Neighborhood Socioeconomic Context and Long-Term Survival After Myocardial Infarction
Background— Neighborhood of residence has been suggested to affect cardiovascular risk above and beyond personal socioeconomic status (SES). However, such data are currently lacking for patients with myocardial infarction (MI). We examined all-cause and cardiac mortality according to neighborhood SES in a cohort of MI patients.
Methods and Results— Consecutive patients ≤65 years of age discharged from 8 hospitals in central Israel after incident MI in 1992 to 1993 were followed up through 2005. Individual data were obtained at study entry, including education, income, and employment. Neighborhood SES was estimated through a composite census-derived index developed by the Israel Central Bureau of Statistics. During follow-up, 326 deaths occurred in 1179 patients. Patients residing in disadvantaged neighborhoods had higher mortality rates, with 13-year survival estimates of 61%, 74%, and 82% in increasing tertiles (Ptrend<0.001). After adjustment for sociodemographic variables, traditional risk factors, MI severity indexes, and individual SES measures, the hazard ratios for death associated with neighborhood SES were 1.47 (95% confidence interval, 1.05 to 2.06) in the lower and 1.19 (95% confidence interval, 0.86 to 1.63) in the middle tertiles compared with the upper tertile (Ptrend=0.02). The respective hazard ratios were even stronger for cardiac death (1.63; 95% confidence interval, 1.09 to 2.25; and 1.41; 95% confidence interval, 0.96 to 2.07). In the final models, neighborhood context and several individual SES measures were concurrently associated with all-cause and cardiac mortality.
Conclusions— Neighborhood SES is strongly associated with long-term survival after MI. The association is partly, but not entirely, attributable to individual SES and clinical characteristics. These data support a multidimensional relationship between SES and MI outcome.
Measures of socioeconomic status (SES) have been identified as important prognostic factors after myocardial infarction (MI).1–14 Yet, because SES is a multidimensional construct comprising various indicators acting at different levels,15,16 the mechanisms by which it affects post-MI prognosis are incompletely understood. Although SES has traditionally been treated as an intrinsic characteristic of individuals, contextual (or macro) effects on health are theoretically and empirically important.17 Indeed, a growing body of evidence indicates that a person’s health may be influenced by the socioeconomic characteristics of the neighborhood in which he or she lives, over and above his or her own SES.18–21 This suggests that characteristics of communities represent more than the aggregation of their residents’ characteristics. Furthermore, as recently pointed out in a critical review of the use of SES in health research, “one size does not fit all,” suggesting that findings from studies that have measured limited aspects of SES should be interpreted with caution.15 Accordingly, a multiple jeopardy hypothesis has been invoked to describe the complex association of SES with cardiovascular disease (CVD) risk but infrequently tested.22–25 This is particularly true after MI; most previous studies were limited to a single SES indicator or a few indicators measured at a single level. Because the effect of each socioeconomic determinant can be either explained in part by or mediated through other determinants,16 current knowledge on this topic is insufficient. This population-based cohort study was undertaken to examine the long-term survival of patients with incident MI according to the SES level of their residential neighborhood after taking into account individual SES measures and other potential confounding factors.
Editorial see p 348
Clinical Perspective on p 383
The Israel Study of First Acute Myocardial Infarction is a longitudinal prospective investigation of the role of sociodemographic, medical, and psychosocial variables measured in patients hospitalized with incident (ie, first ever) MI in long-term clinical outcome, psychosocial adjustment, and quality of life.13,26,27 Between February 15, 1992, and February 15, 1993, a total of 1626 consecutive MI patients ≤ 65 years of age were admitted to 8 Israeli medical centers. These hospitals provide care to the entire population of central Israel (2¼ million in 1992). Of these patients, 81 (5%) died during hospitalization, leaving 1545 eligible candidates for the study, among whom 1521 (98%) consented to participate.
The diagnosis of MI was established by the presence of at least 2 of the following criteria: characteristic chest pain lasting at least 20 minutes, creatine kinase elevation ≥1.5 times the upper limit of normal or creatine kinase-MB fraction >5% when simultaneous reference creatine kinase levels exceeded the upper limit of normal, and ECG changes compatible with Q-wave or non–Q-wave MI. Subjects with prior infarction were excluded.
Data at study entry and during follow-up were obtained through interviews, psychosocial questionnaires, and review of the entire medical record.26,27 The baseline interview was based on a detailed, structured questionnaire prepared by the principal investigator (Y.D.), which was previously examined in a pilot study. All clinical data were verified by a senior cardiologist.
Of the participants in the parent study, 1410 were assigned a census-derived SES score based on their geocoded address at time of enrollment. The remaining subjects could not be linked, mainly because they resided in small villages where such data are unavailable. Patients in these smaller villages were similar to others with regard to measures of personal SES collected from the baseline interview (eg, education, employment, and income). An addition 231 patients were excluded from analysis because of missing information on personal SES indicators (mainly income), leaving 1179 participants with complete data for this study. Nonparticipants did not differ from participants in mean age (54.0 versus 53.7 years; P=0.52), proportion of women (18% versus 19%; P=0.76), average education (11.0 versus 11.0 years; P=0.83), or cumulative survival (71% versus 73% at 13 years; P=0.55). All aspects of the study were approved by the appropriate institutional ethics committees.
