If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Stroke Prevention and Atherosclerosis Research Centre, Robarts Research Institute, Western University, London, Ontario, CanadaSchulich School of Medicine & Dentistry, University of Western Ontario, London, Ontario, Canada
Stroke Prevention and Atherosclerosis Research Centre, Robarts Research Institute, Western University, London, Ontario, CanadaDepartment of Clinical Neurological Sciences, Western University, London, Ontario, Canada
Stroke Prevention and Atherosclerosis Research Centre, Robarts Research Institute, Western University, London, Ontario, CanadaDepartment of Clinical Neurological Sciences, Western University, London, Ontario, CanadaDepartment of Epidemiology and Biostatistics, Schulich School of Medicine & Dentistry, Western University, London, Ontario, Canada
Social deprivation is associated with cardiovascular disease.
Social deprivation has a strong association with atherosclerosis and should be considered an important risk factor.
Intensive preventive therapy can prevent plaque progression irrespective of baseline marginalization index.
Background and aims
Socioeconomic status (SES) is associated with cardiovascular disease. However, the relationship between SES and atherosclerosis is not well documented. This study aims to explore this relationship.
This is a retrospective cohort study in London, Ontario Canada. It includes 6,907 subjects from a vascular prevention centre at baseline, with long term follow up from 1989 to 2021 (total ultrasound examinations 27,103). Using carotid ultrasound, the burden of atherosclerosis was measured as total plaque area (TPA). The Ontario Marginalization Index (OMI) was used to identify SES of participants' neighborhoods. We used a Bayesian hierarchical regression and mixed effects model to identify associations between SES, baseline TPA and plaque progression. In 2003, we implemented more intensive therapy of vascular risk factors after 2003 (called “Treating arteries instead of risk treating factors”); therefore, we compared our findings before and after 2003.
SES was found to have a significant association with TPA, with lower SES associated with higher TPA (adjusted odds ratio [OR] = 2.22, 95% Credible interval [CrI]: 1.37, 3.66). While we observed a higher rate of plaque progression with lower SES in those treated before 2003 (OR = 1.46, 95% CrI:1.04, 2.06), there was no significant association between plaque progression and SES after implementation more intensive therapy (OR = 0.99, 95% CrI: 0.78, 1.27).
SES has a strong association with atherosclerosis and should be considered an important risk factor in clinical practice and vascular disease research. Intensive preventive therapy can prevent plaque progression irrespective of baseline SES.
Knowledge regarding social and environmental risk factors of atherosclerosis is crucial to implementing preventive approaches and allocating public resources. Most research to date has focused on individual-based risk factors that are relatively proximal causes of disease and cannot explain all of the burden of atherosclerosis [
], the independent association between socioeconomic conditions and atherosclerosis requires clarification. The association between socioeconomic status and carotid intima media thickness (IMT) has been investigated in some studies [
]. In 666 participants, age and sex-adjusted IMT were higher in participants from the most deprived areas compared to those from the least deprived areas. The cross-sectional design and using IMT were two major limitations of these studies. IMT is biologically [
] distinct from atherosclerosis; plaque burden assesses true athersoclerosis. IMT also has a very narrow dynamic range (∼5–15 mm, compared to 0–1200 mm2 for Total Plaque Area), so it is much less powerful in regression models [
]. Carotid plaque area burden is therefore a more powerful phenotype than IMT for cohort studies.
Social studies of atherosclerosis are scarce due to the lack of cohort data, and to confounding factors between individual and socioeconomic data. The London, Atherosclerosis, eNvironmental, Demographic and Socioeconomic (LANDS; REB # 119841) study was designed to assess the association between the burden of atherosclerosis, measured as total plaque area (TPA), and social determinants of health. In the current study, we aimed to evaluate the association between social deprivation, the burden of plaque atherosclerosis at baseline, and plaque progression.
