The Dunedin Multidisciplinary Health and Development Study began at the age-3 follow-up in 1975-76, recruiting from a larger perinatal study run by paediatrician Patricia Buckfield. The longitudinal study, originally led by Phil Silva — then a former primary-school teacher and psychology student researching child development — recruited 1,037 of the babies born at Queen Mary Maternity Hospital in Dunedin between 1 April 1972 and 31 March 1973 whose families were still living in the wider Otago region by age 3. The funding kept coming. The study kept going. The babies grew up, took jobs, had children of their own, entered middle age, and continued to participate in detailed assessments every few years. The Dunedin Study, as it eventually became known, is now in its age-52 assessment phase, conducted between 2024 and 2026.
The result is one of the most complete portraits of human development ever assembled. According to the Dunedin Study’s own institutional summary, the 1,037 participants have collectively contributed to more than 1,400 peer-reviewed journal articles, books, and policy reports. At the most recent completed assessment, conducted at age 45 between 2017 and 2019, 94 percent of living study members participated — 938 out of the 997 surviving cohort — making it the highest follow-up rate for a longitudinal study of this design and duration anywhere in the world. Dropout rates of 20 to 40 percent are common over a single decade in comparable studies. Dunedin participants have stayed connected to the study for fifty years, with most travelling back to the research unit at the University of Otago in person for their assessments, even when they have moved overseas. The Dunedin cohort has been described, in a 2018 Science profile, as “the thousand most studied individuals in the world.”
The single most consistent predictor
The most influential single finding to emerge from the study, of the many it has produced, concerns the long-term consequences of a personality trait observable in early childhood. In 2011, an international team led by Terrie Moffitt and Avshalom Caspi of Duke University and King’s College London, working with Richie Poulton at the University of Otago, published a paper in Proceedings of the National Academy of Sciences titled “A gradient of childhood self-control predicts health, wealth, and public safety.” The team had pulled together every measurement of self-control collected from the Dunedin cohort between ages 3 and 11 — observations by examiners, parental and teacher reports, self-reports, behavioural tasks measuring patience and persistence — into a composite score for each child. They then examined which adult outcomes at age 32 were predicted by that single childhood measure.
The answer was: nearly all of them. Children who scored low on self-control as children, regardless of their IQ or their family’s social class, grew up to have substantially worse adult outcomes on a wide range of measures. They had more breathing problems, more gum disease, more sexually transmitted infections, more cardiovascular markers like high cholesterol and blood pressure, and higher levels of inflammation. They had less savings, lower rates of home ownership, more credit card debt, and were more likely to be single parents. They were more likely to have criminal convictions on their record and to be dependent on alcohol, tobacco, cannabis, or harder drugs. The gradient was consistent across the full distribution of self-control scores. The lowest-scoring children had the worst outcomes; the highest-scoring children had the best; and every band in between sat exactly where the gradient predicted. The 2011 paper specifically measured outcomes in three domains — physical health, personal finances, and what the team called “public safety” (criminal convictions and substance dependence). Subjective happiness was not directly measured in that paper, though related Dunedin work on life satisfaction and mental health has produced compatible findings.
What the researchers measured
The “self-control” the Dunedin team was capturing is not a single mysterious quality. According to Duke Today’s coverage of the 2011 paper, it is a cluster of related characteristics that psychologists have studied for decades under various names: conscientiousness, self-discipline, perseverance, the ability to delay gratification, the capacity to plan ahead, impulse control, and the ability to regulate emotions in the service of longer-term goals. The Dunedin children had been assessed on all of these dimensions through behavioural observations, structured tasks, and reports from people who knew them well. A child who scored high on the composite was, broadly, someone who could wait, focus, finish what they started, and consider the consequences of their actions before acting. A child who scored low was, broadly, someone who could not.
The researchers were able to rule out a number of alternative explanations. The pattern was not explained by intelligence — children with high IQs and low self-control still ended up with worse adult outcomes than children with average IQs and high self-control. It was not explained by family socioeconomic status — children from poor families with high self-control outperformed children from wealthy families with low self-control on most measures. Most strikingly, the pattern held within families. The team also analysed data from a separate study of 509 pairs of British twins and found that, within twin pairs, the sibling with lower self-control at age 5 was more likely than their genetically and environmentally similar twin to start smoking by 12, struggle in school, and engage in antisocial behaviour. The effect was operating at the level of the individual, not the household.
Why self-control matters
The mechanism by which childhood self-control translates into adult outcomes is partly intuitive and partly cumulative. The intuitive part is that adolescents and young adults with weaker self-control make more decisions, in any given year, that have negative downstream consequences. They start smoking before they understand the addiction risk. They drop out of education in moments of frustration. They become teenage parents through poorly planned choices about contraception. They commit minor crimes that produce criminal records. Each individual decision is small. Cumulatively, over a decade or two of adolescence and early adulthood, they accumulate into life trajectories that are difficult to reverse. According to National Geographic’s coverage of the study, the Dunedin participants who finished high school without smoking and without becoming parents — i.e., who avoided the three big “snare” decisions of adolescence — had substantially better adult outcomes even when they had started with low self-control. The bad decisions, more than the underlying trait, were what produced the bad outcomes.
The slightly less intuitive part is that even study members who avoided the major snares but had low self-control still showed measurable disadvantages at age 32. The everyday effects of impulsivity — interrupting in conversations, forgetting appointments, failing to follow through on commitments, struggling to save money or stick to medical regimens — produce a continuous accumulation of small costs that add up, over decades, into meaningful differences in health, wealth, and relationships. The Dunedin data captures this drift clearly. Low-self-control adults end up, by their early thirties, in measurably worse positions even when no single dramatic decision can be blamed for the difference.
What this changes
The Moffitt-Caspi paper closed with a note of cautious optimism. The Dunedin participants whose self-control had improved between childhood and adolescence — and a meaningful fraction of them did improve — ended up with substantially better adult outcomes than their early childhood scores would have predicted. Self-control, in other words, is not a fixed trait. It changes, it can be developed, and the improvement transfers to later life outcomes. The implication, the team argued, is that interventions designed to build self-control in early childhood and adolescence could have outsized returns. Several such interventions, including school-based programmes that teach conscientiousness, planning, and emotional regulation, have been evaluated since 2011, with generally promising results. The Dunedin team has continued to publish updates from later assessments, including findings at age 38 and age 45, that have largely confirmed the original 2011 conclusions.
The Dunedin Study itself continues. The age-52 assessment is in progress now, with assessments scheduled to conclude in late 2026. New cohorts of the participants’ own children — the “Next Generation Study” — are being followed in parallel, allowing the research team to examine how patterns of self-control, mental health, and life outcomes propagate across generations. The 1,037 babies born at Queen Mary Hospital in Dunedin between April 1972 and March 1973 have collectively done more to clarify the relationship between childhood experience and adult life than any other group of human beings ever studied. The most reproducible finding is that the children who could control themselves grew up into adults whose lives were measurably better, on nearly every dimension that researchers know how to measure. The hardest part is the part most people already suspected: the trait is much easier to identify in a five-year-old than it is to teach.
