Abstract
Background
Long COVID presents significant health challenges, especially for patients with type 2 diabetes. Emerging evidence suggests that sodium-glucose cotransporter-2 (SGLT2) inhibitors may provide protective effects against COVID-19 complications, but their role in reducing long COVID risk remains unclear.
Methods
Utilizing the TriNetX platform, a retrospective cohort study was conducted among adults with type 2 diabetes diagnosed with COVID-19 between January 1, 2020, and June 30, 2024. Propensity score matching balanced demographic, clinical, and comorbidity profiles between SGLT2 inhibitor users and non-users. Cox proportional hazards regression assessed the risk of long COVID, defined by a spectrum of post-COVID-19 conditions.
Results
Among 5,162 matched pairs, SGLT2 inhibitor use was associated with a significantly lower risk of long COVID (HR = 0.85, 95% CI: 0.79–0.91). In the category of long-COVID symptoms such as abdominal symptoms, anxiety/depression, pain, headache, and cognitive symptoms, there were lower risks observed in the SGLT2 inhibitor group. Subgroup analyses showed consistent risk reduction across different age groups and sexes.
Conclusions
SGLT2 inhibitor use in patients with type 2 diabetes was linked to a reduced risk of long COVID. These findings suggest potential therapeutic benefits beyond glycemic control and highlight the need for further investigation into SGLT2 inhibitors as part of post-COVID-19 management strategies.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12985-025-03054-5.
Keywords: SGLT2 inhibitors, Long COVID, Cognitive symptoms, Type 2 diabetes, Retrospective cohort study
Background
The Coronavirus disease 2019 (COVID-19) pandemic has led to long-term health implications for many individuals, particularly those with underlying chronic conditions [1, 2]. Among these implications, long COVID, also referred to as post-acute sequelae of SARS-CoV-2 infection (PASC), has emerged as a significant public health issue [2]. The World Health Organization (WHO) defines long COVID as the persistence of symptoms for a minimum of three months following acute SARS-CoV-2 infection, in the absence of an alternative diagnosis [3, 4]. These symptoms can impact multiple organ systems and greatly diminish the quality of life for those affected [2, 4]. Recent epidemiological studies suggest that long COVID affects approximately 10–30% of individuals who recover from acute COVID-19, with prevalence varying based on factors such as age, sex, preexisting health conditions, and the severity of the initial infection [4–7]. Long COVID presents a diverse range of symptoms that can persist for months or even years. The most commonly reported symptoms include fatigue, dyspnea, cognitive impairment, persistent cough and chest pain, as well as depression, anxiety, gastrointestinal disturbances, myalgia, and joint pain [5–7]. Notably, individuals with diabetes mellitus, particularly T2DM, are at a heightened risk of developing long COVID compared to the general population. A meta-analysis has indicated that the incidence of long-term COVID-19 among T2DM patients ranges from 25% to 40%, significantly surpassing that of non-diabetic groups [8, 9]. The interplay between long COVID and diabetes is particularly concerning, as the systemic inflammation and metabolic dysregulation associated with post-COVID-19 sequelae can exacerbate glycemic instability, increase insulin resistance, and elevate the risk of cardiovascular complications [8, 10–12]. Additionally, long COVID can negatively affect endothelial function and autonomic regulation, potentially leading to poorer outcomes for patients with preexisting T2DM [13–15].
Long COVID is a multifactorial condition characterized by persistent immune activation, chronic inflammation, endothelial dysfunction, and mitochondrial impairment, leading to systemic symptoms such as fatigue, cognitive dysfunction, and dyspnea. The sustained activation of proinflammatory cytokines including interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), together with NLRP3 inflammasome overactivation, has been implicated in the prolonged post-viral sequelae [5, 16]. Endothelial injury and microthrombosis contribute to cardiovascular and neurological manifestations, while mitochondrial dysfunction promotes oxidative stress and energy imbalance [14].
Although no clinical studies have yet directly evaluated the effects of sodium-glucose cotransporter-2 inhibitors (SGLT2i) on long COVID, accumulating evidence suggests that these agents possess anti-inflammatory, antioxidative, and endothelial-protective properties. SGLT2 inhibitors suppress NLRP3 inflammasome activation, reduce circulating IL-6 and CRP levels, enhance nitric oxide bioavailability, and improve mitochondrial bioenergetics via AMP-activated protein kinase (AMPK) and sirtuin (SIRT1/3) pathways [17, 18]. These mechanisms may theoretically mitigate post-COVID inflammation and microvascular dysfunction. Therefore, this study aimed to explore whether SGLT2 inhibitor use in patients with type 2 diabetes mellitus (T2DM) was associated with a reduced risk of long COVID.
The management of T2DM involves a combination of lifestyle modifications, glucose monitoring, and pharmacologic therapy aimed at optimizing glycemic control and reducing the risk of complications [19]. Pharmacologic treatment is tailored to individual patient characteristics, with therapeutic agents classified into several major categories, including biguanides, sulfonylureas, thiazolidinediones, dipeptidyl peptidase-4 (DPP-4) inhibitors, glucagon-like peptide-1 receptor agonists (GLP-1 RAs), insulin, and sodium-glucose cotransporter-2 (SGLT2) inhibitors [19–21]. SGLT2 inhibitors, a relatively newer class of antidiabetic medications, have gained increasing clinical interest due to their multifaceted benefits beyond glycemic control. These agents work by inhibiting glucose reabsorption in the renal proximal tubules, thereby promoting glycosuria and reducing hyperglycemia. Additionally, SGLT2 inhibitors have been shown to provide significant cardiovascular and renal protective effects, including reducing heart failure hospitalization and slowing the progression of chronic kidney disease. Moreover, these agents possess anti-inflammatory and endothelial-protective properties, which may have implications for mitigating complications related to infections and systemic inflammation [22–24]. Despite the expanding body of evidence supporting the benefits of SGLT2 inhibitors in T2DM, no study to date has explicitly investigated the potential relationship between SGLT2 inhibitor use and the risk of developing long COVID in T2DM patients. Given the anti-inflammatory and metabolic advantages of SGLT2 inhibitors, we hypothesize that their use may confer protective effects against long COVID in individuals with T2DM. This study aims to explore this novel association, providing critical insights into the role of SGLT2 inhibitors in post-viral sequelae.
Methods
Data sources
This retrospective cohort study utilized the TriNetX analytics platform, a web-based database comprising deidentified electronic health records of more than 250 million patients from multiple countries. The database includes a broad spectrum of information such as demographics, diagnoses (coded using ICD-10-CM), medications (coded using RxNorm or Anatomical Therapeutic Chemical codes), procedures (coded using ICD-10-PCS or Current Procedural Terminology), and laboratory measurements (coded using Logical Observation Identifiers Names and Codes, LOINC). Data for this study were specifically sourced from the USA collaborative network within the TriNetX database, encompassing approximately 110 million patients.
