Abstract
Background
Hypertension alone is a poor predictor of the individual risk of cardiovascular disease. Hereditary factors of which hypertension is merely a marker may explain why some hypertensive individuals appear more susceptible to cardiovascular disease, and why some ethnicities have more often seemingly hypertension‐related cardiovascular disease than others. We hypothesize that, in hypertensive individuals, a positive family history of cardiovascular disease identifies a high‐risk subpopulation.
Methods and Results
Healthy Life in Urban Settings (HELIUS) is a cohort study among participants of Dutch, South‐Asian Surinamese, African Surinamese, Ghanaian, Turkish, and Moroccan origin aged 70 years and younger. In participants with hypertension (n=6467), we used logistic regression to assess the association of family history of cardiovascular disease with prevalent stroke and nonstroke cardiovascular disease, adjusting for sex, age, education, and smoking. To detect ethnic differences, we tested for interaction between family history and ethnicity and stratified the analysis by ethnicity. A positive family history was associated with a higher prevalence of nonstroke cardiovascular disease (odds ratio [OR], 2.05; 95% CI, 1.65–2.54) and stroke (OR, 1.62; 95% CI, 1.19–2.20). The strongest association of family history with nonstroke cardiovascular disease was found among the Dutch (OR, 2.47; 95% CI, 1.37–4.44) and with stroke among the African Surinamese (OR, 2.17; 95% CI, 1.32–3.57). The interaction between family history and African Surinamese origin for stroke was statistically significant.
Conclusions
In multiethnic populations of hypertensive patients, a positive family history of cardiovascular disease may be used clinically to identify individuals at high risk for nonstroke cardiovascular disease regardless of ethnic origin and African Surinamese individuals at high risk for stroke.
Keywords: cardiovascular diseases, cerebrovascular disease/stroke, HELIUS study, hypertension, race and ethnicity
Subject Categories: Cardiovascular Disease, Race and Ethnicity, Risk Factors, Hypertension, Epidemiology
Introduction
Hypertension is considered the leading risk factor for cardiovascular disease (CVD), with myocardial infarction and stroke accounting for most cases.1 Because of its high prevalence, hypertension is responsible for the majority of the cases of CVD.2 However, data from longitudinal cohort studies are subject to confounding and cannot clearly show to what extent high blood pressure itself accounts for the observed risk and to what extent it is merely a marker that identifies high‐risk patients. Data from randomized controlled trials of antihypertensive treatment are less subject to confounding, and therefore more reliable than those from longitudinal studies for estimation of the individual risk for CVD originating from hypertension alone. A Cochrane meta‐analysis of 18 placebo‐controlled trials for primary prevention showed that, in patients with untreated hypertension, the yearly incidence of either myocardial infarction or stroke is <1%.3 This translates into a lifetime risk of either myocardial infarction or stroke of <20%, assuming risk increases linearly over the residual lifespan of subjects of similar age and country of origin as those in the trials.4 In other words, over 80% of all untreated hypertensive individuals will experience no event in their lifetime. These data suggest that hypertension does not seem to be a strong risk factor for all individuals but it probably is for certain subpopulations. If we could identify the specific subpopulations in which hypertension itself results in an increased risk, we would be able to better target screening and adapt treatment decisions to the risk level.
Family history of CVD could be used as a tool to identify which hypertensive individuals are at particularly high risk. A positive family history of CVD is an independent predictor of both myocardial infarction5, 6 and stroke.7, 8, 9 Prognostic models for stroke that include both hypertension and family history have predictive value greater than models that include hypertension or family history alone.8 Among individuals with hypertension, family history of CVD has been shown to be independently associated with mortality by ischemic heart disease10 and family history of CVD mortality with incidence of overall CVD.11 Therefore, a positive family history of CVD could reflect an underlying genetic predisposition related to CVD.12
In the current study, we investigated this possible role of family history in hypertensive individuals in a multiethnic population of relatively young age, since the incidence of hypertension‐related CVD differs across ethnic groups13, 14 even among persons with hypertension15, 16 and since young individuals may be erroneously assigned a low risk of CVD. Therefore, better risk estimation in this specific group could lead to better identification of hypertensive individuals at high risk for CVD, especially in multiethnic populations.
To test our hypothesis, we evaluated whether a positive family history of CVD is related to a higher prevalence of nonstroke CVD or stroke in a cross‐sectional study of hypertensive individuals of different ethnic origins.
Methods
Study Population
Healthy Life in Urban Settings (HELIUS) is a cohort study on health among different ethnic groups living in Amsterdam, the Netherlands. The HELIUS study is described in detail elsewhere.13 Briefly, baseline data collection took place in 2011–2015. Participants were randomly sampled from the municipal register, stratified by ethnicity, and included people aged 18 to 70 years of Dutch origin or belonging to one of the ethnic minority groups with a South‐Asian Surinamese, African Surinamese, Ghanaian, Turkish, or Moroccan origin. Data were obtained by questionnaire, physical examination, and collection of biological samples. The study protocols were approved by the ethical review board of the Academic Medical Center of the University of Amsterdam, and all individuals provided written informed consent.
There were 22 165 individuals for whom both questionnaire data and physical measurements were available. We excluded individuals with unknown ethnic origin (n=48) or a Surinamese origin other than African of South Asian (n=500). From the remaining 21 617 participants, after multiple imputation (see Statistical Analysis), we selected 6467 individuals with hypertension, who were included in the analysis.
Definitions
Ethnic origin was defined based on the country of birth of the participant and of his or her parents as recorded in the municipal register of the city of Amsterdam.13 Specifically, a participant is considered to be of non‐Dutch ethnic origin if he or she fulfils either of these criteria: (1) he or she was born abroad and has at least one parent born abroad (first generation); or (2) he or she was born in the Netherlands but both his or her parents were born abroad (second generation).17 The definition therefore accommodates participants of partially mixed ancestry by determining the predominant ethnicity based on the combination of the participant's and his or her parents’ country of birth. The Surinamese group was further classified according to self‐reported ethnic origin into “African,” “South Asian,” or other. Participants were considered of Dutch ethnicity if they and both parents were born in the Netherlands.
