Abstract
Introduction
Recent studies have shown possible protective effects of sodium-glucose co-transporter-2 inhibitors (SGLT-2is) for acute kidney injury (AKI). As SGLT-2is have diuretic properties, uncertainty remains regarding their effect on the risk of AKI during hospital admission, where hypovolemia frequently occurs. Therefore, we assessed the association between (pre)hospital treatment with SGLT-2i and AKI risk in hospitalized patients.
Methods
We performed a matched cohort study. We identified hospitalized patients with prehospital use of SGLT-2i in the Frisius Medical Centre Leeuwarden, between October 21, 2021, and February 12, 2024. These were matched 1:1 to hospitalized patients without prehospital use of SGLT-2i. AKI incidence was the main outcome. Second, SGLT-2i treatment during hospitalization was assessed.
Results
Overall,1,054 hospital admissions were included, 527 with and 527 without prehospital SGLT-2i use. Overall, 34% of patients were women; the median age was 71 years (IQR 65–78). SGLT-2i was indicated for heart failure in 16%, chronic kidney disease 6%, type 2 diabetes 7%, and at least two indications in 72%. Hospital-acquired AKIs occurred in 17.5% admissions with prehospital SGLT-2i treatment and 25.8% among those without (RR = 0.68 [95% CI = 0.53–0.86]). AKI risk was similarly decreased in patients continuing SGLT-2i during hospital admission (RR 0.65 [0.49–0.85]). AKI was associated with increased length of hospital stay (median 7 days, IQR 4–11, versus 3 days, IQR 1–7 without AKI); SGLT-2i use was not. Overall, SGLT-2i treatment did not affect overall mortality. In patients with AKI, mortality was numerically lower in SGLT-2i patients (risk difference = 7.8%, 95% CI –0.9–15.7%), although the confidence interval includes zero.
Conclusion
We observed a 32% lower risk of AKI in hospitalized patients using SGLT-2i at admission. This suggest a protective effect of SGLT-2i against AKI, which remained when SGLT-2i was continued during hospitalization.
Keywords: Acute kidney injury, SGLT-2i
What Is Known
SGLT-2 inhibitors are associated with a reduced risk of acute kidney injury in select patient populations.
What This Study Adds
SGLT-2 inhibitors are associated with a reduced risk of acute kidney injury during hospitalization in a matched cohort.
SGLT-2 inhibitors are associated with a reduced risk of more severe AKI.
Patients who developed AKI experienced a prolonged hospital stay, underscoring the clinical impact of AKI complications.
Introduction
Acute kidney injury (AKI) is one of the most common in-hospital complications [1], affecting approximately one in five adults during hospitalization. It is associated with an increased risk of developing chronic kidney disease (CKD) or progression of pre-existing CKD, particularly when AKI is more severe or prolonged [2, 3]. Preventive strategies and early recognition of AKI are essential as they may improve patient outcomes and reduce long-term renal complications [4].
Sodium-glucose co-transporter-2 inhibitors (SGLT-2is) inhibit glucose and sodium reabsorption in the proximal tubule, leading to increased urinary excretion of both and resulting in osmotic diuresis. SGLT-2is reduce risk of cardiovascular events in patients with and without diabetes [5, 6]. Additionally, they lower the risk of progression of renal disease, hospitalization, and mortality due to heart failure [7].
Recent studies have demonstrated SGLT-2i’s protective effects against AKI risk [8–12]. This is noteworthy, given that SGLT-2i reduce glomerular perfusion through their diuretic effect – a mechanism previously associated with an increased risk of AKI [13]. A meta-analysis of ten studies involving patients with type 2 diabetes mellitus (T2DM) treated with SGLT-2i reported a pooled odds ratio of 0.50 for developing AKI [12]. However, these studies were highly heterogeneous in regards to patients, treatment, and outcomes. To date, RCTs with AKI as pre-specified outcome have not been published. Moreover, most existing studies investigate AKI risk in the outpatient setting, while AKI risk is highest during hospital admission. This raises the question whether SGLT-2i reduce incidence of AKI during hospitalization where the risk of hypovolemia – and consequently AKI – is increased. To address this knowledge gap, we aimed to investigate the association between treatment with SGLT-2i at the time of and during hospitalization and AKI risk in a real-world patient population. Second, we assessed the impact of SGLT-2i treatment on the incidence of AKI within SGLT-2i indication and admission specialty groups, incidence of AKI stages, length of hospital stay, and in-hospital mortality.
