- Email: [email protected]
The association between socioeconomic status and subclinical atherosclerosis in a rural Bangladesh population
Preventive Medicine 102 (2017) 6–11
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
The association between socioeconomic status and subclinical atherosclerosis in a rural Bangladesh population Michael Garshick a, Fen Wu b, Ryan Demmer c, Faruque Parvez d, Alauddin Ahmed e, Mahbub Eunus e, Rabiul Hasan e, Jabun Nahar e, Ishrat Shaheen e, Golam Sarwar e, Moise Desvarieux c, Habibul Ahsan f,⁎⁎, Yu Chen b,⁎ a
Leon H. Charney Division of Cardiology, New York University Langone Medical Center, New York City, NY, USA Department of Population Health and Environmental Health, New York University School of Medicine, New York, NY, USA c Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA d Department of Environrmental Health Sciences, Mailman School of Public Health, Columbia University, New York, NY, USA e U-Chicago Research Bangladesh, Ltd., Dhaka, Bangladesh f Department of Health Studies, Center for Cancer Epidemiology and Prevention, The University of Chicago, Chicago, IL, USA b
a r t i c l e
i n f o
Article history: Received 13 February 2017 Received in revised form 18 May 2017 Accepted 18 June 2017 Available online 21 June 2017 Keywords: Bangladesh South Asian Socioeconomic status Subclinical atherosclerosis Carotid IMT Cardiovascular disease
a b s t r a c t Background. In Bangladesh, CVD accounts for the majority of non-communicable mortality. The purpose of this study was to determine the role of socioeconomic status (SES) on subclinical atherosclerosis measured as carotid intima-media thickness (IMT) in a rural Bangladesh population. Methods. Carotid IMT was measured between 2010 and 2011 in 1022 participants (average age 46, 40% male) randomly selected from the Health Effects of Arsenic Longitudinal Study (HEALS), a population-based prospective cohort study based in rural Bangladesh. SES was measured as occupation type, land ownership, educational attainment, and television ownership. Results. Half of the participants received formal education (53%) and under half owned land (48%) and a television (44%). Women were primarily homemakers (95%) and men worked as factory workers (24%), laborers (18%), or in business (55%). In univariate analysis, those owning greater than one acre of land (p = 0.03), owning a television (p = 0.02), or laborers and business owners compared to factory workers had higher levels of carotid IMT (p b 0.01). In multivariate analysis after adjustment for confounders, only men employed in the business sector had elevated carotid IMT compared to factory workers. The association was strongest in older men (58.7 μm, 95% CI 17.2–100.0, ≥50 years old) compared to younger men (13.7 μm, 95% CI −7.8–35.2, b50 years old). Conclusion. Business sector employment was positively associated with subclinical atherosclerosis after adjustment for confounders. This finding is consistent with evidence from other developing nations suggesting that certain SES factors are independent predictors of CVD. © 2016 Elsevier Inc. All rights reserved.
1. Introduction
Abbreviations: CVD, Cardiovascular disease; SES, Socioeconomic status; Carotid IMT, Carotid intima-media thickness; HEALS, Health Effects of Arsenic Longitudinal Study; DMII, Diabetes mellitus type II; BMI, Body mass index; HTN, Hypertenstion. ⁎ Correspondence to: Y. Chen, Department of Population Health and Environmental Health, New York University School of Medicine, 650 First Avenue, Room 510, New York, NY 10016, USA. ** Corresponding author. E-mail addresses: [email protected] (M. Garshick), [email protected] (F. Wu), [email protected] (R. Demmer), [email protected] (F. Parvez), [email protected] (A. Ahmed), [email protected] (M. Eunus), [email protected] (J. Nahar), [email protected] (G. Sarwar), [email protected] (M. Desvarieux), [email protected] (H. Ahsan), [email protected] (Y. Chen).
http://dx.doi.org/10.1016/j.ypmed.2017.06.022 0091-7435/© 2016 Elsevier Inc. All rights reserved.
Cardiovascular disease (CVD) mortality in South Asian countries such as Bangladesh has risen 30 fold over the past 30 years (Ahsan Karar et al., 2009). CVD is now one of the leading causes of mortality (El-Saharty et al., 2013). The onset of CVD within the South Asian population occurs 5–10 year earlier in life with a 5–10 times higher rate of CVD death prior to the age of 40, compared with other populations (Enas and S., 2001). The mechanisms for this rise in CVD mortality are multifactorial. With regards to lifestyle changes, the increased rates of obesity, metabolic syndrome and change in dietary patterns all associated with progressive industrialization and urbanization has led to worsening cardiometabolic profiles and increased CVD death. As South Asian populations such as the Bangladeshi population modernize, the contribution of socioeconomic status (SES) factors
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
leading to increased CVD requires more careful evaluation. In industrialized nations, CVD mortality is inversely related to SES (Kaplan and Keil, 1993). This relationship in low-income nations however is less clear and studies on the contribution of SES to CVD have shown mixed results. Furthermore, in low income countries, markers of SES may be different than in higher income ones (Vathesatogkit et al., 2014). Typically, CVD takes decades to develop and presents sub-clinically prior to a cardiovascular event. The studying of subclinical atherosclerosis is one potential way to estimate the contribution of various risk factors towards the development of CVD. Carotid intima-media thickness (IMT) is a surrogate measure of atherosclerosis which predicts the risk of subsequent CVD events (Thurston et al., 2014). In developing countries with limited resources such as Bangladesh, it is important to understand the demographics, risk factors and the development of CVD in order to focus preventive measures. Several studies in developed nations have explored the link between SES and carotid IMT and found that low education, low income, or manual occupation were associated with a thicker carotid arterial wall (Diez-Roux et al., 1995; Ebrahim et al., 1999; Lamont et al., 2000; Lynch et al., 1995; Nash et al., 2011; Rosvall et al., 2000) or with faster progression of carotid wall thickness (Lynch et al., 1997). These studies however, do not necessarily translate to underdeveloped or low-income countries. In addition, the contribution of SES within South Asian populations to subclinical atherosclerosis has not yet been described. The purpose of our study was to evaluate the association between SES and carotid IMT within the rural Bangladesh population. 2. Methods 2.1. Study population The Health Effects of Arsenic Longitudinal Study (HEALS) is an ongoing population-based prospective cohort study in Araihazar, Bangladesh, a rural subdistrict of Dhaka (McClintock et al., 2014a). Briefly, between October 2000 and May of 2002 a total of 11,746 married (to reduce loss to follow-up) men and women were recruited between the ages of 18 to 75 years from a well-defined 25-km2 geographical area where they had resided for at least 5 years. Between 2006 and 2008 the HEALS was expanded to include an additional 8287 participants using the same methodology. Study participants underwent demographic and lifestyle data collection using standardized questionnaires. A detailed description of the study protocol has been previously published and described (Ahsan et al., 2006). The overall study participation rate was 97%. Out of the original and expanded cohort members, 800 were randomly selected from the original cohort and 700 participants from the expansion cohort (total n = 1500) to undergo carotid IMT measurements (Chen et al., 2013a). Two hundred and ninety four participants originally selected were unable to complete carotid IMT measurements due to death, move, serious illness (20%) or time constraints (80%). From previously published data on analysis of carotid IMT within this group, the distribution of demographic and lifestyle variables in the study population and in the overall cohort were similar (McClintock et al., 2014a). In total, IMT was measured for 1206 individuals consisting of 600 from the original cohort and 606 from the expansion cohort (McClintock et al., 2014). Informed consent was obtained from study participants and the procedures were approved by the Ethical Committee of the Bangladesh Medical Research Council and the Institutional Review Boards of Columbia University and the University of Chicago. 2.2. Anthropometric and socioeconomic assessment Baseline in-person interviews were performed by trained personnel with detailed questionnaires on lifestyle characteristics. Participants were asked to provide information on demographics, medical co-morbidities (including self-reporting of diabetes mellitus type II, DM II),
7
cigarette smoking, and betel-nut usage. Questions were also answered on SES as defined by television ownership, land ownership, years of education and occupation. In the original dataset occupation status was recorded as daily laborers or farmers, factory workers, business workers or other similar jobs, homemaker, and unemployed for a total of 5 categories of occupation status. The specific occupations of individuals reporting ‘business workers’ or ‘other similar jobs’ were heterogeneous but primarily included merchants, store or factory owners or contractors. Anthropometric measurements including height, weight and blood pressure were performed using standard techniques (Chen et al., 2007; Pierce et al., 2010; Chen et al., 2006). 2.3. Measurement of carotid IMT Carotid IMT measurements were performed between April 2010 and September of 2011 (on average 7.2 years after baseline demographic information was obtained). Detailed methodology for carotid IMT measurements used in this study have been described elsewhere (McClintock et al., 2014a). Briefly, all measurements were obtained by 1 physician trained in sonography on a SonoSite MicroMaxx ultrasound machine (SonoSite, Bothell, WA) equipped with an L38e/10-5 MHz transducer. The protocol used was designed and implemented in the Oral Infections and Vascular Disease Epidemiology Study (INVEST) (Desvarieux et al., 2005). The carotid arteries were scanned longitudinally in 3 segments using the lateral extent of each carotid segment as defined relative to the tip of the flow divider, which is typically the most well defined anatomical reference in the proximity of the carotid bifurcation. The mean of the near and far walls of the maximum common carotid artery IMT of both sides of the neck (mean of twelve measurements) was used as the main outcome variable which is similar to previous reported studies (O'Leary et al., 1999; Kitamura et al., 2004; Touboul et al., 2012). Carotid IMT values are reported in this paper in thousandths of a millimeter (typically reported in tenths of a millimeter) to remain consistent with previously published studies using this same cohort (Chen, et al., 2013a). 2.4. Statistical analysis Of the 1206 participants, we excluded those who were missing data on SES (n = 2), systolic blood pressure (SBP) (n = 172), body mass index (BMI) (n = 180), cigarette use (past and present) (n = 172), and betel nut use (n = 174). The final analysis included a total of 1022 participants. We first conducted descriptive analysis of population characteristics and socioeconomic status across interquartile carotid IMT. Occupation status was divided into 5 categories as described above. Land ownership was modeled as a yes/no variable as well as split into no land, less than or equal to 1 acre and N1 acre. Education status was similarly broken down into yes/no and split into no education, 1 to 5 years of schooling or 6 or greater years, according to the Bangladeshi education system. Interquartile differences were analyzed using a one-way analysis of variance (ANOVA) for continuous variables and chi-squared for categorical variables. Then, we conducted linear regression models with carotid IMT modeled as a continuous dependent variable and each SES attribute modeled as an independent variable. Assumptions of linear regression were checked and none was violated. Next carotid IMT was modeled using multivariable analysis adjusting firstly for age and sex (model 1), and then BMI, SBP, smoking (never, past and present smokers), and betel nut use (never, past and present users) additionally (model 2). These factors are known independent predictors of carotid IMT. The comparison of the two models can help assess to what extent the association between SES and cIMT can be explained by other variables, including potential mediators. Multivariable regression analysis was performed for the overall cohort and stratified for men and women separately. Stratified analysis was also conducted by age within men in an attempt to further assess whether the association between occupation and IMT differed by age. Because in our cohort
8
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
of the initially only statistically significant associations seen between SES and men, we did not perform stratified analysis in women. We also tested interaction between each SES variable with gender and with age in men by entering a cross-product term in the models. Adjustment of the results for DMII was not significantly different and thus was not included in the final multivariate model. All statistical analysis was performed using STATA 11 (Statacorp. College Station, TX). A 2-sided alpha error of b0.05 indicated statistical significance.