A positive family history of CVD was defined as a self‐reported diagnosis of CVD in parents, siblings, or children that occurred at 60 years or younger. This definition is consistent with the one used in the most common risk scores used in clinical practice.18, 19 Information on family history of CVD was considered missing for patients who reported that they did not know their family history of CVD or who did not fill in the corresponding questionnaire item.
Blood pressure was measured as previously described.20 Hypertension was defined as a systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg,1 or the combined report of prior diagnosis of hypertension and current use of antihypertensive agents (Anatomical Therapeutic Chemical [ATC] codes C02, C03, C07, C08, and C09).21
Waist circumference was measured to the nearest 0.1 cm at the level midway between the lower rib margin and the iliac crest. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m2).
Fasting blood samples were drawn and plasma samples were used to determine the concentration of glucose, low‐density lipoprotein and high‐density lipoprotein (HDL) cholesterol, and triglycerides (Roche Diagnostics, Tokyo, Japan). Diabetes mellitus was defined as fasting plasma glucose ≥7 mmol/L or reported use of glucose‐lowering medications (ACT code A10). Dyslipidemia was defined as either reported use of lipid‐lowering agents (ATC code C10) or cholesterol/HDL ratio ≥5.
Educational level was used as an indicator of socioeconomic position, as it has been found to be an indicator of the socioeconomic position most consistently associated with cardiovascular risk across ethnicities in a similar multiethnic sample from a European city.22 It was defined as the highest qualification obtained in the Netherlands or in the country of origin. The categories of educational level were: (1) never been to school or elementary schooling only; (2) lower vocational schooling or lower secondary schooling; (3) intermediate vocational schooling or intermediate/higher secondary schooling; and (4) higher vocational schooling or university. These categories were combined into low education (1st and 2nd category combined) and high education (3rd and 4th category combined).
Individuals were considered smokers if they reported that they were currently smoking or when they had stopped smoking less than 5 years ago.23
Cardiovascular outcomes were based on questionnaire data. Nonstroke CVD was defined as self‐reported myocardial infarction or self‐reported angioplasty or bypass surgery in the coronary circulation or in the lower extremities. Stroke was defined as self‐reported stroke, cerebral infarction, or cerebral hemorrhage.
Statistical Analysis
In the whole HELIUS sample, 3359 participants (15.5%) had missing data for at least one of the variables of interest in this analysis. For each variable, no more than 1.1% of observations were missing, with the only exception of family history of CVD, which was missing in 12.1% of the sample and, in particular, 19.8% of the Ghanaian participants. Before defining our variables and conducting the analysis, under the assumption that data were missing at random, we performed multiple imputation using the chained equations algorithm and generated 10 imputed data sets. Analysis results were pooled across all data sets using Rubin's rules.24 We performed a sensitivity analysis by repeating all analyses with complete case analysis (N=18 032, of which 5202 had hypertension) comparing the effect sizes thus obtained with those obtained from the multiply imputed datasets and found that the results were similar.
We used logistic regression to examine the association of a positive family history of CVD with either nonstroke CVD or stroke as outcomes. First, we analyzed this association in the whole sample without considering ethnicity. Second, we analyzed the relationship separately for each ethnic group by calculating the significance level of the interaction between positive family history of CVD and ethnicity, using Dutch as a reference group, and showing the ethnicity‐specific estimate of the association of family history of CVD with nonstroke CVD and stroke. The Dutch were chosen as a reference because they are the majority ethnic group in the Netherlands, and most non‐Dutch minority groups are known to display a higher risk of CVD than the Dutch.25, 26 The results of logistic regressions are reported as odds ratios (ORs) with 95% CIs. In the regression analysis, we first used a univariate model with only family history as an independent variable (model 1), then adjusted the analysis for age and sex (model 2), and additionally adjusted for smoking status and educational level (model 3). In a fourth model, we additionally adjusted for diabetes mellitus, dyslipidemia, and waist circumference (model 4). Because the covariates added in model 4 are strong determinants of the metabolic syndrome with relevant inheritable components,27 they could reflect at least partly the same inherited predisposition identified by family history of CVD. Therefore, we expected moderate overadjustment to occur in model 4, possibly underestimating the effect size of the association between family history of CVD and the outcome variables.28 We therefore consider model 3 our main analysis model.
Continuous data are reported as mean and SD and categorical data as frequency and proportions. To study differences in baseline characteristics, we used chi‐square test for categorical variables and t test or ANOVA for continuous variables.
Two‐sided P values <.05 were considered significant. Data were imputed and analyzed using R version 3.2.3 (The R Foundation for Statistical Computing, 2015).
Results
Demographic and clinical characteristics of the study population are presented in Table 1, by individuals with or without a positive family history of CVD. The mean age was 52.7 (±9.9) years, and 47.5% were male. There were 7.1% and 3.5% of individuals with a history of nonstroke CVD or stroke, respectively. Moroccans were the least represented ethnic group (10.2% of the sample) and African Surinamese the most represented (27.5% of the sample).
Table 1.