The use of deidentified data in this retrospective analysis exempted the study from Institutional Review Board (IRB) approval for informed consent, as it involved secondary analysis of existing data without intervention or interaction with human subjects, in compliance with the de-identification standard defined in Section § 164.514(a) of the HIPAA Privacy Rule. The process of de-identification was verified through a formal determination by a qualified expert as outlined in Section § 164.514(b)(1) of the HIPAA Privacy Rule. The study was conducted following the Declaration of Helsinki and received approval from the Institutional Review Board of Chung Shan Medical University Hospital (IRB number: CS2-23180).
Study participants
The cohort construction flow chart is depicted in Fig. 1. The study group included individuals aged 18 or older with Type II diabetes (ICD-10-CM = E11) from 2016 to 2019. We then enrolled subjects diagnosed with SARS-CoV-2 infection (COVID-19) between January 1, 2020, and June 30, 2024. This was determined by a positive result from a COVID-19-specific polymerase chain reaction (PCR) test, a positive immunoassay for immunoglobulin in serum or plasma, or a diagnosis of COVID-19 (ICD-10-CM = U07.1) (Supplementary Table 1).
Fig. 1.
Flow-chart of patient selection
The exposure group consisted of those using SGLT2 inhibitors (ATC code: A10BK) 3 months before being diagnosed with COVID-19. The comparison group consisted of individuals who did not use SGLT2 inhibitors within 3 months before or up to 12 months after their COVID-19 diagnosis. The index date was defined as the date of the first COVID-19 diagnosis. Both groups excluded individuals diagnosed with long-COVID-related diseases 3 months before or up to 3 months after the COVID-19 diagnosis and those who died on the same date.
A total of 5,164 patients were included in the exposure group. In the comparison group, 132,161 individuals without the use of SGLT2 inhibitors were identified. Following a 1:1 propensity score matching process based on sex, race, socioeconomic status, body mass index, medical utilization, and comorbidities, 5,162 subjects remained in both the SGLT2 inhibitors and non-SGLT2 inhibitors groups (Fig. 1).
Collection of clinical characters
Baseline characteristics were obtained from records spanning one year before the index date up until one day prior to the index date. Demographic factors of interest included age, sex, race, socioeconomic status, body mass index (BMI), and medical utilization, including ambulatory, emergency, and inpatient encounters. Relevant baseline comorbidities included nicotine dependence, alcohol-related disorders, hypertensive diseases, hyperlipidemia, overweight and obesity, other forms of heart disease, ischemic heart diseases, neoplasms, chronic kidney disease, mental and behavioral disorders due to psychoactive substance use, and cerebrovascular diseases. Medication use encompassed biguanides, glucagon-like peptide-1 receptor agonists (GLP-1 RAs), sulfonylureas, DPP-4 inhibitors, thiazolidinediones, repaglinide/nateglinide (meglitinide analogues), alpha-glucosidase inhibitors, and insulins and analogues (including human and recombinant insulin preparations) (Supplementary Table 2).
To address baseline characteristic differences between the two groups and reduce the impact of confounding factors, propensity score matching was performed using a nearest neighbor greedy matching algorithm with a caliper of 0.1 pooled standard deviations. The matching process was conducted at a 1:1 ratio using the built-in function in TriNetX, considering age, sex, race, socioeconomic status, BMI, medical utilization, comorbidities, and medications.
Primary outcome
The primary outcome of interest in this study was the diagnosis of long COVID, encompassing seven categories of post-COVID-19 conditions: chest/throat pain, abnormal breathing, abdominal symptoms, fatigue, anxiety/depression, pain, headache, cognitive symptoms, and myalgia [25, 26]. The outcome was defined as occurring from the third month after the index date, with a maximum follow-up period of one year. The detailed classification is as follows: the post-COVID-19 condition included unspecified post-COVID-19 conditions and sequelae of other specified infectious and parasitic diseases. The abdominal symptoms included abdominal and pelvic pain, changes in bowel habits, and unspecified diarrhea. The fatigue category included post-viral and related fatigue syndromes, as well as malaise and fatigue. The anxiety/depression category included mood [affective] disorders, anxiety, dissociative, stress-related, somatoform, and other nonpsychotic mental disorders. The headache category included headache, migraine, and other headache syndromes. The cognitive symptoms category included vascular dementia, dementia in other diseases classified elsewhere, unspecified dementia, delirium due to a known physiological condition, other specified mental disorders due to a known physiological condition, Alzheimer’s disease, frontotemporal dementia, neurocognitive disorder with Lewy bodies, mild cognitive impairment of uncertain or unknown etiology, unspecified encephalopathy, somnolence, stupor and coma, other symptoms and signs involving cognitive functions and awareness, dyslexia, and other symbolic dysfunctions not elsewhere classified. Detailed disease codes are listed in Supplementary Table 3.
Sensitivity analysis
To validate the stability of the research findings, three sensitivity analyses were conducted. In the first sensitivity analysis, the definition of the type 2 diabetes mellitus cohort was refined to include both disease diagnosis and the use of hypoglycemic agents, ensuring the accuracy of disease diagnosis. In the second sensitivity analysis, considering that SGLT2 inhibitors are used as second-line therapy for diabetes, the comparison group was composed of patients using DPP-4 inhibitors, another second-line therapy, for comparison. In the third sensitivity analysis, again acknowledging that SGLT2 inhibitors are a second-line therapy, the control group was instead composed of patients using GLP-1 RAs, another second-line therapy, for comparison.
Statistical analysis
All statistical analyses were conducted within the TriNetX platform. The balance of baseline characteristics between the matched cohorts was assessed using standardized mean differences (SMD). Variables with an SMD less than 0.1 were considered well-matched. Cox proportional hazards regression analysis was employed to compare the matched cohorts, providing hazard ratios (HR) along with 95% confidence intervals (CI). Incidence rates for long-COVID were calculated using the Kaplan-Meier method, and the log-rank test was utilized. Subgroup analyses based on age, sex, time period was also conducted. All analyses were conducted using the TrinetX online platform, which employs R 4.0.2 as its background statistical software, alongside Java version 11.0.16 (Oracle Corp) and Python version 3.7 (Python Software Foundation).
Results
Baseline characteristics
Table 1 presents the demographic characteristics of the cohort before and after propensity score matching (PSM). The mean age at index was 61.37 ± 11.28 years in the SGLT2 inhibitor group and 62.14 ± 13.62 years in the non-SGLT2 inhibitor group before matching, with a standardized mean difference (SMD) of 0.061. After matching, the ages were 61.37 ± 11.28 and 61.58 ± 12.28 years, respectively, with an SMD of 0.018. Regarding sex distribution, post-PSM, the percentages of females and males in the SGLT2 inhibitor group were 35.37% and 61.08%, respectively, compared to 34.79% and 61.62% in the non-SGLT2 inhibitor group, with SMDs of 0.012 and 0.011, respectively.
Table 1.