Methods
Study Design and Patients
We conducted a retrospective, single-centre, matched cohort study at the Frisius Medical Centre in the Netherlands. Data were extracted from electronic health records using a predefined query to retrieve relevant demographic, clinical, and laboratory information. Ethical approval and the need for informed consent were waived by the local board of the Frisius Medical Centre given the retrospective nature of the study.
All hospital admissions of adult patients who did not opt out of having their medical records used for research between October 21 2021, and February 12, 2024, were eligible for inclusion in the study. Inclusion criteria were discharge from a clinical ward and a documented indication for SGLT-2i. Exclusion criteria were dialysis therapy before hospital admission, no measurement of serum creatinine during hospital admittance, no available baseline serum creatinine <1 year before hospitalization.
Indication for SGLT-2i
In SGLT-2i users, we determined the indication for prescription: heart failure, type 2 diabetes, and/or CKD. A history of heart failure was identified using registered ICD-10 (International Classification of Diseases and Related Health Problems) codes [14], while a history of diabetes was determined based on a combination of ICD-10 codes and the use of diabetes-specific medications (see online suppl. material Table 4; for all online suppl. material, see https://doi.org/10.1159/000551925). CKD was defined as an eGFR of <60 mL/min/1.73 m2 and/or two measurements of an albumin/creatinine ratio of ≥30 mg/mmol within 1 year.
Outcome and Definitions
The primary outcome of the study was incidence of AKI during hospitalization. The secondary outcomes were the incidence of AKI within SGLT-2i indication and admission specialty groups and AKI stages for admission with and without SGLT-2i treatment. AKI was defined according to the KDIGO criteria [1]. We accepted the most recent creatinine measurement taken up to 1 year prior to admission as the baseline measurement. Urine output was not used as this is not structurally evaluated in routine care. The severity of AKI was classified according to the 2012 KDIGO recommendations as stages 1, 2, or 3 [4].
Finally, length of hospital stay and mortality rates were explored in admissions with and without AKI and stratified by SGLT-2i treatment. Admission specialty, defined as the specialty treating the patient at the moment of AKI, was categorized into 4 groups: surgical specialty, non-surgical specialty, internal medicine, and intensive care (IC) (online suppl. material Table 1). Patients being admitted directly to the IC were classified as such; otherwise, the ward on which AKI occurred was noted. The admission type was categorized as acute or elective. The severity of AKI was classified according to the 2012 KDIGO recommendations as stages 1, 2, or 3 [4].
SGLT-2i Use
Pre-hospital SGLT-2i treatment was defined as an active prescription for SGLT-2i on the day of hospital admission. Controls had no active prescription on the day of hospital admission. Secondly, patients were grouped into four categories: never-users (no pre- and no within-hospital SGLT-2i use), starters (no pre-hospital use but within-hospital use), continuers (pre-hospital and within-hospital use), and stoppers (pre-hospital use but no within-hospital use). In case of AKI, within-hospital use was defined until the time of AKI.
Matching
To minimize bias due to unequal distribution of confounding factors, hospital admissions of patients treated with SGLT-2i and patients not receiving SGLT-2i on the day of hospital admission were matched 1:1 using a combination of exact matching and optimal pairwise propensity score matching without replacement, using the R package MatchIt [15]. Exact matching was performed on the indication for SGLT-2i, reason for admission, admitting speciality, IC admittance, and age in quartiles. Propensity scores were calculated using a multivariable logistic regression model with SGLT-2i treatment as outcome, conditional on gender, use of nephrotoxic medication, antihypertensives, glucose-lowering agents and loop diuretics, baseline kidney function according to KDIGO CKD classification, a history of cardiovascular disease (as defined by ICD-10 scores [14]), and admission type (online suppl. material Tables 2–5). Matching was considered successful as all but eleven eligible hospital admissions with SGLT-2i were matched with a control (Fig. 1), and all standardized mean differences (SMDs) were ≤0.1, indicating sufficient balance between the two groups (Table 1).
