Association of Baseline HbA_1c With Cardiovascular and Renal Outcomes: Analyses From DECLARE-TIMI 58

HbA_1c is an important indicator of glycemic control, providing an estimate of the glycemic burden in the 3 months before the test (1). Interventional trials comparing intensive versus conventional glycemic control in patients with type 2 diabetes demonstrated a robust decline in the occurrence of microvascular complications, predominantly kidney and eye outcomes (2). However, the clinical impact of HbA_1c reduction on the risk of macrovascular diabetic complications has been a source of much deliberation (3–5). While a short-term benefit of intensive glycemic control has not been demonstrated in individual studies, meta-analyses have shown a 15% reduced risk for myocardial infarction, and long-term observation has indicated some benefit when the intervention was implemented at disease onset (6,7). It has been proposed that long-standing, poorly controlled diabetes establishes a negative glycemic legacy, which cannot be easily reversed, and therefore, ameliorating hyperglycemia after the complications have already set in has limited benefit and may possibly pose some risk (8). Setting higher glycemic targets in patients with multiple comorbidities and long-standing diabetes has thus been recommended (9).

Patients with diabetes are at increased risk of macrovascular complications, these being a major cause of morbidity and mortality (10). The cardiovascular impact of the individual glucose-lowering agents has been explored in recent years. While dipeptidyl peptidase 4 inhibitors demonstrated an overall neutral effect, marked cardiovascular benefits have been demonstrated with sodium glucose cotransporter 2 (SGLT2) inhibitors and glucagon-like peptide 1 receptor agonists (11). The cardiovascular benefits observed with these agents are generally attributed to their pleiotropic effects, although a contributory benefit of HbA_1c reduction per se has been suggested (12–14). These findings have led current guidelines to propose prescribing glucagon-like peptide 1 receptor agonists or SGLT2 inhibitors to patients with type 2 diabetes and established or at high risk for atherosclerotic cardiovascular disease (ASCVD) to improve their outcomes (15,16). These recommendations are irrespective of the prevailing HbA_1c and include those with HbA_1c <7%.

The Dapagliflozin Effect on Cardiovascular Events trial (DECLARE-TIMI 58) assessed the cardiovascular and renal outcomes of dapagliflozin versus placebo in a broad population of patients with type 2 diabetes (17). We studied the association of baseline HbA_1c levels with cardiovascular and renal outcomes and in a prespecified analysis, assessed whether the benefits observed with dapagliflozin were consistent across all levels of baseline HbA_1c.

In DECLARE-TIMI 58, 17,160 patients were randomly assigned to receive dapagliflozin 10 mg daily or placebo and followed for a median of 4.2 years. Patients with a HbA_1c of 6.5 to <12.0% and a creatinine clearance of ≥60 mL/min/1.73 m² were eligible for inclusion. The trial included 6,974 patients with established ASCVD (40.6%) and 10,186 with multiple risk factors (MRF) but without ASCVD (59.4%). Participants with MRF were men aged ≥55 years and women aged ≥60 years with at least one additional cardiovascular risk factor, including dyslipidemia, hypertension, or current tobacco use. All patients were treated according to guidelines and regional standards of care for cardiovascular risk factors, including blood pressure, LDL cholesterol, antithrombotic treatment, and HbA_1c. The trial protocol was approved by the institutional review board at each participating site, and all participants provided written informed consent. The design, baseline characteristics, and principal results of this study have been previously published (17–19).

The prespecified dual primary composite efficacy outcomes were cardiovascular death or hospitalization for heart failure (HHF) and major adverse cardiovascular events (MACE), which included the composite of cardiovascular death, myocardial infarction, or ischemic stroke. A prespecified secondary cardiorenal composite outcome included a sustained decrease of ≥40% in eGFR to <60 mL/min/1.73 m², end-stage renal disease, or death as a result of cardiovascular or renal cause. Additional outcomes included components of the primary end points as well as a renal-specific composite outcome that was similar to the cardiorenal outcome excluding cardiovascular death. Analysis of outcomes by HbA_1c categories was prespecified.