Demographic characteristics of SGLT2 inhibitors and non-SGLT2 inhibitors
| Before PSM | After PSM | |||||
|---|---|---|---|---|---|---|
| SGLT2 inhibitors N = 5164 | Non-SGLT2 inhibitors N = 132,161 | SMD | SGLT2 inhibitors N = 5162 | Non-SGLT2 inhibitors N = 5162 | SMD | |
| Age at Index | 61.37 ± 11.28 | 62.14 ± 13.62 | 0.061 | 61.37 ± 11.28 | 61.58 ± 12.28 | 0.018 |
| Sex | ||||||
| Female | 1826 (35.36) | 63,654 (48.16) | 0.262 | 1826 (35.37) | 1796 (34.79) | 0.012 |
| Male | 3155 (61.10) | 63,717 (48.21) | 0.261 | 3153 (61.08) | 3181 (61.62) | 0.011 |
| Unknown Gender | 183 (3.54) | 4790 (3.62) | 0.004 | 183 (3.55) | 185 (3.58) | 0.002 |
| Race | ||||||
| White | 3262 (63.17) | 79,197 (59.93) | 0.067 | 3260 (63.15) | 3235 (62.67) | 0.010 |
| Black or African American | 728 (14.10) | 24,009 (18.17) | 0.111 | 728 (14.10) | 736 (14.26) | 0.004 |
| Asian | 365 (7.07) | 6589 (4.99) | 0.088 | 365 (7.07) | 350 (6.78) | 0.011 |
| American Indian or Alaska Native | 17 (0.33) | 497 (0.38) | 0.008 | 17 (0.33) | 14 (0.27) | 0.011 |
| Native Hawaiian or Other Pacific Islander | 39 (0.76) | 1524 (1.15) | 0.041 | 39 (0.76) | 28 (0.54) | 0.027 |
| Other Race | 237 (4.59) | 6094 (4.61) | 0.001 | 237 (4.59) | 257 (4.98) | 0.018 |
| Unknown Race | 516 (9.99) | 14,251 (10.78) | 0.026 | 516 (10.00) | 542 (10.50) | 0.017 |
| Social economic status | ||||||
|
Persons with potential health hazards related to socioeconomic and psychosocial circumstances |
42 (0.81) | 1660 (1.26) | 0.044 | 42 (0.81) | 36 (0.70) | 0.013 |
| Problems related to housing and economic circumstances | 20 (0.39) | 601 (0.46) | 0.010 | 20 (0.39) | 19 (0.37) | 0.003 |
| Problems related to employment and unemployment | 10 (0.19) | 143 (0.11) | 0.022 | 10 (0.19) | 10 (0.19) | < 0.001 |
| Body Mass Index | 32.56 ± 6.94 | 32.52 ± 7.49 | 0.006 | 32.56 ± 6.94 | 32.95 ± 7.12 | 0.055 |
| < 30 | 1456 (28.20) | 34,851 (26.37) | 0.041 | 1455 (28.19) | 1449 (28.07) | 0.003 |
| ≥ 30 | 2309 (44.71) | 48,916 (37.01) | 0.157 | 2309 (44.73) | 2321 (44.96) | 0.005 |
| Medical utility | ||||||
| Ambulatory | 4376 (84.74) | 103,428 (78.26) | 0.168 | 4374 (84.74) | 4351 (84.29) | 0.012 |
| Emergency | 644 (12.47) | 23,522 (17.80) | 0.149 | 644 (12.48) | 603 (11.68) | 0.024 |
| Inpatient Encounter | 485 (9.39) | 15,109 (11.43) | 0.067 | 485 (9.40) | 444 (8.60) | 0.028 |
| Comorbidities | ||||||
| Nicotine dependence | 225 (4.36) | 6303 (4.77) | 0.020 | 225 (4.36) | 217 (4.20) | 0.008 |
| Alcohol related disorders | 36 (0.70) | 1395 (1.06) | 0.038 | 36 (0.70) | 35 (0.68) | 0.002 |
| Hypertensive diseases | 3808 (73.74) | 76,273 (57.71) | 0.343 | 3806 (73.73) | 3837 (74.33) | 0.014 |
| Hyperlipidemia, unspecified | 2242 (43.42) | 41,730 (31.58) | 0.246 | 2241 (43.41) | 2210 (42.81) | 0.012 |
| Overweight and obesity | 1466 (28.39) | 27,842 (21.07) | 0.170 | 1466 (28.40) | 1472 (28.52) | 0.003 |
| Other forms of heart disease | 1107 (21.44) | 25,906 (19.60) | 0.045 | 1107 (21.45) | 1071 (20.75) | 0.017 |
| Ischemic heart diseases | 1040 (20.14) | 19,845 (15.02) | 0.135 | 1039 (20.13) | 998 (19.33) | 0.020 |
| Neoplasms | 938 (18.16) | 23,496 (17.78) | 0.010 | 937 (18.15) | 894 (17.32) | 0.022 |
| Chronic kidney disease | 862 (16.69) | 20,865 (15.79) | 0.025 | 862 (16.70) | 857 (16.60) | 0.003 |
| Mental and behavioral disorders due to psychoactive substance use | 269 (5.21) | 8275 (6.26) | 0.045 | 269 (5.21) | 267 (5.17) | 0.002 |
| Cerebrovascular diseases | 241 (4.67) | 6991 (5.29) | 0.029 | 241 (4.67) | 253 (4.90) | 0.011 |
| Medications | ||||||
| Biguanides | 3367 (65.20) | 36,846 (27.88) | 0.807 | 3365 (65.19) | 3479 (67.40) | 0.047 |
| Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) | 1794 (34.74) | 13,187 (9.98) | 0.622 | 1792 (34.72) | 1727 (33.46) | 0.027 |
| Sulfonylureas | 1352 (26.18) | 13,107 (9.92) | 0.433 | 1350 (26.15) | 1263 (24.47) | 0.039 |
| Dipeptidyl peptidase 4 (DPP-4) inhibitors | 787 (15.24) | 6419 (4.86) | 0.351 | 785 (15.21) | 741 (14.36) | 0.024 |
| Thiazolidinediones | 303 (5.87) | 2307 (1.75) | 0.217 | 301 (5.83) | 280 (5.42) | 0.018 |
| Repaglinide | 53 (1.03) | 463 (0.35) | 0.082 | 53 (1.03) | 44 (0.85) | 0.018 |
| Alpha glucosidase inhibitors | 21 (0.41) | 105 (0.08) | 0.066 | 20 (0.39) | 15 (0.29) | 0.017 |
| Nateglinide | 10 (0.19) | 144 (0.11) | 0.022 | 10 (0.19) | 10 (0.19) | < 0.001 |
| Insulins and analogues | 1714 (33.19) | 28,024 (21.20) | 0.272 | 1713 (33.19) | 1718 (33.28) | 0.002 |
SMD: Standardized mean difference
The racial composition showed that, post-PSM, the proportions were nearly identical (63.15% in the SGLT2 inhibitor group and 62.67% in the non-SGLT2 inhibitor group, SMD = 0.010). Other racial distributions, including Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, and Other Race, were balanced after matching. Socioeconomic status indicators, such as potential health hazards related to socioeconomic and psychosocial circumstances, after matching, these values were 0.81% and 0.70%, respectively (SMD = 0.013). Problems related to housing and economic circumstances, employment, and unemployment showed minimal differences before and after matching.
