# paper.pdf ## Metadata - PDFFormatVersion=1.5 - Language=en-US - IsLinearized=true - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - pdf:producer=Adobe LiveCycle PDFG ES - xap:createdate=2026-06-02T13:34:33+05:30 - xap:creatortool=Arbortext Advanced Print Publisher - xap:modifydate=2026-06-02T13:34:33+05:30 - xap:metadatadate=2026-06-02T13:34:33+05:30 - xapmm:documentid=uuid:8addd02c-2a8d-4824-a074-0375b52bef83 - xapmm:instanceid=uuid:9735c38c-787b-4f5b-aca7-e8422204bc96 - prism:aggregationtype=Journal - prism:publicationname=Renal Failure - prism:copyright=© 2026 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. - prism:volume=48 - prism:number=1 - prism:pagerange=- - prism:doi=10.1080/0886022X.2026.2680735 - prism:url=https://doi.org/10.1080/0886022X.2026.2680735 - dc:format=application/pdf - dc:identifier=10.1080/0886022X.2026.2680735 - dc:language=en-US - dc:title=Post-transplant anemia as a persistent risk factor for graft dysfunction in kidney transplant recipients - dc:publisher=Taylor & Francis - dc:description=Renal Failure, 2026. doi: 10.1080/0886022X.2026.2680735 - dc:creator=Tsujita Makoto,Matsuoka Yutaka,Uchida Kazuharu,Morozumi Kunio - dc:subject=Post-transplant anemia,kidney transplantation,graft function - jav:journal_article_version=VoR ## Contents ### Page 1 Post-transplant anemia as a persistent risk factor for graftdysfunction in kidney transplant recipients Makoto tsujitaa,bbbbab1.2.3. https://doi.org/10.1080/0886022X.2026.2680735 CONTACT Makoto Tsujita tsujita4850@gmail.com Department of nephrology, Tajimi Clinic, 2–51 Otowa-cho, Tajimi-shi, Gifu 507-0037,Japan ARTICLE HISTORY Received 18 december2025 Revised 20 May 2026 accepted 23 May 2026 This is an Open access article distributed under the terms of the Creative Commons attribution license (), whichpermits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. The terms on which this article has beenpublished allow the posting of the accepted Manuscript in a repository by the author(s) or with their consent. KEYWORDS Post-transplant anemia;kidney transplantation;graft function; [](http://www.tandfonline.com/) [](http://crossmark.crossref.org/dialog/?doi=) CliniCal Study © 2026 The author(s). Published by informa uK limited, trading as Taylor & Francis Group. ### Page 2 - ![Infographic summarizing a study on persistent post-transplant anemia in kidney recipients, featuring study design, findings, and clinical messages. This infographic presents a study on persistent post-transplant anemia \(PTA\) and graft function decline. It includes three panels: "STUDY DESIGN," detailing a cohort of 323 kidney transplant recipients from Masuko Kinen Hospital \(2019-2021\), with a flowchart showing recipients receiving and not receiving ESA therapy; "MAIN FINDING," highlighting that higher time-updated hemoglobin \(Hb\) is linked to better graft function \(eGFRave\) with beta = -1.86/year and beta = 0.76 per 1 g/dL; and "CLINICAL MESSAGE," emphasizing that persistent PTA indicates higher clinical risk. A diagram shows relationships among Hb, eGFR, and PTA.]() ,]. Furthermore, accumulating evidence suggeststhat Pta is negatively associated with long-term clinical outcomes in Ktx recipients, such as graft failure,all-cause mortality, and progressive deterioration of graft function []. Pta is typically classified into early and late periods []. With regard to its impact on graft function,no randomized controlled trials (RCts) have shown the optimal target hemoglobin (Hb) levels for bettergraft function in the early post-transplant period [,]. in the late post-transplant period, in a retrospec-tive cohort study, Heinze et al. demonstrated that increasing Hb levels to above 12.5g/dl witherythropoiesis-stimulating agents (ESas) in Ktx recipients was associated with increased mortality, andthat this effect was significant at Hb concentrations above 14.0 g/dl []. However, our 3-year RCt demon-strated that higher Hb levels (Hb < 13.5 and ≥ 12.5 g/dl; mean