Immigration Status and Sex Differences in Primary Cardiovascular Disease Prevention: A Retrospective Study of 5 Million Adults.

Background We evaluated whether immigration status modified the association between sex and the quality of primary cardiovascular disease prevention in Ontario, Canada. Methods and Results We used a population-based administrative database-derived cohort of community-dwelling adults (aged ≥40 years) without prior cardiovascular disease residing in Ontario on January 1, 2011. In the preceding 3 years, we evaluated screening for hyperlipidemia and diabetes in those not previously diagnosed; diabetes control (HbA1c <7%); and medication use to control hypertension, hyperlipidemia, or diabetes in those with previous diagnosis. We calculated the absolute prevalence difference (APD) between women and men for each metric stratified by immigration status and then determined the difference-in-differences for immigrants compared with long-term residents. Our sample included 5.3 million adults (19% immigrants), with receipt of each metric ranging from 55% to 90%. Among immigrants, women were more likely than men to be screened for hyperlipidemia (APD, 10.8%; 95% CI, 10.5-11.2) and diabetes (APD, 11.5%; 95% CI, 11.1-11.8) and to be treated with medications for hypertension (APD, 3.5%; 95% CI, 2.4-4.5), diabetes (APD, 2.1%; 95% CI, 0.7-3.6) and hyperlipidemia (APD, 1.8%; 95% CI, 0.5-3.1). Among long-term residents, findings were similar except poorer medication use for diabetes (APD, -2.8%; 95% CI, -3.4 to -2.2) and hyperlipidemia (APD, -3.5%; 95% CI, -4.0 to -3.0]) in women compared with men. Conclusions The overall quality of primary preventive care can be improved for all adults, and future research should evaluate the impact of observed equal or better care in women than men, irrespective of immigration status, on cardiovascular disease incidence.

Recent immigrants have a lower risk of cardiovascular disease compared with longer‐term immigrants and nonimmigrants ; however, the magnitude of this health advantage may differ between immigrant women and men because of different postmigration experiences. It is also important to account for the ethnic origins of immigrants as variations in cardiovascular care and outcomes based on ethnicity have been previously described.

Cardiovascular care varies by sex, with women being less likely to be on guideline‐recommended medications for hypertension and hyperlipidemia, and less likely to achieve control of vascular risk factors compared with men. Intersectionality theory suggests that immigrant women may be more vulnerable to these sex disparities compared with other women ; however, little is known on potential sex differences in the quality of primary cardiovascular preventive care among immigrants. We conducted a retrospective cohort study in Ontario, Canada, to compare the quality of primary cardiovascular preventive care in women and men, and to determine if sex differences varied with immigration status. We hypothesized that women would receive poorer quality of primary preventive care compared with men and that this sex difference would be more pronounced in immigrants compared with long‐term residents.

Methods

The data set from this study is held securely in coded form at ICES. Although data‐sharing agreements prohibit ICES from making the data set publicly available, access may be granted to those who meet prespecified criteria for confidential access, available at https://www.ices.on.ca/das.

The Sunnybrook Health Sciences Center Research Ethics Board provided ethics approval for this study. The study uses existing administrative healthcare databases, and individual patient consent is not obtained for their use.

Setting

We used the CANHEART (Cardiovascular Health in Ambulatory Care Research Team) cohort, a population‐based cohort derived from linkage of multiple health administrative databases, to identify all community‐dwelling adults aged 40 to 105 years on January 1, 2011, without prior cardiovascular disease (up to 23‐year look‐back window), residing in Ontario and who were eligible for the provincial health insurance plan for at least 5 years before the inception date. We excluded individuals residing in long‐term care homes because their care needs may differ from those of community‐dwelling adults and because immigrants are less likely than nonimmigrants to reside in a long‐term care home. We also excluded those residing in rural Ontario (geographically defined communities with a population of <10 000) because most immigrants reside in urban Ontario.

Exposures and Outcomes

Our exposures were sex and immigration status. People born outside of Canada and arriving in 1985 or later as per the Immigration, Refugees and Citizenship Canada Permanent Resident database were categorized as immigrants. Those born in Canada or immigrating before 1985 were categorized as long‐term residents. This definition was necessary because data on immigration status were not available before 1985.

