Chronic Treatment with SGLT-2 Inhibitors is Associated with ICU Admission and Disease Severity in Patients with Diabetic Ketoacidosis: A Propensity Score–Matched Cohort Study

doi:10.21203/rs.3.rs-4802040/v1

Introduction: SGLT-2 inhibitors (SGLT-2i) are linked to a higher risk of diabetic ketoacidosis (DKA). However, it is still unclear whether the severity of SGLT-2i associated DKA is higher.

Methods: This is a retrospective cohort study with patients admitted for DKA at a tertiary hospital (2013-2024). Patients were matched by propensity score for age, sex, diabetes duration, type, and ischemic heart disease. ICU admission risk and clinical severity were compared between SGLT-2i users and controls.

Results: The matched sample included 105 subjects (35 SGLT-2i users, 70 controls). The average age was 63.1±15.4 years, and 40 (38.1%) patients were women. ICU admission was higher in the treatment group (65.7% versus 24.6%, p<0.001). A conditional logistic regression showed higher risk of ICU admission in the treatment group (odds ratio 12.7, 95% confidence interval 1.9 - 84.3, p=0.009) after adjusting for confounding factors. The treatment group exhibited less favorable blood gas results (pH 7.10±0.17 vs 7.18±0.16, p=0.024) and shorter symptom duration (2 [1-3] vs 3 [2-7] days, p<0.002). No significant differences were found in diabetes type, ketonemia, creatinine, or DKA precipitating factors.

Conclusion: DKA in patients with diabetes treated with SGLT-2i is associated with more severe acidosis with quicker onset, leading to higher risk of ICU admission compared to patients not receiving this treatment. We recommend temporary discontinuation of SGLT-2i during any acute event until resolution, regardless of diabetes type or the patient's glycemic control.

diabetic ketoacidosis

diabetes

SGLT-2 inhibitor

ICU

Diabetic ketoacidosis (DKA) is a severe and acute complication of diabetes mellitus that results in a high rate (50%) of admission to the intensive care unit (ICU) [1]. The most common causes of DKA in type 1 diabetes (T1D) include treatment omission and infections. In type 2 diabetes (T2D), DKA is caused by acute severe events such as systemic infections or acute cardiovascular events[2, 3].

SGLT-2 inhibitors (SGLT-2i)have revolutionized the management of T2D [4, 5] and have also been used in T1D due to their significant cardiovascular and renal protective benefits[6, 7]. However, one of the major drawbacks of these drugs is the increased risk of DKA associated to their use that has been described in numerous case reports, as well as in several studies and systematic reviews [8–10]. This increased risk has even been observed in non-diabetic patients [11]. Despite the consensus on the DKA risk associated with SGLT-2i [12, 13], the impact of these drugs on DKA severity remains to be studied.

This study aimed to investigate the impact of SGLT-2i treatment on DKA severity and risk of ICU admission in diabetic patients presenting to the Emergency Department. Using a propensity score-matched cohort analysis, we sought to compare outcomes between patients treated with SGLT-2i and a control group of untreated patients.

This is a retrospective cohort study including all patients diagnosed with DKA attended at the Emergency Department of a tertiary hospital in Spain between 2013 and 2024. Eligible participants had to be aged 15 years or older and meet DKA criteria: glucose >200 mg/dL, plasma β-hydroxybutyrate ≥3.0 mmol/L and/or urine ketones >2+, and pH < 7.3 and/or serum bicarbonate <15.0 mmol/L [2, 14]. Patients not meeting these criteria, those with incomplete information regarding personal history, physical examination, vital signs, or treatment, and those directly referred to another Emergency Department were excluded.

The study followed the strengthening the reporting of observational studies in epidemiology (STROBE) guidelines and was approved by the Hospital [BLINDED] Research Ethics Committee (Study Number: 5554–06/24), in accordance with the principles of the Helsinki Declaration.

Procedures

The treatment protocol for each patient with DKA was based on clinical practice guidelines from the American Diabetes Association consensus on the management of hyperglycemic crises [15]. Initial treatment included administering 0.9% saline at 15-20 mL/kg/h during the first hour, followed by 0.45% or 0.9% saline at 250-500 mL/h according to serum sodium levels. When glucose levels dropped below 200 mg/dL, a 5% glucose solution was added. Patients received an initial bolus of regular insulin at 0.1 IU/kg, followed by a continuous infusion at 0.1 IU/kg/h, halving the infusion rate when glucose fell below 200 mg/dL. Sodium bicarbonate (100 mmol) was administered every 2 hours if initial pH was less than 6.9, until pH exceeded 7.0. Upon meeting DKA resolution criteria, subcutaneous insulin glargine was administered and intravenous infusion was discontinued one hour later. Intravenous potassium was administered if serum levels were below 3.3 mEq/L. Decisions on ICU admission were made based on inadequate response to acidosis correction, uncontrollable precipitating events, need for intubation, or individual clinical judgment.

