Article Text

Download PDFPDF

Original article
Sex differences in prodromal symptoms in acute coronary syndrome in patients aged 55 years or younger
  1. Nadia A Khan1,
  2. Stella S Daskalopoulou2,3,
  3. Igor Karp4,
  4. Mark J Eisenberg5,
  5. Roxanne Pelletier2,6,
  6. Meytal Avgil Tsadok2,6,
  7. Kaberi Dasgupta2,3,6,
  8. Colleen M Norris7,
  9. Louise Pilote2,3,6
  10. for the GENESIS PRAXY Team
    1. 1Department of Medicine, Center for Health Evaluation and Outcomes Science, University of British Columbia, Vancouver, British Columbia, Canada
    2. 2Research Institute of the McGill University Health Center, Montreal, Quebec, Canada
    3. 3Division of General Internal Medicine, McGill University Health Center, Montréal, Québec, Canada
    4. 4Department of Social and Preventive Medicine, University of Montréal Hospital Research Center (CRCHUM), University of Montréal, Montréal, Québec, Canada
    5. 5Divisions of Cardiology and Clinical Epidemiology, Jewish General Hospital, McGill University, Montréal, Québec, Canada
    6. 6Division of Clinical Epidemiology, McGill University Health Center, Montréal, Québec, Canada
    7. 7Faculties of Nursing, Medicine and School of Public Health, University of Alberta, Edmonton, Alberta, Canada
    1. Correspondence to Dr Nadia Khan, Internal Medicine, University of British Columbia Faculty of Medicine, 540.7, 1081 Burrard Street, Vancouver, British Columbia, Canada V6Z 1Y6; nakhanubc{at}gmail.com

    Abstract

    Background Studies suggest that young women are at highest risk for failing to recognise early symptoms of acute coronary syndrome (ACS).

    Objectives To examine sex differences in prodromal symptoms occurring days and weeks prior to the acute presentation of ACS. We also examined health-seeking behaviours and prehospital management in young patients.

    Methods Prospective cross-sectional analysis of 1145 patients (368 women) hospitalised for ACS, aged ≤55 years, from the GENdEr and Sex DetermInantS of Cardiovascular Disease: From Bench to Beyond Premature Acute Coronary SYndrome cohort study (January 2009–April 2013). Prodromal symptoms were determined using the McSweeney Acute and Prodromal Myocardial Infarction Symptom questionnaire. Health-seeking behaviour and prehospital care were determined by questionnaires.

    Results The median age was 49 years. The prevalence of prodromal symptoms was high and more women reported symptoms than men (85% vs 72%, p<0.0001). Symptoms were similar between sexes and included unusual fatigue, sleep disturbances, anxiety and arm weakness/discomfort. Chest pain was less common in both sexes (24%). Women were more likely to seek care (49% vs 42%, p=0.04). Among those who sought care, women were more likely to use an ambulance for their ACS compared with men (52% vs 39%). Cardiovascular risk-reduction therapy use was low (≤40%) in all patients and less than half perceived their care provider suspected a cardiac source.

    Conclusions Prior to ACS, women were more likely to experience prodromal symptoms and seek medical attention than men. Prehospital care was generally similar between sexes but demonstrated underutilisation of risk-reduction therapies in at-risk young adults.

    Statistics from Altmetric.com

    Early recognition of acute coronary syndrome (ACS) is critical for timely intervention and improved mortality. Prodromal cardiac symptoms are symptoms occurring in the days and weeks prior to the acute presentation of ACS.1 These symptoms have been shown to predict future cardiovascular events in women.2 As such, there is considerable interest in recognising prodromal symptoms as they may serve as an early warning providing an opportunity to institute prompt diagnostic testing and preventative interventions.

    Most studies detailing prodromal cardiac symptoms and prehospital care were limited to elderly women cohorts.1 However, younger women and men aged 55 years or less, comprise almost a fifth of all ACS cases.3 This younger cohort is at highest risk for failing to recognise ACS symptoms and delayed access to care.4 ,5 Younger women with ACS are more likely to experience non-chest pain cardiac symptoms than men,6 to dismiss their cardiac symptoms to avoid causing a ‘false alarm’,7 to be misdiagnosed from the emergency department,4 ,5 and to have a higher mortality following ACS.8 Younger women also report that healthcare professionals inconsistently respond to prodromal symptoms in women.7 These findings raise concern for poor health-seeking behaviours in young patients and potential sex bias in the assessment and management of prodromal cardiac symptoms. To date, there are sparse data on sex differences in prodromal symptoms, health seeking and prehospital care in a younger cohort. To clarify these issues, we undertook a prospective multinational cohort study to evaluate sex differences in prodromal symptoms including chest pain and non-chest pain symptoms and health-seeking behaviour in women and men aged 55 years or less who had an ACS. We also evaluated sex differences in prehospital management of prodromal symptoms.

