Have you been basking in the warmth of the good press about living alone? Have you been enjoying the acknowledgment that living alone can be a happy, healthy, joyful, and - surprisingly to some - an experience of enhanced connections with other people? Savor it while you can. The backlash is beginning.
"People living alone 'are more depressed,'" proclaimed the BBC. "Living alone linked with depression in young and middle aged adults," said the Huffington Post. The research they are describing is pocked with problems (I'll focus here on just a few of the big ones), but they are easy to miss because the study has some of the superficial characteristics of a good research design.
In the year 2000, the researchers began by surveying several thousand Finns between the ages of 30 and 65. The participants were asked about their living arrangements and demographic characteristics, their social support, and some of their health behaviors. They were never asked any questions about feeling depressed. I'll return to that point later.
Finland keeps tabs on prescription drug use with its National Prescription Register which "covers the entire outpatient population and all reimbursed doctor-prescribed medications in Finland." The researchers tapped into the registry when they first started recruiting participants in 2000 and 2001, and continued to check the purchases of prescription anti-depressants until 2008.
That's the part that sounds great, at first. I'm always preaching the importance of longitudinal research when experimental research is not possible. (We can't randomly assign people to live alone or with other people - they get to choose that for themselves.) The next best thing to an experiment with random assignment is a design in which the same people are studied over time as their life situation changes, whether that situation is marital status or living arrangements or anything else, to see whether the experience of interest also changes accordingly. So if the researchers had followed the same people over the 7 or 8 years, asking them each year about their living arrangements, then looking to see whether their depression changed with their living arrangements, that would have been impressive. Maybe, for instance, people who became widowed and were living alone for the first time in decades did indeed become more depressed, whereas those who, for the first time, could afford that much-coveted place of their own felt less depressed once they had it.
The study did nothing of the sort. The researchers asked about living arrangements only once, when the participants were first recruited. The statement in the Huffington Post that "One-quarter of people living alone filled an antidepressant prescription during the seven-year study, compared to just 16 percent of those who lived with spouses, family, or roommates" is not exactly true. The authors only know whether the participants were living alone during the first year of the study. Over the course of the next seven years, they could have changed living arrangements multiple times for multiple reasons.
During that one time when participants were asked about their living arrangements and the authors checked the drug registry, they found that 9.1% of people living alone filled antidepressant prescriptions, compared to 5.7% living with other people. If you ever took an introductory course in research methods, you know that a study like this tells you nothing about causality.
The Huffington Post story acknowledges that we can't know what causes what - does living alone cause depression or does depression cause living alone? The third option, though, slips past them. Something else could be the motivator behind both living alone and taking antidepressant medications. (Remember, depression was never assessed.)
Here's one possibility: self-sufficiency. Maybe people who are self-sufficient and have a sense of personal mastery (qualities especially important to people who are lifelong singles) prefer to live alone and also take care of themselves when they need to. If they think they might need a prescription drug, they consult with their doctor, and if the doctor agrees, they get the medication. Maybe people who live with others are just as likely to be depressed, but less likely to do anything about it.
Here's a question that is more philosophical than methodological. Suppose you administered a standard measure of depression to those who lived alone and those who lived with others and found no differences whatsoever in how depressed the two groups felt. But when you checked the registry, you found that more of the people living alone (about 9%, compared to about 6%) purchased anti-depressants. What would you conclude? Would you say that the two groups were equally unlikely to have symptoms of depression, so their rates of depression were the same? Or would you count it against the group living alone that they were taking meds, and call them more depressed, even though they scored as no more depressed by the usual ways of measuring depression?
There are analogies in the literature on physical health. Sometimes people are asked questions about how they feel. What if two groups reported equally low levels of pain, but the people in one of the groups took more aspirin? Would you tell them they need to talk to someone about their pain?
The BBC doled out just such advice in their story about this study. Erroneously equating living alone with feeling alone and having no one to talk to, the article said: "Loneliness and isolation results in people having fewer outlets to talk about how they are feeling, which is something that we know can really help to manage and recover from a mental health problem...It is therefore essential that people who live alone are given the most appropriate treatment such as talking therapies..."
Ah, yes, let's have a mass therapeutic intervention into the lives of all of those people living alone. Never mind that only 9% of them are taking antidepressants. (And maybe, as a result, they are not depressed.) About 6% of people living with others are also filling the same prescriptions, but hands off all of them - none of them need help.
Again, I'm not saying that no one finds living alone a depressing experience. I just think it is wrong to make blanket statements unsupported by the evidence. Feelings about living alone - or any other arrangement - are likely to depend on how you arrived at your arrangement and how you really do want to live. The authors acknowledge, in a sentence at the end of their article, that they were "unable to compare individuals who unwillingly lived alone to those living alone through choice." That point, though, never made it into the BBC's article.
[Thanks to Susan Hurt for the heads-up about these media reports.]
UPDATE: Susan Hurt just read this and suggested another important point: Maybe doctors have biased assumptions about people who live alone, and so are quicker to prescribe antidepressants for them than for people who live with others. Happily, Psyngle had a very enlightened physician -- see the Comments section.