Fig. 1.
Patient enrollment diagram.
Table 1.
Baseline characteristics
| | SGLT-2i treatment group (n = 527) | Control group (n = 527) | SMD |
|---|---|---|---|
| Age (mean, SD) | 71.3 (9.5) | 71.4 (10.4) | 0.00*1 |
| Sex, male (%) | 339 (68.1) | 350 (66.4) | 0.03 |
| Medication use | |||
| Number of nephrotoxic medications (%) | | | 0.03 |
| 0 | 135 (25.6) | 151 (28.7) | |
| 1 | 255 (48.4) | 242 (45.9) | |
| >1 | 137 (26.0) | 134 (25.4) | |
| Antihypertensive agents (%) | 459 (87.1) | 449 (85.2) | 0.06 |
| Glucose-lowering agents (%) | 231 (43.8) | 238 (45.2) | −0.03 |
| Loop diuretics (%) | 327 (62.0) | 327 (62.0) | 0.00 |
| Pre-existent kidney function | |||
| G1 (%) | 45 (8.5) | 49 (9.3) | 0.02 |
| G2 (%) | 122 (23.1) | 118 (22.4) | 0.00 |
| G3a (%) | 121 (23.0) | 121 (23.0) | 0.00 |
| G3b (%) | 136 (25.8) | 126 (23.9) | 0.04 |
| G4 (%) | 97 (18.4) | 101 (19.2) | −0.02 |
| G5 (%) | 6 (1.1) | 12 (2.3) | −0.1 |
| Medical history | |||
| Cardiovascular disease (%) | 512 (97.2) | 512 (97.2) | 0.00*2 |
| Diabetes mellitus (%) | 148 (28.1) | 134 (25.4) | |
| Admission specialty | |||
| Surgical specialty (%) | 78 (14.8) | 78 (14.8) | 0.00 |
| Non-surgical specialty (%) | 384 (72.9) | 384 (72.9) | 0.00 |
| Internal medicine (%) | 63 (12.0) | 63 (12.0) | 0.00 |
| Intensive care (%) | 2 (0.4) | 2 (0.4) | 0.00 |
| SGLT-2i indication | |||
| Heart failure (%) | 82 (15.6) | 82 (15.6) | 0.00 |
| T2DM (%) | 38 (7.2) | 38 (7.2) | 0.00 |
| CKD (%) | 29 (5.5) | 29 (5.5) | 0.00 |
| A combination of heart failure, T2DM and/or CKD (%) | 378 (71.7) | 378 (71.7) | 0.00 |
| Admission type | |||
| Acute (%) | 471 (89.4) | 473 (89.8) | 0.00 |
| Elective (%) | 56 (10.6) | 54 (10.2) | 0.00 |
| IC admission | |||
| IC admission during admission (%) | 25 (4.7%) | 25 (4.7%) | 0.00 |
*1Exact matching on age was performed on age quartiles.
*2No matching was performed on a medical history of diabetes.
Statistical Analyses
Descriptive statistics with SMDs are presented for hospital admissions with and without pre-hospital SGLT-2i treatment. For the primary outcome, the incidence of AKI (with 95% confidence interval, CI) in admissions with and without pre-hospital SGLT-2i treatment was reported. The relative risk for AKI in admissions with SGLT-2i treatment versus those without (controls) was estimated using a log-binomial regression model. In a sensitivity analysis, the association between SGLT-2i treatment and AKI was assessed with mixed binary logistic regression considering the clustering of admissions within patients and matched pairs by including two random intercepts.