were associated with higher levels of carotid IMT, compared with the occupation of factory workers. Homemakers were associated with a low level of carotid IMT. In age- and sex-adjusted models (model 1), the association between occupation and carotid IMT remained similar, but the association between land ownership and carotid IMT disappeared. When compared to those with no schooling, there was a significant relationship between a higher educational attainment (≥ 6 years of schooling, p = 0.01) and carotid IMT (model 1, Table 2). After multivariable adjustment for risk factors and potential confounders and mediators (BMI, SBP, cigarette and betel nut usage), the positive relationship observed between television (p = 0.58) or land ownership (p = 0.92) and carotid IMT were no longer present. The positive associations of business occupation and unemployment with carotid IMT remained significant. We further investigated the association between SES indicators and carotid IMT by sex. Among men, after multivariable adjustment, carotid IMT was higher among business workers when compared to factory workers (26.7 μm, 95% CI (6.7–46.8), p b 0.01) (Table 3). There was no significant association between occupation status and carotid IMT in women. We also explored whether the association between SES and carotid IMT differed by age in men (Table 4). The association was stronger in older men (≥ 50 years old; 58.7 μm, 95% CI (17.2–100.0)) compared to younger men (b 50 years old; 13.7 μm, 95% CI (−7.8–35.2)) with an interaction noted for the difference in age (p = 0.048).
3. Results Characteristics of the study population are listed in Table 1. The majority of the participants were b50 years old (65%), female (60%) and had a BMI b25 (80%). Over 1/3 of the participants were active smokers or had a history of smoking (37%); a small percentage of participants had diagnosed hypertension (13%) and 2% were diagnosed with DM II (2%). The majority of participants had at least one or more year of schooling (53%) and under half owned televisions (44%) and land (48%). Women were primarily homemakers (95%) and men worked as factory workers (24%), laborers (18%) or in business (55%). The proportion of men (p b 0.01), older individuals (p b 0.01) and those with elevated blood pressure (b 0.01) increased with increasing quartiles of carotid IMT. Similarly the percentages of those who were ever smokers (p b 0.01), had DMII (P = 0.02) or used betel nut (p b 0.01) increased as well across interquartile carotid IMT. BMI was not associated with increased interquartile carotid IMT (p = 0.77). Univariate analyses of SES variables with demographic and lifestyle variables were shown in Supplemental Table 1. Table 2 shows the association between SES status indicators and carotid IMT in the overall study population. In univariate analysis, owning a television (13.8 μm, p = 0.02) and owning greater than one acre of land (20.5 μm, p = 0.03) were positively associated with higher levels of carotid IMT. Daily laborer, business workers, and unemployment
4. Discussion As each developing country is unique in regards to the contribution of SES to CVD, our study in this Bangladesh cohort is one of the first to evaluate and identify the SES factors associated with subclinical atherosclerosis with the use of carotid IMT. In this study of rural Bangladesh subjects, there were significant univariate associations between carotid IMT and SES factors such as types of employment, land ownership, and
Table 1 Distribution of population characteristics by carotid IMT quartiles. Characteristics
Overall
Quartile 1 (606.4–738.3 μm)
[n = 1022] Carotid IMT (μm) Male sex (%) Age at IMT measurement (years) Body mass index (kg/m2) Cigarette use (%)b Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes (%) Betel nut use (%)c Television ownership (%) Education (years) No education (%) 1–5 years of schooling (%) ≥6 years of schooling (%) Land ownership (%) No land (%) ≤1 acre (%) N1 acre (%) Occupation Factory worker (%) Daily laborer (%)d Business (%) Homemaker (%) Unemployed (%)
807.5 40 46.3 21.1 37 122.4 75.4 2 55 44 3.1 47 31 21 49 52 34 14 10 7 24 57 2
9 1 12 77 0.4
p-Valuea
Quartile 2 (738.8–788.2 μm)
Quartile 3 (788.3–860.8 μm)
Quartile 4 (860.1–1423.3 μm)
[n = 257]
[n = 254]
[n = 257]
[n = 254]
703.8 19 40 21.3 18 118 73.7 1 40 35 3.1 45 32 23 42 60 29 11
764.2 35 44.2 20.9 30 118.6 74.1 0.4 54 44 2.9 51 30 20 51 51 37 12
822.1 44 48.3 21.1 42 122.8 75.3 3 58 51 3.1 46 35 19 52 49 33 18
940.8 63 52.6 21.1 59 129.9 78.4 4 67 46 3.3 48 28 24 52 49 35 16
b0.01 b0.01 b0.01 0.77 b0.01 b0.01 b0.01 0.02 b0.01 b0.01 0.70 0.61 0.39 0.51 0.07 0.06 0.27 0.09
13 5 19 63 0
10 10 28 53 0
9 14 37 25 4
0.24 b0.01 b0.01 b0.01 b0.01
Carotid IMT and demographic data collected from Araihazar Bangladesh between 2010 and 2011. Carotid IMT; Carotid intima-media thickness. a Chi-squared test was used for categorical variables and ANOVA for continuous variables. b Cigarette use includes past and current usage. c Betel nut use includes past and current usage. d Includes daily laborers andfarmers. 28 participants excluded from categorical analysis of land due to unknown amount of land ownership.
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
9
Table 2 Association between socioeconomic status and carotid IMT. Socioeconomic status
N
TV ownership No Yes Occupation Factory worker Daily laborerc Business Homemaker Unemployed Land ownershipb No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
1022 574 448 1022 104 75 246 585 12 994 520 334 140 1022 485 318 219
Multivariable model 1a
Univariable model
Multivariable model 2d
cIMT (μm, CI)
p-Value
cIMT (μm, CI)
p-Value
cIMT (μm, CI)
p-Value
Reference 13.8 (1.9 to 25.7)
0.02
Reference 6.5 (−3.1 to 16.2)
0.18
Reference −2.4 (−12.4 to 6.9)
0.58
Reference 64.4 (37.5 to 91.3) 47.0 (26.5 to 68.1) −20.1 (−38.9 to −1.2) 115.2 (61.0 to 169.3)
b0.01 b0.01 0.04 b0.01
Reference 20.4 (−3.1 to 43.9) 38.1 (20.2 to 56.0) 30.4 (−2.9 to 62.9) 65.4 (18.4 to 112.5)
0.09 b0.01 0.07 b0.01
Reference 22.7 (−0.02 to 45.4) 30.2 (12.8 to 47.6) 29.5 (−1.8 to 60.9) 65.9 (20.5 to 111.2)
0.05 b0.01 0.07 b0.01
Reference 9.7 (−3.6 to 23.0) 20.5 (2.5 to 38.6)
0.15 0.03
Reference 2.5 (−8.2 to 13.3) 6.7 (−7.9 to 21.5)
0.64 0.37
Reference 2.0 (−8.5 to 12.6) −0.8 (−15.2 to 13.7)
0.71 0.92
Reference −2.6 (−16.3 to 11.1) 7.3 (−8.1 to 22.8)
0.71 0.35
Reference 4.1 (−7 to 15.3) 16.7 (3.9 to 29.4)
0.47 0.01
Reference −0.7 (−11.7 to 102.5) 8.6 (−4.3 to 21.5)
0.90 0.19
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariable model 1 adjusted for age and sex. b Multivariable model 2 adjusted for age, sex, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). c Includes daily laborers and farmers. d 28 participants had an unknown amount of land ownership (26 women, 2 men).
