Post-transplant diabetes and prediabetes are common occurrences following kidney transplantation, significantly increasing the risk of cardiovascular events and mortality.1–4 Identifying at-risk patients prior to transplantation is essential as it allows for informed risk counseling, targeted interventions to address modifiable risk factors, and the customization of immunosuppressive therapy without compromising efficacy.1,2

In patients with end-stage renal disease, fasting glucose is less reliable than in the general population due to reduced renal gluconeogenesis, and HbA1c levels may be falsely low.2,5 Oral glucose tolerance test (OGTT)-based studies performed during pretransplant workup6 have reported a prevalence of undiagnosed diabetes mellitus (uDM) between 3% and 8%, and impaired glucose tolerance (IGT) ranging from 20% to 30%,7,9–12 both associated with higher post-transplant glucose metabolic alterations (GMA).7–10 Therefore, the International Consensus on Post-transplant Diabetes2 and the 2020 KDIGO Guidelines5 recommend OGTT screening during pretransplant evaluation. Nevertheless, a recent survey by the ERA Descartes Group revealed significant variability in screening, prevention, and treatment of GMA among transplant centers, with only 13% routinely performing the recommended OGTT prior to transplantation.13

Limited research has investigated the phenotypic characterization of glucose metabolism in non-diabetic transplant candidates. Compared to those with normal glucose tolerance, patients with GMA frequently demonstrate reduced insulin sensitivity and impaired β-cell function.11 Moreover, post-transplant GMA are often observed in individuals who already exhibit inefficient insulin secretion while on the transplant waiting list (WL).7 These findings highlight a potential window for pre-transplant interventions aimed at mitigating both pre- and post-transplant metabolic risk.

Few studies have explored how clinical data might guide OGTT screening in transplant candidates. A Norwegian study found that OGTT in patients with fasting glucose 92–125mg/dL identified 90% of diabetes cases,12 while an Australian study showed poor predictive value of fasting glucose alone (AUC 0.66).10 Moreover, no data currently inform how to monitor glucose abnormalities over time in waitlisted patients.13

The objectives of this study, conducted in a cohort of patients without established diabetes and undergoing OGTT during kidney transplant evaluation, were as follows: (a) to determine the prevalence and metabolic phenotypes of GMA; (b) to identify clinical and biochemical predictors of abnormal OGTT results and assess optimal screening strategies; and (c) to evaluate the one-year trajectory of glucose metabolism abnormalities through repeated OGTT assessments.

Of the total 199 patients, 182 were included, of whom 46 had a repeated OGTT after one year in the WL (Fig. 1).

Fig. 1.

Patient's disposition. WL: waiting list; RT: renal transplantation; OGTT: oral glucose tolerance test.

As shown in Fig. 2, one-third of the patients presented with some form of GMA at study entry, the majority of whom had IGT (23.1%), followed by uDM (6%), and isolated IFG (4.4%). Among the 42 patients with IGT, 7 (16.7%) also exhibited IFG.

Fig. 2.

Prevalence of glucose metabolism disorders in the kidney transplant waiting list. N: normal; IFG: impaired fasting glucose; IGT: impaired glucose tolerance; uDM: unknown diabetes.

Baseline characteristics are presented in Table 1. Chronic hemodialysis was the predominant modality of renal replacement therapy across all groups; however, patients with uDM tended to receive peritoneal dialysis more frequently. The use of statins was significantly higher among patients with any GMA compared with those with normal glucose tolerance.

Metabolic phenotyping of GMA

OGTT-derived parameters of glucose metabolism are shown in Table 2 and Fig. 3. Isolated IFG was mainly associated with reduced insulin sensitivity (Fig 3A, B), whereas IGT and uDM were characterized by impaired insulin secretion (Fig. 3C). Despite these differences, all GMA categories exhibited a significant reduction in the disposition index (Fig. 3D), indicating and overall imbalance between insulin secretion and sensitivity. HbA1c values were comparable among groups. BMI correlated inversely with insulin sensitivity (ρ=−0.49 for Matsuda index; P<0.001) but not with insulin secretion (ρ=0.15; P=0.053).

