Circulation Journal
Online ISSN : 1347-4820
Print ISSN : 1346-9843
ISSN-L : 1346-9843

This article has now been updated. Please use the final version.

How Is Socioeconomic Status Associated With the Incidence of Heart Failure?
Aya HigashiyamaAkira Fujiyoshi
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML Advance online publication

Article ID: CJ-21-0534


An association between social structural factors and the incidence of heart failure (HF) has been reported from prospective observational studies and randomized controlled studies in Western countries.1 In a recent meta-analysis assessing 11 studies, Potter et al reported that low socioeconomic status (SES), including education, income, occupation, and area-based measures, was consistently and independently associated with increased risk of incident HF in high-income nations.1 The pooled hazard ratio (HR) and 95% confidence interval (95% CI) of low SES for incident HF was 1.62 (1.50–1.76) and that of low education alone was 1.66 (1.30–2.10).1 In the conclusion of that meta-analysis, the authors called for further work to elucidate mediators of the observed association and how SES could improve HF risk stratification.1

Article p ????

As an exemplary work by Potter et al,1 any study focused on the effects of social structural factors on states of health or disease is called “social epidemiology”. Many social epidemiologic studies have consistently shown that the distribution of advantage and disadvantage in a society reflects the distribution of health and disease. Social epidemiologists propose to identify societal characteristics that affect the pattern of disease and health distribution in a society and to understand the mechanisms2 in order to improve the health of populations. Thus, crucial questions in social epidemiology are “What is an upper-stream determinant, among socioeconomic factors, that shapes risk factors which in turn directly affect health outcomes?”, and “Through what mechanisms/mediators, does a socioeconomic determinant affect the health outcomes?”.

The association among SES, mediators and the risk of cardiovascular disease (CVD) has not been completely investigated yet. Groenewegen et al presented various possible mechanisms that may explain the relationship between SES and incident CVD; lower SES is associated with a higher prevalence of unfavorable behavioral risk factors, including physical inactivity, poor diet, smoking, and medication non-adherence. In addition, lower SES might be associated with more stress-induced autonomic and neuroendocrine activation, and worse access/utilization of health care, even in countries with universal healthcare systems.3

With regard to HF, previous studies from the Atherosclerosis Risk in Communities (ARIC) cohort in the USA have already revealed some aspects of the inverse association between educational level and incidence of HF such that SES later in adulthood had a more prominent role than earlier SES,4 and that associations between SES and incident HF diagnosis were observed in both inpatient and outpatient settings.5 With respect to mediators (i.e., mechanisms) that would connect educational attainment to the incidence of HF, however, little was known until the appearance of a paper by Lin et al6 in this issue of the Journal.

Lin et al investigate the potential mediators of the causal pathway in the association of low educational attainment with incident HF among a general population in the ARIC cohort. They show an inverse association between educational attainment and HF risk in a total of 12,109 participants with median follow-up of 25 years: hazard ratio (95% confidence interval) of incident HF was 1.41 (1.26–1.57) for low educational attainment, and 1.13 (1.02–1.25) for medium educational attainment, compared with high educational attainment after adjustment for possible confounders. In addition to these confirmatory findings, they report that the association was partially mediated by income, waist-to-hip ratio (obesity), current smoking, body mass index (BMI; obesity), current drinking, sports and physical activity, each of which separately explained 24.3%, 20.2%, 13.8%, 10.1%, 7.7%, 7.3% and 4.5% of the relationship, according to their mediation analysis. Those mediators in combination contributed 56.3% to the total effect of low-to-medium educational attainment on HF risk. The authors concluded that lower educational attainment was associated with increased risk of HF, and income, obesity and current smoking mediated a great proportion of the total effect of the association.

We are concerned, however, about the validity of the authors’ path diagram for the following reasons: lack of dietary components especially those linked to elevated blood pressure (i.e., high sodium and low potassium intake), placing income and other candidate mediators at the same level/time point in the pathway, dual inclusion of BMI and waist-hip ratio as potential mediators despite no direct evidence supporting waist-hip ratio being an independent risk factor from BMI for HF. Owing to those concerns, we are unsure whether we should take the authors’ estimates at their face value. Based on their findings, they propose “comprehensive lifestyle interventions targeting weight loss”, “interventions targeting smoking cessation, such as cessation counseling and pre-cessation medication use.” We agree on the importance of such interventions to lower the risk of not only HF but also other chronic diseases. At the same time, however, we also know that people in low SES are more difficult to be recruited to general population interventions, and they achieve poorer behavioral change following interventions, resulting in poorer health outcomes compared with more affluent participants.7 Because the motivation behind conducting mediation analysis is to improve our understanding of the mechanism(s) through which exposure causes outcome,8 research utilizing mediation analysis should be tailored to identify mediators on which an effective intervention can be administered to improve health outcomes in the context of modern public health. The central questions for such research are 2-fold: “Why people in low SES have worse health outcomes?”, and “What are the targets of effective intervention for better health?”.

As the authors and other ARIC investigators speculated, deprived residential area, fewer social resources, less problem-solving ability/learned effectiveness and less personal control, all may attribute to worse health in the people with low SES.9 In future research, therefore, we suggest those factors along with other novel ones be evaluated as potential mediators and targets for effective interventions. Lin et al’s work is a good start, but there is much to do in closing the health gap across the SES groups. Well-shaped study questions are very much in need for future research in the area.


This article is licensed under a Creative Commons [Attribution-NonCommercial-NoDerivatives 4.0 International] license.