TV ownership (Table 2). The associations of land and TV ownership with carotid IMT levels were largely explained by differences in other predictors of subclinical atherosclerosis including age, sex, BMI, systolic blood pressure, cigarette use and betel nut use. After adjustment of these predictors and potential confounders of CVD risk, elevated carotid IMT among men who were in the business sector remained significant when compared to factory workers. Further stratified analysis by age revealed that the association between being employed in the business sector and IMT was stronger in men over the age of 50 compared to younger men (Table 4). There have been few studies evaluating SES and CVD in the South Asian and Bangladesh population. In a study using 2010–2012 surveillance data from a rural Bangladesh population, non-communicable disease deaths (which including CVD and CVD equivalents) were strongly associated with higher SES groups (Hanifi et al., 2014). When comparing SES status and CVD risk factors in the Bangladesh population, the results are similar. In a 2011 Bangladesh Demographic and Health Survey
study, after adjustment for potential confounders, elevated income and education status were related to HTN and hyperglycemia. (Harshfield et al., 2015). These results, when examined within the context of our study, help to refine in the rural population what predicts subclinical atherosclerosis. By having a job associated with higher overall wealth such as someone employed in the business sector, one is more likely to have subclinical atherosclerosis. Our study is important because CVD mortality remains the leading causing of death in the world, and while mortality is declining in developed countries, in developing ones it is rising (Rasiah et al., 2013). The contribution of SES to CVD depends on the country studied. In the US, as in most developed nations, higher SES is associated with a lower incidence of CVD (Kaplan and Keil, 1993). There have been several studies in developed countries showing associations between lower SES and subclinical atherosclerosis (Thurston et al., 2014; Nash et al., 2011; Lutsey et al., 2008). In the Multi-Ethnic Study of Atherosclerosis (MESA) lower SES and longer duration living in the United States after
Table 3 Association between socioeconomic status and carotid IMT by gender. Multivariable model - mena
Socioeconomic status N TV ownership No Yes Occupation Factory worker Daily laborerb Business Homemaker Unemployed Land ownershipc No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
408 225 183 408 99 73 226 . 10 406 200 148 58 408 161 140 107
cIMT (μm, CI)
Multivariable model - womena p-Value
Reference −3.0 (−20.0 to 14.0)
0.73
Reference 19.6 (−6.6 to 45.7) 26.7 (6.7 to 46.8) . 47.3 (−8.1 to 102.6)
0.14 b0.01 . 0.09
Reference 0.7 (−17.2 to 18.6) 6.5 (−18.8 to 31.7)
. 0.94 0.62
Reference −4.1 (−23.3 to 15.0) 6.8 (−14.4 to 27.9)
. 0.67 0.53
N 614 349 265 614 5 2 20 585 2 588 320 186 82 614 324 178 112
cIMT (μm, CI)
p-Value
Interaction p-value
Reference −2.4 (−13.9 to 9.0)
0.68
0.33
Reference −36.6 (−150.6 to 77.3) −6.4 (−74.6 to 61.8) 2.1 (−58.9 to 63.2) 92.9 (−21.6 to 207.3)
. 0.53 0.85 0.95 0.11
0.29 0.26 . 0.66
Reference 3.4 (−9.3 to 16.1) −6.7 (−23.9 to 10.5)
. 0.53 0.45
0.88 0.09
Reference −0.3 (−13.1 to 13.1) 6.8 (−9.4 to 23.1)
. 0.95 0.41
0.92 0.59
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariate model adjusted for age, sex, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). b Includes daily laborers and farmers. c 28 participants had an unknown amount of land ownership (2 men, 26 women).
10
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
Table 4 Association between socioeconomic status and carotid IMT in men stratified by age. Men ≥ 50 years of agea
Socioeconomic status N TV ownership No Yes Occupation Factory worker Daily laborerb Business Unemployed Land ownership No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
174 91 83 174 27 47 92 8 174 74 67 33 174 72 54 48
cIMT (μm, CI)
Men b 50 years of agea p-Value
N
cIMT (μm, CI)
p-Value
Interaction p-value
Reference −0.6 (−30.3 to 29.1)
0.97
234 134 100
Reference −4.6(−24.8 to 15.7)
0.66
0.78
Reference 30.6 (−15.4 to 76.7) 58.7 (17.2 to 100.0) 92.8 (14.3 to 171.3)
0.19 b0.01 0.02
Reference 26.4 (−6.6 to 59.4) 13.7 (−7.8 to 35.2) −15.8 (−120.4 to 88.8)
0.12 0.21 0.77
0.91 0.04 0.13
Reference 4.1 (−28.8 to 36.9) 22.1 (−19.1 to 63.3)
0.81 0.29
Reference −1.7 (−22.2 to 18.9) −8.9 (−41.1to 23.4)
0.87 0.59
0.82 0.30
Reference −8.1 (−43.4 to 27.2) 24.7 (−13.3 to 61.9)
0.65 0.20
Reference −5.4 (−27.4 to 16.6) −7 (−31.9 to 17.9)
0.63 0.58
0.83 0.20
72 26 134 2 232c 126 81 25 234 89 86 59
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariate model adjusted for age, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). b Includes daily laborers and farmers c 2 participants had an unknown amount of land ownership.
immigration from developing countries were associated with increased carotid IMT (Lutsey et al., 2008). This is in contrast to CVD in developing nations, where the contribution of SES is variable (Reddy and Yusuf, 1998). A large meta-analysis evaluating 29 Asian countries showed inverse relationships between education, income, occupation status and CVD mortality similar to developed countries (Vathesatogkit et al., 2014). Other studies have noted that as developing nation's progress to middle income status and increased wealth with progressive industrialization, the resulting urbanization of the population causes CVD to rise (Yusuf et al., 2001; Clark et al., 2009). The findings from our study indicate that the contribution of SES to subclinical CVD in low-income populations differs and the association depends on the type of SES indicator studied. Other CVD risk factors can explain the relationship between SES markers such as educational status, TV, land ownership with carotid IMT. In our analyses, the positive association between educational attainment and carotid IMT can be explained by differences in blood pressure, cigarette use, betel nut use, and BMI. However, the association between business employment and IMT remained after additional adjustments. There may be other unmeasured specific factors related to SES such as diet, health education, and access to health care that may mediate the association between SES and cIMT. Future studies however, would be needed to assess these hypotheses and identify the mediators. Future studies could also use our described associations to focus on the specific SES indicators to measure in rural populations of other developing nations. The biological plausibility behind the contribution of SES within the context of lifestyle changes and subclinical atherosclerosis in Bangladesh is likely multifactorial. The South Asian population is noted to have accelerated rates of premature coronary artery disease when compared to western counterparts (Islam and Majumder, 2013). Previous epidemiologic studies have estimated the prevalence of atherosclerotic heart disease up to 3.4% in the rural population and 19.6% in urban populations (Islam and Majumder, 2013). These studies however have been small, and it is likely that the prevalence and overall mortality from cardiovascular disease within the Bangladesh population is rising (Ahsan Karar et al., 2009). Additionally, metabolic syndrome, diabetes and obesity are all rising (Bhowmik et al., 2015). Contributing to this rise in cardiovascular disease, in a 2010 Bangladesh survey, 27% of subjects were found to have low levels of activity and when compared to western guidelines of what constitutes adequate exercise, 87% of Bangladeshis