Characteristics of the Hypertensive Study Sample, by Family History of CVD
| Total Sample | Negative Family History | Positive Family History | |
|---|---|---|---|
| N=6467 | n=4461 (69%) | n=2006 (31%) | |
| Male | 3070 (47.5) | 2168 (48.6) | 902 (45)a |
| Age, y | 52.7±9.9 | 52.3±10.1 | 53.6±9.4a |
| Antihypertensive agents | 3166 (49) | 2043 (45.8) | 1124 (56)a |
| Systolic BP, mm Hg | 142.5±17.5 | 143.1±17.4 | 141.5±17.6a |
| Diastolic BP, mm Hg | 86.0±10.3 | 86.3±10.3 | 85.3±10.2a |
| Without antihypertensive agents | |||
| Systolic BP, mm Hg | 149.3±13.9 | 149.6±14.1 | 148.5±13.3a |
| Diastolic BP, mm Hg | 92.2±8.8 | 92.3±8.9 | 91.8±8.5a |
| BMI, kg/m2 | 29.5±5.5 | 29.6±5.4 | 29.5±5.6 |
| Waist circumference, cm | 99.6±12.8 | 99.5±12.6 | 99.7±13.2 |
| Diabetes mellitus | 1287 (19.9) | 844 (18.9) | 443 (22.1)a |
| Dyslipidemia | 2356 (36.4) | 1441 (32.3) | 915 (45.6)a |
| Lipid‐lowering agents | 1419 (21.9) | 812 (18.2) | 606 (30.2)a |
| HDL, mmol/L | 1.31±0.39 | 1.33±0.40 | 1.27±0.38a |
| Triglycerides, mmol/L | 1.33±0.85 | 1.27±0.83 | 1.40±0.87a |
| Without lipid‐lowering agents | |||
| HDL, mmol/L | 1.45±0.44 | 1.47±0.45 | 1.40±0.42a |
| Triglycerides, mmol/L | 1.14±0.77 | 1.10±0.76 | 1.23±0.78a |
| Smokers | 1655 (25.6) | 1056 (23.7) | 600 (29.9)a |
| Educational level (high) | 2698 (41.7) | 1769 (39.6) | 929 (46.3)a |
| Ethnicity | |||
| Dutch | 1026 (15.9) | 633 (14.2) | 393 (19.6)a |
| South‐Asian Surinamese | 1033 (16) | 480 (10.8) | 553 (27.6)a |
| African Surinamese | 1776 (27.5) | 1277 (28.6) | 499 (24.9)a |
| Ghanaian | 1180 (18.2) | 1101 (24.7) | 79 (3.9)a |
| Turkish | 792 (12.3) | 437 (9.8) | 355 (17.7)a |
| Moroccan | 660 (10.2) | 533 (11.9) | 127 (6.3)a |
| Nonstroke CVD | 456 (7.1) | 243 (5.4) | 213 (10.6)a |
| Stroke | 227 (3.5) | 127 (2.8) | 100 (5)a |
Data are the average of 10 multiply imputed datasets. Continuous data are presented as mean±SD and categorical data are presented as number (percentage). BMI indicates body mass index; BP, blood pressure; CVD, cardiovascular disease; HDL, high‐density lipoprotein.
P<0.05 when compared with a negative family history.
Of all individuals, 31% reported a positive family history of CVD. Individuals with a positive family history of CVD had a higher prevalence of almost all traditional risk factors: they were more likely to be treated for hypertension, to have diabetes mellitus or dyslipidemia, and to be smokers, as compared with individuals with a negative family history of CVD. However, the two groups did not differ in BMI or waist circumference. Individuals with a positive family history of CVD were more often highly educated and of Dutch, South‐Asian Surinamese, and Turkish origin and less often of African Surinamese, Ghanaian, and Moroccan origin as compared with individuals with a negative family history of CVD.
The results of the logistic regression to estimate the association of a positive family history of CVD with the risk of either nonstroke CVD or stroke are shown in Table 2. We found that, after adjustment for age, sex, smoking status, and educational level (model 3), individuals with a positive family history of CVD had a significantly higher prevalence of both nonstroke CVD and stroke (OR, 2.05; 95% CI, 1.65–2.54 [P<0.05] and OR, 1.62; 95% CI, 1.19–2.20 [P<0.05], respectively). Further adjustment by diabetes mellitus, dyslipidemia, and waist circumference in model 4 did not significantly attenuate the association (Table 2).
Table 2.
Association of Positive Family History of CVD With Either Nonstroke CVD or Stroke Among Hypertensive Patients
| Nonstroke CVD | |
| Events, No. (%) | 456 (7.1) |
| Positive family history of CVD, OR (95% CI) | |
| Model 1 | 2.07 (1.67–2.55)a |
| Model 2 | 2.02 (1.63–2.50)a |
| Model 3 | 2.05 (1.65–2.54)a |
| Model 4 | 1.75 (1.41–2.18)a |
| Stroke | |
| Events, No. (%) | 227 (3.5) |
| Positive family history of CVD, OR (95% CI) | |
| Model 1 | 1.78 (1.31–2.41)a |
| Model 2 | 1.70 (1.25–2.30)a |
| Model 3 | 1.62 (1.19–2.20)a |
| Model 4 | 1.38 (1.01–1.88)a |
Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: model 2 + smoking status and educational level. Model 4: model 3 + diabetes mellitus, dyslipidemia, and waist circumference. CVD indicates cardiovascular disease; OR, odds ratio.
P<0.05.
In the second part of our analysis, we analyzed the association of a positive family history of CVD with nonstroke CVD or stroke, stratified by ethnicity.
The demographic and clinical characteristics across ethnic backgrounds are shown in Table 3. Individuals with Ghanaian background had the lowest rate of positive family history of CVD (6.7%) and those with South‐Asian Surinamese background had the highest (53.5%).
Table 3.