After matching, the BMIs were 32.56 ± 6.94 and 32.95 ± 7.12, respectively (SMD = 0.055). The distribution of BMI categories (< 30 and ≥ 30) was balanced post-PSM. Medical utility patterns indicated that before PSM, 84.74% of patients in the SGLT2 inhibitor group had ambulatory encounters compared to 78.26% in the non-SGLT2 inhibitor group (SMD = 0.168). After matching, these proportions were 84.74% and 84.29%, respectively (SMD = 0.012). Emergency encounters and inpatient encounters also showed balanced distributions post-PSM.
Regarding comorbidities, nicotine dependence was slightly higher in the non-SGLT2 inhibitor group before matching (4.77% vs. 4.36%, SMD = 0.020), but the difference was negligible post-PSM (4.36% vs. 4.20%, SMD = 0.008). Other comorbidities, such as hypertensive diseases, hyperlipidemia, and ischemic heart diseases, showed significant differences before matching but were balanced after matching. Medication use was also balanced post-PSM. Before matching, a higher percentage of patients in the SGLT2 inhibitor group used medications like biguanides (65.20% vs. 27.88%, SMD = 0.807) and GLP-1 RAs (34.74% vs. 9.98%, SMD = 0.622). After matching, the use of these medications was comparable between the two groups.
Risk of long-COVID
Kaplan-Meier analysis with the log-rank test (p < 0.001) demonstrated a lower incidence of long-COVID in the SGLT2 inhibitors group compared to the non-SGLT2 inhibitors group (Fig. 2). Figure 3 illustrates the risk of developing long-COVID in patients exposed to SGLT2 inhibitors compared to those not exposed. The hazard ratio (HR) for long-COVID at 3 to 6 months was 0.82 (95% CI: 0.74–0.91), at 3 to 9 months was 0.83 (95% CI: 0.77–0.89), and at 3 to 12 months was 0.85 (95% CI: 0.79–0.91). These results indicate a consistently lower risk of long-COVID in patients using SGLT2 inhibitors. In the category of long-COVID symptoms such as abdominal symptoms, anxiety/depression, pain, headache, and cognitive symptoms, there were lower risks observed in the SGLT2 inhibitor group.
Fig. 2.
Kaplan-Meier analysis for risk of long-COVID exposed to SGLT2 inhibitors compared to non-SGLT2 inhibitors
Fig. 3.
Forest plot for risk of long-COVID exposed to SGLT2 inhibitors compared to non-SGLT2 inhibitors
Subgroup analysis
Table 2 provides a subgroup analysis for the risk of long-COVID. The analysis showed that patients aged 18–64 years had a lower risk of long-COVID when exposed to SGLT2 inhibitors (HR = 0.76, 95% CI: 0.70–0.84), while those aged 65 years and older had an HR of 0.88 (95% CI: 0.80–0.98). Sex-specific analysis showed that both female and male patients benefited from SGLT2 inhibitors, with HRs of 0.89 (95% CI: 0.80–0.99) and 0.79 (95% CI: 0.72–0.87), respectively. The analysis over different time periods (2020–2022 vs. 2023–2024) showed consistent results, with a lower risk of long-COVID in the SGLT2 inhibitor group across both periods.…
Table 2.
Subgroup analysis for risk of long-COVID exposed to SGLT2 inhibitors compared to non-SGLT2 inhibitors
| SGLT2 inhibitors | Non-SGLT2 inhibitors | HR (95% C.I.) | SGLT2 inhibitors | Non-SGLT2 inhibitors | HR (95% C.I.) | ||
|---|---|---|---|---|---|---|---|
| Age | 18–64, N = 3046 | ≥ 65, N = 2124 | |||||
| Long-COVID | 850 | 1053 | 0.76 (0.70–0.84) | 667 | 735 | 0.88 (0.80–0.98) | |
| Post COVID-19 condition | 10 | 12 | 0.75 (0.32–1.78) | 10 | 10 | 1.42 (0.45–4.48) | |
| Chest/Throat pain | 164 | 197 | 0.83 (0.67–1.02) | 133 | 126 | 1.09 (0.85–1.39) | |
| Abnormal breathing | 167 | 194 | 0.86 (0.70–1.06) | 172 | 195 | 0.90 (0.73–1.10) | |
| Abdominal symptoms | 212 | 255 | 0.82 (0.69–0.99) | 142 | 157 | 0.92 (0.73–1.15) | |
| Fatigue | 133 | 152 | 0.87 (0.69–1.10) | 119 | 159 | 0.76 (0.60–0.96) | |
| Anxiety/Depression | 267 | 355 | 0.74 (0.63–0.86) | 153 | 211 | 0.72 (0.59–0.89) | |
| Pain | 193 | 286 | 0.66 (0.55–0.79) | 178 | 211 | 0.85 (0.70–1.04) | |
| Headache | 84 | 134 | 0.62 (0.47–0.82) | 33 | 61 | 0.55 (0.36–0.83) | |
| Cognitive symptoms | 36 | 53 | 0.68 (0.44–1.03) | 81 | 91 | 0.90 (0.67–1.22) | |
| Myalgia | 31 | 45 | 0.69 (0.44–1.09) | 31 | 33 | 0.96 (0.59–1.57) | |
| Sex | Female, N = 1821 | Male, N = 3153 | |||||
| Long-COVID | 627 | 692 | 0.89 (0.80–0.99) | 834 | 1003 | 0.79 (0.72–0.87) | |
| Post COVID-19 condition | 10 | 10 | 0.88 (0.29–2.60) | 10 | 10 | 1.25 (0.50–3.18) | |
| Chest/Throat pain | 111 | 134 | 0.84 (0.66–1.09) | 174 | 185 | 0.94 (0.76–1.15) | |
| Abnormal breathing | 133 | 144 | 0.93 (0.74–1.18) | 194 | 230 | 0.84 (0.69–1.02) | |
| Abdominal symptoms | 158 | 165 | 0.97 (0.78–1.21) | 185 | 213 | 0.86 (0.71–1.05) | |
| Fatigue | 97 | 120 | 0.82 (0.63–1.07) | 143 | 161 | 0.89 (0.71–1.11) | |
| Anxiety/Depression | 206 | 276 | 0.74 (0.62–0.89) | 193 | 279 | 0.68 (0.57–0.82) | |
| Pain | 149 | 185 | 0.81 (0.66–1.01) | 207 | 283 | 0.72 (0.60–0.86) | |
| Headache | 67 | 101 | 0.67 (0.49–0.91) | 47 | 78 | 0.60 (0.42–0.86) | |
| Cognitive symptoms | 46 | 58 | 0.80 (0.55–1.18) | 68 | 102 | 0.66 (0.49–0.90) | |
| Myalgia | 24 | 29 | 0.84 (0.49–1.44) | 33 | 37 | 0.89 (0.56–1.43) | |
| Time period | 2020/01/01–2022/12/31, N = 4248 | 2023/01/01–2024/06/30, N =1449 | |||||
| Long-COVID | 1355 | 1551 | 0.82 (0.77–0.89) | 340 | 382 | 0.84 (0.72–0.97) | |
| Post COVID-19 condition | 15 | 20 | 0.74 (0.38–1.45) | 10 | 10 | 0.72 (0.16–3.23) | |
| Chest/Throat pain | 282 | 274 | 1.02 (0.86–1.20) | 51 | 71 | 0.69 (0.48–0.98) | |
| Abnormal breathing | 317 | 340 | 0.92 (0.79–1.07) | 71 | 72 | 0.95 (0.69–1.32) | |
| Abdominal symptoms | 335 | 375 | 0.88 (0.76–1.02) | 77 | 81 | 0.91 (0.67–1.25) | |
| Fatigue | 234 | 276 | 0.83 (0.70–0.99) | 51 | 72 | 0.68 (0.47–0.97) | |
| Anxiety/Depression | 392 | 501 | 0.76 (0.67–0.87) | 85 | 120 | 0.68 (0.51–0.89) | |
| Pain | 340 | 429 | 0.78 (0.67–0.89) | 84 | 90 | 0.90 (0.67–1.21) | |
| Headache | 106 | 157 | 0.66 (0.52–0.85) | 23 | 35 | 0.63 (0.37–1.06) | |
| Cognitive symptoms | 96 | 139 | 0.68 (0.52–0.88) | 31 | 33 | 0.91 (0.56–1.48) | |
| Myalgia | 63 | 70 | 0.89 (0.63–1.25) | 11 | 18 | 0.59 (0.28–1.25) | |
If the patient’s count is 1–10, the results indicate a count of 10
Sensitivity analysis
Supplementary Table 4 presents the results of the first sensitivity analysis, where the DM cohort was defined using disease diagnosis and medication criteria. The study results show that the SGLT2 inhibitor group similarly exhibited significantly lower long-COVID risk, with an HR of 0.84 (95% CI = 0.78–0.91). In the category of long-COVID symptoms such as anxiety/depression, pain, and headache, lower risks were observed in the SGLT2 inhibitor group.