Individual SES data were self-reported at the index hospitalization and included the following variables: family income relative to the national average (categorized as below average, average, and above average), education (years of schooling), pre-MI employment (classified as full time, part time, and none), and living with a steady partner. The last variable was considered a socioeconomic measure on the basis of the Hollingshead 4-factor SES index.28
Neighborhood SES was estimated through a composite index developed and validated by the Israel Central Bureau of Statistics.29 The index, which is based on the 1995 National Census data, is a summary of numerous socioeconomic measures (ie, demographic, schooling and education, standard of living, employment and unemployment, and social benefits; see the online-only Data Supplement) that allows the classification of small geographical units (≈2000 inhabitants) into SES classes using a 20-point scale. Geographic Information System tools were used to assign a SES score to each patient on the basis of his/her geocoded address at the time of the index MI. For ease of presentation and interpretation, approximate tertiles of the score distribution (Figure 1) were used in the analysis (≤10, 11 to 14, and ≥15 points, including 362, 446, and 371 patients, respectively).
The study included patients residing in 512 distinct statistical units (“neighborhoods”). The neighborhoods are located within an area of ≈400 sq miles and are predominantly urban. Medical facilities (hospitals, clinics, first-aid stations) are well distributed in this area of central Israel and serve the entire population according to the National Medical Insurance Law. At any location, the distance to the nearest hospital does not exceed 7 miles. However, the availability of intensive care unit (ICU) and cardiac rehabilitation facilities is heterogeneous. For example, in 1992, the number of ICU beds in the participating hospitals ranged from 5 to 9.
The entire inpatient and outpatient medical records and data obtained through structured detailed interviews were used to ascertain CVD risk factors, MI characteristics, disease severity indexes, and acute management. Measurements recorded at the index hospitalization or at the closest time before or after the index date were considered. Obesity was defined as body mass index ≥30 kg/m2. Diabetes, hypertension, and dyslipidemia were defined clinically. Smoking was classified into (pre-MI) current or noncurrent smoking. Leisure-time physical activity was self-reported and dichotomized into active (ie, regular physical activity) or nonactive (irregular or no physical activity) in the year preceding the index MI. Comorbidity was assessed by the Charlson Index30 and analyzed categorically (no comorbidity for 0 points, moderate comorbidity for 1 to 2 points, and severe comorbidity for ≥3 points). MI characteristics and severity indexes included infarct type and location, Killip class, admission to ICU, and in-hospital complications. Reperfusion therapy and revascularization included thrombolysis, percutaneous transluminal coronary angioplasty, and coronary artery bypass grafting. Revascularization referred to early procedures performed within 45 days of the index date. Self-rated health, a 5-point scale measure, was used to assess general health.27 Patients were asked about their health in the year before the index hospitalization. Poor self-rated health was defined as a score of 1 or 2.
Follow-up began at the date of the index hospitalization and lasted through December 31, 2005 (median, 13.2 years; interquartile range, 12.0 to 13.5 years). Death (ie, date, location, and cause) was ascertained by a senior cardiologist (Y.D.) through various sources, including data from the Israeli Population Registry, death certificates, hospital charts, family physicians, and family members. The Israeli Population Registry has complete ascertainment of death.31 The primary study end points were all-cause and cardiac mortality. The latter was defined as death caused by coronary heart disease or other diseases of the heart.32
Analyses were performed with SAS statistical software version 9.1 (SAS Institute Inc, Cary, NC) and SPSS version 15.0 (SPSS Inc, Chicago, Ill). Baseline characteristics across neighborhood SES tertiles were compared by the χ2 test for categorical variables and ANOVA for continuous variables.
Survival across neighborhood SES tertiles was assessed by the Kaplan-Meier method with right censoring at the time of last follow-up and compared by the log-rank test. Cox proportional-hazards models were constructed to evaluate the hazard ratios (HRs) and 95% confidence intervals (CIs) for all-cause mortality and cardiac mortality associated with neighborhood SES. Because individuals may be more alike with respect to unmeasured characteristics within a given neighborhood than they are between neighborhoods, random-effects models were used, accounting for potential intraneighborhood correlation.33,34 This was done with a SAS macro.35 The proportional-hazards assumption was tested with the Schoenfeld residuals, with no violations detected. Initial adjustment was made for age, sex, and origin. Subsequently, traditional risk factors, MI characteristics, disease severity indexes, and acute management were controlled for. Finally, personal socioeconomic measures were added. There were no missing values in the models evaluated. Model discrimination was assessed by the c statistic, which represents the area under the receiver-operating characteristic curve. Calculating the c statistic from Cox proportional-hazards models was performed through a SAS macro written by W.K. Kremers.36 All P values were 2 tailed.
