Sudden Cardiac Death in Patients With Type 1 Versus Type 2 Diabetes

Abstract

Objective:

To investigate the association between type 1 diabetes mellitus (T1D) and type 2 diabetes mellitus (T2D) with risk of sudden cardiac arrest (SCA).

Methods:

In a prospective community-based study of SCA from February 1, 2002, through November 30, 2019, we ascertained 2771 cases age 18 years of age or older and matched them to 8313 controls based on geography, age, sex, and race/ethnicity. We used logistic regression to evaluate the independent association between diabetes, T1D, T2D, and SCA.

Results:

Patients had a mean age of 64.5±15.9 years, were 33.3% female and 23.9% non-White race. Overall, 36.7% (n=1016) of cases and 23.8% (n=1981) of controls had diabetes. Among individuals with diabetes, the proportion of T1D was 6.5% (n=66) among cases and 2.0% among controls (n=40). Diabetes was associated with 1.5-times higher odds of SCA. Compared with those without diabetes, the odds ratio and 95% CI for SCA was 4.36 (95% CI, 2.81 to 6.75; P<.001) in T1D and 1.45 (95% CI, 1.30 to 1.63; P<.001) in T2D after multivariable adjustment. Among those with diabetes, the odds of having SCA were 2.41 times higher in T1D than in T2D (95% CI, 1.53 to 3.80; P<.001). Cases of SCA with T1D were more likely to have an unwitnessed arrest, less likely to receive resuscitation, and less likely to survive compared with those with T2D.

Conclusion:

Type 1 diabetes was more strongly associated with SCA compared with T2D and had less favorable outcomes following resuscitation. Diabetes type could influence the approach to risk stratification and prevention of SCA.

Diabetes mellitus is independently associated with an increased risk of sudden cardiac arrest (SCA).^1–5 In the United States alone, more than 350,000 lives are lost annually due to SCA.⁶ The steady and substantial rise of diabetes burden in the United States and globally^7,8 lends further urgency to improving the understanding of the diabetes-SCA link, with the overall goal of improving prevention in high-risk subgroups with diabetes.

Although type 1 diabetes (T1D) and type 2 diabetes (T2D) share multiple common pathophysiological consequences, there is evidence to indicate distinct etiologies, disease courses, and demographic differences. Both T1D and T2D are established risk factors for cardiovascular death, early all-cause mortality, and cardiovascular disease.^9–11 However, patients with T1D have a higher risk of myocardial infarction, heart failure hospitalization, atrial fibrillation, and all-cause and cardiovascular mortality compared to those with T2D.^9,12

Previous studies have reported 2- to 10-fold increased risk of SCA with diabetes,¹³ but did not distinguish between types of diabetes. Although T2D accounts for 90% to 95% of all cases of diabetes,¹⁴ whether T1D and T2D confer a similar risk of SCA remains unclear. Those with T1D are likely to be younger at the time of death with major effects on years of potential life lost.¹⁵ Furthermore, small studies have reported cases of “dead-in-bed” syndrome in T1D, where individuals are in apparently good health, only to be found deceased the following morning without an identifiable cause of death.^16–18 Similarly, there is a lack of data regarding SCA outcomes following resuscitation in T1D vs T2D. A detailed understanding of SCA circumstances, resuscitation variables, and survival outcomes are likely to advance the understanding of immediate management following SCA in those with diabetes.

Our primary aim was to determine and compare the association of T1D and T2D with SCA. We hypothesized that, given the differences in the pathophysiology, clinical presentation, and management, the likelihood of having SCA is greater in T1D compared with T2D. Our secondary aim was to examine the SCA circumstances, resuscitation variables and survival outcome of SCA in T1D vs T2D vs the population without diabetes.

METHODS

Study Population

We prospectively identified SCA cases from the ongoing community-based Oregon Sudden Unexpected Death Study (SUDS) capturing events that occurred between February 1, 2002, and November 30, 2019, in the Portland, Oregon, metropolitan area. We have previously published details on SCA ascertainment.^19,20 Briefly, we identified all incident cases of SCA cases via multiple sources, including fire department, ambulance services, local hospital emergency departments, and the county medical examiner’s office. We defined SCA as a sudden and unexpected pulseless condition of likely cardiac etiology if witnessed, and as a sudden unexpected death within 24 hours of last having been seen in a usual state of health if unwitnessed. All identifiable noncardiac causes of death, including trauma, drug overdose, pulmonary embolism, cerebrovascular accident, and chronic terminal illness, were excluded. For this analysis, we included patients who were age 18 years and older, with available medical records and complete information on body weight, height, and race/ethnicity.