Supplementary Tables 5 and 6 present the second and third sensitivity analysis results, respectively, with the comparator groups being DPP4 inhibitors and GLP-1 RAs. The study results indicate that in comparisons with different second-line medications, SGLT2 inhibitors still conducted lower long-COVID risks, with HRs of 0.83 (95% CI = 0.70–0.98) and 0.78 (95% CI = 0.67–0.91), respectively.
Discussion
Long COVID affects 10% to 30% of those recovering from acute COVID-19, with a higher prevalence of 25% to 40% in individuals with type 2 diabetes mellitus (T2DM) [4, 5, 27, 28]. Common symptoms include fatigue (50–80%), dyspnea (20–30%), cognitive issues (30–50%), and cardiovascular complications in about 15% of cases [7, 29]. Treatment for T2DM is personalized and may include biguanides, sulfonylureas, thiazolidinediones, DPP-4 inhibitors, GLP-1 RAs, insulin, and SGLT2 inhibitors [19–21, 30]. Metformin, a biguanide medication, is the first-line treatment for diabetes due to its proven ability to lower blood glucose levels by reducing hepatic glucose production and enhancing insulin sensitivity, all while maintaining a strong safety profile. Sulfonylureas stimulate insulin secretion by closing ATP-sensitive potassium channels in pancreatic β cells, while thiazolidinediones activate peroxisome proliferator-activated receptor-gamma (PPAR-γ) to improve insulin sensitivity. These medications are less commonly used today because sulfonylureas can cause hypoglycemia, and thiazolidinediones may lead to weight gain. DPP-4 inhibitors effectively prolong incretin hormone activity, enhancing glucose-dependent insulin secretion, whereas GLP-1 RAs mimic GLP-1 to promote insulin secretion and slow gastric emptying. Both classes offer significant cardiovascular benefits and aid in weight loss. Insulin therapy is typically reserved for advanced cases or patients who do not achieve adequate glycemic control with oral medications [20]. SGLT2 inhibitors prevent renal glucose reabsorption [23, 31]. The utilization of SGLT2 inhibitors in T2DM management has been increasing steadily over the past decade. Clinical guidelines from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) now recommend SGLT2 inhibitors as a preferred treatment option, particularly in patients with cardiovascular disease, chronic kidney disease, or heart failure [32, 33]. Our study found that SGLT2 inhibitors significantly reduce the risk of long COVID, with HRs of 0.82 to 0.85. This effect held true across various age and sex groups and was confirmed by comparisons with other antihyperglycemic agents. PSM showed balanced baseline characteristics between groups, minimizing confounding factors. These findings suggest that SGLT2 inhibitors may play a dual role in managing T2DM by not only improving glycemic control but also potentially mitigating post-COVID sequelae. Given the high prevalence of long COVID among diabetic patients, clinicians may consider SGLT2 inhibitors as a preferred option in patients at risk for post-viral complications, particularly individuals with additional risk factors such as obesity and cardiovascular disease. These medications could positively influence inflammation, mitochondrial function, and vascular health. Additionally, due to their anti-inflammatory properties, SGLT2 inhibitors may assist in alleviating the persistent effects of SARS-CoV-2 infection.
In epidemiological studies, baseline characteristics often differ between treatment groups due to biases such as confounding by indication [34, 35]. In our research, we found significant disparities between patients prescribed SGLT2 inhibitors and those who were not, particularly regarding age, sex, BMI, and comorbidities. These imbalances can result from clinicians’ preferences in prescribing SGLT2 inhibitors to younger patients or specific health profiles. A study comparing cardiovascular outcomes of SGLT2 inhibitors and sulfonylureas indicated that patients starting on SGLT2 inhibitors were generally younger and had a higher prevalence of comorbidities before matching [36]. To address these discrepancies, we employed PSM, which creates a balanced cohort by matching patients based on observed covariates [37]. After applying PSM, we achieved well-balanced baseline characteristics between the SGLT2 inhibitor and non-SGLT2 inhibitor groups, enhancing the validity of our findings. Similarly, PSM effectively balanced variables in studies assessing the cardioprotective effects of SGLT2 inhibitors. However, PSM only accounts for measured variables, meaning unmeasured confounders could still bias results. Thus, while PSM improves comparability, caution is needed in interpreting findings, and further corroboration through diverse methodologies is advisable.
