Roger, Véronique L.

Objective: To compare two measures of frailty in a community cohort of heart failure (HF) patients. Background: Frailty is increasingly recognized as an important prognostic indicator in HF. Several methods have been proposed to measure frailty. Some conceptualize frailty as a biological phenotype, while others define it as an accumulation of deficits (impairments, disabilities, comorbid diseases), termed the Cumulative Deficit Index (CDI). Each approach has conceptual strengths and limitations; a distinct feature of the CDI is the ability to extract data from medical records. These two methods have never been compared in a community HF cohort. Methods: Olmsted, Dodge and Fillmore County, MN residents with HF between 10/2007 and 12/2010 were prospectively recruited to assess the frailty phenotype. The biological frailty phenotype was defined as ≥3 of the following: unintentional weight loss >10 lbs. in 1 year, physical exhaustion, weak grip strength, and slowness and low activity measured by the SF-12 physical component score. Intermediate frailty was defined as having 1–2 components. The CDI was based on 33 deficits and defined as the proportion of deficits present for each patient. Results: Among 131 patients (mean age 70.5±14, 62% male), according to the biological frailty phenotype definition 18% were frail, 50% intermediate frail and 31% not frail. The CDI ranged from 0.03–0.70, with a mean (SD) of 0.22 (0.12). The CDI was significantly correlated with the frailty phenotype (r=0.59, p<0.001). The mean (SD) of the CDI was 0.35 (0.14), 0.21 (0.10) and 0.15 (0.07) among the frail, intermediate frail and not frail phenotypes (Figure). Conclusion: The CDI is significantly associated with the biological frailty phenotype and can thus be used as an alternative approach to measure frailty in HF. As the components of the biological frailty phenotype are not routinely assessed clinically, the CDI, which can be ascertained from medical records, is a feasible approach to ascertain frailty in large cohorts.

Background: Heart failure (HF) care constitutes an increasing economic burden on the healthcare system, and has become a key focus in the healthcare debate. Costs incurred at the end of life can be particularly high in some populations, however, end of life costs among a community population of HF patients have not been well delineated. Methods: Olmsted County residents with incident HF from 1987-2006 were identified and followed until death. Direct medical costs incurred in the last 6 months prior to death were obtained using population-based administrative data through 2010. For those that died within 6 months of HF diagnosis, costs were included from diagnosis until death. Costs were inflated to 2010 U.S. dollars using the general Consumer Price Index (CPI). Results: A total of 862 HF patients were identified, died during follow-up, and had complete cost data (mean age 79.4 years, 46.4% male, 56.5% EF≥50%). Median costs within the last 6 months of life were $12,069 per person (25 th -75 th percentile $1768 - $30,841, mean $24,052), with the majority accumulated during hospitalizations (median $8849 per person, mean $20,263, 84.2% total costs for population). A total of 596 (69.1%) patients were hospitalized in the last 6 months of life, and 235 (27.3%) died in the hospital. Costs progressively increased in the months prior to death (Figure), with nearly half (43.4%) of total costs in the last 6 months of life accumulated during the month before death. End of life costs did not change over the study period (p=0.154 adjusted for age and sex). Conclusions: End of life costs are high among community HF patients, particularly in the final month of life. Most of these costs are accumulated during hospitalizations, and a large proportion of HF patients died in the hospital. Among patients with HF nearing the end of life, alternative strategies of care that avoid hospitalizations may be particularly cost-saving.

Background: Whether age alone explains the comorbidity burden in heart failure (HF) is unclear. In particular, differences in the burden of co-morbid conditions in HF patients compared to population controls has not been well documented. Methods: The prevalence of 17 chronic conditions defined by the US Department of Health and Human Services were obtained in 1746 incident HF patients from 2000-2010 and controls matched 1:1 on sex and age from Olmsted County, MN. Conditions were ascertained requiring 2 occurrences of a diagnostic code. Logistic regression determined associations of each condition with HF. Results: Among the 1746 matched pairs (mean age 76.2 years, 43.5% men), the prevalence was higher in HF cases for all conditions (p<0.05) except dementia and osteoporosis. After adjusting for all conditions, hypertension, coronary artery disease, arrhythmia, asthma, chronic kidney disease, chronic obstructive pulmonary disease, diabetes, hepatitis, and substance abuse were significantly more common in HF (figure). More than a 2-fold increased odds of hepatitis, arrhythmia, and coronary artery disease was observed among HF cases. Arrhythmia (34.2%), hypertension (31.1%), and coronary artery disease (27.8%) had the largest attributable risk of HF; for example, assuming a causal relationship, if arrhythmias were eliminated, 34% of HF would be avoided. Conclusions: Compared to age- and sex-matched controls, HF patients have a higher prevalence of many chronic conditions, indicating the excess comorbidity in HF is not due to age alone. Some cardiovascular conditions, including arrhythmia, coronary artery disease, and hypertension were more common in HF. Of the non-cardiovascular conditions, hepatitis had the strongest association with HF and was an unanticipated finding that deserves additional investigation. It is important to understand comorbidities as they play a key role in the excess mortality and healthcare utilization experienced by HF patients.