achieved Hb: 12.8 g/dl) were associatedwith better graft outcomes compared to lower Hb levels (Hb < 11.5 and ≥ 10.5 g/dl). Mean differencein estimated glomerular filtration rate (eGFR) was −1.0 ml/min/m2 in high Hb group and −5.1 ml/min/m2in low Hb group. Furthermore, there was no difference in cardiovascular events between the two groups,demonstrating the safety of the high Hb group []. Similarly, Choukroun et al. reported that patients inthe normalization group (Hb 13–15 g/dl; mean achieved Hb: 13.4 g/dl) had more favorable outcomesthan those in the partial correction group (Hb 10.5–11.5 g/dl) []. Both of these RCts indicated thathigher Hb levels were associated with improved graft survival. the Japanese Guideline for Renal anemia in Chronic Kidney disease published in 2015 recommendeda target Hb level of <13.0 g/dl in the late period. Based on these findings, Hb target levels (<13.0,≥12.5 g/dl) were established to maintain graft function stability and have been applied in real clinicalpractice at Masuko Kinen Hospital since 2019. However, real-world data remain limited regarding whetherpersistent Pta despite ESa therapy is associated with subsequent graft dysfunction. therefore, we exam-ined the association between persistent Pta and graft function decline over 2 years in routine outpatientpractice, including recipients receiving ESa therapy. Materials and methods Study design a single-center retrospective observational study was conducted at Masuko Kinen Hospital in Japan toexamine the association between persistent Pta and subsequent graft function trajectory in recipients≥1 year after Ktx. Patients were divided into those receiving and not receiving ESa therapy, and changes ### Page 3 in kidney function over 2 years were compared between the two groups. Patients who needed ESa weregiven subcutaneous Epoetin Beta Pegol (Mircera®; Chugai Pharmaceutical Co. ltd., Japan) 0–250 μg ateach visit, and the doses and interval were adjusted according to Hb level and its target. (<13.0, ≥12.5 g/dl). Measurements of kidney function Creatinine-based eGFR (eGFRcre) was calculated using the equation published by the Japanese Societyof nephrology []: 194 × (age)−0.287 × (sCr)−1.094, including a correction factor of 0.739 for women. CystatinC-based eGFR (eGFRcys) was also calculated using the equation published by the Japanese Society ofnephrology []: (104 × cystatin-C−1.019 × 0.996age)-8 for men; (104 × 0.929 × cystatin-C−1.019 × 0.996age)-8for women. averaged eGFR (eGFRave) was calculated as the average of eGFRcre and eGFRcys. Our pre-vious study demonstrated that eGFRave may better reflect estimated kidney function in Japanese recip-ients with Ktx []. in this study, therefore, eGFRcre and eGFRave were assessed. Participants all outpatients at Masuko Kinen Hospital in 2019 were recruited. after excluding 14 patients, including 7lost to follow-up due to relocation, 5 receiving renal replacement therapy (including 2 reinitiating dialysis),and 2 deaths, a total of 323 outpatients were enrolled in this study. shows the study flowchart. the demographic and clinical characteristics of the patients were retrieved from the hospital elec-tronic medical records. informed consent for the data collection was waived due to the retrospectivestudy design. Statistical analysis data were expressed as means ± standard deviation, 95% confidence intervals (Cis), or medians (inter-quartile ranges). differences in variables between groups were evaluated using the Wilcoxon rank-sum ![Flowchart of patient enrollment at Masuko Kinen Hospital, showing 323 outpatients, 14 exclusions, and ending with 73 patients per group after propensity score matching. Flowchart illustrating the enrollment of 323 outpatients at Masuko Kinen Hospital, divided into 215 not receiving ESA and 108 receiving ESA. It details 14 exclusions: 7 due to relocation, 2 deaths, and 5 receiving renal replacement therapy. Final totals are 210 not receiving ESA and 99 receiving ESA, with 73 patients in each group after propensity score matching.]