Growth charts for babies

Are the growth charts that we currently use inaccurate? Recent press reports about new growth charts from the World Health Organization imply that they are, particularly for breast fed babies. These charts are an exciting development, but are our current charts really as inadequate as the press would have us believe?

The first widely used growth reference in the United Kingdom was produced nearly 40 years ago,¹ followed by the chart from the National Center for Health Statistics in the United States,² which has been used ever since as WHO's international standard. These established the value of plotting measurements on growth charts in order to properly assess growth and nutritional status. However, problems with their accuracy were recognised 20 years ago; the growth of breast fed and formula fed infants, when plotted on either chart, rose steadily in the first few weeks and then fell by around one centile space (2/3 standard deviation).³ This is possibly because most of the children on whom the charts were based were born in the 1950s and fed fairly crude breast milk substitutes. Secular trends to increasing height and gender discrepancies also rendered the childhood charts less valid.⁴,⁵ Because of these limitations, the new United Kingdom 1990 charts and the charts from the US Centers for Disease Control were developed.⁴,⁶ These are based on larger, more recent datasets, constructed by using statistical rather than visual smoothing, and seem to fit infant growth patterns far better.⁷,⁸ In the United Kingdom, the new charts were rapidly adopted in primary care. Some paediatricians were reluctant to leave their more familiar charts behind, but a consensus group of the Royal College of Paediatrics and Child Health has now recommended that only the new United Kingdom 1990 charts should be used in infancy.⁹

Charts to date have simply described growth at any one time in the reference population. This is the simplest approach methodologically, but it raises difficult issues. In affluent populations, increasing rates of obesity mean that the proportion of children above the upper centiles is rising, but there is understandable resistance to updating charts to reflect this. In the developing world, in contrast, chronic and intergenerational undernutrition means that average growth is suboptimal, making the construction of local reference charts difficult if not unethical.

The alternative to a reference chart is a growth standard based only on the growth of healthy children in optimal conditions. This describes how children should grow, rather than how they actually grow, and is what WHO set out to produce.¹⁰ They collected data in six centres worldwide (Brazil, Ghana, India, United States, Norway, Oman), with those from poorer countries represented by subjects from affluent communities, and recruited only non-smoking mothers willing to breast feed exclusively for four months. Because of these stringent criteria, the study plan expected that as few as 20% might be eligible and willing to participate. This raises concerns that the infants studied may be different in other ways from the rest of the population. However, the resulting data (not yet in the public domain) are said to show great similarities in growth across all six study centres, which is in keeping with earlier studies,³,⁷ indicating that differences in height between rich and poor populations reflect environmental far more than genetic variation.

Attention has been focused particularly on the differences in growth between breast fed and formula fed infants. Slightly slower growth has been consistently seen in exclusively breast fed, compared with formula fed, infants in observational studies, which has led to the suggestion that breast fed infants should have separate charts.¹¹ However, a strong argument exists that such differences are not actually caused by the differing feeding mode, since faster growth was seen in trial populations exposed to breast feeding promotion,¹² with the likelier explanation being reverse causation: relatively large babies feed more, are more demanding, and are thus less likely to remain exclusively breast fed.¹² Rather than picking out breast fed infants as exceptional, all charts should be based on breast fed infants since they are the biological norm, but such data do need to be unbiased. For the first time, WHO could supply such data, provided that most infants were retained in the study and breast feeding rates were well maintained, assisted by active lactation support programmes provided in all study sites.

Should the United Kingdom move over to the new WHO charts when they are released? We already have much improved charts that reflect the growth of breast fed and formula fed infants far better than the old WHO charts, and persuading colleagues to renounce the earlier inaccurate charts has already proved hard enough. Will it be worth the effort to make a change from one much improved chart to another? The key questions, which can be answered only once the peer review process has been completed, will be how much the rigour of the WHO method has succeeded in producing a true blueprint for optimum growth and how well the new charts actually fit to the growth of infants in the United Kingdom.

If they pass the test, for the first time infants worldwide could be compared with the same growth standard, with breast fed infants rightly established as the norm with which all other infants should be compared.