Methods

The present study was a prospective, controlled cohort study. From February 2005 through June 2008, we recruited bereaved spouses or partners or parents (74 spouses and 6 parents) of deceased patients from critical care units in 5 urban hospitals. The bereaved subjects were recruited and studied within 2 weeks of bereavement and at 6 months. Concurrently, we enrolled 80 nonbereaved relatives of patients using the same hospitals as the bereaved. Of the participants, 3 (2 bereaved and 1 nonbereaved) were excluded from the present analysis because of underlying atrial fibrillation. Therefore, 78 bereaved and 79 nonbereaved participants were included in the final analysis. The institutional ethics committee approved the study protocol.

The bereaved spouses or parents who met the eligibility criteria at the participating institutions were contacted by a social worker or study investigator at the hospital or by telephone within 72 hours after the bereavement. Interested subjects were assessed at home within 2 weeks of their bereavement (average 11 days, range 3 to 15). Of those who met the inclusion criteria, 60% agreed to participate; 3 bereaved (4%) withdrew from the study before the 6-month assessment and 1 died before the final assessment from multiorgan failure. The control group consisted of family members of patients, who were recruited from clinical areas (spouses or parents) such as day surgery, orthopedic, or endoscopy clinics of the same hospitals as the bereaved. The initial evaluation of nonbereaved participants was approximately 1 month after hospital discharge of their hospitalized family member.

The bereaved and nonbereaved participants were excluded if they reported severe illness (ie, respiratory, heart, liver, renal failure), malignancy, history of coagulopathy, cognitive impairment, psychosis, or immunosuppressive illness, were a nursing home resident, were unable to speak or read English, or had baseline rhythm of atrial fibrillation. We excluded nonbereaved subjects who had experienced the death of a close family member or friend in the previous 2 years. Only 1 bereaved participant had experienced a second close bereavement in the previous 2 years.

The symptoms of depression were measured using the Center for Epidemiologic Studies Depression (CES-D) questionnaire, and the symptoms of anxiety and anger were measured using the Spielberger State Anxiety and Anger scales. Social support was also measured using a validated questionnaire (Social Support Questionnaire-6), and the morning blood cortisol level was collected and analyzed, as described previously. The participants were instructed to undertake their usual daily activity and completed an activity diary every 30 minutes throughout the waking hours of the monitoring period. The activity level was documented in the diary on a 1 to 10 Likert scale with greater levels, indicating more activity.

Five-lead Medilog AR4 Holter monitors (Schiller, Australia) were used for 24-hour monitoring and analysis conducted using Medilog Prima 3 software (Schiller, Australia). All RR interval time series were first edited automatically, after which careful manual editing was performed by visual inspection. All abnormal beats were manually checked and arrhythmia strips classified by a cardiologist who was unaware of the study group. After editing of the RR intervals, the RR interval spectrum was computed for the entire recording interval. A Fast Fourier transform algorithm was used to estimate the power spectrum densities of HR variability. The frequency domain measures of RR interval variability were computed by integrating the point power spectrum over the frequency intervals. The power spectra were quantified by measuring the areas in the following frequency bands: (1) overall band (0.00 to 0.40 Hz range), (2) ultra-low-frequency power (<0.0033 Hz), (3) very-low-frequency power (0.0033 to <0.04 Hz), (4) low-frequency (LF) power (0.04 to 0.15 Hz), and (5) high-frequency (HF) power (0.15 to <0.40 Hz).

The time domain measures included the standard deviation of the NN intervals index (SDNNi), an estimate of the changes in HR due to cycles <5 minutes and sensitive to short-term changes in HRV; the square root of the mean squared differences of successive intervals (rMMSD), and number of interval differences of successive NN intervals >50 ms divided by the number of NN intervals (pNN50), which similar to the rMMSD, is associated with parasympathetic modulation of the HR.

Differences in the sociodemographic data were analyzed using Student’s t test for interval data and the chi-square test for categorical data. The primary analysis comparing the bereaved to nonbereaved at the acute assessment used the t test or Mann–Whitney U test, as appropriate. The longitudinal changes from acute to 6 months were evaluated using paired analysis (paired t test or McNemar test, as appropriate). The skewed HRV power parameters were logarithmically transformed before statistical testing. To explore the relations between HR and HRV variables with the measure of psychological stress and cortisol, Spearman’s rank correlation and Pearson’s correlation tests were used. The data were analyzed using SPSS for Windows, version 17.0 (SPSS, Chicago, Illinois).

Results

The bereaved and nonbereaved had similar sociodemographic characteristics, including the level of social support ( Table 1 ). Most (n = 72) were spouses or partners of the deceased, and 6 were parents. For bereaved spouses, the number of years as a couple ranged from 3 to 62 years (mean 38). The cause of death for the deceased family members of bereaved participants was CVD in 21 (cardiac arrest in 17, heart failure in 2, aortic aneurysm in 2), respiratory failure in 17, cerebral hemorrhage in 13, multiorgan failure in 9, cancer in 9, liver disease in 4, trauma in 2, suicide in 2, and meningitis in 1.

