Abstract
To study the incidence and aetiology of sudden cardiac death (SCD) in 1- to 35-year-olds in Sweden from 2000 to 2010.
We used the database of the Swedish National Board of Forensic Medicine and the Swedish Cause of Death Registry and identified SCD cases by review of forensic files and death certificates. We identified 552 individuals with SCD in 1- to 35-year-olds; 156 (28%) were women. In 393 (71%), a forensic autopsy had been performed; in 131 (24%), a clinical autopsy had been performed; in 28 (5%) with no autopsy, a cardiac disease was diagnosed before death. The incidence of SCD per 100 000 person-years was 1.3 in 1- to 35-year-olds and 1.8 in 15- to 35-year-olds. In women, the incidence rates yearly decreased during the study period by 11% (95% confidence interval 6.6–14.2). The most common aetiology in 1- to 35-year-olds was sudden arrhythmic death syndrome (31%) and coronary artery disease (15%). In cases with forensic autopsy, death occurred during daily activity (48%), sleep (38%), and physical activity (14%); death was unwitnessed in 60%. Co-morbidity in 15- to 35-year-olds, e.g. psychiatric disorder, obesity, or diabetes, was present in 93/340 (27%) (73 men).
The incidence of SCD among 1- to 35-year-olds in Sweden during 2000–10 was 1.3 per 100 000 person-years (28% women); incidence was decreasing in women. Sudden arrhythmic death syndrome was the most common diagnosis. Co-morbidity such as psychiatric disorders and obesity was common among men.
The incidence of sudden cardiac death among 1- to 35-year-olds in Sweden during 2000–10 was 1.3 per 100 000 person-years (156 women and 396 men).
The incidence among 15- to 35-year-olds was 1.8 per 100 000 person-years during the study period. Compared with data from 1992 to 1999, the incidence was higher but had been the same if the same inclusion criteria had been used (0.9 per 100 000 person-years).
The incidence showed a decreasing trend over time in women during the study period.
Sudden arrhythmic death syndrome was the most frequent diagnosis (31%), followed by coronary artery disease (15%), myocarditis (14%), and unspecified cardiomyopathy (12%).
Hypertrophic cardiomyopathy (5%), arrhythmogenic right ventricular cardiomyopathy (4%), and dilated cardiomyopathy (4%) were less frequent than commonly reported.
Co-morbidities such as obesity and psychiatric disease were present in 27% of 15- to 35-year-olds and was four-fold more common in men.
Introduction
Sudden cardiac death (SCD) among the young is an uncommon and tragic event. The incidence varies between 0.7 and 3.7 per 100 000 person-years, and most patients have a previously unrecognized heart disease.1–5 These cases are found by the exclusion of unnatural, non-cardiac, and non-sudden deaths. There is a challenge involved in selecting ‘true’ cases of SCD. First, it can be difficult to assess whether the ‘sudden’ and ‘unexpected’ criteria are met, especially in cases with a diagnosed heart disease.6 Second, the diagnosis is often unclear even when an autopsy is performed. Sometimes the cause is obvious, such as a major coronary thrombus, but there may also be only minor findings of uncertain relevance for the death. The diagnostic problem in the investigation of SCD was described by the British pathologist Davis: ‘Giving the cause of sudden death for purposes of death certification is an exercise in probabilities rather than certainties’.7 Approximately 50% of sudden deaths remain unexplained after autopsy.8 These deaths may have a cardiac background, such as ion channel diseases or cardiomyopathies with undetectable changes, but may also have non-cardiac causes, such as epilepsy, diabetes, or the use of drugs not detected in toxicological analyses.
Third, it is difficult to know who should be included if the deceased has other diseases. Should individuals with co-morbidity such as psychiatric disease and arrhythmic medication be included? The recently published European Society of Cardiology (ESC) guidelines discuss ‘arrhythmic risk in selected populations’ such as patients with psychiatric diseases, muscular dystrophies, pregnant patients, and athletes.6 This suggests that such risk groups should be included in the SCD population.
During the last decade, different efforts have been made to prevent SCD in Sweden. These have included a thorough investigation in cases with cardiac-related symptoms; insertion of appropriate treatment, e.g. implantable cardioverter defibrillator; routines concerning family investigation in cases of SCD; ECG screening of athletes; and early cardiopulmonary resuscitation in out-of-hospital cardiac arrest.9
We have studied the incidence and aetiology of SCD in the young in Sweden during 2000–10, described some characteristic features of this population, and compared the results with a study from the previous decade.4
Methods
Definitions
Sudden death: non-traumatic, unexpected fatal event occurring within 1 h of the onset of symptoms in an apparently healthy subject. If death is not witnessed, the definition applies when the victim was in good health 24 h before the event.
Sudden unexpected death syndrome (SUDS): sudden death without an apparent cause and in which an autopsy has not been performed.