    Methods

    Study design/subjects

    The study design is a prospective cross-sectional analysis of participants of the GENESIS PRAXY (GENdEr and Sex DetermInantS of Cardiovascular Disease: From Bench to Beyond Premature Acute Coronary SYndrome) cohort study (January 2009 to April 2013). The details of the GENESIS PRAXY study are published elsewhere.9 Briefly, patients aged 55 years or less, admitted with ACS to urban tertiary care and community hospital coronary care units, intensive care units or cardiology wards were enrolled in 24 sites in Canada, one in the USA and one in Switzerland. ACS diagnosis was determined by the treating physician based on standard criteria for symptoms, signs, ECG findings and elevations in cardiac enzymes.6

    Data collection

    Prodromal symptom presentation

    Prodromal symptom presentation for ACS was determined in all participants during hospitalisation for ACS using the McSweeney Acute and Prodromal Myocardial Infarction Symptom Survey (MAPMISS).10 MAPMISS assesses the presence and intensity of 30 prodromal symptoms preceding the onset of ACS. The intensity (mild, moderate or severe) and frequency (daily, several times per week, at least once per week) of each of the symptoms was also self-reported in MAPMISS. Validation studies for the MAPMISS prodromal symptoms in women demonstrated good content and test-retest reliability (r=0.92).11 Prodromal symptoms were stratified into chest pain prodromal symptoms and non-chest pain prodromal symptoms. Chest pain was defined as generalised chest discomfort or pain, centred high on chest or left breast. In addition, patients were asked about additional symptoms they experienced in open text. However, as these additional symptoms were only rarely noted, we only report the MAPMISS prodromal symptoms here. We defined a prodromal symptom as being present if the intensity was at least moderate and the symptom occurred at least once or more times per week before their ACS.

    Demographic, gender role, clinical and psychosocial characteristics

    Patient age, socioeconomic status as indicated by annual household income (high if >$C50 000), education level (postsecondary education vs less than postsecondary) and ethnicity (white vs other) was determined by self-report. Clinical characteristics and health behaviours were determined from questionnaires and medical chart reviews. Obesity was defined as a body mass index ≥30 kg/m2. Depression and anxiety symptoms preceding the index hospitalisation for ACS were assessed using the Hospital Anxiety and Depression Scale with a score of >8 on each subscale indicating probable depression or anxiety.12

    ACS severity and coronary anatomy

    ACS severity was assessed by ACS type (ST segment elevation myocardial infarction (STEMI) versus non-ST segment elevation myocardial infarction (NSTEMI) or unstable angina), as determined by medical record; peak troponin levels (tertiles above normal value for both troponin T and troponin I assays) obtained from day of admission from the medical record; presence and extent of coronary stenosis determined from angiogram results (obtained in 88% of cohort); and Global Registry of Acute Coronary Events (GRACE) score, a validated score used to predict inhospital and long-term mortality or re-infarction in patients with STEMI and NSTEMI.13

    Health-seeking behaviours and prehospital management

    Whether patients sought medical care prior to emergency department visit for index ACS was determined through the patient self-reported questionnaire. We evaluated arrival by ambulance and outpatient cardiovascular medications use prior to ACS as indices of prehospital management. Outpatient use of ACE inhibitor, angiotensin receptor blockers, β-blockers, antiplatelet agents, lipid-lowering therapies (statins) and nicotine replacement therapies were determined from the hospital medical record. We also explored interventions by care providers (any tests ordered, medication changes, referral to specialist, referral to hospital) and patient perceptions of whether their care provider suspected a heart problem, determined by patient self-reported questionnaires.

