database prospectively captures data at pre-specified time points. Living kidney donors were assessed at transplantation, first discharge, and annually thereafter. The KTRs were evaluated at transplantation, first discharge, 6 months post-transplantation, and then annually.
Baseline demographics, comorbidities, and transplant-related variables (including human leukocyte antigen [HLA] typing and desensitization status) were recorded at the time of transplantation. Laboratory measurements, including serum creatinine levels, were obtained during scheduled visits.
The transplantation date was defined as the index date. BP measurements were obtained at each patient’s clinic following routine care procedures. BP was measured at each scheduled visit by trained staff using validated automated or calibrated manual devices with appropriate cuff sizes.
Participants were seated with back supported and rested for ≥ 5 minutes; caffeine, exercise, and smoking were avoided for at least 30 minutes beforehand. Kidney function was evaluated using estimated glomerular filtration rate (eGFR), which was calculated using the 2009 Chronic Kidney Disease Epidemiology Collaboration equation [17]. Proteinuria was deemed present if any of the following were met: 1) urine protein excretion > 300 mg/24 hr or a urine protein-to-creatinine ratio > 300 mg/gCr in either a spot urine sample or a 24-hour collection; 2) urine albumin excretion > 30 mg/24 hr or an albumin-to-creatinine ratio > 30 mg/gCr in either a spot sample or a 24-hour collection; or 3) proteinuria ≥ 1+ on a spot urine dipstick test [18].
Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared (kg/m2).
Study workflow
The study workflow is shown in Supplementary Fig. 1. Eligible living kidney donors were used to identify post-donation SBP trajectory classes based on serial SBP measurements obtained during follow-up.
Each paired recipient was then linked to the SBP trajectory class of their matched donor. Recipient analyses were performed based on matched donor SBP trajectory class, with separate evaluation of recipients with and without baseline hypertension. Donor clinical outcomes during follow-up were summarized separately as supplementary characterization of the SBP trajectory classes.
Trajectory analyses
To characterize the distinct longitudinal SBP patterns, we fitted latent class growth models to serial SBP values collected at every visit from baseline to the end of follow-up. After model convergence, posterior class membership probabilities were computed for each participant, and individuals were assigned to the class with the highest probability. The optimal number of classes and the functional form of time (linear, quadratic, or cubic) were selected using: 1) information criteria (Akaike information criterion and Bayesian information criterion), favoring lower values; 2) an average posterior probability > 0.70 within each class; and 3) maximization of the class-membership log-likelihood [19].
A total of 9,379 SBP measurements were obtained during a median 2.86 (1.61–4.49) year follow-up after donation, and they were included in the modeling (Supplementary Fig. 2). Estimated SBP at donation, estimated annual SBP change, and estimated SBP at 3 years were derived from latent class mixture modeling, whereas the absolute SBP increase from donation to 3 years was calculated using the observed SBP measurements.
Outcome measures
The primary outcome was BP deterioration, defined state-dependently as incident hypertension among recipients without baseline hypertension and uncontrolled hypertension (SBP ≥ 140 mmHg on 2 consecutive visits), among those with baseline hypertension [20]. Key secondary endpoints included annual eGFR decline (mL/min/1.73 m2 per year) modeled with linear mixed models, graft loss, and all-cause death. Graft loss was defined as progression to ESKD requiring maintenance dialysis or retransplantation.
Death-censored graft loss was defined as graft loss events censored at the time of death. Hypertension was defined as a documented physician diagnosis at or before the visit or current use of any anti-hypertensive medication. Diabetes was defined as a documented diagnosis by a physician, current treatment with glucose-lowering therapy (oral agents or insulin), or prescribed dietary therapy.
All-cause graft loss included graft failure from any cause, including return to dialysis, retransplantation, or death with a functioning graft [21]. Participants were followed up until the first occurrence of death, graft failure, loss to follow-up, or end of the study, whichever occurred first.
Statistical analysis
Continuous variables were summarized as mean ± standard deviation and categorical variables as counts (percentages). Group comparisons were performed using analysis of variance for continuous variables and the chi-square test for categorical variables. Correlates of donor SBP trajectory membership at the time of transplantation were examined using multinomial logistic regression and modeling assignments to class 2 (reference: class 1).
Variables with P < 0.10 in univariable screening were entered into multivariable models. Incidence rates were calculated as events per accumulated person-time and presented per 1,000 person-years. Recipient cumulative event curves were estimated using the Kaplan–Meier method and compared using log-rank tests.