2. Patients and methods
2.1 Study database
This retrospective cohort study included a large sample of patients from London, Ontario, Canada, referred to the Stroke Prevention and Atherosclerosis Research Centre (SPARC). The study population includes those who were referred for stroke prevention, carotid stenosis, or for a personal or family history of vascular disease. The SPARC database contains information on carotid plaque burden measured as Total Plaque Area (TPA) and stenosis, progression of plaque and stenosis, risk factors (e.g. age, sex, blood pressure), and outcomes (e.g., stroke, death, myocardial infarction).
All cases were treated and followed at SPARC with a shift toward more intensive therapy in high-risk patients after 2003 (an approach that we called “Treating arteries rather than treating risk factors”) [
]. “Treating arteries” is a multi-domain approach in which the goal of therapy is not merely ot achieve target levels of risk factors such as blood pressure and LDL-C; the objective is to stop progression of plaque or achieve regression of plaque [
]. The process incorporates lifestyle modification, and intensive medication, enhanced by positive self-perceived patient feedback. This begins with showing patients images of their plaques, showing them on a graph that their arteries are much worse than expected for age and sex, and explaining that plaque regression is achievable and reduces risk. This motivates patients to comply with lifestyle and medication. Two trials have confirmed that this approach markedly improves compliance in vascular prevention [
]. In vascular prevention clinics across Argentina, this approach, implemented in 2010, resulted in a decline in the annual risk of vascular events among patients are >65 from 5.85% in 2011 to 2.56% in 2015 [
]. We defined TPA as the sum of the cross-sectional areas of all plaques detected in both carotid arteries between the clavicle and the angle of the jaw, measured by a registered ultrasound technologist at SPARC. Plaque was defined as a local thickening of the intima greater than 1 mm. We used a cutoff of 1.19 cm2, the top quartile of TPA, to define high risk patients, with a 5-year risk of stroke, myocardial infarction or vascular death of 19.5%. Plaque progression was defined as an increase of more than 0.05 cm2 per year, the median change in the study population [
]. Disseminated areas are the smallest geographic unit described by Statistics Canada. Using ArcMAP software version 10.8.1 in the LANDS study, postal codes were used to relate patient data to the dissemination areas in which they were located. In most populated regions such as cities, Canadian postal codes, which contain a combination of numbers and letters, can identify regions as local as to within one of two houses on one side of a street.
2.4 Socioeconomic status
We used the Ontario Marginalization Index (OMI) to identify socioeconomic status [
]. The OMI uses data from the 2016 Statistics Canada Census to calculate multiple variables to represent marginalization in areas across Ontario. It divides marginalization into four categories: material deprivation, dependency, ethnic concentration, and residential instability. The OMI has been used frequently to identify association between different comorbidities and SES and has demonstrated consistency to represent health inequities in Ontario, Canada [
]. The OMI has 5 categories; category 1 represents the least marginalized and 5 the most marginalized area.
2.5 Statistical analysis
Results are presented as mean ± standard deviation (SD) or median with interquartile range (IQR), according to their distribution. After a univariate data analysis, we entered variables with a significant or borderline p-value (p < 0.1) into the multiple regression model. We adjusted the final model for age, sex, diabetes, smoking status, and systolic blood pressure. The year of carotid plaque measurement was considered as a random effect to account for possible variability due to assessment periods. We used Bayesian hierarchical regression to examine the association between deprivation level and TPA. To assess the association between SES with plaque progression we used a mixed effects model. Any increase greater than 0.05 cm2 in the TPA over one year was defined as plaque progression.
Data regarding the burden of atherosclerosis (TPA) have a nested structure with clusters of participants recruited on a yearly basis. Therefore, we performed a 3-level mixed effects model in which observations are at level 1, patients are at level 2 and regions of assessment are at level 3 [
]. We then specified two Bayesian models regarding the distribution of TPA; an ordinal and a Bernoulli distribution (based on a TPA value of 1.19 cm2). We selected the final model as the one with the lowest PSIS-LOO (Pareto smoothed importance-sampling leave-one-out cross-validation) [
]. To ensure the accuracy of sample size, likelihood, and priors, we performed simulation analysis. Statistical analyses were conducted using the RStan and brms packages in the R 4.2 environment.
