Article Text
PDF
Abstract
Objective Evidence on sex differences in outcomes after developing coronary heart disease (CHD) has focused on recurrent CHD, all-cause mortality or revascularisation. We assessed sex disparities in subsequent major adverse cardiovascular events (MACE) in adults surviving their first-time CHD.
Methods Using a population-based cohort obtained from the Clinical Practice Research Datalink (CPRD GOLD) linked to hospitalisation and death records in the UK, we identified 143 702 adults (aged ≥18 years) between 1 January 1998 and 31 December 2017 with no prior history of MACE. MACE outcome was a composite of recurrent CHD, stroke, peripheral vascular disease, heart failure and cardiovascular-related mortality. Multivariable models (Cox and competing risks regressions) were used to assess differences between sexes.
Results There were 143 702 adults with any incident CHD (either angina, myocardial infarction or coronary revascularisation). Women (n=63 078, 43.9%) were older than men (median age, 73 vs 66 years). First subsequent MACE outcome was observed in 91 706 (63.8%). Women had a significantly lower risk of MACE (hazard ratio (HR), 0.68 (95% CI 0.67 to 0.69); sub-hazard ratio (HRsd), 0.71 (0.70 to 0.72), respectively) and recurrent CHD (n=66 543, 46.3%) (HR, 0.60 (0.59 to 0.61); HRsd, 0.62 (0.61 to 0.63)) when compared with men after incident CHD. However, women had a significantly higher risk of stroke (n=5740, 4.0%) (HR, 1.26 (1.19 to 1.33); HRsd, 1.32 (1.25 to 1.39)), heart failure (n=7905, 5.5%) (HR, 1.09 (1.04 to 1.15); HRsd, 1.13 (1.07 to 1.18)) and all-cause mortality (n=29 503, 20.5%) (HR, 1.05 (1.02 to 1.07); HRsd, 1.11 (1.08 to 1.13)).
Conclusions After incident CHD, women have lower risk of composite MACE and recurrent CHD outcomes but higher risk of stroke, heart failure, and all-cause mortality compared with men.
- coronary heart disease
- major adverse cardiovascular events
- secondary prevention
- sex difference
- competing risks
Data availability statement
The data supporting the findings of this study are available from the Clinical Practice Research Datalink (CPRD). Restrictions apply to the availability of these data, which were used under licence for the current study, hence are not publicly available.
Statistics from Altmetric.com
- coronary heart disease
- major adverse cardiovascular events
- secondary prevention
- sex difference
- competing risks
Introduction
Coronary heart disease (CHD) is a global public health problem1 and remains a major cause of early morbidity and mortality despite advances in treatment and public health.2 With many individuals surviving their initial CHD presentation, there is a growing population with established CHD with a substantially high risk of subsequent cardiovascular events or death.3 The residual high risk in these individuals persists despite optimal therapy.4 Greater and more nuanced understanding of their risk of subsequent events is needed to enable more targeted secondary prevention strategies.
A large body of evidence has outlined differences in the clinical presentation,5 6 diagnosis7 and management/treatment8–10 between men and women with an established diagnosis of CHD. Women with established CHD may have a lower probability of coronary revascularisation procedures9 and a higher mortality outcome compared with men.9 Most research examining sex differences in patients’ outcomes with CHD or CHD subtypes has focused primarily on recurrent CHD or CHD subtypes, all-cause mortality, revascularisation or outcomes in the first year after CHD.9 11 However there remains considerable uncertainty about wider experience of composite cardiovascular outcomes such as major adverse cardiovascular events (MACE) (recurrent CHD, stroke, peripheral vascular disease (PVD), heart failure and cardiovascular-related mortality) after incident CHD.
In this population-based cohort study we used multiple databases of electronic health records (EHRs) from primary care consultations, secondary care (hospital admissions and procedure-level data), and the national death registry, known to be representative of the UK population. We sought to estimate sex disparities in first subsequent MACE outcome in adults with any incident CHD.
