Sudden Cardiac Death as First Manifestation of Heart Disease in Women: The Oregon Sudden Unexpected Death Study 2004-2016

The incidence of sudden cardiac death (SCD) declined significantly between 1950 and 1999,¹ mirroring a decline in overall cardiovascular mortality. Recently, however, cardiovascular mortality rates have increased.² To assess sex-specific trends in SCD incidence, we evaluated out-of-hospital SCD cases among adults (age ≥ 18y) between February 1, 2004 and January 31, 2016, from the Oregon SUDS (Oregon Sudden Unexpected Death Study), an ongoing community-based study of SCD in the Portland, Oregon metropolitan area. Methods for this study have been published previously.³ This study was approved by the institutional review boards of all participating institutions.

This analysis included all cases attended by the region’s two-tiered emergency medical services system, both deceased subjects (on-scene or after attempted resuscitation) and survivors of cardiac arrest. Cases were adjudicated using all available records, emergency medical services records, and medical examiner’s records. Cases were identified as SCD when a sudden, unexpected pulseless condition of likely cardiac origin occurred (if witnessed); or if unwitnessed, a sudden unexpected death in an individual seen in normal health within 24 h of arrest.

We obtained Multnomah County, Oregon annual population age-stratified denominators from the US Census Bureau.⁴ We divided the 12-y period from February 1, 2004, to January 31, 2016, into three 4-y intervals: 2004 to 2007, 2008 to 2011, and 2012 to 2015. Sex-specific SCD incidence was calculated year by year and in the 3 time periods. Last, we evaluated rates of SCD with known heart disease and those with SCD as first manifestation of heart disease. Patients with available records from regional health systems (80% of all cases) were included for analysis of past medical history. Diagnosis of prearrest cardiac disease was based on review of each patient’s prearrest records, and autopsy if available. We defined “SCD as first manifestation of heart disease” to include all patients who had none of the following conditions or procedures before SCD: cardiomyopathy, heart failure, severe valvular heart disease, moderate or severe left ventricular hypertrophy, atrial fibrillation, sustained ventricular arrhythmias, permanent pacemaker, congenital heart disease, cardiac surgery, obstructive coronary artery disease, previous myocardial infarction, or percutaneous coronary or valve interventions.

Analyses were performed using SAS 9.4 (SAS Institute Inc., Cary, NC). Rates and their 95% CIs were calculated using the PROC GENMOD function with a Poisson distribution. Differences in rates across the 3 time periods and their 95% CIs were calculated using the PROC NLMIXED function with a Poisson distribution. All tests were 2-tailed; P<0.05 indicated significance.

From February 2004 through January 2016, there were 2,938 SCDs among Multnomah County, Oregon residents age ≥18 y (average adult population, 572 000). Thirty-seven percent were female; the age was 70 ± 16 y among women and 65 ± 15 y among men. During this time, the incidence of SCD in Multnomah County, Oregon, exhibited a U-shaped pattern, with a nadir in 2011. In a year-by-year analysis, SCD rates were significantly lower in years 2007, 2008, 2010, 2011, and 2012 relative to 2004, with 95% CIs outside the bounds of the 95% CI for the rate in 2004 (Figure). Regarding the rebound, rates in 2013 and 2015 were significantly higher than the rate in 2011, with non-overlapping 95% CIs.

Figure. Sudden cardiac death incidence rates in women and men.

Panel A. Annual sudden cardiac death incidence rates in Multnomah County, Oregon among women and men.

Panel B. Colored bars show the incidence of sudden cardiac death without previously diagnosed heart disease in each four-year period among women and men, with 95% confidence intervals.

Among women across the 3 time periods, rates of SCD declined by 30% between the first and second 4-year time periods from 34 (95% CI, 31–38) to 24 (95% CI, 21–27) per 100 000 (P<0.001), and increased by 27% to 30 (95% CI, 27–33) per 100 000 (p=0.002) between the second and third periods. The subset with SCD as first manifestation of heart disease (Figure) accounted for 58% of the total rebound in SCD incidence from periods 2 to 3. There was no change in incidence over time for SCD occurring with identifiable preexisting heart disease.

Among men, SCD rates declined by 21% between the first and second 4-year time periods from 63 (95% CI, 59–68) to 50 (95% CI, 46–54) per 100 000 (P<0.001) and increased by 17% between the second and third periods to 58 (95% CI, 54–62) per 100 000 (p=0.006). In men, the subset of SCD occurring in the setting of identifiable prearrest heart disease was responsible for 55% of the rebound in overall SCD incidence.

Reasons for the rebound in SCD incidence and its sex-specific characteristics are likely complex. SCD in women is less likely to be explained by left ventricular dysfunction and obstructive coronary artery disease than in men, whereas other causes may be more common, including left ventricular hypertrophy by ECG criteria and nonischemic causes.^{3, 5} Given these findings, although improvements in treatment of traditional cardiovascular risk factors remain important in both sexes, focus on these factors alone may be less effective in preventing SCD in women than in men. Targets for reducing SCD risk among women need to be identified and should be an important focus of future investigation.