We selected controls from patients enrolled in the Kaiser Permanente Northwest Region health maintenance organization between 2002 and 2019. Individuals were eligible for inclusion if their age was 18 years or older, with complete information on body weight, height, and race/ethnicity, and if they were residents of the Portland, Oregon, metropolitan area. Stratified random sampling was used to select controls from the eligible patient population to achieve a 1:3 case-control ratio in each stratum. The size of each stratum was determined by the number of SCA cases in each sex, age (5-year intervals), and race/ethnicity (Non-Hispanic White, Black, Hispanic, and “Other”) group. This study was approved by the institutional review boards at Oregon Health and Science University and Cedars-Sinai Medical Center. All survivors of SCA provided informed consent, and this requirement was waived for nonsurvivors.

We captured diabetes from medical records for both cases and controls, and used standard definitions according to the Standards of Medical Care in Diabetes — 2019 from the American Diabetes Association).²¹ We classified any diabetes (yes/no) as a diagnosis of T1D, T2D, diabetes mellitus (unspecified type), or based on hemoglobin A1C greater than or equal to 6.5%. We further classified patients into specific T1D or T2D categories: All T1D patients had a documented history of T1D and were using insulin therapy.²² The T2D category included those with a documented history of T2D and remaining subjects with diabetes (unspecified type), except those with other documented subtypes of diabetes (gestational, monogenic, or secondary diabetes).²¹

We used information from medical records to define covariates which have been previously described.^19,23,24 For this analysis, we harmonized variable definitions between cases and controls and included the following: body mass index (BMI), chronic heart failure, coronary artery disease (CAD), hypertension, hyperlipidemia, chronic obstructive pulmonary disease, asthma, chronic kidney disease (CKD), and atrial fibrillation/flutter. Race and ethnicity data for SCA cases were obtained from death certificates, medical examiner reports, medical records, and emergency medical service pre-hospital care reports. Race/ethnicity was categorized into four categories: White; Black; Hispanic; and “Other,” which included Asian, American Indian/Alaska Native, and Hawaiian/Pacific Islander. “Hispanic” included any individual reporting Hispanic ethnicity, regardless of race. Race/ethnicity in controls was captured from administrative and medical records. To match cases to controls, we used the four-category race/ethnicity variable. During statistical analysis we combined the Hispanic and Other categories due to small numbers, into one Other category.

Statistical Analysis

We performed univariate comparisons using Pearson χ² tests for categorical variables and independent samples Student’s t test for continuous variables. We used an unconditional logistic regression model to report odds ratios (ORs) and 95% CIs comparing the risk of SCA in those with diabetes compared with those without diabetes. In addition to presenting unadjusted ORs, we adjusted for the variables of age, sex, race/ethnicity, and BMI in model 1; and in model 2 we additionally adjusted for chronic heart failure, CAD, hypertension, hyperlipidemia, chronic obstructive pulmonary disease, asthma, CKD, and atrial fibrillation/flutter. We tested for multiplicative interactions by sex, race, and presence of CAD. For the second analysis limited to those with SCA, we used logistic regression models with the same adjustment for covariates to report the association between diabetes status and resuscitation variables. All statistical analyses were performed using SAS 9.4 (SAS Institute Inc) or STATA version 15.0 (StataCorp LLC) and 2-sided statistical tests with P less than .05 were considered significant.

RESULTS

We matched 2771 SCA cases to 8313 controls. Overall, patients had a mean age of 64.5±15.9 years, 33.3% were female, and 23.9% were a non-White race. Table 1 displays clinical characteristics by case status. Compared with controls, those with SCA were more likely to have heart failure, CAD, hypertension, CKD, and atrial fibrillation, and less likely to have hyperlipidemia. In total, 36.7% (n=1016) of cases and 23.8% (n=1981) of controls had diabetes. Among individuals with diabetes, the proportion of T1D was 6.5% (n=66) among cases and 2.0% among controls (n=40; P<.001).