Methods
Data source
This prospective population-based cohort study used the UK Clinical Practice Research Datalink (CPRD GOLD) database of anonymised longitudinal primary care EHRs,12 linked to secondary care hospitalisation data (Hospital Episode Statistics (HES)),13 national mortality data (Office for National Statistics (ONS))14 and social deprivation data (2015 Index of Multiple Deprivation (IMD)).15 Individuals included in the CPRD GOLD database, from a network of general practices across the UK, are representative of the UK general population in terms of sex, age and ethnicity,12 thereby validating CPRD GOLD for epidemiological research.
Study population
We identified a cohort of individuals with any incident non-fatal CHD in either primary care (CPRD GOLD) or secondary care (HES) data between 1 January 1998 and 31 December 2017, so long as the patient has at least 12 months of registration at the practice, the diagnosis was made after the first 12 months of their current registration period,16 the practice was deemed to be contributing ‘up-to-standard’ data (online supplemental methods), and the patient’s CPRD record had linkage to HES. CHD was defined as angina, myocardial infarction (MI) or coronary revascularisation (coronary bypass surgery or coronary angioplasty)17 (online supplemental table 1 for codes used in identifying both incident and outcome events). Individuals with a history of any stroke, PVD or heart failure before incident CHD were excluded. The study flow diagram is presented in online supplemental figure 1.
Supplemental material
Descriptive characteristics of the study population
Outcome measures
First subsequent MACE after incident CHD was the primary outcome. MACE was defined as a composite of recurrent CHD, any stroke, PVD, heart failure or cardiovascular-related mortality, based on record from across the linked data sources (CPRD, HES or ONS registry). All-cause mortality was considered as a secondary outcome.
The study cohort and outcomes were identified from CPRD using Read codes, from HES using the International Classification of Diseases, Tenth Revision codes, and the Office of Population Censuses and Surveys Classification of Surgical Operations and Procedures revision 4.6 for procedure codes. All code lists used are available for download from https://portal.caliberresearch.org.18
Cohort demographics and baseline characteristics
Age was defined at the time of incident CHD. Ethnicity was categorised into six groups: Asian, Black, Mixed, Other, White and unknown.19 To describe socioeconomic status (SES), the 2015 English IMD15 linked to the individual’s residential postcode was used. IMD is a weighted mean across seven domains, hence offers a single score to describe the concept of deprivation, categorised into quintiles (from quintile 1 (least deprived group) to quintile 5 (most deprived group)). Medication prescription (issue of prescription) at baseline was defined as a prescription within 12 months before incident CHD. For cholesterol (low-density lipoprotein (LDL), high-density lipoprotein (HDL) and total), body mass index (BMI) and blood pressure measures (diastolic and systolic), the most recent values/measures within 24 months before incident CHD were used. All other comorbidities were defined based on the latest record before incident CHD.
Statistical analysis
The Shapiro-Wilk test was used to assess normality of distribution for continuous variables. Mann-Whitney U test for continuous data and χ2 test for categorical data were used to compare baseline characteristics between men and women. The level of missing values ranged between 17.5% for blood pressure measures and 62.7% for LDL cholesterol. Details on the proportion of missingness and differences in characteristics between those with and without missing data are provided in online supplemental tables 2 and 3. To estimate missing values for BMI, systolic and diastolic blood pressures, HDL cholesterol, LDL cholesterol and total cholesterol levels, multiple imputation by chained equations was used to generate 10 imputed data sets using all the other available patient variables and all the outcomes.20 The imputed data sets were pooled into a single data set using Rubin’s rules.21 Age-standardised prevalence for comorbidities and prescribed medications at baseline were obtained by using the study population to standardise the prevalence across men and women.
