Eminent New Zealand social scientist Professor Richie Poulton, who has died aged 61, was director of the world-famous Dunedin Study for more than 20 years. In this story, first published in the Listener in January 2008, Rebecca Macfie spent time with Poulton, study founder Professor Phil Silva and the team who, by closely monitoring key aspects of the lives of 1000 people from birth, discovered empirical evidence on how early childhood experience shapes the adults we become.
Dunedin in the early 1970s was part of what some would say was a kinder, gentler New Zealand. Mothers received the Family Benefit for every child. There was next to no unemployment. People left their doors unlocked at night.
New Zealand was then the ninth-richest country in the world. We had the fourth-highest level of education spending, the third-highest number of hospital beds, the sixth-highest number of GPs per head of population. It was God’s own country and the best place in the world to bring up kids.
But Phil Silva wasn’t so sure that all was as well as we liked to think with the state of the nation’s children.
An educational psychologist and former teacher, Silva had been involved in a small study of 225 children alongside Dunedin paediatrician Patricia Buckfield, whose main interest was in the impact of new high-tech birth interventions on child development. Buckfield had gathered data on every baby born at Dunedin’s Queen Mary obstetric hospital between 1967 and 1973 – 12,000 in all.
But when Silva assessed a group of the children a few years later, he made some worrying findings: a large number had hearing, vision, language, behavioural and other problems that had gone completely undetected.
Silva decided to take a closer look at a subset of Buckfield’s sample: those children born in the year from April 1, 1972, and whose mothers still lived in the city three years later. The parents were tracked down and asked if they would mind bringing their children in for a health and development assessment. Astonishingly, 91% of those who fitted the criteria were happy to spare the time to help: 1037 three-year-olds were brought in for half a day of tests and measurements. The Dunedin Multidisciplinary Health and Development Study was born.
There was precious little money at Silva’s disposal. The assessments were done at the Sunday school and condemned manse of Knox Church in George St. Dozens of volunteers and a handful of doctors helped out. Each child was assessed against a range of measures: height, weight and head circumference were measured; language skills were assessed; profiles of behaviour were compiled; whether the boys had been circumcised was recorded. Data was also collected about the mothers: their health status, reading ability, personality, child-rearing attitudes.
But one set of assessments was never going to be enough to satisfy Silva’s curiosity. So when the kids were five, their parents were asked to bring them back for a follow-up. Of the original 1037 children, 96% came. The data gathered at age three was updated, and new measurement criteria – discipline, nutrition, dental health – were introduced.
By 1978, only five years after he had begun the study, Silva was able to shock the nation with his findings: one in four children had delayed language development or behavioural problems; 15% had glue ear; the researchers had identified one child who was almost blind and another who might have died of undetected melanoma.
New Zealanders, Silva declared, cared more about their cars than their kids. Cars were required by law to be regularly checked; if children were treated with similar diligence, problems would be picked up in time to prevent lasting damage.
Editorial writers harrumphed in agreement. Politicians took note. Funding began to trickle in with a little more certainty. Silva had launched what the current director of the Dunedin study, Richie Poulton, calls a “juggernaut”.
That colourful language aside, what Silva had done was to lay down a bedrock of data upon which an emerging three-dimensional model of the human condition would be built by hundreds of researchers from a multitude of disciplines over the next three decades.
The parents were asked to bring their children back, at ages seven, nine and 11, to Silva’s Knox Church base. (Every Friday night during assessment periods, his helpers would have to remove their charts and equipment to make way for Sunday school, and set it all out again on Monday morning.)
By the time the children returned at age 13, the study had taken up residence in an abandoned building tucked behind the Dental School. It was (and remains) less ivory tower than grungy dive, but at least it was permanent. A Space Invaders machine was installed for the young study members to use in idle moments.
They came back again at age 15, 18 and 21, by which time the tests and questions lasted for a full day. At each visit, researchers updated the core data, but also delved more deeply into the study members’ health and wellbeing.
At nine, their awareness of alcohol was assessed; at 11, they gave a blood sample; at 13, they were asked about delinquent behaviour; at 15, they had an echocardiogram; at 18, they took a personality test; at 21, they were questioned about assault, sex and substance use. And much, much more besides.
Along the way, parents, teachers, and friends were questioned, and – provided study members agreed – information was gathered from police and medical files.
After the age-21 assessment, the study was described by Bruce Scoggins, then director of the key funder, the Health Research Council, as “the richest archive the world will have on human development”.
By then, fewer than two-thirds of the original sample of 1037 people still lived in Dunedin, but 97% returned to volunteer yet more detail on their health and wellbeing.