Background: The presence of frailty increases the risk of long-term mortality for a myriad of reasons, but its association with cardiac-specific mortality among the elderly undergoing percutaneous coronary intervention (PCI) is unknown. We sought to determine the relationship between frailty and cause-specific (cardiac versus non-cardiac) mortality among elderly patients undergoing PCI. Methods: A prospective cohort of elderly (>65 years) patients undergoing PCI at Mayo Clinic between October 2005 and September 2008 were tested for frailty using the Fried criteria. Vital status was prospectively monitored including evaluation of death certificates, telephone interviews and review of autopsy/medical records. Results: Frailty data was obtained in 428 (84%) of 508 elderly patients (median age 74.2 years [SD 6.4, range 65-100]; 59% male). Of these, 92 (21%) patients were frail while 231 (54%) had intermediate frailty and 105(25%) were not frail. Frail patients were older than intermediate and non-frail patients (77.6 vs 74.3 vs 71.0 years; p<0.001) and had more comorbidities (mean Charlson index 4.6 vs 3.9 vs 2.8; p<0.001). There were 113 deaths (26% of cohort) during 5.1 median years of follow-up, of which 77 (68%) were classified as non-cardiac and 36 (32%) which were classified as cardiac. Frailty at baseline was associated with poor overall survival compared to non-frail patients (HR 6.1, 95% CI 3.2, 11.6; p<0.001). The estimated effect of frailty was nearly 3 times greater for cardiac mortality (HR 13.1, 95% CI 3.0, 57.0; p=0.001) than non-cardiac mortality (HR 4.6, 95% CI 2.2, 9.7; p<0.001). These findings remained significant following multivariable adjustment. Conclusions: Frailty is common among elderly patients undergoing PCI and is associated with poor overall survival and both cardiac and non-cardiac causes of death.

Introduction: Proton pump inhibitors (PPIs) are used by an estimated 29 million Americans. PPIs increase the levels of asymmetrical dimethylarginine, a known risk factor for cardiovascular disease (CVD). Data from a select population of patients with CVD suggest that PPI use is associated with an increased risk of stroke, heart failure, and coronary heart disease. The impact of PPI use on incident CVD is largely unknown in the general population. Hypothesis: We hypothesized that PPI users have a higher risk of incident total CVD, coronary heart disease, stroke, and heart failure compared to nonusers. To demonstrate specificity of association, we additionally hypothesized that there is not an association between use of H 2 -blockers - another commonly used class of medications with similar indications as PPIs - and CVD. Methods: We used the Rochester Epidemiology Project’s medical records-linkage system to identify all residents of Olmsted County, MN on our baseline date of January 1, 2004 (N=140217). We excluded persons who did not grant permission for their records to be used for research, were <18 years old, had a history of CVD, had missing data for any variable included in our model, or had evidence of PPI use within the previous year.We followed our final cohort (N=58175) for up to 12 years. The administrative censoring date for CVD was 1/20/2014, for coronary heart disease was 8/3/2016, for stroke was 9/9/2016, and for heart failure was 1/20/2014. Time-varying PPI ever-use was ascertained using 1) natural language processing to capture unstructured text from the electronic health record, and 2) outpatient prescriptions. An incident CVD event was defined as the first occurrence of 1) validated heart failure, 2) validated coronary heart disease, or 3) stroke, defined using diagnostic codes only. As a secondary analysis, we calculated the association between time-varying H 2 -blocker ever-use and CVD among persons not using H 2 -blockers at baseline. Results: After adjustment for age, sex, race, education, hypertension, hyperlipidemia, diabetes, and body-mass-index, PPI use was associated with an approximately 50% higher risk of CVD (hazard ratio [95% CI]: 1.51 [1.37-1.67]; 2187 CVD events), stroke (hazard ratio [95% CI]: 1.49 [1.35-1.65]; 1928 stroke events), and heart failure (hazard ratio [95% CI]: 1.56 [1.23-1.97]; 353 heart failure events) compared to nonusers. Users of PPIs had a 35% greater risk of coronary heart disease than nonusers (95% CI: 1.13-1.61; 626 coronary heart disease events). Use of H 2 -blockers was also associated with a higher risk of CVD (adjusted hazard ratio [95% CI]: 1.23 [1.08-1.41]; 2331 CVD events). Conclusions: PPI use is associated with a higher risk of CVD, coronary heart disease, stroke and heart failure. Use of a drug with no known cardiac toxicity - H 2 -blockers - was also associated with a greater risk of CVD, warranting further study.