The analysis included 1179 incident MI patients with a mean age of 54 years (SD, 8 years); 19% were women. The percentage of patients with <12 years of education was 51%, whereas 47% reported a below-average income and 24% were unemployed before MI. As expected, CVD risk factors were highly prevalent at baseline, particularly smoking (51%), hypertension (38%), and physical inactivity (72%). Fewer than half (43%) of the patients received thrombolytic therapy on hospital admission, and 24% underwent revascularization within 45 days of hospitalization. Patient characteristics across neighborhood SES tertiles are presented in Table 1. On average, patients residing in lower-SES neighborhoods were less educated and more likely to be of Middle Eastern/North African origin, to live alone, and to report a below-average income and unemployment. They also had higher prevalence of diabetes and smoking, were less frequently engaged in physical activity, and had more severe disease, as indicated by higher Killip class, more comorbidities, and poorer self-rated health compared with patients living in higher-SES neighborhoods.
|Characteristic||Neighborhood SES Tertiles||P|
|Lower (n=362), %||Middle (n=446), %||Upper (n=371), %|
|CABG indicates coronary artery bypass graft; PTCA, percutaneous transluminal coronary angioplasty.|
|Age, mean±SD, y||54.2±8.0||53.0±8.6||54.1±7.9||0.09|
|Middle Eastern/North African origin||46||36||16||<0.001|
|Education, mean±SD, y||9.1±4.8||10.9±3.9||12.9±3.6||<0.001|
|Living with a steady partner||83||88||88||0.05|
|Cardiovascular risk factors|
|Leisure-time physical activity||21||29||34||<0.001|
|MI characteristics, comorbidity, and acute management|
|CABG w/in 45 d||4||6||8||0.09|
|PTCA w/in 45 d||14||13||27||<0.001|
|Charlson comorbidity index||0.008|
|Killip class >1||27||19||20||0.01|
|Poor self-rated health||24||20||14||0.002|
Neighborhood Socioeconomic Context and Outcomes
Over a total of 13 535 person-years of follow-up, 326 patients died; 233 of these deaths (71%) were cardiac deaths. Overall survival differed substantially by SES category, with a dose-response pattern of a relationship between neighborhood SES and long-term survival after MI (Figure 2). The 13-year survival estimates in the lower through upper neighborhood SES tertiles were 61%, 74%, and 82%, respectively (P for trend <0.001). In random-effects Cox analyses accounting for possible spatial autocorrelation, the HRs for death in the lower versus upper SES tertiles were 2.37 (95% CI, 1.75 to 3.20) after adjustment for age, sex, and origin; 1.75 (95% CI, 1.28 to 2.40) after the addition of classic risk factors, MI characteristics, disease severity indexes, and acute management; and 1.47 (95% CI, 1.05 to 2.06) after further adjustment for individual SES measures (Table 2). An even stronger association was noted for cardiac death, with respective estimates of 2.91 (95% CI, 2.02 to 4.21), 2.09 (95% CI, 1.42 to 3.05), and 1.63 (95% CI, 1.09 to 2.45). Tests for linear trend revealed statistically significant decreases in all-cause (χ2 change, 5.51; P=0.02) and cardiac mortality (χ2 change, 5.45; P=0.02) with increasing neighborhood SES tertiles in the final models. The c statistics for the final models were 0.75 for all-cause mortality and 0.76 for cardiac mortality, which indicate a fair to good discriminatory ability.
|Adjustment||Neighborhood SES, HR (95% CI)||P for Trend|
|First (Lower) Tertile (n=362)||Second Tertile (n=446)||Third Tertile (n=371)|
|Model 1: adjusted for age, sex, origin, CVD risk factors (hypertension, diabetes mellitus, dyslipidemia, smoking, and physical activity), and disease severity indexes (admission to ICU, anterior MI, comorbidity index, Killip class, coronary artery bypass graft within 45 days, percutaneous coronary angioplasty within 45 days, and self-rated health). Model 2: model 1 plus individual-level socioeconomic measures (education, income, pre-MI employment, and living with a steady partner).|
|Unadjusted||2.41 (1.80–3.23)||1.40 (1.04–1.90)||1 (Reference)||<0.001|
|Age, sex, and origin||2.37 (1.75–3.20)||1.45 (1.07–1.98)||1 (Reference)||<0.001|
|Model 1||1.75 (1.28–2.40)||1.28 (0.93–1.75)||1 (Reference)||<0.001|
|Model 2||1.47 (1.05–2.06)||1.19 (0.86–1.63)||1 (Reference)||0.02|
|Unadjusted||2.98 (2.09–4.26)||1.78 (1.23–2.58)||1 (Reference)||<0.001|
|Age, sex, and origin||2.91 (2.02–4.21)||1.81 (1.25–2.63)||1 (Reference)||<0.001|
|Model 1||2.09 (1.42–3.05)||1.56 (1.07–2.28)||1 (Reference)||<0.001|
|Model 2||1.63 (1.09–2.45)||1.41 (0.96–2.07)||1 (Reference)||0.02|
Multiple Socioeconomic Measures and MI Prognosis
The multivariable-adjusted associations of various SES determinants with all-cause and cardiac mortality are shown in Table 3. Education, employment, income, living with a steady partner, and neighborhood SES were all strongly associated with both end points (model 1). Furthermore, the association of most of these measures persisted even on simultaneous adjustment for all other SES indicators (model 2). These data demonstrate the multidimensional importance of SES, and neighborhood context as an integral part of it, in long-term survival after MI.