They came again at 26 and, most recently, at 32. By then, almost a quarter lived overseas, yet 96% of the 1014 still alive (23 have died through illness, accident or suicide) took part. As in previous years, researchers travelled around the country’s jails to assess those who were incarcerated.
By the end of this year when they will be approaching 38, they will again start filing up the tatty stairway and through the dowdy, utilitarian interview rooms of the study headquarters for a day of scrupulously timetabled 50-minute quantitative interviews and tests.
They will arrive before 8.30am and finish – having disclosed intimate details of their health and happiness, sex life and relationships, work and finances, illegal activities and achievements – after 5.00pm. The old Space Invaders machine will whirr into action for some light entertainment over morning tea and lunch.
Since 1994, about 400 study members have also been visited in their homes for a separate one-off assessment with their first-born three-year-old children. Late last year, their 15-year-old offspring started being assessed. And between 2003 and 2006, about 95% of the surviving parents of study members – 1900 people – took part in two-hour interviews about their own health and wellbeing.
As each layer of data has been meticulously collected, researchers have been able to hunt out more correlations, draw new parallels, bring hard evidence to bear on old assumptions and explode myths.
It turned out that birth complications – the focus of Buckfield’s concern – didn’t set kids up for problems in later life, although low birthweight was found to be associated with low IQ. Thumbsucking was found to do no harm; birth order made no difference to later development; preschoolers whose mothers worked were not disadvantaged.
Delayed language development in toddlers was linked to low IQ and reading deficits down the track. Glue ear – regarded in the 70s by ear, nose and throat clinicians as harmless – was shown to have a lasting impact on behaviour and reading ability. Lax and inconsistent discipline in childhood was linked with higher rates of mental illness at age 15. The longer a person lived in an area with fluoridated water, the better their teeth.
For years, paediatric textbooks had insisted that breastfeeding protected against the development of asthma and allergies. The Dunedin study not only showed – somewhat controversially – that assumption to be incorrect, but even that breastfed children may be at an elevated risk of those conditions.
The study has lent scientific weight to the notion that the child is father of the man. Broad personality traits are laid down by the age of three: under-controlled toddlers grow up to be impulsive, unreliable and antisocial; inhibited three-year-olds are more likely to become unassertive and depressed adults; well-adjusted three-year-olds tend to become well-adjusted adults. Socially isolated children are more likely to develop health problems as adults.
Growing up poor leads to significantly worse cardiovascular and oral health and higher rates of tobacco and alcohol dependence at the age of 26 – and upward social mobility does not undo the damage. Victims of childhood maltreatment have elevated rates of inflammation in adulthood, which is a risk factor for heart disease and respiratory illness.
Latterly, the Dunedin study has opened a new scientific frontier with revelations about the way genes and environmental factors work together to influence behaviour. As outlined in the Listener (“Nature via nurture”, November 17, 2007), it has been shown that boys who have been abused in childhood and carry a certain variant of a gene involved in the breakdown of mood-regulating neurotransmitters such as dopamine and serotonin are three times as likely to be convicted of serious crime by age 26 as those who endured similar childhood abuse but had a different variant of that gene. Those with a particular variant of a serotonin-controlling gene are twice as likely to suffer depression when faced with traumatic events. Teenage dope-smokers with a particular genotype are at heightened risk of psychosis in their twenties.
And for some babies lucky enough to be born with a certain variant of the fatty acid-processing gene, FADS2, breastfeeding will boost their IQ by around six points; those without that variant will gain no IQ advantage from breastfeeding.
None of these groundbreaking findings into the way nature works via nurture would have been possible without the deep history of data amassed by the study over more than three decades.
Some of the most far-reaching insights to have emerged from the study relate to the origins of criminal behaviour. One of the leading lights in this work has been Professor Terrie Moffitt, of the Institute of Psychiatry at the University of London’s King’s College, who is also an associate director of the Dunedin study.
Last year Moffitt was awarded the Stockholm Prize – criminology’s Nobel-Prize equivalent – largely for work based on the Dunedin study, which she has been involved with since 1985. From the early 90s, she was able to start teasing out two distinct subsets of delinquents – those who got involved in crime as teenagers as a rite of passage and then grew out of it, and those who progress from thumping and biting other kids at preschool, through car conversion in adolescence to rape and murder in adulthood. For these so-called “life-course persistent” criminals, Moffitt has shown that a constellation of factors in very early childhood has combined to set them on their hellish pathway: low IQ, attention problems and hyperactivity, the last of these possibly rooted in neuro-psychological deficiencies. As these problems cause relationships with parents and teachers to deteriorate, the downward spiral accelerates and they become ensnared in a life of deviancy by illicit drug-taking, illiteracy, poor work histories and stints in jail.
