How a pioneering study of child health has influenced a generation of parents

25 September 2019 - The Guardian 

At a bright green table scattered with toys and juice cups sit Charlotte and her six-year-old daughter Amelia, while two-year-old Isabella races a yellow plastic shopping trolley across the floor. Being here is much more fun than school, announces Amelia, as she shows off the animal plaster that covers her (apparently painless) blood test to a smiling Professor Jean Golding.

The emeritus professor of paediatrics and perinatal epidemiology at the University of Bristol is founder of the hugely influential Avon Longitudinal Study of Pregnancy and Childhood (Alspac), better known as Children of the 90s (CO90s). She has just turned 80 and has just published her latest paper on child health.

Charlotte, who doesn’t want her surname used, belongs to the original cohort of children, her mother recruited in pregnancy, while Amelia and Isabella are the third generation – the so-called Children of the Children of the 90s (COCO90s). Their arrival has made this study unique for the length, breadth and depth of its data on medical, social, genetic, environmental, and inter-generational factors affecting our health and wellbeing. Now involving 27,600 people, a biobank of 1.2m samples (including DNA, urine and placenta), and countless questions and answers, it has led to more than 2,000 research papers and had a significant impact on policy.

The UK and US governments’ 1990s Back to Sleep campaigns (advising that babies sleep more safely on their backs) were informed by Alspac work, as was the US government shift to advise women to eat fish in pregnancy. And data from CO90s babies revealed that using peanut oil on infants’ skin – then common in baby creams – increased the risk of allergic sensitisation.

One of Golding’s favourite discoveries was that kids can be too clean: “ultra-clean babies are more likely to get asthma”, she says. And that household chemicals (especially aerosols) can do direct harm to children – though she adds, “I’m not sure that these results have had enough impact [on policy and behaviour].”

Golding’s idea for Alspac was considered “crazy… impossible”, she admits, when she first proposed following every pregnancy in the Bristol area with a due date between April 1991 and December 1992, and then studying the parents and children as they grew. The big funders, including the Medical Research Council (MRC), refused to commit to “this monster”, says Golding, waiting almost a decade before getting involved in what Bristol University had started. “And scientists said it would never work,” she adds. “They said we should concentrate on one topic.”

The whole point was to collect as much information as possible – including data they didn’t yet have a use for – maximising the ability to make unexpected connections. She wanted it to be open access too, and Alspac now has more than 800 registered users in dozens of countries. In the last six years, 1,000 COCO90s have been added to the cohort, and a further 2,500 are expected in the next five years as the CO90s tip into their 30s.

The study’s 2,000th research paper, published this spring, found today’s new mothers to be more educated and less likely to smoke than their own mothers, but with higher body mass index and higher cholesterol. More had caesarean deliveries and heavier babies whom they were more likely to be breastfeed. The most striking difference emerged from another Alspac study, says Golding: “a depressing increase in (prenatal) depression in the mothers” – up by more than 50%.

The reasons aren’t yet clear, but Alspac is working on it. In the meantime, the finding has led to calls for better mental health screening in pregnancy and greater availability of treatment.

As a child, Golding spent months at a time in hospital and contracted polio at the age of 13, which left her permanently disabled. Being ill so much meant she “spent a lot of time watching,” she says. “So I really am an observer.”

Clearly clever at school, she only realised many years later that her teachers were fighting over which of them would get the kudos of sending her to read their subject at a top university. Her interests were zoology, psychology and chemistry, but the maths teacher convinced her that she would be unable to do lab and field work - “I had a calliper and on uneven ground I would fall down” - so she went to Oxford to read maths.

“I attended the first statistics lecture and thought ‘this is rubbish’. They kept saying ‘approximate’. I don’t like approximate. I like to know exactly.” She switched to another branch of maths. She uses stats now of course? “Yes,” she agrees, “but with care!”

She stood down as scientific director of Alspac aged 65, but her research is going strong. This month she published a paper showing that the children of mothers who take paracetamol in pregnancy (around half of mothers-to-be in Europe and America) are more likely to have behavioural difficulties, including hyperactivity in their early years. Though these particular difficulties resolve by age eight, Golding says other associations may come to light and advises, “pregnant women should reduce their intake of drugs, including over-the-counter drugs, whenever possible”.

Her research results often bring surprises, she says. In her current work on “fascinating” transgenerational effects, she recently found that if the grandmother smoked in pregnancy (yes, the grandmother), the grandchild was likely to have a higher than expected birthweight but a slightly raised risk of autism and obesity. Traumatic events cascade down the generations, sometimes in unexpected ways. “Shock in one generation can be linked to a higher rates of obesity later on,” she says. “And one study involving grandparents who had been in one of the great famines of the second world war found that their sons have better mental health than others.” Golding is now studying families with refugees, war deaths and similar challenges in their background and is “open to any results”.

Her other main focus is locus of control. This identifies people as either “internals”, who feel they have some control over their fate, and “externals”, who think life happens to them and there isn’t much point in trying to change it. “The main psychologist in the field pushed for us to include questions on this from the start,” says Golding, “and it turns out to have a big effect on the way people live and parent.

Clearly, Golding isn’t retiring any time soon. “I think I have enough to keep me going until I’m 100,” she says. Meanwhile, younger colleagues are developing new techniques for direct collection of data – including via smart devices and miniature head cameras – and examining social media, both for its use in data gathering and the effect it has on us.

“We have followed people from an era when there was no real access to the internet, to it being inescapable,” says Alspac’s principal investigator, professor Nic Timpson. One ongoing piece of research is looking at the link between material on the internet and self-harm and suicide. It has so far found that 22% of young people have viewed self-harm and suicide-related content online and 70% of those who have self-harmed with suicidal intent have accessed such material. Already presented to an all-party parliamentary committee, the work is feeding into development of mental health policy, as well as a joint approach with the Samaritans to try to get YouTube to change what it allows on its site.

Alspac’s role, Timpson believes, is to “make observations, explain them to government, push for specific policies and see practical change”. It’s a powerful model that has already proved its worth.

Charlotte is gathering her daughters ready to go home. She has always been proud to be one of the Children of the 90s, she says, and she hopes Amelia and Isabella will be too. With luck, she adds, she’ll see their children joining as the COCOCO90s, bringing still more insight and improvement to our health and wellbeing.