It has taken half a century, but insights into mortality from a survey on health and nutrition are finally bearing fruit, writes Barbara Millar

Back in 1951, Isabella Leitch said 'it would be of great interest to trace accurately the further history of well and ill-grown people in terms of living and dying and causes of death'.

Ms Leitch was a nutritionist involved in the Carnegie Survey of health and nutrition in Britain, conducted on the eve of the Second World War. She was also a far-sighted woman. Fifty years on, her idea has borne fruit.

Scientists from Bristol University's department of social medicine have traced more than 4,500 children first studied by scientists back in the 1930s. And they have started to find associations between nutritional status in childhood and both cardiovascular disease and cancer mortality in adulthood.

The original survey was carried out between 1937 and 1939 after the Carnegie Trust gave a grant to the Aberdeen-based Rowett Research Institute to look at the state of health and nutrition in Britain.

The grant was awarded following the publication by Sir John Boyd Orr, director of the institute, of his report Food, Health and Income , which claimed that one-third of the British population was too poor to buy enough food to maintain health.

The survey was carried out at 16 centres in England and Scotland, including Bethnal Green in London, Liverpool, Barrow-in-Furness, Yorkshire, Dundee, Edinburgh and Aberdeen, which were chosen to give a mix of rural and urban areas.

Households were chosen to include large and small families; children ranged in age from 0-16 years; families' incomes ranged from 10 shillings per head per week to 45 shillings per head per week and there were families in which the man was unemployed.

In total, more than 1,300 families took part, with more than 7,900 individuals providing dietary information and more than 3,000 children receiving a medical examination. Another 1,200 children whose diet was looked at were not physically examined.

All the food available for each family for a week was weighed and recorded. Every purchase of food during the survey week was also recorded, as were meals taken outside the home (usually school meals). An inventory of all the food in the home was made at the end of the week.

Dietary information was subsequently analysed to produce per capita consumption of calories, fat, protein, carbohydrate, iron, calcium, phosphorus and vitamins A, B1 and C.

The physical examination of the children, carried out by two doctors, included their height and weight, an assessment of physical deformities including pigeon chest, knock knees, rickets and flat feet and a search for signs of chronic disease such as bronchitis and TB. Teeth and hearing were also examined.

'The children we examined in 1937-39 are now in their late 60s, 70s and 80s, ' says Professor John Pemberton, one of the two doctors involved, now retired and living in Sheffield. 'The survey was very unusual for its time.'

A quarter of the families underwent a repeat dietary assessment over the course of the study, to look at whether the seasonal availability of some foods affected overall diet.

At five centres interventions were made - at three of them food supplements were given to children at school, and in the other two the families were given extra cheese, butter, milk and fruit, Professor Pemberton recalls.

For controls, suitably matched families were selected and serial measures of growth and health were made in children in the intervention and control groups.

The outbreak of the Second World War meant that analysis of the survey results was delayed until 1955, although 'interim conclusions' were made available to the government and influenced wartime food policy.

Then the survey data was stored away somewhere in the basement of the Rowett Institute until 1989, when Stephen Frankel, head of the department of social medicine at Bristol University, and his colleague George Davey Smith tracked it down.

'There was a great interest in looking at people's early lives and how this influenced later disease risk, ' says David Gunnell, senior lecturer in epidemiology and public health at Bristol University.

'This data offered the opportunity to look at the impact of nutritional status in childhood on the incidence of diseases in adulthood.'

Tracking down those involved in the survey has not been easy, but over the past 10 years the Bristol researchers have found 3,500 of the original 4,973 children. Almost 1,000 have died in the intervening years and 500 could not be traced.

Researchers used the NHS Central Register in Southport and its equivalent in Edinburgh to trace the children.

The original analysis showed that the poorest children had their growth stunted by poverty. Revisiting this data has enabled Dr Gunnell and his colleagues to identify 'a consistent link' between the youngsters who were comparatively short and a risk of dying of coronary heart disease in later life. Taller children, particularly those with long legs, appear to suffer a significantly greater risk of cancer in later life.

To take the work further, a postal questionnaire has now been sent out to ascertain current health and diet among surviving survey recruits, and there are plans to carry out detailed clinical examinations on a sample of them.

'This should reveal new facts about the influences of childhood development and subsequent impact, ' says Dr Gunnell. 'It will also allow us to make practical recommendations on diets that will not only make the children healthy but also the later adults as well.'