Kidney function and glycemic control in solid-organ transplant recipients with pretransplant versus post-transplant diabetes: a cross-sectional study of a predominantly non-kidney transplant cohort
Diabetes contributes substantially to morbidity after solid-organ transplantation, but comparisons of kidney function and glycemic control between pretransplant diabetes mellitus (PDM) and post-transplant diabetes mellitus (PTDM) remain limited, particularly in cohorts in which most recipients have received a non-kidney organ. This single-center, retrospective, cross-sectional study included 113 solid-organ transplant recipients with diabetes, each assessed at a single outpatient visit (index visit) between May 2017 and May 2026. Kidney function, albuminuria, glycemic control and treatment patterns were described at that visit. The comparison between PDM and PTDM was the primary analysis; comparisons of early- versus late-onset PTDM and of sodium-glucose cotransporter 2 (SGLT2) inhibitor users versus non-users were exploratory. Sequential multivariable linear models for estimated glomerular filtration rate (eGFR), log10(urinary albumin-to-creatinine ratio [UACR] + 1) and glycated hemoglobin (HbA1c) were specified from an explicit causal framework (Model 1, unadjusted; Model 2, adjusted for age, sex, transplanted organ and time since transplantation; Model 3, additionally adjusted for body mass index, glucocorticoid and calcineurin-inhibitor use). Analyses were repeated after excluding kidney transplant recipients and stratified by transplanted organ. P values were adjusted within each table by the Holm and Benjamini–Hochberg procedures. The cohort included liver (51.3%), kidney (36.3%) and heart (12.4%) transplant recipients; 61.6% had PTDM. Median time since transplantation was 9 [5–16] years, mean eGFR was 60.77 ± 28.39 mL/min/1.73 m² and mean HbA1c was 7.40 ± 1.84%; 49.5% had an eGFR below 60 mL/min/1.73 m² and 53.2% an HbA1c above 7% at the index visit. Compared with PTDM, patients with PDM had a shorter time since transplantation (6.0 [2.0–11.5] vs. 11.0 [7.0–19.0] years; P < 0.001; Holm-adjusted P < 0.001), were more often in eGFR categories G3–G5 (28 (68.3) vs. 25 (37.3); Holm-adjusted P = 0.048), were nominally older (median 64.0 [56.5–69.5] vs. 60.0 [52.0–65.0] years; P = 0.037; Holm-adjusted P = 0.887) and had a lower mean eGFR (53.06 ± 21.82 vs. 65.78 ± 31.02 mL/min/1.73 m²; P = 0.014; Holm-adjusted P = 0.369), whereas no statistically significant difference in HbA1c was detected ( P = 0.978). In the primary adjusted model (Model 2) eGFR remained 12.7 mL/min/1.73 m² higher in PTDM (95% CI 1.2 to 24.1; P = 0.030), and was attenuated to 7.8 mL/min/1.73 m² (95% CI − 4.9 to 20.5; P = 0.223) in Model 3. After exclusion of kidney transplant recipients the unadjusted difference was 19.2 mL/min/1.73 m² (95% CI 6.6 to 31.7; P = 0.003) and the Model 2 difference 10.5 mL/min/1.73 m² (95% CI − 3.1 to 24.0; P = 0.128). Albuminuria (UACR ≥ 30 mg/g) was present in 45.9% of the 98 patients with an available measurement. No statistically significant difference in UACR between diabetes groups remained after adjustment for multiple comparisons (median 19.5 [4.8–52.4] vs. 28.1 [11.0–223.3] mg/g; unadjusted P = 0.037; Holm-adjusted P = 0.887). In an exploratory model, LDL cholesterol was associated with UACR (1.20-fold per 10 mg/dL; 95% CI 1.10 to 1.31; P < 0.001). No statistically significant differences in kidney function or glycemic control were detected between early- and late-onset PTDM, although confidence intervals were wide. 40% of the patients with albuminuria were receiving an SGLT2 inhibitor. Solid-organ transplant recipients with diabetes evaluated a median of 9 years after transplantation showed a high prevalence of reduced eGFR, albuminuria and HbA1c above target at a single assessment. Reduced eGFR was frequently non-albuminuric, whereas albuminuria identified a distinct subgroup at high KDIGO risk, of whom only 40% were receiving an SGLT2 inhibitor. The lower eGFR in PDM was not explained by age, sex, transplanted organ or time since transplantation, was attenuated after further adjustment for BMI and current immunosuppressive exposure and after exclusion of kidney recipients, and is additionally subject to survivorship and prevalent-user selection; the differences should therefore be interpreted as describing two differently selected populations rather than an effect of diabetes type.
Authors
- Humberto Batista Ferreira (ORCID: https://orcid.org/0000-0002-6435-1606)
- Amanda Gonçalves Duarte
- Márcio Weissheimer Lauria (ORCID: https://orcid.org/0000-0001-7558-1231)
- YUITI PEDRO YAMASHITA
- ESTEVAO FARIAS
- LARISSA BERNARDES IZZA (ORCID: https://orcid.org/0009-0006-2369-2822)
- Ithallo Ricardo Ensser
Institutions
- Universidade Federal de Minas Gerais (BR)
- Hospital das Clínicas da Universidade Federal de Minas Gerais (BR)
Publication Details
- Journal
- Diabetology & Metabolic Syndrome
- Published
- 2026-11-21
- DOI
- https://doi.org/10.1186/s13098-026-02359-x
- Primary Topic
- Renal Transplantation Outcomes and Treatments
- Type
- article
- Field-Weighted Citation Impact
- 0.00