Capillary ketone (ß-hydroxybutyrate) strips FreeStyle Optium ß-Ketone (Abbot, Chicago, Illinois) were used when available. Venous and arterial blood gas analysis was performed using the GEM® Premier™ 5000 blood gas analyzer (Salzburg, Austria).

Data Collection

A comprehensive set of sociodemographic and clinical variables were collected from patient´s medical records registered upon initial Emergency Department assessment:

Variables from Medical history: Age, sex, diabetes duration, recent glycated hemoglobin (HbA1c), history of ischemic heart disease, stroke, diabetic retinopathy, chronic kidney disease, amputations, hypertension, dyslipidemia, HIV/AIDS, active cancer treatment, psychiatric history, diabetes type, eligibility for ICU admission, and treatment with SGLT-2i.

Physical Examination and symptoms: Mean arterial pressure, respiratory rate, temperature at admission, oxygen saturation, body mass index, Glasgow Coma Scale, dyspnea, vomiting, confusion, weight loss, days from symptom onset, gastroenteritis, missed doses of medication, non-gastroenteritis infections, cardiovascular events.

Results from Basic Laboratory Analysis in the Emergency Department: Hemoglobin, hematocrit, white blood cell count, platelets, creatinine, urea, sodium, potassium, total bilirubin, C-reactive protein (CRP), procalcitonin, pH, PCO2, bicarbonate, base excess, L-lactate, and capillary ketones.

Statistical analysis

Quantitative variables are represented as mean and standard deviation (SD) or median and interquartile range (IQR) and categorical variables as number of events and percentage.

To address the lack of randomization between the study groups and to minimize confounding bias, the 1:2 nearest neighbor propensity score matching method (caliper width 0.015 of the standard deviation [SD] of the logit propensity score) was employed to estimate the effect of treatment on the DKA risk profile [16]. Propensity scores were estimated using logistic regression using age, sex, duration of diabetes, type of diabetes, presence of retinopathy, and history of ischemic heart disease. The matching procedure excluded one patient treated with SGLT-2i, resulting in n = 35 individuals who were matched 1:2 with control group subjects (n = 70) within a 12-month window of the event date.

The propensity score distributions between the SGLT-2i treatment and control groups were well-aligned (p = 0.985; Supplementary S1). Bivariate differences between these groups were calculated using t-tests or Mann-Whitney U tests for quantitative variables depending on whether they were normally distributed or not, and Chi-square tests for categorical variables.

Additionally, potential confounding factors were assessed based on compliance with three confounding criteria in the study variables [17]. The first criterion determined whether the confounder was an independent risk factor for the outcome (ICU admission), regardless of the intervention. This was evaluated by significant association between the factor and outcome in the control group. The second criterion assessed whether the confounder was associated with the exposure (SGLT-2i treatment), through mean comparison and Spearman correlation analysis. The third criterion examined whether the confounder was not a consequence of the exposure (SGLT-2i treatment).

Finally, a conditional logistic regression model for matched data, adjusted for confounding variables, was performed to study the association between the risk of ICU admission in DKA patients and treatment with SGLT-2 inhibitors.
Statistical analysis was performed using R version 4.0.3 [18] and STATA 17.0 Basic Edition (Lakeway Drive, TX, USA). Statistical significance was set at p < 0.05.

Fig 1 shows a flowchart detailing patient recruitment. Out of 214 patients meeting DKA criteria between 2013 and 2024, 180 were not diabetes onset cases and were candidates for ICU admission. Among them, 36 were receiving SGLT-2i treatment at the time of DKA appearance, while 144 were not under this treatment. Propensity score matching excluded one patient from the SGLT-2i treatment group, resulting in a final sample of 35 patients on SGLT-2i treatment during DKA and 70 controls without SGLT-2i treatment.

Mean age of the study population was 63.1 ± 15.4 years, and 40 (38.1%) patients were males. Of all events, 30 (28.6%) were attributed to T1D, 71 (67.6%) to T2D, and 4 (3.8%) to pancreatic diabetes. A total of 40 events (38.1%) needed ICU admission.

Differences between patients treated with SGLT-2 inhibitors and control untreated patients.

Differences between the SGLT-2i treatment group and the control group were examined for variables from the medical history, and for those obtained in the physical examination and the emergency laboratory analysis.