Sudden cardiac death: sudden death when a congenital, or acquired, potentially fatal cardiac condition was known to be present during life; or autopsy has identified a cardiac or vascular anomaly as the probable cause of the event; or no obvious extra-cardiac causes have been identified by post-mortem examination and therefore an arrhythmic event is a likely cause of death.
Sudden arrhythmic death syndrome (SADS): both autopsy and toxicology investigations are inconclusive, the heart is structurally normal at gross and histological examination, and non-cardiac aetiologies are excluded.6
Inclusion criteria
We included cases meeting the criteria of SCD/SADS. Individuals who were successfully resuscitated after out-of-hospital cardiac arrest but died hours or days later in hospital were also included. Cases with aborted sudden death and cases with SUDS were not included.
Description of databases
We used the database of the Swedish National Board of Forensic Medicine (SFM) and the Swedish Cause of Death Registry (SDR). All cases with a forensic autopsy are registered in SFM. It contains the coded death diagnoses and short information about clinical details and circumstances of death. The documents needed for further information, i.e. forensic and police reports, are available in six forensic departments. Interviews of eyewitnesses, family members, and medical records in cases with a medical history are included in the forensic files.
All deaths in Sweden are registered in SDR. It contains the coded death diagnoses and the type of autopsy being performed, if any, but no circumstances around death. The death certificates can be ordered from the Swedish National Board of Health and Welfare. They contain the diagnoses, place of death, and the duration of the acute condition and of pre-existing diseases. Information varies, particularly as to whether death was ‘sudden and unexpected’.
Autopsy
According to Swedish regulations, the police are almost always involved in unexplained sudden death, especially in the young. The police normally order a forensic autopsy to be performed. Some cases might not be reported to the police, and sometimes the police have no interest of further investigation. In these cases, the doctor responsible for death certification can call for a clinical autopsy in the hospital. In cases with a known potentially lethal disease, the doctor can refrain from a post-mortem examination.
All forensic autopsies are performed or supervised by a limited number (∼20) of experienced forensic pathologists, located in six regional departments. Autopsy is performed according to several consensus documents such as ‘Principles and Procedures of Forensic Autopsy/Basis for National Guidelines’ issued by the Swedish National Board of Forensic Medicine (1994) and a newly implemented Swedish accreditation standard protocol. Clinical autopsies are performed in numerous hospitals in Sweden. The procedures are based on tradition, and there is no generally accepted standard.
Selection of the sudden cardiac death study group in Swedish National Board of Forensic Medicine database
We searched for all deaths with a forensic autopsy in persons aged 1–35 years in Sweden in 2000–10 (Figure 1). (i) Unnatural deaths were excluded. These correspond to the following diagnosis codes in the International Classification of Diseases (ICD)-10 system: V01–Y98 or F10–F19 (murder, suicide, accidents, and intoxication) (Figure 1). Accidents that were considered to be caused by heart disease were included. (ii) Non-cardiac and non-sudden deaths were then excluded manually. This included subarachnoid haemorrhage, diabetes coma, tumours, epilepsy, infections, pulmonary emboli, abusers, and decomposed corpse. Cases with ‘unknown’ cause of death, ICD code R96–R99, were not excluded. (iii) In remaining cases, the forensic records were ordered and independently reviewed by two of the authors (A.W. and P.K.). More details about the deceased were now obtained. We excluded individuals with clinical information about diseases or medication likely causative to their deaths. This included abuse, unstable epilepsy, individuals with a high load of medication such as anti-anxiety medication and opioids, and blood alcohol content (BAC) of ≥1.5%. Co-morbidity not considered as a likely cause of death by the forensic physician was included, e.g. diabetes, obesity, and psychiatric disease. All with forensic autopsy had a toxicological examination.
The selection procedure of the SCD population in 1- to 35-year-olds in the Swedish Forensic Registry and in the Swedish Cause of Death Registry during 2000–10. SUDS, sudden unexpected death syndrome; SFM, database of the Swedish National Board of Forensic Medicine, ‘Swedish Forensic Registry’; SDR, Swedish Cause of Death Registry.
Selection of the sudden cardiac death study group in Swedish Cause of Death Registry
Swedish Cause of Death Registry was searched for all deaths in persons aged 1–35 years during the study period (Figure 1). (i) Unnatural deaths, i.e. with ICD-10 code V01–Y98 or F10–F19 (murder, suicide, accidents, and intoxication), and non-cardiac diagnoses were excluded. Accidents that were considered to be caused by heart disease were included. Cases with ‘unknown’ cause of death’, ICD code R96–R99, were not excluded. (ii) In remaining cases, the death certificates were ordered (397 missing) and independently and manually reviewed by two of the authors (A.W. and E.S.). Non-cardiac and non-sudden deaths were then excluded, as were cases with unknown cause of sudden death without autopsy, and deaths with forensic autopsy. To assess the validity of death certificates among those cases that did not undergo forensic autopsy, we obtained a random sample of 20% of the records from the clinical autopsies. In most cases, death was witnessed and they had a history of heart dysfunction. Approximately one out of seven toxicological analyses, rarely contributing to causation of the fatality, were performed (most often alcohol analysis). The information contained in the death certificates was quite similar to the autopsy protocols.