Table 2.

Parameters of glucose metabolism in the fasting state and following an OGTT across different glucose metabolic alterations groups.

	Normal (n=121)	IFG (n=8)	IGT (n=42)	uDM (n=11)	Total (n=182)	P value
HbA1c (%)	5 (4.9–5.2)	5.4 (5.1–5.6)	5.1 (4.9–5.4)	5.3 (4.8–6)	5.1 (4.9–5.3)	0.087
Glucose 0 (mg/dL)	82 (76–88)	103 (100–111.3)	85 (78.8–93.3)	89 (79–110)	83 (77–90)	<0.001a; 0.053c; 0.003d
Glucose 30 (mg/dL)	132 (116.5–149)	174 (152.8–193)	152 (134.8 –173.8)	171 (161–195)	139 (121–162)	0.001a,b; <0.001c
Glucose 120 (mg/dL)	114 (93.5–128)	119 (98.5–132.75)	157 (144.8–170.5)	249 (221–272)	126 (103–144)	<0.001b,c,e; 0.001d
Insulin 0 (μU/mL)	7.4 (5.3–7.4)	14 (12.3–23)	7.4 (5.3–11.2)	7.1 (5.7–10.3)	7.6 (5.3–11.4)	0.002a; 0.004d; 0.018e
Insulin 30 (μU/mL)	42.6 (26.2–62.2)	75.8 (43.7–100)	36.4 (24.1–58.3)	22.3 (12.5–42.7)	41.4 (24.9–61.6)	0.045d; 0.002e
Insulin 120 (μU/mL)	33.5 (21.9–54.5)	83.6 (34.1–138.8)	44.3 (34.5–92.5)	49.3 (24.8–84.9)	38.7 (25.2–61.4)	0.034a; 0.007b
HOMA-IR	1.4 (1–2.1)	3.4 (3.2–6.2)	1.5 (1–2.3)	1.7 (1.1–2.8)	1.5 (1–2.3)	<0.001a; 0.003d; 0.028e
Insulinogenic Index	12.1 (7.2–21.9)	15 (10.3–25.6)	7.4 (4.8–13)	3.4 (2.2–7.3)	10.2 (6.4–17.8)	<0.001c; 0.012b; 0.002e
Matsuda Insulin Sensitivity Index	17.6 (12.6 –25.7)	7.2 (6.8–9)	16.1 (10.1–22)	12.6 (10.1–23.5)	16.4 (11.4–24.1)	<0.001a; 0.012d
Disposition Index	223.7 (141.8–320.9)	119.4 (95.4–137.5)	124.9 (78.2–175.7)	46.3 (32.7–79.2)	168.4 (110.2–268)	0.018a; <0.001b,c; 0.043f

IFG: impaired fasting glucose; IGT: impaired glucose tolerance; uDM: unknown diabetes. Glucose 0, Gluc 30, Glucose 120, Insulin 0, Insulin 30 and Insulin 120: fasting, and 30- and 120-min plasma glucose and insulin levels after an OGTT: oral glucose tolerance tests.

a

N versus IFG.

b

N versus IGT.

c

N versus uDM.

d

IFG versus IGT.

e

IFG versus uDM.

f

IGT versus uDM.

Fig. 3.

Metabolic phenotype of different glucose metabolic disorders. IFG: impaired fasting glucose; IGT: impaired glucose tolerance; uDM: unknown diabetes.

