Decades-Long, 1,000 Strong Dunedin Study Asks Why People Are the Way They Are
What ultimately shapes who we are as humans, nature or nurture? Arguments back and forth over that question have raged on for centuries. A person who excels in school can chalk those academic accolades up to good genes, some say. No, no, others argue. Success or failure hinges on childhood experiences, for better or for worse.
The folly of that debate was made clear when a scientific team led by Professors of Psychology and Neuroscience Terrie Moffitt and Avshalom Caspi showed in a 2002 Science paper that people who had been treated poorly as children didn’t necessarily grow up to mistreat others. Likewise, people who had an “antisocial” variant in a neurotransmitter pathway gene called MAOA didn’t necessarily become violent. Rather, it was individuals who were both mistreated during childhood and carried the variant that were most likely to repeat the cycle of violence.
Their result was among the first to show what many understood intuitively: genes and environments don’t work in isolation. The landmark finding was one in a long line based on more than 30 years of data documenting the health, circumstances and lifestyles of 1,037 babies born in 1972 in Dunedin, New Zealand, an endeavor called the Dunedin Multidisciplinary Health and Development Study. The study that started with a focus on the effects of environmental factors on human development has since evolved and grown to include genetic and genomic data, to explore how genes and environments interact, and to inform policy decisions that improve people’s lives.
Life as a Lab
Since the beginning, Dunedin researchers have checked in with that original group of New Zealanders every two or three years, watching as they grew from childhood to adolescence and on to adulthood. Meticulous records of every study member’s whereabouts and careful attention to their individual needs have kept even the highest risk volunteers coming back – in fact, 96 percent of them participated in the most recent assessment, which just ended last spring.
“Most longitudinal studies only see half of what is happening in the real world,” says Dunedin Study director Richie Poulton, a professor at the University of Otago in New Zealand. “But the people who drift away are not a random group; they are the ones who are on the run from the law, in and out of institutions, who can hardly get out of bed in the morning, let alone travel great distances to participate in a study. We make every effort to catch the true range of life experiences and the problems that result, and that gives us faith in our findings.”
No matter where in the world those study subjects call home, when it comes time for an assessment phase, they are brought back to Dunedin to spend the day in a ten-room research unit. Each room is specially designed for a specific arm of the project: the mental health room has a psychiatrist ready to conduct a neuropsychiatric interview, the cardiovascular room has an exercise bike and tools for measuring blood pressure and heart rate, the asthma room has equipment for performing respiratory tests. In other rooms, researchers capture data on oral health, injuries, sexual and reproductive health, substance abuse, cognitive functioning, and everyday psychosocial issues like money management.
Moffitt got involved with the Dunedin Study when she finished graduate school in 1985. In her first project she gave the study subjects, who were 13 at the time, a number of simple neuropsychological tests to assess their cognitive abilities. When she and Caspi administered the same tests 25 years later, they found that the IQs of the subjects who began smoking marijuana heavily as teenagers had dropped by eight points.
There are as yet no brain scans showing how that drug use affected brain development (those tests weren’t even available in the 1980s). But Moffitt and her team have now applied for federal funding that would allow her to do neuroimaging on the subjects now that they are 38-year-old adults.
“Most longitudinal studies only see half of what is happening in the real world. We make every effort to catch the true range of life experiences and the problems that result, and that gives us faith in our findings.”
“The beauty of a longitudinal study like this is you can follow people through time and establish what comes first and what comes after,” Moffitt said. “The downside is that, of necessity, some of your most important measures were taken 25 or 30 years ago, and science marches on.”
As Associate Director of the Dunedin study, Moffitt is constantly looking five to ten years into the future, scanning the horizon for the next big thing in science and technology. The researchers began collecting DNA in 1996, at a time when many behavioral scientists didn’t fully grasp the power of genetic information. In fact, when Moffitt let the NIH know what they were doing, she quickly got a call from her program officer with a warning not to use federal money to collect DNA. She was told that the scientific benefits didn’t outweigh the ethical risks. But Moffitt and Caspi had already made up their minds, based on a talk they’d heard geneticist Robert Plomin give months earlier on the complex origins of mental illness.