Ethical considerations
The study was approved by the Regional Ethical Review Board at Umeå University.
Statistical analysis
The SCD incidence by sex, age, and year was calculated from the numbers of SCD in the study population divided by the number of residents in Sweden for each sex, age, and year, according to Statistics Sweden.10 The incidence by age group was calculated by dividing the total number of SCD cases in each age group between 2000 and 2010 by the number of person-years in each age group. χ2 (Yates-corrected) test was used to detect significant differences between groups, and the level of statistical significance was set at P< 0.05. Poisson regression was used to test trends by calendar year and age.
Results
The sudden cardiac death study group
During the 11-year review period from 2000 to 2010, there were 15 533 deaths in age group 1–35 years; a forensic autopsy was performed in 9182 (59%); a clinical autopsy was performed in 1061 (7%). After selection procedures in both databases, SCD study group ultimately included 552 cases (156 women), 393 (71%) with a forensic autopsy, 131 (24%) with a clinical autopsy, and 28 (5%) without an autopsy but with a pre-known cardiac disease (Figure 1). By cross checking, we found that eight cases of the forensic study group were not listed in SDR, which could be due to errors in the coding procedures in the National Board of Health and Welfare.
Incidence
The mean population in Sweden during the study period of 2000–10 was 9.10 million inhabitants, and the mean number of 1- to 35-year-olds was 3.90 million persons. This corresponds to 42.9 million person-years during the 11-year period in age group 1–35. The incidence of SCD per 100 000 person-years was 1.3 in age group 1–35 years. The corresponding incidence figures were 0.5 in age group 1–14 years and 1.8 in age group 15–35 years (Table 1). The incidence rates of SCD decreased over time with on average 11% per year in women (95% confidence interval 6.6–14.2). In men, there was no change in the rates of SCD over time (Figure 2). As for age, in men there was an average of 7.7% increase per year with increasing age (95% confidence interval 5.8–9.7) and in women the increase was 5.7% per year (95% confidence interval 2.5–9.1) (Figure 3).
Incidence of sudden cardiac death in persons aged 1–35 years in Sweden during 2000–10
| Age (years) | No. of SCD | Million person-years | Incidence per 100 000 person-years |
|---|---|---|---|
| 1–14 | 76 | 16.1 | 0.5 |
| 15–35 | 476 | 26.8 | 1.8 |
| 1–35 | 552 | 42.9 | 1.3 |
| Age (years) | No. of SCD | Million person-years | Incidence per 100 000 person-years |
|---|---|---|---|
| 1–14 | 76 | 16.1 | 0.5 |
| 15–35 | 476 | 26.8 | 1.8 |
| 1–35 | 552 | 42.9 | 1.3 |
SCD, sudden cardiac death.
Incidence of sudden cardiac death in persons aged 1–35 years in Sweden during 2000–10
| Age (years) | No. of SCD | Million person-years | Incidence per 100 000 person-years |
|---|---|---|---|
| 1–14 | 76 | 16.1 | 0.5 |
| 15–35 | 476 | 26.8 | 1.8 |
| 1–35 | 552 | 42.9 | 1.3 |
| Age (years) | No. of SCD | Million person-years | Incidence per 100 000 person-years |
|---|---|---|---|
| 1–14 | 76 | 16.1 | 0.5 |
| 15–35 | 476 | 26.8 | 1.8 |
| 1–35 | 552 | 42.9 | 1.3 |
SCD, sudden cardiac death.
Trend analysis of sudden cardiac death in persons aged 1–35 years in Sweden during 2000–10.
Trends in sudden cardiac death by age and sex in persons aged 1–35 years in Sweden during 2000–10.
Comparison with a previous incidence study in 15- to 35-year-olds
The incidence of SCD per 100 000 person-years was 0.9 in 15- to 35-year-olds in a previous nationwide Swedish study including only cases with forensic autopsy, without co-morbidity, and with BAC < 0.1%.4 Unlike the previous study, we have now included SCD cases with no forensic autopsy identified from SDR (n = 137), with co-morbidity (n = 93) (Table 3), and with BAC ≥ 0.1% (n = 6). If these cases were excluded, then the incidence in 15- to 35-year-olds would have been the same as in the previous study (0.9 per 100 000 person-years).