    Statistical analysis

    Baseline characteristics, prodromal symptoms, health-seeking behaviours and prehospital management were compared between groups (men vs women; those with and without prodromal symptoms; those with prodromal chest pain symptoms vs non-chest pain symptoms; and those who sought medical care vs those who did not among patients reporting prodromal symptoms) using Fisher's exact test and Student's t-tests for continuous variables, respectively. There were some missing baseline data: postsecondary education (1% men; 1% women), household income 16%, troponin values (2% men; 8% women) and symptoms (0–1.9%). We evaluated risk factors including specific prodromal symptoms that were independently associated with health-seeking behaviour using logistic regression modelling. Logistic regression models to estimate ORs and their 95% CIs were adjusted for age, sex and traditional cardiovascular risk factors including previous myocardial infarction and symptoms. Given the large number of symptoms, symptoms were removed in a backwards-stepwise elimination and models were refit until remaining p values for symptoms held a significance level of 0.10. However, we retained sociodemographic variables and clinical variables in the model that may affect symptom presentation regardless of level of significance (ie, sex, age, household income, traditional cardiac risk factors, previous acute myocardial infarction, depression and anxiety). For the exploratory assessment of medical management that patients reported, some data were missing (referral to hospital 27%, referral to specialist 27%, tests done 12%, no intervention 6%, medication change 27% and physician suspected a medical problem 6%). Missing data for medical management section were due to a clerical error in the first version of the questionnaire. This was corrected once identified in the questionnaire. Missing data values were excluded from the analysis. All analyses were performed using SAS V.9.2 (SAS Institute, Cary, North Carolina, USA).

    Results

    The GENESIS-PRAXY cohort was comprised of 1213 patients of whom 1145 (368 (32%) women) had both complete chart and medical record review available for this analysis (94% of original cohort). The median age of women and men was approximately 49 years and most patients were white. There were 876 patients (76%) who experienced at least one prodromal symptom that was of moderate to severe intensity and the symptom occurred once or more times per week before their ACS. Of these patients, 313 were women.

    Sex differences in prodromal symptoms and patient characteristics

    Women were more likely to report prodromal symptoms compared with men (85% vs 72%, p<0.0001) (table 1) and reported a greater number of symptoms (6 vs 4, p<0.0001). The most common prodromal symptoms were similar in women and men and included feeling unusually tired or fatigued, having difficulty sleeping, anxiety, heart racing, arm weakness, numbness or discomfort, and indigestion. Prodromal chest pain was less common, occurring in approximately 24% of patients and did not differ in frequency by sex (table 2).

    Table 1

    Baseline characteristics according to sex and prodromal symptoms N (%)

    Table 2

    Prodromal cardiac symptoms according to sex, N (%)

    Patients with prodromal symptoms had a high burden of cardiovascular risk factors (table 1). In women, hypertension and dyslipidaemia were more prevalent in those with prodromal symptoms than those without. Anxiety and depression were also more common in men and women with prodromal symptoms than those without symptoms.

    For women, severity of ACS, number of vessels obstructed and location of coronary artery stenosis were similar between those with and without symptoms. However, men with prodromal symptoms were less likely to have a STEMI or higher troponin values than men without these symptoms.

    Health-seeking behaviours and prehospital management

    Overall healthcare seeking for prodromal symptoms was generally low among those with prodromal symptoms. Women were more likely to seek medical care compared with men (49% vs 42%, p=0.04) (table 3). When the prodromal symptoms included chest pain, women and men frequently and similarly sought medical care (81% vs 74%, respectively) (figure 1). However, in the absence of chest pain, few men and women sought care (39% vs 30%, respectively, p=0.02). After adjusting for differences in symptoms and presence of cardiovascular risk factors, there were no sex differences in seeking medical care for prodromal symptoms. Factors associated with seeking care included the presence of chest pain, feeling intensely tired, weakness in arms, increased frequency of headaches and depression (table 4 and figure 2).

    Table 3

    Prehospital health-seeking behaviours in patients with prodromal symptoms and according to chest pain status, N (%)

    Table 4

    Multivariate analysis of seeking care*

    Figure 1

    Prehospital health-seeking behaviours in patients with prodromal symptoms.

    Figure 2

    Multivariate analysis of seeking care.

    Outpatient use of cardiovascular risk-reduction therapies was low in both those who did and did not seek prehospital medical care for prodromal symptoms (table 5). However, women who sought care were more likely to be taking acetyl salicylic acid and men who sought care were more likely to be taking antiplatelet agents, statins and β-blockers compared with their counterparts who did not seek care. There were no sex differences in use of these therapies among those with prodromal symptoms. Women who sought medical care were more likely to arrive to hospital via ambulance than men for ACS.