|SES Characteristic||HR (95% CI)|
|All-Cause Death (n=326)||Cardiac Death (n=233)|
|Model 1||Model 2||Model 1||Model 2|
|Model 1: adjusted for age, sex, origin, hypertension, diabetes, dyslipidemia, smoking, physical activity, ICU, anterior MI, comorbidity index, Killip class, coronary artery bypass graft within 45 days, percutaneous coronary angioplasty within 45 days, and self-rated health. Model 2: model 1 plus all other SES measures shown in the table.|
|Education, 3 y||0.85 (0.78–0.92)||0.91 (0.84–0.99)||0.79 (0.72–0.87)||0.86 (0.78–0.96)|
|Below average||1.59 (1.14–2.20)||1.07 (0.75–1.54)||1.84 (1.23–2.75)||1.15 (0.74–1.78)|
|Average||0.98 (0.67–1.42)||0.87 (0.59–1.27)||1.14 (0.72–1.79)||0.96 (0.60–1.53)|
|Above average||1 (Reference)||1 (Reference)||1 (Reference)||1 (Reference)|
|Unemployment||2.04 (1.56–2.68)||1.74 (1.31–2.31)||2.08 (1.51–2.87)||1.71 (1.22–2.39)|
|Part time||1.51 (0.98–2.32)||1.35 (0.87–2.09)||1.48 (0.87–2.50)||1.33 (0.78–2.26)|
|Full time||1 (Reference)||1 (Reference)||1 (Reference)||1 (Reference)|
|Living with a partner||0.59 (0.44–0.79)||0.66 (0.49–0.88)||0.61 (0.44–0.87)||0.70 (0.49–0.99)|
|Neighborhood SES tertile|
|1 (Lower)||1.75 (1.28–2.40)||1.47 (1.05–2.06)||2.09 (1.42–3.05)||1.63 (1.09–2.45)|
|2||1.28 (0.93–1.75)||1.19 (0.86–1.63)||1.56 (1.07–2.28)||1.41 (0.96–2.07)|
|3||1 (Reference)||1 (Reference)||1 (Reference)||1 (Reference)|
In this prospective cohort study of community-dwelling patients <65 years of age discharged from hospital after incident MI and followed up for a median of 13 years, neighborhood SES was strongly related to survival in a dose-response manner. Adjustment for various individual socioeconomic measures and established prognostic factors attenuated the association, yet excess risk persisted for living in a disadvantaged neighborhood. The prognostic importance of neighborhood SES was even stronger with regard to cardiac mortality. In the final models, several SES measures, including neighborhood context, were concurrently associated with all-cause and cardiac mortality, supporting a multiple jeopardy phenomenon.
Comparison With Other Studies
Primary Prevention Studies
A number of primary prevention studies have shown an increased CVD risk associated with poor area-level SES after controlling for individual-level measures.22–25 For example, Diez Roux et al22 have used a summary neighborhood score based on US Census data and found an association between residence in a disadvantaged neighborhood and increased incidence of coronary heart disease after adjustment for personal income, education, and occupation. Comparably, Sundquist et al,23 using follow-up data of 25 319 subjects in Sweden, have shown that low levels of both neighborhood education and neighborhood income predicted incident coronary heart disease after adjustment for age, sex, and individual income and education (HRs of 1.25 and 1.23 for the lowest and highest quintiles, respectively).
Socioeconomic position in post-MI research has traditionally been treated as an intrinsic characteristic of individuals. Several studies have shown that low-SES patients, usually defined on the basis of personal income, educational attainment, and occupational status, typically present with more severe disease as indicated by longer history of CVD, higher prevalence of risk factors and comorbidity, and more complicated MI.5,8–10,13 Some,5,10,11,13 but not all,14,37 studies have further revealed that despite their higher-risk profiles, low-SES patients are less likely to receive reperfusion therapy and evidence-based medications (ie, aspirin, statins, and β-blockers) and to be treated with invasive cardiac procedures, particularly angiography and percutaneous coronary interventions. Nearly all previous studies eventually found an increased mortality risk associated with low personal SES.4,5,8–11,13
Many studies have used area-based SES measures as proxies for individual measures.3,6,7,12,38–40 These studies generally reported findings similar to those of studies that used personal SES measures, although the observed effect sizes were usually smaller. The weaker regression coefficients may be due in part to the inherent exposure misclassification resulting from using aggregate geographic data to proxy individual SES.41
Only a few studies have attempted to evaluate the combined impact of personal-level and neighborhood-level SES on mortality after MI. Winkleby et al42 have assessed the independent association of neighborhood deprivation with coronary heart disease incidence and case fatality rate in Sweden. In multivariable models, neighborhood deprivation was significantly associated with both coronary heart disease incidence and 1-year case fatality after adjustment for 7 personal socioeconomic factors. However, no clinical data were available in this study, which leaves substantial room for confounding. The roles of personal-level education and neighborhood-level income in survival after MI were evaluated in a recent cohort study (median follow-up, 13 months) conducted in Olmsted County, Minnesota.14 Both poor individual education and low neighborhood income were associated with a worse clinical presentation, and the latter was shown to be a powerful predictor of mortality even after adjustment for personal education and a variety of potential confounding factors. However, because education was the sole personal SES characteristic measured in the study, it is likely that neighborhood income has acted, at least partly, as a proxy for unmeasured characteristics of individual SES. In summary, previous post-MI research was generally restricted to the prognostic role of either individual-level or neighborhood-level SES measures. The few studies that attempted to evaluate the combined effects of personal- and neighborhood-level SES lacked essential data on either personal SES determinants or clinical characteristics and were limited by a relatively short follow-up period.
