Divorced people have a different health profile than those who stay married. By examining the health of people who divorce, along with those who stay married, we can learn something that is often overlooked about the implications of marriage for health.
In the article below we’ll explore how poorly executed studies sometimes result in propagating widespread belief in claims that may not be true, particularly around marriage and health. We’ve all seen the “statistics” that tell us “marriage makes people healthier.” However, in the vast majority of studies which are used to make that claim, people who are currently married, or people in their first marriages, or people who got married and stayed married, are compared to single people.
Do you see what’s missing from those groups of married people? None of them include people who got married and then got divorced. (Widowed people are excluded, too, but that’s a topic for a different day.) The resulting group, without the people who got divorced, is a biased sample.
The lesson based on the biased studies is supposed to be, get married and you will get healthier. But the people who got divorced did get married; they just didn’t stay that way. If you get married, you could end up divorced. And if you marry and then divorces, what does that mean for your health? This is a critical omission, as a substantial percentage of people fall into this category, and therefore, it is a vital part of the picture.
What we have gleaned thus far is that people who get divorced are less healthy than they were before – at least at first. And if you compare people who are currently married to those who are divorced and those who have always been single, it is the divorced group that typically has the worst health.
Instead of just looking at the health of divorced people just after they divorce, and comparing it to before they divorced, we can also look at the health question a different way: Is the health of married people a factor in whether they get divorced? For example, are married people who are especially healthy more likely to get divorced? Or is the reverse true – that married people who are especially unhealthy are more likely to get divorced? Or does it make no difference? It doesn’t take long to see that the issues are not so easily teased apart, and thus conclusions are difficult to draw. Similarly, it is not easy to determine whether the issues are causally related or simply correlated.
A longitudinal study of about 10,000 Dutch adults looked at the health of people over the course of various marital transitions. At the start of the study, the participants were asked about their health in general. They were also asked if they had each of 13 health complaints (such as regularly having an upset stomach, or often feeling tired) and 23 chronic conditions (such as serious heart disease or heart attack, or chronic obstructive lung disease).
The participants were followed for 4.5 years. In the key comparison, the authors looked at the health of the people who started out married but were divorced by the end of the study – that is, 4.5 years later.
The results were interesting, but perhaps not in the way the authors expected. Married people who said that they had 4 or more of the 13 health complaints were 1.5 times as likely to be divorced by the end of the study than those who had fewer than 4 health complaints. The findings were even stronger for chronic conditions. Married people with 2 or more of the 23 chronic were more than twice as likely to end up divorced.
The authors ruled out, statistically, other factors that could account for their results, such as changes in health with age. They also did analyses in which they only looked at divorces that occurred in the last 2 years of the study instead of all 4.5 years. None of that mattered. The married people who were particularly unhealthy (as self-reported) were still most likely, by a lot, to end up divorced by the end of the study. (Of course, because health was self-reported and not measured directly by medical staff, it may well be that certain personality types are more likely to complain about health or imagine themselves as having an ailment when presented with a survey. We aren’t necessarily talking about hypochondriacs, but some personality types are predisposed to giving themselves a clean bill of health, and others the opposite. Perhaps those personality types that tend to self-report an unhealthy state also have personalities which contribute to a sub-optimal living environment with their spouse, increasing divorce rates. But self-reports of health are valid and cannot be attributed entirely to personality.) In future studies, it would be helpful to include objective measures of health instead of, or in addition to, people’s reports of their health.
Now let’s go back to where I started. It is likely that we will continue to read about studies that compare currently married people to single people, or people in their first marriages to single people, or people who got married and stayed married to single people. We will continue to be told: “Look, the married people are healthier! Get married and you will be healthier, too.” When that happens, keep in mind that all these different comparisons take all of the divorced people out of the marriage group. Also recall the study I’ve been describing. Statistics can be misused to fit a specific agenda. Those kinds of claims cannot be accepted uncritically, because they don’t really tell the whole story.
The married people who got divorced in the Dutch study were much more likely to have health complaints and chronic conditions than the married people who did not get divorced. What researchers are typically doing is setting aside all those people who were especially unhealthy and ended up divorced, and then comparing the remaining married people (the ones skimmed off the top) to single people. And then they want to tell you that getting married makes people healthier.
Sometimes it is even worse than that. Sometimes researchers lump the divorced people in with the lifelong single people, and compare the resulting group to the people who are left in the married group (the currently married people, or the people in first marriages, or the people who got married – whether for the first time or not – and stayed married). This is not a fair way to handle the data. Lifelong single people are typically healthier than divorced people. So the social scientists, in these instances, take the least healthy people – people who did get married and then got divorced – out of the marriage group and reassign them to the not-married group. Then they say, “see, marriage makes people healthier.” This is an inaccurate conclusion. They should instead be comparing everyone who ever got married to people who stayed single. It can also be useful to report the results separately for single, married, and divorced.
Here’s something remarkable. Even when social scientists use cheater techniques like these – and they use them constantly – they don’t always find that marriage makes people healthier. Recent studies, often more sophisticated methodologically than their predecessors, are not showing that people who get married get healthier than they were when they were single. I described one of those studies in “Health benefits unlikely even from the longest marriages.” Another study found that people described their overall health as somewhat worse after they married than when they were single.
I know it sounds intuitively reasonable to compare the health status of people who are currently married to people who are not married. If the married people are doing better health-wise than those who are not married, it seems to make sense to conclude that marriage improved their lives. But that’s cheating, because it does not include the people who got married and later divorced.
Next time you hear about a study that compares married people to people who are not married, remember that someone is trying to pull a fast one on you. They may not be deliberately trying to mislead you. They may actually think that such a comparison counts as good science, or they may have an agenda. I hope you understand that to be believable and accurate, studies need to be conducted and interpreted in a rigorous manner.