Characteristics of the Hypertensive Study Sample, Stratified by Ethnicity
| N=6467 | Dutch | South‐Asian Surinamese | African Surinamese | Ghanaians | Turks | Moroccans |
|---|---|---|---|---|---|---|
| n=1026 | n=1033 | n=1776 | n=1180 | n=792 | n=660 | |
| Male | 612 (59.6) | 505 (48.9) | 694 (39.1) | 507 (43) | 430 (54.2) | 322 (48.8) |
| Age, y | 56.1±10.8 | 54.2±9.5 | 53.9±9.4 | 49.8±8.0 | 49.2±9.9 | 51.3±10.6 |
| Antihypertensive agents | 424 (41.4) | 565 (54.7) | 962 (54.2) | 611 (51.8) | 350 (44.2) | 253 (38.3) |
| Systolic BP, mm Hg | 138.4±16.1 | 144.0±18.1 | 142.8±17.2 | 145.1±17.7 | 140.8±17.3 | 141.5±17.2 |
| Diastolic BP, mm Hg | 83.6±9.8 | 85.3±9.6 | 86.6±9.9 | 89.4±10.6 | 84.6±10.4 | 82.5±9.9 |
| Without antihypertensive agents | ||||||
| Systolic BP, mm Hg | 149.5±13.8 | 149.1±14.3 | 150.0±14.6 | 151.9±14.7 | 145.4±11.6 | 148.1±11.9 |
| Diastolic BP, mm Hg | 91.5±8.9 | 92.1±8.8 | 93.2±8.6 | 94.0±9.2 | 92.0±7.8 | 88.8±8.4 |
| BMI, kg/m2 | 27.8±4.8 | 28.1±4.7 | 29.6±5.8 | 29.6±4.9 | 32.1±5.8 | 31.0±5.2 |
| Waist circumference, cm | 99.5±13.1 | 97.8±11.4 | 98.4±13.9 | 97.4±11.2 | 104.6±12.8 | 103.4±11.7 |
| Diabetes mellitus | 96 (9.3) | 350 (33.9) | 323 (18.2) | 165 (14) | 169 (21.3) | 184 (28) |
| Dyslipidemia | 408 (39.8) | 579 (56) | 459 (25.9) | 223 (18.9) | 413 (52.1) | 272 (41.2) |
| Lipid‐lowering agents | 210 (20.4) | 407 (39.4) | 307 (17.3) | 155 (13.1) | 192 (24.3) | 148 (22.5) |
| HDL, mmol/L | 1.32±0.38 | 1.23±0.36 | 1.42±0.42 | 1.56±0.43 | 1.16±0.31 | 1.21±0.31 |
| Triglycerides, mmol/L | 1.45±0.71 | 1.47±1.01 | 1.03±0.53 | 0.84±0.44 | 1.65±0.94 | 1.46±0.88 |
| Without lipid‐lowering agents | ||||||
| HDL, mmol/L | 1.49±0.44 | 1.29±0.36 | 1.52±0.44 | 1.63±0.46 | 1.21±0.33 | 1.27±0.37 |
| Triglycerides, mmol/L | 1.31±0.78 | 1.35±0.94 | 0.95±0.61 | 0.82±0.52 | 1.57±0.87 | 1.27±0.80 |
| Smokers | 322 (31.4) | 294 (28.4) | 592 (33.3) | 84 (7.2) | 268 (33.8) | 95 (14.4) |
| Educational level (high) | 698 (68) | 397 (38.5) | 905 (51) | 310 (26.3) | 189 (23.8) | 199 (30.1) |
| Positive family history of CVD | 393 (38.3) | 553 (53.5) | 499 (28.1) | 79 (6.7) | 355 (44.8) | 127 (19.2) |
| Nonstroke CVD | 65 (6.3) | 122 (11.8) | 80 (4.5) | 62 (5.2) | 93 (11.7) | 35 (5.3) |
| Stroke | 36 (3.5) | 55 (5.4) | 76 (4.3) | 25 (2.2) | 21 (2.7) | 13 (1.9) |
Data are the average of 10 multiply imputed datasets. Continuous data are presented as column mean±SD and categorical data as number (percentage). BMI indicates body mass index; BP, blood pressure; CVD, cardiovascular disease; HDL, high‐density lipoprotein.
In every ethnic group, the distribution of vascular risk factors between individuals with positive and negative family history of CVD reflected that observed in the total sample. The only exceptions were HDL levels in individuals untreated with lipid‐lowering medications, which among the Turks were lower among those with negative family history; and triglyceride levels in untreated individuals, which among the African Surinamese were higher among those with negative family history (data not shown).
The ethnicity‐specific associations of family history with nonstroke CVD and with stroke are shown in Table 4.
Table 4.
Ethnicity‐Specific Association of Positive Family History of CVD With Either Nonstroke CVD or Stroke Among Hypertensive Patients
| Dutch n=1026 | South‐Asian Surinamese n=1033 | African Surinamese n=1776 | Ghanaians n=1180 | Turks n=792 | Moroccans n=660 | |
|---|---|---|---|---|---|---|
| Nonstroke CVD | ||||||
| Events, No. (%) | 65 (6.3) | 122 (11.8) | 80 (4.5) | 62 (5.2) | 93 (11.7) | 35 (5.3) |
| Positive family history of CVD, OR (95% CI) | ||||||
| Model 1 | 2.65 (1.47–4.79)a | 1.42 (0.94–2.16) | 1.93 (1.19–3.14)a | 1.45 (0.51–4.10) | 1.73 (1.04–2.86)a | 1.09 (0.42–2.83) |
| Model 2 | 2.61 (1.45–4.69)a | 1.53 (1.00–2.36)a | 2.02 (1.24–3.28)a | 1.42 (0.50–4.05) | 1.94 (1.17–3.22)a | 1.09 (0.41–2.89) |
| Model 3 | 2.47 (1.37–4.44)a | 1.60 (1.03–2.47)a | 2.10 (1.29–3.42)a | 1.54 (0.54–4.41) | 1.86 (1.12–3.08)a | 1.16 (0.44–3.09) |
| Model 4 | 2.24 (1.24–4.05)a | 1.42 (0.91–2.22) | 1.94 (1.17–3.22)a | 1.33 (0.45–3.93) | 1.87 (1.11–3.14)a | 1.28 (0.48–3.42) |
| Stroke | ||||||
| Events, No. (%) | 36 (3.5) | 55 (5.4) | 76 (4.3) | 25 (2.2) | 21 (2.7) | 13 (1.9) |
| Positive family history of CVD, OR (95% CI) | ||||||
| Model 1 | 1.01 (0.47–2.13) | 1.73 (0.95–3.15) | 2.08 (1.27–3.41)a | 1.60 (0.33–7.74) | 1.21 (0.49–3.01) | 0.97 (0.16–5.86) |
| Model 2 | 0.95 (0.45–2.02) | 1.77 (0.97–3.24) | 2.08 (1.27–3.42)a | 1.57 (0.33–7.60) | 1.24 (0.50–3.08) | 0.96 (0.16–5.76) |
| Model 3 | 0.88 (0.41–1.86) | 1.80 (0.98–3.32) | 2.17 (1.32–3.57)a | 1.63 (0.33–7.98) | 1.14 (0.46–2.84) | 1.00 (0.17–6.01) |
| Model 4 | 0.80 (0.38–1.71) | 1.56 (0.84–2.88) | 2.02 (1.21–3.36)a | 1.46 (0.29–7.31) | 1.15 (0.46–2.87) | 1.11 (0.18–6.75) |
Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: model 2 + smoking status and educational level. Model 4: model 3 + diabetes mellitus, dyslipidemia, and waist circumference. CVD indicates cardiovascular disease; OR, odds ratio. Interaction family history of CVD—ethnicity for nonstroke CVD: not statistically significant. Interaction family history of CVD—ethnicity for stroke: statistically significant for African Surinamese (P<0.05) in models 3 and 4.