Baseline characteristics are reported as frequencies and percentages for categorical variables and as median (interquartile range) for continuous variables by HbA_1c group. P-trend values for categorical variables were calculated from the Cochran-Armitage trend test, and continuous variables were calculated from the Jonckheere-Terpstra test. Baseline and efficacy analyses were performed on an intention-to-treat basis. The risk of outcome by baseline categorical HbA_1c was calculated using Cox regression models with HbA_1c of 7 to <8% set as the reference group, and a P-trend value was calculated for each outcome analyzed. The models were adjusted for age, sex, BMI, diabetes duration, eGFR, history of heart failure, and insulin, metformin, sulfonylurea, and statin use, as well as for randomized treatment arm and ASCVD risk category in models not restricted to a specific treatment arm or risk category. The risk for each outcome was also considered, analyzing HbA_1c as a continuous variable, and adjusted hazard ratios (HRs) with 95% CIs were calculated per 1% higher HbA_1c. An interaction term was included to test for heterogeneity of the effect of HbA_1c on outcomes between the MRF and ASCVD subgroups.

The effect of dapagliflozin on the incidence of the outcomes within each HbA_1c subgroup was calculated using Cox regression models that included the randomization stratification factor of baseline hematuria and the risk category (ASCVD or MRF) for the overall population, and we report the HRs and 95% CIs. When the effect of dapagliflozin within HbA_1c subgroups was considered as a further subset by risk category (ASCVD or MRF), models included only the stratification factor of baseline hematuria status, as described in the study design article (18). To test for heterogeneity of effect, an interaction term was included in the Cox regression model. Probability and 95% CIs of outcomes by HbA_1c were graphed in a continuous model implementing a restricted cubic spline for HbA_1c.

Mixed models for repeated measures in HbA_1c were analyzed to produce least squares mean estimates in each treatment and baseline HbA_1c group. Models included baseline HbA_1c values, hematuria status, risk category, treatment, visit, and the interaction of treatment and visit. Three-way interactions were calculated assessing the interaction of baseline HbA_1c group, visit, and treatment allocation to dapagliflozin versus placebo.

There was no statistical adjustment for multiple comparisons. P < 0.05 was considered statistically significant. All analyses were performed using SAS 9.4 software (SAS Institute, Cary, NC).

Demographic, clinical, and laboratory characteristics of patients by baseline HbA_1c are shown in Supplementary Table 1. With higher baseline HbA_1c, patients were more likely to be younger and female, to have a longer diabetes duration, and to be using insulin or sulfonylureas. They were less likely to be using metformin or statins. Patients with higher HbA_1c were also more likely to have a history of heart failure, be current smokers, and have had a higher eGFR at baseline. The baseline characteristics of patients in the ASCVD and MRF populations are shown in Supplementary Tables 2 and 3, respectively, and followed similar patterns.

In the overall study population, higher baseline HbA_1c was associated with an increased risk of cardiovascular death/HHF (adjusted HR 1.12 [95% CI 1.06–1.19] per 1% higher HbA_1c), driven by an increased risk of cardiovascular death. The risk of MACE and cardiorenal and renal-specific outcomes was greater with higher HbA_1c. The risk for HHF did not differ by baseline HbA_1c (Table 1 and Fig. 1A). Analyses of these outcomes in the subgroups of patients with established ASCVD and with MRF but no established ASCVD revealed some heterogeneity in the association of HbA_1c with outcomes in these two populations (Table 1 and Fig. 1B and C). Specifically, the risk of MACE was greater with higher HbA_1c in the MRF but not the ASCVD subgroup (P-interaction = 0.0064). This was driven by a stronger association of HbA_1c with cardiovascular death and stroke in the MRF than in the ASCVD populations (P-interactions = 0.0670 and 0.0004, respectively). The risk for the cardiorenal and renal-specific outcomes was more prominently increased with higher HbA_1c in the MRF versus ASCVD subgroup (P-interaction = 0.0093 and 0.0704). Of note, the risk of adverse outcomes was higher in the ASCVD subgroup than in the MRF subgroup (Fig. 1B and C).

In the overall study population, dapagliflozin led to a reduction in cardiovascular death/HHF, driven by a reduction in HHF. The cardiorenal and renal-specific outcomes were also markedly reduced with dapagliflozin. These outcomes were consistently reduced irrespective of baseline HbA_1c (P-interaction > 0.05) (Fig. 2 and Supplementary Fig. 1). Similar trends were observed in the ASCVD and MRF subgroups (Supplementary Figs. 2 and 3). Plotting the risk for outcomes by HbA_1c as restricted cubic splines with dapagliflozin versus placebo revealed that the benefits of dapagliflozin were generally maintained across all HbA_1c levels (Fig. 3).