What This Study Adds
To the best of our knowledge, this work is the first to evaluate the role of neighborhood socioeconomic context in long-term post-MI prognosis after accounting for multiple personal measures of SES and clinical confounders. A positive relationship was found between individual and neighborhood SES, which supports previous reports.41,43,44 Furthermore, the results suggest an additional contribution of residential neighborhood to mortality risk beyond individual SES. Risk of death, particularly cardiac death, was significantly higher in patients living in disadvantaged neighborhoods compared with their more advantaged counterparts in a dose-response manner. Notably, repetition of findings in a new context adds support to a causal relationship.45 Our finding of a role for neighborhood SES in long-term survival after MI supports previous reports from primary prevention studies and raises the possibility of a causal relationship. Furthermore, our study population has the distinct advantage of being ethnically, culturally, and socioeconomically diverse, an ideal setting for such research.
The strong association shown in this study between neighborhood SES and post-MI prognosis is in accordance with a contextual effect. The alternative explanation would be that neighborhood SES merely reflected unmeasured characteristics of individual SES. However, it is unlikely that this is the main explanation for the findings because we minimized the possibility of residual confounding by personal SES by simultaneously adjusting for education, income, employment, and living with a spouse, all of which are known prognostic factors after MI.9,10,13,46
Theoretically, neighborhood features that may affect MI outcome can be classified into characteristics of the physical environment and characteristics of the social environment.47 In our study, patients with higher neighborhood SES were more likely to have been hospitalized in an ICU and to have undergone revascularization, even though they were typically healthier to begin with and suffered less severe MI. This might be related to unequal hospital resources; the availability of ICU facilities has been higher in our study for hospitals serving more advantaged populations (around Tel Aviv) compared with those serving the less affluent. Similarly, cardiac rehabilitation facilities have been more available in more advantaged areas. The observed inequality in the provision of health services supports previous reports showing pronounced socioeconomic effects on access to specialized cardiac services, use of invasive procedures, and treatment with evidence-based therapies after MI.3,39,40
Neighborhood effects can also be related to characteristics of the social environment, which include social norms and prevailing attitudes toward health and health-related behavior (eg, smoking, diet, and physical activity) and features of the social connections within neighborhoods such as social cohesion and support.47 Such connections may lead to better sharing of health-related information and expertise gained from experience with the healthcare system. Hypothetically, patients from high-SES neighborhoods and their families may have been more vigilant in their responses to symptoms, more knowledgeable regarding various treatment options, and more assertive at obtaining them. Alternatively, it cannot be ruled out that among neighborhoods with lower SES, there might be additional sources of stress (eg, crime) and consequent anxiety among patients in these areas that might further aggravate risk. These mechanisms should be the focus of future research.
Although personal SES was measured more directly and more comprehensively than in many previous studies, the information was self-reported and sometimes included a limited number of categories, both of which may result in some misclassification. In addition, socioeconomic changes occurring after hospital discharge, including residential mobility, were not accounted for. Yet, baseline SES measures were recently shown to predict long-term mortality in this cohort effectively.13 By nature of the length of follow-up, the participants were enrolled in the early 1990s. Because changes have taken place both in the definition of MI48 and in coronary intervention practice,49 confirmation of our findings in more recent cohorts is needed. Seventy-eight percent of the original enrollees were included in the analysis. The primary reason for exclusion was missing values on SES measures, the exposure of interest. However, nonparticipants did not differ from participants in any of the sociodemographic or clinical variables evaluated, including long-term survival. This argues against substantial selection bias. Finally, although causes of death were carefully ascertained through multiple sources of information, misclassification of cardiac deaths is possible.
Our findings indicate the importance of enhancing preventive/rehabilitative healthcare services for post-MI patients residing in deprived neighborhoods. These services should be applied at the community level. Cardiac rehabilitation programs, previously shown to be cost-effective both from a community perspective and from the patients’ perspective,50 need to be modified to meet patients’ and community needs and to overcome the barriers for participation. Further research is warranted to identify the most relevant features of residential environments that affect MI outcome. These studies should include the elderly MI population and incorporate psychosocial measures such as depression, social support, and sense of coherence.
This study suggests a role for neighborhood context in long-term survival after MI beyond individual SES, thereby supporting a multidimensional association between SES and post-MI outcome.