Second, the primary analysis was repeated for SGLT-2i treatment in four groups based on SGLT-2i treatment at the time of and during admission (never-users, starters, continuers, stoppers). Furthermore, the incidence of AKI was reported for SGLT-2i treated and non-treated groups stratified for indication and admission specialty. The proportion of AKIs within each AKI severity category was reported in admissions with and without SGLT-2i treatment. Finally, the median length of hospital stay (with interquartile range) and mortality rates were reported in the two SGLT-2i treatment groups and separately in those with and without AKI.
Results
Study Population Characteristics
We identified 9,914 admissions meeting the inclusion criteria, of whom 538 were of SGLT-2i users. After matching, the analysis included 527 admissions with SGLT-2i treatment (348 unique patients) and 527 matched controls (472 unique patients) (Fig. 1). Mean (SD) age was 71 years [10], with 66–68% males in each group. The indications for SGLT-2i were heart failure (16%), T2DM (7%), and CKD (6%), while 72% of patients had combined indications. Most admissions were to non-surgical specialties (73% in both groups). Nephrotoxic medication use during admission was similar in both groups: in the SGLT-2i group, 26% of patients used none, 48% used one, and 26% used more than one nephrotoxic medication during hospitalization. In the control group, this was 29%, 46%, and 25%, respectively. Most patients were admitted in an emergency setting, with only 10% admitted electively.
Primary Outcome
Among SGLT-2i-treated patients, 92 AKIs were observed in 527 admissions (17.5%, 95% CI = 14.2–20.7) versus 136 AKIs in 527 admissions (25.8%, 95% CI = 22.1–29.5) in the control group. The relative risk of AKI for SGLT-2i treatment versus no treatment (control) is 0.68 (95% CI = 0.53–0.86) (Table 2, Fig. 2).
Table 2.
AKI incidence and relative risk of AKI for pre-admission treatment with SGLT-2i versus no pre-hospital SGLT-2i treatment
| Pre-admission SGLT-2i treatment | AKI incidence, N (%) (95% CI) | RR (95% CI) |
|---|---|---|
| No (n = 527) | 136 (25.8%) | Reference |
| (22.1–29.5) | ||
| Yes (n = 527) | 92 (17.5%) | 0.68 (0.53–0.86) |
| (14.3–20.7%) |
Fig. 2.
AKI incidence and RRs for AKI for SGLT-2i treatment and in-hospital treatment groups. Represented are RR (relative risk) and CIs: relative risk of AKI for starters, continuers, and stoppers versus those who were never treated with SGLT-2i.
In a sensitivity analysis, the clustering of admissions within patients and matched pairs ranged from modest to substantial. The estimated protective effect for SGLT-2i treatment on AKI did become stronger when adjusted for this clustering (online suppl. material Table 8).
Secondary Outcomes
Patients who continued SGLT-2i treatment throughout hospitalization had a significantly lower risk of AKI compared to patients who were never treated with SGLT-2i (i.e., no pre-hospital and no within-hospital treatment) (16.6% vs. 25.8%, RR 0.65; 95% CI 0.49–0.85) (Table 3, Fig. 2).
Table 3.
AKI incidence and relative risk of AKI for groups defined on both pre-hospital and within hospital treatment of SGLT-2i
| SGLT-2i treatment group | AKI incidence, N (%) (95% CI) | RR (95% CI) |
|---|---|---|
| Never-users (n = 471) | 121 (25.7%) | Reference |
| (21.7–29.6%) | ||
| Starters (n = 56) | 15 (26.8%) | 1.04 (0.66–1.65) |
| (15.8–40.3%) | ||
| Continuers (n = 397) | 66 (16.6%) | 0.65 (0.49–0.85) |
| (13.1–20.7%) | ||
| Stoppers (n = 130) | 26 (20.0%) | 0.78 (0.53–1.13) |
| (13.5–27.9%) |
RR, relative risk of AKI for starters, continuers, and stoppers versus those who were never treated with SGLT-2i (reference group); CI, confidence interval (Wald).
