One of the myths about single people is that they "don't have anyone," whereas married people do. The same simple-minded thinking assumes that married people have social support and single people do not. I was pleasantly surprised when Amy pointed me to a study in which the researchers actually recognized that both married and unmarried people could have good social support or poor social support.
The study was based on a survey of more than 100,000 adults from 16 states, with most analyses focusing on 967 of the participants who had active epilepsy. The authors (full reference is at the end) set up their study by claiming that "several theories suggest that social support, especially from marriage, has a positive impact on health" and that previous surveys have found that PWE (people with epilepsy) "are more likely to report being never married compared to those without epilepsy."
You can see right away where this is going: Pity the poor people with epilepsy who are especially likely to end up without that special social support from marriage that is especially likely to have a positive impact on health. If you read just the abstract of the article or just the one-paragraph summary, you might think that their study provided empirical support for tut-tutting unmarried PWE. Living Single readers know full well, though, that I always read closely and thoroughly, and study the numbers for myself.
In the survey, participants were simply asked whether they were married (or part of an unmarried couple) or not. The "not" group included divorced, widowed, separated, and never married. So, as is too often the case, this study tells us nothing about the implications of getting married for health or well-being. It compares only the currently coupled with everyone else, including those who got married and hated it and got divorced.
The authors did, though, ask all participants this question about social support: "How often do you get the social and emotional support you need?" Then they separated both the married and the unmarried group into those who had good social support (they were the ones who said that they "always" or "usually" got the support they needed) and those who had poor social support (they said they "sometimes," "rarely," or "never" got the support they needed).
Remember that the authors expected to find that "social support, especially from marriage" would have "a positive impact on health." So they asked participants to rate their health, and they also asked them about their general satisfaction with their lives.
The authors could have statistically compared each of four groups (married with good support, married with poor support, unmarried with good support, unmarried with poor support) to each of the other groups. Instead, they only reported statistical tests comparing each of the 3 other groups to the group of married people with good support.
The statistic they reported is an "odds ratio." If a group has an odds ratio of 1, that means that the members of that group have the same odds of good health or life satisfaction as do the married people with good support. A number less than 1 means that the group is doing less well than the married with good support, and a number greater than 1 means that the group is doing better than the married with good support. In the numbers below, an asterisk means that the difference is statistically significant.
Comparisons to Married People with Good Social Support
0.48 married with poor support
0.98 unmarried with good support
1.51 unmarried with poor support
0.33 married with poor support
0.57 unmarried with good support
0.22* unmarried with poor support
(All of these analyses control for factors such as age, gender, race, income, and mental health, so differences in those factors cannot explain the results In analyses in which mental health was not controlled, the results looked statistically worse for the married people with poor support.)
Let's consider the health results first. The fact that there are no asterisks means that none of the groups differed in health from the married with good support. If the groups were identical, the odds ratio would be 1. Look at the unmarried with good support - their health is essentially identical to the health of the married group with good support. With regard to health, there is nothing special about the support that married people get. Unmarried people with good support have health that is just as good.
Now look at the health of the unmarried people with poor support. Statistically, it does not differ from that of the married with good support. In fact, it trends toward looking better. The unmarried with poor support also look healthier than the married with poor support, though the authors do not report statistics for that comparison.
In life satisfaction, the unmarried with good support do not differ significantly from the married with good support (though they trend toward being a bit less satisfied). Only the unmarried with poor support report statistically lower life satisfaction than the married with good support. The unmarried with poor support looks similar in life satisfaction to the married with poor support, but again, the authors did not report statistical tests for that comparison.
When the authors get to the discussion section, they concede that their results suggest that "the quality of social support received is more influential than being married by itself." Friends, family, and other personal relationships, they acknowledge, may fulfill needs for social and emotional support.
The study has its limitations but it is a big step forward to ask people in all marital and relationship statuses about the social and emotional support they are getting, rather than just assume that married people rule because they are married. In this study of people with active epilepsy, the unmarried with good social support were not significantly different in their health or in their life satisfaction from married people with good social support.
Elliott, J. O., Charyton, C., Sprangers, P., Lu, B., & Moore, J. L. (2011). The impact of marriage and social support on persons with active epilepsy. Epilepsy & Behavior, 20, 533-538.