Background: Little is known about the association between coronary artery disease (CAD) and the risk of heart failure (HF) after myocardial infarction (MI), and whether it differs by reduced (HFrEF) or preserved (HFpEF) ejection fraction (EF) has yet to be determined. Subjects and Methods: Olmsted County, Minnesota residents (n=1,924; mean age, 64 years; 66% male) with first MI diagnosed in 1990-2010 and no prior HF were followed through 2013. Framingham Heart Study criteria were used to define HF, which was further classified according to EF (applying a 50% cutoff). The extent of angiographic CAD was defined at index MI according to the number of major epicardial coronary arteries with ≥50% lumen diameter obstruction. Fine & Gray and Cox proportional hazards regression models were used to assess the association of CAD categories with incidence of HF, and multiple imputation methodology was applied to account for the 19% with missing EF data. Results: During a mean (SD) follow-up of 6.7 (5.9) years, 594 patients developed HF. Adjusted for age and sex, with death considered a competing risk, the cumulative incidence rates of HF among patients with 1- (n=581), 2- (n=622), and 3-vessel disease (n=721) were 11.2%, 14.6% and 20.5% at 30 days; and 18.1%, 22.3% and 29.4% at 5 years after MI, respectively. The increased risk of HF with greater number of occluded vessels was only modestly attenuated after further adjustment for patient and MI characteristics, and did not differ materially by EF (Table). Conclusions: The extent of angiographic CAD expressed by the number of diseased vessels is independently associated with HF incidence after MI. The association is evident promptly after MI and applies to both HFrEF and HFpEF.

Background: Differences in the prevalence and duration of co-morbid conditions in atrial fibrillation (AF) patients compared to population controls have not been well documented. Methods: The prevalence and duration of 17 chronic conditions defined by the US Department of Health and Human Services, as well as anxiety, obesity, and smoking status, was obtained in a random sample of 1430 patients with incident AF from 2000-2010 and 1430 controls from Olmsted County, MN. Controls were matched to cases 1:1 on sex and age (within 5 years). Chronic conditions were ascertained electronically requiring 2 occurrences of a diagnostic code; the duration of each condition (up to 25 years) was calculated. Logistic regression determined associations of each condition with AF after adjustment for all other conditions. Results: Among the 1430 matched pairs (median age 76 years, 48.6% men), the prevalence of chronic conditions was higher in AF cases compared to controls for all conditions except asthma, dementia, depression, hepatitis, and osteoporosis (figure). However, the duration of the conditions were similar in AF compared to controls, except for hypertension (median duration 12.3 and 9.9 years in AF cases and controls, respectively; p=0.002). After adjusting for all other conditions, obesity, hypertension, congestive heart failure, coronary artery disease, chronic kidney disease, and chronic obstructive pulmonary disease remained significantly more common in AF compared to controls (figure). Conditions with the largest attributable risk of AF were hypertension (25.4%), coronary artery disease (17.7%), and congestive heart failure (12.3%). Conclusions: AF patients have a higher prevalence of many chronic conditions compared to population controls. However, besides hypertension, these comorbidities do not develop earlier in AF. Nevertheless, the excess comorbidity burden in AF is important to characterize and understand as it may partly explain the excess mortality and healthcare utilization experienced by AF patients.