The primary outcomes were screening for hyperlipidemia and diabetes in patients without these diagnoses and glycemic control (HbA1c <7%) in those with diabetes in the 3 years before cohort inception. We also evaluated whether medications to control hypertension, hyperlipidemia, and diabetes were filled at least once in the year prior among patients aged >64 years (between January 1, 2010, and December 31, 2010) in those with relevant diagnoses. We used the Ontario Laboratories Information System database to determine screening of hyperlipidemia and screening and control of diabetes, and the Ontario Drug Benefit database to determine medication use.

We obtained information on demographic characteristics (age and neighborhood‐level income), comorbidities (hypertension, diabetes, hyperlipidemia, atrial fibrillation, chronic obstructive pulmonary disease, congestive heart failure), Charlson comorbidity index (categorized as tertiles), and, on a subsample linked to survey data, self‐reported vascular risk factors (physical activity, smoking status, and obesity status based on self‐reported height and weight, unhealthy diet, alcohol consumption, and stress). Details on data sources and operationalization of these variables are provided in Table S1.

Statistical Analysis

We compared differences in demographics, comorbidities, and self‐reported vascular risk factors between women and men, stratified by immigration status, using the chi‐square test for categorical variables and the Kruskal‐Wallis test for continuous variables, and by reporting standardized difference, with values >0.10 suggesting a potentially meaningful difference. Standardized differences express the difference in means or prevalence between 2 populations as a proportion of the pooled standard deviation.

For each outcome, we used generalized linear models with the binomial distribution and identity link function to calculate absolute prevalence differences (APD) in women compared with men. We undertook these analyses separately for immigrants and long‐term residents, reporting unadjusted estimates for the outcome of medication use and age‐adjusted estimates for the screening and control outcomes. We calculated the difference‐in‐difference estimates by taking the difference in APD between women and men among immigrants (first difference) and that among long‐term residents (second difference) to evaluate the modifying role of immigration status on sex‐outcome associations.

World Region of Immigrants and Years Lived in Ontario

To evaluate if the region of origin of immigrants (which closely relates to ethnicity of immigrants) had an influence on the observed sex differences in the quality of preventive care, we calculated estimates of association for each quality metric comparing women to men based on the following 7 world regions: Africa, Caribbean, East Asia, Latin America, Middle East, South Asia, and Western countries. We calculated the difference‐in‐differences estimate for each world region with long‐term residents as the reference group.

Finally, we evaluated whether sex differences in primary preventive care varied among immigrants based on the number of years lived in Ontario (<10 years or ≥10 years).

Results

We included 5.3 million adults (19% immigrants). The proportion of women among immigrants was slightly lower than in long‐term residents (51.5% versus 53.4%; P<0.001) (Table). Compared with men, women in both immigrant and long‐term resident groups had a lower prevalence of diabetes and hyperlipidemia and a higher prevalence of hypertension (Table). Information on self‐reported vascular risk factors, obtained through linkage with self‐reported community health surveys, was available for only 0.7% of the total sample. Among both immigrants and long‐term residents, women were less likely than men to have an unhealthy diet, be a current smoker, or have heavy alcohol use, but were more likely to be physically inactive (Table). Immigrant women were more likely to report being obese than immigrant men, whereas the reverse was observed among long‐term residents (Table).

Table 1

Baseline Characteristics of 5.3 Million Ontario Adults (≥40 years) Without Cardiovascular Disease

Characteristics of interest	Immigrants (N=984 978)		Long‐term residents (N=4 352 340)
	Women n=502 905 (51.5)	Men n=482 073 (48.5)	Women n=2 323 935 (54.5)	Men n=2 028 405 (45.5)
Median age, Q1–Q3	50 (45–60)	50 (45–57)	56 (48–67)	55 (47–64)
Neighborhood‐level income, n (%)
Lowest quintile	128 970 (25.6)	122 313 (25.4)	372 257 (16.0)	323 560 (16.0)
Highest quintile	66 691 (13.3)	62 107 (12.9)	547 364 (23.6)	483 694 (23.8)
Comorbidities, n (%)
Hypertension	147 532 (29.3)	128 990 (26.8)	874 481 (37.6)	709 875 (35.0)
Diabetes	63 785 (12.7)	65 935 (13.7)	282 151 (12.1)	278 767 (13.7)
Hyperlipidemia	113 836 (22.6)	142 853 (29.6)	584 285 (25.1)	574 884 (28.3)
Atrial fibrillation	3912 (0.8)	3858 (0.8)	46 108 (2.0)	47 719 (2.4)
CHF	3173 (0.6)	2380 (0.5)	27 870 (1.2)	23 390 (1.2)
COPD	3980 (0.8)	5945 (1.2)	77 020 (3.3)	67 914 (3.3)
Charlson comorbidity, n (%)
Medium	12 122 (2.4)	9745 (2.0)	92 206 (4.0)	76 497 (3.8)
High	12 278 (2.4)	9845 (2.0)	103 400 (4.4)	94 572 (4.7)

