Picture a kitchen table at 7 p.m. on a Tuesday: one plate, one fork, a phone propped against a glass.
That quiet scene is now the default for roughly one in four American adults on any given day. It has been creeping in for years. And a sweeping new global study says the habit carries a measurable cost that surprises most people who hear it.
So what is actually happening at the level of brain and behavior when people stop eating together?
Why eating together works on wellbeing the way a paycheck does
Shared meals do something that most social rituals cannot: they are both physically synchronizing and emotionally anchoring at the same time. Sitting down together sets a rhythm of looking, pausing, passing and responding that mirrors the call-and-response patterns linked to social bonding. The body reads the shared table as a signal of inclusion, and researchers find that the wellbeing boost it delivers is statistically comparable to the boost from economic security itself.
That is why researchers treat meal frequency as a behavioral measure of social connection rather than a lifestyle detail. A shared meal is countable in a way that “feeling close to others” is not. Either lunch happened with another person or it did not, making it a cleaner instrument than most survey questions about belonging.
So when analysts want a quick read on whether a population is socially connected, the dinner question cuts through the noise better than almost anything else they have tried.
What the data looks like across a generation of American dinners
Using two decades of time-use records, researchers found that roughly 1 in 4 Americans reported eating all of their meals alone the previous day in 2023, an increase of 53 percent since 2003. The rise started well before the pandemic, with a notable spike in 2011, and has continued climbing especially sharply among adults under 35, whose rates of eating all meals alone have more than doubled over that period.
Meanwhile, the restaurant industry is tracking the same shift from the other side of the table. Solo reservations at full-service restaurants rose 22 percent in the third quarter of 2025 compared with a year earlier, according to restaurant technology company Toast’s Q3 2025 Restaurant Trends Report.
More tables for one being booked does not mean the habit is harmless.
The scale and the number that stopped researchers
The findings draw on Gallup World Poll data across 142 countries and territories, covering more than 150,000 respondents, making them among the broadest meal-sharing data ever assembled. The dataset was large enough to test whether the pattern held across regions with very different incomes, cultures and employment rates. It did, almost everywhere. (The American time-use portion of the research drew on roughly 235,000 responses collected by the American Time Use Survey from 2003 to 2023.)
Researchers found that whether someone shared a meal last week predicted their overall life evaluation more reliably than knowing whether they were unemployed, placing meal sharing on par with income and employment status as a predictor of wellbeing. Sharing 13 meals a week appears optimal, but every additional shared meal contributes, and the benefit compounds as the number rises.
Where the pattern holds and where it gets complicated
One complication the researchers themselves flag is direction of causality. The study shows a strong correlation between shared meals and happiness but does not establish whether sharing meals causes happiness or whether happy people tend to share more meals. Co-author Micah Kaats said both are likely true, but which factor dominates is a question for future research.
Geography adds another wrinkle. The United States ranks 69th out of 142 countries for shared meals, while Senegal, Gambia, Malaysia and Paraguay top the global rankings, with residents sharing 11 or more meals with others each week. These differences are not fully explained by income, education or employment. Culture at the table is doing real independent work.
Living alone does not fully account for the American trend either. Dining alone has climbed sharply even among people who live with others: in 2023, nearly 70 percent of those living alone ate all meals solo, compared to 20 percent of people who lived with housemates nearby. That second figure is the one that most surprised the team.
The reading also matters for how Americans think about restaurants as social destinations. When a meal is delivered to a sofa instead, the setting that normally turns eating into a social event disappears entirely.
What a single shared meal a week actually changes in practice
The good news is that the threshold is low. People who ate even one meal with others in the past week scored measurably higher on life evaluations than those who ate every meal alone. Getting to zero shared meals is the costliest move; going from zero to one recovers most of the gap.
That framing shifts the conversation away from grand lifestyle overhauls. A work lunch that used to feel like a chore, or a Sunday dinner that keeps getting skipped, carries more psychological weight than most people assign to it when deciding whether to bother. The barrier is rarely desire; it is inertia and scheduling.
The research does not say eating alone is always harmful. What it measures is the population-level cost of a steady disappearance of shared meals from American life, week after week, year after year. Two decades of that pattern is something the data can now measure precisely. And the fix, the study suggests, is simpler than most people expect: one extra seat at the table, even with someone you barely know, is enough to move the dial.
