You are here

Article Text

Article menu

Download PDFPDF

Heart failure and cardiomyopathies
Original research
Lower socioeconomic status predicts higher mortality and morbidity in patients with heart failure
  1. Benedikt Schrage1,2,3,
  2. Lars H Lund1,
  3. Lina Benson1,
  4. Davide Stolfo1,4,
  5. Anna Ohlsson5,
  6. Ragnar Westerling5,
  7. Dirk Westermann2,3,
  8. Anna Strömberg6,
  9. Ulf Dahlström6,
  10. Frieder Braunschweig1,
  11. João Pedro Ferreira7,8,
  12. Gianluigi Savarese1
  1. 1 Department of Medicine, Karolinska Institute, Stockholm, Sweden
  2. 2 German Center for Cardiovascular Research (DZHK), Partner Site Hamburg/Lübeck/Kiel, Hamburg, Germany
  3. 3 Department of Cardiology, University Heart and Vascular Center Hamburg, Hamburg, Germany
  4. 4 Cardiovascular Department, 'Ospedali Riuniti' and University of Trieste, Trieste, Italy
  5. 5 Department of Public Health and Caring Sciences, Uppsala University, Uppsala, Sweden
  6. 6 Department of Medical and Health Science, Linköping University, Linköping, Sweden
  7. 7 Centre d’Investigations Cliniques Plurithématique 1433, Université de Lorraine and CHU de Nancy, INSERM UMR1116, Vandoeuvre-les-nancy, France
  8. 8 F-CRIN INI-CRCT Cardiovascular and Renal Clinical Trialists, Vandoeuvre-les-Nancy, France
  1. Correspondence to Dr Gianluigi Savarese, Department of Medicine, Karolinska Institute, Stockholm 17176, Sweden; gianluigi.savarese{at}ki.se

Abstract

Objective It is not fully understood whether and how socioeconomic status (SES) has a prognostic impact in patients with heart failure (HF). We assessed SES and its association with patient characteristics and outcomes in a contemporary and well-characterised HF cohort.

Methods Socioeconomic risk factors (SERF) were defined in the Swedish HF Registry based on income (low vs high according to the annual median value), education level (no degree/compulsory school vs university/secondary school) and living arrangement (living alone vs cohabitating).

Results Of 44 631 patients, 21% had no, 33% one, 30% two and 16% three SERF. Patient characteristics strongly and independently associated with lower SES were female sex and no specialist referral. Additional independent associations were older age, more severe HF, heavier comorbidity burden, use of diuretics and less use of HF devices. Lower SES was associated with higher risk of HF hospitalisation/mortality, and overall cardiovascular and non-cardiovascular events. These associations persisted after extensive adjustment for patient characteristics, treatments and care. The magnitude of the association increased linearly with the increasing number of coexistent SERF: HR (95% CI) 1.09 (1.05 to 1.13) for one, 1.16 (1.12 to 1.20) for two and 1.22 (1.18 to 1.28) for three SERF (p<0.01).

Conclusions In a contemporary and well-characterised HF cohort and after comprehensive adjustment for confounders, lower SES was linked with multiple factors such as less use of HF devices and age, but most strongly with female sex and lack of specialist referral; and associated with greater risk of morbidity/mortality.

  • heart failure

https://doi.org/10.1136/heartjnl-2020-317216

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Heart failure (HF) affects approximately 2%–3% of the Western population and is increasing in prevalence. It is characterised by high mortality/morbidity, impaired quality of life and an heavy economic burden on healthcare systems worldwide.1 Several factors, such as demographic and clinical characteristics linked with more severe HF, comorbidities and poorly implemented use of treatment contribute to explain the impaired prognosis characterising the HF population.2 However, these important factors only partially account for the total risk of adverse outcomes observed in HF.3

    Lower socioeconomic status (SES) is associated with poorer outcome in the general population as well as in patients with cardiovascular disease.4 It might also play an important role in HF disease progression and contribute to explain the total risk of adverse outcomes, which is not linked with well-established prognostic factors.

    Previous studies have highlighted an interplay between different measures of SES and outcomes in HF.5–7 However, the lack of important phenotypic information and of granular data on use of care and of HF treatments might have led to overestimation of the prognostic role of SES, which might rather be explained by unmeasured confounding, for example, be a risk marker simply for less use of HF therapies.5–7 How SES impacts on prognosis is not exactly known. Socioeconomic disparities might foster differences in the distribution of comorbidities which are prognostically relevant in HF (eg, diabetes, hypertension, etc), and/or might limit the access to healthcare, influence the likelihood of receiving/using life-saving treatments and also influence lifestyle patterns.