No differences were observed between the SGLT-2i treatment group and the control group for medical history variables such as age, sex, disease duration, chronic complications, history of ischemic heart disease, substance use, or oncological history. The only statistically significant difference was observed in HbA1c levels, which were higher in the treatment group compared to the control group (8.0% [7.3 – 10.3] vs 10.8 % [9.1–13.1], p < 0.001). Results from the remaining medical history variables are presented in Table 1.

Table 1. Sociodemographic and clinical characteristics of the study population according to treatment with SGLT-2 inhibitors.

Variable	Obs n= 105	Control n =70	SGLT-2i n=35	P value
Age (years)	63.1± 15.4	63.8 ± 20.7	61. ± 13.2	0.513
Sex (woman)	40 (38.1%)	27 (38.5%)	13 (37.1%)	0.887
Type Diabetes Mellitus Type 1 Diabetes Type 2 Diabetes Pancreatic Diabetes	30 (28.6%) 71 (67.6%) 4 (3.8%)	18 (25.7%) 49 (70.0%) 3 (4.3%)	12 (34.2%) 22 (62.9%) 1 (2.8%)	0.359 0.461 0.718
Duration of diabeyes (years)	10.3 (5.0 – 20.3)	9.8 (1.8 – 20.3)	13.3 (8.4 – 20.4)	0.100
HbA1c (mmol/mol %)	9.9 (8 -12.8)	10.8 (9.1–13.1)	8.0 (7.3 – 10.3)	<0.001
Ischemic heart disease (%)	14 (13.3%)	8 (11.4%)	6 (17.1%)	0.417
Stroke (%)	6 (5.7%)	5 (7.2%)	1 (2.9%)	0.364
Limb amputation(%)	4 (3.8%)	3 (4.2%)	1 (2.9%)	0.718
eGFR < 60 ml/min	14 (13.3%)	11 (15.7%)	3 (8.5%)	0.310
Diabetic retinopathy(%)	21 (20.0%)	15 (21.4%)	6 (17.1%)	0.605
Alcohol (%)	15 (14.4%)	12 (17.1%)	3 (8.8%)	0.257
Smoking habit (%)	23 (21.9%)	15 (21.4%)	8 (22.9%)	0.867
Oncological treatment (%)	9 (8.6%)	7 (10.0%)	2 (5.7%)	0.460

HbA1c: glycated hemoglobin. eGFR: estimated Glomerular Filtration Rate. Statistically significant differences are highlighted in bold.

The median duration of symptoms was shorter in the SGLT-2i-treated group compared to the control group (2 [1 - 3] days vs3 [2 - 7] days, p < 0.002). Additionally, there was a non-significant trend towards greater weight loss since symptom onset in the control group compared to the SGLT-2i group (23.5% vs. 8.8%, p = 0.072), consistent with a longer duration of symptoms.

No differences were observed in the proportion of dyspnea, fever at home, vomiting, or mental confusion; neither in DKA triggering factors. It is noteworthy that the proportion of non-gastroenteritis infections, including urinary tract infections, was nearly identical in both groups (26.1% in the control group vs. 28.6% in SGTL-2i treated patients, p = 0.787). Furthermore, there were no statistically significant differences in heart rate, blood pressure, temperature, respiratory rate, oxygen saturation, Glasgow Coma Scale, or capillary ketone levels (only available for 60 out of all the patients in the study population). All information regarding the variables assessed in the anamnesis and physical examination is presented in Table 2.

Table 2: Anamnesis and physical examination variables according to treatment with SGLT-2 inhibitors

Variable	Obs n= 105	Control n =70	SGLT-2i n=35	P value
Dyspnoea (%)	23 (23.1%)	16 (23.1%)	7 (20.0%)	0.711
Fever at home(%)	9 (8.6%)	4 (5.8%)	5 (14.3%)	0.146
Vomiting (%)	34 (32.7%)	20 (29.0%)	14 (40.0%)	0.258
Clouded consciusness (%)	34 (32.6%)	23 (33.3%)	11 (31.4%)	0.845
Weight loss(%)	19 (18.6%)	16 (23.5%)	3 (8.8%)	0.072
Days with symptoms	2 (1 - 5)	3 (2 - 7)	2 (1 - 3)	0.002
Ketone bodies in blood*	4.9 (4 – 6.4)	4.9 (4 – 6.4)	4.2 (3.8 – 6.4)	0.704
Gastroenteritis (%)	9 (8.6%)	8 (11.6%)	1 (2.9%)	0.134
Infection other than gastroenteritis (%)	28 (26.9%)	18 (26.1%)	10 (28.6%)	0.787
Dose Omission (%)	22 (21.2%)	14 (20.3%)	8 (22.9%)	0.762
Cardiovascular event (%)	11 (10.6%)	8 (11.6%)	3 (8.6%)	0.636
Mean blood pressure (mmHg)	91.4 ± 18.7	91.9 ± 19.0	90.5 ± 18.7	0.752
Cardiac frequency (bpm)	99 (85 – 113)	96 (84 – 112)	102 (85 -114)	0.450
Temperature (ºC)	36.0 ± 0.9	36.1 ± 0.9	35.9 ± 0.8	0.228
Breaths/min	15 (12 - 22)	12 (12 - 20)	18 (12 - 26)	0.230
SatO₂ (%)	95 ± 5	95.2 ± 6.0	95.7 ± 3.9	0.704
Glasgow Scale	15 (14 - 15)	15 (14 - 15)	15 (14 - 15)	0.883