P<0.05.
With regard to the first outcome, a positive family history of CVD was associated with a higher prevalence of nonstroke CVD in most ethnicities (model 3 in Table 4). Expressed in OR (95% CI), the point estimates for the association ranged from 1.16 (0.44–3.09) among the Moroccans to 2.47 (1.37–4.44) among the Dutch. Further adjustment for diabetes mellitus, dyslipidemia, and waist circumference in model 4 did not significantly alter the effect sizes or interethnic differences.
With regard to the second outcome, a positive family history of CVD was also associated with a higher prevalence of stroke in most ethnicities (model 3, Table 4). The OR (95% CI) for the association ranged from 0.88 (0.41–1.86) among the Dutch to 2.17 (1.32–3.57) among the African Surinamese. Further adjustment in model 4 did not significantly alter the effect sizes or interethnic differences.
To evaluate whether the differences between the ethnic groups in the association between family history of CVD and both outcomes were significant, the regression model included the interaction between family history and ethnicity. For the outcome nonstroke CVD, the interaction was not statistically significant. For the outcome stroke, the interaction was statistically significant in models 3 and 4, with the association between family history and stroke being stronger in the African Surinamese hypertensive participants than in the other ethnic groups (P<0.05). To further investigate this ethnic group, we used an interaction model to test for sex differences in this association among the African Surinamese (Table 5). While the interaction was not statistically significant, the sex‐specific ORs were higher in women (OR, 3.12 [95% CI, 1.61–5.64] for model 3) than in men (OR, 1.15 [95% CI, 0.43–3.06]).
Table 5.
Sex‐Specific Association of Positive Family History of CVD With Stroke Among Hypertensive Patients of African Surinamese Ethnicity
| African Surinamese Men (n=694) | African Surinamese Women (n=1082) | |
|---|---|---|
| Stroke | ||
| Events, No. (%) | 29 (4.2) | 47 (4.3) |
| Positive family history of CVD, OR (95% CI) | ||
| Model 1 | 1.12 (0.42–3.01) | 2.92 (1.59–5.36) |
| Model 2 | 1.15 (0.43–3.08) | 2.89 (1.57–5.32)a |
| Model 3 | 1.15 (0.43–3.06) | 3.12 (1.68–5.80)a |
| Model 4 | 0.96 (0.34–2.66) | 3.06 (1.59–5.87)a |
Model 1: unadjusted. Model 2: adjusted for age and sex. Model 3: model 2 + smoking status and educational level. Model 4: model 3 + diabetes mellitus, dyslipidemia, and waist circumference. Interaction family history—sex: not statistically significant. CVD indicates cardiovascular disease; OR, odds ratio.
P<0.05.
Discussion
Our findings support the hypothesis that a positive family history of CVD independently identifies subgroups at higher risk for nonstroke CVD among patients with hypertension, and show that it specifically identifies hypertensive individuals at higher risk of stroke in women of the African Surinamese ethnic group.
Vascular disease among hypertensive individuals with a positive family history of CVD was found nearly twice as often than among hypertensive individuals with a negative family history, even after adjustment for notable environmental risk factors.
In the analysis on the sample stratified by ethnic groups, effect sizes in the fully adjusted model displayed large variations across ethnic groups. With regard to nonstroke CVD, the effect size of a positive family history of CVD was most pronounced for the Dutch and absent in the Moroccans. The other ethnicities were in between. Since the interaction term between family history of CVD and ethnicity for this outcome was not statistically significant, we cannot conclude that the Dutch had a higher risk for nonstroke CVD as compared with the other ethnic groups. With regard to stroke, the African Surinamese displayed the strongest association: in contrast to the other ethnic groups, African Surinamese hypertensive individuals with a positive family history of CVD are approximately twice as likely to report stroke as those with a negative family history of CVD. The interaction between family history of CVD and ethnicity for this outcome was statistically significant. While there was no statistically significant interaction between sex and family history of CVD in this ethnic group, the sex‐specific effect sizes revealed that this result seems to be driven by individuals of female sex.