    Therefore, we investigated in a large, well-characterised and contemporary HF cohort (1) the clinical phenotype of patients with HF with different SES; (2) whether lower SES is associated with poorer outcome and (3) whether this potential association might be explained by specific patient characteristics, or different use of care or treatments.

    Methods

    Study protocol and setting

    The design of the Swedish Heart Failure Registry (SwedeHF, www.SwedeHF.se) has been previously described.8 Briefly, patients with clinician-judged HF have been enrolled since 11 May 2000. Approximately 80 variables are recorded at discharge from hospital or after outpatient visit.

    For this analysis, SwedeHF was linked to the National Patient Registry which provided data on additional comorbidities and the outcomes HF, cardiovascular, non-cardiovascular hospitalisation, to Statistics Sweden which provided data on education level and income and to the Cause of Death Registry which provided all-cause, cardiovascular, non-cardiovascular mortality data.

    Patient and public involvement

    Patients or the public were not involved in this research project.

    Determination of SES

    In this study, SES has been broadly defined, including different factors indicating the level of social and economic resources.

    Income: higher/lower income was defined as individual annual income at the index date > or ≤median annual income of all SwedeHF patients registered in the same year, respectively.

    Education level: higher education was defined as secondary school degree or higher whereas lower education as compulsory school degree or no degree.

    Living arrangement: defined as living alone versus cohabitating.

    SES was graded based on the prevalence/coexistence of the socioeconomic risk factors (SERF) measured in our cohort, that is, lower income, lower education level, living alone. Therefore, a patient could have 0, 1, 2 or 3 SERF.

    Patients

    Patients registered in SwedeHF between 11 May 2000 and 31 December 2016 with available data on income, education level, living arrangement and ejection fraction (which is a categorised variable in SwedeHF, ie, <30%, 30%–39%, 40%–49% and ≥50%), were included in the analyses. Patients who died during the hospitalisation/visit linked to their SwedeHF registration were excluded. If the same patient had multiple registrations in SwedeHF, the last one was selected, since more representative of contemporary care and socioeconomic context. The index date was defined as either the day of the outpatient visit or the day of hospital discharge.

    Statistical analyses

    Patient characteristics

    Baseline characteristics were compared across different SES (ie, different number of coexistent SERF) by analysis of variance or Kruskal-Wallis tests for continuous and by χ2 for categorical variables.

    To evaluate patient characteristics which were independently associated with lower SES, a multivariable ordinal logistic regression model was fitted, with the number of coexistent SERF as the dependent variable and the baseline characteristics reported in table 1 as covariates. The proportional odds assumption was visually verified and met. ORs and 95% CIs were calculated for each potential predictor.

    View this table:
    Table 1

    Baseline characteristics stratified by number of socioeconomic risk factors

    Outcome analysis

    The primary outcome was a composite of time to first HF hospitalisation or all-cause death. Secondary outcomes were HF hospitalisation (with censoring at death) and all-cause death, separately, cardiovascular events (time to first cardiovascular hospitalisation or cardiovascular death, with censoring for non-cardiovascular death) and non-cardiovascular events (time to first non-cardiovascular hospitalisation or non-cardiovascular death; with censoring for cardiovascular death).

    To investigate the association between SES and outcomes, the Kaplan-Meier method was used to estimate unadjusted survivor functions in patients stratified by the number of coexistent SERF. Separate Cox regression models were fitted with (1) no adjustment; (2) adjustment for demographic characteristics; (3) for demographic+clinical characteristics; (4) for demographic+clinical characteristics+comorbidities; (5) for demographic+clinical characteristics+comorbidities+treatments (42 variables in total). HR and 95% CI were calculated. The proportional hazard assumption was verified by Schoenfeld residuals and met. To assess whether SES was similarly associated with the different outcomes, the respective HRs were compared based on their betas/SEs by t-test.

    In both multivariable ordinal logistic and Cox regression models, missing data were handled by chained-equations multiple imputation (R-package mice). Analyses were performed in 10 imputed datasets and then the estimates were combined by Rubin’s rules. Variables used for the multiple imputation are marked in table 1.

    All statistical analyses were performed by R V.3.5.3. A two-sided p value <0.05 was considered as statistically significant.

    Variable definitions are provided in online supplementary table 1, missing data are shown in online supplementary table 2.

    Supplemental material

    Results

    Prevalence of socioeconomic risk factors

    Between 11th May 2000 and 31st December 2016, there were 130 420 registrations from 76 506 unique patients in SwedeHF. After applying inclusion/exclusion criteria, 44 641 patients were eligible for this analysis.