Aetiology
The distribution of diagnoses by age group showed that SADS was the most frequent finding in <30 years old (123/337, 36%) (Table 2). In 30- to 35-year-olds, coronary artery disease (CAD) was the most common diagnosis (62/215, 29%). Sudden arrhythmic death syndrome was more common in women (60/156, 38%) than in men (110/396, 28%) (P= 0.02). In CAD, there was no significant difference between women (16/156, 10%) and men (65/396, 16%) (P= 0.09). Unspecific cardiomyopathy (UCM), hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), and arrhythmogenic right ventricular cardiomyopathy (ARVC) were together more common in men (107/396, 27%) than in women (27/156, 18%) (P= 0.02). In UCM, the structural changes in the myocardium described as insignificant enlargement, a few mononuclear cells and some fibrotic changes, did not meet the criteria for HCM, DCM, or ARVC. Totally 83 of 552 (15%) had a pre-known cardiac disease: congenital heart disease (CHD; n = 44); DCM (n = 9); HCM UCM; (n = 7),4 Wolff–Parkinson–White syndrome (n = 4); long QT syndrome (LQTS) (n = 4); myocardial infarction (n = 4); ischaemic cardiac disease (n = 3); ARVC (n = 2); myocarditis (n = 1); and AV-block III and pacemaker (n = 1). We found one case with LQTS who drowned and one case with SADS who had a traffic accident.
Diagnoses in sudden cardiac death in persons aged 1–35 years in Sweden during 2000–10
| SCD (n = 552) diagnoses no. (%) | Women/men (n = 156/n = 396) no. (%) | 1–14 years (n = 76) no. (%) | 15–29 years (n = 261) no. (%) | ≥30 years (n = 215) no. (%) | Mean age (24.8) years |
|---|---|---|---|---|---|
| SADS 170 (31) | 60 (38)/110 (28) | 31 (41) | 92 (35) | 47 (22) | 22.3 |
| CAD 81 (15) | 16 (10)/65 (16) | 2 (3) | 17 (7) | 62 (29) | 31.3 |
| Myocarditis 75 (14) | 23 (15)/52 (13) | 13 (17) | 46 (18) | 16 (7) | 22.0 |
| UCM 66 (12) | 16 (10)/50 (13) | 3 (4) | 32 (12) | 31 (14) | 27.4 |
| CHD 56 (10) | 21 (13)/35 (9) | 12 (16) | 27 (10) | 17 (8) | 22.4 |
| HCM 28 (5) | 5 (3)/23 (6) | 7 (9) | 15 (6) | 6 (3) | 22.0 |
| Dissecting TAA 28 (5) | 8 (5)/20 (5) | 2 (3) | 9 (3) | 17 (8) | 28.8 |
| DCM 20 (4) | 0 (0)/20 (5) | 1 (1) | 7 (3) | 12 (6) | 29.7 |
| ARVC 20 (4) | 6 (4)/14 (4) | 2 (3) | 14 (5) | 4 (2) | 24.2 |
| CAA 8 (1) | 1 (1)/7 (2) | 3 (4) | 2 (1) | 3 (1) | 20.3 |
| SCD (n = 552) diagnoses no. (%) | Women/men (n = 156/n = 396) no. (%) | 1–14 years (n = 76) no. (%) | 15–29 years (n = 261) no. (%) | ≥30 years (n = 215) no. (%) | Mean age (24.8) years |
|---|---|---|---|---|---|
| SADS 170 (31) | 60 (38)/110 (28) | 31 (41) | 92 (35) | 47 (22) | 22.3 |
| CAD 81 (15) | 16 (10)/65 (16) | 2 (3) | 17 (7) | 62 (29) | 31.3 |
| Myocarditis 75 (14) | 23 (15)/52 (13) | 13 (17) | 46 (18) | 16 (7) | 22.0 |
| UCM 66 (12) | 16 (10)/50 (13) | 3 (4) | 32 (12) | 31 (14) | 27.4 |
| CHD 56 (10) | 21 (13)/35 (9) | 12 (16) | 27 (10) | 17 (8) | 22.4 |
| HCM 28 (5) | 5 (3)/23 (6) | 7 (9) | 15 (6) | 6 (3) | 22.0 |
| Dissecting TAA 28 (5) | 8 (5)/20 (5) | 2 (3) | 9 (3) | 17 (8) | 28.8 |
| DCM 20 (4) | 0 (0)/20 (5) | 1 (1) | 7 (3) | 12 (6) | 29.7 |
| ARVC 20 (4) | 6 (4)/14 (4) | 2 (3) | 14 (5) | 4 (2) | 24.2 |
| CAA 8 (1) | 1 (1)/7 (2) | 3 (4) | 2 (1) | 3 (1) | 20.3 |
SCD, sudden cardiac death; SADS, sudden arrhythmic death syndrome; CAD, coronary artery disease; UCM, unspecific cardiomyopathy; CHD, congenital heart disease; HCM, hypertrophic cardiomyopathy; TAA, thoracic aortic aneurysm; DCM, dilated cardiomyopathy; ARVC, arrhythmogenic right ventricular cardiomyopathy; CAA, coronary artery anomaly.