The absence of a statistically significant interaction between family history of CVD and ethnicity for the outcome nonstroke CVD, despite considerable differences in the ethnicity‐specific ORs, could be due to limited power, since our cohort was rather young and had only few events per ethnic group. Indeed, a post hoc power calculation showed that, for the outcome stroke, the power for the ORs of family history of CVD observed in Dutch and African Surinamese in model 3 to be significantly different was about 60%, while, for the outcome nonstroke CVD and the ORs observed in Dutch and Moroccans, the power is only about 40%.
The differences in the effect size of the relationship between family history of CVD and the two outcomes across the 4 models are small. It can be noted that adjustment by age does not lead here to a strong reduction in the effect size of family history of CVD because our sample was relatively young, and the independent association of age with vascular events was therefore less evident.
The difference across ethnicities in the relationship between family history and future CVD might be explained by an underlying genetic predisposition, which is different for each ethnicity. It is well known that individuals of African background often have more severe organ damage from hypertension as compared with individuals of Caucasian background.16, 29, 30 A positive family history also reflects a genetic background responsible for CVD, and because positive family history is associated with CVD even after adjustment for environmental confounders, this genetic background seems likely to be responsible for the increased risk for CVD. By dividing our cohort by ethnic background and positive family history, the genetic susceptibility becomes more pronounced, resulting in the identification of hypertensive individuals at particular risk because of their genetic background.
Study Limitations
Our study has some limitations. First, in studies of an associative nature it is impossible to distinguish cause from consequence. Nevertheless, a genetic predisposition can be assumed to be present from birth. Accordingly, if a positive family history is associated with vascular disease, this could reflect a specific genetic background. While lifestyle factors shared in families could also play a role, they are still less likely than genetic factors to be shared by first‐degree relatives.31
Second, self‐reported family history of CVD may be inaccurate. Nevertheless, it has good discriminatory power: validation studies of the accuracy of offspring‐reported parental events found likelihood ratios >8.0 for parental heart attack and stroke in medical records32 and >10.0 for stroke self‐reported by the parents themselves.33 Moreover, we were able to gather specific information on which event had occurred, in which family member, and at what age; and questionnaires in HELIUS were completed with the help of a trained interviewer in a considerable proportion of the minority ethnicities, ranging from 27% of the Surinamese to 40% of the Ghanaians. Nevertheless, the high number of missing values among the Ghanaians might also reflect greater inaccuracy among this population. However, individuals who did not recall their family history might have had a positive family history, so that the number of individuals with a positive family history was probably underestimated. In contrast, it is less likely that cases of positive family history are misclassified, because of the specific nature of the questions asked. We do not believe that our variable on this issue is less reliable than any other similar study.
Third, the HELIUS cohort was relatively young, including individuals aged 18 to 70 years. In our selection of hypertensive participants, the average age was 52.7 years. Since several participants who will develop CVD may not yet have developed it at the time of the inclusion, early‐onset CVD might be overrepresented as compared with older‐age CVD. Since early‐onset CVD is more strongly associated with genetic predispositions than late‐onset CVD,6 we may have overestimated the impact of the genetic predisposition reflected by a positive family history of CVD.
Fourth, ethnic minority groups in the Netherlands have a higher case fatality rate than those of Dutch origin for both stroke34 and nonstroke CVD.35 If fatal cases among hypertensive individuals of minority ethnic groups were associated with positive family history of CVD, the absence of those cases from the sample may have introduced sampling bias. Interethnic differences may therefore be different if those cases could be included.
Fifth, the HELIUS study was conducted in Amsterdam, the Netherlands, and only included the most common ethnic groups of this city.13 While these ethnic groups are also important minority groups in several other European countries,13 they do not include other ethnic groups important on a global level, such as those from Southern America and Eastern Asia.
Conclusions
A positive family history of CVD in African Surinamese hypertensive individuals is associated with a higher risk of stroke. Therefore, clinicians working with multiethnic hypertensive populations may consider preventive measures or more intense screening in individuals of this ethnicity with a positive family history of CVD if a predictive value of family history for nonstroke and stroke CVD in similar populations and ethnicities is confirmed by longitudinal studies.
Sources of Funding
The HELIUS study was funded by the Academic Medical Center Amsterdam, the Public Health Service of Amsterdam, the Dutch Heart Foundation, the Netherlands Organization for Health Research and Development (ZonMw), and the European Union (FP‐7).
Disclosures
None.
Acknowledgments
The HELIUS study is conducted by the Academic Medical Center of Amsterdam and the Public Health Service of Amsterdam. Both organizations provided core support for HELIUS. We gratefully acknowledge the AMC Biobank for their support in biobank management and high‐quality storage of collected samples. We are most grateful to the participants of the HELIUS study and the management team, research nurses, interviewers, research assistants, and other staff who have taken part in gathering the data of this study.