    As many as 9362 patients (21%) had no SERF, 14 600 patients (33%) had one SERF, 13 594 patients (30%) had two SERF and 7085 patients (16%) had all three SERF. Figure 1 illustrates the inter-relationship across the different SERF.

    Figure 1
    Figure 1

    Inter-relationship between socioeconomic risk factors.

    Baseline characteristics

    Patients with lower SES were more likely older and female, with higher ejection fraction, more severe HF (ie, higher New York Heart Association functional class, N-terminal pro-B-type natriuretic peptide levels) and higher comorbidity burden. They were less likely to receive guideline-recommended HF therapies and to have follow-up in specialty care, but more likely to receive diuretics. Similar results were shown when analyses were stratified by ejection fraction subtype or by individual SERF (see online supplementary tables 3–8).

    Results in table 1 are unadjusted. Therefore, we performed multivariable ordinal logistic regression to identify patient characteristics independently associated with lower SES (figure 2).

    Figure 2
    Figure 2

    Independent associations between patient characteristics and socioeconomic status. ORs from the multivariable ordinal logistic regression are shown on the X-axis, with an OR <1.0 indicating an association with higher socioeconomic status and an OR >1.0 indicating an association with a lower socioeconomic status. Specialty care follow-up refers to planned follow-up in a specialty care (eg, cardiology, internal medicine) institution. Similarly, nurse-led HF follow-up refers to planned follow-up in a nurse-led HF clinic. CRT, cardiac resynchronisation therapy; eGFR, estimated glomerular filtration rate (calculated by Chronic Kidney Disease Epidemiology Collaboration formula); HFmrEF, heart failure with mid-range ejection fraction; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; HHF, heart failure hospitalisation; ICD, implantable cardioverter-defibrillator; MAP, mean arterial pressure; MRA, mineralocorticoid receptor antagonist; NTproBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; RASI, renin-angiotensin-system inhibitor; TIA, transient ischaemic attack.

    Among the variables considered in our analyses, female sex (OR 2.97, 95% CI 2.85 to 3.08, p<0.01) and no referral to specialty care follow-up (OR 1.50, 95% CI 1.44 to 1.56, p<0.01) showed strong associations with lower SES. Other important patient characteristics linked with lower SES were older age, more severe HF, higher comorbidity burden (including smoking, ischaemic heart disease, obesity, liver disease, diabetes, pulmonary disease, chronic kidney disease, anaemia, atrial fibrillation, neurologic/psychiatric disorders), more use of diuretics but less use of HF devices. Similar results were obtained when patients were stratified by ejection fraction subtype (see online supplementary figures 1–3).

    Outcome

    Over a median follow-up of 1.98 (IQR 0.40–4.76) years, 30 402 (68.1%) all-cause death/HF hospitalisation events occurred, corresponding to an event rate of 231 (95% CI 229 to 234) per 1000 patient-years. Unadjusted event rates of the composite outcome increased together with worsening SES: 157 (95% CI 153 to 161) per 1000 patient-years for no SERF, 207 (95% CI 203 to 212) for one SERF, 275 (95% CI 269 to 279) for two SERF and 358 (95% CI 348 to 367) for all three SERF (p<0.001; figure 3).

    Figure 3
    Figure 3

    Kaplan-Meier curves for heart failure hospitalisation or all-cause death according to the number of socioeconomic risk factors. Kaplan-Meier curves are shown per number of socioeconomic risk factors (income ≤median, living alone, compulsory school degree/no degree). SERF, socioeconomic risk factor.

    Multivariable Cox regression models showed lower SES to be significantly associated with a higher risk of HF hospitalisation/all-cause death. This association persisted after adjustment for demographics, clinical characteristics, comorbidities and treatments. Risk of outcome steadily increased with the increasing number of coexistent SERF. In the fully adjusted model, as compared with no SERF, HR for one SERF was 1.09 (95% CI 1.05 to 1.13; p<0.001); 1.16 (95% CI 1.12 to 1.20; p<0.001) for two SERF and 1.22 (95% CI 1.18 to 1.28, p<0.001) for three SERF (figure 4). Similar results were observed when all the different clusters/combinations of SERF and individual SERF were analysed and when analyses were performed stratifying by ejection fraction subtype (see online supplementary table 9 and figures 4–5). We explored sex interactions for the associations between SERF and outcome showing that the relative risk in females was higher as compared with males when all SERF were prevalent (p interaction <0.01, online supplementary figure 6).