The level of physical activity, as recorded during 24-hour monitoring, was similar between the bereaved and nonbereaved at the initial assessment (mean ± SE 5.1 ± 0.2 vs 4.9 ± 0.2, p = 0.23; possible range 1 to 10). Among the bereaved, no difference was found in the activity level from acute assessment to 6 months (5.1 ± 0.2 vs 5.3 ± 0.2; p = 0.33). HR-lowering medication was used by 8 of the bereaved (β blockers in 3 and calcium channel blockers in 5) at baseline and by 3 of the nonbereaved controls (β blocker in 1 and calcium channel blockers in 2; p = 0.16).

The symptoms of depression, anxiety, and anger were all elevated acutely in the bereaved compared to the nonbereaved and reduced in the bereaved at 6 months compared to at the acute assessment ( Table 1 ). Additionally, the morning cortisol levels were greater in the bereaved participants than in the nonbereaved acutely (median 312 mmol/L, interquartile range 241 to 414; vs 271 mmol/L, interquartile range 222 to 353; p = 0.01). The cortisol levels did not change from the acute to the 6-month assessment in the bereaved (median 326 mmol/L, interquartile range 237 to 400; p = 0.47) or the nonbereaved (median 279 mmol/L, interquartile range 229 to 361; p = 0.39).

The results of 24-hour HR analysis are listed in Table 2 . At the initial acute assessment, the bereaved had a greater 24-hour average and minimum HR and more atrial premature beats and episodes of nonsustained (>3 consecutive beats) supraventricular tachycardia (SVT) compared to the nonbereaved ( Table 2 ). Similar between-group differences were observed after excluding the 8 bereaved and 3 nonbereaved participants taking HR-lowering medications (data not shown). At 6 months, the average and minimal HR and episodes of SVT were all significantly lower in the bereaved participants compared to their initial acute assessment ( Table 2 ). However, there were no differences in HR or arrhythmia in the nonbereaved from the acute assessment to the 6-month assessment ( Table 2 ). There were too few ventricular premature complexes or episodes of ventricular arrhythmias in either bereaved or nonbereaved to enable statistical comparison.

At the initial acute assessment, the bereaved had lower SDNNi, total power, very-low-frequency, and LF power than the nonbereaved ( Table 3 ). In the bereaved, the SDNNi, rMMSD, pNN50, and LF and HF power parameters were all greater at 6 months compared to the initial acute assessment. No significant differences were seen in any HRV measures from the initial assessment to the 6-month assessment in the nonbereaved participants ( Table 3 ).

Table 3

Time and frequency heart rate variability measures in bereaved and nonbereaved participants from initial assessment (acute) to 6 months (mo) follow-up

Variable	Bereaved		Nonbereaved		p Value
	Acute	6 mo	Acute	6 mo	B vs NB, Acute	B, Acute to 6 mo	NB, Acute to 6 mo
SDNNi (ms)	45.4 (35.8–61.7)	46.0 (40.8–69.4)	49.9 (42.8–59.3)	49.8 (42.3–59.4)	0.02	0.02	0.38
RMSSD (ms)	30.1 (18.8–44.0)	36.5 (23.9–53.5)	26.8 (22.2–39.8)	30.4 (19.9–38.8)	0.15	0.04	0.74
PNN50 (ms)	5.6 (1.6–12.0)	7.2 (3.17–14.2)	5.2 (2.4–9.9)	5.5 (2.4–10.2)	0.49	0.04	0.54
Total power (ms )	2,225 (1378–3578)	2,700 (1892–6171)	2,935 (1980–4063)	2,891 (1950–3843)	0.03	0.08	0.23
Ultra-low-frequency power (ms )	294 (169–524)	320 (189–947)	404 (212–608)	262 (238–485)	0.07	0.62	0.13
Very-low-frequency power (ms )	1347 (853–2,104)	1489 (1,110–3,373)	1677 (1,185–2,421)	1705 (1,176–2,353)	0.046	0.06	0.24
Low-frequency power (ms )	386 (128–582)	521 (298–1,041)	521 (295–753)	544 (303–804)	0.01	0.02	0.42
High-frequency power (ms )	152 (109–258)	257 (138–573)	177 (137–296)	177 (130–326)	0.06	0.002	0.89
LF/HF power ratio	2.1 (1.6–2.9)	2.1 (1.5–2.9)	2.4 (1.8–3.9)	2.3 (1.7–3.2)	0.076	0.487	0.924

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Detection of Previously Undiagnosed Atrial Fibrillation in Patients With Stroke Risk Factors and Usefulness of Continuous Monitoring in Primary Stroke Prevention
2. Intravascular Ultrasound Analysis of Plaque Characteristics and Postpercutaneous Coronary Intervention Catheterization Outcomes According to the Remodeling Pattern in Narrowed Saphenous Vein Grafts
3. Incidence, Predictors, and Implications of Access Site Complications With Transfemoral Transcatheter Aortic Valve Implantation
4. Asymptomatic Microvascular Angina?

Stay updated, free articles. Join our Telegram channel

Join