() Figure 1. Flowchart of this study. ### Page 4 test or paired or unpaired Student’s t test, as appropriate. the primary analysis used a linear mixed-effectsmodel fitted to repeated eGFR measurements obtained at baseline and at 1 and 2 years. time and Hbwere included as fixed effects, with Hb treated as a time-varying covariate, and a patient-specific randomintercept was included to account for within-patient correlation. time was modeled as a continuousvariable (0, 1, and 2 years), and a patient-specific random intercept was included to account forwithin-patient correlation. the model was adjusted for ESa use, age, sex, baseline eGFRave (time = 0),and baseline urine protein, which was included a priori as a fixed effect to account for baseline diseaseseverity. as a supportive analysis, the same longitudinal model was also applied using eGFRcre. Supportive analyses also included linear regression analyses to examine the association of baselineclinical factors with the 2-year change in kidney function. in addition, propensity score matching wasperformed to reduce differences in clinical and demographic characteristics between recipients receivingand not receiving ESa therapy. Propensity scores were estimated using a logistic regression model incor-porating age, sex, eGFRave, transferrin saturation, and urinary protein. Matching was performed using 1:1nearest-neighbor matching with a caliper width of 0.20, without replacement. the balance betweengroups before and after matching was assessed by descriptive comparisons of baseline characteristics.after matching, multivariable regression was performed as a supportive analysis to account for residualconfounding and improve precision. For a supportive subgroup analysis restricted to recipients receiving ESa therapy, persistent anemiawas defined as Hb < 12.0 g/dl at two or more of the three assessments during the 2-year observationperiod. this cutoff was selected as a clinically pragmatic threshold for supportive subgroup analysis andwas not intended as a definitive guideline-based threshold. in this subgroup, multivariable linear regres-sion was used to examine the association between persistent anemia and the 2-year change in eGFRaveafter adjustment for age, sex, baseline eGFRave, and baseline Hb. all p values were two-sided. a p value of <0.05 was considered statistically significant. all statisticalanalyses were performed using JMP version 17 (SaS institute, inc., Cary, nC). Results Baseline characteristics the demographic and clinical characteristics of the study cohort are shown in. a total of 323kidney transplant recipients were included, with 215 recipients not receiving ESa therapy and 108 recip-ients receiving ESa therapy. at baseline, recipients receiving ESa therapy had lower Hb levels (12.0 ± 0.9vs. 13.4 ± 1.2g/dl) and lower eGFRave (39.3 ± 13.6 vs. 51.1 ± 14.0 ml/min/1.73 m2) than those not receiv-ing ESa therapy, indicating substantial baseline differences between the groups. after propensity scorematching, most baseline characteristics became more comparable, although Hb levels remained lower inrecipients receiving ESa therapy (). during the 2-year follow-up, no acute rejection episodes wereidentified in this cohort. likewise, no cytomegalovirus or Epstein–Barr virus infection associated withanemia was identified during the study period. Trends in Hb and eGFRave during this study in the overall cohort and in the propensity score-matched cohort in the overall cohort and the propensity score matched cohort, trends in Hb and eGFRave during thisstudy are shown in. in the overall cohort, the proportion of patients with a Hb level of≥12 g/dl was 58.3% at baseline, 44.9% at 1 year, and 33.7% at 2 years. Primary longitudinal analysis using a linear mixed-effects model in the linear mixed-effects model using eGFRave, eGFRave declined over time (β = −1.86 per year; 95%Ci, −2.20 to −1.53; p < 