    Figure 4
    Figure 4

    Association between socioeconomic status and heart failure hospitalisation or all-cause death. Models were adjusted for demographic factors: age >75 vs ≤75 years, sex, registration 2000–10 vs 2011–16; clinical factors: outpatient, ejection fraction, New York Heart Association class, N-terminal pro-B-type natriuretic peptide >2545 vs ≤2545 pg/mL (median), hospitalisation for heart failure within the last 12 months, heart rate ≥70 vs <70 bpm, mean arterial pressure ≥90 vs <90 mm Hg, duration of heart failure <6 months vs ≥6 months; comorbidities: current smoking, obesity, estimated glomerular filtration rate <60 vs ≥60 mL/min/1.73 m2, liver disease, cancer within past 3 years, anaemia, atrial fibrillation, diabetes mellitus, prior revascularisation, ischaemic heart disease, arterial hypertension, valvular heart disease, stroke/transient ischaemic attack, pulmonary disease, neurologic/psychiatric disease, history of bleeding, history of angina; treatments: implantable cardioverter defibrillator, cardiac resynchronisation therapy, beta-blocker, renin-angiotensin-system inhibitors, mineralocorticoid receptor antagonists, digoxin, stations, anticoagulants, antiplatelets, nitrates, diuretics, planned nurse-led heart failure follow-up, planned specialty care follow-up. Reference category was not having any socioeconomic risk factors. HRs are shown for comparison of having 1, 2 or 3 socioeconomic risk factors versus not having any socioeconomic risk factors.

    In the secondary outcome analysis, lower SES was associated with an increased risk of all-cause mortality but not of HF hospitalisation, and was associated with a similarly increased risk of cardiovascular versus non-cardiovascular events (see online supplementary figure 7–8).

    Discussion

    In this large, contemporary HF cohort, we observed that: (1) patients with lower SES had more severe HF, but were less likely treated with HF therapies and to be referred to specialty care follow-up; and (2) lower SES was associated with higher mortality/morbidity, which was only partially attenuated by extensive adjustments for multiple patient and treatment characteristics. Two critical and quantitatively very strong associations with lower SES were female sex and no referral to specialist follow-up.

    Patient characteristics according to SES

    Women with HF are older, have more variable symptoms and longer delay to prognosis,9 and in most10 11 though not all12 studies women are underserved by many HF interventions, especially devices. Even independent of these factors, women had an almost threefold greater odds of having low SES, and low SES was associated with worse outcomes, independent of sex and other factors. Taken together, these data suggest that strong and formal efforts are needed to improve HF care in women. Similarly, lower SES was strongly and independently linked to not being referred to specialist follow-up care, even independent of therapy and other factors. It has previously been described that specialist care is critical to receiving appropriate HF drug initiation and up-titration as well as device therapy.8 10 Additionally, these findings suggest that in this era of rapidly increasing HF prevalence together with diminishing resources, it is critical to ensure access to specialist care (eg, cardiology or internal medicine). Furthermore, in recent years, multiple new drugs, catheter-based interventions and procedures have proven beneficial in heart failure with reduced ejection fraction (HFrEF), and treatment optimisation is becoming increasingly difficult for the generalist. Thus, it has been suggested that essentially all patients with symptomatic HFrEF should be referred to the HF team for appropriate selection of HF interventions.13 14

    Another key finding was that patients with lower SES had more severe HF but were less likely treated with HF therapies and referred to specialty care follow-up. In principle, there should be no inequalities in quality or affordability of medical care in a universal healthcare system, such as in Sweden. Nevertheless, healthcare inequalities for patients with different SES have been previously reported in this and similar healthcare systems.4 This inconsistency might be at least partially explained by the fact that patients are requested to pay a fee for drug dispensation or healthcare access. Although costs are limited, fees may affect the likelihood of seeking care or collecting prescribed drugs for patients with a low income. The higher comorbidity burden, also linked to a poorer lifestyle, could be a likely explanation for the observed more severe HF in patients with lower SES versus higher SES. Indeed, in our study, several comorbidities known to be linked with both HF disease progression and overall impaired prognosis, for example, diabetes, renal disease, ischaemic heart disease, anaemia and atrial fibrillation, were independently associated with lower SES.2 However, lower SES drives disease progression in patients with HF, and is associated with incident HF in the general population.15 Important factors such as alcohol and drug abuse, smoking, unhealthy diet, physical inactivity and systemic inflammatory response can contribute per se or by fostering the onset of comorbidities linked to HF onset.16–19 Poor treatment of or adherence to therapies for comorbidities (eg, hypertension, diabetes, dyslipidaemia), delayed access to revascularisation in case of myocardial infarction leading to greater myocardial damage, are further factors which may contribute to explain the association between lower SES and both incident and more severe HF.20 21