HbA_1c levels following baseline were lower with dapagliflozin versus placebo at all time points in all HbA_1c subgroups (P < 0.01) (Supplementary Fig. 4). In the higher versus lower baseline HbA_1c subgroups, the difference in HbA_1c between dapagliflozin and placebo was greater (P-trend interaction = 0.0006 at month 48).

Our study sheds more light on the association of HbA_1c with cardiovascular and renal outcomes in a broad population of patients with type 2 diabetes and increased risk of cardiovascular disease. Overall, higher baseline HbA_1c was associated with a higher rate of adverse cardiovascular outcomes and with a higher rate of adverse renal outcomes. Notably, the association of HbA_1c with outcomes was stronger in the MRF versus the ASCVD subgroup. The risk for HHF was not modified by baseline HbA_1c overall or in the ASCVD or MRF subgroup. Nonetheless, dapagliflozin reduced the risk of cardiovascular death/HHF, HHF, adverse renal outcomes across all HbA_1c categories with no heterogeneity.

In a large observational study, higher HbA_1c was associated with a greater risk of mortality, heart failure, and macrovascular complications, yet patients with prior stroke or myocardial infarction were excluded from this analysis (20). In the Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus–Thrombolysis in Myocardial Infarction 53 (SAVOR-TIMI 53) trial, higher baseline HbA_1c was associated with a greater risk of macrovascular complications, yet not of HHF (21). In the Trial Evaluating Cardiovascular Outcomes With Sitagliptin (TECOS), baseline and time-updated HbA_1c were each associated with a higher risk of HHF and additional macrovascular outcomes (22). To the contrary, an association between baseline or 1-month HbA_1c levels and the risk of MACE was not found in the Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care (EXAMINE) trial (23). Of note, participants in the EXAMINE trial were of highest cardiovascular risk among these three dipeptidyl peptidase 4 inhibitor cardiovascular outcome studies. These patients were within 15–90 days of an acute coronary syndrome and had the highest rate of three-point MACE among the three trials (24,25). Thus, it is possible that their enhanced cardiovascular risk masked the modest impact of glycemic control (23).

In the current study, higher baseline HbA_1c was associated with a higher risk of MACE in the MRF but not the ASCVD subgroup. A differential association of glycemic control and cardiovascular outcomes has also been reported in a large observational study, whereby the risk of cardiovascular events was lower with lower HbA_1c in patients in the low to moderate comorbidity subgroup but not in the high comorbidity subgroup (26). These data may support the notion that while glycemic control is associated with cardiovascular outcomes in lower-risk patients, the role of glycemia is minimized with increased cardiovascular risk. This further highlights the importance of prescribing drug classes that have an independent effect in reducing cardiovascular risk to high-risk patients with diabetes, irrespective of their prevailing glycemic control.

HHF was not associated with baseline HbA_1c in the overall study population or within the ASCVD or MRF subgroups. These observations are in contrast to multiple observational studies indicating a greater risk of HHF with higher HbA_1c, with some indicating a U-shaped association (20,27,28). It is possible that our cohort of patients may be at higher risk for HHF; thus, while diabetes and poor glycemic control are well-established risk factors for heart failure, once clinical or subclinical heart failure has ensued, the association between glycemia and disease exacerbation is diminished.

The risk of adverse renal outcomes was strongly associated with baseline HbA_1c in the MRF subgroup, yet the magnitude of this association was again much smaller in the ASCVD subgroup. The renal-specific outcome comprised hard renal outcomes, including a sustained decrease of ≥40% in eGFR to <60 mL/min/1.73 m², end-stage renal disease, and renal death. While reducing HbA_1c has been clearly shown to reduce macroalbuminuria, evidence supporting its role in improving eGFR-based renal outcomes has been limited (2,29,30). Improved glycemic control appears to have a role in the primary prevention of diabetic kidney disease, particularly when implemented early (31). While the eGFR in the ASCVD subgroup was only slightly lower than that of the MRF subgroup, and the urinary albumin-to-creatinine ratio somewhat higher (19), it is possible that in the presence of prevailing ASCVD, subclinical chronic kidney disease has already ensued, limiting the association between baseline glycemia on kidney outcomes.