Diagnoses in sudden cardiac death in persons aged 1–35 years in Sweden during 2000–10
| SCD (n = 552) diagnoses no. (%) | Women/men (n = 156/n = 396) no. (%) | 1–14 years (n = 76) no. (%) | 15–29 years (n = 261) no. (%) | ≥30 years (n = 215) no. (%) | Mean age (24.8) years |
|---|---|---|---|---|---|
| SADS 170 (31) | 60 (38)/110 (28) | 31 (41) | 92 (35) | 47 (22) | 22.3 |
| CAD 81 (15) | 16 (10)/65 (16) | 2 (3) | 17 (7) | 62 (29) | 31.3 |
| Myocarditis 75 (14) | 23 (15)/52 (13) | 13 (17) | 46 (18) | 16 (7) | 22.0 |
| UCM 66 (12) | 16 (10)/50 (13) | 3 (4) | 32 (12) | 31 (14) | 27.4 |
| CHD 56 (10) | 21 (13)/35 (9) | 12 (16) | 27 (10) | 17 (8) | 22.4 |
| HCM 28 (5) | 5 (3)/23 (6) | 7 (9) | 15 (6) | 6 (3) | 22.0 |
| Dissecting TAA 28 (5) | 8 (5)/20 (5) | 2 (3) | 9 (3) | 17 (8) | 28.8 |
| DCM 20 (4) | 0 (0)/20 (5) | 1 (1) | 7 (3) | 12 (6) | 29.7 |
| ARVC 20 (4) | 6 (4)/14 (4) | 2 (3) | 14 (5) | 4 (2) | 24.2 |
| CAA 8 (1) | 1 (1)/7 (2) | 3 (4) | 2 (1) | 3 (1) | 20.3 |
| SCD (n = 552) diagnoses no. (%) | Women/men (n = 156/n = 396) no. (%) | 1–14 years (n = 76) no. (%) | 15–29 years (n = 261) no. (%) | ≥30 years (n = 215) no. (%) | Mean age (24.8) years |
|---|---|---|---|---|---|
| SADS 170 (31) | 60 (38)/110 (28) | 31 (41) | 92 (35) | 47 (22) | 22.3 |
| CAD 81 (15) | 16 (10)/65 (16) | 2 (3) | 17 (7) | 62 (29) | 31.3 |
| Myocarditis 75 (14) | 23 (15)/52 (13) | 13 (17) | 46 (18) | 16 (7) | 22.0 |
| UCM 66 (12) | 16 (10)/50 (13) | 3 (4) | 32 (12) | 31 (14) | 27.4 |
| CHD 56 (10) | 21 (13)/35 (9) | 12 (16) | 27 (10) | 17 (8) | 22.4 |
| HCM 28 (5) | 5 (3)/23 (6) | 7 (9) | 15 (6) | 6 (3) | 22.0 |
| Dissecting TAA 28 (5) | 8 (5)/20 (5) | 2 (3) | 9 (3) | 17 (8) | 28.8 |
| DCM 20 (4) | 0 (0)/20 (5) | 1 (1) | 7 (3) | 12 (6) | 29.7 |
| ARVC 20 (4) | 6 (4)/14 (4) | 2 (3) | 14 (5) | 4 (2) | 24.2 |
| CAA 8 (1) | 1 (1)/7 (2) | 3 (4) | 2 (1) | 3 (1) | 20.3 |
SCD, sudden cardiac death; SADS, sudden arrhythmic death syndrome; CAD, coronary artery disease; UCM, unspecific cardiomyopathy; CHD, congenital heart disease; HCM, hypertrophic cardiomyopathy; TAA, thoracic aortic aneurysm; DCM, dilated cardiomyopathy; ARVC, arrhythmogenic right ventricular cardiomyopathy; CAA, coronary artery anomaly.
Activity at death in 1- to 35-year-olds
Information about activity at the time of death was obtained in 377 of 393 cases (96%) with forensic autopsy. Death occurred during daily activity (181/377, 48.0%), sleep (144/377, 38%), and physical activity (52/377, 14%) (Figure 4). Death during physical activity was more common in cases with structural findings (41/244, 17%) than in SADS (11/133, 8%) (P = 0.03). If death occurred during sleep, then SADS was a more likely diagnosis (63/133, 47%) than a diagnosis with structural findings (81/244, 33%) (P = 0.009). Death was witnessed in 40%. In cases without a forensic autopsy, 106/159 (67%) died in the emergency department or in a hospital ward, compared with 59/393 (15%) in cases in which a forensic autopsy was performed. There were nine athletes in the SCD study group: four with HCM [basketball (n = 2),2 ice-hockey and football], two with SADS (ice-hockey and floorball), two with CAD (running and skiing), and one with myocarditis (triathlon). Sudden cardiac death in athletes occurred in relation to sports activity (HCM and CAD), during sleep (SADS), or during daily activity (myocarditis).