    Another potential explanation for our finding might be linked to the association between SES and health literacy. Lower health literacy, which is more likely in patients with lower SES, might limit self-care in HF and therefore contribute to disease progression.22 Additionally, lower SES and lower health literacy can also be barriers to adequate use of healthcare resources, as they negatively influence the use of preventive measures or treatments in several cardiovascular diseases,22 23 or simply delay the first contact with primary/secondary care and prevent an adequate follow-up. Finally, present bias and naïveté, where patients avoid short-term inconvenience or discomfort at the cost of long-term complications, has been described in conditions such as type 1 diabetes24 and may be linked with SES.

    Morbidity/Mortality according to SES

    An association between SES and outcome has been previously reported in HF, but with limited adjustment for important clinical factors and treatments likely to influence the relationship between SES and prognosis.5–7 In this study, we showed that the association between SES and mortality/morbidity persisted even after extensive adjustments for relevant confounders, including demographics, clinical characteristics, comorbidities and treatments. We combined different indicators of SES and thus we could show that the magnitude of SES-associated risk of outcomes was greater when several SERF coexisted. Importantly, our definition of SES included classical measures such as income and education, and living status, which is a measure for social connectedness. This is of relevance, as it has been shown that individuals with lower social connectedness might not use the full range of available medical measures (ranging from preventive to therapeutic measures).23 Our findings highlight the importance of assessing SES in patients with HF, and lead to important questions regarding potential mechanisms or mediators.

    As we reported for the association between SES and more severe HF, it is unlikely that worse morbidity/mortality with lower SES might be mainly explained by differences in affordability or availability of care in the Swedish healthcare system. Dedicated follow-up care, as well as comorbidity burden and HF severity have been considered for adjustments and, thus, might not explain the independent association between SES and morbidity/mortality. Furthermore, lower SES was associated with all-cause mortality but not HF hospitalisation, which might be explained by the adjustments for several variables linked with HF severity. This is consistent with UK data showing lower SES to be associated with a higher risk of all-cause death, but not with HF hospitalisation, in 1802 patients with HFrEF.25 As for the primary outcome, we observed increased risk of cardiovascular and non-cardiovascular events in patients with lower SES. Health behaviours, such as unhealthy diet, sedentary lifestyle, alcohol/drug abuse, which have not been considered, might contribute to explain the overall worse prognosis in patients with lower SES.26–29 Additionally, for adjustments we considered treatments as prescribed, but lower health literacy, social support and self-care which are more likely in lower SES could have negatively impacted on adherence leading to higher mortality/morbidity.22 Limited patient participation to follow-up care might have prevented the correct up-titration of HF treatment in those with lower SES, which is known to be linked to worse outcomes.30 Additionally, lower SES is linked to a lower likelihood of being engaged in preventive measures,23 and comorbidity status might have further worsened over time, which could not be captured in our analysis where comorbidities were considered at the baseline. Finally, although we adjusted for several cardiovascular comorbidity-related therapies (eg, antiplatelets), we could not adjust for treatments required for many other comorbidities considered and not considered in this analysis.

    Strengths and limitations

    Strengths include the use of a large, well-characterised and contemporary HF cohort with data on several measures of SES, which allowed to perform extensive adjustments, and a detailed characterisation of patients with HF according to SES.

    Limitations are linked with the observational nature of this analysis, and therefore with residual/unmeasured confounding, which is likely to explain the significant association between SES and outcome. SES was defined based on income, education and living arrangement, but different definitions might provide different results. Income and living arrangement, but not education, are dynamic and might thus have varied during the follow-up, resulting in misclassification. Few variables such as New York Heart Association functional class and N-terminal pro-B-type natriuretic peptide levels had a considerable amount of missing data (25.6% and 64.2% missing, respectively), which were handled by multiple imputation. Data on accessibility to specialty care were not available. Notably, clustering of such facilities in neighbourhoods with a higher SES might contribute to explain less referral to specialty care in patients with lower SES. Finally, our study is based on a national cohort and generalisability to other countries/healthcare systems might be limited.

    Conclusions and clinical implications

    In this large and contemporary HF cohort from universal tax-financed healthcare system, patients with lower SES had more severe HF but were less likely to receive HF treatments and care. Lower SES was associated with worse prognosis, even after extensive adjustment for demographics, clinical characteristics, comorbidities and treatments.