3.1 Demographic data
Descriptive statistics of the data used in the final analysis are provided in Table 1. Data are available for 6,874 patients (mean age: 63.2 ± 13.5 years), including 1,916 (27.9%) with a TPA >1.19 cm2. Patients were followed from 1 to a maximum of 26 ultrasound examinations, with 27,103 measurements of TPAs in total. There were significant similarities in OMI class 1 and 2, as well as in OMI 3 and 4 with regard to patients’ age, sex, risk factors and TPA. Therefore, to optimize sample size, we combined OMI 1 to 2 and 3 to 4 and then compared them with OMI 5 as the most marginalized area.
Table 1Demographic data according to socioeconomic status.
OMI 1 & 2
OMI 3 &4
63.79 ± 13.51
62.88 ± 13.98
62.78 ± 13.36
Body mass index (kg/m2)
27.86 ± 5.35
28.33 ± 5.5
28.67 ± 6.76
Systolic blood pressure (mmHg)
141.28 ± 20.59
141.86 ± 21.02
141.44 ± 19.99
Mean arterial pressure (mmHg)
99.90 ± 11.11
99.36 ± 11.89
99.73 ± 11.21
172.28 ± 56.94
174.65 ± 45.08
173.38 ± 48.21
Lipid lowering treatment
Total plaque area ≥1.19 mm2
Plaque progression (change in median)*
0.018 [-0.01, 0.07]
0.017 [-0.011, 0.07]
0.016 [-0.008, 0.071]
Plaque progression before 2003 (median)
0.04 [0, 0.145]
0.03 [-0.015, 0.094]
0.05 [0, 0.15]
Plaque progression after 2003
0.013 [-0.015, 0.06]
0.03 [-0.015, 0.065]
0.011 [-0.012, 0.06]
Plaque progression >0.05 cm2 per year n (%)
Plaque progression >0.05 cm2 per year before 2003
Plaque progression >0.05 cm2 per year after 2003 n (%)
3.2 The association between deprivation index and plaque area
The level of socioeconomic status had a significant association with TPA. In the non-adjusted Bayesian binary logistic model those with a lower socioeconomic status had a larger TPA (OR = 1.18, 95% CrI: 1.06, 1.34). After adjusting for confounding variables including age, sex, smoking status, diabetes, and systolic blood pressure, the odds ratio of a higher TPA was significantly higher in the most marginalized areas as compared to the least marginalized neighborhoods (OR = 1.86, 95% CrI:1.07, 3.24). This adjusted odds ratio was similar to the effect of sex on the burden of atherosclerosis (Table 2). We did not observe any significant difference in TPA of those living in OMI 3 and 4 as compared to 1 and 2 (Table 2).
Table 2The association of deprivation and the burden of atherosclerosis (total Plaque area): results of the multiple Bayesian regression model.
3.3 The association between deprivation index and plaque progression
In the multiple logistic Bayesian regression model, there was no difference in plaque progression according to study neighborhoods (OR = 1.23, 95% CrI: 0.94, 1.59). However, the time-deprivation interaction effects were statistically significant; therefore, we performed a subgroup analysis comparing cases before and after 2003, the year we began implementing the “Treating Arteries” approach. Those living in the most marginalized area had more plaque progression (OR = 1.46, 95% CrI: 1.04, 2.06; Reference group: OMI 1&2) prior to 2003. After 2003, there was no association between neighborhood deprivation index and plaque progression (OR = 0.99, 95% CrI: 0.78, 1.27).
Our study has important clinical and public health implications. In a cohort of 6,907 cases with long term ultrasound follow-up (27,103 ultrasounds), the burden of atherosclerosis had a strong association with neighborhood deprivation. While plaque progression was more common in those living in most marginalized areas prior to 2003, after implementing intensive therapy based on “Treating Arteries”, there was no longer an association between deprivation and plaque progression.
Several studies found a close association between SES and cardiovascular disease incidence and outcomes [
] adult participants. However, limitations of IMT were discussed above.