Randomization to dapagliflozin versus placebo reduced the risk of cardiovascular death/HHF, driven by a reduction in HHF, and led to a reduction in the rate of adverse renal outcomes (17). This observation persisted at all levels of baseline HbA_1c, including in those with HbA_1c <7%. No heterogeneity of the effect of dapagliflozin by baseline HbA_1c was noted in the ASCVD or MRF subgroups. Notably, the effect of dapagliflozin on HbA_1c was greater with higher HbA_1c, yet the cardiovascular and renal benefits were consistently maintained irrespective of baseline HbA_1c. Similar results have been noted with empagliflozin and canagliflozin, indicating the modest contributory role, if any, of glycemic control per se on the benefits of SGLT2 inhibitors on cardiovascular and renal outcomes (32–34). In a recently published mediation analysis, a limited role of baseline HbA_1c and its change on the beneficial effects of dapagliflozin on HHF has been demonstrated as well (35). These data support that the cardiac and renal beneficial effect of dapagliflozin persist irrespective of baseline HbA_1c, even in those seemingly on target with an HbA_1c <7%. We have also recently demonstrated that these benefits persist irrespective of baseline glucose-lowering agents (36). Moreover, a reduced risk of HHF and adverse renal outcomes with SGLT2 inhibitors has been recently demonstrated in patients without diabetes, further supporting the HbA_1c-independent benefits of SGLT2 inhibitors (37–39).

Several limitations of our analysis should be considered. The association of HbA_1c with cardiovascular and renal outcomes is post hoc and, thus, should be viewed as hypothesis generating. Moreover, the number of events were quite small in some of the HbA_1c subgroups, particularly in the MRF population, and we did not adjust for multiple comparisons. Finally, only the baseline HbA_1c was considered in these analyses, and we did not account for the impact of HbA_1c changes occurring over time on cardiovascular and renal outcomes.

In conclusion, our study demonstrates an association of high baseline HbA_1c with adverse cardiovascular and renal outcomes, which were mainly observed in the MRF and not the ASCVD subgroup, highlighting the greater importance of glycemic control early on, before complications ensue. However, the beneficial effect of dapagliflozin on HHF and adverse renal outcomes was consistently observed in the overall population and in both subgroups. These data highlight the importance of including SGLT2 inhibitors in the treatment regimen of all patients with type 2 diabetes with or at high risk for ASCVD, irrespective of glycemic control.