When stratified by the clinical indication for SGLT-2i or by admission specialty, patients treated with SGLT-2i consistently had lower risk of AKI compared to controls (Table 4; online suppl. Table 6). Lastly, SGLT-2i-treated patients were less likely to progress to higher AKI stages compared to controls (stage 1: 79% vs. 63%, stage 2: 12% vs. 27%, stage 3: 9% vs. 10%) (Table 5).
Table 4.
AKI incidence per SGLT-2i indication and pre-admission treatment
| Indication | SGLT-2i treatment | No AKI | AKI | Total |
|---|---|---|---|---|
| HF | No | 76 (92.7%) | 6 (7.3%) | 82 |
| Yes | 75 (91.5%) | 7 (8.5%) | 82 | |
| Total | 151 (92.1%) | 13 (7.9%) | 164 | |
| DM2 | No | 34 (89.5%) | 4 (10.5%) | 38 |
| Yes | 38 (100%) | 0 (0%) | 38 | |
| Total | 72 (94.7%) | 4 (5.3%) | 76 | |
| CKD | No | 23 (79.3%) | 6 (20.7%) | 29 |
| Yes | 28 (96.6%) | 1 (3.4%) | 29 | |
| Total | 51 (87.9%) | 7 (12.1%) | 58 | |
| >1 | No | 258 (68.3%) | 120 (31.7%) | 378 |
| Yes | 294 (77.8%) | 84 (22.2%) | 378 | |
| Total | 552 (73.0%) | 204 (27.0%) | 756 |
Table 5.
AKI stage per pre-admission SGLT-2i treatment
| SGLT-2i treatment | AKI stage 1 | AKI stage 2 | AKI stage 3 | Total |
|---|---|---|---|---|
| No | 85 (62.5%) | 37 (27.2%) | 14 (10.3%) | 136 |
| Yes | 73 (79.3%) | 11 (12.0%) | 8 (8.7%) | 92 |
| Total | 158 (69.3%) | 48 (21.1%) | 22 (9.6%) | 228 |
In patients with AKI, hospital admissions were longer (median length of hospital stay 7 days, IQR: 4–11) than of patients without AKI (median 3 days, IQR: 1–7). No difference in length of hospital stay was seen in admissions with SGLT-2i treatment and controls (Table 6). In-hospital mortality was higher in admissions with AKI than those without (12.3% vs. 2.5%). In admissions with AKI, mortality was numerically higher in controls (15.4%) as compared to admissions with pre-hospital SGLT-2i treatment (7.6%; risk difference = 7.8%, 95% CI = −0.9% to 15.7%), although this difference is uncertain since the CI includes zero. Such a difference was not observed in admissions without AKI (Table 6).
Table 6.
Length of hospital stay and in-hospital mortality rates for patients with and without AKI and further stratified by pre-admission SGLT-2i treatment
| | No AKI | AKI | ||||
|---|---|---|---|---|---|---|
| All | SGLT-2i treatment | control group | All | SGLT-2i treatment | control group | |
| Length of hospital stay, median (IQR) | 3 (1–7) | 3 (1–7) | 3 (1–7) | 7 (4–11) | 7 (5–13) | 6 (3–10) |
| Mortality, n (%, 95% CI) | 21 (2.5%, 1.6–3.8%) | 13 (3.0%, 1.7–4.9%) | 8 (2.0%, 1.0–3.8%) | 28 (12.3%, 8.5–17.0%) | 7 (7.6%, 3.5–14.4%) | 21 (15.4%, 10.1–22.2%) |
Discussion
We observed a 32% lower AKI incidence in admissions of hospitalized patients treated with SGLT-2i at the time of admission, compared to matched patients without SGLT-2i. This effect remained consistent in patients in whom SGLT-2i treatment was continued during admission. Furthermore, patients treated with SGLT-2i at admission showed less progression to more severe AKI stages. Although hospital length of stay did not differ, SGLT-2i treatment at admission showed numerically lower mortality among patients with AKI compared to patients not treated with SGLT-2i, but the sample size precludes firm conclusions on this outcome.