TABLE 1.

Characteristics of SCA Cases and Non-SCA Controls^a,b

	SUDS SCA cases (N=2771)	Kaiser controls (N=8313)	P
Matched variables
Age, y	64.9±16.1	64.3±15.9	.07
Male sex	1848 (66.7)	5544 (66.7)	1.00
Race			.78
White	2118 (76.4)	6317 (75.9)
African-American	255 (8.1)	710 (8.5)
Other	428 (15.5)	1286 (15.5)
Medical history
Diabetes	1016 (36.7)	1981 (23.8)	<.001
Type 1 diabetesc	66 (6.5)	40 (2.0)	<.001
Type 2 diabetesc	950 (93.5)	1941 (98.0)	<.001
Body mass index, kg/m2	30.4±9.7	29.3±6.5	<.001
Chronic heart failure	1001 (36.1)	849 (10.2)	<.001
Coronary artery disease	1204 (43.4)	1653 (19.9)	<.001
Hypertension	1867 (67.4)	4733 (56.9)	<.001
Hyperlipidemia	1251 (45.2)	4190 (50.4)	<.001
COPD or asthma	710 (25.6)	2032 (24.4)	.21
Chronic kidney disease	653 (25.0)	1277 (15.4)	<.001
Atrial fibrillation/flutter	647 (23.4)	980 (11.8)	<.001

^aCOPD, chronic obstructive pulmonary disease; SCA, sudden cardiac arrest; SUDS, Sudden Unexpected Death Study.

^bValues are mean ± SD or n (%).

^cProportion of all subjects with diabetes.

Table 2 shows the clinical characteristics of those with diabetes, stratified by type of diabetes and case-control status. In unadjusted comparisons among those with T1D, those with SCA (compared with controls) had a higher BMI and a higher prevalence of heart failure, CAD, and atrial fibrillation. In unadjusted comparisons among those with T2D compared with controls, SCA patients were slightly younger; had higher BMI; higher prevalence of heart failure, CAD, CKD, and atrial fibrillation; and lower prevalence of hyperlipidemia.

TABLE 2.

Characteristics of Diabetic Subjects, Stratified by SCA Cases and Non-SCA Controls^a,b

	Type 1 diabetes			Type 2 diabetes
	SUDS SCA cases	Kaiser controls	P	SUDS SCA cases	Kaiser controls	P
n	66	40		950	1941
Age, y	60.1±15.4	62.3±17.5	.49	68.0±13.1	69.8±12.3	<.001
Age categories, y			.92			.001
18–40	8 (12.1)	5 (12.5)		22 (2.3)	16 (0.8)
41–64	29 (43.9)	16 (40.0)		341 (35.9)	648 (33.4)
65–105	29 (43.9)	19 (47.5)		587 (61.8)	1277 (65.8)
Male sex	41 (62.1)	29 (72.5)	.27	608 (64.0)	1370 (70.6)	<.001
Non-White race	23 (34.9)	8 (20.0)	.10	254 (26.7)	581 (29.9)	.07
Body mass index, kg/m2	33.0±8.3	28.1±5.7	<.001	33.1±10.5	31.6±7.1	<.001
Chronic heart failure	36 (54.6)	3 (7.5)	<.001	469 (49.4)	358 (18.4)	<.001
Coronary artery disease	34 (51.5)	12 (30.0)	.03	540 (56.8)	650 (33.5)	<.001
Hypertension	53 (80.3)	32 (80.0)	.97	811 (85.4)	1607 (82.8)	.08
Hyperlipidemia	39 (59.1)	27 (67.5)	.39	593 (62.4)	1479 (76.2)	<.001
COPDa or asthma	17 (25.8)	8 (20.0)	.50	278 (29.3)	557 (28.7)	.75
Chronic kidney disease	42 (63.6)	18 (45.0)	.06	395 (41.6)	547 (28.2)	<.001
Atrial fibrillation/flutter	24 (36.4)	7 (17.5)	.04	270 (28.4)	330 (17.0)	<.001

^aCOPD, chronic obstructive pulmonary disease; SCA, sudden cardiac arrest; SUDS, Sudden Unexpected Death Study.