Associations between donor SBP trajectory classes and clinical outcomes were evaluated using multivariate Cox proportional hazards models adjusted for age, sex, BMI, ESKD etiology, time on dialysis, positive crossmatch, maintenance immunosuppressants, and living-related donors. The results of the Cox analysis are presented as hazard ratios (HRs) and 95% confidence intervals (CIs). Linear mixed-effects models were used to evaluate the longitudinal changes in recipient eGFR during follow-up.
The models included fixed effects for age, sex, BMI, ESKD etiology, dialysis duration, history of CVD, smoking, SBP, baseline donor eGFR, HLA mismatch number, ABO incompatibility, positive crossmatch, donor-specific antibodies, induction and maintenance immunosuppression, time (linear), and donor SBP trajectory class, with a random intercept for participants. Differences in the eGFR slope were examined using the interaction between SBP trajectory classes and time. Latent class growth modeling and all statistical analyses were conducted in R (version 4.1.3; R Foundation for Statistical Computing, Vienna, Austria) using the lcmm package.
All tests were 2-sided, with a significance threshold of P < 0.05.
RESULTS
Identification of post-donation SBP trajectory patterns
Longitudinal SBP measurements of donors after donation were analyzed using latent class mixture modeling to identify the SBP trajectories (Supplementary Table 1). We identified 2 distinct SBP trajectory classes: 3,654 donors (95.1%) in class 1 and 190 donors (4.9%) in class 2. Fig.
1 shows the changes in SBP after donation. Both class 1 and class 2 donors showed increased SBP during 8 years of follow-up. However, class 2 donors showed a steeper increase than class 1 donors.
The estimated SBP at donation, estimated annual SBP change, and estimated SBP at 3 years, as well as the absolute SBP increase from donation to 3 years were significantly higher in class 2 donors than in class 1 donors (Supplementary Table 2).
Baseline characteristics of donors at the time of donation based on SBP trajectories
The baseline characteristics of donors at the time of donation were compared between SBP trajectory classes 1 and 2 (Table 1). Class 2 donors had a lower prevalence of males and a higher prevalence of smokers than that of class 1 donors. Class 2 donors had higher BMI, SBP, diastolic blood pressure (DBP), and serum glucose and uric acid levels, and lower eGFR compared to that of class 1 donors at the time of donation.
Predictors of donor SBP trajectory class 2
Relevant factors associated with class 2 donors were identified using baseline characteristics at the time of donation (Table 2). In the univariable analysis, female sex, higher BMI, smoking status, higher SBP, lower eGFR, and higher serum glucose levels were associated with donors in the class 2 trajectory. In the multivariable analysis, higher BMI, SBP, and serum glucose level at the time of donation were independently associated with class 2 donors.
Comparison of donor outcomes between class 1 and class 2 SBP trajectory
The cumulative incidences of new-onset hypertension (Supplementary Fig. 3A), new-onset diabetes (Supplementary Fig. 3B), and eGFR < 60 mL/min/m2 (Supplementary Fig.
3C) were significantly higher in class 2 donors than in class 1 donors (all P < 0.001). However, the cumulative incidence of new-onset proteinuria did not differ between the two groups (P = 0.245; Supplementary Fig. 3D).
Baseline characteristics of paired KTRs based on donor SBP trajectories
The baseline characteristics of paired KTRs were compared between donor SBP trajectory classes 1 and 2 (Supplementary Table 3). Paired recipients of class 2 donors had lower HLA mismatch numbers and a more prevalent use of tacrolimus than that of paired recipients of class 1 donors. Other characteristics were not significantly different between the 2 groups.
Incidence rate and risk of incident and uncontrolled hypertension of paired KTRs based on donor SBP trajectories
Table 3 shows the incidence rate and Cox proportional hazard regression analysis of the risk of incident and uncontrolled hypertension in paired KTRs, based on donor SBP trajectories. Among recipients who did not have hypertension before KT (n = 445), recipients of class 2 donors had a higher incidence of new-onset hypertension than that of recipients of class 1 donors (222 per 1000 person-years versus 84 per 1,000 person-years; P = 0.001). Compared with that of recipients of class 1 donors, recipients of class 2 donors had a significantly higher risk of new-onset hypertension in the univariable analysis.
After adjustments for age, sex, BMI, dialysis duration, positive crossmatch, and living-related donor, recipients of class 2 donors had an independently higher risk of incident hypertension compared to those of class 1 donors (HR, 2.70; 95% CI, 1.44 to 5.08; P < 0.002; Table 3). In a full-adjusted model including age, sex, BMI, etiology of ESKD, dialysis duration, positive crossmatch, maintenance immunosuppressants, and living-related donor, recipients of class 2 donors had an independently higher risk of incident hypertension compared to those of class 1 donors (HR, 2.47; 95% CI, 1.31 to 4.66; P < 0.001; Supplementary Table 4). Restricted cubic spline analyses showed an association between the adjusted HRs of incident hypertension and the posterior probabilities of donor SBP trajectory classification (Supplementary Fig.