“He said that the genetics of mental illness is too important to leave to geneticists, because they are going to approach it from a purely biological perspective and not take into account the environmental factors that influence how people make their way into mental illness, such as losing their parents young, enduring overwhelming traumas, or facing grinding daytoday stress,” said Moffitt, who found other funds to support the DNA-gathering part of the research. “We collected the DNA because we recognized its value, even though we still had no idea what we were going to be able to do with it.”
Nature via Nurture
That decision put the Dunedin team in a good position when the human genome revolution began in earnest. Most traditional geneticists use a “case-control” approach to discover disease-causing genes, comparing the genetic code of people who already have a disease with those who are well. In a different and equally valid approach, the behavioral psychologists at Dunedin started with people exposed to the causes of the disease and then followed them to see what distinguished those who went on to develop the disease from those who did not. The approach is what led them to their landmark discovery connecting the MAOA gene variant with the outcomes of child abuse.
“That paper received much fanfare and spurred many colleagues to look for these types of interactions in their own work,“ Poulton explained. “There were also some colleagues who were skeptical; they thought that it was luck and wouldn’t happen again in our lifetimes.”
Within a year, the Dunedin researchers published a second paper, demonstrating the same basic phenomenon with a different gene. Among people who had endured childhood trauma, those with the shorter version of the serotonin transporter gene were more likely to suffer depression than those with the longer version.
They followed with work establishing links between the COMT gene, marijuana use, and the development of psychosis; the FADS2 gene, breastfeeding, and an increase in IQ; and the p53 gene, tobacco smoking, and an accelerated decline in lung function. Each of these discoveries helped to explain a vexing question: Why do people turn out so differently even when they’ve had very similar experiences?
Some argue that the Dunedin study lacks power to separate cause and effect due to its uncontrolled, observational nature. But Jay Belsky, a professor of human development at the University of California-Davis, says that this argument misses the value of longitudinal studies in providing real-life insight into the human condition.
“Nobody is going to let you do a controlled experiment where you mess up people’s lives and see what the long term consequences are. Astronomers can’t move the galaxies around, geologists can’t move the land around, and life course observers are working with something they can’t move around either,” says Belsky, who is assessing the experiences of the Dunedin study subjects as they become parents. “Even a randomized controlled trial, which is the gold standard of cause and effect, has its own problems, in that the people who participate are not necessarily representative of the real world population. Every study has its limitations and its strengths.”
Genomics of Human Experience
The real strength of the Dunedin Study lies in the almost 40 years of extensively detailed information amassed on virtually every aspect of human health and experience, from cholesterol levels and physical fitness, to intelligence and criminal activity, to sexual health and infertility. As technologies for probing the human genome and epigenome evolve, researchers can continue to tap into this unique resource to explore new theories about the causes of disease and in new ways.
“We are in an excellent position to look at a genetic influence to see how it unfolds over time to shape the course of behavior,” Caspi explains. “Let’s say that genome-wide association studies reveal hits in particular regions of the genome; the question then becomes: What do they mean? When do the effects of a discovered genetic variant come into play? How do they shift the progression of a particular behavior over a lifetime?”
Since the Dunedin Study began, it has generated more than 1,200 research articles, reports, books and book chapters over a wide range of topics. Among those pertaining to genetics, the focus has shifted in recent years from a proof of principle that genes and environment interact to an understanding of how that interaction gives rise to problems in life.
Three of the most recent papers have begun to make the leap from single genes to genomes. In them, the Dunedin researchers compiled all the known genetic vulnerabilities for a given condition – obesity, smoking and asthma to start – to create a combined risk score for each Dunedin participant. For instance, they added up the 32 variants associated with a risk of obesity to give individuals a score between 1 and 32. Once they had assigned risk scores, they compared them with what they knew about who had become obese, a smoker, or an asthmatic.
Children with a high genomic risk score for obesity began to gain body fat years earlier than their peers, putting them on a path to become obese in adulthood, they found. While a high genomic risk score for smoking did not predict who would take up the habit, it did predict how quickly they became dependent on tobacco once the first cigarette was lit. The genomic risk scores for asthma separated those children with chronic asthma from those with only intermittent symptoms.