^bValues are mean ± SD or n (%).

We found significant associations between diabetes and SCA (Table 3). In the fully adjusted analysis, the odds of SCA were 1.52 times higher in patients with any diabetes type compared with those without diabetes (adjusted OR, 1.52; 95% CI, 1.36 to 1.70; P<.001). Comparisons of diabetes stratified by type with controls showed that the OR (95% CI) for SCA was 4.36 (95% CI, 2.81 to 6.75; P<.001) in those with T1D and 1.45 (95% CI, 1.30 to 1.63; P<.001) in those with T2D. Within the diabetes subgroup (n=2997), the odds of having SCA were 2.4 times higher in T1D than in T2D (OR, 2.41; 95% CI, 1.53 to 3.80; P<.001). We also evaluated myocardial repolarization from archived electrocardiograms (ECGs). Archived ECGs were available for 43.2% of SCA cases. Myocardial repolarization indices measured as QTc and JTc intervals on the resting 12-lead ECG were not found to be different between T1D (n=32) and T2D SCA cases (n=485, data not shown). Cases of SCA with diabetes (n=517) had longer JTc interval compared with cases without diabetes (n=659; mean 359±41 ms vs 352±42 ms; P=.007). There were no significant differences in QTc interval (467±41 ms vs 462±143 ms; P=.46).

TABLE 3.

Association Between Diabetes and Sudden Cardiac Arrest^a

	Odds ratio (95% CI) for SCA (vs non-SCA)
	Unadjusted	P	Model 1b	P	Model 2c	P
Full cohort
No diabetes	1.00 (reference)		1.00 (reference)		1.00 (reference)
Type 1 diabetes	5.95 (4.01–8.85)	<.001	5.84 (3.93–8.69)	<.001	4.36 (2.81–6.75)	<.001
Type 2 diabetes	1.77 (1.61–1.94)	<.001	1.69 (1.53–1.86)	<.001	1.45 (1.30–1.63)	<.001
Combined type 1 or type 2 diabetes	1.85 (1.69–2.04)	<.001	1.78 (1.61–1.96)	<.001	1.52 (1.36–1.70)	<.001
Diabetics only
Type 1 diabetes	3.37 (2.26–5.03)	<.001	3.20 (2.13–4.80)	<.001	2.41 (1.53–3.80)	<.001
Type 2 diabetes	1.00 (reference)		1.00 (reference)		1.00 (reference)

^aSCA, sudden cardiac arrest.

^bModel 1: Adjusted for age, sex, race/ethnicity, and body mass index.

^cModel 2: Adjusted for model 1 + chronic heart failure, coronary artery disease, hypertension, hyperlipidemia, chronic obstructive pulmonary disease, asthma, chronic kidney disease, and atrial fibrillation/flutter.

The associations between diabetes and SCA differed by sex and by CAD, with a stronger association in females vs males and in those without CAD vs those with a history of CAD (Table 4). Females with diabetes had 2.1-times higher odds of SCA compared with females without diabetes, whereas males with diabetes had 1.3 times higher odds of SCA compared with males without diabetes (P for interaction <.001). This association was stronger in those with T1D than in those with T2D, although our estimates may be imprecise due to small numbers of sex-stratified T1D SCA cases. In those without a history of CAD, the presence of diabetes was associated with 1.6-fold higher odds of SCA; whereas in those with CAD the presence of diabetes was associated with 1.3-fold higher odds of SCA (P for interaction <.001). Similar to the main results and for both those with CAD and without CAD, those with T1D had a stronger association with SCA compared with those with T2D. Associations were similar among the three race/ethnicity categories and interaction by race was not significant.

TABLE 4.