    Our findings call for more awareness on health inequalities in HF. Identifying factors leading to worse outcomes in patients with lower SES may contribute to attenuate health disparities. Additionally, assessing SES-related factors, such as health literacy and social support, could improve risk stratification in HF. Specifically, there were two findings of critical importance that can be easily addressed: (1) female sex was associated with a nearly threefold higher odds of low SES, suggesting that extra effort should be expended on optimising care for women with HF; and (2) referral to specialist follow-up was less common with lower SES, suggesting that proper specialty care should be offered to and ideally accepted by all patients who need it and that community interventions in primary care setting are needed.

    Key messages

    What is already known on this subject?

    • Lower socioeconomic status is suggested to be associated with high risk of morbidity/mortality in heart failure, although it is debated whether this association is mediated by differences in heart failure severity or differences in quality of care.

    What might this study add?

    • This study showed that lower socioeconomic status is associated with a 9%–22% higher risk of morbidity/mortality independently of heart failure severity and quality of care.

    How might this impact on clinical practice?

    • These findings call for more awareness on health inequalities in heart failure.

    • Identifying factors leading to worse outcomes in patients with lower socioeconomic status may contribute to attenuate health disparities.

    References

    1. ↵
      2. Savarese G ,
      3. Lund LH
      . Global public health burden of heart failure. Cardiac failure review: Radcliffe Cardiology, 2017: 7–11.
    2. ↵
      2. Ponikowski P ,
      3. Voors AA ,
      4. Anker SD , et al
      . ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J 2016;2016:2129–200.
      OpenUrl
    3. ↵
      2. Magnussen C ,
      3. Niiranen TJ ,
      4. Ojeda FM , et al
      . Sex-Specific Epidemiology of Heart Failure Risk and Mortality in Europe: Results From the BiomarCaRE Consortium. JACC Heart Fail 2019;7:204–13.doi:10.1016/j.jchf.2018.08.008 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30819375
      OpenUrlAbstract/FREE Full Text
    4. ↵
      2. Mackenbach JP ,
      3. Stirbu I ,
      4. Roskam A-JR , et al
      . Socioeconomic inequalities in health in 22 European countries. N Engl J Med 2008;358:2468–81.doi:10.1056/NEJMsa0707519 pmid:http://www.ncbi.nlm.nih.gov/pubmed/18525043
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Lindenauer PK ,
      3. Lagu T ,
      4. Rothberg MB , et al
      . Income inequality and 30 day outcomes after acute myocardial infarction, heart failure, and pneumonia: retrospective cohort study. BMJ 2013;346:f521. doi:10.1136/bmj.f521 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23412830
      OpenUrlAbstract/FREE Full Text
    6. ↵
      2. Schjødt I ,
      3. Johnsen SP ,
      4. Strömberg A , et al
      . Socioeconomic Factors and Clinical Outcomes Among Patients With Heart Failure in a Universal Health Care System. JACC Heart Fail 2019;7:746–55.doi:10.1016/j.jchf.2019.06.003 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31466671
      OpenUrlAbstract/FREE Full Text
    7. ↵
      2. Lawson CA ,
      3. Zaccardi F ,
      4. Squire I , et al
      . 20-year trends in cause-specific heart failure outcomes by sex, socioeconomic status, and place of diagnosis: a population-based study. Lancet Public Health 2019;4:e406–20.doi:10.1016/S2468-2667(19)30108-2 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31376859
      OpenUrlPubMed
    8. ↵
      2. Savarese G ,
      3. Carrero J-J ,
      4. Pitt B , et al
      . Factors associated with underuse of mineralocorticoid receptor antagonists in heart failure with reduced ejection fraction: an analysis of 11 215 patients from the Swedish heart failure registry. Eur J Heart Fail 2018;20:1326–34.doi:10.1002/ejhf.1182 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29578280
      OpenUrlPubMed
    9. ↵
      2. Lund LH ,
      3. Mancini D
      . Heart failure in women. Med Clin North Am 2004;88:1321–45.doi:10.1016/j.mcna.2004.03.003 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15331319
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. Lund LH ,
      3. Braunschweig F ,
      4. Benson L , et al
      . Association between demographic, organizational, clinical, and socio-economic characteristics and underutilization of cardiac resynchronization therapy: results from the Swedish heart failure registry. Eur J Heart Fail 2017;19:1270–9.doi:10.1002/ejhf.781 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28176416
      OpenUrlPubMed
    11. ↵
      2. Lund LH ,
      3. Trochu J-N ,
      4. Meyns B , et al
      . Screening for heart transplantation and left ventricular assist system: results from the screening for advanced heart failure treatment (SEE-HF) study. Eur J Heart Fail 2018;20:152–60.doi:10.1002/ejhf.975 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28960673
      OpenUrlPubMed
    12. ↵
      2. Lainščak M ,
      3. Milinković I ,
      4. Polovina M , et al
      . Sex- and age-related differences in the management and outcomes of chronic heart failure: an analysis of patients from the ESC HFA EORP heart failure long-term registry. Eur J Heart Fail 2020;22:92–102.doi:10.1002/ejhf.1645 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31863522
      OpenUrlPubMed
    13. ↵
      2. Thorvaldsen T ,
      3. Benson L ,
      4. Ståhlberg M , et al
      . Triage of patients with moderate to severe heart failure: who should be referred to a heart failure center? J Am Coll Cardiol 2014;63:661–71.doi:10.1016/j.jacc.2013.10.017 pmid:http://www.ncbi.nlm.nih.gov/pubmed/24161453