Activity at the time of sudden cardiac death in 377 persons aged 1–35 years in Sweden during 2000–10. TAA, thoracic aortic aneurysm; ARVC, arrhythmogenic right ventricular cardiomyopathy; SADS, sudden arrhythmic death syndrome.
Body mass index and co-morbidity in 15- to 35-year-olds
Body mass index (BMI, kg/m²) was calculated in 310/340 (91%) with a forensic autopsy. The mean value of BMI in men was 26.4 (range 17.3–52.9), in women 25.6 (range 15.6–41.6), and in all 26.0, compared with 23.4 in the general Swedish population of the same age.11 The highest mean BMI were observed in DCM (28.8) and in UCM (28.3). All diagnoses, except HCM and CHD, had a mean value of BMI ≥ 25, which is the limit of overweight. Two persons with obesity were treated for obstructive sleep apnoea (OSA). Altogether 93/340 (27%) had at least one co-morbidity, and 73 of these (78%) were men. Psychiatric disease occurred in 33 cases, and half of those in SADS. Individuals with psychiatric disease were treated with antidepressants (n = 16), antipsychotic medication (n = 9), or with both drugs (n = 8). Unspecific cardiomyopathy had the highest proportion of co-morbidity (23/48, 48%) (Table 3).
Co-morbidity in relation to diagnoses in persons aged 15–35 years with sudden cardiac death in Sweden during 2000–10, all with a forensic autopsy
| Co-morbidity | SADS (n = 115) no. (%) | UCM (n = 48) no. (%) | CAD (n = 58) no. (%) | Myocarditis (n = 47) no. (%) | HCM (n = 16) no. (%) | DCM (n = 9) no. (%) | Othera (n = 47) no. (%) | All cases (n = 340) no. (%) |
|---|---|---|---|---|---|---|---|---|
| Psych. disease | 16 (14) | 7 (15) | 3 (5) | 2 (4) | 2 (13) | 0 (0) | 3 (6) | 33 (10) |
| BMIb ≥30 | 15 (13) | 12 (25) | 9 (16) | 9 (19) | 3 (19) | 3 (33) | 0 (0) | 51 (15) |
| Fatty liver | 5 (4) | 5 (10) | 2 (3) | 2 (4) | 0 (0) | 0 (0) | 2 (4) | 16 (5) |
| Diab. mellitus | 1 (1) | 2 (4) | 4 (7) | 1 (2) | 0 (0) | 1 (11) | 0 (0)/ | 9 (3) |
| DM | 4 (3) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (2)/ | 5 (1) |
| Epilepsy | 2 (2) | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) | 4 (1) |
| Cases with ≥1 co-morbidity | 28 (24) | 23 (48) | 16 (28) | 11 (23) | 5 (31) | 4 (44) | 6 (13) | 93 (27) |
| Co-morbidity | SADS (n = 115) no. (%) | UCM (n = 48) no. (%) | CAD (n = 58) no. (%) | Myocarditis (n = 47) no. (%) | HCM (n = 16) no. (%) | DCM (n = 9) no. (%) | Othera (n = 47) no. (%) | All cases (n = 340) no. (%) |
|---|---|---|---|---|---|---|---|---|
| Psych. disease | 16 (14) | 7 (15) | 3 (5) | 2 (4) | 2 (13) | 0 (0) | 3 (6) | 33 (10) |
| BMIb ≥30 | 15 (13) | 12 (25) | 9 (16) | 9 (19) | 3 (19) | 3 (33) | 0 (0) | 51 (15) |
| Fatty liver | 5 (4) | 5 (10) | 2 (3) | 2 (4) | 0 (0) | 0 (0) | 2 (4) | 16 (5) |
| Diab. mellitus | 1 (1) | 2 (4) | 4 (7) | 1 (2) | 0 (0) | 1 (11) | 0 (0)/ | 9 (3) |
| DM | 4 (3) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (2)/ | 5 (1) |
| Epilepsy | 2 (2) | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) | 4 (1) |
| Cases with ≥1 co-morbidity | 28 (24) | 23 (48) | 16 (28) | 11 (23) | 5 (31) | 4 (44) | 6 (13) | 93 (27) |
SADS, sudden arrhythmic death syndrome; UCM, unspecific cardiomyopathy; CAD, coronary artery disease; HCM, hypertrophic cardiomyopathy; DCM, dilated cardiomyopathy; DM, dystrophia myotonica.
aArrhythmogenic right ventricular cardiomyopathy (2), congenital heart disease (2), coronary artery anomaly (1), and dissecting thoracic aortic aneurysm (1).
bBMI was calculated in 310 cases.
