Temporal Trends and Factors Associated With Diabetes Mellitus Among Patients Hospitalized With Heart Failure

Findings From Get With the Guidelines-Heart Failure Registry

Justin B. Echouffo-Tcheugui, MD, PhD; Haolin Xu, MS; Adam D. DeVore, MD; Phillip J. Schulte, PhD; Javed Butler,MD, MPH, MBA; ClydeW. Yancy, MD,MSc; Deepak L. Bhatt, MD, MPH; Adrian F. Hernandez, MD, MHS; Paul A. Heidenreich, MD, MS; Gregg C. Fonarow, MD


Am Heart J. 2016;182(12):9-20. 

In This Article

Abstract and Introduction


Background The contribution of diabetes to the burden of heart failure (HF) remains largely undescribed. Assessing diabetes temporal trends among US patients hospitalized with HF and their relation with quality measures in real-world practice can help to define this burden.

Methods Using data from the Get With the Guidelines–Heart Failure registry, we assessed temporal trends in diabetes prevalence among patients with HF and in subgroups with reduced ejection fraction (HFrEF; EF < 40%), borderline EF (HFbEF; 40% ≤ EF < 50%), or preserved EF (HFpEF; EF ≥ 50%), hospitalized between 2005 and 2015. Logistic regression was used to assess whether in-hospital outcomes and HF quality of care were related to trends.

Results Among 364,480 HF hospitalizations, 160,171 had diabetes (44.0% overall, 41.8% in HFrEF, 46.7% in HFbEF, 45.5% in HFpEF). There was a temporal increase in diabetes frequency in HF patients (43.2%-45.8%; P trend < .0001), including among those with HFrEF (42.0%-43.6%; P trend < .0001), HFbEF (46.0%-49.2%; P trend < .0001), or HFpEF (43.6%-46.8%, P trend < .0001). Diabetic patients had a longer hospital stay (adjusted odds ratio 1.14, 95% CI 1.12–1.16), but lower in-hospital mortality (adjusted odds ratio 0.93 [0.89–0.97]) compared with those without diabetes, with limited differences in quality measures. Temporal trends in diabetes were not associated with in-hospital mortality or length of stay. There were no temporal interactions of most HF quality measures with diabetes status.

Conclusions Approximately 44% of hospitalized HF patients have diabetes, and this proportion has been increasing over the past 10 years, particularly among those patients with new-onset HFpEF.


Heart failure (HF) is associated with a substantial burden of morbidity and mortality in the United States, with the number of Americans currently suffering from this condition projected to increase by 46% by the year 2030.[1] The risk of HF among people with diabetes mellitus (diabetes) is higher compared with that of the general population.[2,3] Hence, diabetes and HF often coexist, posing unique clinical challenges.[4] The burden of diabetes in the United States has been growing continuously, with 14.3% of adults American with diabetes, of whom up to 36.4% remained undiagnosed in 2012.[5] These numbers are projected to increase by at least 30% by 2050.[6] There is a paucity of national data on recent patterns and factors associated with coexisting diabetes among persons with HF. Insights into the evolving burden of diabetes among HF patients may be of major public health and economic importance, as this may offer the opportunity to enhance outcomes among HF patients, especially given the prohibitive costs of care for HF and diabetes, respectively.[7,8] Knowledge of these trends has potential implications for systematic screening for diabetes among those with HF patients and vice versa. Furthermore, our evolving understanding of the diabetes-related cardiomyopathy indicates that it may occur with or without systolic dysfunction.[9] Hence, it is important to understand time patterns of diabetes in clinically important subgroups of HF patients with different type of HF (preserved vs borderline vs reduced left ventricular ejection fraction [LVEF]). Furthermore, the influence of the presence of diabetes on treatment and short-term outcomes in HF patients with comorbid diabetes has not been evaluated in detail in previous studies.[10,11]

The current study uses the framework of the Get With The Guidelines–Heart Failure (GWTG-HF) program registry to assess recent time trends in the prevalence of clinically diagnosed diabetes among US adults primarily hospitalized with HF, including by HF type (HF with preserved ejection fraction [HFpEF], borderline EF [HFbEF], and reduced EF [HFrEF]), and the influence of coexisting diabetes mellitus on in-hospital outcomes and quality measures.