Values in parentheses represent proportion unless otherwise specified. CHF indicates congestive heart failure; and COPD, chronic obstructive pulmonary disease.

On the basis of information obtained by linkage with the Canadian Community Health Surveys. Charlson comorbidity index divided into low, medium, and high categories based on tertiles.

Long‐term resident women received the most care and immigrant men received the least care. Among immigrants, compared with men, women were more likely to be screened for hyperlipidemia (67.0% versus 55.5%; age‐adjusted APD, 10.8%; 95% CI, 10.5–11.2) and diabetes (74.7% versus 64.0%; age‐adjusted APD, 11.5%; 95% CI, 11.1–11.8]); equally likely to have glycemic control (69.0% versus 67.7%; age‐adjusted APD, 95% CI, 0.8%; 95% CI, 0.0–1.1); and more likely to be treated with medications for hypertension (65.2% versus 61.8%; APD, 3.5%; 95% CI, 2.4–4.5]), diabetes (56.1% versus 53.9%; APD, 2.1%; 95% CI, 0.7–3.6), and hyperlipidemia (56.6% versus 54.8%; APD, 1.8%; 95% CI, 0.5–3.1) (Figure 1). Among long‐term residents, findings were similar, except women were less likely than men to be treated with medications for diabetes (53.6% versus 62.1%; APD, −2.8%; 95% CI, −3.4 to −2.2) and hyperlipidemia (58.6% versus 62.1%; APD, −3.5%; 95% CI, −4.0 to −3.0), and, among those with diabetes, women were more likely than men to achieve glycemic control (73.2% versus 69.7%; APD, 3.2%; 95% CI, 2.7–3.6). Immigration status did not modify the sex‐outcome association, except for medication use (Figure 1).

Figure 1

Sex differences in cardiovascular preventive care comparing women with men among immigrants and long‐term residents, and the difference‐in‐difference estimates comparing sex differences among immigrants and long‐term residents.

*In people aged >64 years with a relevant diagnosis.

Results by World Region of Immigrants and Years Lived in Ontario

Findings of equal or better care in women compared with men were generally similar across immigrant groups from different world regions, with variable magnitude, except for the use of medications for hyperlipidemia and diabetes, which was less common in women than men among immigrants from Africa (Figure 2). The sex differences in primary preventive care did not vary among immigrants on the basis of their years lived in Ontario, except for a relative improvement in the use of medications for hyperlipidemia and diabetes in women compared with men among immigrants who have been in Ontario longer (Figure S1).

Figure 2

Primary cardiovascular preventive care in immigrant women compared with men based on region of origin of immigrants.

 

Discussion

In this population‐based cohort of over 5 million people, the overall quality of primary cardiovascular preventive care in women was similar to or better than that in men, and this was true in both immigrants and long‐term residents.

These findings are consistent with some previous studies that have shown that women are more likely than men to have adequate control of diabetes and to be screened for vascular risk factors. Improvements in awareness of cardiovascular disease in women over time may explain the favorable results for women in our study. Further studies are needed to identify patient‐, physician‐, or organization‐level drivers of the observed sex‐specific variation in the quality of primary preventive care.

Immigration status‐sex differences in the prevalence of hypertension have been observed in the United States, and we found similar differences in medication use for hyperlipidemia and diabetes. Our finding of a lack of variation in sex differences by immigration status for other outcomes supports the need to evaluate this in other jurisdictions. An explanation may be that all Ontario residents, including immigrants, are covered for hospital and essential physician services, investigations ordered by physicians, and medications in those aged >64 years.