0.001). Higher Hb levels were associated with higher eGFRave (β = 0.76 per 1g/dl;95% Ci, 0.44–1.09; p < 0.001), whereas higher baseline urine protein levels were associated with lowereGFRave (β = −2.49; 95% Ci, −3.60 to −1.38; p < 0.001). Baseline eGFRave was significantly associated with ### Page 5 subsequent eGFRave measurements (β = 0.93; 95% Ci, 0.90–0.96;p < 0.001). ESa use was not inde-pendently associated with eGFRave (β = 0.17; 95% Ci, −0.80 to 1.14;p = 0.731). Similarly, age and sexwere not independently associated with eGFRave (p = 0.811 and 0.335, respectively). in the linear mixed-effects model using eGFRcre, similar findings were observed. eGFRcre declinedover time (β = −1.56 per year; 95% Ci, −1.90 to −1.23; p < 0.001), and higher Hb levels were independently Table 1. Characteristics of this study. | non-eSa (n = 215) | eSa (n = 108) | | - | - | | Male sex, n (%) | 149 (69.3) | 63 (58.3) | | age, years | 51.2 ± 13.2 | 51.3 ± 13.4 | | Body mass index, kg/m2 | 23.3 ± 4.3 | 22.4 ± 5.3 | | Duration after KTx, months | 120 [60, 192] | 168 [116, 220] | | Cause of chronic kidney disease, n (%)Diabetic nephropathy | 8 (3.7) | 4 (3.7) | | Systolic blood pressure, mmHg | 128.2 ± 8.7 | 127.6 ± 13.9 | | laboratory examinations | | Hemoglobin, g/dl | 13.4 ± 1.2 | 12.0 ± 0.9 | | Serum creatinine, mg/dl | 1.30 ± 0.40 | 1.59 ± 0.61 | | Cystatin C, mg/l | 1.49 ± 0.41 | 1.87 ± 0.65 | | eGFRcre, ml/min/1.73 m2 | 47.5 ± 12.6 | 38.3 ± 12.7 | | eGFRcys, ml/min/1.73 m2 | 55.0 ± 16.6 | 40.6 ± 15.9 | | eGFRave, ml/min/1.73 m2 | 51.1 ± 14.0 | 39.3 ± 13.6 | | Transferrin saturation, % | 31.8 ± 11.6 | 40.2 ± 12.8 | | Bicarbonate, mmol/l | 27.5 ± 3.0 | 26.1 ± 4.0 | | intact parathyroid hormone, pg/ml | 57 [41, 80] | 63 [39, 95] | | Corrected calcium, mg/dl | 9.4 ± 0.4 | 9.4 ± 0.5 | | Phosphate, mg/dl | 3.0 ± 0.5 | 3.3 ± 0.6 | | urine protein, g/gCre | 0.11 [0.07, 0.22] | 0.18 [0.08, 0.38] | | immunosuppressive drugs | | Calcineurin inhibitor use, n (%) | 215 (100.0) | 108 (100.0) | | Mycophenolate mofetil use, n (%) | 207 (71.0) | 85 (68.0) | | everolimus use, n (%) | 52 (17.8) | 22 (17.6) | | antihypertensive drugs | | angiotensin ii receptor blocker use, n (%) | 169 (76.6) | 92 (73.6) | | Other antihypertensive drugs, n (%) | 135 (46.2) | 54 (43.2) | | Statin use, n (%) | 189 (64.7) | 82 (65.6) | | Dose of erythropoiesis stimulating agents (μg / month) | － | 50 [25, 75] | | non-eSa: not receiving eSa therapy, eSa: receiving eSa therapy | Table 2. Characteristics of the study after propensity score matching. | non-eSa (n = 73) | eSa (n = 73) | p Value | | - | - | - | | Male sex, n (%) | 46 (63.0) | 46 (63.0) | 1.000 | | age, years | 52.8 ± 13.5 | 50.5 ± 12.9 | 0.271 | | Body mass index, kg/m2 | 21.8 ± 3.1 | 22.5 ± 3.1 | 0.210 | | Duration after KTx, months | 170 [107, 240.5] | 180 [120, 220] | 0.819 | | Cause of chronic kidney disease, n (%)Diabetic nephropathy | 2 (2.7) | 2 (2.7) | 1.000 | | Systolic blood pressure, mmHg | 130.0 ± 9.3 | 127.8 ± 10.2 | 0.180 | | laboratory examinations | | Hemoglobin, g/dl | 13.2 ± 0.9 | 12.0 ± 0.8 | <0.001 | | Serum creatinine, mg/dl | 1.47 ± 0.43 | 1.43 ± 0.41 | 0.561 | | Cystatin C, mg/l | 1.63 ± 0.45 | 1.66 ± 0.46 | 0.692 | | eGFRave, ml/min/1.73 m2 | 43.7 ± 12.4 | 42.7 ± 11.5 | 0.614 | | Transferrin saturation, % | 38.3 ± 11.6 | 37.2 ± 10.0 | 0.524 | | Bicarbonate, mmol/l | 27.1 ± 3.4 | 26.4 ± 3.6 | 0.227 | | intact parathyroid hormone, pg/ml | 70 [39.5, 94] | 62 [36, 87.5] | 0.489 | | Corrected calcium, mg/dl | 9.6 ± 0.5 | 9.4 ± 0.5 | 0.501 | | Phosphate, mg/dl | 3.1 ± 0.6 | 3.2 ± 0.5 | 0.092 | | urine protein, g/gCre | 0.12 [0.08, 0.28] | 0.19 [0.08, 0.37] | 0.128 | | immunosuppressive drugs | | Calcineurin inhibitor use, n (%) | 73 (100.0) | 73 (100.0) | 1.000 | | Mycophenolate mofetil use, n (%) | 49 (67.1) | 51 (69.9) | 0.729 | | everolimus use, n (%) | 15 (20.6) | 15 (20.6) | 1.000 | | antihypertensive drugs | | angiotensin ii receptor blocker use, n (%) | 50 (68.5) | 54 (73.9) | 0.465 | | Other antihypertensive drugs, n (%) | 29 (39.7) | 35 (48.0) | 0.404 | | Statin use, n (%) | 56 (76.7) | 49 (67.1) | 0.269 | | non-eSa: not receiving eSa therapy, eSa: receiving eSa therapy | ### Page 6 associated with higher eGFRcre (β = 0.89 per 1g/dl; 95% Ci, 0.56–1.21; p < 0.001). Higher baseline urineprotein levels were also associated with lower eGFRcre (β = −2.20; 95% Ci, −3.30 to −1.10;p < 0.001),whereas ESa use was not independently associated with eGFRcre (β = 0.38; 95% Ci, −0.58 to 1.34;p = 0.439). age and sex were likewise not independently associated with eGFRcre (p = 0.653 and 0.972,respectively). Changes in kidney function during 2 years in the overall cohort and in the propensityscore-matched cohort in the overall cohort, changes in eGFRave (ΔeGFRave, ml/min/1.73m2) during the 2 years of follow-upwere −3.12 (95% Ci, −3.89 to −2.36) in patients not receiving ESa therapy treatment and −4.73 (95% Ci,−5.89 to −3.57) in those receiving ESa therapy, with a statistically significant difference (p = 0.021)(). By multivariate analysis, Hb was a significant factor influencing ΔeGFRave, whereas ESa usedid not have a significant impact on ΔeGFRave, as shown in. in the propensity score matchedcohort (n = 146), ΔeGFRave were − 3.0 (95% Ci, −4.1 to −1.93) and −5.66 (95% Ci, −7.1 to −4.3) (p = 0.003,with adjustment for baseline eGFRave), as shown in. in a supportive analysis restricted to recipients receiving ESa therapy, persistent anemia, defined asHb <12.0 g/dl at two or more of the three assessments during the 2-year observation period, was inde-pendently associated with a greater decline in eGFRave over 2 years (β = −3.26; 95% Ci, −6.13 to −0.38;p = 0.027). ![Two panels of line graphs comparing Hb levels and eGFR for Non-ESA and ESA groups over time. Panel \(a\) displays two line graphs: the top shows Hb levels \(g/dL\) over time for Non-ESA \(dashed line, stable near 13\) and ESA \(dotted line, declining\). The bottom shows eGFRave \(mL/min/m²\), with Non-ESA stable around 51 and ESA declining. Panel \(b\) presents similar graphs; the top reveals stable Hb levels close to 13 for Non-ESA and a slight drop for ESA. The bottom graph indicates a decline in eGFRave for both groups, with ESA showing a sharper decrease. Error bars are included.]() Figure 2. (a) Trends in Hemoglobin and eGFRave during this study pre-propensity score matching. The horizontalreference line in the hemoglobin panel indicates Hb 12 g/dl, which was used as a pragmatic threshold in the supportiveanalysis. error bars represent 95% confidence intervals Cis. (b) Trends in Hemoglobin and eGFRave during this studypost-propensity score matching. The horizontal reference line in the hemoglobin panel indicates Hb 12 g/dl, which wasused as a pragmatic threshold in the supportive analysis. error bars represent 95% confidence intervals (Cis). non-eSa:not receiving eSa therapy, eSa: receiving eSa therapy. ### Page 7 ![Bar chart comparing ?eGFRave \(mL/min/m²\) for Non-ESA and ESA groups, showing values around -3 and -4 with error bars. Bar chart displays ?eGFRave \(mL/min/m²\) for two groups: Non-ESA and ESA. Non-ESA averages around -3, while ESA is at approximately -4. Error bars represent variability, with P = 0.021 indicating statistical significance.]() Figure 3. Changes in eGFRave during 2 years. non-eSa: not receiving eSa therapy, eSa: receiving eSa therapy. ![Bar chart comparing eGFR changes for Non-ESA and ESA groups, with P = 0.003. Bar chart comparing ?eGFRave \(mL/min/m²\) for Non-ESA and ESA groups. Non-ESA has a mean around -3, and ESA is lower, nearly -6. Error bars show variability, and P-value of 0.003 is noted below the bars.]