4). An increase in the posterior probability of class 2 donors (> 0.5) was associated with a continuously increased risk of new-onset hypertension, whereas a lower probability of class 2 donor classification, that is, a higher probability of class 1 donor classification, was associated with a lower risk of new-onset hypertension.
Among recipients with pre-existing hypertension (n = 3,399), the incidence rate of uncontrolled hypertension was similar between recipients of class 1 and class 2 donors (50 per 1,000 person-years vs. 49 per 1,000 person-years). In the Cox proportional hazards regression analysis, recipients of class 2 donors did not have a higher risk of incident hypertension than those of class 1 donors in both univariable and multivariable analyses.
Association between donor SBP trajectories and paired recipient outcomes
Table 4 shows the association between donor SBP trajectories and changes in eGFR of paired recipients. Among recipients without hypertension before KT (n = 445), the eGFR slope of recipients of class 2 donors did not differ from that of recipients of class 1 donors in both univariable and multivariable analyses. Among recipients who had hypertension before KT (n = 3,399), recipients of class 2 donors had a significantly steeper decline in eGFR than those of class 1 donors in univariable analysis.
Recipients of class 2 donors remained to have a significantly steeper decline in eGFR compared to that of class 1 donors (coefficient ß = −0.67; 95% CI, −1.08 to −0.26; P = 0.001), after taking into account the age, sex, BMI. dialysis duration, SBP, baseline donor eGFR, HLA mismatch number, and, positive crossmatch. In a full-adjusted model including age, sex, BMI, etiology of ESKD, dialysis duration, history of CVD, smoking, SBP, baseline donor eGFR, HLA mismatch number, ABO incompatibility, positive crossmatch, donor-specific antibodies, and induction and maintenance immunosuppression, recipients of class 2 donors remained to have a significantly steeper decline in eGFR compared to that of class 1 donors (coefficient ß = −0.68; 95% CI, −1.08 to −0.27; P = 0.001; Supplementary Table 5).
The cumulative incidences of all-cause graft loss (Supplementary Fig. 5A) and death-censored graft loss (Supplementary Fig. 5B) did not differ between recipients of class 1 and class 2 trajectory donors.
DISCUSSION
Using a nationwide cohort of living donors and paired KTRs, we evaluated the longitudinal SBP trajectories of donors after donation, and their clinical impact on donor and recipient outcomes. A distinct subset of donors who showed a steeper increase in SBP was identified, and these donors had a higher incidence of new-onset hypertension and diabetes and eGFR < 60 mL/min/1.73 m2. Paired recipients of these donors had a higher risk of new-onset hypertension among those without pretransplant hypertension and a greater decline in eGFR among those with pretransplant hypertension.
These findings show that a post-donation SBP trajectory characterized by a steeper increasing SBP pattern is associated with adverse donor and recipient outcomes, which underscores the importance of longitudinal donor BP monitoring and supports its potential role in risk stratification of both donors and recipients.
We characterized the longitudinal patterns of donor SBP changes and identified two distinct trajectory groups. This finding is clinically informative because post-donation SBP has often been interpreted using single time-point measurements or binary hypertension diagnoses, which may obscure meaningful heterogeneity in how BP changes after nephrectomy. Our results suggest that SBP follows a broadly linear upward drift over time; however, the rate of increase differs across individuals, resulting in subgroups with a steeper increase.
Importantly, the distinction between classes was driven primarily by slope rather than by complex nonlinear shapes, implying that a progressively widening separation in SBP and cumulative BP burden can emerge with a longer follow-up [1822].
Class 2 donors, defined as those with a steeper SBP increase, had higher metabolic or cardiovascular risks at baseline in terms of BP, BMI, smoking, and serum glucose levels. The independent relevant factors for class 2 donors were higher SBP, BMI, and serum glucose level at the time of donation. This is consistent with previous reports that demonstrated that higher BMI [89] and higher fasting blood glucose [2324] before donation are risk factors for hypertension after donor nephrectomy.
These findings show that class 2 donors have preexisting risks of increasing SBP at the time of donation. Based on the Kidney Disease Improving Global Outcomes guidelines, donor candidates should be counseled on lifestyle practices to modify risks for hypertension and CVD, and these interventions should be initiated before donation and maintained lifelong [25]. Therefore, donors with higher baseline SBP, BMI, and serum glucose level should receive counseling on cardiovascular and metabolic risk reduction before and after donor nephrectomy, as they may experience a steeper increase in SBP following donation.
