One in Four American Adults Ate Every Meal Alone, and the Cost Is About the Same as Losing a Job

This is correlational and the authors do not claim otherwise. People who are depressed, ill, bereaved or newly alone eat by themselves more, and every one of those conditions independently lowers a life rating. A finding that solitary eaters score lower is entirely compatible with eating alone being a symptom rather than a cause.

There is a second limit worth naming. Eating alone is not the same as being lonely. Somebody who lives by themselves, eats lunch at their desk by preference and spends four evenings a week with other people is counted here as a solitary eater. The measure is a proxy, and it is a good one precisely because it is easy to answer accurately, which is not a small virtue in survey work.

Why the meal, specifically

Because it is the last piece of social structure most people still have by default.

A shared meal has a beginning and an end that neither party controls, it lasts long enough for a conversation to develop past its opening, and it requires no organiser, no invitation and no stated purpose. Almost nothing else in adult life has that combination. Which is why it keeps turning up as the variable when researchers look for what has thinned out.

It also connects two findings we have covered separately in the last month. A nationally representative study found that social media contacts a person had never met in real life doubled the odds of loneliness, while contacts who were close friends changed nothing either way, which suggests mediated contact is not doing the job unmediated contact does. And the alcohol study we looked at yesterday found wine drinkers dying less often than spirits drinkers, with the researchers' own leading explanation being that wine is drunk with a meal and spirits often are not.

Three different methods, three different questions, and the same thing keeps appearing in the results. It is not the drink and it is not the feed. It is whether anyone was sitting there.

What to do with it

Very little, and that is the point.

The interventions that follow from this are not products. Eat with somebody. Not a scheduled monthly dinner that takes three weeks of messages to arrange, which is the version most people default to and which is doing something else. An ordinary meal, often, with whoever is around.

We wrote about the physical places where that used to happen without arrangement in the return of the third place. The reason those places matter is the same reason this study finds what it finds: they lowered the cost of ordinary company to almost nothing, and the cost has gone up.