Funding and Duality of Interest. The sponsors of DECLARE-TIMI 58 was initially AstraZeneca and Bristol-Myers Squibb, and AstraZeneca later became the sole sponsor of the study. DECLARE-TIMI 58 was a collaboration between the funder and two academic research organizations (TIMI Study Group and Hadassah Medical Organization). A.C. reports grants and personal fees from AstraZeneca and Novo Nordisk and personal fees from Boehringer Ingelheim, Eli Lilly, Pfizer, and Medial EarlySign. S.D.W. discloses grants from AstraZeneca, Bristol-Myers Squibb, Sanofi Aventis, and Amgen; grants and personal fees from Arena, Daiichi Sankyo, Eisai, Eli Lilly, and Janssen; grants and consulting fees from Merck (also his spouse is employed by Merck); and personal fees from Aegerion, Allergan, AngelMed, Boehringer Ingelheim, Boston Clinical Research Institute, Icon Clinical, Lexicon, St. Jude Medical, Xoma, Servier, AstraZeneca, and Bristol-Myers Squibb. O.M. reports grants and personal fees from AstraZeneca, Bristol-Myers Squibb, and Novo Nordisk and personal fees from Eli Lilly, Sanofi, Merck Sharp & Dohme, Boehringer Ingelheim, Johnson & Johnson, and Novartis. S.A.M. and E.L.G. report research grant support through Brigham and Women’s Hospital from Abbott, Amgen, Aralez, AstraZeneca, Bayer, Daiichi Sankyo, Eisai, GlaxoSmithKline, Intarcia, Janssen, MedImmune, Merck, Novartis, Pfizer, Poxel, Quark Pharmaceuticals, Roche, Takeda, The Medicines Company, and Zora Biosciences. D.L.B. discloses the following relationships: advisory boards for Boehringer Ingelheim, Cardax, CellProthera, Cereno Scientific, Elsevier Practice Update Cardiology, Janssen, Level Ex, Medscape Cardiology, MyoKardia, NirvaMed, Novo Nordisk, PhaseBio, PLx Pharma, Regado Biosciences, and Stasys; board of directors for Boston VA Research Institute, Society of Cardiovascular Patient Care, and TOBESOFT; inaugural chair of the American Heart Association Quality Oversight Committee; data monitoring committees for the Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute, for the PORTICO trial, funded by St. Jude Medical, now Abbott), Boston Scientific (chair, PEITHO trial), Cleveland Clinic (including for the CENTERA THV System in Intermediate Risk Patients Who Have Symptomatic, Severe, Calcific, Aortic Stenosis [ExCEED] trial, funded by Edwards), Contego Medical (chair, PERFORMANCE 2), Duke Clinical Research Institute, Mayo Clinic, Mount Sinai School of Medicine (for the ENVISAGE trial, funded by Daiichi Sankyo), Novartis, and Population Health Research Institute; honoraria from the American College of Cardiology (ACC) (senior associate editor, Clinical Trials and News, ACC.org, and chair, ACC Accreditation Oversight Committee), Arnold and Porter law firm (work related to Sanofi/Bristol-Myers Squibb clopidogrel litigation), Baim Institute for Clinical Research (formerly Harvard Clinical Research Institute; Randomized Evaluation of Dual Antithrombotic Therapy with Dabigatran versus Triple Therapy with Warfarin in Patients with Nonvalvular Atrial Fibrillation Undergoing Percutaneous Coronary Intervention [RE-DUAL PCI] clinical trial steering committee funded by Boehringer Ingelheim; AEGIS-II executive committee funded by CSL Behring), Belvoir Publications (editor in chief, Harvard Heart Letter), Canadian Medical and Surgical Knowledge Translation Research Group (clinical trial steering committees), Cowen and Company, Duke Clinical Research Institute (clinical trial steering committees, including for A Trial Comparing Cardiovascular Safety of Degarelix Versus Leuprolide in Patients With Advanced Prostate Cancer and Cardiovascular Disease [PRONOUNCE] trial, funded by Ferring Pharmaceuticals), HMP Global (editor in chief, Journal of Invasive Cardiology), Journal of the American College of Cardiology (guest editor, associate editor), K2P (co-chair, interdisciplinary curriculum), Level Ex, Medtelligence/ReachMD (continuing medical education [CME] steering committees), MJH Life Sciences, Piper Sandler, Population Health Research Institute (for the Cardiovascular Outcomes for People Using Anticoagulation Strategies [COMPASS] operations committee, publications committee, steering committee, and U.S. national coleader, funded by Bayer), Slack Publications (chief medical editor, Cardiology Today’s Intervention), Society of Cardiovascular Patient Care (secretary/treasurer), and WebMD (CME steering committees); other for Clinical Cardiology (deputy editor), NCDR’s ACTION Registry Steering Committee (chair), and Veterans Affairs Cardiovascular Assessment, Reporting and Tracking Research and Publications Committee (chair); research funding from Abbott, Afimmune, Amarin, Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Cardax, CellProthera, Cereno Scientific, Chiesi, CSL Behring, Eisai, Ethicon, Faraday Pharmaceuticals, Ferring Pharmaceuticals, Forest Laboratories, Fractyl, Garmin, HLS Therapeutics, Idorsia, Ironwood, Ischemix, Janssen, Lexicon, Eli Lilly, Medtronic, MyoKardia, NirvaMed, Novartis, Novo Nordisk, Owkin, Pfizer, PhaseBio, PLx Pharma, Regeneron, Roche, Sanofi, Stasys, Synaptic, The Medicines Company, and 89Bio; royalties from Elsevier (editor, Cardiovascular Intervention: A Companion to Braunwald’s Heart Disease); site coinvestigator for Abbott, Biotronik, Boston Scientific, CSI, St. Jude Medical (now Abbott), Philips, and Svelte; trustee for the American College of Cardiology; and unfunded research for FlowCo, Merck, and Takeda. L.A.L. has received research funding from, has provided CME on behalf of, and/or has acted as an advisor to AstraZeneca, Bayer, Boehringer Ingelheim, Eli Lilly, Janssen, Lexicon, Merck, Novo Nordisk, Pfizer, Sanofi, and Servier. D.K.M. reports honoraria for trial leadership from Boehringer Ingelheim, Sanofi Aventis, Merck & Co., Pfizer, Akebia, AstraZeneca, Novo Nordisk, Esperion, AbbVie, Eli Lilly USA, Eidos, Arena, Dynavax, Lexicon, Otsuka, and CSL Behring and honoraria for consultancy from Eli Lilly USA, Boehringer Ingelheim, Merck & Co., Novo Nordisk, Applied Therapeutics, Metavant, Sanofi Aventis, Afimmune, CSL Behring, and Bayer. J.P.H.W., outside the submitted work, has grants, personal fees for lectures, and consultancy fees (paid to his institution) from AstraZeneca and Novo Nordisk; personal fees for lectures and consultancy fees (paid to his institution) from Boehringer Ingelheim, Janssen, Eli Lilly, Mundipharma, Napp, Sanofi, and Takeda; and consultancy fees (paid to his institution) from Rhythm Pharmaceuticals and Wilmington Healthcare. I.A.M.G.-N. and A.M.L. are employees at BioPharmaceuticals R&D, AstraZeneca. M.S.S. reports research support from Abbott, Amgen, Anthos Therapeutics, AstraZeneca, Bayer, Daiichi-Sankyo, Eisai, Intarcia, IONIS, Medicines Company, MedImmune, Merck, Novartis, Pfizer, and Quark Pharmaceuticals and has received consulting fees from Althera, Amgen, Anthos Therapeutics, AstraZeneca, Beren Therapeutics, Bristol-Myers Squibb, CVS Caremark, DalCor, Dr. Reddy’s Laboratories, Fibrogen, IFM Therapeutics, Intarcia, MedImmune, Merck, Moderna, Novo Nordisk, and Silence Therapeutics. I.R. reports personal fees from AstraZeneca, Bristol-Myers Squibb, Boehringer Ingelheim, Concenter BioPharma and Silkim, Eli Lilly, Merck Sharp & Dohme, Novo Nordisk, Orgenesis, Pfizer, Sanofi, SmartZyme Innovation, Panaxia, FuturRx, Insuline Medical, Medial EarlySign, CameraEyes, Exscopia, Dermal Biomics, Johnson & Johnson, Novartis, Teva, GlucoMe, and DarioHealth. No other potential conflicts of interest relevant to this article were reported.