      OpenUrlFREE Full Text
    14. ↵
      2. Crespo-Leiro MG ,
      3. Metra M ,
      4. Lund LH , et al
      . Advanced heart failure: a position statement of the heart failure association of the European Society of cardiology. Eur J Heart Fail 2018;20:1505–35.doi:10.1002/ejhf.1236 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29806100
      OpenUrlPubMed
    15. ↵
      2. Potter EL ,
      3. Hopper I ,
      4. Sen J , et al
      . Impact of socioeconomic status on incident heart failure and left ventricular dysfunction: systematic review and meta-analysis. Eur Heart J Qual Care Clin Outcomes 2019;5:169–79.doi:10.1093/ehjqcco/qcy047 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30295783
      OpenUrlCrossRefPubMed
    16. ↵
      2. Hawkins NM ,
      3. Jhund PS ,
      4. McMurray JJV , et al
      . Heart failure and socioeconomic status: accumulating evidence of inequality. Eur J Heart Fail 2012;14:138–46.doi:10.1093/eurjhf/hfr168 pmid:http://www.ncbi.nlm.nih.gov/pubmed/22253454
      OpenUrlCrossRefPubMed
    17. ↵
      2. Jousilahti P ,
      3. Salomaa V ,
      4. Rasi V , et al
      . Association of markers of systemic inflammation, C reactive protein, serum amyloid A, and fibrinogen, with socioeconomic status. J Epidemiol Community Health 2003;57:730–3.doi:10.1136/jech.57.9.730 pmid:http://www.ncbi.nlm.nih.gov/pubmed/12933781
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Schaufelberger M ,
      3. Rosengren A
      . Heart failure in different occupational classes in Sweden. Eur Heart J 2007;28:212–8.doi:10.1093/eurheartj/ehl435 pmid:http://www.ncbi.nlm.nih.gov/pubmed/17185309
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      2. Aggarwal M ,
      3. Bozkurt B ,
      4. Panjrath G , et al
      . Lifestyle Modifications for Preventing and Treating Heart Failure. J Am Coll Cardiol 2018;72:2391–405.doi:10.1016/j.jacc.2018.08.2160 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30384895
      OpenUrlFREE Full Text
    20. ↵
      2. Muñoz D ,
      3. Uzoije P ,
      4. Reynolds C , et al
      . Polypill for cardiovascular disease prevention in an underserved population. N Engl J Med 2019;381:1114–23.doi:10.1056/NEJMoa1815359 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31532959
      OpenUrlPubMed
    21. ↵
      2. Udell JA ,
      3. Desai NR ,
      4. Li S , et al
      . Neighborhood socioeconomic disadvantage and care after myocardial infarction in the National cardiovascular data registry. Circ Cardiovasc Qual Outcomes 2018;11:e004054. doi:10.1161/CIRCOUTCOMES.117.004054 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29848476
      OpenUrlAbstract/FREE Full Text
    22. ↵
      2. Magnani JW ,
      3. Mujahid MS ,
      4. Aronow HD , et al
      . Health literacy and cardiovascular disease: fundamental relevance to primary and secondary prevention: a scientific statement from the American heart association. Circulation 2018;138:e48–74.doi:10.1161/CIR.0000000000000579 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29866648
      OpenUrlAbstract/FREE Full Text
    23. ↵
      2. Stafford M ,
      3. von Wagner C ,
      4. Perman S , et al
      . Social connectedness and engagement in preventive health services: an analysis of data from a prospective cohort study. Lancet Public Health 2018;3:e438–46.doi:10.1016/S2468-2667(18)30141-5 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30143472
      OpenUrlPubMed
    24. ↵
      2. Wang Y ,
      3. Sloan FA
      . Present bias and health. J Risk Uncertain 2018;57:177–98.doi:10.1007/s11166-018-9289-z pmid:http://www.ncbi.nlm.nih.gov/pubmed/31244508
      OpenUrlPubMed
    25. ↵
      2. Witte KK ,
      3. Patel PA ,
      4. Walker AMN , et al
      . Socioeconomic deprivation and mode-specific outcomes in patients with chronic heart failure. Heart 2018;104:993–8.doi:10.1136/heartjnl-2017-312539 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29386325
      OpenUrlAbstract/FREE Full Text
    26. ↵
      2. Mellis AM ,
      3. Athamneh LN ,
      4. Stein JS , et al
      . Less is more: negative income shock increases immediate preference in cross commodity discounting and food demand. Appetite 2018;129:155–61.doi:10.1016/j.appet.2018.06.032 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29959952
      OpenUrlPubMed
    27. ↵
      2. Stringhini S ,
      3. Carmeli C ,
      4. Jokela M , et al
      . Socioeconomic status and the 25 × 25 risk factors as determinants of premature mortality: a multicohort study and meta-analysis of 1·7 million men and women. Lancet 2017;389:1229–37.doi:10.1016/S0140-6736(16)32380-7 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28159391
      OpenUrlCrossRefPubMed
    28. ↵
      2. Beard E ,
      3. Brown J ,
      4. West R , et al
      . Associations between socio-economic factors and alcohol consumption: a population survey of adults in England. PLoS One 2019;14:e0209442. doi:10.1371/journal.pone.0209442 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30716098
      OpenUrlPubMed
    29. ↵
      2. Casetta B ,
      3. Videla AJ ,
      4. Bardach A , et al
      . Association between cigarette smoking prevalence and income level: a systematic review and meta-analysis. Nicotine Tob Res 2017;19:1401–7.doi:10.1093/ntr/ntw266 pmid:http://www.ncbi.nlm.nih.gov/pubmed/27679607
      OpenUrlPubMed
    30. ↵
      2. Santema BT ,
      3. Ouwerkerk W ,
      4. Tromp J , et al
      . Identifying optimal doses of heart failure medications in men compared with women: a prospective, observational, cohort study. Lancet 2019;394:1254–63.doi:10.1016/S0140-6736(19)31792-1 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31447116
      OpenUrlPubMed

