Article Text

Original research
Temporal trends and sex differences in sudden cardiac death in the Copenhagen City Heart Study
  1. Frederik Nybye Ågesen1,
  2. Thomas Hadberg Lynge1,
  3. Paul Blanche1,2,3,
  4. Jytte Banner4,
  5. Eva Prescott5,6,
  6. Reza Jabbari1,
  7. Jacob Tfelt-Hansen1,4
  1. 1 Department of Cardiology, University of Copenhagen, Rigshospitalet, Copenhagen, Denmark
  2. 2 Department of Cardiology, University of Copenhagen, Gentofte, Denmark
  3. 3 Section of Biostatistics, University of Copenhagen, Copenhagen, Denmark
  4. 4 Department of Forensic Medicine, Faculty of Health and Medical Sciences, Copenhagen, Denmark
  5. 5 Department of Cardiology, University of Copenhagen, Bispebjerg Hospital, Copenhagen, Denmark
  6. 6 Copenhagen City Heart Study, Frederiksberg Hospital, Copenhagen, Denmark
  1. Correspondence to Dr Frederik Nybye Ågesen, Department of Cardiology, University of Copenhagen, Rigshospitalet, Copenhagen, Denmark; frederik.nybye.aagesen{at}regionh.dk

Abstract

Objective More knowledge about the development of sudden cardiac death (SCD) in the general population is needed to develop meaningful predictors of SCD. Our aim with this study was to estimate the incidence of SCD in the general population and examine the temporal changes, demographics and clinical characteristics.

Methods All participants in the Copenhagen City Heart Study were followed from 1993 to 2016. All death certificates, autopsy reports and national registry data were used to identify all cases of SCD.

Results A total of 14 562 subjects were included in this study. There were 8394 deaths with all information available, whereof 1335 were categorised as SCD. The incidence of SCD decreased during the study period by 41% for persons aged 40–90 years, and the standardised incidence rates decreased from 504 per 100 000 person-years (95% CI 447 to 569) to 237 per 100 000 person-years (95% CI 195 to 289). The incidence rate ratio of SCD between men and women ≤75 years was 1.99 (95% CI 1.62 to 2.46). The proportion of SCD of all cardiac deaths decreased during the observation period and decreased with increasing age. Men had more cardiovascular comorbidities (OR 1.34, 95% CI 1.07 to 1.68, p<0. 01), and SCD was the first registered manifestation of cardiac disease in 50% of all cases.

Conclusion The incidence of SCD in the general population has declined significantly during the study period but should be further investigated for more recent variations as well as novel risk predictors for persons with low to medium risk of SCD.

  • epidemiology
  • risk factors
  • arrhythmias
  • cardiac
  • coronary artery disease
  • global burden of disease

Data availability statement

Deidentified data are not publicly available but may be obtained after permission directly from the Copenhagen City Heart Study (anja.lykke.madsen@regionh.dk).

Statistics from Altmetric.com

Introduction

The estimated incidence of sudden cardiac death (SCD) in the Western general population has been reported to be between 50 and 130 deaths per 100 000 person-years, corresponding to up to 20% of all mortality.1–4 Although coronary artery disease (CAD) is the cause of death in almost 80% of SCD cases, death is the first manifestation of cardiac disease in half of all SCD events.5–7 In addition, most SCDs are unwitnessed and occur at home, which is why SCD prevention is essential to reduce mortality.8 9

Previous preventative measures mainly focus on individuals with a high risk of SCD, that is, left ventricular ejection fraction ≤35, but as half of the SCD cases in the general population are without known cardiac disease, it is of importance to investigate and identify risk factors for SCD in populations with low to medium risk.5 10 To find meaningful predictors of SCD, the burden in the general population needs to be examined in longitudinal studies.5 11 The temporal trends within the last decades are especially of interest, as studies have observed variations in the otherwise decreasing incidence of SCD depending on sex and cardiac history.7 12 13

Therefore, we have conducted this study with the aim of estimating the incidence and temporal trends of SCD and identifying the demographic, clinical and sex differences of SCD in a defined longitudinal community-based cohort. Ultimately, establishing more knowledge about the population at risk and the changes over time may promote new approaches to reduce the risk of SCD.