(J Am Heart Assoc. 2016;5:e004260 doi: 10.1161/JAHA.116.004260)
References
- 1. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Bohm M, Christiaens T, Cifkova R, De Backer G, Dominiczak A, Galderisi M, Grobbee DE, Jaarsma T, Kirchhof P, Kjeldsen SE, Laurent S, Manolis AJ, Nilsson PM, Ruilope LM, Schmieder RE, Sirnes PA, Sleight P, Viigimaa M, Waeber B, Zannad F. 2013 ESH/ESC practice guidelines for the management of arterial hypertension. Blood Press. 2014;23:3–16. [DOI] [PubMed] [Google Scholar]
- 2. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair‐Rohani H, Amann M, Anderson HR, Andrews KG, Aryee M, Atkinson C, Bacchus LJ, Bahalim AN, Balakrishnan K, Balmes J, Barker‐Collo S, Baxter A, Bell ML, Blore JD, Blyth F, Bonner C, Borges G, Bourne R, Boussinesq M, Brauer M, Brooks P, Bruce NG, Brunekreef B, Bryan‐Hancock C, Bucello C, Buchbinder R, Bull F, Burnett RT, Byers TE, Calabria B, Carapetis J, Carnahan E, Chafe Z, Charlson F, Chen H, Chen JS, Cheng AT, Child JC, Cohen A, Colson KE, Cowie BC, Darby S, Darling S, Davis A, Degenhardt L, Dentener F, Des Jarlais DC, Devries K, Dherani M, Ding EL, Dorsey ER, Driscoll T, Edmond K, Ali SE, Engell RE, Erwin PJ, Fahimi S, Falder G, Farzadfar F, Ferrari A, Finucane MM, Flaxman S, Fowkes FG, Freedman G, Freeman MK, Gakidou E, Ghosh S, Giovannucci E, Gmel G, Graham K, Grainger R, Grant B, Gunnell D, Gutierrez HR, Hall W, Hoek HW, Hogan A, Hosgood HD III, Hoy D, Hu H, Hubbell BJ, Hutchings SJ, Ibeanusi SE, Jacklyn GL, Jasrasaria R, Jonas JB, Kan H, Kanis JA, Kassebaum N, Kawakami N, Khang YH, Khatibzadeh S, Khoo JP, Kok C, Laden F, Lalloo R, Lan Q, Lathlean T, Leasher JL, Leigh J, Li Y, Lin JK, Lipshultz SE, London S, Lozano R, Lu Y, Mak J, Malekzadeh R, Mallinger L, Marcenes W, March L, Marks R, Martin R, McGale P, McGrath J, Mehta S, Mensah GA, Merriman TR, Micha R, Michaud C, Mishra V, Mohd Hanafiah K, Mokdad AA, Morawska L, Mozaffarian D, Murphy T, Naghavi M, Neal B, Nelson PK, Nolla JM, Norman R, Olives C, Omer SB, Orchard J, Osborne R, Ostro B, Page A, Pandey KD, Parry CD, Passmore E, Patra J, Pearce N, Pelizzari PM, Petzold M, Phillips MR, Pope D, Pope CA III, Powles J, Rao M, Razavi H, Rehfuess EA, Rehm JT, Ritz B, Rivara FP, Roberts T, Robinson C, Rodriguez‐Portales JA, Romieu I, Room R, Rosenfeld LC, Roy A, Rushton L, Salomon JA, Sampson U, Sanchez‐Riera L, Sanman E, Sapkota A, Seedat S, Shi P, Shield K, Shivakoti R, Singh GM, Sleet DA, Smith E, Smith KR, Stapelberg NJ, Steenland K, Stockl H, Stovner LJ, Straif K, Straney L, Thurston GD, Tran JH, Van Dingenen R, van Donkelaar A, Veerman JL, Vijayakumar L, Weintraub R, Weissman MM, White RA, Whiteford H, Wiersma ST, Wilkinson JD, Williams HC, Williams W, Wilson N, Woolf AD, Yip P, Zielinski JM, Lopez AD, Murray CJ, Ezzati M, AlMazroa MA, Memish ZA. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet. 2012;380:2224–2260. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Wright JM, Musini VM. First‐line drugs for hypertension. Cochrane Database Syst Rev. 2009;(3):CD001841. [DOI] [PubMed] [Google Scholar]
- 4. Organisation for Economic Co‐operation and Development (OECD) . Life expectancy at 65 (indicator). 2016. doi:10.1787/0e9a3f00‐en. Accessed November 22, 2016.
- 5. Andresdottir MB, Sigurdsson G, Sigvaldason H, Gudnason V. Fifteen percent of myocardial infarctions and coronary revascularizations explained by family history unrelated to conventional risk factors. The Reykjavik Cohort Study. Eur Heart J. 2002;23:1655–1663. [DOI] [PubMed] [Google Scholar]
- 6. Prabhakaran D, Jeemon P. Should your family history of coronary heart disease scare you? Mt Sinai J Med. 2012;79:721–732. [DOI] [PubMed] [Google Scholar]
- 7. Kubota M, Yamaura A, Ono J, Itani T, Tachi N, Ueda K, Nagata I, Sugimoto S. Is family history an independent risk factor for stroke? J Neurol Neurosurg Psychiatry. 1997;62:66–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Mvundura M, McGruder H, Khoury MJ, Valdez R, Yoon PW. Family history as a risk factor for early‐onset stroke/transient ischemic attack among adults in the United States. Public Health Genomics. 2010;13:13–20. [DOI] [PubMed] [Google Scholar]
- 9. Seshadri S, Beiser A, Pikula A, Himali JJ, Kelly‐Hayes M, Debette S, DeStefano AL, Romero JR, Kase CS, Wolf PA. Parental occurrence of stroke and risk of stroke in their children: the Framingham study. Circulation. 2010;121:1304–1312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Williamson C, Jeemon P, Hastie CE, McCallum L, Muir S, Dawson J, Walters M, Sloan W, Morrison D, Dominiczak AF, Pell J, Padmanabhan S. Family history of premature cardiovascular disease: blood pressure control and long‐term mortality outcomes in hypertensive patients. Eur Heart J. 2014;35:563–570. [DOI] [PubMed] [Google Scholar]
- 11. Hamer M, Chida Y, Stamatakis E. The role of conventional and novel mechanisms in explaining increased risk of cardiovascular events in offspring with positive parental history. J Hypertens. 2009;27:1966–1971. [DOI] [PubMed] [Google Scholar]
- 12. Neutel JM. Why lowering blood pressure is not enough: the hypertension syndrome and the clinical context of cardiovascular risk reduction. Heart Dis. 2000;2:370–374. [PubMed] [Google Scholar]