Malcolm Sears, a professor at McMaster University in Canada who consults on the respiratory health arm of the Dunedin Study, says that these studies show that genetic risk factors exert their effects early in life and could give clinicians an opportunity to intervene before the damage is irreparable. The opportunity to improve a particular outcome – to take a person off a bad trajectory and put them on a better one – underpins the work of every researcher involved in the Dunedin Study, and especially the leadership of Moffitt and Poulton.
Changing the Course of Fate
“As the director, I think of myself as the chief guardian of these 1,000 lives,” Poulton said. “Too often as researchers, you toil away, often in isolation, then you publish and get your academic accolades, and you talk to your small fraternity of researchers that might number the people at a city bus stop, and then it doesn’t go much further. I want my legacy to be that the generosity of people who participated in this study was rewarded in terms of having an impact on public health.”
Because the Dunedin Study has spanned so many years, many of the study subjects have children who are already benefiting from policy changes in completely unexpected ways. For example, the Injury Prevention Research Unit, which over the years has studied everything from playground falls to car crashes, found that the long cords on electric tea kettles were a hazard to children who could pull the pots of boiling water onto themselves. After that report came out, New Zealand changed the product standards to restrict the length of those cords and address a major cause of burns in young children.
The impact of the Dunedin Study has also been felt on this side of the pond, where research findings have informed many rewrites of the mental health professionals’ guidebook or DSM (Diagnostic and Statistical Manual of Mental Disorders). Work on the developing teen brain led to the overturning of the death penalty for juveniles by the U.S. Supreme Court, an accomplishment that garnered Moffitt the 2007 Stockholm Prize, the Nobel equivalent for research in criminology. In fact, the Dunedin Study is represented in a great number of health policy documents, from the U.S. Surgeon General’s report on adolescent mental health, to the most recent series of reports on child development from the American Academy of Pediatrics, to a report on medical marijuana from the Executive Office of the President.
One of the overarching themes emerging from the Dunedin Study is that early life adversities can affect important aspects of life, like academic achievement or cardiovascular health, decades later. Lack of nourishment as a baby, trauma as a child, or excessive drug use as a teenager all can have life-altering ramifications.
“Why are we spending all this money trying to make people’s lives healthier in their thirties, forties and fifties when actually there is much more bang for your buck when you intervene earlier in life?” Belsky asked. “This is where I think Moffitt and Caspi are on the leading edge of showing that people who are dying in their sixties are doing so not just because they are unlucky, but because of their life history.”
Tim Strauman, who recruited Caspi and Moffitt when he was chair of Duke’s Department of Psychology and Neuroscience, explains their impact this way: “The work of Professors Caspi and Moffitt with the Dunedin study has revolutionized how scientists think about human nature…They have shown that gene-environment interaction is the rule, not the exception, and that understanding the origins and life course of particular traits or disorders is complicated and challenging – but possible nonetheless. Their findings have stimulated an enormous new research literature in which multiple disciplines and levels of analysis are used to take on the most vexing questions, like predicting who will become schizophrenic or depressed and why.”
Dunedin researchers are now investigating how adversity and health in the first half of the life course can influence how aging unfolds. Moffitt and Caspi are about to submit a paper showing that chronic stress – from disorders like PTSD or persistent depression – can erode telomeres, those caps at the ends of our chromosomes that have been linked with cellular aging. If funding allows, the latest in neuroimaging technology will soon give them a new window on mind-body connection through aging.
When the Dunedin Study was first conceived in 1972, no one could have imagined that the research would eventually turn from measures of height, weight, and psychological development to measures of wrinkles, bone loss, and dementia. But the constantly shifting nature of the work – as well as its longevity – is what makes it so unique and exciting.
“As a scientist, this work keeps me energized,” Caspi said. “Once upon a time I was interested in child health, then puberty, then heart disease, and now cognitive decline. It is certainly a different way of doing science. I let the study members guide me to the important problems.”