    Footnotes

    • Twitter @giasav

    • Contributors BS and GS wrote the first draft of the manuscript. LHL, DS, AO, RW, DW, AS, UD, FB, LB and JPF edited and revised for important intellectual content. All the authors approved the submitted version of the manuscript.

    • Funding This study received support from the EU/EFPIA Innovative Medicines Initiative 2 Joint Undertaking BigData@Heart grant (n° 116074). BS is funded by the German Research Foundation.

    • Competing interests BS reports personal fees from AstraZeneca. GS reports grants and personal fees from Vifor, grants from Boehringer Ingelheim, personal fees from Societa' Prodotti Antibiotici, grants from MSD, grants and personal fees from AstraZeneca, grants from Pharmacosmos, grants from Boston Scientific, personal fees from Roche, personal fees from Servier, personal fees from Medtronic, personal fees from Cytokinetics, grants from Novartis, personal fees from Genesis. UD reports grants and honoraria/consultancies from AstraZeneca, Honoraria/consultancies from Novartis and Amgen and grants from Boehringer Ingelheim. AS reports honoraria from Novartis. DW reports honorary from AstraZeneca, Bayer, Berlin-Chemie and Novartis. LHL reports personal fees from Merck, personal fees from Sanofi, grants and personal fees from Vifor-Fresenius, grants and personal fees from AstraZeneca, grants and personal fees from Relypsa, personal fees from Bayer, grants from Boston Scientific, grants and personal fees from Novartis, personal fees from Pharmacosmos, personal fees from Abbott, grants and personal fees from Mundipharma, personal fees from Medscape, personal fees from Myokardia, grants and personal fees from Boehringer Ingelheim, outside the submitted work.

    • Patient consent for publication Not required.

    • Ethics approval SwedeHF as well as this analysis were approved by a multisite ethics committee. The study was conducted in accordance with the Declaration of Helsinki. Individual patient consent was not required, but patients were informed of entry into the registry and allowed to opt out.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data availability statement Data are available on reasonable request. The data that support the findings of this study are available from the corresponding author, provided that data sharing is permitted by European Union General Data Protection Regulation regulations and appropriate ethics committees.