Q&A: George Davey Smith on Children of the 90s

George Davey Smith

The Avon Longitudinal Study of Parents and Children (ALSPAC), otherwise known as ‘Children of the 90s’, is a large-scale longitudinal study that collects biological, psychological and environmental data from around 14, 000 children and their families.

After years of being questioned, one of the ‘children’, Emma James, poses her own questions to the study’s Scientific Director, Professor George Davey Smith.

I’ve recently been looking back on my own lifetime involvement with Children of the 90s. How did you come to be a part of ALSPAC?

I’d been involved with the study, though not as the director, for quite a long period of time. I’m very interested in early life influences on development and health and disease, for example, how things that happen to us before we’re conceived or during pregnancy can influence our foetal development. Similarly, how things we’re exposed to in early life can relate to our health and disease later in life. Clearly, a birth cohort that recruits mothers when they’re pregnant is the ideal way to look at these things.

As an epidemiologist, I’m interested in things you can change that might improve health and development. I’m interested in genetic differences as ways of learning about modifiable processes (those things we can change, like our diet, as opposed to our gender or age, which we can’t). Children of the 90s has always been ahead of its time in collecting genetic material and having a large sample size to look at these issues.

As participants, our main contact with the research team is at the data collection stage. I’ve always wondered, how many researchers are working with our data exactly?

In total, there are nearly a thousand research projects registered to use the data and there are several hundred active projects at the moment. That’s fantastic, as there’s a huge commitment from participants and a great amount of time and effort spent in collecting the data. The more that ALSPAC can contribute to scientific understanding and processes in ways that can improve prospects of future generations the better. It’s the ultimate mark that everyone’s time and effort has been spent in a worthwhile way.

Are there particular connections you’ve been looking for from the beginning?

That’s a very interesting question. Among the extraordinary foresights of Jean Golding [who established ALSPAC] was the fact that she asked scientists from all different fields to suggest questions that could be asked, and collected data on everything she could. In many cases there was just no hypothesis at the time. The obvious examples are some of the genetic findings. For example, ALSPAC was key to identifying FTO, the first common genetic variant (carried by 10-20 per cent of the population) related to obesity. No-one had any idea that this gene would be involved in obesity. It was known at the time as ‘fused toes’, because all that was known about it was that, in mice, mutations caused the toes to be fused. But then it turned out in humans to be important in relation to obesity. There have been lots of such unexpected findings.

Many participants don’t like giving blood, or get too busy to fill in questionnaires. If we choose not to take part in a particular aspect of the study, how much does that affect the science?

Firstly, it produces a smaller sample size, which is bad news because in science we need as big a sample size as possible to have the greatest power and statistical ability to demonstrate things. The second issue is that the study sample becomes much less representative. You need a representative spread of people, behaviours, and types of people to be able to get results that can be generalised to other populations. So any non-response makes the results less valuable.

Some participants think that once they’ve stopped responding then we don’t want them back. But of course we do! Many of the things we look at relate to an exposure at one time, such as relating your birth weight to your glucose levels at age 18. If you haven’t filled out questionnaires in the interim then that doesn’t make that any less valuable. There are also very good statistical ways of taking into account periods where people haven’t responded. So if you’ve ever responded or have been included in Children of the 90s then we’d love to have you back at any time!

Has lack of participation ever been an issue?

Well, one issue is when people don’t respond to questions where we’re asking their permission to do things like link our data to GP records. We know from follow-ups that for the vast majority of people this isn’t because they don’t want this to happen, it’s just because they’re busy. So it’s a problem when there are consent issues. If people are happy for their data to be used as widely as possible for very valuable science, it’s important that they respond to those sorts of consent requests.

What are your plans for the study?

We’re going to include the fathers and partners more. The mothers were offered a clinical examination recently and this autumn we’re going to start a clinic for the fathers. They haven’t taken part to the same level as the mothers in the past, although they still have filled out many, many questionnaires about themselves and their children for us!

Many of the participants have brothers and sisters, so we’re working towards including their siblings, although obviously only if they want to be involved, Everything only happens if people are interested in taking part.

The other exciting thing is that some of the participants are beginning to have children themselves, so we are getting ready to include the next generation. Children of the 90s is the best study in the world looking at how what happens during pregnancy relates to what happens to children, but there’s also a lot of interest in things that are transmitted from one generation to the next. So we’re really excited about the prospect of the study spanning three generations.

Will the next generation be tested in the same ways we were?

We’re planning similar sorts of approaches, but of course technologies have developed considerably over twenty years. For example, we have digital technology for recording data rather than paper questionnaires, including the use of sound recording and digital film capture. We’re working hard to develop and think about how we can incorporate new technologies of data capture and data collection into the study.

Many of us are away at university now, and might in a few years be settling down elsewhere in the country. How will you keep in touch with us?

This is obviously an important issue, and one reason for looking at new ways of data capture, rather than having people come to see us as they do now. We’re looking into the use of web-based data capture, telephone, mobile phone apps and postal samples. However, we do think that many people will maintain family links in Bristol, so we’re going to try to set up a more flexible system whereby we can see people when they come back at Christmas, Easter or other holidays.

Finally, if you could discover anything from the data, what would it be?

That’s a really good question! As an epidemiologist, I’d like to find some novel way of reducing the risk of disease. However, it’s not as though I have one very particular nut I’m trying to crack. For me, developing ways to get reliable information on how modifiable factors can influence health and development is what’s most exciting.

Emma James

Emma James is a summer intern at the Wellcome Trust and has been an ALSPAC participant all her life. Read her account of being a ‘Child of the 90s’.

ALSPAC is funded by the Wellcome Trust and the Medical Research Council.