Relationships are the best investment of your life

In the autumn of 1938, as Europe held its breath, a team of Harvard researchers began measuring everything in 268 second-year students: lung capacity, blood pressure, physical build, even skull shape. They were searching for the formula for vital success, suspecting it would be written in biology. Almost simultaneously, across town, another team was tracking 456 boys from Boston's poorest neighborhoods, many raised in tenements without hot water. No one imagined that those two projects would eventually merge into the longest-running study ever conducted on adult life: 724 men accompanied year after year for more than eight decades, through their wars, weddings, divorces, promotions, relapses, and grandchildren.

The study is still active. It is called the Harvard Study of Adult Development, is on its fourth director—psychiatrist Robert Waldinger—and has outlived its own founders, who died without ever knowing the answer to the question they had raised. When Waldinger is asked to summarize 87 years of files, analyses, and interviews, his response disappoints anyone expecting a molecule: the best predictor of a healthy and satisfied old age is not cholesterol, nor income, nor fame achieved. It is the quality of relationships.

What 87 Years Teach Us

It is worth stating this with the precision of data, because the phrase sounds like a waiting room poster. Of everything the study had measured in those men at age 50, the variable that best anticipated their health at age 80 was their satisfaction with their relationships. Better than blood tests. Those who were happiest with their people at age 50 reached age 80 in better physical and mental condition. Waldinger has summarized this with unacademic bluntness: loneliness kills, and its power is comparable to that of smoking or alcoholism.

A legitimate objection arises naturally: 724 men—only men, almost all white, half of them from Harvard—are not the world. True, and the team itself acknowledges this. That is why the data matters above all because of the company it keeps.

308,849 People Later

In 2010, psychologist Julianne Holt-Lunstad, of Brigham Young University, asked the question in reverse: instead of tracking a group for decades, she gathered everything science had already measured regarding relationships and mortality. One hundred and forty-eight studies, 308,849 people, and an average follow-up of seven and a half years. The result, published in PLoS Medicine, has become one of the most cited statistics in public health: individuals with strong social relationships showed a 50% higher likelihood of survival during the studied period than those who were isolated. Placed alongside classic risk factors, the effect proved comparable to quitting smoking, and greater than that of obesity or physical inactivity.

There is a detail of that work that is almost never shared, and which matters most here: the average age of the participants was 63.9 years. This evidence was not gathered on university campuses. It was gathered, for the most part, from people who were right where you are now.

The WHO Takes Note—and Debunks a Cliché

In June 2025, the World Health Organization published the report of its Commission on Social Connection, co-chaired by Vivek Murthy, former Surgeon General of the United States. The figures in the report are sobering: one in six people worldwide experiences loneliness, and it is associated with approximately 871,000 deaths annually—about one hundred every hour. Sustained disconnection is linked to an increased risk of stroke, heart disease, and diabetes.

And then there is the finding that dismantles the cliché. If you were asked to point to the loneliest group on the planet, your mental image would likely have gray hair. The WHO data says otherwise: the highest rates of loneliness are among adolescents and young adults—between 17% and 21% of people aged 13 to 29. Loneliness is not an attribute of age; it is a human experience that comes and goes with circumstances. What does grow with age is the risk of isolation—having objectively fewer contacts—which affects up to one in three older adults. These are distinct concepts, and distinguishing between them already changes the conversation: we talk about this in detail when discussing the three lonelinesses.

What This Data Does Not Say

This Notebook has the habit of detailing the full nuances, so let us get to it. First: all of this is association, not proven causality. No one can—nor should—take ten thousand people and randomly assign them to isolation to see what happens, so science works with observational data, and in it, the arrow also points the other way: prior health shapes social life just as social life shapes health. Second: studies mostly measure the quantity and structure of relationships, and rarely their quality; Holt-Lunstad herself warns that a toxic relationship provides no protection—it is associated with the opposite. Third: averages describe vast populations, not individual biographies. That a solid bond is associated with better health in 308,849 people does not sign any contract with you.

And even with all that on the table, the convergence is striking: an 87-year study, a meta-analysis of 148 studies, and a global commission from the WHO, looking at the same phenomenon through three different methods, point in the exact same direction. In science, that is called consistency. And it deserves to be taken seriously.

Social Fitness

Waldinger and his co-director, psychologist Marc Schulz, have coined a useful expression to anchor all this: social fitness. The idea is simple and slightly uncomfortable: relationships are not an asset one owns, but a practice one trains. They are less like owning a home and more like muscles: neglected, they atrophy in silence, painlessly, until one day they are needed and are not there.

Retirement tests this idea like no other stage of life. For decades, a large part of your social life came built-in: colleagues, clients, mid-morning coffee, hallway conversations. No one warns you that with your last paycheck, that invisible infrastructure disappears as well. The bonds that remain—and the new ones—begin to depend on something that wasn’t as necessary before: intention. Calling for no reason. Suggesting a walk. Sustaining the friendship that is no longer sustained solely by shared routine.

The good news is that intention pays off more now than ever, because it is accompanied by something that was once scarce: time and discernment to choose whom to spend it with. You do not need fifty friends. The data—and experience—point to depth being far more important than quantity: someone with whom to think out loud, someone to tell the truth to, someone with whom to share a page you enjoyed.

Two of the Five Verbs

In the second stage of the Senior journey, that of Anthropology, we address a question that seems to belong to a different topic: who am I beyond the roles I have played? The stage covers the five verbs that define human existence—to be, to think, to love, to create, to coexist—and there we discover something that the science in this article independently confirms: two of the five verbs do not exist in the singular. To love and to coexist can only be lived with others. Those who ask themselves who they are without their titles end up finding that a large part of the answer has names and surnames. That is why in the NOMBA community, learning takes place in company: not as a backdrop for the program, but as part of the core subject. An idea read alone is an idea; once discussed, it is a bond.

Someone to Call

Of those 268 students whose skulls were measured in 1938 in search of the formula for success, those who reached their nineties in good health did not carry it in their biology. They carried it at the kitchen table, in their Tuesday friends, in the phone that was answered. Eighty-seven years of science to confirm what your experience perhaps already suspected: that the best portfolio you own is not traded on any market.

Compound interest also exists in affection. Every relationship nurtured today is patient capital. And there is, almost certainly, a phone call you have been putting off for weeks. The data cannot make it for you.

Cover image generated with artificial intelligence.