The study’s design reflects a pragmatic response to a structural gap in Japanese health data. Unlike the United States, where donor outcomes can be tracked through linked national records such as the Scientific Registry of Transplant Recipients, Japan has no routine individual-level linkage between its transplant registry and its dialysis registry. The researchers therefore had to construct their estimate from aggregate cross-tabulations, combining the Japanese Renal Data Registry’s relatively new item on donor history among incident dialysis patients with published counts of living-donor transplants stretching back to 1999. This period approach, which applies the dialysis rates observed in 2019 through 2024 to every interval since donation, is a well-established epidemiological technique, but it carries an inherent assumption: that the risk of kidney failure in a given decade after donation is similar across donation cohorts. The authors checked this assumption and found no meaningful differences in rates between donation cohorts, which lends some support to the approach even though it cannot be verified with certainty.

The statistical machinery behind the headline figure deserves attention because it illustrates how epidemiologists handle uncertainty when perfect data are unavailable. The team used cause-specific hazards calculated on death-censored donor-years, meaning that years in which a donor died were removed from the denominator so that the estimate reflects the risk of dialysis among living donors rather than being diluted by deaths. These hazards were then converted into a cumulative incidence using the Aalen–Johansen estimator, which properly accounts for death as a competing risk — a donor who dies before developing kidney failure can never reach that outcome, and ignoring this would overstate the dialysis risk. The authors then produced three estimates that bracket the truth: an unadjusted figure of 0.5 percent, a death-censored net estimate of 1.3 percent, and a primary adjusted estimate of 1.1 percent that allocates interval-unknown cases proportionally and imputes donor cases among patients whose donor status was never recorded. The honesty of this bounding approach is notable; rather than presenting a single number as definitive, the authors explicitly state that the resulting range of roughly 0.5 to 2 percent is not a confidence interval and that values above 2 percent cannot be excluded if the sensitivity of the registry item falls below 0.6.

The demographic profile of Japanese donors makes these findings particularly relevant to counselling conversations in that country. More than 40 percent of living donations in Japan come from spouses, who are unrelated donors, and roughly half of donors are aged 60 or older at the time of donation. This age structure differs substantially from that of the United States, where donors tend to be younger at donation. The comparison with American data illustrates this point vividly: the net Japanese estimate of 1.3 percent at 20 years is 2.7 times the US figure of 0.49 percent, but when the comparison is restricted to donation before age 60, the ratio narrows to 1.7, and it falls further when Japanese estimates are compared against white or unrelated American donors specifically. In other words, much of the apparent gap between the two countries is explained by the simple fact that Japanese donors are older, and kidney failure risk rises steeply with both age and time since donation. This is a crucial nuance for anyone interpreting the numbers: the elevated Japanese figure is not necessarily evidence that Japanese donors face intrinsically higher risks, but rather that the population being measured is older on average.

The authors are also candid about the limits of comparing donors with the general population. Donors are selected for good health, so their baseline risk should sit well below that of the general population, which includes people with diabetes and chronic kidney disease — conditions that account for a large share of dialysis initiations in Japan as elsewhere. American data suggest donors have roughly 0.3 times the lifetime kidney failure risk of the general population, so a standardized incidence ratio near 1, which is what the Japanese analysis found at 1.23, would already imply some excess risk relative to what selection alone would predict. Yet the authors caution that this ratio depends heavily on the accuracy of the donor-history item in the dialysis registry: with 41 percent of incident dialysis patients having no recorded donor status, the standardized incidence ratio falls below 1 if the positive predictive value of the recorded item reaches 0.8. The data therefore cannot settle whether Japanese donors carry an excess risk or whether donor selection in Japan is less stringent than elsewhere — a question made more pointed by the observation that nearly a quarter of recent Japanese donors have hypertension, most of it treated. The authors’ recommendation to focus on absolute risk rather than ratios is a methodologically sound one for counselling purposes.

Several of the limitations the authors describe are worth unpacking for readers who want to understand what the estimate can and cannot capture. Misclassification runs in both directions: six of the 94 recorded donors had a documented kidney transplant, suggesting their kidney failure may have occurred in the transplanted organ or after re-transplantation rather than in the remaining native kidney, and two had an incompatible primary disease. Excluding these cases lowers the estimate slightly to 1.0 percent. More troubling is the reverse direction — donors recorded as non-donors are undetectable by design, and the registry has no code for kidney failure occurring in a solitary kidney, so some genuine donor cases may be buried in the “other” primary disease category, where donor history was 14 times as frequent as in diabetic nephropathy. The registry also counts only dialysis, meaning donors who received a pre-emptive transplant, died before reaching dialysis, or chose conservative management of advanced kidney disease are missed entirely. This last omission is particularly consequential for older donors, among whom conservative management is more common, so the 1.6 percent figure for donation at age 60 or older likely understates the true burden of severe kidney disease in that group.

The finding that over half of recorded dialysis initiations occurred more than 20 years after donation carries significant implications for how the numbers should be communicated. The exploratory extension of the analysis, using transplant denominators going back to 1971, suggests a cumulative risk of about 2.2 percent at 30 years — roughly one in 50 donors. Because today’s donors are older than the cohorts whose second and third post-donation decades the study could observe, the 20-year figure probably understates the lifetime risk that current donors will actually face. The authors’ standardization of recent age-specific estimates to the current donor age mix yields 1.3 percent at 20 years, supporting this interpretation. For prospective donors, the practical message is that kidney failure risk after donation is a moving target that grows with advancing age, and that a counselling conversation framed only around a 20-year horizon may substantially understate the relevant lifetime picture, particularly for someone donating in their sixties.

What the study does not claim is equally important. The authors emphasize that these figures represent the absolute risk of dialysis after donation, not the risk attributable to donation itself. Disentangling how much kidney failure among donors would have occurred anyway, had they not donated, requires comparison against a carefully matched non-donor population, which the aggregate registry data cannot provide. The authors also note that the figures do not by themselves change who may be eligible to donate; rather, they equip prospective Japanese donors with a concrete, nationally derived number for the consent process and support the case for monitoring kidney function well beyond 20 years after donation, especially in older donors. Registry linkage, which the authors identify as the path forward, would enable precise age- and sex-specific estimates and resolve the uncertainties that currently force the analysis to rely on bounding assumptions. Until then, the Japanese figures join those from the United States and Norway as part of a small but growing international evidence base, offering donors in each country numbers calibrated to their own population’s demographics and health system rather than imported from abroad.

Subject of Research: Long-term risk of dialysis initiation among living kidney donors in Japan

Article Title: Risk of dialysis initiation within 20 years after living kidney donation in Japan: a period estimate from the national dialysis and transplant registries

Article References: Torigoe, K., Tsuyuki, T., Iwata, M., Otsuka, E., Tsuji, K., Yamashita, A., Torigoe, M., Kitamura, M., Muta, K., Imamura, R., & Nishino, T. (2026). Risk of dialysis initiation within 20 years after living kidney donation in Japan: a period estimate from the national dialysis and transplant registries. Transplant International, 39, Article 17852. https://doi.org/10.3389/ti.2026.17852

Image Credits: AI Generated

DOI: 10.3389/ti.2026.17852

Keywords: living kidney donation, dialysis, kidney failure, Japan, registry study, transplantation, nephrology, donor risk, end-stage renal disease, informed consent, cumulative incidence, aging donors