Incidence rates with 95% CI for first subsequent MACE, its individual constituents and all-cause mortality end points were calculated by dividing the number of incident outcomes by the total person-years at risk. Kaplan-Meier curves accompanied by HRs from Cox proportional hazard regression models were used to analyse the time-to-event outcomes. Competing risk analysis, which provides the cause-specific HR (or sub-HR), was used to calculate the cumulative incidence of the outcomes. The method proposed by Fine and Gray22 was used to estimate the association of sex with the subhazard of MACE (or the specific individual constituent of MACE) and all-cause mortality. Non-cardiovascular-related mortality was considered a competing risk for MACE outcome. For both Cox and competing risks models, results are presented for models adjusted for age (model 1) and models adjusted for age, SES, smoking status, BMI, blood pressure (diastolic and systolic), total cholesterol level, history of alcohol problem, diabetes mellitus, dyslipidaemia, cancer, chronic kidney disease (CKD), hypertension, atrial fibrillation (AF), depression and a family history of cardiovascular disease (CVD) (model 2). The composite MACE outcome was further analysed using a win ratio approach,23 first described by Finkelstein and Schoenfeld,24 which prioritises fatal outcome(s) (ie, cardiovascular-related death) over less severe or non-fatal outcomes (ie, recurrent CHD, stroke, PVD and heart failure) for composite outcome. The R package, WWR, was used for the win ratio analysis. In a sensitivity analysis, subsequent outcomes within 30 days were considered as representing or relating to the same incident CHD event.25 Analyses were, therefore, restricted to subsequent outcomes occurring after 30 days of incident CHD. All statistical analyses were performed using Stata SE V.16.1 and R V.4.0.3. An alpha level of 0.05 was used for all analyses.
Patient and public involvement
Patients or the public were not involved in the design, conduct or reporting. We plan on involving patient groups in the dissemination of our research findings.
Results
There were a total of 166 068 individuals aged 18 years and over with any incident CHD between 1998 and 2017 in either CPRD GOLD or HES. Of these individuals 22 366 with a record of a major adverse event prior to their incident CHD event were excluded from the analysis. The study, therefore, included a cohort of 143 702 individuals 18 years and over with incident CHD and no prior record of MACE.
Baseline characteristics
The median follow-up time was 13.4 years (IQR: 8.4–17.7 years). The cohort comprised 63 078 (43.9%) women, who were older than men (median age of 73 vs 66 years, p≤0.001). Detailed descriptive characteristics of the study cohort presented by sex are shown in table 1.
After adjustment for age, women had a higher prevalence of the following comorbidities and known risk factors at the time of incident CHD when compared with men: CKD (9.3% vs 8.2%), depression (25.6% vs 13.0%), dyslipidaemia (12.9% vs 11.2%), family history of CVD (27.7% vs 21.0%), hypertension (46.9% vs 40.6%), hypothyroidism (10.9% vs 2.8%), migraine (9.1% vs 3.2%) and rheumatoid arthritis (2.5% vs 1.2%). Within 12 months prior to the incident CHD, women had a higher number of prescriptions for antiarrhythmic, antidepressant, antiepileptic, antihypertensive, antiplatelet, beta-blockers, corticosteroid, diuretics, and both low-intensity and high-intensity statins, after adjusting for age. Online supplemental table 4 details the age-adjusted prevalence for comorbidities, risk factors and prescribed medications.
First subsequent MACE outcome
Most first subsequent major adverse outcomes occurred within 2 years of incident CHD, with the median time to outcome ranging from 0.11 years (IQR: 0.02–0.81) for recurrent CHD to 2.54 years (IQR: 0.63–5.83) for subsequent stroke event. Of the 143 702 individuals with incident CHD, 91 706 (63.8%) had a MACE (men: 55 087 (68.3%) vs women: 36 619 (58.1%)), 66 543 (46.3%) had a recurrent CHD, 5740 (4.0%) strokes, 1624 (1.1%) PVD, 7905 (5.5%) heart failure, 9894 (6.9%) cardiovascular death and 29 503 (20.5%) all-cause death, occurring after the incident CHD events. Figure 1 and online supplemental figure 2 show the distribution of individuals with major adverse outcomes, by sex and across 5-year age bands.