After PSM matching, we observed that the use of SGLT2 inhibitors in patients with T2DM was associated with a modest reduction in the risk of developing long COVID compared to non-SGLT2 inhibitor users. Although the overall HRs were close to unity (HR = 0.94, 95% CI: 0.87–1.00 for six months; HR = 0.92, 95% CI: 0.84–1.00 for three to six months), specific long COVID symptoms such as fatigue, anxiety/depression, headache, cognitive symptoms, and myalgia were significantly less prevalent among SGLT2 inhibitor users. Additionally, a lower mortality rate was observed in this group (HR = 0.68, 95% CI: 0.50–0.92), suggesting potential survival benefits. Several mechanisms may underlie these observations. Firstly, SGLT2 inhibitors have demonstrated anti-inflammatory properties. A systematic review and meta-analysis indicated that SGLT2 inhibition reduces inflammatory markers, including interleukin-6, tumor necrosis factor-α, and C-reactive protein. Chronic inflammation and immune dysregulation are key features of long-COVID. SGLT2 inhibitors suppress the NLRP3 inflammasome, a major driver of IL-1β and IL-18 production, reducing systemic inflammation and endothelial dysfunction [17, 38]. In addition, these inhibitors decrease pro-inflammatory cytokines such as IL-6, TNF-α, and MCP-1, which are highly elevated in T2DM and long-COVID patients [38, 39]. Modulation of CD4 + and CD8 + T cell responses further contributes to restoring immune homeostasis, preventing prolonged hyperinflammation [16, 39]. SGLT2 inhibitors may exert protective effects against long-COVID through multiple biological pathways. Mechanistically, these agents inhibit the NLRP3 inflammasome and downregulate proinflammatory cytokines including interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP), thereby alleviating systemic inflammation that sustains long-COVID symptoms [38, 39]. In addition, SGLT2i improve mitochondrial biogenesis and oxidative metabolism through activation of AMP-activated protein kinase (AMPK) and sirtuin (SIRT1/3) signaling, enhancing cellular energy efficiency and reducing oxidative stress [40]. These mechanisms may help relieve post-viral fatigue and skeletal muscle weakness, which are among the most frequent clinical features of long-COVID. Moreover, SGLT2i enhance endothelial nitric oxide bioavailability, reduce vascular inflammation, and suppress platelet activation, thereby improving endothelial function and preventing microthrombotic complications that contribute to cardiovascular and neurological sequelae [38, 39]. Importantly, SGLT2 inhibitors also possess neuroprotective and anti-inflammatory properties within the central nervous system by modulating glial activation and restoring mitochondrial redox balance, which may mitigate cognitive impairment (“brain fog”) and neuropsychiatric manifestations such as anxiety and depression [41]. Collectively, these multifaceted mechanisms suggest that SGLT2 inhibitors can counteract the chronic inflammation, endothelial dysfunction, and mitochondrial dysregulation that drive long-COVID pathogenesis. Secondly, the reduction in fatigue and cognitive symptoms may be related to the impact of SGLT2 inhibitors on oxidative stress and mitochondrial function. SGLT2 inhibitors promote ketogenesis, increasing circulating β-hydroxybutyrate (BHB), which serves as an alternative energy source and reduces oxidative stress and systemic inflammation [18, 40]. This metabolic shift enhances mitochondrial efficiency, alleviating post-viral fatigue and myopathy frequently seen in long-COVID [17, 18]. Activation of AMP-activated protein kinase (AMPK) and sirtuins (including SIRT1 and SIRT3) further enhances mitochondrial biogenesis and autophagy, mitigating oxidative damage and neuroinflammatory stress, which are implicated in post-COVID cognitive dysfunction [42, 43]. Research has shown that SGLT2 inhibitors can improve mitochondrial efficiency and reduce oxidative damage, which are factors associated with fatigue and cognitive impairment in long COVID patients [18, 40]. Thirdly, SGLT2 inhibitors confer cardiovascular and microvascular benefits. They enhance endothelial function, increase nitric oxide bioavailability, reduce vascular resistance and oxidative stress, thus ameliorating microvascular dysfunction that contributes to fatigue, dyspnea, and neurocognitive impairment in long-COVID [42, 44]. Additionally, they exert anti-thrombotic effects by reducing platelet activation and fibrinogen levels, lowering the risk of microthrombosis, which is a hallmark of COVID-19-associated vascular complications [45]. In addition, these effects may alleviate symptoms such as headache and myalgia associated with long COVID. Furthermore, the neuroprotective effects of SGLT2 inhibitors may play a role in reducing the incidence of anxiety, depression, and cognitive impairment. Studies have suggested that SGLT2 inhibitors exert beneficial effects by modulating neuroinflammatory pathways and improving brain energy metabolism, thereby mitigating neuropsychiatric symptoms associated with long-COVID [46]. In conclusion, our findings indicate that SGLT2 inhibitors may provide a modest protective effect against long-COVID in patients with T2DM. These medications can help alleviate symptoms like fatigue, anxiety, and cognitive impairment, potentially reducing mortality. The present study provides compelling evidence that the use of SGLT2 inhibitors is associated with a reduced risk of long-COVID in patients with T2DM after PSM. Subgroup analyses further reveal that the protective effect of SGLT2 inhibitors varies by age, sex, and disease onset period, suggesting that multiple factors influence the observed outcomes. Younger patients (18–64 years) showed significant risk reduction (HR = 0.76) in long-COVID outcomes, particularly in cognitive dysfunction, fatigue, and neuroinflammatory symptoms. This is likely due to enhanced metabolic resilience and immune plasticity, which amplify the anti-inflammatory and mitochondrial benefits of SGLT2 inhibitors [47, 48]. In contrast, older individuals (≥ 65 years) exhibited minimal risk reduction (HR = 0.88), potentially due to advanced endothelial dysfunction and reduced metabolic flexibility, limiting the effectiveness of SGLT2 inhibitors in preventing long-term complications [49]. The reduction in long-COVID risk observed in both younger and older populations suggest a broad applicability across different age groups, which is critical given the heterogeneity in long-COVID presentation. Male patients benefited more from SGLT2 inhibition, significantly reducing mortality (HR = 0.79) and lower incidence of neuropsychiatric sequelae. This may be attributed to greater baseline endothelial dysfunction and oxidative stress in males, which SGLT2 inhibitors effectively counteract [41]. Female patients exhibited a more modest reduction in long-COVID risk (HR = 0.89), possibly due to endogenous estrogen-mediated vascular protection, which may overshadow the cardiovascular benefits of SGLT2 inhibitors due to differences in metabolic or immune system regulation. Notably, the findings remained consistent across study periods (2020–2022 vs. 2023–2024), reinforcing the robustness of the association despite changes in viral variants, vaccination rates, and treatment strategies over time. The efficacy of SGLT2 inhibitors appeared to vary depending on the COVID-19 variant period. Infections with earlier variants (Alpha, Delta) were associated with more severe systemic inflammation and endothelial injury, where SGLT2 inhibitors provided substantial protection against vascular damage and immune dysregulation [50]. In contrast, later variants (Omicron), which predominantly affect the upper respiratory tract, showed reduced benefit from SGLT2 inhibition, suggesting that its effects may be more relevant in preventing severe inflammatory and cardiovascular complications rather than mild respiratory syndromes [50]. This study demonstrates that SGLT2 inhibitors reduce long COVID risk in T2DM patients, with stronger benefits in younger individuals and males. Younger patients showed significant reductions in cognitive dysfunction and fatigue, while older patients had minimal benefit, likely due to endothelial dysfunction. Males benefited more, possibly due to higher baseline oxidative stress. The protective effect remained consistent across study periods, highlighting SGLT2 inhibitors as a promising strategy for long-COVID prevention, particularly in younger and male patients.
