Just how stressful is "getting caught"?

The Ashley Madison data breach could have caused hearts to break in more ways than one

Photo by Kelly Sikkema on Unsplash

What happened with AshleyMadison.com?

Ashley Madison, a dating website that was famous in the 2010s for encouraging and facilitating extra-marital affairs, is the subject of a new docuseries that serves to remind us of this controversial chapter in internet history.

In 2015, a data breach at Ashley Madison revealed the identities of millions of the site’s users who were on the site. After the breach, anyone could search the data to see if someone they knew had an account on the site. Though an examination of the data suggested that the site only had a handful of real women, there seemed to be about 20 million men who were actively looking for affairs—and who could now get caught.

We imagine that once the news of the breach broke, many of those 20 million men were suddenly sweating bullets. Was the stress of getting caught enough to cause a heart attack?

Stress and health (and soccer)

We’ll return to the Ashley Madison affair shortly, but first, let’s talk about stress and our health.

Because our brain is heavily connected to our entire body—including our internal organs (just think of that feeling in your stomach when you get bad news)—psychological stress, anxiety, and negative health outcomes often go hand in hand. Heart attacks, which have been known to manifest during times of emotional stress, are one of those bad outcomes.

The challenge, as for many questions of lifestyle and health, is figuring out to what extent stress can cause heart attacks as opposed to simply being correlated with heart attacks—since people with more stress in their lives may also be at increased risk of heart attacks for reasons besides stress alone (e.g., people with high stress may be more likely to smoke cigarettes, which also increases risk of heart attacks).

How can we scientifically evaluate if stress causes heart attacks? Without purposefully stressing people out in an (unethical) randomized controlled trial, one way we can get at causation is through natural experiments, where people may be exposed to stress in a way that’s essentially random. When the timing of increased stress is completely unrelated to the health outcome in question, it’s as good as random. This means that if we look at a group of people before the stressor and after the stressor (in an analysis called an interrupted time series), we can attribute any changes to that group to the stressor, assuming no other major stressors occurred at the same time.

Let’s take an example from across the pond. During the 1998 World Cup, England played in four soccer matches, winning one and losing two before losing again in a nail-biter shootout with Argentina, knocking them out of the tournament. In a study published in the British Medical Journal, researchers wanted to know whether the stress of these soccer games was enough to cause heart problems. They found that within two days of the stressful shoot-out loss to Argentina, the rate of heart attacks was 25% higher compared to baseline. Meanwhile, there was no change in the rate of stroke, traffic injuries, or deliberate self harm. There was also no increase in the rate of heart attacks during the less stressful World Cup matches where England played. It seemed that the stress of that match could be leading to heart attacks—researchers estimated about 55 heart attacks attributable to the World Cup at the hospitals they studied. Other studies have found similar findings.

Did the Ashley Madison breach lead to heart attacks?

With millions of men on the site, it seemed plausible to us that the stress of getting caught with an account could lead to heart attacks.

With the caveat that this was a quick look at the data in a study that did not go through the peer review process of studies we often discuss, we (with Brown University economist Jaemin Woo and Harvard cardiologist André Zimmerman) took a look at insurance claims for hospitalizations across the U.S. We focused on men aged 40-60 years—the demographic that was both on the site and at higher baseline risk for heart attacks. The graph below shows the percentage of hospital admissions for heart attack in the days surrounding the release of the identities of Ashley Madison users:

“reldate” is the number of days since the Ashley Madison data was published online, with negative numbers representing the days before. “Linear prediction” is the percentage of hospitalizations for heart attack based on a simple statistical model.

While we mostly saw statistical noise, there was a particularly high amount of heart attacks the day before the identities were published. While the argument could be made that the anticipation of the release was stressful (hackers warned that the data would be released), if it was truly due to the Ashley Madison data breach, we would expect to see the increase persist the day of and probably the days after the release. This suggested to us that the positive blip was likely due to something else.

In short, it didn’t seem like the stress of getting caught was enough to lead to population-level increases in heart attacks among men aged 40-60. Now, this quick analysis is not without some limitations, so we should take it with a grain of salt. For example, perhaps the data breach wasn’t large enough to lead to detectable changes in population-level heart attacks, even if the breach did have an effect for some. Or the types of men including in the data we studied weren’t included in the data breach, in which case we wouldn’t expect to see a change in heart attacks.

Either way, it certainly didn’t seem to us like there was a population-level increase in heart attacks on par with the World Cup study in England. It’s hard to decipher what this might tell us about the stress-related health effects of the men with accounts on the site at the time, but perhaps that’s a question best left unanswered.

NB: We hope you enjoyed this week’s post and if you did, please share with one other person! We’ll plan to do more posts like this in the future, where we share the answers to quirky questions that we’ve asked — questions that we thought were interesting but not reveling enough to pursue beyond a “quick and dirty” first pass. Our goal is to give you a window into an otherwise black box of how ideas come about and where they go. Till next week.