Methods

Study design

This study was a retrospective study in which we included all participants in the Copenhagen City Heart Study (CCHS) alive on 1 January 1993 and followed participants for subsequent SCD using a comprehensive approach. We reviewed all deaths that occurred in the CCHS from 1993 to 2016. The method has previously been used to identify SCD cases in Denmark.14 15

Copenhagen City Heart Study

The CCHS was initiated in 1976 and recruited a random age-stratified and sex-stratified sample of individuals in the general population aged 20–93 years living in a district (Østerbro) of the Danish capital, Copenhagen. The aim of the CCHS was to describe the distribution of cardiovascular risk factors and diseases in the general population. Out of the 19 329 invited, 14 223 participated in the first examination round in 1976–1978, and at three later follow-ups, 4751 new participants were added to the cohort (online supplemental table 1). All participants gave informed consent at enrolment and the study is still ongoing.

Supplemental material

Data sources

The Danish National Health Service provides tax-financed public healthcare for all Danish residents. Free medical care is guaranteed for all visits to general practitioners, outpatient clinics, emergency departments and public hospitals.

All Danish citizens are issued a unique personal identification number which is linked across all Danish national registries and used in the CCHS. This study was based on two national registries: (1) the Danish Cause of Death Register, which contains information regarding the cause and circumstances of deaths since 1970 including the original death certificates in which the immediate, contributory and underlying causes of death are documented, and (2) the Danish National Patient Register, which contains information on all inpatient activities since 1977 and outpatient visits since 1994. All available hospital discharge summaries and autopsy reports from forensic departments and pathology departments were included and accessed digitally.

Death certificates

In Denmark, death certificates are issued by a medical doctor who determines the most likely cause of death based on all available information, including all medical records. The immediate, contributory and underlying causes of death are registered with a possibility to specify the timing of the causes. The police are contacted if the death was sudden and unexpected or the person was found dead, as they will decide whether a medicolegal external examination should be performed. The external examination is performed by a medical officer of public health and a police officer, who can access emergency medical service (EMS) records, medical records, police reports including eyewitness statements and statements from the general practitioner. The information from the investigation is included in the supplementary information field on the death certificate, which makes Danish death certificates highly suitable for the identification of sudden and unexpected death.14 15

Studied deaths

All deaths from 1 January 1993 to 31 December 2016 in the CCHS were reviewed case-by-case with all the above-mentioned available information. Incomplete or missing death certificates excluded the participant from the study. Two physicians reviewed all information independently to identify all potential sudden deaths (SDs). In case of disagreement, a third reviewer was involved to re-evaluate the case and reach a consensus. Information from the available death certificates, the Danish National Patient Register, the Danish Cause of Death Register, autopsy reports and hospital discharge summaries were used to identify all cases of SCD according to the definitions described below. Death certificates before 1993 have not been digitised and were not obtained.

Definitions

We defined SCD as a sudden, unexpected and natural death, in witnessed cases as an acute change in cardiovascular status leading to death <1 hour, and for unwitnessed cases as a person seen alive and functioning normally <24 hours before being found dead. SCD was subdivided into three categories: definite, probable and possible SCD. Definite SCD was defined as a person with an autopsy with cardiac or unknown cause of death, or a death with confirmed ventricular arrhythmia preceding death. Probable SCD was defined as a presumed cardiac origin in non-autopsied cases after review of all available information with an established time frame from change in cardiovascular status to death. Possible SCD was defined as a presumed cardiac origin in non-autopsied cases after review of all available information without a fully established time frame from change in cardiovascular status to death.

In case of uncertainty regarding the cause of death in autopsied cases, the entire case was discussed with a forensic cardiopathologist. All persons included as possible SCD were free of any severe and/or chronic diseases expected to result in an impending death. Deaths occurring in the hospital were categorised as non-SCD unless an autopsy report or the death certificate clearly stated that the death was sudden and unexpected in a patient hospitalised due to a non-severe disease; therefore, all possible SCDs occurred in non-hospitalised persons. A death was categorised as non-sudden if the person was found in a state of decomposition.

The first manifestation of cardiac disease was defined as having no cardiac diagnoses registered prior to the death date. International Classification of Disease codes (ICD-codes) defining previous cardiac disease and cardiovascular disease (CVD) are available in online supplemental table 2.

Patient and public involvement

Patients and the public were not involved in any way in the study design and conduct, dissemination of results or evaluation of the study since all cases are deceased participants in the cohort study.