did not meet them (Islam and Majumder, 2013; Hayes et al., 2002). The traditional Bangladesh diet, already poor in fruits and vegetables and high in grains is undergoing a further transition with the rise of fast food culture and western style foods high in saturated fats (Islam and Majumder, 2013). Given these lifestyle related changes it is not surprising that in our study, individuals engaged in business sector were more prone to subclinical atherosclerosis. Our paper has several limitations. This is a post-hoc analysis of an existing prospective cohort initially designed to study arsenic exposure, well water and CVD. However, since the HEALS is a population-based cohort study, the resources allow the evaluation of other risk factors for CVD. We have also previously shown significant associations with specific dietary patterns (Chen et al., 2006; Jiang et al., 2014; Chen et al., 2013b; McClintock et al., 2014b), cigarette smoking (Ahn et al., 2013), betel nut chewing (McClintock et al., 2014a; Heck et al., 2012), and anthropometric measures (Pierce et al., 2010; Chen et al., 2014; Ge et al., 2014) with CVD. Most of the women in the study were homemakers, and thus we cannot apply these findings to women who work in the business sector. Additionally, because this cohort was designed in a rural population, the applicability to urbanized areas is limited. Based on limited previous literature, we made assumptions about what constitutes SES and thus we may have missed some markers of SES that more accurately reflected our cohort characteristics. Our definition of businessmen is heterogeneous, including mostly individuals owning small grocery stores, clothing stores, small restaurants, and food-related business in the area, thus it may be hard to pinpoint one specific business sector to focus CVD prevention efforts on. Finally we used carotid IMT as a surrogate endpoint for overall cardiovascular disease, and elevated carotid IMT may not translate to the more important endpoint of clinically significant CVD. Regardless of these limitations, this analysis of SES and subclinical atherosclerosis within a Bangladesh cohort is novel, and deserves attention given the rising burden of CVD and the need for more information of what drives this CVD risk. Overall, our study found that in a rural Bangladesh population, elevated SES as defined by businessmen had elevated risk of subclinical atherosclerosis, and the association was stronger in men older than the age of 50. Education status, TV and amount of land ownership showed positive trends with carotid IMT, but the association can largely be explained by differences in other risk factors. These findings are novel and represent new information on one association of potential
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
drivers of CVD within Bangladesh. Given the rising CVD burden, it will be necessary to focus preventive cardiovascular services on specific subpopulations such as those employed in small business. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ypmed.2017.06.022. Conflict of Interst The authors report that there are no potential conflicts of interst. Funding This study was supported by grants from the National Institutes of Health (R01 ES017541, P42 ES010349, and P30 ES000260). References Ahn, J., et al., 2013. Human gut microbiome and risk for colorectal cancer. J. Natl. Cancer Inst. 105 (24), 1907–1911. Ahsan Karar, Z., Alam, N., Kim Streatfield, P., 2009. Epidemiological transition in rural Bangladesh, 1986–2006. Glob. Health Action 2. Ahsan, H., et al., 2006. Health Effects of Arsenic Longitudinal Study (HEALS): description of a multidisciplinary epidemiologic investigation. J. Expo. Sci. Environ. Epidemiol. 16 (2), 191–205. Bhowmik, B., et al., 2015. Comparison of the prevalence of metabolic syndrome and its association with diabetes and cardiovascular disease in the rural population of Bangladesh using the modified National Cholesterol Education Program Expert Panel Adult Treatment Panel III and International Diabetes Federation definitions. J Diabetes Investig 6 (3), 280–288. Chen, Y., et al., 2006. Nutritional influence on risk of high blood pressure in Bangladesh: a population-based cross-sectional study. Am. J. Clin. Nutr. 84 (5), 1224–1232. Chen, Y., et al., 2007. Arsenic exposure from drinking water, dietary intakes of B vitamins and folate, and risk of high blood pressure in Bangladesh: a population-based, crosssectional study. Am. J. Epidemiol. 165 (5), 541–552. Chen, Y., et al., 2013a. Arsenic exposure from drinking water, arsenic methylation capacity, and carotid intima-media thickness in Bangladesh. Am. J. Epidemiol. 178 (3), 372–381. Chen, Y., et al., 2013b. Prospective investigation of major dietary patterns and risk of cardiovascular mortality in Bangladesh. Int. J. Cardiol. 167 (4), 1495–1501. Chen, Y., et al., 2014. A prospective study of arm circumference and risk of death in Bangladesh. Int. J. Epidemiol. 43 (4), 1187–1196. Clark, A.M., et al., 2009. Socioeconomic status and cardiovascular disease: risks and implications for care. Nat. Rev. Cardiol. 6 (11), 712–722. Desvarieux, M., et al., 2005. Periodontal microbiota and carotid intima-media thickness: the Oral Infections and Vascular Disease Epidemiology Study (INVEST). Circulation 111 (5), 576–582. Diez-Roux, A.V., et al., 1995. Social inequalities and atherosclerosis. The atherosclerosis risk in communities study. Am. J. Epidemiol. 141 (10), 960–972. Ebrahim, S., et al., 1999. Carotid plaque, intima media thickness, cardiovascular risk factors, and prevalent cardiovascular disease in men and women: the British Regional Heart Study. Stroke 30 (4), 841–850. El-Saharty, S.A., Zunaid, Karar, Koehlmoos, Tracey L.P., Engelgau, Michael M., 2013. Tackling noncommunicable diseases in Bangladesh: now is the time. Health, Nutrition & Population (SASHN). Enas, E., S., A., 2001. Coronary artery disease in Asian Indians: an update and review. International Journal of Cardiology 1 (2). Ge, W., et al., 2014. Association between anthropometric measures of obesity and subclinical atherosclerosis in Bangladesh. Atherosclerosis 232 (1), 234–241.
11
Hanifi, S.M., Mahmood, S.S., Bhuiya, A., 2014. Cause-specific mortality and socioeconomic status in Chakaria, Bangladesh. Glob. Health Action 7, 25473. Harshfield, E., et al., 2015. Association of hypertension and hyperglycaemia with socioeconomic contexts in resource-poor settings: the Bangladesh Demographic and Health Survey. Int. J. Epidemiol. 44 (5), 1625–1636. Hayes, L., et al., 2002. Patterns of physical activity and relationship with risk markers for cardiovascular disease and diabetes in Indian, Pakistani, Bangladeshi and European adults in a UK population. J. Public Health Med. 24 (3), 170–178. Heck, J.E., et al., 2012. Betel quid chewing in rural Bangladesh: prevalence, predictors and relationship to blood pressure. Int. J. Epidemiol. 41 (2), 462–471. Islam, A.K., Majumder, A.A., 2013. Coronary artery disease in Bangladesh: a review. Indian Heart J. 65 (4), 424–435. Jiang, J., et al., 2014. Association of major dietary patterns and blood pressure longitudinal change in Bangladesh. J. Hypertens. (in press). Kaplan, G.A., Keil, J.E., 1993. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation 88 (4 Pt 1), 1973–1998. Kitamura, A., et al., 2004. Carotid intima-media thickness and plaque characteristics as a risk factor for stroke in Japanese elderly men. Stroke 35 (12), 2788–2794. Lamont, D., et al., 2000. Risk of cardiovascular disease measured by carotid intima-media thickness at age 49-51: lifecourse study. BMJ 320 (7230), 273–278. Lutsey, P.L., et al., 2008. Associations of acculturation and socioeconomic status with subclinical cardiovascular disease in the multi-ethnic study of atherosclerosis. Am. J. Public Health 98 (11), 1963–1970. Lynch, J., et al., 1995. Socioeconomic status and carotid atherosclerosis. Circulation 92 (7), 1786–1792. Lynch, J., et al., 1997. Socioeconomic status and progression of carotid atherosclerosis. Prospective evidence from the Kuopio Ischemic Heart Disease Risk Factor Study. Arterioscler. Thromb. Vasc. Biol. 17 (3), 513–519. McClintock, T.R., et al., 2014a. Association between betel quid chewing and carotid intima-media thickness in rural Bangladesh. Int. J. Epidemiol. 43 (4), 1174–1182. McClintock, T.R., et al., 2014b. Major dietary patterns and carotid intima-media thickness in Bangladesh: the HEAL Study. Public Health Nutr. (in press). Nash, S.D., et al., 2011. Socioeconomic status and subclinical atherosclerosis in older adults. Prev. Med. 52 (3–4), 208–212. O'Leary, D.H., et al., 1999. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N. Engl. J. Med. 340 (1), 14–22. Pierce, B.L., et al., 2010. A prospective study of body mass index and mortality in Bangladesh. Int. J. Epidemiol. 39 (4), 1037–1045. Rasiah, R., et al., 2013. Cardiovascular disease risk factors and socioeconomic variables in a nation undergoing epidemiologic transition. BMC Public Health 13 (p. 886). Reddy, K.S., Yusuf, S., 1998. Emerging epidemic of cardiovascular disease in developing countries. Circulation 97 (6), 596–601. Rosvall, M., et al., 2000. Occupational status, educational level, and the prevalence of carotid atherosclerosis in a general population sample of middle-aged Swedish men and women: results from the Malmo Diet and Cancer Study. Am. J. Epidemiol. 152 (4), 334–346. Thurston, R.C., et al., 2014. Low socioeconomic status over 12 years and subclinical cardiovascular disease: the study of women's health across the nation. Stroke 45 (4), 954–960. Touboul, P.J., et al., 2012. Mannheim carotid intima-media thickness and plaque consensus (2004–2006–2011). An update on behalf of the advisory board of the 3rd, 4th and 5th watching the risk symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004, Brussels, Belgium, 2006, and Hamburg, Germany, 2011. Cerebrovasc. Dis. 34 (4), 290–296. Vathesatogkit, P., Batty, G.D., Woodward, M., 2014. Socioeconomic disadvantage and disease-specific mortality in Asia: systematic review with meta-analysis of populationbased cohort studies. J. Epidemiol. Community Health 68 (4), 375–383. Yusuf, S., et al., 2001. Global burden of cardiovascular diseases: part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization. Circulation 104 (22), 2746–2753.