The following investigators and institutions took part in the Israel Study Group on First Acute Myocardial Infarction: Yaacov Drory, MD, principal investigator, Department of Rehabilitation, Sackler Medical School, Tel Aviv University, Tel Aviv; Yeheskiel Kishon, MD, Michael Kriwisky, MD, and Yoseph Rosenman, MD, Wolfson Medical Center, Holon; Uri Goldbourt, PhD, Hanoch Hod, MD, Eliezer Kaplinsky, MD, and Michael Eldar, MD, Sheba Medical Center, Tel Hashomer; Itzhak Shapira, MD, Amos Pines, MD, Margalit Drory, MSW, Arie Roth, MD, Shlomo Laniado, MD, and Gad Keren, MD, Tel-Aviv Sourasky Medical Center, Tel-Aviv; Daniel David, MD, Morton Leibowitz, MD, and Hana Pausner, MD, Meir Medical Center, Kfar Sava; Zvi Schlesinger, MD, and Zvi Vered, MD, Assaf Harofeh Medical Center, Zerifin; Alexander Battler, MD, Alejandro Solodky, MD, and Samuel Sclarovsky, MD, Beilinson Medical Center, Petach Tikvah; Izhar Zehavi, MD, and Rachel Marom-Klibansky, MD, Hasharon Medical Center, Petah Tikvah; and Ron Leor, MD, Laniado Medical Center, Netanya. The authors are also indebted to Zalman Kaufman, MSc, for assistance with the Geographic Information System analysis.
Source of Funding
This work was supported in part by the Israel National Institute for Health Policy and Health Services Research.
- 1 Weinblatt E, Ruberman W, Goldberg JD, Frank CW, Shapiro S, Chaudhary BS. Relation of education to sudden death after myocardial infarction. N Engl J Med. 1978; 299: 60–65.CrossrefMedlineGoogle Scholar
- 2 Tofler GH, Muller JE, Stone PH, Davies G, Davis VG, Braunwald E. Comparison of long-term outcome after acute myocardial infarction in patients never graduated from high school with that in more educated patients: Multicenter Investigation of the Limitation of Infarct Size (MILIS). Am J Cardiol. 1993; 71: 1031–1035.CrossrefMedlineGoogle Scholar
- 3 Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med. 1999; 341: 1359–1367.CrossrefMedlineGoogle Scholar
- 4 Salomaa V, Miettinen H, Niemela M, Ketonen M, Mahonen M, Immonen-Raiha P, Lehto S, Vuorenmaa T, Koskinen S, Palomaki P, Mustaniemi H, Kaarsalo E, Arstila M, Torppa J, Kuulasmaa K, Puska P, Pyorala K, Tuomilehto J. Relation of socioeconomic position to the case fatality, prognosis and treatment of myocardial infarction events: the FINMONICA MI Register Study. J Epidemiol Community Health. 2001; 55: 475–482.CrossrefMedlineGoogle Scholar
- 5 Rao SV, Kaul P, Newby LK, Lincoff AM, Hochman J, Harrington RA, Mark DB, Peterson ED. Poverty, process of care, and outcome in acute coronary syndromes. J Am Coll Cardiol. 2003; 41: 1948–1954.CrossrefMedlineGoogle Scholar
- 6 Rao SV, Schulman KA, Curtis LH, Gersh BJ, Jollis JG. Socioeconomic status and outcome following acute myocardial infarction in elderly patients. Arch Intern Med. 2004; 164: 1128–1133.CrossrefMedlineGoogle Scholar
- 7 Tonne C, Schwartz J, Mittleman M, Melly S, Suh H, Goldberg R. Long-term survival after acute myocardial infarction is lower in more deprived neighborhoods. Circulation. 2005; 111: 3063–3070.LinkGoogle Scholar
- 8 Lee JR, Paultre F, Mosca L. The association between educational level and risk of cardiovascular disease fatality among women with cardiovascular disease. Womens Health Issues. 2005; 15: 80–88.CrossrefMedlineGoogle Scholar
- 9 Rasmussen JN, Rasmussen S, Gislason GH, Buch P, Abildstrom SZ, Kober L, Osler M, Diderichsen F, Torp-Pedersen C, Madsen M. Mortality after acute myocardial infarction according to income and education. J Epidemiol Community Health. 2006; 60: 351–356.CrossrefMedlineGoogle Scholar
- 10 Alter DA, Chong A, Austin PC, Mustard C, Iron K, Williams JI, Morgan CD, Tu JV, Irvine J, Naylor CD. Socioeconomic status and mortality after acute myocardial infarction. Ann Intern Med. 2006; 144: 82–93.CrossrefMedlineGoogle Scholar
- 11 Bernheim SM, Spertus JA, Reid KJ, Bradley EH, Desai RA, Peterson ED, Rathore SS, Normand SL, Jones PG, Rahimi A, Krumholz HM. Socioeconomic disparities in outcomes after acute myocardial infarction. Am Heart J. 2007; 153: 313–319.CrossrefMedlineGoogle Scholar
- 12 Chang WC, Kaul P, Westerhout CM, Graham MM, Armstrong PW. Effects of socioeconomic status on mortality after acute myocardial infarction. Am J Med. 2007; 120: 33–39.CrossrefMedlineGoogle Scholar
- 13 Gerber Y, Goldbourt U, Drory Y. Interaction between income and education in predicting long-term survival after acute myocardial infarction. Eur J Cardiovasc Prev Rehabil. 2008; 15: 526–532.CrossrefMedlineGoogle Scholar
- 14 Gerber Y, Weston SA, Killian JM, Therneau TM, Jacobsen SJ, Roger VL. Neighborhood income and individual education: effect on survival after myocardial infarction. Mayo Clin Proc. 2008; 83: 663–669.CrossrefMedlineGoogle Scholar