Statistics

The inter-reviewer agreement was calculated as the ratio of the total number of case agreements divided by the total number of cases before the consensus on disagreement cases was reached.

Incidence rates were calculated by considering both the late entries into the cohort (ie, left truncation) and lost to follow-up (ie, right censoring). Exact Poisson 95% CIs were computed for incidence rates and rate ratios. We stratified analysis on age (groups) and sex. Temporal incidence trends were calculated over 5-year periods to account for year-to-year variations. Directly standardised (age-adjusted) rate and rate ratios were computed using the European Standard Population, and CIs were computed using gamma and F-distribution, all referenced and presented in online supplemental figure 1.

The clinical characteristics of subjects who died from SCD are presented stratified by sex and SCD status. Continuous variables are presented as the median and IQR. Clinical characteristics were compared between SCD groups using Fisher’s exact test for categorical variables and Wilcoxon rank-sum test for continuous variables.

A two-sided p value of <0.05 was considered statistically significant. The data analyses were performed using SAS software, V.7.1 of the SAS Enterprise Guide (SAS Institute) and R V.3.6.016

Results

Study population

A total of 14 562 persons aged ≥20 years participating in the CCHS in 1993–2016 were included in the study. Throughout the 228 715 person-years of follow-up (median follow-up time 16.7 years, maximum 23.9 years), 8555 died. After exclusion of cases with missing (n=43, 0.5%) and incomplete (n=118, 1.4%) death certificates, the remaining 8394 deaths were reviewed.

After the assessment of all available data, 1335 SCDs (15.9% of all deaths) were identified (figure 1). Of these, 105 (8%) were definite SCDs, 438 (33%) probable SCDs and 792 (59%) possible SCDs. The inter-reviewer agreement between physicians was 91%.

Figure 1

Flow chart of the identification of sudden cardiac deaths in participants in the Copenhagen City Heart Study from 1993 to 2016.

Incidence and temporal trends of SCD

There was a statistically significant 41% decrease in the incidence of SCD from 1993 to 2016 (figure 2). Standardised incidence rates for SCD in persons aged 40–90 years decreased from 504 per 100 000 person-years (95% CI 447 to 569) to 237 per 100 000 person-years (95% CI 195 to 289). The temporal change in incidence was equal between sexes. By restricting to only definite and probable SCD in persons aged 40–90 years, the standardised incidence rate in 2012–2016 was 91 per 100 000 person-years (95% CI 64 to 128). In comparison, during the same period and age group, non-SCD (all other death) declined by 8% (from 2858 to 2638 deaths per 100 000) and all cardiovascular deaths declined by 21% (from 1284 to 1016 deaths per 100 000). The proportion of SCD of all deaths in persons aged 40–90 years declined from 23.6% to 15.3%.

Figure 2

Temporal change in the incidence rate of sudden cardiac death per age group from 1993 to 2016. Data are expressed as incidence rates and 95% CIs. PY, person-years.

The incidence rate of SCD in the cohort increased with increasing age in both sexes (figure 3). Overall, men had a proportionally higher incidence of SCD compared with women for all ages; an incidence rate ratio of 5.6 (95% CI 2.4 to 14.9) was found in persons aged 50–60 years, declining with increasing age (online supplemental table 3). The incidence rate ratio between men and women ≤75 years of age was 1.99 (95% CI 1.62 to 2.46).

Figure 3

The incidence rate of sudden cardiac death from 1993 to 2016 per age group stratified by sex. Data are expressed as incidence rates and 95% CIs. PY, person-years.

Clinical characteristics of SCD cases

Clinical characteristics of all SCD cases in 1993–2016 are shown in table 1. In total, 15.1% of all deaths among women were SCD compared with 16.9% among men. There were significantly more men in the definite SCD category (56% vs 44%, p=0.04). There was no variance of detected shockable rhythm in relation to SCD between men and women (4% vs 3%, p=0.05).

Table 1

Clinical characteristics of sudden cardiac death cases by sex

Comparing comorbidities between sexes showed a higher prevalence of CVD among men (OR 1.34, 95% CI 1.07 to 1.68, p<0.01), especially with a male predominance of previously diagnosed CAD and myocardial infarction.