The “life-course persistent” are only 10% of the male population but account for half of all crime, and also carry more than their share of ill health. They’re two or three times more at risk of heart disease, stroke and gum disease and, by the same margin, more likely to be smokers and suffer injuries.
Moffitt says the Dunedin study paved the way for understanding the roots of criminal behaviour in childhood.
“Before our reports from the Dunedin study appeared, the standard approach to crime control was to put offenders into prison in their late twenties, after they had already done considerable damage to society and their victims.
“But after the Dunedin findings revealed the strong childhood roots of violence, most governments in developed countries are now implementing early childhood programmes to prevent criminal lifestyles from taking root, before victims are harmed. I think most pundits in the US, the UK, and Canada would agree that the Dunedin findings were highly influential in shifting the focus today from prison to prevention.”
The sheer scope of the study’s findings over the years is little short of breathtaking. More than 1000 scholarly papers have been published, documenting in “glorious detail”, as Poulton says, “the risk pathways to violence and madness; to fatness and sadness; from gum disease to heart disease; from [chronic obstructive pulmonary disease] to herpes; from failure in school, and the workplace, and in relationships”.
Moffitt says no other study worldwide is as scientifically powerful, and it is increasingly famous outside New Zealand. Sir Michael Rutter, professor of developmental psychopathology at London’s King’s College – a man regarded as a global doyen of developmental science – considers it the “best ever” longitudinal study. He says it wins respect for its methodological rigour and the researchers’ “outstanding record of using the study to tackle really important scientific issues that have important implications for policy and practice”.
Key to that scientific credibility and international standing is the extraordinary loyalty of study members. No other longitudinal study in the world matches its retention rate of 96% at age 32.
Luck has something to do with it: Dunedin in the early 70s was close-knit, homogeneous and intensely loyal to its university and medical school. It was large enough to produce a study population that (although whiter and slightly better off than the nation as a whole) would produce findings that could be generalised to New Zealand and other developed countries, yet small enough that one in every 10 households had a child or relative in the study.
Because money was scarce in the early years, up to 100 volunteers were involved in helping out and they all served as “wonderful ambassadors”, adding to the study’s profile around town, says Silva.
But more important than luck and circumstance was Silva’s sheer doggedness and commitment. He recalls that when the study members were 13, participation dropped off (only 82% came in for assessments). Urgent remedial action was required.
“I realised the most important asset we had was our sample and it was more important for me to be out there in the field, catching them and bringing them in.”
It makes it sound as if he was out there hauling truants into his paddy wagon, which he wasn’t. But he did go to great lengths to keep them in the net: at one stage he tracked a young man down at his dope plot in South Canterbury. He says his scariest moment was knocking on the door of the White Power headquarters in Christchurch to persuade a study member to come down to Dunedin for an assessment. (“They all treated me with extreme courtesy and cups of tea.”)
When the participants were 15, Silva had the use of a friend’s bright red Jaguar, which became the official collection vehicle for those he thought were at risk of slipping out of his grasp. “We just couldn’t afford to lose them.”
Silva and Poulton believe a reservoir of goodwill and trust has been earned not only through persistence, but because study members have been treated with equity and consideration. They are able to refuse any test or question and each step in the assessment process is explained. They are also kept up to date with the steady flow of findings to which they have contributed.
Confidentiality is paramount – Poulton won’t even say exactly how many have committed suicide, because the families affected would know who he is referring to. Similarly, he won’t say how many are in prison. Nothing must erode the trust that’s been built up over the years.
Says Silva, who retired in 2000 and carries the title director emeritus: “We respect them, and they know that. We enjoy them, and they know that. We make sure we look after them when they are here. And we listen to them. They want to talk about these things and we don’t judge. We hear some pretty horrendous things, but we don’t judge … It’s been a real privilege to be privy to all this knowledge, and especially the trust. And the parents have been wonderful at rounding up the kids and encouraging them to come, at great inconvenience to themselves very often.”
Today, study members seem to require little inducement to come back every six years for assessments. There’s nothing in it for them other than a return flight and a couple of nights’ accommodation for those who have moved away from Dunedin (two-thirds of the original cohort), lunch and $100 for the day to cover expenses.
Enormous effort goes in to maintaining contact and removing obstacles to participation. Susan Newlove, who was employed as a part-time tracer for 18 months leading up to the age-32 assessment, patiently worked the phone and sometimes went door-knocking in a bid to pick up the scent of people who had gone, no address.