SatO2: Oxygen saturation. Statistically significant differences are highlighted in bold.* Only available for 60 out of all the patients in the study population.

Lastly, regarding laboratory analysis results, the SGLT-2i treatment group exhibited a lower pH compared to the control group (7.10 ± 0.17 vs. 7.18 ± 0.16, p = 0.024) (Fig 2). This group also showed trends towards lower base excess (–17.28 ± 7.84 vs. -14.69 ± 7.43, p = 0.103) and higher potassium levels (4.9 ± 0.8 vs. 4.6 ± 0.0, p = 0.09), although these differences were not statistically significant.

On the other hand, hematocrit was higher in the SGLT-2i treatment group than in the control group (43.69 ± 6.35% vs. 40.59 ± 7.11%, p = 0.032), and baseline glucose levels in the emergency department were higher in the control group than in the treatment group (549 ± 220 mg/dL vs. 338 ± 194 mg/dL, p < 0.001). No statistically significant differences were observed in renal function, hematologic data, or inflammatory parameters (Table 3).

Table 3. Parameters from basic laboratory analysis according to treatment with SGLT-2 inhibitors.

Variable	Obs n= 105		Control n =70		iSGLT-2 n=35		P value
Hemoglobin (g/L)	13.68	± 2.29	13.60	± 2.39	13.85	± 2.12	0.593
Hematocrit (%)	41.65	± 6.99	40.59	± 7.11	43.69	± 6.36	0.032
Platelets (mm³)	295914	± 136550	295624	± 128943	296485	± 152425	0.975
Leukocytes (mm³)	16176.63	± 17182.70	16767.39	± 20484.48	15012	± 7286	0.625
Neutrophils (mm³)	12062	± 6918	12054	± 7182	12077	± 6475	0.987
Lymphocytes (mm³)	1685	± 1267	1758	± 1461	1544	± 762	0.420
Monocytes (mm³)	862	± 636	800	± 508	981	± 824	0.175
Glucose (mg/dL)	478	± 233	549	± 220	338	± 194	<0.001
Urea (mg/dL)	62.03	± 39.64	64.94	± 42.12	56.21	± 33.99	0.297
Creatinine (mg/dL)	1.30	± 0.58	1.35	± 0.60	1.20	± 0.55	0.214
Sodium (mEq/L)	134.47	± 7.07	134.01	± 7.57	135.37	± 5.98	0.357
Potassium (mEq/L)	4.73	± 0.87	4.6	± 0.9	4.9	± 0.8	0.09
Total bilirubin (mg/dL)	0.52	± 0.40	0.55	± 0.43	0.46	± 0.33	0.287
CRP (ng/dL)*	6.76	± 9.32	7.60	± 9.08	5.76	± 9.69	0.461
Procalcitonin (ng/mL)**	0.81	± 1.69	1.00	± 2.12	0.52	± 0.68	0.405
pH	7.15	± 0.16	7.18	± 0.16	7.10	± 0.17	0.024
BE (Base Excess)	-15.58	± 7.63	-14.69	± 7.43	-17.28	± 7.84	0.103
PCO2 (mmHg)	29.83	± 2.22	29.76	± 12.26	29.94	± 12.33	0.944
Bicarbonate (mmol/L)	11.21	± 6.33	11.72	± 6.32	10.23	± 6.33	0.261
L-lactate (mmol/L)	3.27	± 2.25	3.06	± 1.89	3.67	± 2.80	0.200

CRP: C-reactive protein. Statistically significant differences are highlighted in bold. *Only available for 57 out of all the patients in the study population. ** only available for 37 out of all the patients in the study population

SGLT-2 inhibitors and ICU

A comparative analysis showed that the proportion of DKA patients that underwent ICU admission was higher in those treated with SGLT-2i compared to the control group (65.7% vs 24.6%, p < 0.001) (Fig 3).

The proportion of ICU admissions was higher in the group treated with SGLT-2 inhibitors compared to the control group (65.7% vs. 24.6%, p < 0.001).