Distribution of first subsequent major adverse outcomes by sex and 5-year age group for patients with incident CHD. CHD, coronary heart disease; MACE, major adverse cardiovascular events.
Cumulative incidence function plots for first subsequent major adverse outcomes.
Incidence rate for clinical outcomes
The overall incidence rate for MACE was 25.18 per 100 person-years (95% CI 25.02 to 25.34), with a higher incidence rate in men compared with women (31.03 vs 19.62 per 100 person-years). Table 2 details the sex variation in the incidence of the constituent MACE outcomes. In comparing women with men, the age-adjusted and SES-adjusted sex-specific incidence rate ratio for MACE was 0.58 (0.57–0.59), for recurrent CHD 0.52 (0.51–0.53), for stroke 1.22 (1.16–1.29), for PVD 0.88 (0.80–0.97), for heart failure 1.00 (0.96–1.05), for CVD-related death 0.89 (0.85–0.93) and for all-cause mortality 0.92 (0.90–0.94).
Incidence of first subsequent major adverse outcomes (N=143 702)
Sex difference and clinical outcomes
After adjusting for age, socioeconomic and smoking status, BMI, blood pressure, total cholesterol, history of alcohol problem, diabetes, dyslipidaemia, cancer, CKD, hypertension, AF, depression and family history of CVD, in both Cox and competing risks models (table 3) women had a significantly lower risk of first subsequent MACE (HR, 0.68 (95% CI 0.67 to 0.69); sub-HR (HRsd), 0.71 (95% CI 0.70 to 0.72), respectively) and recurrent CHD (HR, 0.60 (95% CI 0.59 to 0.61); HRsd, 0.62 (95% CI 0.61 to 0.63)) when compared with men after incident CHD. Women, however, had a significantly higher risk of any stroke (HR, 1.26 (95% CI 1.19 to 1.33); HRsd, 1.32 (95% CI 1.25 to 1.39)), heart failure (HR, 1.09 (95% CI 1.04 to 1.15); HRsd, 1.13 (95% CI 1.07 to 1.18)) and all-cause mortality (HR, 1.05 (95% CI 1.02 to 1.07); HRsd, 1.11 (95% CI 1.08 to 1.13)).
Risk of first subsequent major adverse outcome for women compared with men (reference category)
The cumulative incidence function (figure 2 and online supplemental figure 3) and Kaplan-Meier curves (figure 3 and online supplemental figure 4) as well as the adjusted Kaplan-Meier cumulative incidence curves (online supplemental figure 5) for MACE and its constituent outcomes illustrate women have a higher incidence of subsequent stroke, heart failure and all-cause mortality over a 10-year follow-up period.
Kaplan-Meier plots for first subsequent major adverse outcomes. CHD, coronary heart disease; MACE, major adverse cardiovascular events.
To describe the effect of being a woman on the fatal outcome (cardiovascular-related death) in the composite MACE as compared with the non-fatal outcomes (recurrent CHD, stroke, PVD and heart failure), the win ratio was 1.331 (95% CI 1.329 to 1.331).
Sensitivity analysis
For the sensitivity analysis, 7566 (5.3%) individuals who died within 30 days of incident CHD were excluded. There were 76 571 subsequent MACE outcomes recorded after 30 days of incident CHD for the remaining 136 326 individuals. The median time from incident CHD to subsequent outcome after 30 days ranged from 0.58 years (IQR: 0.21–2.25) for recurrent CHD to 2.98 years (IQR: 0.85–6.59) for all-cause mortality (online supplemental table 5). After full adjustment, in both Cox and competing risks models (online supplemental table 6) women had a significantly lower risk of first subsequent MACE (HR, 0.70 (95% CI 0.69 to 0.71); HRsd, 0.71 (95% CI 0.70 to 0.72), respectively) and recurrent CHD (HR, 0.63 (95% CI 0.62 to 0.64); HRsd, 0.64 (95% CI 0.63 to 0.65)) when compared with men after incident CHD. Women, however, had a significantly higher risk of any stroke (HR, 1.21 (95% CI 1.15 to 1.28); HRsd, 1.27 (95% CI 1.20 to 1.34)) and all-cause mortality (HR, 1.01 (95% CI 0.98 to 1.04); HRsd, 1.08 (95% CI 1.05 to 1.11)). Similar sex differences were observed when the analysis was done by incident CHD time period (1998–2007 and 2008–2017) (online supplemental table 7) and when the analysis was restricted to 61 167 individuals with incident MI (online supplemental table 8).