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
The association between socioeconomic status and subclinical atherosclerosis in a rural Bangladesh population Michael Garshick a, Fen Wu b, Ryan Demmer c, Faruque Parvez d, Alauddin Ahmed e, Mahbub Eunus e, Rabiul Hasan e, Jabun Nahar e, Ishrat Shaheen e, Golam Sarwar e, Moise Desvarieux c, Habibul Ahsan f,⁎⁎, Yu Chen b,⁎ a
Leon H. Charney Division of Cardiology, New York University Langone Medical Center, New York City, NY, USA Department of Population Health and Environmental Health, New York University School of Medicine, New York, NY, USA c Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA d Department of Environrmental Health Sciences, Mailman School of Public Health, Columbia University, New York, NY, USA e U-Chicago Research Bangladesh, Ltd., Dhaka, Bangladesh f Department of Health Studies, Center for Cancer Epidemiology and Prevention, The University of Chicago, Chicago, IL, USA b
a r t i c l e
i n f o
Article history: Received 13 February 2017 Received in revised form 18 May 2017 Accepted 18 June 2017 Available online 21 June 2017 Keywords: Bangladesh South Asian Socioeconomic status Subclinical atherosclerosis Carotid IMT Cardiovascular disease
a b s t r a c t Background. In Bangladesh, CVD accounts for the majority of non-communicable mortality. The purpose of this study was to determine the role of socioeconomic status (SES) on subclinical atherosclerosis measured as carotid intima-media thickness (IMT) in a rural Bangladesh population. Methods. Carotid IMT was measured between 2010 and 2011 in 1022 participants (average age 46, 40% male) randomly selected from the Health Effects of Arsenic Longitudinal Study (HEALS), a population-based prospective cohort study based in rural Bangladesh. SES was measured as occupation type, land ownership, educational attainment, and television ownership. Results. Half of the participants received formal education (53%) and under half owned land (48%) and a television (44%). Women were primarily homemakers (95%) and men worked as factory workers (24%), laborers (18%), or in business (55%). In univariate analysis, those owning greater than one acre of land (p = 0.03), owning a television (p = 0.02), or laborers and business owners compared to factory workers had higher levels of carotid IMT (p b 0.01). In multivariate analysis after adjustment for confounders, only men employed in the business sector had elevated carotid IMT compared to factory workers. The association was strongest in older men (58.7 μm, 95% CI 17.2–100.0, ≥50 years old) compared to younger men (13.7 μm, 95% CI −7.8–35.2, b50 years old). Conclusion. Business sector employment was positively associated with subclinical atherosclerosis after adjustment for confounders. This finding is consistent with evidence from other developing nations suggesting that certain SES factors are independent predictors of CVD. © 2016 Elsevier Inc. All rights reserved.
1. Introduction
Abbreviations: CVD, Cardiovascular disease; SES, Socioeconomic status; Carotid IMT, Carotid intima-media thickness; HEALS, Health Effects of Arsenic Longitudinal Study; DMII, Diabetes mellitus type II; BMI, Body mass index; HTN, Hypertenstion. ⁎ Correspondence to: Y. Chen, Department of Population Health and Environmental Health, New York University School of Medicine, 650 First Avenue, Room 510, New York, NY 10016, USA. ** Corresponding author. E-mail addresses: [email protected] (M. Garshick), [email protected] (F. Wu), [email protected] (R. Demmer), [email protected] (F. Parvez), [email protected] (A. Ahmed), [email protected] (M. Eunus), [email protected] (J. Nahar), [email protected] (G. Sarwar), [email protected] (M. Desvarieux), [email protected] (H. Ahsan), [email protected] (Y. Chen).
http://dx.doi.org/10.1016/j.ypmed.2017.06.022 0091-7435/© 2016 Elsevier Inc. All rights reserved.
Cardiovascular disease (CVD) mortality in South Asian countries such as Bangladesh has risen 30 fold over the past 30 years (Ahsan Karar et al., 2009). CVD is now one of the leading causes of mortality (El-Saharty et al., 2013). The onset of CVD within the South Asian population occurs 5–10 year earlier in life with a 5–10 times higher rate of CVD death prior to the age of 40, compared with other populations (Enas and S., 2001). The mechanisms for this rise in CVD mortality are multifactorial. With regards to lifestyle changes, the increased rates of obesity, metabolic syndrome and change in dietary patterns all associated with progressive industrialization and urbanization has led to worsening cardiometabolic profiles and increased CVD death. As South Asian populations such as the Bangladeshi population modernize, the contribution of socioeconomic status (SES) factors
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
leading to increased CVD requires more careful evaluation. In industrialized nations, CVD mortality is inversely related to SES (Kaplan and Keil, 1993). This relationship in low-income nations however is less clear and studies on the contribution of SES to CVD have shown mixed results. Furthermore, in low income countries, markers of SES may be different than in higher income ones (Vathesatogkit et al., 2014). Typically, CVD takes decades to develop and presents sub-clinically prior to a cardiovascular event. The studying of subclinical atherosclerosis is one potential way to estimate the contribution of various risk factors towards the development of CVD. Carotid intima-media thickness (IMT) is a surrogate measure of atherosclerosis which predicts the risk of subsequent CVD events (Thurston et al., 2014). In developing countries with limited resources such as Bangladesh, it is important to understand the demographics, risk factors and the development of CVD in order to focus preventive measures. Several studies in developed nations have explored the link between SES and carotid IMT and found that low education, low income, or manual occupation were associated with a thicker carotid arterial wall (Diez-Roux et al., 1995; Ebrahim et al., 1999; Lamont et al., 2000; Lynch et al., 1995; Nash et al., 2011; Rosvall et al., 2000) or with faster progression of carotid wall thickness (Lynch et al., 1997). These studies however, do not necessarily translate to underdeveloped or low-income countries. In addition, the contribution of SES within South Asian populations to subclinical atherosclerosis has not yet been described. The purpose of our study was to evaluate the association between SES and carotid IMT within the rural Bangladesh population. 2. Methods 2.1. Study population The Health Effects of Arsenic Longitudinal Study (HEALS) is an ongoing population-based prospective cohort study in Araihazar, Bangladesh, a rural subdistrict of Dhaka (McClintock et al., 2014a). Briefly, between October 2000 and May of 2002 a total of 11,746 married (to reduce loss to follow-up) men and women were recruited between the ages of 18 to 75 years from a well-defined 25-km2 geographical area where they had resided for at least 5 years. Between 2006 and 2008 the HEALS was expanded to include an additional 8287 participants using the same methodology. Study participants underwent demographic and lifestyle data collection using standardized questionnaires. A detailed description of the study protocol has been previously published and described (Ahsan et al., 2006). The overall study participation rate was 97%. Out of the original and expanded cohort members, 800 were randomly selected from the original cohort and 700 participants from the expansion cohort (total n = 1500) to undergo carotid IMT measurements (Chen et al., 2013a). Two hundred and ninety four participants originally selected were unable to complete carotid IMT measurements due to death, move, serious illness (20%) or time constraints (80%). From previously published data on analysis of carotid IMT within this group, the distribution of demographic and lifestyle variables in the study population and in the overall cohort were similar (McClintock et al., 2014a). In total, IMT was measured for 1206 individuals consisting of 600 from the original cohort and 606 from the expansion cohort (McClintock et al., 2014). Informed consent was obtained from study participants and the procedures were approved by the Ethical Committee of the Bangladesh Medical Research Council and the Institutional Review Boards of Columbia University and the University of Chicago. 2.2. Anthropometric and socioeconomic assessment Baseline in-person interviews were performed by trained personnel with detailed questionnaires on lifestyle characteristics. Participants were asked to provide information on demographics, medical co-morbidities (including self-reporting of diabetes mellitus type II, DM II),
7
cigarette smoking, and betel-nut usage. Questions were also answered on SES as defined by television ownership, land ownership, years of education and occupation. In the original dataset occupation status was recorded as daily laborers or farmers, factory workers, business workers or other similar jobs, homemaker, and unemployed for a total of 5 categories of occupation status. The specific occupations of individuals reporting ‘business workers’ or ‘other similar jobs’ were heterogeneous but primarily included merchants, store or factory owners or contractors. Anthropometric measurements including height, weight and blood pressure were performed using standard techniques (Chen et al., 2007; Pierce et al., 2010; Chen et al., 2006). 2.3. Measurement of carotid IMT Carotid IMT measurements were performed between April 2010 and September of 2011 (on average 7.2 years after baseline demographic information was obtained). Detailed methodology for carotid IMT measurements used in this study have been described elsewhere (McClintock et al., 2014a). Briefly, all measurements were obtained by 1 physician trained in sonography on a SonoSite MicroMaxx ultrasound machine (SonoSite, Bothell, WA) equipped with an L38e/10-5 MHz transducer. The protocol used was designed and implemented in the Oral Infections and Vascular Disease Epidemiology Study (INVEST) (Desvarieux et al., 2005). The carotid arteries were scanned longitudinally in 3 segments using the lateral extent of each carotid segment as defined relative to the tip of the flow divider, which is typically the most well defined anatomical reference in the proximity of the carotid bifurcation. The mean of the near and far walls of the maximum common carotid artery IMT of both sides of the neck (mean of twelve measurements) was used as the main outcome variable which is similar to previous reported studies (O'Leary et al., 1999; Kitamura et al., 2004; Touboul et al., 2012). Carotid IMT values are reported in this paper in thousandths of a millimeter (typically reported in tenths of a millimeter) to remain consistent with previously published studies using this same cohort (Chen, et al., 2013a). 2.4. Statistical analysis Of the 1206 participants, we excluded those who were missing data on SES (n = 2), systolic blood pressure (SBP) (n = 172), body mass index (BMI) (n = 180), cigarette use (past and present) (n = 172), and betel nut use (n = 174). The final analysis included a total of 1022 participants. We first conducted descriptive analysis of population characteristics and socioeconomic status across interquartile carotid IMT. Occupation status was divided into 5 categories as described above. Land ownership was modeled as a yes/no variable as well as split into no land, less than or equal to 1 acre and N1 acre. Education status was similarly broken down into yes/no and split into no education, 1 to 5 years of schooling or 6 or greater years, according to the Bangladeshi education system. Interquartile differences were analyzed using a one-way analysis of variance (ANOVA) for continuous variables and chi-squared for categorical variables. Then, we conducted linear regression models with carotid IMT modeled as a continuous dependent variable and each SES attribute modeled as an independent variable. Assumptions of linear regression were checked and none was violated. Next carotid IMT was modeled using multivariable analysis adjusting firstly for age and sex (model 1), and then BMI, SBP, smoking (never, past and present smokers), and betel nut use (never, past and present users) additionally (model 2). These factors are known independent predictors of carotid IMT. The comparison of the two models can help assess to what extent the association between SES and cIMT can be explained by other variables, including potential mediators. Multivariable regression analysis was performed for the overall cohort and stratified for men and women separately. Stratified analysis was also conducted by age within men in an attempt to further assess whether the association between occupation and IMT differed by age. Because in our cohort
8
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
of the initially only statistically significant associations seen between SES and men, we did not perform stratified analysis in women. We also tested interaction between each SES variable with gender and with age in men by entering a cross-product term in the models. Adjustment of the results for DMII was not significantly different and thus was not included in the final multivariate model. All statistical analysis was performed using STATA 11 (Statacorp. College Station, TX). A 2-sided alpha error of b0.05 indicated statistical significance.