To better characterize the effect of SGLT-2i exposure on ICU admission, a conditional logistic regression was performed incorporating treatment variables and confounding factors. Analysis of potential confounding factors revealed that none of them was significantly associated with ICU admission among those patients not treated with SGLT-2i, whereas HbA1c was the sole variable meeting confounding criteria for association with SGLT-2i treatment(Supplementary Material S2). Therefore, the conditional logistic regression model for paired data was adjusted for HbA1c.

The model assessing ICU admission in DKA patients showed a higher risk for patients treated with SGLT-2i compared to those untreated, after adjusting for HbA1c at admission (odds ratio [OR] 12.7, 95% confidence interval [CI] 1.9 - 84.3, p = 0.009 for SGLT-2i treatment and OR 0.64, 95% CI 0.45 - 0.91, p = 0.013 for HbA1c at admission).

The aim of this study was to investigate the impact of chronic treatment with SGLT-2i on disease severity and the risk of ICU admission in individuals presenting to the Emergency Department with DKA. Our data, derived from a propensity score-matched cohort study, revealed that chronic treatment with SGLT-2i is associated with an increased risk of ICU admission probably related to more severe acidosis, despite subjects treated with SGLT-2 inhibitors present a more favorable glucose profile.

SGLT-2i are widely used medications that have demonstrated efficacy in reducing glycemia, as well as blood pressure, body weight, albuminuria, cardiovascular events, and even mortality in individuals with T2D [4, 19, 20]. Consistent cardiovascular benefits have also been observed in individuals without diabetes [21, 22]. Although there is less evidence in T1D, real-world studies have shown benefits of SGLT-2i treatment in HbA1c reduction, weight loss, and insulin requirements in T1D patients[7]. Promising beneficial results have also been shown for this treatment in small-sample studies with individuals using open-loop insulin pumps[23, 24].

However, while some older studies did not observe an increase in DKA in individuals treated with SGLT-2i [25], in recent years, mounting evidence has supported that SGLT-2i increase the risk of DKA [8–10, 26–28]. The issue of increased risk of DKA has been addressed in various studies evaluating the efficacy and safety of SGLT-2 inhibitors in recent years. This has led to many research efforts and even international consensus recommendations advocating for individualized drug administration to prevent DKA [6, 12, 29, 30].

Several factors have been associated with an increased risk of DKA in individual treated with SGLT-2i, such as reductions in insulin dose, increased insulin demand, metabolic stress, low carbohydrate intake, female sex, and latent autoimmune diabetes of adulthood [10]. To mitigate biases arising from the lack of randomization, our analysis was adjusted using propensity score matching to ensure there were no differences in various variables associated with an increased risk of DKA in patients under therapy with SGLT-2i, such as alcohol consumption, missed doses, infections, and cardiovascular events [9, 27]. Nevertheless, despite similar personal histories and triggering factors in both groups, lower pH levels, shorter duration of symptoms at home, and a significantly higher ICU admission rate were observed in the group treated with SGLT-2i compared to the control group. These findings align with the observed increase in blood ketone levels associated with SGLT-2i, which underlies the elevated risk of DKA documented in the literature[31]. This increase can lead to more severe acute clinical presentations of DKA that occur faster in individuals with closer-to-target glycemic control compared to those not treated with SGLT-2i. Our data suggested that the severity of DKA in individuals treated with SGLT-2i is increased independently of diabetes type or chronic disease control.

The possible mechanisms of ketoacidosis induction by SGLT-2i are multifaceted. In addition to the increased risk of urinary tract infections and reduced insulin doses [32], treatment with SGLT-2i has been linked to elevated levels of beta-hydroxybutyrate [31]. SGLT2i reduce blood glucose levels by increasing urinary glucose excretion, which in turn reduces insulin secretion from pancreatic β-cells. The decrease in circulating insulin levels results in a reduction in insulin's anti-lipolytic activity and a consequent stimulation of production of free fatty acids, which are converted into beta-hydroxybutyrate through β-oxidation in the liver. Additionally, insulin stimulates the activity of acetyl-CoA carboxylase, which produces malonyl-CoA, a potent inhibitor of carnitine palmitoyltransferase-I (CPT-I). Since CPT-I promotes the transport of fatty acids into the mitochondria and thus increases the rate of β-oxidation, the decrease in circulating insulin levels also promotes ketone body production through CPT-I activation [33]. Furthermore, the administration of SGLT2i stimulates glucagon secretion, either as a side effect mediated by the reduction in insulin secretion or as a direct effect of SGLT2i on pancreatic α-cells. Since glucagon inhibits acetyl-CoA carboxylase and thus increases CPT-I activity in the liver, the upregulation of glucagon secretion also likely contributes to overproduction of beta-hydroxybutyrate[34].This may explain the trend towards more severe acidosis and higher ICU admission rates observed among patients treated with SGLT-2i compared to the control group in our study.