The first sensitivity analysis of the T2DM cohort, which required both diagnosis and medication use, found that SGLT2 inhibitors were linked to a lower risk of long-COVID (HR = 0.84, 95% CI = 0.78–0.91). This association extended to symptoms like anxiety and depression, suggesting pharmacologic mechanisms at play, such as reduced chronic inflammation and improved endothelial function. In a second analysis comparing SGLT2 inhibitors to DPP-4 inhibitors, SGLT2 inhibitors also showed a lower long-COVID risk (HR = 0.83, 95% CI = 0.70–0.98). SGLT2 inhibitors provide systemic anti-inflammatory benefits and enhance vascular health, which may be crucial in preventing long-COVID complications. The third analysis, comparing SGLT2 inhibitors with GLP-1 RAs, again found SGLT2 inhibitors linked to a lower risk of long-COVID (HR = 0.78, 95% CI = 0.67–0.91). Compared to GLP-1 RAs, which primarily exert metabolic effects through incretin pathways, SGLT2 inhibitors may provide additional benefits through systemic anti-inflammatory effects and endothelial protection. This distinction may explain the observed greater reduction in long COVID risk with SGLT2 inhibitors. Compared to GLP-1 RAs, which primarily exert metabolic effects through incretin pathways, SGLT2 inhibitors may provide additional benefits through systemic anti-inflammatory effects and endothelial protection. This distinction may explain the observed greater reduction in long COVID risk with SGLT2 inhibitors. These findings reinforce the idea that SGLT2 inhibitors may offer significant protective effects through anti-inflammatory, endothelial, and metabolic benefits. Future research should explore these mechanisms further and identify which patients benefit most from SGLT2 inhibition.
This study indicates that SGLT2 inhibitors may reduce the risk of long COVID in patients with T2DM, but several limitations should be acknowledged. The retrospective design restricts the ability to draw causal conclusions, and unmeasured factors such as diet, physical activity, and medication adherence could introduce bias. Reliance on TriNetX electronic health records may lead to misclassification of long COVID cases, making it essential for future research to incorporate patient-reported outcomes and standardized diagnostic criteria. Additionally, key epidemiological factors like vaccination status and antiviral therapy, which are known to influence COVID-19 outcomes, were not included in this analysis. The variability among SARS-CoV-2 variants raises further questions about the effectiveness of SGLT2 inhibitors against different strains. The findings may specifically apply to T2DM patients, suggesting a need to explore the relevance of these results for individuals with prediabetes or metabolic syndrome. Finally, the lack of biomarker data limits the understanding of the mechanisms at play, underscoring the need for future studies to include inflammatory and endothelial markers. Despite these limitations, the study highlights the potential role of SGLT2 inhibitors in mitigating long COVID, warranting further research and validation.
Conclusions
Our findings indicate that SGLT2 inhibitors may have a significant role in mitigating the risk of long COVID among patients with T2DM, particularly with respect to symptoms such as abdominal discomfort, anxiety/depression, pain, headache, and cognitive symptoms. This study provides substantial epidemiological evidence; however, further randomized controlled trials are essential to validate these associations and to ascertain whether SGLT2 inhibitors should be recommended for the prevention of long COVID in high-risk populations. Moreover, the dynamic nature of SARS-CoV-2 variants underscores the need for ongoing research to evaluate the persistence of these protective effects across different viral strains and vaccination statuses.
Supplementary Information
Acknowledgements
The authors sincerely thank all the participants and professionals who contributed to the TriNetX network.
Abbreviations
- COVID-19
Coronavirus disease 2019
- DPP-4
Inhibitors dipeptidyl peptidase-4 inhibitors
- GLP-1
RAs glucagon-like peptide-1 receptor agonists
- PSM
Propensity score matching
- SGLT2
Inhibitor sodium-glucose cotransporter 2 inhibitor
- SMD
Standardized mean differences
- T2DM
Type 2 diabetes mellitus
Author contributions
**Han-Wei Yeh: ** This author designed the study, search the literature, responsible for data interpretation, prepare the manuscript draft and was approved the final version of the manuscript. **Chung-Hsien Chaou: ** This author designed the study, was responsible for data interpretation and approved the final version of the manuscript. **Shun-Fa Yang: ** This author designed the study and approved the final version of the manuscript. **Yu-Hsun Wang: ** This author was responsible for data collection, data analysis, performed the statistical analyses and approved the final version of the manuscript. **Yu-Hsiang Kuan: ** This author designed the study, search the literature, responsible for data interpretation, prepare the manuscript draft and was approved the final version of the manuscript. **Chao-Bin Yeh: ** This author designed the study, was responsible for data interpretation; prepare the manuscript and approved the final version of the manuscript.
Funding
This research was funded by grants from the Chung Shan Medical University Hospital (CSH-2025-D-005).
Data availability
The findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the principles of the Declaration of Helsinki and received approval from the Institutional Review Board of Chung Shan Medical University Hospital (IRB number: CS2-23180). Consent to participate was waived, as TriNetX had obtained a waiver of informed consent from the Western Institutional Review Board. All data and statistical summaries were presented in aggregated form, and no identifiable personal information was accessible.
Consent for publication
All data were anonymized, and no personally identifiable information was accessible or shared. Only aggregated results were reported; therefore, informed consent was waived.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Yu-Hsiang Kuan, Email: kuanyh001@gmail.com.
Chao-Bin Yeh, Email: sky5ff@gmail.com.
References
- 1.Del Rio C, Collins LF, Malani P. Long-term health consequences of COVID-19. JAMA. 2020;324(17):1723–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Leung TYM, et al. Short- and potential long-term adverse health outcomes of COVID-19: a rapid review. Emerging Microbes & Infections. 2020;9(1):2190–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Centers for Disease Control and Prevention. Long COVID basics. February 17, 2025]; Available from: https://www.cdc.gov/covid/long-term-effects/index.html
- 4.Organisation for Economic Co-operation and Development. The impacts of long COVID across OECD countries. 2024 February 17, 2025]; Available from: https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/06/the-impacts-of-long-covid-across-oecd-countries_f662b21c/8bd08383-en.pdf
- 5.Davis HE, et al. Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol. 2023;21(3):133–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Robertson MM, et al. The epidemiology of long coronavirus disease in US adults. Clin Infect Dis. 2023;76(9):1636–45. [DOI] [PubMed] [Google Scholar]
- 7.Wong MC, et al. Epidemiology, symptomatology, and risk factors for long COVID symptoms: population-based, multicenter study. JMIR Public Health Surveill. 2023;9:e42315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Heald AH, et al. The prevalence of long COVID in people with diabetes mellitus-evidence from a UK cohort. EClinicalMedicine. 2024;71:102607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Norouzi M, et al. Type-2 diabetes as a risk factor for severe COVID-19 infection. Microorganisms. 2021. 10.3390/microorganisms9061211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kazakou P, et al. Diabetes and COVID-19; a bidirectional interplay. Front Endocrinol (Lausanne). 2022;13:780663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Man DE, et al. Insulin resistance in long COVID-19 syndrome. J Pers Med. 2024. 10.3390/jpm14090911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Varghese NE, et al. Unraveling the nexus: Post-COVID hyperglycaemia and diabetes-a comprehensive review of the intricate interplay and implications for glycemic control. J Young Pharm. 2024;16(3):456–60. [Google Scholar]