were associated with higher levels of carotid IMT, compared with the occupation of factory workers. Homemakers were associated with a low level of carotid IMT. In age- and sex-adjusted models (model 1), the association between occupation and carotid IMT remained similar, but the association between land ownership and carotid IMT disappeared. When compared to those with no schooling, there was a significant relationship between a higher educational attainment (≥ 6 years of schooling, p = 0.01) and carotid IMT (model 1, Table 2). After multivariable adjustment for risk factors and potential confounders and mediators (BMI, SBP, cigarette and betel nut usage), the positive relationship observed between television (p = 0.58) or land ownership (p = 0.92) and carotid IMT were no longer present. The positive associations of business occupation and unemployment with carotid IMT remained significant. We further investigated the association between SES indicators and carotid IMT by sex. Among men, after multivariable adjustment, carotid IMT was higher among business workers when compared to factory workers (26.7 μm, 95% CI (6.7–46.8), p b 0.01) (Table 3). There was no significant association between occupation status and carotid IMT in women. We also explored whether the association between SES and carotid IMT differed by age in men (Table 4). The association was stronger in older men (≥ 50 years old; 58.7 μm, 95% CI (17.2–100.0)) compared to younger men (b 50 years old; 13.7 μm, 95% CI (−7.8–35.2)) with an interaction noted for the difference in age (p = 0.048).
3. Results Characteristics of the study population are listed in Table 1. The majority of the participants were b50 years old (65%), female (60%) and had a BMI b25 (80%). Over 1/3 of the participants were active smokers or had a history of smoking (37%); a small percentage of participants had diagnosed hypertension (13%) and 2% were diagnosed with DM II (2%). The majority of participants had at least one or more year of schooling (53%) and under half owned televisions (44%) and land (48%). Women were primarily homemakers (95%) and men worked as factory workers (24%), laborers (18%) or in business (55%). The proportion of men (p b 0.01), older individuals (p b 0.01) and those with elevated blood pressure (b 0.01) increased with increasing quartiles of carotid IMT. Similarly the percentages of those who were ever smokers (p b 0.01), had DMII (P = 0.02) or used betel nut (p b 0.01) increased as well across interquartile carotid IMT. BMI was not associated with increased interquartile carotid IMT (p = 0.77). Univariate analyses of SES variables with demographic and lifestyle variables were shown in Supplemental Table 1. Table 2 shows the association between SES status indicators and carotid IMT in the overall study population. In univariate analysis, owning a television (13.8 μm, p = 0.02) and owning greater than one acre of land (20.5 μm, p = 0.03) were positively associated with higher levels of carotid IMT. Daily laborer, business workers, and unemployment
4. Discussion As each developing country is unique in regards to the contribution of SES to CVD, our study in this Bangladesh cohort is one of the first to evaluate and identify the SES factors associated with subclinical atherosclerosis with the use of carotid IMT. In this study of rural Bangladesh subjects, there were significant univariate associations between carotid IMT and SES factors such as types of employment, land ownership, and
Table 1 Distribution of population characteristics by carotid IMT quartiles. Characteristics
Overall
Quartile 1 (606.4–738.3 μm)
[n = 1022] Carotid IMT (μm) Male sex (%) Age at IMT measurement (years) Body mass index (kg/m2) Cigarette use (%)b Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Diabetes (%) Betel nut use (%)c Television ownership (%) Education (years) No education (%) 1–5 years of schooling (%) ≥6 years of schooling (%) Land ownership (%) No land (%) ≤1 acre (%) N1 acre (%) Occupation Factory worker (%) Daily laborer (%)d Business (%) Homemaker (%) Unemployed (%)
807.5 40 46.3 21.1 37 122.4 75.4 2 55 44 3.1 47 31 21 49 52 34 14 10 7 24 57 2
9 1 12 77 0.4
p-Valuea
Quartile 2 (738.8–788.2 μm)
Quartile 3 (788.3–860.8 μm)
Quartile 4 (860.1–1423.3 μm)
[n = 257]
[n = 254]
[n = 257]
[n = 254]
703.8 19 40 21.3 18 118 73.7 1 40 35 3.1 45 32 23 42 60 29 11
764.2 35 44.2 20.9 30 118.6 74.1 0.4 54 44 2.9 51 30 20 51 51 37 12
822.1 44 48.3 21.1 42 122.8 75.3 3 58 51 3.1 46 35 19 52 49 33 18
940.8 63 52.6 21.1 59 129.9 78.4 4 67 46 3.3 48 28 24 52 49 35 16
b0.01 b0.01 b0.01 0.77 b0.01 b0.01 b0.01 0.02 b0.01 b0.01 0.70 0.61 0.39 0.51 0.07 0.06 0.27 0.09
13 5 19 63 0
10 10 28 53 0
9 14 37 25 4
0.24 b0.01 b0.01 b0.01 b0.01
Carotid IMT and demographic data collected from Araihazar Bangladesh between 2010 and 2011. Carotid IMT; Carotid intima-media thickness. a Chi-squared test was used for categorical variables and ANOVA for continuous variables. b Cigarette use includes past and current usage. c Betel nut use includes past and current usage. d Includes daily laborers andfarmers. 28 participants excluded from categorical analysis of land due to unknown amount of land ownership.
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
9
Table 2 Association between socioeconomic status and carotid IMT. Socioeconomic status
N
TV ownership No Yes Occupation Factory worker Daily laborerc Business Homemaker Unemployed Land ownershipb No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
1022 574 448 1022 104 75 246 585 12 994 520 334 140 1022 485 318 219
Multivariable model 1a
Univariable model
Multivariable model 2d
cIMT (μm, CI)
p-Value
cIMT (μm, CI)
p-Value
cIMT (μm, CI)
p-Value
Reference 13.8 (1.9 to 25.7)
0.02
Reference 6.5 (−3.1 to 16.2)
0.18
Reference −2.4 (−12.4 to 6.9)
0.58
Reference 64.4 (37.5 to 91.3) 47.0 (26.5 to 68.1) −20.1 (−38.9 to −1.2) 115.2 (61.0 to 169.3)
b0.01 b0.01 0.04 b0.01
Reference 20.4 (−3.1 to 43.9) 38.1 (20.2 to 56.0) 30.4 (−2.9 to 62.9) 65.4 (18.4 to 112.5)
0.09 b0.01 0.07 b0.01
Reference 22.7 (−0.02 to 45.4) 30.2 (12.8 to 47.6) 29.5 (−1.8 to 60.9) 65.9 (20.5 to 111.2)
0.05 b0.01 0.07 b0.01
Reference 9.7 (−3.6 to 23.0) 20.5 (2.5 to 38.6)
0.15 0.03
Reference 2.5 (−8.2 to 13.3) 6.7 (−7.9 to 21.5)
0.64 0.37
Reference 2.0 (−8.5 to 12.6) −0.8 (−15.2 to 13.7)
0.71 0.92
Reference −2.6 (−16.3 to 11.1) 7.3 (−8.1 to 22.8)
0.71 0.35
Reference 4.1 (−7 to 15.3) 16.7 (3.9 to 29.4)
0.47 0.01
Reference −0.7 (−11.7 to 102.5) 8.6 (−4.3 to 21.5)
0.90 0.19
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariable model 1 adjusted for age and sex. b Multivariable model 2 adjusted for age, sex, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). c Includes daily laborers and farmers. d 28 participants had an unknown amount of land ownership (26 women, 2 men).