The precise mechanisms by which SGLT-2is prevent AKI remain unclear, although several complementary mechanisms are believed to contribute. Firstly, by reducing proximal tubular glucose and sodium reabsorption, SGLT-2is lower local energy demand, decreasing local reactive oxygen species generation and tubular hypoxia. Furthermore, SGLT-2 inhibition is reported to reduce pro-apoptotic, inflammatory, and profibrotic pathways associated with mitochondrial dysfunction and reactive oxygen species production, further mitigating oxidative stress and tubular injury [16]. This is reflected in reductions in relevant markers associated with tubular damage [17, 18].
Thirdly, SGLT-2i are hypothesized to restore tubuloglomerular feedback by increasing sodium delivery to the macula densa, which enhances adenosine-mediated afferent arteriolar vasoconstriction, lowering intraglomerular pressure and protecting the glomerulus from hyperfiltration injury [19]. Lastly, off-target mechanisms improving cardiovascular outcomes in SGLT-2i use are hypothesized to contribute to reduced AKI incidence [16].
The addition of SGLT-2i to the management of heart failure, T2DM, and CKD has been shown to improve cardiovascular outcomes. Although initial concerns were raised about their potential to increase AKI risk due to their diuretic effects and associated reduction of extracellular volume, studies in patients with T2DM, heart failure, or CKD receiving SGLT-2i showed a reduced risk of AKI [9, 20–23]. Our study adds valuable insights by demonstrating that in a real-world setting, treatment of SGLT-2i is associated with a lower risk of AKI during hospitalization in a matched population. These findings contribute to the growing body of evidence supporting the renoprotective effects of SGLT-2i and highlights their potential to reduce the risk of AKI in hospitalized patients.
A key strength of our study is the inclusion of hospitalized patients, a group which is underrepresented in prior studies [9, 12, 20, 23]. Unlike most previous studies, which were post hoc analyses from randomized trials, our analysis used real-world data. Furthermore, by applying the KDIGO definition based on serum creatinine values, we mitigate the risk of underreporting AKI events. Even if AKI was not clinically diagnosed, these criteria capture relevant events, enhancing accuracy and reliability. However, as urine output is not routinely monitored and not used as AKI criterion in this study, we may have missed some AKI cases. Lastly, by matching on key determinants of AKI risk, we minimized bias from measured confounders and strengthened the internal validity of the comparisons.
An additional strength is our evaluation of in-hospital SGLT-2i treatment patterns. We observed a consistent reduction in AKI risk among patients who continued SGLT-2i treatment during hospitalization compared to patients who were never treated with SGLT-2i. This suggests that routine discontinuation at admission may be unnecessary and that continued therapy could even be protective. No significant reduction in AKI risk was seen in patients who initiated SGLT-2i during hospitalization. This may be a reflection of the decline in eGFR directly following initiation [10, 24], or suggests a delay in renoprotective effects of SGLT-2i. However, it is important to note our study was not designed to appropriately investigate the impact of discontinuing or starting SGLT-2i during hospitalization, so no conclusions can be drawn from this analysis.
Because SGLT-2i use may be associated with other antihypertensive or diuretic medications, we examined AKI incidence across medication subgroups (online suppl. material Table 7). Within each group, SGLT-2i use was consistently associated with lower AKI risk. Although absolute AKI rates differed, the relative reduction with SGLT-2i remained consistent, suggesting the association is not driven by concomitant medication use.
Our results should be interpreted with some caution. Although we applied matching to control for key confounders, the retrospective nature of our study means bias may remain. In particular, clinical decisions to discontinue SGLT-2i treatment during hospital admission may reflect unmeasured factors. Although we matched on key clinical characteristics associated with AKI risk, residual indication bias may persist if physicians avoided initiating SGLT-2i therapy based on unmeasured factors or perceived clinical vulnerability not captured in our matching variables, or if (pre-hospital) initiation of SGLT-2i is associated with other unmeasured treatment factors that could affect AKI risk. Finally, despite our reasonable sample size, the model to estimate the relative risk with adjustment for clustering within patients and matched pairs did not converge. The alternative model used in the sensitivity analysis, which accounts for this clustering, suggests that the crude estimate of a 32% relative risk reduction is more likely an underestimation rather than an overestimation of the effect. Further, the study was underpowered to study rare outcomes such as mortality.