() Figure 4. Changes in eGFRave during 2 years post-propensity score matching. non-eSa: not receiving eSa therapy, eSa:receiving eSa therapy. Table 3. associations of clinical parameters at baseline with ΔeGFRave. | univariate | Multivariate | | - | - | | β | t | p | β | t | p | | age | 0.03 | 0.49 | 0.628 | 0.01 | 0.17 | 0.866 | | Sex, female | 0.11 | 2.01 | 0.045 | 0.09 | 1.66 | 0.098 | | Body mass index | 0.09 | 1.51 | 0.131 | 0.11 | 1.84 | 0.066 | | Duration after KTx | −0.08 | −1.39 | 0.165 | −0.05 | −0.85 | 0.395 | | Original Disease, DKD | −0.05 | −0.23 | 0.504 | −0.01 | −0.23 | 0.821 | | Systolic blood pressure | −0.03 | −0.51 | 0.613 | −0.05 | −0.83 | 0.409 | | Hemoglobin, baseline | 0.17 | 3.02 | 0.003 | 0.16 | 2.28 | 0.023 | | eGFRave, baseline | − 0.08 | −1.31 | 0.19 | −0.27 | −3.99 | <0.001 | | Bicarbonate | 0.06 | 1.07 | 0.287 | 0.12 | 1.9 | 0.059 | | PTH | −0.004 | −0.08 | 0.938 | 0.02 | 0.41 | 0.683 | | Transferrin saturation | 0.03 | 0.58 | 0.564 | 0.03 | 0.43 | 0.666 | | eSa use | −0.13 | −2.33 | 0.021 | −0.07 | −0.96 | 0.338 | | urinary protein | −0.25 | −4.59 | <0.001 | −0.29 | −5.13 | <0.001 | ### Page 8 Discussion in the present real-world outpatient cohort of kidney transplant recipients ≥1 year after transplantation,the primary analysis using a linear mixed-effects model showed that higher time-updated Hb levels wereindependently associated with better longitudinal graft function, whereas ESa use itself was not inde-pendently associated with graft function trajectory. this association was observed after adjustment fortime, baseline eGFR, age, sex, and baseline urine protein. in supportive analyses based on 2-year changein eGFR, recipients with lower Hb levels showed greater graft function decline in both the overall cohortand the propensity score-matched cohort. Furthermore, in a supportive analysis restricted to recipientsreceiving ESa therapy, persistent anemia was independently associated with a greater decline in eGFRaveover 2 years. taken together, these findings suggest that the clinically relevant phenotype is not ESaexposure itself, but persistent Pta, including persistent anemia despite ESa therapy, which may helpidentify recipients at higher risk of subsequent graft dysfunction in routine clinical practice. in this con-text, Hb likely reflects underlying graft pathology, inflammation, or ESa resistance rather than beingdirectly causal. Several RCts have evaluated Hb targets in Ktx recipients receiving ESa therapy, revealing that higherHb targets are associated with improved renal functional outcomes in some settings [,,], althoughoptimal targets remain controversial. Based on this evidence, our center adopted a relatively higher Hbtarget in routine practice. However, in real-world outpatient care, achieving and maintaining such higherHb levels proved challenging, and, on average, Hb levels remained lower in recipients receiving ESatherapy than in those not receiving ESa therapy. therefore, this study cohort was not designed to eval-uate the safety or efficacy of achieving a higher Hb target (e.g. ≥13 g/dl) through ESa therapy. Rather,our findings should be interpreted as associative, highlighting that persistent Pta despite ESa therapymay indicate a high-risk clinical phenotype associated with graft dysfunction. the association between ESa use and graft function should be interpreted cautiously in our observa-tional setting. ESa therapy is initiated and continued based on anemia severity; therefore, indication biasand residual confounding may have persisted despite propensity score matching and multivariableadjustment. unmeasured factors, such as inflammation, iron-restricted erythropoiesis, ESa hyporespon-siveness/bone marrow resistance, and underlying allograft pathology, may contribute to both persistentanemia and graft dysfunction []. importantly, Hb levels remained lower in recipients receiving ESa ther-apy even after propensity score matching, indicating that the matched comparison showed a differencebetween non-anemic recipients and those with persistent anemia despite ESa therapy. accordingly, ourfindings do not establish a direct, causal detrimental effect of ESa but instead support an association,underscoring the clinical importance of identifying and characterizing persistent