TV ownership (Table 2). The associations of land and TV ownership with carotid IMT levels were largely explained by differences in other predictors of subclinical atherosclerosis including age, sex, BMI, systolic blood pressure, cigarette use and betel nut use. After adjustment of these predictors and potential confounders of CVD risk, elevated carotid IMT among men who were in the business sector remained significant when compared to factory workers. Further stratified analysis by age revealed that the association between being employed in the business sector and IMT was stronger in men over the age of 50 compared to younger men (Table 4). There have been few studies evaluating SES and CVD in the South Asian and Bangladesh population. In a study using 2010–2012 surveillance data from a rural Bangladesh population, non-communicable disease deaths (which including CVD and CVD equivalents) were strongly associated with higher SES groups (Hanifi et al., 2014). When comparing SES status and CVD risk factors in the Bangladesh population, the results are similar. In a 2011 Bangladesh Demographic and Health Survey
study, after adjustment for potential confounders, elevated income and education status were related to HTN and hyperglycemia. (Harshfield et al., 2015). These results, when examined within the context of our study, help to refine in the rural population what predicts subclinical atherosclerosis. By having a job associated with higher overall wealth such as someone employed in the business sector, one is more likely to have subclinical atherosclerosis. Our study is important because CVD mortality remains the leading causing of death in the world, and while mortality is declining in developed countries, in developing ones it is rising (Rasiah et al., 2013). The contribution of SES to CVD depends on the country studied. In the US, as in most developed nations, higher SES is associated with a lower incidence of CVD (Kaplan and Keil, 1993). There have been several studies in developed countries showing associations between lower SES and subclinical atherosclerosis (Thurston et al., 2014; Nash et al., 2011; Lutsey et al., 2008). In the Multi-Ethnic Study of Atherosclerosis (MESA) lower SES and longer duration living in the United States after
Table 3 Association between socioeconomic status and carotid IMT by gender. Multivariable model - mena
Socioeconomic status N TV ownership No Yes Occupation Factory worker Daily laborerb Business Homemaker Unemployed Land ownershipc No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
408 225 183 408 99 73 226 . 10 406 200 148 58 408 161 140 107
cIMT (μm, CI)
Multivariable model - womena p-Value
Reference −3.0 (−20.0 to 14.0)
0.73
Reference 19.6 (−6.6 to 45.7) 26.7 (6.7 to 46.8) . 47.3 (−8.1 to 102.6)
0.14 b0.01 . 0.09
Reference 0.7 (−17.2 to 18.6) 6.5 (−18.8 to 31.7)
. 0.94 0.62
Reference −4.1 (−23.3 to 15.0) 6.8 (−14.4 to 27.9)
. 0.67 0.53
N 614 349 265 614 5 2 20 585 2 588 320 186 82 614 324 178 112
cIMT (μm, CI)
p-Value
Interaction p-value
Reference −2.4 (−13.9 to 9.0)
0.68
0.33
Reference −36.6 (−150.6 to 77.3) −6.4 (−74.6 to 61.8) 2.1 (−58.9 to 63.2) 92.9 (−21.6 to 207.3)
. 0.53 0.85 0.95 0.11
0.29 0.26 . 0.66
Reference 3.4 (−9.3 to 16.1) −6.7 (−23.9 to 10.5)
. 0.53 0.45
0.88 0.09
Reference −0.3 (−13.1 to 13.1) 6.8 (−9.4 to 23.1)
. 0.95 0.41
0.92 0.59
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariate model adjusted for age, sex, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). b Includes daily laborers and farmers. c 28 participants had an unknown amount of land ownership (2 men, 26 women).
10
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
Table 4 Association between socioeconomic status and carotid IMT in men stratified by age. Men ≥ 50 years of agea
Socioeconomic status N TV ownership No Yes Occupation Factory worker Daily laborerb Business Unemployed Land ownership No land One acre or less Greater than one acre Education status No education 1–5 years of schooling ≥6 years of schooling
174 91 83 174 27 47 92 8 174 74 67 33 174 72 54 48
cIMT (μm, CI)
Men b 50 years of agea p-Value
N
cIMT (μm, CI)
p-Value
Interaction p-value
Reference −0.6 (−30.3 to 29.1)
0.97
234 134 100
Reference −4.6(−24.8 to 15.7)
0.66
0.78
Reference 30.6 (−15.4 to 76.7) 58.7 (17.2 to 100.0) 92.8 (14.3 to 171.3)
0.19 b0.01 0.02
Reference 26.4 (−6.6 to 59.4) 13.7 (−7.8 to 35.2) −15.8 (−120.4 to 88.8)
0.12 0.21 0.77
0.91 0.04 0.13
Reference 4.1 (−28.8 to 36.9) 22.1 (−19.1 to 63.3)
0.81 0.29
Reference −1.7 (−22.2 to 18.9) −8.9 (−41.1to 23.4)
0.87 0.59
0.82 0.30
Reference −8.1 (−43.4 to 27.2) 24.7 (−13.3 to 61.9)
0.65 0.20
Reference −5.4 (−27.4 to 16.6) −7 (−31.9 to 17.9)
0.63 0.58
0.83 0.20
72 26 134 2 232c 126 81 25 234 89 86 59
cIMT; Carotid intima-media thickness. CI; Confidence interval. a Multivariate model adjusted for age, BMI, systolic blood pressure, cigarette use (past and present) and betel nut use (past and present). b Includes daily laborers and farmers c 2 participants had an unknown amount of land ownership.
immigration from developing countries were associated with increased carotid IMT (Lutsey et al., 2008). This is in contrast to CVD in developing nations, where the contribution of SES is variable (Reddy and Yusuf, 1998). A large meta-analysis evaluating 29 Asian countries showed inverse relationships between education, income, occupation status and CVD mortality similar to developed countries (Vathesatogkit et al., 2014). Other studies have noted that as developing nation's progress to middle income status and increased wealth with progressive industrialization, the resulting urbanization of the population causes CVD to rise (Yusuf et al., 2001; Clark et al., 2009). The findings from our study indicate that the contribution of SES to subclinical CVD in low-income populations differs and the association depends on the type of SES indicator studied. Other CVD risk factors can explain the relationship between SES markers such as educational status, TV, land ownership with carotid IMT. In our analyses, the positive association between educational attainment and carotid IMT can be explained by differences in blood pressure, cigarette use, betel nut use, and BMI. However, the association between business employment and IMT remained after additional adjustments. There may be other unmeasured specific factors related to SES such as diet, health education, and access to health care that may mediate the association between SES and cIMT. Future studies however, would be needed to assess these hypotheses and identify the mediators. Future studies could also use our described associations to focus on the specific SES indicators to measure in rural populations of other developing nations. The biological plausibility behind the contribution of SES within the context of lifestyle changes and subclinical atherosclerosis in Bangladesh is likely multifactorial. The South Asian population is noted to have accelerated rates of premature coronary artery disease when compared to western counterparts (Islam and Majumder, 2013). Previous epidemiologic studies have estimated the prevalence of atherosclerotic heart disease up to 3.4% in the rural population and 19.6% in urban populations (Islam and Majumder, 2013). These studies however have been small, and it is likely that the prevalence and overall mortality from cardiovascular disease within the Bangladesh population is rising (Ahsan Karar et al., 2009). Additionally, metabolic syndrome, diabetes and obesity are all rising (Bhowmik et al., 2015). Contributing to this rise in cardiovascular disease, in a 2010 Bangladesh survey, 27% of subjects were found to have low levels of activity and when compared to western guidelines of what constitutes adequate exercise, 87% of Bangladeshis
