Intended for healthcare professionals

Papers

National trends in birth weight: implications for future adult disease

BMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6939.1270 (Published 14 May 1994) Cite this as: BMJ 1994;308:1270
  1. C Power
  1. Epidemiology and Biostatistics, Institute of Child Health, London WC1N 1EH.
  • Accepted 11 January 1994

Birth weight is a powerful predictor of infant survival and is associated with morbidity, childhood growth, and adult height. Recently it has been argued that intrauterine conditions, represented by birth weight, have long term effects on adult conditions such as ischaemic heart disease and non-insulin dependent diabetes.1 thus national trends in this health indicator are of great interest.

Sources of information and results

In Scotland birth weight is reported on Scottish Maternity Records (SMR2) and collated by the Scottish Health Service Common Services Agency. The agency provided information on birth weights of liveborn infants since 1975. For England and Wales, routinely collected data on live births were obtained from the Office of Population Censuses and Surveys since 1983, when records were considered to be complete.2

The table shows an increasing percentage of heavier birth weights; regression of birth weights >3500g against year indicates that heavier births increased by 0.40% a year (95% confidence interval 0.37% to 0.44%) in Scotland and by 0.35% (0.11% to 0.60%) in England and Wales. Over this period the proportion of births 2500-2999g and 3000-3499g steadily decreased and very low weight births (<1500g) increased, with the proportion of other low weight births (<2500g) remaining unchanged.

Trends in birth weight. Figures are percentages of live births unless stated otherwise

View this table:

Comment

In recent years the birth weight distribution has shifted noticeably towards heavier birth weights.3 Although generally related to infant and childhood outcomes, such changes may have wider implications if hypotheses linking intrauterine development with adult health1 are confirmed. Good quality national data on birth weight are desirable for future monitoring; hence, the rise in incomplete data for England and Wales since 1989 is regrettable.

Annual increases in birth weight in England, Wales, and Scotland have been small and gradual, as in other countries such as Norway, Sweden, and the United States.3 Even so, the impact of such changes may not be insignificant in relation to adult health. Risks associated with birth weight, as for other cardiovascular risk factors, seem to be continuous and graded: standardised mortality ratios for cardiovascular disease decrease from 119 for weights of <5.5 lbs (2500g) at birth to 74 for those >8.5lbs (3900g).4 Risk of disease is not confined to a high risk group. When risk is diffused throughout the population a large proportion of the population affected by a small risk may have a greater impact on population levels of disease than a smaller proportion at high risk. The distributional shift shown here for birth weight could therefore have substantial and favourable effects on future population levels of disease.

Although the general trend is towards heavier births, very low birth weights (<1500g) are also increasing. The trend for Scotland is similar to that for England and Wales.5 Possible explanations include an increasing proportion of triplets and higher order births, and changes in classification of non-registrable miscarriages to registered births.5 Very low weight births are important in relation to infant deaths, but as they constitute a small proportion of survivors (under 1%) this increase is unlikely to have a major impact on population levels of adult disease.

Recent birth weight trends should not engender complacency, especially as we can only speculate as to why changes have occurred. Research on intergenerational influences on birth weight suggests that current trends may be as much a product of grandmaternal as of current circumstances. Increases in height and reductions in smoking in women (from 49% in 1972 to 34% in 1990 among 25-34 year olds) may contribute to the recent birthweight trend, while increases in the proportion of first births would tend to depress it. Within Britain at least the association between birth weight and socioeconomic circumstances is well documented. Further improvements in birth weight depend, in particular, on the weights of babies from lower social class backgrounds increasing to resemble those in higher classes.

Dr Susan Cole and the Scottish Health Service Common Services Agency provided the Scottish data. Colleagues in the Department of Epidemiology, Institute of Child Health, gave constructive comments.

References