Discussion
Within a population-based cohort, we show there are sex disparities in the risk of developing first subsequent MACE and its individual constituent events in adults with any incident CHD. Women are less likely to have a MACE or recurrent CHD as a first subsequent event after incident CHD when compared with men. However, women are more likely to have stroke, heart failure or death from any cause after incident CHD.
The risk profiles of men and women have been shown to substantially differ when diagnosed with CHD26 and fare much differently after incident CHD. The cause of disparities is multifaceted, relating to differences in baseline cardiovascular profile, access to care, use of resources and evidence-based guidelines, and social as well as environmental factors.8 27 Previous studies have frequently been based on selected cohorts from trials, registries or individuals with specific type of CHD.9 26 Consistent with our findings, a study of 3779 patients from the Euro Heart Survey of Stable Angina reported women have a higher risk of death even after multivariable adjustment.9 However, in a study of 30 977 outpatients with stable coronary artery disease from the ProspeCtive observational LongitudinAl RegIstry oF patients with stable coronary arterY disease (CLARIFY) register, similar event rates in men and women for the composite outcome of cardiovascular death, non-fatal MI or stroke at 1-year follow-up were observed after adjustment for baseline differences.26 Although 22.6% of the CLARIFY study patients were women, women were more likely to have diabetes and hypertension, consistent with our findings.
Population-based studies, such as our study using data representative of the UK population, provide real-world evidence regarding sex differences in outcomes for patients with incident CHD.28 It is by considering disparities across individuals from the whole spectrum of CHD that the full burden of subsequent MACE outcome can be captured and accurate distinctions made between men and women. Most studies have focused on sex differences in mortality outcome—differences in age, comorbidities and treatment use between men and women have largely explained the sex differences in mortality outcome.11 Studies have also differed in the methodological approach used in assessing sex differences—logistic regression26 as opposed to survival analysis.
The analysis of survival (time-to-event) data plays a key role in cardiovascular research and competing events are prevalent.29 A competing event (eg, death from non-cardiovascular cause) hinders or changes the possibility of observing the outcome of interest (eg, death from cardiovascular-related death). Koller et al 29 found a large majority of clinical studies neglected the competing risks process despite the studies having populations susceptible to competing risks. Failure to account correctly for these competing events results in the overestimation of probabilities for the incidence of outcomes.30 Our analyses illustrated the overestimation of the risk of first subsequent MACE and its constituents when using a standard Cox model. Our study demonstrates the importance of accounting for competing events. The impact of incorrectly treating competing events has practical importance as clinical decisions often rely on an individual’s risk of a disease event or outcome.31
Combining multiple types of clinical outcomes into a single composite outcome is common in clinical research.32 The usual analysis of time to first occurrence of any event in the composite outcome treats individual constituent outcomes as being equally important despite differences in clinical relevance and severity. The novel approach, win ratio,23 provides a useful alternative for analyses of composite outcomes, addressing the limitations of usual first event analysis. Win ratio requires a ranking of outcomes by severity but does not require assigning a specific weight to each outcome. As shown in our study, women have more fatal outcome in composite MACE than men.