LANDS is a large cohort study of atherosclerosis with access to multiple individual-based and social risk factors. In the current study, we found a significant inverse association between a lower SES and the burden of atherosclerosis, measured by total plaque area. The reasons that SES may affect endothelial function and lead to atherosclerosis is not clear. A higher rate of vascular disease in lower SES groups can be partially explained by a higher rate of vascular risk factors among those with a lower SES [
]. While smoking and diabetes were more prevalent in more marginalized areas of our study, the association between deprivation and TPA was independent of vascular risk factors. It is possible that SES affects atherosclerosis via other pathways such as psychosocial stress and stress-related hypertension [
]. In a cohort study (the Multi-Ethnic Study of Atherosclerosis; n = 412 adults free of hypertension at baseline; follow-up: 6.5 years), the level of urinary stress hormones had a close association with incident hypertension and vascular events [
]. A psychological stress task (the Stroop task) was shown to significantly increase systolic and diastolic blood pressure and plasma norepinephrine in patients with hypertension. Amlodipine blunted the rise in blood pressure, but increased the plasma norepinephrine significantly [
A novel aspect of this study is that we followed our cases for an average of 3.5 years (range: 2–26 years). While those living in the most marginalized areas had a higher burden of plaque at their first visit, this did not necessarily predict more progression of plaque with intensive therapy: we observed that after initiating intensive therapy in 2003 (Treating Arteries) there was no difference in the rate of plaque progression in relation to the OMI categories. This is an important finding emphasizing the importance of adequate therapy in all cases regardless of the baseline SES. The success of “Treating Arteries” in eliminating differences in plaque progression may be partly explained by two factors: 1. The Ontario health care system provides physician visits, investigations free of charge in all patients, and pays for medication in persons who are unemployed or cannot afford their medication; therefore, our patients were able to follow our treatment plans 2. Blood pressure in our patients is well controlled by application of physiologically individualized therapy based on renin/aldosterone phenotyping, which has been shown in a controlled trial to markedly improve blood pressure control in patients resistant to usual therapy [
Our study has some limitations. While we did not collect individual based data regarding SES, the OMI provides SES data based on dissemination units in Ontario. Dissemination units are the smallest geographical unit described in Statistics Canada, containing only 400–700 persons, and most likely represent the true SES of our participants. In the current study, we used OMI according to the 2016 census. While there is a possibility of time mismatch between OMI and ultrasound assessments, this is a systematic limitation for all participants and should not affect our final results. In addition, there is a consistency in the indices across census years with limited or no change in quintiles of marginalization between 2006, 2011 and 2016 [
]. SPARC is not a population-based study with a random selection from the community. However, approximately 40% of our participants from London and its nearby region were referred by family physicians for primary prevention of vascular disease. Hence, we believe that our sample is fairly representative of high-risk middle-aged and older adults from the general population living in our region. Finally, atherosclerosis is due to several non-mutually exclusive mechanisms from traditional risk factors to genetic background [
]) Adjusting for many cofounders is difficult and requires further investigation. LANDS is a large cohort that can address many of these questions.
In summary, SES has a clear association with the burden of atherosclerosis. This is an important finding that explains in part the importance of social factors on vascular disease risk. Despite a higher burden of atherosclerosis at baseline in those with lower SES, plaque progression is not inevitable for these patients and intensive treatment of vascular risk factors regardless of SES may reduce the risk of plaque progression.
CRediT authorship contribution statement
Amadene B. Woolsey: Data gathering, Formal analysis, Writing – original draft, Writing – review & editing. Shahram Arsang-Jang: Consultant for data analysis, Writing – original draft, Writing – review & editing. J. David Spence: Conceiving the study, Writing – review & editing. Daniel G. Hackam: Writing – review & editing. M. Reza Azarpazhooh: Writing the first and final draft, data gathering, Formal analysis, editing, study design.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Traditional risk factors as the underlying cause of racial disparities in stroke: lessons from the half-full (empty?) glass.