- 15 Braveman PA, Cubbin C, Egerter S, Chideya S, Marchi KS, Metzler M, Posner S. Socioeconomic status in health research: one size does not fit all. JAMA. 2005; 294: 2879–2888.CrossrefMedlineGoogle Scholar
- 16 Lahelma E, Martikainen P, Laaksonen M, Aittomaki A. Pathways between socioeconomic determinants of health. J Epidemiol Community Health. 2004; 58: 327–332.CrossrefMedlineGoogle Scholar
- 17 Rose G. Sick individuals and sick populations. Int J Epidemiol. 1985; 14: 32–38.CrossrefMedlineGoogle Scholar
- 18 Anderson RT, Sorlie P, Backlund E, Johnson N, Kaplan GA. Mortality effects of community socioeconomic status. Epidemiology. 1997; 8: 42–47.CrossrefMedlineGoogle Scholar
- 19 Diez-Roux AV. Bringing context back into epidemiology: variables and fallacies in multilevel analysis. Am J Public Health. 1998; 88: 216–222.CrossrefMedlineGoogle Scholar
- 20 Pickett KE, Pearl M. Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical review. J Epidemiol Community Health. 2001; 55: 111–122.CrossrefMedlineGoogle Scholar
- 21 Robert SA. Socioeconomic position and health: the independent contribution of community context. Annu Rev Sociol. 1999; 25: 489–516.CrossrefGoogle Scholar
- 22 Diez Roux AV, Merkin SS, Arnett D, Chambless L, Massing M, Nieto FJ, Sorlie P, Szklo M, Tyroler HA, Watson RL. Neighborhood of residence and incidence of coronary heart disease. N Engl J Med. 2001; 345: 99–106.CrossrefMedlineGoogle Scholar
- 23 Sundquist K, Winkleby M, Ahlen H, Johansson SE. Neighborhood socioeconomic environment and incidence of coronary heart disease: a follow-up study of 25,319 women and men in Sweden. Am J Epidemiol. 2004; 159: 655–662.CrossrefMedlineGoogle Scholar
- 24 Stjarne MK, Fritzell J, De Leon AP, Hallqvist J. Neighborhood socioeconomic context, individual income and myocardial infarction. Epidemiology. 2006; 17: 14–23.CrossrefMedlineGoogle Scholar
- 25 Smith GD, Hart C, Watt G, Hole D, Hawthorne V. Individual social class, area-based deprivation, cardiovascular disease risk factors, and mortality: the Renfrew and Paisley Study. J Epidemiol Community Health. 1998; 52: 399–405.CrossrefMedlineGoogle Scholar
- 26 Drory Y. Medical Outcome and Psychosocial Adjustment in Post Myocardial Infarction Patients: 10–13 Years Follow-Up [in Hebrew]. Tel Aviv, Israel: Drory; 2008: 1–409.Google Scholar
- 27 Gerber Y, Benyamini Y, Goldbourt U, Drory Y. Prognostic importance and long-term determinants of self-rated health after initial acute myocardial infarction. Med Care. 2009; 47: 342–349.CrossrefMedlineGoogle Scholar
- 28 Hollingshead AB. Hollingshead Four Factor Index of Social Status. New Haven, Conn: Yale University; 1975.Google Scholar
- 29 Burck L, Feinstein Y. Characterization and Classification of Geographical Units by the Socioeconomic Level of the Population. Jerusalem: Israel Central Bureau of Statistics; 2000. Series of 1995 Census of Population and Housing Publications, No. 13.Google Scholar
- 30 Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987; 40: 373–383.CrossrefMedlineGoogle Scholar
- 31 Modan M, Karp M, Bauman B, Gordon O, Danon YL, Laron Z. Mortality in Israeli Jewish patients with Type 1 (insulin-dependent) diabetes mellitus diagnosed prior to 18 years of age: a population based study. Diabetologia. 1991; 34: 515–520.CrossrefMedlineGoogle Scholar
- 32 Gerber Y, Jacobsen SJ, Frye RL, Weston SA, Killian JM, Roger VL. Secular trends in deaths from cardiovascular diseases: a 25-year community study. Circulation. 2006; 113: 2285–2292.LinkGoogle Scholar
- 33 Pankratz VS, de Andrade M, Therneau TM. Random-effects Cox proportional hazards model: general variance components methods for time-to-event data. Genet Epidemiol. 2005; 28: 97–109.CrossrefMedlineGoogle Scholar
- 34 Hougaard P. Survival models for heterogeneous populations derived from stable distributions. Biometrika. 1986; 73: 387–396.CrossrefGoogle Scholar
- 35 Shu Y, Klein JP. A SAS macro for the positive stable frailty model. Available at: http://www.mcw.edu/biostatistics/Research/Software.htm. Accessed October 7, 2009.Google Scholar
- 36 Macros. Available at: http://mayoresearch.mayo.edu/mayo/research/biostat/upload/survcstd.sas. Accessed October 7, 2009.Google Scholar
- 37 Lang T, Ducimetiere P, Arveiler D, Amouyel P, Ferrieres J, Ruidavets JB, Montaye M, Haas B, Bingham A. Is hospital care involved in inequalities in coronary heart disease mortality? Results from the French WHO-MONICA Project in men aged 30–64. J Epidemiol Community Health. 1998; 52: 665–671.CrossrefMedlineGoogle Scholar
- 38 Barakat K, Stevenson S, Wilkinson P, Suliman A, Ranjadayalan K, Timmis AD. Socioeconomic differentials in recurrent ischaemia and mortality after acute myocardial infarction. Heart. 2001; 85: 390–394.CrossrefMedlineGoogle Scholar