Co-morbidity in relation to diagnoses in persons aged 15–35 years with sudden cardiac death in Sweden during 2000–10, all with a forensic autopsy
| Co-morbidity | SADS (n = 115) no. (%) | UCM (n = 48) no. (%) | CAD (n = 58) no. (%) | Myocarditis (n = 47) no. (%) | HCM (n = 16) no. (%) | DCM (n = 9) no. (%) | Othera (n = 47) no. (%) | All cases (n = 340) no. (%) |
|---|---|---|---|---|---|---|---|---|
| Psych. disease | 16 (14) | 7 (15) | 3 (5) | 2 (4) | 2 (13) | 0 (0) | 3 (6) | 33 (10) |
| BMIb ≥30 | 15 (13) | 12 (25) | 9 (16) | 9 (19) | 3 (19) | 3 (33) | 0 (0) | 51 (15) |
| Fatty liver | 5 (4) | 5 (10) | 2 (3) | 2 (4) | 0 (0) | 0 (0) | 2 (4) | 16 (5) |
| Diab. mellitus | 1 (1) | 2 (4) | 4 (7) | 1 (2) | 0 (0) | 1 (11) | 0 (0)/ | 9 (3) |
| DM | 4 (3) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (2)/ | 5 (1) |
| Epilepsy | 2 (2) | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) | 4 (1) |
| Cases with ≥1 co-morbidity | 28 (24) | 23 (48) | 16 (28) | 11 (23) | 5 (31) | 4 (44) | 6 (13) | 93 (27) |
| Co-morbidity | SADS (n = 115) no. (%) | UCM (n = 48) no. (%) | CAD (n = 58) no. (%) | Myocarditis (n = 47) no. (%) | HCM (n = 16) no. (%) | DCM (n = 9) no. (%) | Othera (n = 47) no. (%) | All cases (n = 340) no. (%) |
|---|---|---|---|---|---|---|---|---|
| Psych. disease | 16 (14) | 7 (15) | 3 (5) | 2 (4) | 2 (13) | 0 (0) | 3 (6) | 33 (10) |
| BMIb ≥30 | 15 (13) | 12 (25) | 9 (16) | 9 (19) | 3 (19) | 3 (33) | 0 (0) | 51 (15) |
| Fatty liver | 5 (4) | 5 (10) | 2 (3) | 2 (4) | 0 (0) | 0 (0) | 2 (4) | 16 (5) |
| Diab. mellitus | 1 (1) | 2 (4) | 4 (7) | 1 (2) | 0 (0) | 1 (11) | 0 (0)/ | 9 (3) |
| DM | 4 (3) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (2)/ | 5 (1) |
| Epilepsy | 2 (2) | 0 (0) | 0 (0) | 2 (4) | 0 (0) | 0 (0) | 0 (0) | 4 (1) |
| Cases with ≥1 co-morbidity | 28 (24) | 23 (48) | 16 (28) | 11 (23) | 5 (31) | 4 (44) | 6 (13) | 93 (27) |
SADS, sudden arrhythmic death syndrome; UCM, unspecific cardiomyopathy; CAD, coronary artery disease; HCM, hypertrophic cardiomyopathy; DCM, dilated cardiomyopathy; DM, dystrophia myotonica.
aArrhythmogenic right ventricular cardiomyopathy (2), congenital heart disease (2), coronary artery anomaly (1), and dissecting thoracic aortic aneurysm (1).
bBMI was calculated in 310 cases.
Discussion
The incidence of SCD during 2000–10 in 1- to 35-year-olds was 1.3 per 100 000 person-years with a decreasing trend in women. During the 11-year survey period, 2.5 times more men than women died from SCD. Sudden arrhythmic death syndrome (31%) and CAD (15%) were the most frequent diagnoses. To our knowledge, this is the most comprehensive study of SCD in a nationwide setting, including 42.9 million person-years.
To be able to determine the incidence of SCD and frequency of different disturbances causing these fatalities, the method of case selection is fundamental. We primarily included all fatalities in this age group and thereafter excluded non-sudden and non-cardiac deaths. Cases with SUDS (no autopsy performed and no apparent cause) were excluded because among them there are epilepsy, diabetes, intoxications, and subarachnoid haemorrhage that otherwise wrongly would be included. This might cause a certain underestimation of SCD. Incidence studies that include all ages often use out-of-hospital resuscitation registries to select the cases, and SUDS and SIDS (sudden infant death syndrome) are usually included. This leads to higher incidence figures; in a study from Germany, the incidence of SCD was 7.4 per 100 000 person-years in age group 0–35 years.12
We have used the SCD definition from ESC guidelines, i.e. included cases with a post-mortem examination or a pre-known cardiac disease.6 The SCD incidence found in this study is in the lower range of comparable studies. This can partly be explained by a strict exclusion of non-SCD cases. Incidence data in the literature vary between 0.7 and 3.7 per 100 000 person-years.1–5 The inclusion/exclusion criteria are rarely reported in detail. The study with the highest incidence included cocaine addicts.3 We call for a discussion on the difficulties of including accurate cases, e.g. how to assess if death is sudden and unexpected in cases with chronic heart disease, how to assess if cases with diseases that can cause sudden death (e.g. epilepsy and diabetes) should be included. Different approaches when it comes to abuse without acute signs of intoxication probably cause major differences in incidence rates of SCD, as abuse accounts for a high percentage of sudden death in this age group. We have chosen not to include abusers because of the high risk of drug-related death. In a few cases, SCD may falsely be classified as accidents. Given the low frequency of SCD in 1- to 35-year-olds and the high autopsy rate of accidents (86%), it is unlikely that the exclusion of accidents will influence the total outcome. We included two cases that after thorough investigation were considered to be caused by heart disease.