This study underscores the necessity to be specially cautious during any acute event in patients undergoing treatment with SGLT-2i, irrespective of other factors. These precautions are already implemented in patients undergoing surgical interventions, in whom the medication has to be discontinued 24 hours before the procedure [35]. In order to prevent DKA, we suggest extending these recommendations to any concurrent health event, halting the medication until the condition resolves. This measure could mitigate significant complications given the rapid onset, intensity, and need for ICU admission associated with DKA in individuals treated with SGLT-2i.

This study has several limitations. First, it was a single-center retrospective study with inevitable sample selection bias. Second, the study was based on a chohort design, with inherent risk of observation and confounding biases. However, these potential biases were minimized through applications of a propensity score matching protocol. Third, ketone levels, a critical component in studying the etiopathogenesis of DKA, were not measured in a significant portion of the population, thus limiting the quality of these data. Fourth, the impact of socioeconomic variables, which are crucial to disease control [36] and associated complications [37], was not assessed. Lastly, the time spanning from drug initiation to DKA onset was not recorded because, unlike in other studies [27, 38], "Nuestro objetivo no fue analizar las causas de la CAD inducida por iSGLT-2 nor to determine whether iSGLT-2 treatment is associated with the increased incidence of CAD observed in recent years[39].

In summary, DKA in patients with diabetes is associated with more severe acidosis with a quicker onset in those patients treated with SGLT-2i compared to untreated patients. These characteristics lead to a higher risk of ICU admission. We recommend temporary discontinuation of SGLT-2i during any acute event until its resolution, regardless of diabetes type or the patient's glycemic control.

Author Contribuition Information

All authors have reviewed and approved the data, contributed to the development and approval of the manuscript, and acknowledged the decision to submit the manuscript for publication.

Competing Interests

The authors declare they have no financial interests.

Ethics approval and consent to participate

The Research Ethics Committee of Hospital de La Princesa, Madrid, (Study Number: 5554–06/24),) approved this study and waived informed consent from patients. The research was conducted according to the Declaration of Helsinki.

Authors’ Contributions

F.S.-V and J.A.M.-A were involved in the study design, researched data, contributed to discussion, and wrote/edited the article. C.M.-O. performed statistical analyses. M.S.T.-S.. M.L.-R and M.M. wrote/edited the article. J.J.R.-L., V.N.M., S.A., S.G.-C., C.A., A.W.T.., .E.C.-L. and C.S.-L. were involved in data collection and reviewed/edited the article.

Acknowledgments

We appreciate helpful comments made by Manuel Gómez Gutiérrez

Funding

This work was funding by Proyectos de Investigacion en Salud PI19/00584, PI22/01404 and PMP22/00021 (funded by Instituto de Salud Carlos III), iTIRONET- P2022/BMD7379 (funded by Comunidad de Madrid), and co-financed by FEDER funds to MM.