- 39 Philbin EF, McCullough PA, DiSalvo TG, Dec GW, Jenkins PL, Weaver WD. Socioeconomic status is an important determinant of the use of invasive procedures after acute myocardial infarction in New York State. Circulation. 2000; 102 (suppl): III-107–III-115.LinkGoogle Scholar
- 40 Rathore SS, Berger AK, Weinfurt KP, Feinleib M, Oetgen WJ, Gersh BJ, Schulman KA. Race, sex, poverty, and the medical treatment of acute myocardial infarction in the elderly. Circulation. 2000; 102: 642–648.CrossrefMedlineGoogle Scholar
- 41 Soobader M, LeClere FB, Hadden W, Maury B. Using aggregate geographic data to proxy individual socioeconomic status: does size matter? Am J Public Health. 2001; 91: 632–636.CrossrefMedlineGoogle Scholar
- 42 Winkleby M, Sundquist K, Cubbin C. Inequities in CHD incidence and case fatality by neighborhood deprivation. Am J Prev Med. 2007; 32: 97–106.CrossrefMedlineGoogle Scholar
- 43 Geronimus AT, Bound J. Use of census-based aggregate variables to proxy for socioeconomic group: evidence from national samples. Am J Epidemiol. 1998; 148: 475–486.CrossrefMedlineGoogle Scholar
- 44 Steenland K, Henley J, Calle E, Thun M. Individual- and area-level socioeconomic status variables as predictors of mortality in a cohort of 179 383 persons. Am J Epidemiol. 2004; 159: 1047–1056.CrossrefMedlineGoogle Scholar
- 45 Hill AB. The environment and disease: association or causation? Proc R Soc Med. 1965; 58: 295–300.MedlineGoogle Scholar
- 46 Case RB, Moss AJ, Case N, McDermott M, Eberly S. Living alone after myocardial infarction: impact on prognosis. JAMA. 1992; 267: 515–519.CrossrefMedlineGoogle Scholar
- 47 Diez Roux AV, Kershaw K, Lisabeth L. Neighborhoods and cardiovascular risk: beyond individual-level risk factors. Curr Cardiovasc Risk Rep. 2008; 2: 175–180.CrossrefGoogle Scholar
- 48 Roger VL, Killian JM, Weston SA, Jaffe AS, Kors J, Santrach PJ, Tunstall-Pedoe H, Jacobsen SJ. Redefinition of myocardial infarction: prospective evaluation in the community. Circulation. 2006; 114: 790–797.LinkGoogle Scholar
- 49 Gerber Y, Rihal CS, Sundt TM III, Killian JM, Weston SA, Therneau TM, Roger VL. Coronary revascularization in the community: a population-based study, 1990 to 2004. J Am Coll Cardiol. 2007; 50: 1223–1229.CrossrefMedlineGoogle Scholar
- 50 Oldridge N, Furlong W, Perkins A, Feeny D, Torrance GW. Community or patient preferences for cost-effectiveness of cardiac rehabilitation: does it matter? Eur J Cardiovasc Prev Rehabil. 2008; 15: 608–615.CrossrefMedlineGoogle Scholar
Neighborhood of residence has been suggested to affect cardiovascular risk over and above personal socioeconomic status (SES). However, such data are currently lacking for patients with myocardial infarction (MI). We examined long-term survival according to neighborhood SES in a cohort of myocardial infarction patients. Consecutive patients (n=1179) ≤65 years of age discharged from 8 hospitals in central Israel after incident (ie, first ever) myocardial infarction in 1992 to 1993 were followed up for a median of 13 years. Personal socioeconomic and clinical data were obtained at study entry. Neighborhood SES was estimated through a composite census-derived index developed and validated by the Israel Central Bureau of Statistics. Neighborhood SES was strongly related to survival in a dose-response manner; patients residing in disadvantaged neighborhoods had higher mortality rates. Adjustment for personal SES measures and established prognostic factors attenuated the association, yet excess risk persisted for living in a disadvantaged neighborhood. The association was even stronger with regard to cardiac mortality. In the final fully adjusted models, several personal SES measures, including education, employment, and living with a steady partner, as well as neighborhood socioeconomic context, were concurrently associated with all-cause and cardiac mortality. Thus, our data support a multidimensional relationship between SES and myocardial infarction survival and indicate the importance of enhancing preventive/rehabilitative healthcare services for post–myocardial infarction patients residing in deprived neighborhoods. Further research is needed to identify the most relevant features of residential environments that affect cardiovascular incidence and prognosis.
See the Acknowledgments for a complete list of participating medical centers and investigators. The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.882555/DC1.
See the Acknowledgments for a complete list of participating medical centers and investigators.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.109.882555/DC1.