Losing a patient—and the family, too?

People were more likely to have an MI shortly after a significant death than in the six months prior.

Being broken-hearted is more typically imagined as an emotional state than a medical diagnosis. But according to a recent study, a very sad experience—specifically, the loss of a loved one—may actually trigger cardiac problems.

The study, published in Circulation on Jan. 24, included almost 2,000 patients who had been hospitalized for an acute myocardial infarction (MI). The patients were asked whether they had been notified about the death of a relative, friend or other significant person in their lives during the past year.

In a case-crossover analysis, researchers found that people were more likely to have an MI shortly after a significant death than in the six months prior. The risk was highest within 24 hours of the loved one's death and declined steadily from there.

ACP Hospitalist recently talked to lead author Elizabeth Mostofsky, ScD. Dr. Mostofsky, who is a research fellow in the department of epidemiology at the Harvard School of Public Health in Boston, Mass., discussed the study's results, future research and how hospitalists can incorporate this new information into their interactions with patients and families.

Q: What motivated this study?

A: There's been a lot of anecdotal evidence about an association between losing somebody and risk of heart attack or death. There have been a lot of studies looking at the long-term risk, mostly from spousal bereavement and all-cause mortality. We wanted to look at whether there was an acute increased risk of having a heart attack in a shorter time frame.

Q: How did the results compare to your expectations?

A: Based on the prior literature on spousal bereavement and the risk of mortality in the following months and years, we expected to find an association, but we were surprised that the risk spiked so quickly and we were surprised by the magnitude of the effect. We found that there was a 21-fold increased risk of having a heart attack in the 24 hours after losing a loved one compared to other times. The risk actually remained elevated for up to a month, but it declines over time to about fourfold by the end of the month.

Q: What are the clinical implications of the results?

A: The takeaway message is that bereaved individuals, and their family and friends, should be aware of the heightened risk to make sure that somebody who is dealing with loss of a loved one takes care of himself or herself. That includes taking regular medication and having the social support to deal with the anxiety and the grief and the depression of losing a loved one.

The message for clinicians is that if a person that recently experienced a loss experiences symptoms of a heart attack, the symptoms should not be ignored and attributed to the fact that they're dealing with a stressful period in their life. These symptoms should be evaluated, since they may in fact be signaling a heart attack, and the person should receive immediate medical attention.

Q: Any advice for hospitalists about how to apply these results in their interactions with patients' family members?

A: Hospitalists should be aware of this issue and remind close friends and family of a patient to take their regular medications. Furthermore, if a hospitalist notices that the friends and relatives are experiencing symptoms of stress, such as nausea, dizziness or chest pain, this may signal that they are having a heart attack.

[The study] has led to a lot of conversation among some of the physicians here about [whether] you really want to scare bereaved individuals. We haven't come out with any conclusive comment on whether relatives of patients should be warned that they are at risk in the next few days and weeks. We don't want to seem alarmist.

Q: Could some kind of intervention, such as grief counseling, potentially reduce the risk?

A: Psychosocial interventions seem plausible, but there hasn't been any conclusive evidence. There haven't been any randomized clinical trials to address whether psychosocial interventions would be helpful. Our collaborators are actually doing a study to look at whether interventions with medication may be helpful—giving people statins or aspirin in those days after bereavement.

Q: What's your next step in research on this topic?

A: Our study has two limitations. One is that we're asking heart attack patients to recall whether they've lost somebody in their life and if so, when. It would be great if we could use administrative data that has actual confirmation of these deaths. The other problem is that we only had 19 people who were exposed [to a death] on the day before their heart attack. It would be great if we had a larger sample so that we could look at whether the risk varies for losing a child or a parent or a sibling or a spouse, and also to look at whether the cause of death has an effect on the magnitude of risk. The next step for us is going to be a large prospective study using data from a large national registry.