SCD was the first registered manifestation of cardiac disease in 50% of all SCD cases. Women were more likely to have no previous cardiac disease (OR 1.26, 95% CI 1.02 to 1.56, p=0.04). Definite SCD cases were more likely to have a previous diagnosis of a CVD compared with probable and possible SCD (OR 1.87, 95% CI 1.19 to 2.94, p=0.02).

The proportion of SCD among all cardiac deaths was 41%. The proportion of SCD among all cardiac deaths in different age groups was: 53% in <60 years, 51% in 60–70 years, 43% in 70–80 years and 38% in >80 years. During the observation period, the proportion of SCD among all cardiac deaths decreased from 46.8% (95% CI 0.43 to 0.51) to 38.8% (95% CI 0.34 to 0.44, p=0.02).

The clinical characteristics and circumstances of death stratified by SCD subgroups during the study period are shown in online supplemental table 4.

Causes of SCD

The causes of SCD according to information from autopsy reports, death certificates and the Danish Cause of Death Register are shown in figure 4. Among all deaths, the most common causes of death were cancer (27.3%), non-SCD (15.8%) and SCD (15.9%).

Figure 4

Causes of sudden cardiac death in the Copenhagen City Heart Study according to information from autopsy reports, death certificates and the Danish Cause of Death Register.

Table 1 and figures 2–4 are reproduced using only definite and probable SCD in online supplemental table 5 and online supplemental figures 2–4, respectively.

Discussion

We have assessed the incidence of SCD in a large longitudinal cohort followed for 24 years, where we have described incidence rates, temporal trends, clinical and demographical characteristics of SCD cases. Overall, 15.9% of all deaths were SCD, which corresponds with previous studies. During the study period, this proportion declined from 23.6% to 15.3%. The SCD incidence was twice as high in men ≤75 years of age, and half of all SCD cases had no diagnosed cardiac disease prior to death.

From a societal perspective, the impact of SCD is noticeable, as almost half of all deaths in the background population occur before the retirement age with a considerable loss of potential life years.2 17 Subsequently, the temporal trends of SCD and the pathophysiological findings are of importance, as SCD is a leading cause of death as observed in this study. In previous studies, differences in case ascertainment methods, background populations and study periods largely explain the varying incidence rates.1 3 7 9 12 The incidence observed in this study is high compared with previous studies but is mostly explained by the influence of age, sex and study period.1 3 We saw a substantial decrease in older age groups and statistically insignificant trends in persons <60 years. In a previous study, we have observed a decline in the incidence of SCD in persons aged 1–35 years in the Danish population from 2000 to 2009.18 The higher incidence of SCD among men is consistent with previous observational studies, and comparable decreases in SCD incidence have been observed in Finland and Canada since 1998 and 2003, respectively.12 19 Compared with out-of-hospital cardiac arrest (OHCA) studies which share many of the same cases, SCD studies include more cases, as in-hospital deaths are included and OHCA is defined by either involvement of bystanders conducting cardiopulmonary resuscitation and/or defibrillation or EMS attendance.19

The cause of the observed temporal decrease in incidence of SCD is multifaceted. First, significant advances in the knowledge on prevention of CVD have emerged during the study period with identification and treatment of CVD risk factors such as physical inactivity, obesity and hypertension.18 20 Second, improvement in the diagnosis and treatment of acute cardiac disease both prehospital and in-hospital has advanced immensely and possibly impacted the incidence of SCD in Denmark.21

We observed 50% of SCD cases with no known cardiac disease prior to death according to the national registries, which is in accordance with previous studies.5 6 9 This was more predominant in women likely due to a higher prevalence of cardiovascular comorbidities among men and possible underdiagnosing of CAD in women.22 This supports the call for preventive strategies for seemingly low to medium SCD risk populations.

We observed significantly more men dying in the emergency department. Haukilahti et al 19 found that this may partly be explained by a higher proportion of ischaemic SCD in men and the age difference between sexes. Moreover, the fact that men more often die in the emergency department may explain the higher proportion of men-definite SCD cases found in our study.

Previous studies examining the occurrence of SCD by using only ICD-codes from death certificates in the USA are known to overestimate the incidence of SCD.23–26 However, compared with other countries’ death certificates, the Danish death certificates contain a supplementary text field, where circumstances surrounding the death as well as the underlying, immediate and contributing causes of death are specified. Furthermore, the Danish National Patient Register ensures a complete follow-up and information from both in-hospital and outpatient clinics, which enables us to identify terminal patients, other previous hospitalisations and specified diagnoses. Additionally, compared with other countries, coroners and medical examiners in the Copenhagen area are forensic pathologists trained in death investigation with a national registry controlling the death certificates.27 Consequently, the method used in this study differs markedly from a method that relies solely on ICD-codes.