did not meet them (Islam and Majumder, 2013; Hayes et al., 2002). The traditional Bangladesh diet, already poor in fruits and vegetables and high in grains is undergoing a further transition with the rise of fast food culture and western style foods high in saturated fats (Islam and Majumder, 2013). Given these lifestyle related changes it is not surprising that in our study, individuals engaged in business sector were more prone to subclinical atherosclerosis. Our paper has several limitations. This is a post-hoc analysis of an existing prospective cohort initially designed to study arsenic exposure, well water and CVD. However, since the HEALS is a population-based cohort study, the resources allow the evaluation of other risk factors for CVD. We have also previously shown significant associations with specific dietary patterns (Chen et al., 2006; Jiang et al., 2014; Chen et al., 2013b; McClintock et al., 2014b), cigarette smoking (Ahn et al., 2013), betel nut chewing (McClintock et al., 2014a; Heck et al., 2012), and anthropometric measures (Pierce et al., 2010; Chen et al., 2014; Ge et al., 2014) with CVD. Most of the women in the study were homemakers, and thus we cannot apply these findings to women who work in the business sector. Additionally, because this cohort was designed in a rural population, the applicability to urbanized areas is limited. Based on limited previous literature, we made assumptions about what constitutes SES and thus we may have missed some markers of SES that more accurately reflected our cohort characteristics. Our definition of businessmen is heterogeneous, including mostly individuals owning small grocery stores, clothing stores, small restaurants, and food-related business in the area, thus it may be hard to pinpoint one specific business sector to focus CVD prevention efforts on. Finally we used carotid IMT as a surrogate endpoint for overall cardiovascular disease, and elevated carotid IMT may not translate to the more important endpoint of clinically significant CVD. Regardless of these limitations, this analysis of SES and subclinical atherosclerosis within a Bangladesh cohort is novel, and deserves attention given the rising burden of CVD and the need for more information of what drives this CVD risk. Overall, our study found that in a rural Bangladesh population, elevated SES as defined by businessmen had elevated risk of subclinical atherosclerosis, and the association was stronger in men older than the age of 50. Education status, TV and amount of land ownership showed positive trends with carotid IMT, but the association can largely be explained by differences in other risk factors. These findings are novel and represent new information on one association of potential
M. Garshick et al. / Preventive Medicine 102 (2017) 6–11
drivers of CVD within Bangladesh. Given the rising CVD burden, it will be necessary to focus preventive cardiovascular services on specific subpopulations such as those employed in small business. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ypmed.2017.06.022. Conflict of Interst The authors report that there are no potential conflicts of interst. Funding This study was supported by grants from the National Institutes of Health (R01 ES017541, P42 ES010349, and P30 ES000260). References Ahn, J., et al., 2013. Human gut microbiome and risk for colorectal cancer. J. Natl. Cancer Inst. 105 (24), 1907–1911. Ahsan Karar, Z., Alam, N., Kim Streatfield, P., 2009. Epidemiological transition in rural Bangladesh, 1986–2006. Glob. Health Action 2. Ahsan, H., et al., 2006. Health Effects of Arsenic Longitudinal Study (HEALS): description of a multidisciplinary epidemiologic investigation. J. Expo. Sci. Environ. Epidemiol. 16 (2), 191–205. Bhowmik, B., et al., 2015. Comparison of the prevalence of metabolic syndrome and its association with diabetes and cardiovascular disease in the rural population of Bangladesh using the modified National Cholesterol Education Program Expert Panel Adult Treatment Panel III and International Diabetes Federation definitions. J Diabetes Investig 6 (3), 280–288. Chen, Y., et al., 2006. Nutritional influence on risk of high blood pressure in Bangladesh: a population-based cross-sectional study. Am. J. Clin. Nutr. 84 (5), 1224–1232. Chen, Y., et al., 2007. Arsenic exposure from drinking water, dietary intakes of B vitamins and folate, and risk of high blood pressure in Bangladesh: a population-based, crosssectional study. Am. J. Epidemiol. 165 (5), 541–552. Chen, Y., et al., 2013a. Arsenic exposure from drinking water, arsenic methylation capacity, and carotid intima-media thickness in Bangladesh. Am. J. Epidemiol. 178 (3), 372–381. Chen, Y., et al., 2013b. Prospective investigation of major dietary patterns and risk of cardiovascular mortality in Bangladesh. Int. J. Cardiol. 167 (4), 1495–1501. Chen, Y., et al., 2014. A prospective study of arm circumference and risk of death in Bangladesh. Int. J. Epidemiol. 43 (4), 1187–1196. Clark, A.M., et al., 2009. Socioeconomic status and cardiovascular disease: risks and implications for care. Nat. Rev. Cardiol. 6 (11), 712–722. Desvarieux, M., et al., 2005. Periodontal microbiota and carotid intima-media thickness: the Oral Infections and Vascular Disease Epidemiology Study (INVEST). Circulation 111 (5), 576–582. Diez-Roux, A.V., et al., 1995. Social inequalities and atherosclerosis. The atherosclerosis risk in communities study. Am. J. Epidemiol. 141 (10), 960–972. Ebrahim, S., et al., 1999. Carotid plaque, intima media thickness, cardiovascular risk factors, and prevalent cardiovascular disease in men and women: the British Regional Heart Study. Stroke 30 (4), 841–850. El-Saharty, S.A., Zunaid, Karar, Koehlmoos, Tracey L.P., Engelgau, Michael M., 2013. Tackling noncommunicable diseases in Bangladesh: now is the time. Health, Nutrition & Population (SASHN). Enas, E., S., A., 2001. Coronary artery disease in Asian Indians: an update and review. International Journal of Cardiology 1 (2). Ge, W., et al., 2014. Association between anthropometric measures of obesity and subclinical atherosclerosis in Bangladesh. Atherosclerosis 232 (1), 234–241.
11
Hanifi, S.M., Mahmood, S.S., Bhuiya, A., 2014. Cause-specific mortality and socioeconomic status in Chakaria, Bangladesh. Glob. Health Action 7, 25473. Harshfield, E., et al., 2015. Association of hypertension and hyperglycaemia with socioeconomic contexts in resource-poor settings: the Bangladesh Demographic and Health Survey. Int. J. Epidemiol. 44 (5), 1625–1636. Hayes, L., et al., 2002. Patterns of physical activity and relationship with risk markers for cardiovascular disease and diabetes in Indian, Pakistani, Bangladeshi and European adults in a UK population. J. Public Health Med. 24 (3), 170–178. Heck, J.E., et al., 2012. Betel quid chewing in rural Bangladesh: prevalence, predictors and relationship to blood pressure. Int. J. Epidemiol. 41 (2), 462–471. Islam, A.K., Majumder, A.A., 2013. Coronary artery disease in Bangladesh: a review. Indian Heart J. 65 (4), 424–435. Jiang, J., et al., 2014. Association of major dietary patterns and blood pressure longitudinal change in Bangladesh. J. Hypertens. (in press). Kaplan, G.A., Keil, J.E., 1993. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation 88 (4 Pt 1), 1973–1998. Kitamura, A., et al., 2004. Carotid intima-media thickness and plaque characteristics as a risk factor for stroke in Japanese elderly men. Stroke 35 (12), 2788–2794. Lamont, D., et al., 2000. Risk of cardiovascular disease measured by carotid intima-media thickness at age 49-51: lifecourse study. BMJ 320 (7230), 273–278. Lutsey, P.L., et al., 2008. Associations of acculturation and socioeconomic status with subclinical cardiovascular disease in the multi-ethnic study of atherosclerosis. Am. J. Public Health 98 (11), 1963–1970. Lynch, J., et al., 1995. Socioeconomic status and carotid atherosclerosis. Circulation 92 (7), 1786–1792. Lynch, J., et al., 1997. Socioeconomic status and progression of carotid atherosclerosis. Prospective evidence from the Kuopio Ischemic Heart Disease Risk Factor Study. Arterioscler. Thromb. Vasc. Biol. 17 (3), 513–519. McClintock, T.R., et al., 2014a. Association between betel quid chewing and carotid intima-media thickness in rural Bangladesh. Int. J. Epidemiol. 43 (4), 1174–1182. McClintock, T.R., et al., 2014b. Major dietary patterns and carotid intima-media thickness in Bangladesh: the HEAL Study. Public Health Nutr. (in press). Nash, S.D., et al., 2011. Socioeconomic status and subclinical atherosclerosis in older adults. Prev. Med. 52 (3–4), 208–212. O'Leary, D.H., et al., 1999. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N. Engl. J. Med. 340 (1), 14–22. Pierce, B.L., et al., 2010. A prospective study of body mass index and mortality in Bangladesh. Int. J. Epidemiol. 39 (4), 1037–1045. Rasiah, R., et al., 2013. Cardiovascular disease risk factors and socioeconomic variables in a nation undergoing epidemiologic transition. BMC Public Health 13 (p. 886). Reddy, K.S., Yusuf, S., 1998. Emerging epidemic of cardiovascular disease in developing countries. Circulation 97 (6), 596–601. Rosvall, M., et al., 2000. Occupational status, educational level, and the prevalence of carotid atherosclerosis in a general population sample of middle-aged Swedish men and women: results from the Malmo Diet and Cancer Study. Am. J. Epidemiol. 152 (4), 334–346. Thurston, R.C., et al., 2014. Low socioeconomic status over 12 years and subclinical cardiovascular disease: the study of women's health across the nation. Stroke 45 (4), 954–960. Touboul, P.J., et al., 2012. Mannheim carotid intima-media thickness and plaque consensus (2004–2006–2011). An update on behalf of the advisory board of the 3rd, 4th and 5th watching the risk symposia, at the 13th, 15th and 20th European Stroke Conferences, Mannheim, Germany, 2004, Brussels, Belgium, 2006, and Hamburg, Germany, 2011. Cerebrovasc. Dis. 34 (4), 290–296. Vathesatogkit, P., Batty, G.D., Woodward, M., 2014. Socioeconomic disadvantage and disease-specific mortality in Asia: systematic review with meta-analysis of populationbased cohort studies. J. Epidemiol. Community Health 68 (4), 375–383. Yusuf, S., et al., 2001. Global burden of cardiovascular diseases: part I: general considerations, the epidemiologic transition, risk factors, and impact of urbanization. Circulation 104 (22), 2746–2753.