Association Between Diabetes and Sudden Cardiac Arrest, Stratified by Sex and Coronary Artery Disease^a,b

	Odds ratio (95% CI) for SCA (vs non-SCA)
Stratified by sex	Females	P	Males	P	P for interaction
No diabetes	1.00 (reference)		1.00 (reference)
Type 1 diabetes	6.70 (3.06–14.7)	<.001	3.54 (2.08–6.03)	<.001	.0002
Type 2 diabetes	1.99 (1.64–2.43)	.01	1.26 (1.10–1.45)	.006
Combined type 1 or type 2 diabetes	2.09 (1.72–2.54)	<.001	1.32 (1.15–1.51)	.0001	<.001
Stratified by CAD	No CAD	P	CAD	P	P for interaction
No diabetes	1.00 (reference)		1.00 (reference)
Type 1 diabetes	4.55 (2.62–7.90)	<.0001	3.18 (1.54–6.56)	.005	.0009
Type 2 diabetes	1.53 (1.32–1.78)	.03	1.27 (1.07–1.52)	.04
Combined type 1 or type 2 diabetes	1.61 (1.49–1.86)	<.001	1.31 (1.10–1.56)	.003	<.001

^aCAD, coronary artery disease; SCA, sudden cardiac arrest.

^bModel adjusted for age, sex, race/ethnicity, body mass index, chronic heart failure, CAD, hypertension, hyperlipidemia, chronic obstructive pulmonary disease, asthma, chronic kidney disease, and atrial fibrillation/flutter.

For our secondary aim, we limited our sample to the 2771 SCA cases. In Table 5, we have listed the health characteristics by diabetes status at the time of SCA. Those with T1D or T2D had a worse cardiovascular profile at the time of SCA compared with those without diabetes. Those with T1D were younger and more likely to have CKD compared with those with T2D and those without diabetes.

TABLE 5.

Characteristics of Sudden Cardiac Arrest Cases, Stratified by Diabetes Status^a,b

	Type 1 diabetes	Type 2 diabetes	No diabetes
n	66	950	1755
Age, y	60.1±15.4c,d	68.0±13.1c	63.5±17.3
Age categories, y
18–40	8 (12.1)d	22 (2.3)d	167 (9.5)
41–64	29 (43.9)d	341 (35.9)c	739 (42.1)
65–105	29 (43.9)d	587 (6l.8)c	849 (48.4)
Male sex	41 (62.1)	608 (64.0)c	1199 (68.3)
Non-White race	23 (34.9)c	254 (26.7)c	376 (21.4)
Body mass index, kg/m2	33.0±8.3c	33.1±10.5c	28.9±8.9
Chronic heart failure	36 (54.6)c	469 (49.4)c	496 (28.3)
Coronary artery disease	34 (51.5)c	540 (56.8)c	630 (35.9)
Hypertension	53 (80.3)c	811 (85.4)c	1003 (57.2)
Hyperlipidemia	39 (59.1)c	593 (62.4)c	619 (35.3)
COPDa or asthma	17 (25.8)	278 (29.3)c	415 (23.7)
Chronic kidney disease	42 (63.6)c,d	395 (41.6)c	256 (14.6)
Atrial fibrillation/flutter	24 (36.4)c	270 (28.4)c	353 (20.1)

^aCOPD, chronic obstructive pulmonary disease.

^bValues are mean ± SD or n (%).

^cP<.05 compared to those with no diabetes.

^dP<.05 between diabetes type.

Of the 2771 SCA events, 59.1% of individuals with T1D had an unwitnessed event, significantly higher than those with T2D (43.8%) and those without diabetes (45.0%), as depicted in Figure 1. Following this same pattern, those with T1D had a significantly lower rate of attempted resuscitation (72.8%) compared with those with T2D (79.8%) and those without diabetes (76.6%). Limiting the data to only those with resuscitation attempted, we observed that those with T1D had the highest percentage of nighttime arrest, the lowest percentage to present with a shockable rhythm, and were less likely to survive to hospital discharge (10.4% survival compared with 13.2% and 17.6%, respectively) compared with those with T2D and those without diabetes. Figure 2A shows the adjusted ORs for resuscitation variables. Individuals with T1D were two times more likely (P=.008) to have an unwitnessed SCA compared with those without diabetes, and 1.7 times more likely (P=.04) compared with those with T2D. Those with T1D were also half as likely (P=.04) to have resuscitation attempted and were approximately two times more likely to die before hospital discharge compared with those with T2D and without diabetes. Descriptively, those with T1D tended to have SCA between the hours of 6 am and 10 am compared with those with T2D and no diabetes (Figure 2B).