BACKGROUND Epidemiological cohorts can assist with post-marketing surveillance as they provide community-based outcomes data. Concern was raised about the adverse impact of sulfonylureas (SU) drugs on the outcome of diabetes after myocardial infarction (MI). We evaluated the outcome of MI according to treatment among diabetic persons compared to non-diabetic counterparts in a community-based cohort. METHODS All MIs that occurred in Olmsted County, MN between 1979 and 2002 were validated by standardized epidemiological including cardiac pain, biomarkers and ECG Minnesota coding. Diabetes mellitus (DM) was defined by the National Diabetes Data Group criteria. RESULTS Among 2732 persons with MI (mean age 70 ± 12 years, 44% men), 486 (18%) had DM. Subjects with DM were treated with second generation SU (SU2) (24%), insulin (47%) or diet alone (22%) Diabetic subjects were more likely to be overweight, hyperlipidemic and to receive reperfusion therapy during hospitalization than non-diabetic subjects. One year after the MI, 518 subjects died. Diabetic subjects taking SU2 experienced better survival than subjects without diabetes, while subjects treated with diet alone or insulin experienced worse survival. (Figure ). After adjusting for age, sex, reperfusion and duration of DM, an excess risk of death was detected among persons with insulin-treated DM, particularly those with long-standing DM. No association was found for DM patients treated with SU2 or diet. CONCLUSION These data do not support the concern for an adverse impact of SU2 on survival and underscore the potential role of community studies for population surveillance.

Background: Whether patients with STEMI with longer times from symptom onset to hospital presentation are less likely to be treated with any reperfusion therapy or treated with less urgency resulting in longer door-to-balloon and door-to-needle times is not known. The relationship between delay in hospital presentation and the quality of reperfusion therapy for patients with STEMI has not been examined Methods: We constructed 3 cohorts of STEMI patients to analyze use of any reperfusion (n=440,398), door-to-balloon time (n=67,207), and door-to-drug time (n=183,441) as a function of delay in hospital presentation. We constructed multivariable generalized linear models for each outcome to estimate the associations between delay in hospital presentation adjusted for all patient and hospital characteristics. Results: In adjusted analysis, longer times from symptom onset to hospital presentation was associated with lower use of any reperfusion therapy (p<0.0001 for trend). For early presenters with times of ≤1 hour, >1 to 2 hours, and >2 to 3 hours, 77%, 77%, and 73% of patients received any reperfusion therapy, respectively, and late presenters with times >9 to 10 hours, >10 to 11 hours, and >11 to 12 hours were treated with any reperfusion therapy in 53%, 50%, and 46%, respectively. Delay in hospital presentation was associated with longer drug-to-balloon and door-to-drug times (p<0.0001 for trend). For time intervals of ≤ 1 hour, >1 to 2 hours, >2 to 3 hours, >9 to 10 hours, >10 to 11 hours, and >11 to 12 hours, patients with STEMI were treated with door-to-balloon times of 99, 101, 106, 123, 125, and 123 minutes respectively. For delay time intervals of ≤ 1 hour, >1 to 2 hours, >2 to 3 hours, >9 to 10 hours, >10 to 11 hours, and >11 to 12 hours, patients were treated with door-to-drug times of 32, 34, 36, 46, 44, and 46 minutes, respectively. Conclusions: Longer time from symptom onset to hospital presentation has important implications for subsequent treatment and was associated with lower use of any reperfusion therapy and longer door-to-balloon and door-to-drug times. Longer delay in hospital presentation is associated with worse system performance for reperfusion therapy quality and represents an opportunity to improve quality of care for STEMI patients.

Background Patients with heart failure and preserved ejection fraction (HFpEF) have diastolic dysfunction, but are traditionally considered to have normal left ventricular (LV) systolic function. However, ventricular remodeling can result in preservation of EF despite abnormal myocardial contractility. Methods We performed echo-Doppler characterization of LV chamber and myocardial systolic properties in a population-based study, comparing patients with HFpEF (N=244) to healthy controls (CON, N=617), and hypertensives without HF (HTN, N=719), then examined long term outcome. Results All subjects had a normal EF (>50%). However, systolic chamber function, measured by wall stress-corrected endocardial fractional shortening (sc-eFS), was impaired in HFpEF (96±12%) compared to both CON (100±8%, p<0.0001) and HTN (108±11%, p<0.0001). Myocardial contractility, assessed by wall stress-corrected midwall shortening (sc-mFS), was also reduced in HFpEF (91±13%) compared to CON (100±10%, p<0.0001) and HTN (105±12%, p<0.0001). HTN had increased sc-eFS and sc-mFS compared with both HFpEF and CON (p<0.0001). In HFpEF, impaired sc-mFS was associated with increased mortality, independent of age (Figure ), while EF and sc- eFS were not. Conclusions Despite preservation of EF, unselected HFpEF patients from the community have significantly impaired systolic chamber function and depressed myocardial contractility. Abnormal myocardial contractility in HFpEF is associated with increased mortality. These data suggest that myocardial systolic dysfunction contributes to the pathophysiology of HFpEF and may represent a potential therapeutic target.