In the present study, the incidence in 15- to 35-year-olds was 1.8 per 100 000 person-years. This is higher compared with incidence data from 1992 to 1999 in this age group (0.9 per 100 000 person-years), but had been the same if the same inclusion criteria had been used.4 The total incidence of SCD has not decreased in Sweden despite medical advances and great efforts in the society. In Sweden, three million people are now trained in cardiopulmonary resuscitation (CPR), and a recent study showed that bystander CPR more than doubled the 30-day survival rate after an out-of-hospital cardiac arrest.9 We saw some downward trend in women during the study period. In this study, we have included co-morbidity that was four-fold more common among men, which may be the reason why the incidence was not declining among men. It is known that obesity and psychiatric disease entail a higher risk of SCD. In a Danish study, patients with prior psychiatric hospital contact had a four-fold increased risk of SCD.13 Obesity leads to abnormal cardiac function and increased electrical instability in the myocardium, which may cause SCD.14
In accordance with other studies, SADS was the most frequent cause of death.2,4 Sudden arrhythmic death syndrome can partly be due to inherited cardiac diseases such as channelopathies and cardiomyopathies with changes not detected by autopsy.15,16 Antipsychotic drugs and antidepressants may also in therapeutic doses prolong the QT duration and lead to Torsade de pointes and SCD.17 Sudden arrhythmic death syndrome was more common in women (P = 0.02). Tester et al. showed that one-third of young SADS cases had verified genetic causes and that LQTS-associated mutations in SCD were more common in women.15 The frequency of CAD, myocarditis, and UCM was almost the same (12–15%). In our experience, the entity UCM was used when there were slight and undefined histological findings and probably some similar cases in other studies are labelled as unspecified myocarditis. There may be an association with obesity. A Finnish study showed that cardiomyopathy was the most frequent cause of death in cases with extreme obesity (BMI > 40 kg/m2), followed by CAD.18 The low frequency of HCM, ARVC, and DCM (3–5%) compared with that in our previous study4 may be due to better medical treatment and increased use of implantable cardioverter defibrillator. It also becomes more common with a family investigation in these cases.
To a high degree (almost 40%), death occurred during sleep, also reported in other studies.2,4 The mechanisms of nocturnal death is largely unknown. People with OSA have a higher risk of dying in their sleep.19 In this study, two had OSA, but undiagnosed cases are common. In long QT type 3 and Brugada syndrome, it is believed that an increased vagal activity and/or a decreased sympathetic drive during sleep trigger a malignant ventricular arrhythmia.20 Similar mechanisms may be involved when SCD is triggered by drugs affecting the conduction system.17 Deaths related to physical activity were uncommon (14%). There were few athletes in the study (n = 9), which may have several causes, but is partly due to pre-participation screening programmes in many sports in Sweden.
In total, we have studied SCD for 19 years in Sweden.4 We have used a selection process that we believe will find ‘true’ cases of SCD. This is a prerequisite for continued research and prevention of these deaths.
Strengths and weaknesses of the study
The studied cohort represents cases from the entire country during 11 years, which rules out the common problem with a selected group of individuals. The selection of included subjects is thoroughly reported. The high number of autopsies and the totally low risk of missed cases are important factors for the strength of the study. We may miss some individuals with SCD in cases without autopsy and due to the lack of information about the time between symptom and death. Also, there was no re-examination of the histological analyses associated with the autopsies.
Conclusion
The incidence of SCD during 2000–10 in 1- to 35-year-olds was 1.3 per 100 000 person-years. During the 11-year survey period, the incidence showed a decreasing trend over time in women. In 15- to 35-year-olds, the incidence was the same as during 1992–99 if the same inclusion criteria had been used. Co-morbidities such as obesity and psychiatric diagnoses were common, especially among men.
Funding
This study was supported by grants from the Medical faculty at Umeå University and from the Norrbotten County Council.
Conflict of interest: none declared.