    Table 5

    Prehospital cardiovascular risk-reduction treatment status on admission among those who sought medical care versus those who did not among patients with at least one prodromal symptom regardless of intensity, N (%)

    In our exploratory analysis, most patients seeking care reported that their physician ordered tests, changed medications, referred them to specialists or to hospital (figure 1). Overall, there were no sex differences in prehospital care interventions undertaken by the care provider or in patient's perception of their physician suspecting a cardiac source. However women that did not have chest pain reported that they felt their physician was significantly less likely to suspect a cardiac problem (35% vs 52%, p=0.02) and more often reported no intervention done (13% vs 3%, p=0.01) compared with men. There were no sex differences in other prehospital interventions among patients without chest pain.

    Discussion

    This study had several key findings. Most young patients with ACS report non-specific prodromal symptoms and women report these symptoms slightly more often than men. Approximately half of patients with prodromal symptoms sought medical care with the presence of chest pain rather than sex being significantly associated with seeking care. Prehospital care was generally similar between sexes overall but cardiovascular risk-reduction therapy use remained low despite the high prevalence of cardiovascular risk factors and complaints of prodromal symptoms in these patients.

    Prodromal symptoms are common among the young and women experienced these symptoms more often than men. These data highlight the potential importance of detecting early warning symptoms especially in women as they have a higher burden of cardiovascular risk factors and similar ACS severity compared with their non-prodromal symptom counterparts. Sex differences in symptom presentation in this young cohort is consistent with older patient studies.1 ,2 ,14 However, unlike the elderly,1 prodromal symptoms in the young were largely non-specific, such as intense and unusual tiredness/fatigue, sleeping disturbances and anxiety with chest pain occurring less frequently. These results also contrast acute symptoms of ACS in young patients where there are high rates of chest pain.6 Underlying reasons for sex differences in experiencing prodromal symptoms are unclear. Women with prodromal symptoms were more likely to have a family history of premature cardiac disease that may have increased their awareness of these symptoms. Women have been noted to be more attentive to symptoms for other conditions15 and therefore may also have been more attentive given their significantly higher burden of cardiovascular disease risk factors than women without prodromal symptoms.

    There have been increasing public awareness programmes especially targeted to women, to improve recognition and healthcare seeking for ACS symptoms. In this study, although women were more likely to seek medical care than men, this sex difference was explained by differences in symptom presentation. Most women and men who experienced prodromal chest pain similarly sought medical care. In contrast, women and men with the more common, non-chest pain symptoms were less likely to seek care despite their high burden of risk factors. This low rate of care seeking may be due to misattribution of non-specific symptoms in younger patients. In a qualitative study, competing priorities, and failure to accurately assess personal cardiac risk were also thought to underlie poor health-seeking behaviours for prodromal and acute symptoms of ACS in women.14 Studies suggest that men tend to underuse health services or disclose distressing symptoms.16 ,17

    Popular opinion and findings from qualitative studies7 ,18 raise concern that women may receive less investigation or interventions when seeking medical care. In our study, we found no sex difference in receipt of cardiovascular risk-reduction therapy, referral for investigations, medication changes or referrals to specialists or hospital when patients reported prodromal symptoms. Women were also more likely to use ambulance services at time of ACS. Although several studies identified sex differences in prescribing risk-reduction therapies and in receiving invasive cardiac investigations,18 ,19 more recent studies demonstrate a narrowing in this care gap.20

    Although there was a significantly higher utilisation of cardiovascular risk-reduction therapies among those who sought care, overall usage was low in all groups. This failure to optimise cardiovascular risk-reduction therapy is concerning given the high burden of cardiovascular risk factors and patient's complaint for prodromal symptoms including chest pain. Over 40% of patients seeking care for prodromal symptoms were smokers, yet none were prescribed smoking-cessation therapy. Reasons for underprescribing of therapies are likely complex and may be due to expected co-management with referred specialists, contraindications for therapy, patient preferences for delaying therapy or suspected non-adherence.21 Patients may have already been at target blood pressure and lipid levels at the time of prodromal symptom presentation, however, antiplatelet agents were also not initiated in the majority of patients. Care providers may not recognise the prodromal symptoms given their largely non-specific nature, existing risk scores may underestimate cardiovascular risk in younger patients22 or providers may consider delaying therapy until confirmation of diagnoses.