Strengths and limitations
This study has a number of strengths. First is the size and representativeness of the CPRD GOLD data set12; this large retrospective population-based study used primary care data linked to hospital and mortality records, allowing us to assess sex-related differences in major CVD events and mortality occurrence after incident CHD. Second, we used an incident cohort, which reflects current practice and avoids the distorting influences of bias present in cohorts with prevalent major adverse events. We acknowledge limitations generally inherent in studies using EHRs. These include missing data in EHRs, including CPRD GOLD. Potential ascertainment and information bias are acknowledged. The coded definitions of outcomes and CHD incident diagnosis used in this study are, however, well established due to the pay-for-performance scheme (Quality and Outcome Framework) which has improved documentation/coding for cardiovascular conditions and associated risk factors.17 33 The potential for misclassification bias is, therefore, not likely. The subtyping of CHD in both primary care (CPRD GOLD) and secondary care (HES) databases is not reliable34 and hence unable to assess differences for CHD subtypes. The use of ‘softer’ CHD codes in primary care data is yet to be validated.35
Conclusions
CHD remains the leading cause of mortality globally. Improved understanding of outcomes in patients with CHD is key to reduce the disease burden. In this large population-based cohort study of patients with any type of incident CHD, we identified after appropriate adjustments for confounders a lower risk of MACE and recurrent CHD in women when compared with men. However, there was a higher risk of stroke, heart failure and all-cause mortality in women. As more people are surviving their incident CHD events, further attention to all patients with incident CHD is needed to narrow this range of sex disparities in major subsequent clinical outcomes. Improving the standard and equity of care for women and men with incident CHD should recognise a ‘one size fits all’ approach may not hold.
Key messages
What is already known on this subject?
Sex differences exist in the presentation, treatment and outcomes of individuals with incident coronary heart disease (CHD).
Most studies have focused on sex differences in recurrent CHD, all-cause mortality or revascularisation.
What might this study add?
The study provides evidence on sex differences in the first subsequent composite major adverse cardiovascular events and constituent outcomes in individuals with any incident CHD using a large population-based cohort.
How might this impact on clinical practice?
As more people are surviving their incident CHD events, further attention to all patients with incident CHD is needed to narrow this range of sex disparities in major subsequent clinical outcomes.
Improving the standard and equity of care for women and men with incident CHD should recognise a ‘one size fits all’ approach may not hold.
Data availability statement
The data supporting the findings of this study are available from the Clinical Practice Research Datalink (CPRD). Restrictions apply to the availability of these data, which were used under licence for the current study, hence are not publicly available.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by the Independent Scientific Advisory Committee of the Medicines and Healthcare products Regulatory Agency (protocol number 19_023R).
Acknowledgments
We thank the practices that contributed to CPRD GOLD.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @rkakyea, @dataevan, @StephenFWeng, @DrRiyazPatel, @fwasselbergs
FWA and NQ contributed equally.
Contributors RKA, SW, FWA and NQ were involved in the design and planning of the study. RKA conducted the main statistical analysis and wrote the first draft of the manuscript. All authors contributed to the interpretation of the data, writing of the manuscript and critical revisions. RKA is the guarantor.
Funding RKA is funded by a National Institute for Health Research School for Primary Care Research (NIHR SPCR) PhD Studentship Award, supervised by JK, FWA and NQ.
Disclaimer The views expressed are those of the authors and not necessarily those of the NIHR, the NHS, or the Department of Health and Social Care.
Competing interests RKA currently holds an NIHR-SPCR funded studentship (2018-2021). SW is currently an employee of Janssen R&D. NQ was a member of the most recent NICE Familial Hypercholesterolaemia and Lipid Modification Guideline Development Groups (CG71 and CG181). NQ and SW have previously received honorarium from AMGEN. RSP has funding from the British Heart Foundation and the National Institute for Health Research. FWA is supported by UCL Hospitals NIHR Biomedical Research Centre. The remaining authors have no competing interests.
Provenance and peer review Not commissioned; externally peer reviewed.
Author note Additional references can be found in online supplemental file 1.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.