Thinking Longitudinally and Long-term

There are many things we take for granted these days that were once mysteries. To take a specific example, relevant to the rest of this post, I’d include the fact that smoking during pregnancy has implications for both the immediate and long-term health of the foetus/child. But this discovery came about because of serendipity in a questionnaire design drawn up by some determined social scientists, medics and more back in the 1950s. The finding arose from the so-called 1958 Cohort Study. If you haven’t heard of this – or the other Cohort studies, of which more below – it’s not unduly surprising as their existence isn’t particularly well known, although they probably should be accorded ‘National Treasure’ status.

I was very familiar with the 1958 cohort, as I’ll explain, but the existence of several more longitudinal studies, additional cohorts, I had not really taken on board until I read a fascinating recent book by Helen Pearson (features editor at Nature and a regular on the Guardian pages), called The Life Project. It turns out that the first of these studies was set up in the heady post-war years, funded by the MRC albeit on a shoe string, back in 1946. Known as the National Survey of Health and Development, it interviewed mothers shortly after the birth of their children, the children being those born within a single week in March 1946. Although originally intended to examine maternal health and the immediate effect on birth outcomes, follow-ups of this cohort of children have continued at regular intervals since, thereby providing an enormous wealth of data for the medical profession and social scientists to pore over. Indeed, very early on, data from the study was used to inform the NHS as it was set up in 1948. It also dramatically highlighted the consequences of inequality, with still births – for example – far higher in mothers from low-income families. The cost of inequality on health and outcomes is of course a story that is still very much with us today, a point each of the successive cohort studies has continued to ram home.

The aims of the next British cohort study were rather different. 12 years on and again all the babies born in a  single week in 1958 were studied, but this time specifically with a view to following their long-term development. Known as the National Child Development Study (NCDS) it also continues to this day as a longitudinal study. I feel a very personal connection with this study, not because I was one of the children under study but because as a gap year student I briefly worked on the project as an allegedly ‘statistical assistant’, although my knowledge of statistics up to that point was precisely nil: statistics did not form part of either the pure or applied maths A levels I did. I was not specifically involved in the exploration of the link between maternal smoking and low birthweight babies, but I most certainly was involved with the low birthweight part of the cohort. My diary of the time is frustratingly short on the detail of what I did, but it mainly involved reading off data from the earliest form of computer print-outs to find the babies, and enter the information on index cards. My memory says most of the time was spent trying to locate these low birthweight babies (and the triumph when I tracked one down that had gone ‘missing’), to sort out those who had died at birth or within the first month, and any link with what was then described as ‘educationally sub-normal’ or ‘severely sub-normal’.

This was in 1971, when language was different and my knowledge of social science as a discipline was zero. I learned a lot from my couple of months there (and return visits over the next two summers), by which time follow up studies at 7 and 11 had been completed. Computing was just blossoming and one of my senior colleagues spent most of the time worrying about how to compress the data onto manageable magnetic tapes, an edifying experience to watch. I learned how to use a ‘card sorter’ to sort my punchcards; I even learned some very basic statistical tests, at a level probably familiar to GCSE students these days. I watched a mechanical calculator go up in flames, and I got exposed to ideas of social science which probably colour me to this day. (Occasionally, when the day job and research have been going badly, I have thought I ought to go off and satisfy my social conscience by becoming a social worker.)

I was lucky to have such an educational job for a period in my gap year. So, for full disclosure, I should state I got this job entirely through ‘contacts’. The junior viola player in the school orchestra, who seemed to look up to me a lot as the leader in our group of 3 violists, told her mother that I was looking for a job and she was one of the medical doctors involved with the project. A phone call later – not even, as I recall, an interview – and there I was sitting happily in the National Children’s Bureau where the research was based. I thought nothing of that at the time, except to be grateful, but things would be unlikely to work that way now.

So, the 1958 cohort – nearly 60 and the study is still going strong – is close to my heart. Subsequent cohorts were initiated in 1970 and 2000, although others never got off the ground; hence the 30 year gap when researchers had hoped to stick with the 12 year periodicity. The former  is focussing on development issues again, hence with a medical bias. The Millennium cohort is more social science driven, aiming at a broad-ranging set of data examining the diversity of backgrounds into which the cohort are born and how that influences what happens thereafter. Child development still features, but so do the details of the family background and life highlighting, as each previous study has, how social background confers advantages or disadvantages which essentially don’t subsequently go away. Inequality is of course still rampant and basically not getting any better, something those who set up the original 1946 study might be bitterly disappointed to learn. Everything demonstrates that being born into a family without much will, on average, impact on health and wealth for the rest of life.

I recommend this book as an eye-opener to British society over the past 60 years. The research encompassed by the four studies has laid bare many ideas about our health and development and highlighted linkages that directly impact on our everyday lives. It isn’t always cheerful reading, but it is illuminating about how the best laid research plans can go astray and how a sudden unanticipated shower of money may require pulling all-nighters to write the research case; it reminds us that political whims may impact, or even shut down, your pet project; and illustrates – as if we needed it today – that evidence is never necessarily sufficient to change minds or policy.

This entry was posted in Research, Science Culture and tagged , , , . Bookmark the permalink.