    This study has several strengths including our focus on a young ACS population, collection of rich, clinical data in detail and use of standardised prodromal symptom questionnaires. However, there are several limitations to note. First, our patient cohort only included patients diagnosed with ACS and therefore we did not include patients reporting prodromal cardiac symptoms who did not ultimately develop ACS, or those who did not seek care for their ACS. Care provider interventions and investigations for our exploratory analyses were reported by patients rather than by physicians and so may be under-reported. However, studies of recall of investigations like ECG and medications by patients were high.23 ,24 Also, we were unable to discern what was the exact nature of the investigations ordered or medication changes. Nevertheless, we were able to identify medications prescribed on arrival to the emergency department reflecting outpatient medication management. Finally, the MAPMISS prodromal questionnaire is validated in women only.11 We included additional symptoms as well as open-ended questions for symptoms and men rarely reported symptoms not included in the MAPMISS questionnaire.

    Conclusion

    Prodromal symptoms are common among those with ACS with young women more often reporting these symptoms than men. Most prodromal symptoms in the young are non-specific and not often highlighted in public ACS awareness campaigns. These patients had a high burden of cardiovascular risk factors and therefore prompt recognition and medical attention in at-risk patients may provide an earlier opportunity to initiate risk-reduction therapy. Despite popular opinion that women's cardiac complaints are not taken as seriously as men's, we found no evidence of sex differences in the care received from physicians for prodromal symptoms. However, few received cardiac risk-reduction therapy prior to ACS suggesting an unrealised care gap for initiating preventive therapy during a potentially critical time window.

    Key messages

    What is already known on this subject?

    • Young women are at risk for failing to have their acute coronary syndrome (ACS) symptoms recognised but less is known about prodromal symptoms that occur in the days and weeks prior to ACS.

    • Young women report that healthcare professionals respond inconsistently to their symptom complaints prior to ACS in qualitative study.

    What might this study add?

    • Women were more likely than men to report prodromal symptoms and to seek medical attention.

    • Although generally similar between sexes, there was low use of cardiovascular risk-reduction therapy among those who sought medical attention prior to ACS.

    How might this impact on clinical practice?

    • The presence of prodromal symptoms in patients aged 55 years or younger with multiple cardiac risk factors may provide an opportunity to ensure cardiac risk factor control and education on the need to seek immediate care if they experience acute symptoms of ACS.