Conclusion
Our real-world data suggest a substantial reduction in the risk of AKI during hospitalization in high-risk patients using SGLT-2i, suggesting a significant renoprotective effect. This reduction did not correspond with shorter hospital stay, and the effect of SGLT-2i use on mortality risk in patients with AKI remains uncertain. Further randomized trials are needed to validate our observations, particularly to determine whether the initiation of SGLT-2i during hospitalization can further reduce the risk of AKI and if SGLT-2i should be discontinued after AKI has occurred. Effects on mortality should be assessed in larger populations.
Acknowledgments
The authors thank all colleagues involved in data management and patient care at the Frisius Medical Centre and thank the AKI-PATROL consortium members for their help in making this research possible.
Statement of Ethics
This study was conducted in accordance with the Declaration of Helsinki. Because it involved retrospective analysis of data from electronic medical records, the Medical Ethical Committee of the Frisius Medical Centre classified the study as non-WMO and exempt from formal ethical review (date of nWMO declaration 22-11-2023). The requirement for active informed consent was waived by the Frisius Medical Centre. Data from patients who had opted out of research use were excluded, and all data were handled in compliance with privacy regulations.
Conflict of Interest Statement
The authors declare no conflicts of interest.
Funding Sources
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contributions
De Martines and Aydinoglu drafted the manuscript. Westra performed the data extraction and cleaning. De Jong performed the statistical analyses. Gant and Adema conceived the study and contributed to the study design and interpretation of findings, and provided critical revisions to the manuscript. All authors reviewed and approved the final version of the manuscript.
Funding Statement
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon request.
Supplementary Material.
Supplementary Material.
References
- 1. Susantitaphong P, Cruz DN, Cerda J, Abulfaraj M, Alqahtani F, Koulouridis I, et al. World incidence of AKI: a meta-analysis. Clin J Am Soc Nephrol. 2013;8(9):1482–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Veltkamp DMJ, Porras CP, Gant CM, Groenestege WMT, Kok MB, Verhaar MC, et al. Long-term risks of adverse kidney outcomes after acute kidney injury: a systematic review and meta-analysis. Nephrol Dial Transplant. 2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Jensen SK, Heide-Jørgensen U, Gammelager H, Birn H, Christiansen CF. Acute kidney injury duration and 20-Year risks of CKD and cardiovascular disease. Kidney Int Rep. 2024;9(4):817–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–84. [DOI] [PubMed] [Google Scholar]
- 5. Zinman B, Wanner C, Lachin JM, Fitchett D, Bluhmki E, Hantel S, et al. Empagliflozin, cardiovascular outcomes, and mortality in type 2 diabetes. N Engl J Med. 2015;373(22):2117–28. [DOI] [PubMed] [Google Scholar]
- 6. Teo YH, Teo YN, Syn NL, Kow CS, Yoong CSY, Tan BYQ, et al. Effects of Sodium/Glucose cotransporter 2 (SGLT2) inhibitors on cardiovascular and metabolic outcomes in patients without diabetes mellitus: a systematic review and meta-analysis of randomized-controlled trials. J Am Heart Assoc. 2021;10(5):e019463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Zelniker TA, Wiviott SD, Raz I, Im K, Goodrich EL, Bonaca MP, et al. SGLT2 inhibitors for primary and secondary prevention of cardiovascular and renal outcomes in type 2 diabetes: a systematic review and meta-analysis of cardiovascular outcome trials. Lancet. 2019;393(10166):31–9. [DOI] [PubMed] [Google Scholar]