Changes in glucose metabolism alterations over one year in the WL

The progression of glucose metabolism abnormalities (GMA) over one year was assessed using a repeat OGTT (Fig. 4). Although the overall distribution of GMA categories at one year showed minimal change, 17 patients (37%) exhibited transitions: 10 experienced deterioration, whereas 7 showed improvement. None of the patients with normal glucose tolerance progressed to overt diabetes (uDM); however, 8 developed a prediabetic state (7 IGT and 1 isolated IFG). Those who progressed (n=8) had a significantly higher baseline BMI (27.5kg/m2 [IQR 25.6–31.7] vs. 24.1kg/m2 [IQR 21.3–25.8]; P=0.04), lower insulin sensitivity (Matsuda index 15.7 [IQR 11.4–24.5] vs. 22.9 [IQR 15.5–37]; P=0.049), and a higher proportion were treated with statins (75% vs. 32%; P=0.047) compared to those who remained with normal glucose metabolism (n=17).

Fig. 4.

Evolution glucose metabolism abnormalities (GMA) after one year on the kidney transplant waiting list. Those who progressed are indicated with red arrows, whereas those who improved are marked with green arrows. N; normal; IFG: impaired fasting glucose; IGT: impaired glucose tolerance, uDM: undiagnosed diabetes mellitus.

The majority of patients with isolated IFG or IGT showed reversion to normal tolerance at one year. There were no significant differences in any of the evaluated risk parameters between patients who reverted (n=7) and those who did not (n=5). Only one patient progressed to uDM, which occurred in an individual with baseline IGT.

In this cohort of non-diabetic patients on the kidney transplant WL, we found that approximately one in three patients had some form of GMA (Fig. 1). Among these, only 4% exhibited IFG, while the remaining 29% showed abnormalities on the OGTT: 23% had IGT and 6% met criteria for uDM. These findings are consistent with previous studies in this population, falling within the reported prevalence range.9–12 Importantly, the presence of IGT or uDM during pre-transplant evaluation is linked to a higher risk of developing post-transplant GMA.7–10 Given their potential reversibility, these results emphasize the importance of initiating preventive measures against post-transplant GMA prior to transplantation.

We employed OGTT with plasma insulin level determination for more accurate individual assessment of insulin sensitivity and secretion compared to single-point fasting values.11 These indices help delineate pathomechanisms of disturbed glucose metabolism useful for interventions and risk reduction. Patients with IGT or uDM exhibited a similar phenotype characterized by significantly lower insulin secretion compared to those with normal tolerance (Fig. 3C). Notably, urea and other uremic toxins decrease insulin secretion by increasing oxidative stress in beta cells.17 Patients with isolated IFG demonstrated normal insulin secretion (Fig. 3C) with markedly lower insulin sensitivity (Fig. 3A and B), consistent with insulin resistance at the hepatic level leading to elevated fasting plasma glucose.18 Lastly, all groups with GMA showed a significant decrease in the disposition index, indicating inefficient insulin secretion relative to insulin resistance which is common in the uremic state.19 Therefore, interventions enhancing insulin sensitivity, such as therapeutic exercise20 and weight control, may help prevent post-transplant GMA.

Owing to the limited diagnostic sensitivity of fasting glucose and HbA1c, the OGTT is recommended as a screening tool during pre-transplant evaluation.2,5 However, its implementation increases logistical complexity and resource utilization, and in a European survey was routinely performed in only 13% of centers.13 Thus, studies focusing on clinical and laboratory data that can potentially obviate the need for routine OGTT would streamline pre-transplant assessments. Patients with abnormal OGTT results (IGT or uDM) tended to be older, exhibited higher fasting plasma glucose levels, and were more frequently treated with statins (Table 3). Notably, in multivariate analysis, statin therapy emerged as the only variable significantly associated with an increased risk of abnormal OGTT (Table 4). The potential diabetogenic effect of statins, though outweighed by their cardiovascular benefits, has been extensively investigated. Two recent meta-analyses reported an increased risk of developing diabetes of 9–12% in randomized clinical trials21 and up to 44% in observational studies.22 In a prospective study, statin therapy was associated with a 46% increased risk of type 2 diabetes, attributed to reduced insulin sensitivity and impaired insulin secretion.23 In murine models, β-cell-specific deletion of HMG-CoA reductase led to decreased β-cell mass and impaired insulin secretion.24 However, in the present study, differences in age and BMI between statin users and non-users (Supplementary Table 1) limit the ability to draw definitive conclusions about the independent diabetogenic effect of statins. Given that statin therapy independently increases the risk of post-transplant diabetes,25 routine OGTT screening may be warranted in statin-treated patients awaiting transplantation to enable timely non-pharmacological and/or pharmacological interventions. In all cases, patients identified with GMA should be advised to initiate non-pharmacological interventions, including dietary modification and increased physical activity.