Takahashi K, Uenishi N, Sanui M et al (2024) Clinical profile of patients with diabetic ketoacidosis and hyperglycemic hyperosmolar syndrome in Japan: a multicenter retrospective cohort study. Acta Diabetol 61:117–126. https://doi.org/10.1007/s00592-023-02181-1
Dhatariya KK (2019) Defining and characterising diabetic ketoacidosis in adults. Diabetes Res Clin Pract. https://doi.org/10.1016/j.diabres.2019.107797. 155:
Benoit SR, Zhang Y, Geiss LS et al (2018) Trends in Diabetic Ketoacidosis Hospitalizations and In-Hospital Mortality — United States, 2000–2014. MMWR Morb Mortal Wkly Rep 67:362–365. https://doi.org/10.15585/mmwr.mm6712a3
Romera I, Ampudia-Blasco FJ, Pérez A et al (2016) Eficacia y seguridad de empagliflozina en combinación con otros hipoglucemiantes orales en pacientes con diabetes mellitus tipo 2. Endocrinología y Nutrición 63:519–526. https://doi.org/10.1016/j.endonu.2016.06.003
Pujante P, Ares J, Maciá C et al Efficacy of sodium glucose cotransporter 2 inhibitors as an adjunct treatment for patients with diabetes type 2. Medicina Clínica. https://doi.org/10.1016/j.medcli.2018.09.019
Palanca A, van Nes F, Pardo F et al (2022) Real-world Evidence of Efficacy and Safety of SGLT2 Inhibitors as Adjunctive Therapy in Adults With Type 1 Diabetes: A European Two-Center Experience. Diabetes Care 45:650–658. https://doi.org/10.2337/dc21-1584
Durán-Martínez M, Azriel S, Doulatram-Gamgaram VK et al (2024) Real-world safety and effectiveness of dapagliflozin in people living with type 1 diabetes in Spain: The Dapa-ON multicenter retrospective study. Diabetes Metab 50:101501. https://doi.org/10.1016/j.diabet.2023.101501
Colacci M, Fralick J, Odutayo A, Fralick M (2022) Sodium-Glucose Cotransporter-2 Inhibitors and Risk of Diabetic Ketoacidosis Among Adults With Type 2 Diabetes: A Systematic Review and Meta-Analysis. Can J Diabetes 46:10–15e2. https://doi.org/10.1016/j.jcjd.2021.04.006
Meyer EJ, Gabb G, Jesudason D (2018) SGLT2 Inhibitor–Associated Euglycemic Diabetic Ketoacidosis: A South Australian Clinical Case Series and Australian Spontaneous Adverse Event Notifications. Diabetes Care 41:e47–e49. https://doi.org/10.2337/dc17-1721
Palmer BF, Clegg DJ (2021) Euglycemic Ketoacidosis as a Complication of SGLT2 Inhibitor Therapy. Clin J Am Soc Nephrol 16
Hayes AG, Raven LM, Viardot A et al (2023) SGLT2 Inhibitor–Induced Ketoacidosis in a Patient Without Diabetes. Diabetes Care 47:e4–e5. https://doi.org/10.2337/dc23-1903
Danne T, Garg S, Peters AL et al (2019) International Consensus on Risk Management of Diabetic Ketoacidosis in Patients With Type 1 Diabetes Treated With Sodium–Glucose Cotransporter (SGLT) Inhibitors. Diabetes Care 42:1147–1154. https://doi.org/10.2337/dc18-2316
Wolfsdorf JI, Ratner RE (2019) SGLT Inhibitors for Type 1 Diabetes: Proceed With Extreme Caution. Diabetes Care 42:991–993. https://doi.org/10.2337/dci19-0008
Savage MW, Dhatariya KK, Kilvert A et al (2011) Joint British Diabetes Societies guideline for the management of diabetic ketoacidosis. Diabet Med 28:508–515. https://doi.org/10.1111/j.1464-5491.2011.03246.x
Kitabchi AE, Umpierrez GE, Murphy MB, Kreisberg RA (2006) Hyperglycemic Crises in Adult Patients With Diabetes: A consensus statement from the American Diabetes Association. Diabetes Care 29:2739–2748. https://doi.org/10.2337/dc06-9916
Rassen JA, Shelat AA, Myers J et al (2012) One-to-many propensity score matching in cohort studies. Pharmacoepidemiol Drug Saf 21:69–80. https://doi.org/10.1002/pds.3263
Rothman KJ, Lash TL, Greenland S (2008) Modern epidemiology /. In: Kenneth J, Rothman (eds) Sander Greenland; Timothy L. Lash, 3rd edn. Lippincott Williams & Wilkins, Philadelphia
Team RC (2013) R: A language and environment for statistical computing
Zinman Bernard W, Christoph LJM, Empagliflozin et al Cardiovascular Outcomes, and Mortality in Type 2 Diabetes. N Engl J Med 373:2117–2128. https://doi.org/10.1056/NEJMoa1504720
Neal Bruce P, Vlado, Mahaffey Kenneth W et al Canagliflozin and Cardiovascular and Renal Events in Type 2 Diabetes. N Engl J Med 377:644–657. https://doi.org/10.1056/NEJMoa1611925
Anker SD, Butler J, Filippatos G et al (2021) Empagliflozin in Heart Failure with a Preserved Ejection Fraction. N Engl J Med 385:1451–1461. https://doi.org/10.1056/NEJMoa2107038
McMurray JJV, Solomon SD, Inzucchi SE et al (2019) Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction. N Engl J Med 381:1995–2008. https://doi.org/10.1056/NEJMoa1911303
