What connection does to you

In the autumn of 1938, researchers at Harvard began following 268 college sophomores, hoping to watch something no study had watched before: whole lives, from the inside, as they happened.¹

Around the same time, a second project began tracking 456 teenage boys from Boston’s poorest neighborhoods — lives chosen precisely for being nothing like Harvard’s. In time the two groups were folded into a single project, 724 men in all, and the Harvard Study of Adult Development became the longest study of adult life ever conducted.¹

The study outlived its founders. It followed the men to war and back, through marriages and careers, children and divorces, fortunes and relapses. Researchers read their medical records, drew their blood, scanned their brains, interviewed their wives, and sat in their living rooms asking the same patient questions, decade after decade. The men became bricklayers and doctors, factory workers and lawyers. One became President of the United States.¹

It has never stopped. The study is still running today, now following more than 1,300 of the original men’s children.¹ And after nearly nine decades of data, its central finding fits in a single sentence.

One study, even one this long, is still one study. The harder question is whether connection shows up everywhere else researchers have looked.

It does. It shows up most starkly in the mortality tables.

In 2010, psychologist Julianne Holt-Lunstad and her colleagues pooled 148 studies that had followed 308,849 people for an average of seven and a half years, asking a blunt question: do people with stronger social relationships actually live longer?³

The size of the effect rivaled quitting smoking, and exceeded the effects of obesity and physical inactivity.⁴ Not metaphorically — statistically. In the aggregate, the strength of your relationships predicts whether you will still be alive in a decade about as well as cigarettes do.

Five years later, the same team measured the inverse: what the absence of connection costs.⁵

And one more detail, easy to miss and hard to forget: in that same analysis, the effects were stronger in people younger than 65.⁵ Loneliness is not an old-age problem. If anything, the young carry more of its weight.

By 2023, the evidence was loud enough that the U.S. Surgeon General declared loneliness an epidemic, warning that the mortality impact of being socially disconnected is “similar to that caused by smoking up to 15 cigarettes a day.”⁶

So connection is among the strongest levers a person can pull on their own wellbeing. Which leaves one practical question for a century that mostly types at itself: does it matter how we connect?

It turns out it matters enormously.

Everything that matters about “I’m fine” lives outside the letters. The pitch that falls away at the end. The half-second before fine. The breath. The smile — or its absence — that you can somehow hear. Researchers call these paralinguistic cues. The rest of us call it how someone sounds. And this channel, it turns out, is where empathy does most of its work.

In 2017, Yale psychologist Michael Kraus ran five experiments with 1,772 people, testing how accurately they could read what others were feeling: watching and listening, watching only, or listening only. Listening only won. People judged others’ emotions most accurately when the voice was all they had — because a voice forces attention onto the cues that actually carry feeling.⁷

The body agrees. When researchers gave 68 girls a stressful public-speaking task and then let them contact their mothers, the girls who heard her voice — in person or by phone — released oxytocin, the hormone of bonding, while their stress hormone cortisol fell. The girls who exchanged instant messages with her got neither. Hormonally, texting their own mother was indistinguishable from no contact at all.⁸

Then why do we keep typing? Because we mispredict. In a series of experiments published in 2021, Amit Kumar and Nicholas Epley asked people to reconnect with an old friend. Most expected a phone call to feel awkward, and chose to write instead. But the people randomly assigned to call reported feeling significantly more connected than those who typed — and no more awkward.⁹

We route around the very medium that bonds us, to avoid a discomfort that never arrives.