- 13.Aljadah M, et al. Clinical implications of COVID-19-related endothelial dysfunction. JACC: Advances. 2024;3(8):101070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wu X, et al. Damage to endothelial barriers and its contribution to long COVID. Angiogenesis. 2024;27(1):5–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yanai H, et al. The significance of endothelial dysfunction in long COVID-19 for the possible future pandemic of chronic kidney disease and cardiovascular disease. Biomolecules. 2024. 10.3390/biom14080965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Low RN, Low RJ, Akrami A. A review of cytokine-based pathophysiology of long COVID symptoms. Front Med (Lausanne). 2023;10:1011936. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kim SR, et al. SGLT2 inhibition modulates NLRP3 inflammasome activity via ketones and insulin in diabetes with cardiovascular disease. Nat Commun. 2020;11(1):2127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mone P, et al. Effects of SGLT2 inhibition via empagliflozin on cognitive and physical impairment in frail diabetic elders with chronic kidney disease. Pharmacol Res. 2024;200:107055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Deol R, Bashir DS. Exploring the complications of TYPE 2 DIABETES MELLITUS: pathophysiology and management strategies. EPRA Int J Res Dev (IJRD). 2024;9(7):1–10. [Google Scholar]
- 20.Kalra S, Aggarwal S, Khandelwal D. Thyroid dysfunction and type 2 diabetes mellitus: screening strategies and implications for management. Diabetes Ther. 2019;10(6):2035–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nitzke D, et al. Increasing dietary fiber intake for type 2 diabetes mellitus management: a systematic review. World J Diabetes. 2024;15(5):1001–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Fu WJ, et al. Emerging role of antidiabetic drugs in cardiorenal protection. Front Pharmacol. 2024;15:1349069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hsia DS, Grove O, Cefalu WT. An update on sodium-glucose co-transporter-2 inhibitors for the treatment of diabetes mellitus. Curr Opin Endocrinol Diabetes Obes. 2017;24(1):73–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Takahashi Y, et al. Long-term benefit of SGLT2 inhibitors to prevent heart failure hospitalization in patients with diabetes, with potential time-varying benefit. Clin Transl Sci. 2024;17(12):e70088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pfaff ER, et al. Coding long COVID: characterizing a new disease through an ICD-10 lens. BMC Med. 2023;21(1):58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Taquet M, et al. Incidence, co-occurrence, and evolution of long-COVID features: a 6-month retrospective cohort study of 273,618 survivors of COVID-19. PLoS Med. 2021;18(9):e1003773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Cohen J, van der Meulen Y, Rodgers. An intersectional analysis of long COVID prevalence. Int J Equity Health. 2023;22(1):261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Coste J, et al. Difference in long COVID prevalence due to different definitions and long COVID related risk factors. Eur J Public Health. 2024;34(Suppl 3):ckae144811. [Google Scholar]
- 29.Mudgal SK, et al. Pooled prevalence of long COVID-19 symptoms at 12 months and above follow-up period: a systematic review and meta-analysis. Cureus. 2023;15(3):e36325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lipscombe L, et al. Pharmacologic glycemic management of type 2 diabetes in adults: 2020 update. Can J Diabetes. 2020;44(7):575–91. [DOI] [PubMed] [Google Scholar]
- 31.Farazul H, et al. Navigating the therapeutic landscape of SGLT2 inhibitors in diabetes management: exploring efficacy and emerging concerns. Exploration of Medicine; 2024.
- 32.Adhikari R, et al. National trends in use of Sodium-Glucose Cotransporter-2 inhibitors and Glucagon-like Peptide-1 receptor agonists by cardiologists and other specialties, 2015 to 2020. J Am Heart Assoc. 2022;11(9):e023811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Seidu S, et al. SGLT2 inhibitors - the new standard of care for cardiovascular, renal and metabolic protection in type 2 diabetes: a narrative review. Diabetes Ther. 2024;15(5):1099–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chang HC, et al. Sodium glucose transporter 2 inhibitors versus Metformin on cardiovascular and renal outcomes in patients with diabetes with low cardiovascular risk: a nationwide cohort study. J Am Heart Assoc. 2024;13(8):e032397. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ueda P, et al. Sodium glucose cotransporter 2 inhibitors and risk of serious adverse events: nationwide register based cohort study. BMJ. 2018;363:k4365. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Dave CV, et al. Risk of cardiovascular outcomes in patients with type 2 diabetes after addition of SGLT2 inhibitors versus sulfonylureas to baseline GLP-1RA therapy. Circulation. 2021. 10.1161/CIRCULATIONAHA.120.047965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Das SR, et al. 2020 expert consensus decision pathway on novel therapies for cardiovascular risk reduction in patients with type 2 diabetes: a report of the American College of Cardiology solution set oversight committee. J Am Coll Cardiol. 2020;76(9):1117–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wang D, et al. The effect of sodium-glucose cotransporter 2 inhibitors on biomarkers of inflammation: a systematic review and meta-analysis of randomized controlled trials. Front Pharmacol. 2022;13:1045235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Scisciola L, et al. Anti-inflammatory role of SGLT2 inhibitors as part of their anti-atherosclerotic activity: data from basic science and clinical trials. Front Cardiovasc Med. 2022;9:1008922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Koutentakis M. The ketogenic effect of SGLT-2 inhibitors—beneficial or harmful? J Cardiovasc Dev Dis. 2023. 10.3390/jcdd10110465. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Pawlos A, et al. Neuroprotective effect of SGLT2 inhibitors. Molecules. 2021. 10.3390/molecules26237213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Sotoudeheian M. Targeting SIRT1 by scopoletin to inhibit XBB.1.5 COVID-19 lifecycle. Current Reviews in Clinical and Experimental Pharmacology. 2025;20(1):4–13. [DOI] [PubMed] [Google Scholar]
- 43.Zhang P, Zhao T, Zhou W. The clinical significance of SIRT3 in COVID-19 patients: a single center retrospective analysis. Ann Clin Lab Sci. 2021;51(5):686–93. [PubMed] [Google Scholar]
- 44.Mroueh A, et al. SGLT2 expression in human vasculature and heart correlates with low-grade inflammation and causes eNOS-NO/ROS imbalance. Cardiovasc Res. 2024. 10.1093/cvr/cvae257. [DOI] [PubMed] [Google Scholar]
- 45.Stanger L, et al. Comparison of the anti-platelet and anti-thrombotic effects of the dual SGLT1/2 inhibitor Sotagliflozin to the relatively selective SGLT2 inhibitor empagliflozin. Blood. 2024;144(Supplement 1):3933–3933. [Google Scholar]
- 46.Xu SW, Ilyas I, Weng JP. Endothelial dysfunction in COVID-19: an overview of evidence, biomarkers, mechanisms and potential therapies. Acta Pharmacol Sin. 2023;44(4):695–709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Goedeke L, et al. SGLT2 inhibition alters substrate utilization and mitochondrial redox in healthy and failing rat hearts. J Clin Invest. 2024. 10.1172/JCI176708. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Troise D, et al. mTOR and SGLT-2 inhibitors: their synergistic effect on age-related processes. Int J Mol Sci. 2024. 10.3390/ijms25168676. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Huerne K, et al. Epidemiological and clinical perspectives of long COVID syndrome. American Journal of Medicine Open. 2023;9:100033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Chu RYK, et al. A global scale COVID-19 variants time-series analysis across 48 countries. Front Public Health. 2023;11:1085020. [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 findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.