Although alcohol consumption has been reported to affect BP [2627] and vascular cell function [28], information on alcohol intake of donors was not available in our study, precluding analysis of this factor. In addition, none of the class 2 donors had a history of CVD. Nevertheless, in the absence of brain imaging, subclinical cerebrovascular abnormalities, such as silent lacunar infarctions or other subtle ischemic changes, may have remained undetected and may have influenced subsequent outcomes.
Donor SBP trajectories were associated with adverse donor outcomes. As class 2 donors are characterized by a steeper increase in SBP, it is not surprising that these donors have a higher incidence of new-onset hypertension than that of class 1 donors. The higher incidence of new-onset diabetes may be related to a higher pre-donation BMI [29].
A higher BMI in kidney donors is likely to reflect their metabolic susceptibility, which may predispose donors to diabetes after donation, a finding similar to that in the general population. The higher incidence of eGFR < 60 mL/min/1.73 m2 in class 2 donors may be attributable to the greater burden of post-donation hypertension and diabetes, conditions that have been linked to accelerated eGFR decline in prior studies [93031]. Furthermore, nephrectomy itself imposes physiological alterations, including hyperfiltration, upregulation of the renin-angiotensin-aldosterone system, and changes in vascular tone [7].
The remaining kidneys of living donors have adequate functional reserve capacity, which enables compensatory, adaptive hyperfiltration, typically increasing their function in the months post-donation [32]. We speculate that class 2 SBP trajectory donors may have decreased functional reserve capacity because of rapidly increasing SBP levels, in addition to higher arterial stiffness and BMI.
Several studies have demonstrated the close relationship between donor and recipient outcomes. KTRs of living donors who subsequently developed ESKD also had a higher graft failure rate [11], and post-donation SBP was associated with graft failure [10]. However, little is known about the effect of post-donation longitudinal BP trajectory of donors on the outcomes of paired recipients.
In this study, we found that the donor SBP trajectory affected BP of paired KTRs. Among KTRs without pre-transplant hypertension, recipients of class 2 donors had an increased risk of incident hypertension. These findings suggest that the two donor kidneys share risk factors such as subclinical microvascular dysfunction or kidney pathology, which may have contributed to the development of incident hypertension in paired recipients.
However, among KTRs with pretransplant hypertension, recipients of class 2 donors did not have an increased risk of uncontrolled hypertension. Therefore, a class 2 donor relationship may not substantially influence the severity or treatment response to hypertension.
We also evaluated the association between post-donation SBP trajectory and changes in the eGFR of paired recipients. Among recipients who had hypertension before KT, recipients of class 2 donors showed a more rapid decline in eGFR compared to those of class 1 donors, while among recipients without pretransplant hypertension, recipients of class 2 donors did not show a faster eGFR decline. This finding suggests that the recipient baseline hypertension status modifies the relationship between the donor SBP trajectory and graft function.
One plausible explanation is that class 2 donor kidneys may have been more susceptible to BP changes, such as arterial stiffness or vascular dysfunction. If class 2 donor kidneys were transplanted into recipients with preexisting hypertension, the interaction between the donor and recipient could have resulted in increased intraglomerular pressure and endothelial dysfunction of the graft kidney, leading to a more rapid decline in eGFR. This interpretation is supported by previous studies showing that subclinical donor kidney susceptibility identified on implantation biopsy may persistently influence subsequent graft function [33].
This study has some limitations. First, confounding factors, such as lifestyle practices, alcohol consumption, changes in the metabolic parameters, or subclinical cerebrovascular abnormalities of donors were not included in the analysis. Second, the KOTRY database did not capture the history of transient ischemic attack or other cerebrovascular events among recipients.
Consequently, we were could not assess whether the shared microvascular or vascular risk profile has clinically meaningful manifestations in this population, which warrants further investigation in future studies. Third, data on the types and numbers of anti-hypertensive medications were not available for inclusion in the analyses [34]. Fourth, the relatively small sample size in the class 2 donor group may increase the risk of model overfitting and limit the statistical power of the between-group comparison.
Fifth, a longer follow-up period is warranted to evaluate the impact of the donor SBP trajectory on donor ESKD development or graft failure in paired recipients. However, the strengths of this study include the use of a nationwide cohort and comprehensive analysis to assess the contribution of longitudinal donor SBP trajectories to post-donation donor and recipient outcomes.
In conclusion, a distinct post-donation SBP trajectory of living donors, characterized by a steeper increase in SBP, was associated with incident hypertension and diabetes, low donor eGFR, incident hypertension in KTRs without pre-transplant hypertension, and greater eGFR decline in KTRs with pre-transplant hypertension. These findings suggest that post-donation SBP trajectories capture the shared risk across the donor–recipient pair and are clinically relevant markers of recipient prognosis.