Excluding statin therapy from the logistic regression model, age and fasting glucose remained significantly associated with abnormal OGTT (Table 4, Model 2), showing modest diagnostic accuracy in the ROC curve analysis (Table 5). However, patients who lacked all three risk factors (age and fasting glucose below optimal cut-offs, and no statin therapy) had substantially lower odds of abnormal OGTT (0.13; 95% CI: 0.05–0.4). This risk-stratified approach could potentially reduce the need for routine OGTT in 26% of waitlisted patients in our study.

According to ERA Descartes survey data, 57% of centers repeat some screening method annually, and there is little data on the longitudinal course of glucose metabolism in waitlisted patients.2,13 In our study, only 22% showed worsening after one year, mainly to prediabetes (Fig. 4). Notably, worsening among patients initially with normal metabolism was associated with BMI, reduced insulin sensitivity, and statin use. These findings suggest that annual OGTT screening may be justified in patients with an initially normal OGTT who are overweight or obese, or on statin therapy. Conversely, most patients with prediabetes (IFG or IGT) reverted to normal glucose tolerance, without any accompanying differences in the study's risk factors. Given the small sample size (n=12; Fig. 4), additional studies are warranted in a larger population.

This study has limitations, including its cross-sectional design, which precludes causal inference between clinical predictors and abnormal OGTT. Moreover, the high transplantation rate limited follow-up OGTT availability, warranting cautious interpretation. Additionally, the reproducibility of OGTT is limited, as repeated test in the same individual may yield discordant results.26 Lastly, the predominantly Caucasian patient cohort limits generalizability to other racial groups.

In conclusion, one-third of waitlisted patients without diabetes exhibit abnormal OGTT results. Age, fasting glucose levels, and statin therapy were identified as independent risk factors. Patients younger than the optimal discriminatory age threshold (51 years), with fasting glucose levels below 100mg/dL, and not receiving statin therapy (representing one in four patients) have a low likelihood of abnormal OGTT results, and routine testing may therefore be unnecessary. Conversely, annual OGTT screening may be warranted in patients with an initially normal OGTT particularly if they are overweight, obese, or undergoing statin therapy.

E.T-G., A.T., and E.P. designed the study; E.T.-G., A.T., L.P.-T., and E.P. revised and approved the analytical protocol; E.T.-G., L.P.-T., C.R.R.-L., A.F.- H., E.P.-C., A.M.d.V.G., A.G.-D, D.M.-M., C.A.-S., A.E.R.-R., A.M.R.-G., M.M., S.G.-R., A.A.-G., A.G.-R., C.R.-A., A.P.R.-H., E.d.B.-R. and D.H.-M. collected the data; E.T-G., A.T., and E.P. analyzed the data; E.T.-G. and A.T. made the tables and figures; E.T.-G. and A.T. wrote the first draft of the manuscript; E.P. and D.H.-M. revised and amended the manuscript; all authors approved the final version of the manuscript.

This study was supported by grants from Instituto de Salud Carlos III (PI16-02151, PI19/01187, ICI21/00042, RD16/0009/0031 and RICORS2040 RD24/0004/0025).

The authors declare that no financial or personal relationships with other people or organizations influenced the work reported in this manuscript. A.T. received support for attending the 2023 Spanish Society of Nephrology Congress and honoraria for educational lectures from Alnylam and Chiesi. All other authors declare no conflicts of interest.

The data that support the tables and plots within this manuscript and other findings of this study are available from the corresponding author upon reasonable request.