FIGURE 1.

Resuscitation variables for sudden cardiac arrest cases. SCA, sudden cardiac arrest; VF, ventricular fibrillation; VT, ventricular tachycardia.

FIGURE 2.

Association between diabetes and resuscitation variables for sudden cardiac arrest cases. A, Odds ratios and CI from a model adjusted for age, sex, and race/ethnicity, body mass index, and comorbidities (chronic heart failure, coronary artery disease, hypertension, hyperlipidemia, chronic obstructive pulmonary disease or asthma, chronic kidney disease, and atrial fibrillation/flutter). B, Time of sudden cardiac arrest by diabetes status. SCA, sudden cardiac arrest.

DISCUSSION

In this community-based case-control study, adults with diabetes had higher odds of SCA than those without diabetes. Among individuals with diabetes, T1D was more strongly associated with SCA compared with T2D. These associations remained significant after adjusting for common cardiovascular risk factors, and the association was stronger in females that it was in males. Additionally, we found that those with T1D had less favorable resuscitation circumstances and worse survival outcomes compared with those with T2D and no diabetes. Individuals with T1D were more likely to have an unwitnessed cardiac arrest, no resuscitation attempted, and a lower survival rate. These findings have important implications for sudden death prevention.

Previous studies have reported individuals with diabetes have a 2- to 10-fold increased risk of SCA compared with those without diabetes,¹³ and that SCA was the leading cause of death among younger persons (1–35 years of age) with diabetes,¹⁵ but have failed to distinguish between the types of diabetes. Our study found 1.5-fold higher odds of SCA in adults with any diabetes, but also found that these odds were nearly 2.5 times higher in T1D compared with those with T2D. These findings in the context of SCA follow a pattern similar to other cardiovascular-focused studies that report differential risks among patients with diabetes; those with T1D have a higher risk for myocardial infarction, hospitalized heart failure, atrial fibrillation, and all-cause and cardiovascular mortality compared with those with T2D.^9,12 Our finding that SCA risk is higher in females with diabetes, particularly females with T1D, appears to be novel and previously unreported. Our number of 66 cases of SCA in those with T1D is a robust number of cases compared with previous observational studies. Future studies should include larger, sex-stratified analyses to confirm this finding.

Several mechanisms have been proposed to account for the increased risk of SCA in those with diabetes, and it is likely that risk is conferred both directly and indirectly. Diabetes is independently associated with increased SCA risk.^1–5,25 However, patients with diabetes also have a significantly increased prevalence of other SCA risk factors including obesity,²⁶ hypertension,²⁷ renal disease,²⁸ and coronary disease.²⁹ In our analysis, the prevalence of these risk factors were higher in SCA cases compared with controls, but were similar among T1D and T2D SCA cases, despite the T1D patients being younger on average at the time of the event. This may indicate that those with T1D are developing severe disease such as premature heart failure and CAD, at a younger age compared with both those with T2D and those without diabetes. Alternatively, there may be additional pathways to SCA risk that are unique to T1D. In our analysis, the association between diabetes and SCA remained when stratified by CAD status, and was stronger in the no CAD group in both T1D and T2D, suggesting that there are other pathways between diabetes and SCA. Taken together, our findings suggest that the higher SCA risk in T1D vs T2D is not entirely explained by differences in clinical profile. Future studies should assess and compare the effectiveness of primary prevention implantable cardioverter-defibrillators in those with T1D and T2D, especially those identified to be at higher risk, such as females with T1D.

There are several potential differences in pathophysiology between T1D and T2D that could suggest differential risk of SCA. Co-conditions common in diabetes such as hypoglycemia, cardiac autonomic neuropathy, and polypharmacy are all associated with abnormally prolonged ventricular repolarization measured as QTc,^13,30 which is an important predictor of SCA risk for all ages.³¹ A potential exaggeration of this effect in T1D has not been evaluated. Higher rates of complicated diabetes, especially when adjusted for age, have specific implications for potential role of autonomic abnormalities that could lead to ventricular arrhythmias. We have previously shown a loss of the interaction between heart rate variability and heart rate in those with diabetes who develop SCA.³² Lastly, in younger patients, genetic contribution to disease is generally greater, and modes of inheritance of T1D vs T2D also differ. It is possible that those with T1D and T2D who carry variants in repolarization controlling genes such as HERG (LQTS2) and KCNQ1 (LQTS2) are at increased risk of SCA¹³; however, this has not yet been studied. More work is needed in the specific context of a T1D/SCA genetic connection.