Pta in routine outpa-tient follow-up. Prior trials evaluated prespecified Hb targets under controlled protocols [,,], whereas our cohortreflects routine outpatient practice in long-term Ktx recipients. thus, our results complement the trialevidence by highlighting the prognostic relevance of persistent anemia in real-world care. althougheGFRave is not widely used internationally, we selected it a priori based on our previous validation studyin Japanese kidney transplant recipients, in which it better reflected kidney function than creatinine-basedeGFR alone. nevertheless, generalizability to non-Japanese populations and other clinical settings remainsuncertain. From a practical standpoint, our findings indicate that persistent Pta, which often requires ESa ther-apy, may help identify recipients at higher risk of subsequent graft function decline during long-termoutpatient follow-up. these results support closer monitoring of graft function and systematic evaluationof potentially modifiable contributors to anemia, such as iron status and the inflammatory milieu, whileacknowledging that specific treatment strategies or Hb targets cannot be determined from this observa-tional study. accordingly, these findings should be interpreted as hypothesis-generating rather thanpractice-changing, because clinically important hard renal outcomes were infrequent during the 2-yearobservation period. the observed association between lower Hb levels and graft dysfunction may reflect several overlap-ping pathways. Persistent anemia may reflect renal oxygen delivery and tissue hypoxia and can beaccompanied by hemodynamic alterations; it may also serve as a surrogate marker of underlying ### Page 9 processes that impair both erythropoiesis and graft function, such as chronic inflammation, iron-restrictederythropoiesis, and chronic allograft injury [,]. Because histopathology and comprehensivetime-updated inflammatory assessments were not systematically available, the role of these mechanismsin our findings remains speculative, warranting further investigation. Because Hb may also lie on thepathway linking declining graft function, inflammation, and ESa treatment, the possibility oftime-dependent confounding and reverse causation should be acknowledged. in particular, worseninggraft function may itself contribute to subsequent anemia. although the present longitudinal model wasdesigned to better account for time-varying Hb levels, future studies using lagged analyses or marginalstructural models may help further clarify the temporal and causal structure of this association. this study has several limitations. First, this was a single-center retrospective study, which may limit thegeneralizability of our findings. Second, the sample size was modest. third, Hb targets were not fully stan-dardized and might have differed among the attending physicians. Fourth, residual confounding cannot beexcluded, particularly because clinically important time-updated variables, such as ferritin, inflammatorymarkers, and histopathological findings, were not systematically available. Fifth, clinically important hardrenal outcomes, such as graft failure or marked eGFR decline, were infrequent during the 2-year observationperiod, limiting event-based analyses. Sixth, the supportive analyses, including the propensity score-matchedand ESa-restricted analyses, should be interpreted cautiously as complementary evidence to the main lon-gitudinal findings. Finally, ESa exposure was modeled primarily as treatment status rather than as a quanti-tative variable; therefore, cumulative dose, dose intensity, and responsiveness to ESa were not fully evaluated. Conclusion Persistent Pta, rather than ESa use itself, was associated with graft function decline in kidney transplantrecipients. these findings suggest that Hb likely reflects underlying graft pathology, inflammation, or ESaresistance rather than being directly causal. Acknowledgments We sincerely thank all