Recently a child in the UK was referred for assessment of microcephaly. The referring doctor was surprised to receive a letter back from the specialist stating that the child did not have microcephaly. The UK, like other countries, has now adopted the World Health Organization (WHO) Growth Charts. Although the measures for length and weight are quite close to pre-existing UK charts, there has been a very substantial change in the head circumference centiles. These are relatively similar at birth but by the age of 1 year the 2nd centile is approximately 2cm smaller than before. Putting it another way, at that age a head circumference that was on the 2nd centile on the old charts would now be between the 25th and 50th centile on the new charts, or a child following the new 2nd centile would be demonstrating a marked acquired, or postnatal, microcephaly on the old charts. These differences become even more marked by the age of 2 years.1,2 There has been a less marked change at the top end of the chart, but a child who was on the 98th centile would now be above the 99.6th and therefore have macrocephaly.

Comparisons with previous versions of the French and German charts show similar findings (B Echenne, personal communication 2010; F Heinen, personal communication 2010). In the USA a recently published community study of 75 412 young children found poor correlations with both the WHO and the 2000 Centers for Disease Control and Prevention charts, but reasonable correlation with an older chart from 1977. Eighteen per cent of children exceeded the WHO 95th centile at 2 years.3 France and the USA are also officially adopting the WHO charts.

The new growth charts are the result of a major initiative, the Multicentre Growth Reference Study, which was undertaken between 1997 and 2003 in 8440 children from Brazil, Ghana, India, Norway, Oman, and the USA. The WHO states that the data were designed to describe ‘how children should grow’ in optimal circumstances. Therefore, the children involved were selected, in that their mothers had followed healthy practices such as breastfeeding and avoiding smoking, and they were raised in environments that minimized poor diet and infection, as these are well recognized constraints on growth.1 Much of the discussion on the website concerns growth in terms of length, weight, and feeding practices. Surprisingly little is devoted to head growth. This is particularly odd given that length and height centiles have if anything shifted slightly upwards. Why not the head as well?

The lack of comment cannot help but give the impression that head, and therefore brain, growth is of less interest than growth in length and weight. However, the body itself prioritizes brain growth, since if a child becomes increasingly calorie-deprived weight velocity falls, then height velocity, and head growth is preserved until last. This is most commonly observed in the developed world in the ‘relative head sparing’ of newborns with intrauterine growth retardation. Furthermore, head size correlates with function. Using Nelhaus charts, children with an IQ above 70 and head circumference below −2SD (i.e. below the 2.3rd centile) had a lower academic achievement than those above −2SD.4 Using data from 28 820 children in the National Collaborative Perinatal Project, which had values slightly lower than the Nelhaus charts, a head circumference smaller than −2SD at 7 years had a fourfold increased association with an IQ below 70, and below −3SD the risk was increased 20 fold.5 Transposing this to the WHO charts, 50% of US children with a head circumference below the 2nd centile could have an IQ below 70.

Clinical practice will be affected in several countries. With the WHO charts fewer children will be referred for assessment of microcephaly, and more for macrocephaly. The WHO centiles only apply to children below the age of 5 years. After that age there will be a choice between dramatically shifting several centiles back to the old charts, or not being assessed against a norm. Since head size is partly familial, one essential clinical assessment of a child with a head size above the 98th or below the 2nd centile is to measure their parents’ heads – but what are the new adult norms? Finally, microcephaly can be a useful diagnostic clue in a child with other neurological concerns, with many possible pathologies.6 For example, acquired microcephaly in a young female could raise the possibility of Rett syndrome. It is well known that many young females with Rett syndrome have head sizes in the normal range, but with the WHO charts most will, so paediatricians will need to be alert for decelerating head growth within the normal range instead.7 However, in the US study deceleration was found in 12% of children below 1 year.3

In advising carers of the associations between the head circumference centile, function such as IQ, and possible pathologies, paediatricians will need to check from which chart the reference data were derived. This is a curious omission from the recent practice parameter on microcephaly, which cites data from the studies above but recommends use of the WHO charts.6 Some clinicians may prefer to use locally derived head growth charts. Alternatively, in many countries children older than 1 year with a head circumference below the 25th centile on the WHO charts may need to be regarded as an at risk group, but not most of those with head sizes above the 95th.


  1. Top of page
  2. References