In 1938, a longitudinal study began at Harvard University that has since become the longest continuous study of human development ever conducted. The Grant Study, named after the philanthropist who funded it, enrolled 268 sophomores from the Harvard classes of 1939 through 1944. A separate project, the Glueck Study, was launched in 1940 at Harvard Law School by the criminologist Sheldon Glueck, Roscoe Pound Professor of Law, who was studying juvenile delinquency. The Gluecks enrolled 500 inner-city Boston boys who had been remanded to juvenile correctional facilities, along with 500 matched non-delinquent boys from the same neighbourhoods to serve as controls. The two studies ran independently for decades before being combined in the 1970s into what is now known as the Harvard Study of Adult Development, currently directed by the Harvard Medical School psychiatrist Robert Waldinger, who took over the directorship in 2004.
The study’s most-cited finding is striking. According to the Harvard Gazette’s 2017 feature on the study at its 80-year mark, “close relationships, more than money or fame, are what keep people happy throughout their lives.” Waldinger summarised the finding in his widely-circulated TED Talk: “The people who were the most satisfied in their relationships at age 50 were the healthiest at age 80.” Relationship satisfaction at age 50 turned out to be a better predictor of physical health at 80 than cholesterol levels measured at the same point. It was also a better predictor of late-life outcomes than social class, IQ, or genetic background. The Harvard men and the inner-city Boston men, despite very different starting circumstances, showed essentially the same pattern.
How relationships affect the ageing brain
The mechanism by which close relationships might protect against cognitive decline is not fully understood, but the broader literature has converged on a few plausible pathways. According to a 2022 BMC Public Health meta-analysis of social relationships and cognitive decline, strong social ties appear to operate through at least three distinct mechanisms. The first is direct cognitive engagement — conversation, joint problem-solving, and the ongoing mental work of maintaining relationships requires the kind of cognitive load that helps preserve neural function. The second is stress regulation — close relationships buffer against the chronic cortisol elevation that damages the hippocampus and accelerates cognitive ageing. The third is behaviour propagation — people in close relationships tend to maintain healthier habits and respond more rapidly to early warning signs of cognitive change.
The cumulative effect, across decades, is substantial. The 2024 Lancet Commission on Dementia Prevention identified social isolation as one of 14 modifiable risk factors for dementia, with approximately 4.6 percent of dementia cases attributable to social isolation. Combined with the 13 other modifiable factors — hearing loss, depression, smoking, obesity, alcohol use, hypertension, traumatic brain injury, physical inactivity, diabetes, air pollution, low educational attainment, untreated vision loss, and uncontrolled LDL cholesterol — the Commission estimated that approximately 45 percent of dementia cases globally could in principle be prevented or delayed through modification of these risks. Social isolation is one of the factors that has historically received the least public health attention, despite the strength of the evidence for its causal role.
The limitations
The Harvard Study of Adult Development has substantial limitations that should be acknowledged. The original cohorts were entirely male, predominantly white, and concentrated in a single American city during a specific historical period. The findings may not generalise straightforwardly to women, to other ethnic or cultural backgrounds, or to the very different patterns of family, friendship, and community connection that characterise the early 21st century. The findings are also correlational rather than experimental — the study cannot prove that maintaining close relationships causes better health and cognition; it can only show that the two are strongly associated across 87 years of data.
What can be said with confidence is that, across the longest continuous study of human lives ever conducted, the single strongest predictor of late-life flourishing was not what most of the participants had spent their middle years optimising for. It was not professional accomplishment, not wealth, not physical fitness, not intellectual achievement. It was the quality of the relationships the participants maintained with the people closest to them. According to a Harvard Crimson profile of Waldinger and his co-director Marc Schulz of Bryn Mawr College published shortly after the 2023 release of their book The Good Life, the implications for individual decision-making are concrete: Waldinger noted that “the people who were the happiest, the healthiest, all that good stuff were the people who were more active, who were more intentional” about making relationships. Investments in relationships in middle adulthood compound over time, with measurable returns on health and cognition decades later. The investment that pays the highest cumulative return, by the time a person reaches their eighties, is the time and emotional attention they spent on the people they loved in their fifties.