the staff at Masuko Kinen Hospital for their cooperation with this study. Author contributions CRedit: Makoto Tsujita: data curation, Formal analysis, investigation, Writing – original draft; Yutaka Matsuoka:investigation, Project administration, Writing – review & editing; Kazuharu Uchida: investigation, Project administra-tion, Supervision, Writing – review & editing; Kunio Morozumi: Conceptualization, Project administration, Supervision,Writing – review & editing. Disclosure statement the authors have declared that there is no conflict of interest. Funding statement none. Human and animal rights this study was performed per the declaration of Helsinki, and the Ethics Committee approved the research protocolof the Masuko Kinen Byoin Research institute (approval no. MR6-12). Because of a retrospective cohort study,informed consent from individual participants was waived. Data availability statement the data supporting the findings of this study are available from the corresponding author upon reasonable request.the data are not publicly available due to privacy and ethical restrictions, in accordance with the taylor & FrancisShare upon Reasonable Request data policy. ### Page 10 References [] Kawada n, Moriyama t, ichimaru n, et al. negative effects of anemia on quality of life and its improvement bycomplete correction of anemia by administration of recombinant human erythropoietin in posttransplant pa-tients. Clin Exp nephrol. 2009;13(4):355–360. doi:. [] Bonomini M, di liberato l, Sirolli V. treatment options for anemia in kidney transplant patients: a review.Kidney Med. 2023;5(8):100681. doi:. [] yamamoto H, nishi S, tomo t, et al. 2015 Japanese Society for dialysis therapy: guidelines for renal anemia inchronic kidney disease. Ren Replace ther. 2017;3(1):1–46. doi:. [] Van Biesen W, Vanholder R, Veys n, et al. Efficacy of erythropoietin administration in the treatment of anemiaimmediately after renal transplantation. transplantation. 2005;79(3):367–368. doi:. [] Pile t, Raftery M, thuraisingham R, et al. treating posttransplant anemia with erythropoietin improves qualityof life but does not affect progression of chronic kidney disease. Exp Clin transplant. 2020;18(1):27–33. doi:. [] Heinze G, Kainz a, Hörl WH, et al. Mortality in renal transplant recipients given erythropoietins to increasehaemoglobin concentration: cohort study. BMJ. 2009;339(1):b4018–b4018. doi:. [] tsujita M, Kosugi t, Goto n, et al. the effect of maintaining high hemoglobin levels on long-term kidney func-tion in kidney transplant recipients: a randomized controlled trial. nephrol dial transplant. 2019;34(8):1409–1416. doi:. [] Choukroun G, Kamar n, dussol B, et al. Correction of postkidney transplant anemia reduces progression of al-lograft nephropathy. J am Soc nephrol. 2012;23(2):360–368. doi:. [] Matsuo S, imai E, Horio M, et al. Revised equations for estimated GFR from serum creatinine in Japan. am JKidney dis. 2009;53(6):982–992. doi:. [] tsujita M, Goto n, yamamoto t, et al. How to estimate kidney function in kidney transplant recipients with mildto moderate kidney impairment: comparison of estimated glomerular filtration (eGFR) values betweencreatinine-based GFR equations and cystatin C-based GFR equations for Japanese population. Clin Exp nephrol.2014;18(1):130–134. doi:. [] Obi y, ichimaru n, Sakaguchi y, et al. Correcting anemia and native vitamin d supplementation in kidney trans-plant recipients: a multicenter, 2 x 2 factorial, open-label, randomized clinical trial. transpl int. 2021;34(7):1212–1225. doi:. [] Camaschella C, nai a, Silvestri l. iron metabolism and iron disorders revisited in the hepcidin era. Haematologica.2020;105(2):260–272. doi:. [] Fine lG, norman Jt. Chronic hypoxia as a mechanism of progression of chronic kidney diseases: from hypoth-esis to novel therapeutics. Kidney int. 2008;74(7):867–872. doi:. [] Wang B, li Zl, Zhang yl, et al. Hypoxia and chronic kidney disease. EBioMedicine. 2022;77:103942. doi:.