Bronwyn Driver is the study’s in-house travel agent, arranging flights from all over the world and doing her best to accommodate requests to spend Christmas in Dunedin with family, or to coincide with old mates who are also in town for an assessment. For the two years leading up to and during an assessment period, she’s available by cellphone 24 hours a day to solve any problems that crop up. “It’s taxing, but very, very rewarding.”
Now everyone wants to stake a claim to this goldmine of information. But Poulton, a clinical psychologist who was employed as Silva’s deputy in 1995 and took over as director in 2000 at the age of 37, is a conscientious guardian of both the study members and the wealth of data in his care. The study must remain faithful to its six key domains – mental, cardiovascular, oral, pulmonary, sexual and musculoskeletal health – and not be tempted into distraction.
“We stick to our strengths. Opportunistically you could add a 10-minute assessment about some new thing and generate 10 papers, but we don’t need that. We’re into getting some depth and making linkages across the disciplines, so it’s about doing the assessments better, more cleverly.”
It costs $1.5 million a year to run the study, and Poulton has costed it down to one-minute blocks. Researchers who make the cut must come with their own funding – for a 50-minute session they need $250,000.
Funding has got easier since those early days when Silva couldn’t be sure he was able to pay the rent to Knox Church, but Poulton says even now the study has to constantly prove itself to funding agencies. The Health Research Council, and its predecessor the Medical Research Council, a key backer since 1977, awards funding in six-year blocks. Money has also come from agencies such as ACC and the Health and Education Ministries. About a third comes from overseas – the prestigious US National Institutes of Health has been a funder since 1985, and, more recently, the UK Medical Research Council has also backed it.
After 35 years, Silva says the study is “just at the beginning”. He’s excited by the potential for research into the interplay between genes and the environment to bring new insights into human behaviour.
But in the meantime, some absolutes have emerged: love is necessary, but not sufficient – children also need luck, he says. For instance, a mentor in a child’s life can make an enormous difference. And the more experiences a parent gives a child, the better, particularly when combined with language – simple things like mucking around in a puddle or climbing a tree are as valuable as expensive outings.
Yet, the more the Dunedin study has probed the mysteries of human development, the more remote a simple recipe for health and wellbeing seems to have become. Yes, says Poulton, some things are clear: children need affection, boundaries and affirmation; no child should be abused; satisfactory housing and nutrition are important; poverty is not propitious for a child’s development because it is often associated with risks such as high levels of adult stress.
But beyond this, life remains extraordinarily complex, and black-and-white answers are scarce. “It’s like putting things in a cake mixer and stirring it up,” says Poulton. “We’re trying to work out exactly how much flour, how much sugar. What say you put it in the oven too late? What if the temperature is five degrees off? It’s the subtleties, the grey areas. The grey is where all the gold lies. This is the value of these developmental studies – they give you a window on these complexities.
“I have a sense that I will be long gone and there will still be so many questions.”
If there is a singular lesson to be drawn from more than three decades of Dunedin data, it’s this: what happens very early in life has far-reaching implications for how people’s lives turn out, and the roots of criminality, and poor mental and physical health lie in childhood.
And although the wisdom of early intervention has become widely accepted – thanks not least to the years of Dunedin findings – much more is needed.
“There’s probably an understandable weighting in social policy to the period in life when the problems appear to emerge,” says Poulton. “But everything we know from studies like these is that you can see the roots of problems much earlier. So I would take a portion of the investment that is channelled into early adulthood and later life and really invest in the first decade or 15 years. If you do that the returns will be enormous. If you try to deal with things [after problems emerge], the horse will be well and truly out of the barn.
“Remember dental health clinics at school? We need behavioural health clinics in every school … I think the argument is compelling. If I was the government and I had half a billion dollars to spend, I’d train, over two generations, a whole cadre of clinical psychologists for the school setting. They’re good at evaluation, at intervention, at group teaching, at individual therapy.”
Mental illness is associated with the third greatest loss of life and productivity, after cardiovascular disease and cancer. Yet the Dunedin findings have shown that, among those with a psychiatric diagnosis at 26, half had had a previous diagnosis by the age of 15.
“In other words, if you intervened early, at the first emergence, you could avert up to half the burden associated with mental-health disorders.”
But one-off interventions are unlikely to be sufficient. “A really troubled kid is unlikely to be turned around by the time they get to school, even if you intervene early with the family. So they need to be moved into the next set of supports, and once they outgrow those, the next set after that. That sounds like a lot of supports – but let’s look at the costs if you don’t. They’re enormous.
“It’s clear that the current way of running things with ambulances at the bottom of cliffs is doomed to failure.”