Data analyses were done by the academic TIMI Study Group, which has access to the complete study database, allowing independent analyses of the results, any discrepancies were resolved by discussion. The DECLARE-TIMI 58 publication committee made the decision to submit for publication. The funder was involved in the study design, data collection, data analysis, interpretation, and writing of this report.

Author Contributions. A.C. wrote the original draft of the manuscript. A.C., S.D.W., O.M., S.A.M., E.L.G., I.Y., A.R., D.L.B., L.A.L., D.K.M., J.P.H.W., I.A.M.G.-N., A.M.L., M.S.S., and I.R. contributed to the investigation and to the writing, review, editing of the manuscript. A.C., S.D.W., O.M., S.A.M., D.L.B., L.A.L., D.K.M., J.P.H.W., I.A.M.G.-N., A.M.L., M.S.S., and I.R. contributed to the conceptualization and provided supervision. A.C., S.D.W., O.M., I.A.M.G.-N., A.M.L., M.S.S., and I.R. contributed to the project administration. A.C., S.D.W., S.A.M., E.L.G., I.A.M.G.-N., A.M.L., M.S.S., and I.R. provided validation. S.D.W., S.A.M., E.L.G., M.S.S., and I.R. contributed to the formal analysis. S.D.W., S.A.M., I.A.M.G.-N., A.M.L., M.S.S., and I.R. contributed to the data curation. S.D.W., I.A.M.G.-N., A.M.L., M.S.S., and I.R. acquired funding. A.C. and I.R. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the 81st Scientific Sessions of the American Diabetes Association, 25–29 June 2021.