Our study found that SCA resuscitation circumstances and subsequent survival differed in those with T1D compared with both T2D as well as those without diabetes. Although our study contained small numbers of patients with T1D, indicators for worse survival outcomes were clear and included less witnessed arrest, less resuscitation attempted, and a higher prevalence of a nonshockable initial rhythm. Although those with T1D tended to have more nighttime SCA, they were much more likely to experience an event between the hours of 6 am and 10 am compared with T2D and patients without diabetes. Time of arrest and initial rhythm were captured only in those with resuscitation attempted, and should be interpreted in the context of observing that those with T1D were more likely to experience an unwitnessed arrest and less likely to have resuscitation attempted.

Unique to T1D, small studies have reported cases of dead-in-bed syndrome, defined as death in a generally well individual with T1D in the absence of cardiovascular disease who suffers an unwitnessed death, overnight, and the autopsy provides no clear cause of death.^16–18 A review from Australia found two-thirds of sudden unexpected deaths, or approximately 14% of all-cause mortality in patients with T1D younger than the age 40 years to be attributed to dead-in-bed syndrome.³³ Tsujimoto et al³⁴ reported that during episodes of severe hypoglycemia, the incidence of abnormal QT prolongation was significantly higher in the early morning (4 am-10 am) than at other times of the day, irrespective of diabetes status. Although the exact causes and mechanisms leading to dead-in-bed syndrome are unknown, there is growing evidence that both nocturnal hypoglycemia and autonomic neuropathy may play a role.³³ Serum glucose levels are highly modified by the SCA event itself making it difficult to ascertain the role of nocturnal hypoglycemia. Autonomic neuropathy in diabetes can cause impaired parasympathetic activity resulting in sympathetic predominance and subsequent high resting heart rate, reduction in diurnal heart rate variation, and prolonged QTc interval.^16,18,33 Although dead-in-bed syndrome is more common in males,^16–18 we have previously found that females were more likely to have nighttime SCA events and that these are associated with decreased survival.³⁵ Studies with larger numbers are needed to investigate these findings, and they should include sex-specific analyses.

Study Limitations

Our findings should be interpreted in the context of an observational, case-control study. This is a study of SCA compared with controls from a single large US community and these findings may not be generalizable to other geographic regions. Furthermore, we have a relatively small number of people with T1D whose health care circumstances and management may not be generalizable to the entire United States. We were unable to collect all the socioeconomic variables and therefore could not incorporate these in our analysis. In a subgroup of SCA cases, we may have incomplete medical record information because SCA can present without warning and as the first manifestation of illness. We had relatively small numbers of individuals with T1D, and so did not have power to perform sex-specific analyses for resuscitation circumstances. All patients were 18 years or older, thus excluding a portion of the T1D population. For those with diabetes, we did not have information on the duration of diabetes, the level of diabetes control in each individual or current hemoglobin A1c levels. Our study strengths relate to the availability of detailed prehospital emergency medical service records, including if witnessed, the time of cardiac arrest, and initial presenting rhythm.

CONCLUSION

Individuals with T1D and T2D have a higher risk of SCA than those without diabetes. In adults with diabetes, T1D had higher odds of SCA compared with those with T2D. Additionally, we found that those with T1D had less favorable resuscitation circumstances and outcomes compared with those with T2D and to those without diabetes. Type of diabetes could determine the approach to prevention of sudden death in this condition and larger studies in diverse regions are needed to confirm and further investigate these findings.

Abbreviations and Acronyms:

BMI

body mass index

CAD

coronary artery disease

CKD

chronic kidney disease

OR

odds ratio

SCA

sudden cardiac arrest

SCD

sudden cardiac death

T1D

type 1 diabetes

T2D

type 2 diabetes