In Denmark, since 1971, forensic autopsies have been mandatory in deaths among drug misusers with subsequent toxicological screening for suspected drugs.28 This limits the risk of misclassification of SCD deaths from overdose due to drug abuse, but this may be under-recognised in probable and possible SCD cases.

Andersen et al 29 found that the non-responders in the CCHS had a higher mortality rate within the first 10 years compared with the included participants in the CCHS, but the difference in mortality was insignificant after 10 years of observation. The non-responders may differentiate significantly from the responders examined in other aspects than mortality, suggesting selection bias and a healthy responder effect. In addition, the area in Copenhagen where the cohort was selected is more affluent compared with other parts of Copenhagen. The results should, therefore, be interpreted in this context.

Despite a thorough review of all available data, the burden of SCD found in our study carries a large degree of uncertainty due to a high percentage of at-home deaths (74%) and many unwitnessed deaths. The lack of information immediately prior to a possible SCD is a limitation when studying and developing preventative strategies for SCD. Systematically collected EMS records were not available throughout the study period and were only available if it was described in the death certificate. Autopsies are not mandatory in Denmark in the event of SD, which largely explains the low autopsy rate among SCD cases (5%) in this study. The cause of death in non-autopsied SCD cases is based on information from the death certificates and national registries, and the estimates should, therefore, be interpreted with caution, as information from death certificates is often inaccurate.30 The time frame from change in cardiovascular status to death was not fully established in 59% of SCD cases (possible SCD). These were mainly healthy individuals who were found dead, where we cannot rule out the possibility that some were not seen alive and functioning normally within 24 hours prior to death. However, we were able to identify and exclude persons with a terminal illness or long-term hospitalisation prior to death, thereby eliminating individuals with an expected death.

Conclusions

During the study period, the incidence of SCD declined by 41% in individuals aged 40–90 years. In persons aged ≤75 years, the incidence of SCD among men was twice as high as compared with women. Men had more comorbidities and more often died in the emergency department. This cohort will form the basis for further studies of SCD risk factors in the general population.

Key messages

What is already known on this subject?

  • The incidence of sudden cardiac death (SCD) is significant in the general population. Preventative measures for cardiovascular diseases have significantly decreased the mortality within the last decades, but the temporal changes in SCD in relation to other deaths have not been investigated as well in recent years.

What might this study add?

  • Even though we observed a decrease in the incidence of SCD by 41% in persons 40–90 years during the study period (1993–2016), 15.3% of all deaths in 2012–2016 were SCD. In context, all non-SCD deaths declined by 8% and all cardiovascular deaths declined by 21% through the same period. The SCD incidence is twice as high for men compared with women ≤75 years, and SCD is the first manifestation of cardiac disease in half of all cases.

How might this impact on clinical practice?

  • SCD is an important contributor to total mortality in all age groups in the general population, also among persons not previously diagnosed with cardiovascular disease. Future research should focus on early identification of persons at risk in the general population that would benefit from further investigation and targeted prevention.

Data availability statement

Deidentified data are not publicly available but may be obtained after permission directly from the Copenhagen City Heart Study (anja.lykke.madsen@regionh.dk).

Ethics statements

Ethics approval

This study was approved by the Danish Data Protection Agency (P-2019-159) and the Danish Patient Safety Authority (3-3013-2730/1).

Acknowledgments

The authors thank all the participants in the Copenhagen City Heart Study.

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

  • RJ and JT-H are joint senior authors.

  • Contributors JT-H, RJ and EP wrote the study protocol. JT-H, RJ, THL and FNÅ reviewed death certificates. FNÅ wrote the first draft of the manuscript. All authors contributed to the further development of the manuscript, discussed the results and helped to complete the manuscript for publication. All authors have approved the final version of the manuscript for submission.

  • Funding This project has received funding from the European Union’s Horizon 2020 research and innovation programme ESCAPE-NET (grant number: 733381).

  • Competing interests None declared.

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

  • 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.

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.

Linked Articles