    References

    Footnotes

    • Collaborators The GENESIS PRAXY Team: Louise Pilote, MD, MPH, PhD, Divisions of General Internal Medicine and Clinical Epidemiology, McGill University Health Center, Montréal, QC, Canada; Igor Karp, MD, MPH, PhD, University of Montréal Hospital Research Centre, CRCHUM and Department of Social and Preventive Medicine, University of Montréal, Montréal, QC, Canada; Simon Bacon, PhD, Concordia University and Research Centre, Hôpital du Sacré-Coeur de Montréal, Montréal, QC, Canada; Jafna L Cox, BA, MD, FRCPC, FACC, Department of Medicine and of Community Health and Epidemiology, Dalhousie University, Halifax, NS, Canada; Kaberi Dasgupta, MD, MSc, FRCPC, Research Institute of the McGill University Health Center, Montréal, QC, Canada; Stella S Daskalopoulou, MD, MSc, PhD, Research Institute of the McGill University Health Center, Montréal, QC, Canada; Mark J. Eisenberg, MD, MPH Jewish General Hospital, McGill University, Montréal, QC, Canada; James Engert, PhD, Research Institute of the McGill University Health Center, Montréal, QC, Canada; William Ghali, MD, MPH, FRCPC, University of Calgary, Calgary, AB, Canada; Karin Humphries, MBA DSc, University of British Columbia, Vancouver, BC, Canada; Nadia Khan, MD, MSc, University of British Columbia, Vancouver, BC, Canada; Kim Lavoie, PhD, University of Quebec at Montréal, UQAM and Research Centre, Hôpital du Sacré-Coeur de Montréal, Montréal, QC, Canada; Colleen Norris, RN, PhD, University of Alberta, Edmonton, AB, Canada; Doreen Rabi, MD, FRCPC, MS, University of Calgary, Calgary, AB, Canada; Derek So, MD, FRCPC, FACC, University of Ottawa Heart Institute, Ottawa, ON, Canada; Ken D Stark, PhD, Department of Kinesiology, University of Waterloo, Waterloo, ON, Canada; Vicky Tagalakis, MD, FRCPC, MSc, McGill University, Divisions of Internal Medicine and Centre for Clinical Epidemiology and Community Studies, Jewish General Hospital, Montréal, QC, Canada; Meytal Avgil TSADOK, PhD, Research Institute of the McGill University Health Center, Montréal, QC, Canada; Roxanne Pelletier, PhD, Research Institute of the McGill University Health Center, Montréal, QC, Canada; George Thanassoulis, MD, FRCPC, Research Institute of the McGill University Health Center, Montréal, QC, Canada; Avi Shimony, MD, Jewish General Hospital, McGill University, Montréal, QC, Canada; Dr Krishan Ramanthan, St Paul's Hospital, Vancouver, British Columbia, Canada; Dr Jan Kornder, Surrey Memorial Hospital, Surrey, British Columbia, Canada; Drs Doreen Rabi and Todd Anderson, Libin Cardiovascular Institute of Alberta, University of Calgary, Calgary, Alberta, Canada; Drs. Colleen Norris and Michelle Graham, University of Alberta and the Mazankowski Alberta Heart Institute, Edmonton, Alberta, Canada; Dr Derek So, University of Ottawa Heart Institute, Ottawa, Ontario, Canada; Dr Madhu Natarajan, McMaster University/Hamilton Health Sciences (General Site), Hamilton, Ontario, Canada; Dr Mike Rokoss, McMaster University/Hamilton Health Sciences (Juravinski Site); Hamilton, Ontario, Canada; Dr Michele Turek, Ottawa Hospital, Ottawa, Ontario, Canada; Dr Asim Cheema, St Michael's Hospital, Toronto, Ontario, Canada; Dr Shahar Lavi, London Health Sciences Centre, London, Ontario, Canada; Dr Sherryn Roth, The Scarborough Hospital, General Division, Scarborough, Ontario, Canada; Dr Thao Huynh, Hôpital Général de Montréal, Montréal, Québec, Canada; Dr Viviane Nguyen, Hôpital Royal Victoria, Montréal, Québec, Canada; Dr Mark Eisenberg, Hôpital Général Juif-Sir Mortimer B Davis, Montréal, Québec, Canada; Dr Julie Méthot, Institut universitaire de cardiologie et de pneumologie de Québec (Hôpital Laval), Québec, Québec, Canada; Dr Michel Doucet, Hôpital du Sacré-Coeur de Montréal, Montréal, Québec, Canada; Dr Martine Montigny, Cité de la Santé de Laval, Laval, Québec, Canada; Dr Samer Mansour, Hôtel Dieu du Centre Hospitalier de l'Université de Montréal, Montréal, Québec, Canada; Dr Claude Lauzon, Centre de santé et de services sociaux de la région de Thetford, Thetford Mines, Québec, Canada; Dr Tomas Cieza, CSSS Chicoutimi, Chicoutimi, Québec, Canada; Dr Michel Nguyen, Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, Québec, Canada; Dr François Grondin, CSSS Alphonse Desjardins (CHAU—Hôtel-Dieu de Lévis), Lévis, Québec, Canada; Dr Jafna Cox, Queen Elizabeth II Health Science Centre, Halifax, Nova Scotia, Canada; Dr Peter Fong, The New Brunswick Heart Centre Research Initiative and The New Brunswick Heart Centre, New Brunswick, Canada; Dr Dhananjai Menzies, Basset Healthcare, Cooperstown, New York, USA; Dr Nicolas Rodondi, Inselspital, University of Bern, Switzerland and Lausanne University Hospital, Lausanne, Switzerland.

    • Contributors NAK, LP, IK, SSD, MJE, CMN contributed to the design of the study. LP and MAT contributed to data collection and NAK, LP, RP, MAT contributed towards analysis. All authors contributed to interpretation of the results, and meaningful contribution to writing and accepting the final manuscript. NAK and LP had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    • Funding This study was funded by the Canadian Institutes of Health Research (CIHR) and the Heart and Stroke Foundations of Quebec, Nova Scotia, Alberta, Ontario and Yukon and British Columbia, Canada. NAK is funded by a Michael Smith Foundation for Health Research Career Scientist award. LP, KD and SSD are supported by the Fonds de recherche du Québec (FRQS) award. SSD is supported through FRQS—Sociétéquébécoised'hypertensionarterielle—Jacques de Champlain Clinician Scientist Career Award. IK is a CIHR New Investigator and an FRSQ Junior 1 Scholar. LP holds a James McGill Chair in medicine.

    • Competing interests None declared.

    • Ethics approval McGill Ethics Board and at all local IRBs.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Investigators interested in working with these data should please contact Jasmine Poole, study coordinator: jpoole@epimgh.mcgill.ca or Dr Louise Pilote, Principal Investigator: louise.pilote@mcgill.ca.

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.