Sex, race, and ethnicity differences in primary preventive care have been previously described, but immigration status is generally not accounted for. Compared with men, African immigrant women in our study, but not Caribbean immigrant women, were less likely to be on cholesterol‐lowering drugs. We found variation in the magnitude of difference in screening for vascular risk factors between women and men by region of origin of immigrants, but these were not significantly different. Further, our findings of slight improvement in the use of cholesterol‐lowering and antihyperglycemic medications in immigrant women compared with men with longer duration of stay suggests that acculturation could play a role and needs further evaluation using appropriate measures of acculturation. Future research should evaluate drivers of these variations based on region of origin, which may include sex‐based health‐seeking patterns of immigrant groups based on their country of origin and characteristics of health systems.

Most guidelines recommend screening for hyperlipidemia and diabetes in people aged >40 years at least once in 3 years, and our finding of 60% to 75% screening for these risk factors in the overall sample suggests the need to improve guideline‐recommended preventive care. The format of the guidelines, the language used, and the lack of absolute risk differences reporting are some factors associated with poor uptake of guideline recommendations in clinical practice. Organizational change and physician and patient education are potential avenues to improve screening and treatment of vascular risk factors. Finally, targeted screening in relatives of people with cardiovascular disease and use of e‐health tools or decision‐support software in primary care practices have also been shown to improve screening rates.

Strengths and Limitations

Our study is strengthened by using routinely collected data to determine screening and control of vascular risk factors, and drug claims to capture drug use among almost the entire population of a province.

Study limitations include a lack of medication data in patients aged <65 years. Further work in younger adults is needed, as prior data have suggested lower use of antihyperglycemic drug in women compared with men for cost‐cutting purposes in this population. A potential explanation for an overall lower rate of screening and treatment of vascular risk factors could be attributable to incomplete data in administrative databases; however, the incompleteness is unlikely to vary by sex or immigration status. For example, HbA1c measurements in the prior 3 years were available in about 90% of people with diabetes, without significant differences in this testing either by sex or immigration status. We could only evaluate self‐reported measures of vascular risk factors in <1% of the study sample; however, these data were derived from cross‐sectional surveys of representative Ontarians, allowing us to draw meaningful conclusions. Furthermore, among immigrants, we were unable to rule out cardiovascular disease occurring before migration, leading to potential misclassification as receiving primary preventive care. We also did not have information on factors that might affect use of preventive care, such as education or occupation. Although we did not have individual‐level income data, we used neighborhood‐level income as a proxy for socioeconomic status. Our data sources only allowed us to identify immigrants who arrived in 1985 or later, and so our findings are most generalizable to recent immigrants (ie, those arriving within the past 3 decades). We also do not have information on postmigration patterns; however, prior work suggests that only a minority of immigrants who landed in Ontario between 1991 and 2006 had moved to other provinces. Our study cohort was assembled in 2011, and it is possible that there have been changes in patterns of preventive care since that time. However, to our knowledge, these are the most recent data on sex differences in primary cardiovascular preventive care at a population level.

Implications of Our Findings

Contrary to our hypothesis, the observed sex differences in screening and treatment favored women over men, whereas immigration status did not significantly modify the association between sex and primary cardiovascular preventive care. These findings may suggest the importance of adequate healthcare coverage in eliminating immigration status–specific healthcare disparities in primary cardiovascular preventive care. Additionally, population‐level efforts to improve the overall quality of primary preventive care for all are needed, and future projects should evaluate the impact of the observed sex differences in primary preventive care on cardiovascular disease incidence and outcomes.

Sources of Funding

This study was funded from Heart and Stroke Foundation of Canada (Grant‐in‐Aid 19‐26333). The study was supported by ICES, which is funded in part by an annual grant from the Ontario Ministry of Health and Ministry of Long‐Term Care. This study was based on data compiled by ICES. Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information and Immigration, Refugees and Citizenship Canada. The analyses, opinions, results, and conclusions reported in this article are those of the authors and are independent from ICES, the funding and data sources. No endorsement by ICES, the Ministry of Health, Ministry of Long‐Term Care, Canadian Institute for Health Information, or Immigration, Refugees and Citizenship Canada, is intended or should be inferred.

Disclosures

Drs Austin and Kapral are supported by Mid‐Career Investigator Awards and Dr Yu by a New Investigator Award from the Heart and Stroke Foundation. Dr Kapral holds the Lillian Love Chair in Women’s Health from the University Health Network/University of Toronto, Canada. Dr Vyas holds a fellowship award from the Canadian Institutes of Health Research. The remaining authors have no disclosures to report.

Table S1

Figure S1

Click here for additional data file.