- 13. Stronks K, Snijder MB, Peters RJ, Prins M, Schene AH, Zwinderman AH. Unravelling the impact of ethnicity on health in Europe: the HELIUS study. BMC Public Health. 2013;13:402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. van Oeffelen AA, Agyemang C, Stronks K, Bots ML, Vaartjes I. Incidence of first acute myocardial infarction over time specific for age, sex, and country of birth. Neth J Med. 2014;72:20–27. [PubMed] [Google Scholar]
- 15. Chaturvedi N, Bulpitt CJ, Leggetter S, Schiff R, Nihoyannopoulos P, Strain WD, Shore AC, Rajkumar C. Ethnic differences in vascular stiffness and relations to hypertensive target organ damage. J Hypertens. 2004;22:1731–1737. [DOI] [PubMed] [Google Scholar]
- 16. Okin PM, Kjeldsen SE, Julius S, Dahlof B, Devereux RB. Racial differences in sudden cardiac death among hypertensive patients during antihypertensive therapy: the LIFE study. Heart Rhythm. 2012;9:531–537. [DOI] [PubMed] [Google Scholar]
- 17. Stronks K, Kulu‐Glasgow I, Agyemang C. The utility of ‘country of birth’ for the classification of ethnic groups in health research: the Dutch experience. Ethn Health. 2009;14:255–269. [DOI] [PubMed] [Google Scholar]
- 18. Assmann G, Cullen P, Schulte H. Simple scoring scheme for calculating the risk of acute coronary events based on the 10‐year follow‐up of the prospective cardiovascular Munster (PROCAM) study. Circulation. 2002;105:310–315. [DOI] [PubMed] [Google Scholar]
- 19. Hippisley‐Cox J, Coupland C, Vinogradova Y, Robson J, Minhas R, Sheikh A, Brindle P. Predicting cardiovascular risk in England and Wales: prospective derivation and validation of QRISK2. BMJ. 2008;336:1475–1482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Agyemang C, Kieft S, Snijder MB, Beune EJ, van den Born BJ, Brewster LM, Ujcic‐Voortman JJ, Bindraban N, van Montfrans G, Peters RJ, Stronks K. Hypertension control in a large multi‐ethnic cohort in Amsterdam, the Netherlands: the HELIUS study. Int J Cardiol. 2015;183C:180–189. [DOI] [PubMed] [Google Scholar]
- 21. Guidelines for ATC classification and DDD assignment 2013. 2012.
- 22. Fischbacher CM, Cezard G, Bhopal RS, Pearce J, Bansal N. Measures of socioeconomic position are not consistently associated with ethnic differences in cardiovascular disease in Scotland: methods from the Scottish Health and Ethnicity Linkage Study (SHELS). Int J Epidemiol. 2014;43:129–139. [DOI] [PubMed] [Google Scholar]
- 23. Mannan H, Stevenson C, Peeters A, Walls H, McNeil J. Framingham risk prediction equations for incidence of cardiovascular disease using detailed measures for smoking. Heart Int. 2010;5:e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Janssen KJ, Donders AR, Harrell FE Jr, Vergouwe Y, Chen Q, Grobbee DE, Moons KG. Missing covariate data in medical research: to impute is better than to ignore. J Clin Epidemiol. 2010;63:721–727. [DOI] [PubMed] [Google Scholar]
- 25. Agyemang C, van Oeffelen AA, Norredam M, Kappelle LJ, Klijn CJ, Bots ML, Stronks K, Vaartjes I. Ethnic disparities in ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage incidence in the Netherlands. Stroke. 2014;45:3236–3242. [DOI] [PubMed] [Google Scholar]
- 26. Bos V, Kunst AE, Keij‐Deerenberg IM, Garssen J, Mackenbach JP. Ethnic inequalities in age‐ and cause‐specific mortality in the Netherlands. Int J Epidemiol. 2004;33:1112–1119. [DOI] [PubMed] [Google Scholar]
- 27. Pollex RL, Hegele RA. Genetic determinants of the metabolic syndrome. Nat Clin Pract Cardiovasc Med. 2006;3:482–489. [DOI] [PubMed] [Google Scholar]
- 28. Schisterman EF, Cole SR, Platt RW. Overadjustment bias and unnecessary adjustment in epidemiologic studies. Epidemiology. 2009;20:488–495. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Agyemang C, Addo J, Bhopal R, Aikins Ade G, Stronks K. Cardiovascular disease, diabetes and established risk factors among populations of sub‐Saharan African descent in Europe: a literature review. Global Health. 2009;5:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Lackland DT. Racial differences in hypertension: implications for high blood pressure management. Am J Med Sci. 2014;348:135–138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Elis A, Pereg D, Tirosh A, Shochat T, Tekes‐Manova D, Lishner M. Family history of cardiovascular disease does not predict risk‐reducing behavior. Eur J Cardiovasc Prev Rehabil. 2008;15:325–328. [DOI] [PubMed] [Google Scholar]
- 32. Murabito JM, Nam BH, D'Agostino RB Sr, Lloyd‐Jones DM, O'Donnell CJ, Wilson PW. Accuracy of offspring reports of parental cardiovascular disease history: the Framingham Offspring Study. Ann Intern Med. 2004;140:434–440. [DOI] [PubMed] [Google Scholar]
- 33. Oygarden H, Fromm A, Sand KM, Eide GE, Thomassen L, Naess H, Waje‐Andreassen U. Can the cardiovascular family history reported by our patients be trusted? The Norwegian Stroke in the Young Study. Eur J Neurol. 2016;23:154–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Agyemang C, Vaartjes I, Bots ML, van Valkengoed IG, de Munter JS, de Bruin A, Berger‐van Sijl M, Reitsma JB, Stronks K. Risk of death after first admission for cardiovascular diseases by country of birth in the Netherlands: a nationwide record‐linked retrospective cohort study. Heart. 2009;95:747–753. [DOI] [PubMed] [Google Scholar]
- 35. van Oeffelen AA, Agyemang C, Stronks K, Bots ML, Vaartjes I. Prognosis after a first hospitalisation for acute myocardial infarction and congestive heart failure by country of birth. Heart. 2014;100:1436–1443. [DOI] [PubMed] [Google Scholar]