- 8. Gilbert RE, Thorpe KE. Acute kidney injury with sodium-glucose co-transporter-2 inhibitors: a meta-analysis of cardiovascular outcome trials. Diabetes Obes Metab. 2019;21(8):1996–2000. [DOI] [PubMed] [Google Scholar]
- 9. Menne J, Dumann E, Haller H, Schmidt BMW. Acute kidney injury and adverse renal events in patients receiving SGLT2-inhibitors: a systematic review and meta-analysis. PLoS Med. 2019;16(12):e1002983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Heerspink HJL, Oshima M, Zhang H, Li J, Agarwal R, Capuano G, et al. Canagliflozin and kidney-related adverse events in type 2 diabetes and CKD: findings from the randomized CREDENCE trial. Am J Kidney Dis. 2022;79(2):244–56.e1. [DOI] [PubMed] [Google Scholar]
- 11. Thiele K, Rau M, Hartmann NUK, Möller M, Möllmann J, Jankowski J, et al. Empagliflozin reduces markers of acute kidney injury in patients with acute decompensated heart failure. ESC Heart Fail. 2022;9(4):2233–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Shi Y, Dong S, Xu Y, Ji D, Yang L, Zhao R. Acute kidney injury events in patients with diabetes using sodium glucose transporter 2 inhibitors: a meta-analysis of cohort studies. Acta Diabetol. 2024;61(6):755–64. [DOI] [PubMed] [Google Scholar]
- 13. Ejaz AA, Mohandas R. Are diuretics harmful in the management of acute kidney injury? Curr Opin Nephrol Hypertens. 2014;23(2):155–60. [DOI] [PubMed] [Google Scholar]
- 14.Available from: https://www.icd10data.com/ICD10CM/Codes2025
- 15. Ho D, Imai K, King G, Stuart EA. MatchIt: nonparametric preprocessing for parametric Causal inference. J Stat Softw. 2011;42(8):1–28. [Google Scholar]
- 16. Nakao Y, Mori M, Mori Y, Bonventre JV. SGLT2 inhibitors and acute kidney injury. Nephrol Dial Transpl. 2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Dharia A, Khan A, Sridhar VS, Cherney DZI. SGLT2 inhibitors: the sweet success for kidneys. Annu Rev Med. 2023;74:369–84. [DOI] [PubMed] [Google Scholar]
- 18. Palmer BF, Clegg DJ. Kidney-Protective effects of SGLT2 inhibitors. Clin J Am Soc Nephrol. 2023;18(2):279–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Thomson SC, Rieg T, Miracle C, Mansoury H, Whaley J, Vallon V, et al. Acute and chronic effects of SGLT2 blockade on glomerular and tubular function in the early diabetic rat. Am J Physiol Regul Integr Comp Physiol. 2012;302(1):R75–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Nuffield Department of Population Health Renal Studies GroupSGLT2 inhibitor Meta-Analysis Cardio-Renal Trialists' Consortium . Impact of diabetes on the effects of sodium glucose co-transporter-2 inhibitors on kidney outcomes: collaborative meta-analysis of large placebo-controlled trials. Lancet. 2022;400(10365):1788–801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zhao M, Sun S, Huang Z, Wang T, Tang H. Network meta-analysis of novel glucose-lowering drugs on risk of Acute Kidney injury. Clin J Am Soc Nephrol. 2020;16(1):70–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Vukadinović D, Abdin A, Anker SD, Rosano GMC, Mahfoud F, Packer M, et al. Side effects and treatment initiation barriers of sodium-glucose cotransporter 2 inhibitors in heart failure: a systematic review and meta-analysis. Eur J Heart Fail. 2022;24(9):1625–32. [DOI] [PubMed] [Google Scholar]
- 23. Mavrakanas TA, Tsoukas MA, Brophy JM, Sharma A, Gariani K. SGLT-2 inhibitors improve cardiovascular and renal outcomes in patients with CKD: a systematic review and meta-analysis. Sci Rep. 2023;13(1):15922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Jongs N, Chertow GM, Greene T, McMurray JJV, Langkilde AM, Correa-Rotter R, et al. Correlates and consequences of an acute change in eGFR in response to the SGLT2 inhibitor dapagliflozin in patients with CKD. J Am Soc Nephrol. 2022;33(11):2094–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to privacy reasons but are available from the corresponding author upon request.