Haidar A, Yale J-F, Lovblom LE et al (2021) Reducing the need for carbohydrate counting in type 1 diabetes using closed-loop automated insulin delivery (artificial pancreas) and empagliflozin: A randomized, controlled, non-inferiority, crossover pilot trial. Diabetes Obes Metabolism 23:1272–1281. https://doi.org/10.1111/dom.14335
Garcia-Tirado J, Farhy L, Nass R et al (2022) Automated Insulin Delivery with SGLT2i Combination Therapy in Type 1 Diabetes. Diabetes Technol Ther 24:461–470. https://doi.org/10.1089/dia.2021.0542
Wang Y, Desai M, Ryan PB et al (2017) Incidence of diabetic ketoacidosis among patients with type 2 diabetes mellitus treated with SGLT2 inhibitors and other antihyperglycemic agents. Diabetes Res Clin Pract 128:83–90. https://doi.org/10.1016/j.diabres.2017.04.004
Haddadin R, Aboujamra D, Iraninezhad H (2023) Sodium-Glucose Cotransporter-2 Inhibitor-Induced Euglycemic Diabetic Ketoacidosis in a Type 2. Diabet Patient Cureus 15:e51184. https://doi.org/10.7759/cureus.51184
Zhao Z, Zhao F, Zhang Y et al (2023) Risk factors of dapagliflozin-associated diabetic ketosis/ketoacidosis in patients with type 2 diabetes mellitus: A matched case-control study. Diabetes Res Clin Pract. https://doi.org/10.1016/j.diabres.2023.110236. 196:
Peters AL, Buschur EO, Buse JB et al (2015) Euglycemic Diabetic Ketoacidosis: A Potential Complication of Treatment With Sodium–Glucose Cotransporter 2 Inhibition. Diabetes Care 38:1687–1693. https://doi.org/10.2337/dc15-0843
Danne T, Edelman S, Frias JP et al (2021) Efficacy and safety of adding sotagliflozin, a dual sodium-glucose co-transporter (SGLT)1 and SGLT2 inhibitor, to optimized insulin therapy in adults with type 1 diabetes and baseline body mass index ≥ 27 kg/m2. Diabetes Obes Metabolism 23:854–860. https://doi.org/10.1111/dom.14271
Siegmund T, Ampudia-Blasco FJ, Schnell O (2020) Two clinical cases of adjunctive use of a SGLT-2 inhibitor in type 1 diabetes. Diabetes Res Clin Pract 162. https://doi.org/10.1016/j.diabres.2020.108131
Boeder S, Davies MJ, McGill JB et al (2024) Beta-Hydroxybutyrate Levels and Risk of Diabetic Ketoacidosis in Adults with Type 1 Diabetes Treated with Sotagliflozin. Diabetes Technology & Therapeutics. https://doi.org/10.1089/dia.2023.0605
Halimi S, Vergès B (2014) Adverse effects and safety of SGLT-2 inhibitors. Diabetes Metab 40:S28–S34. https://doi.org/10.1016/S1262-3636(14)72693-X
Ogawa W, Sakaguchi K (2016) Euglycemic diabetic ketoacidosis induced by SGLT2 inhibitors: possible mechanism and contributing factors. J Diabetes Invest 7:135–138. https://doi.org/10.1111/jdi.12401
Kibbey RG (2015) SGLT-2 inhibition and glucagon: Cause for alarm? Trends in Endocrinology &. Metabolism 26:337–338. https://doi.org/10.1016/j.tem.2015.05.011
Handelsman Y, Henry RR, Bloomgarden ZT et al (2016) American Association of Clinical Endocrinologists and American College of Endocrinology Position Statement on the Association of SGLT-2 Inhibitors and Diabetic Ketoacidosis. Endocr Pract 22:753–762. https://doi.org/10.4158/EP161292.PS
Sebastian-Valles F, Martínez-Alfonso J, Arranz Martin JA et al (2024) Impact of socioeconomic status on chronic control and complications of type 1 diabetes mellitus in users of glucose flash systems: a follow-up study. BMC Med 22:37. https://doi.org/10.1186/s12916-024-03254-w
Talbo MK, Katz A, Dostie M et al (2022) Associations Between Socioeconomic Status and Patient Experience With Type 1 Diabetes Management and Complications: Cross-sectional Analysis of a Cohort From Québec, Canada. Can J Diabetes 46:569–577. https://doi.org/10.1016/j.jcjd.2022.02.008
Jiménez-Castillo RA, Carrizales-Sepúlveda EF, Román-Montalvo MA et al (2020) Clinical outcomes and biochemical characteristics of a Hispanic cohort of patients with diabetic ketoacidosis: 10-Year experience in an emergency department. Endocrinología Diabetes y Nutrición 67:216–218. https://doi.org/10.1016/j.endinu.2019.09.007
Rodríguez Escobedo R, Lambert C, Huidobro Fernández B et al Cetoacidosis diabética al diagnóstico de diabetes mellitus tipo 1 en Asturias entre 2011 y 2020: influencia de la duración de los síntomas en la prevalencia de cetoacidosis y en la pérdida de peso. e202310090. Rev Esp Salud Pública [Internet]. 26 de octubre de 2023 [citado 25 de julio de 2024];97:13 páginas. Disponible en: https://ojs.sanidad.gob.es/index.php/resp/article/view/170

Chronic Treatment with SGLT-2 Inhibitors is Associated with ICU Admission and Disease Severity in Patients with Diabetic Ketoacidosis: A Propensity Score–Matched Cohort Study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Declarations

References

Supplementary Files

Status:

Version 1