Article Text


Epidemiology and policy
P1-285 Use of linear spline models to describe bi-ethnic differences in early childhood growth: findings from the born in Bradford birth cohort study
  1. E Petherick1,2,
  2. L Fairley2,
  3. L Howe3,
  4. D Lawlor3,
  5. K Tilling4,
  6. J Wright2
  1. 1Paediatric Epidemiology Group, University of Leeds, Leeds, UK
  2. 2Bradford Institute for Health Research, Bradford, UK
  3. 3MRC Centre for Causal Analyses in Translational Epidemiology, Bristol University, Bristol, UK
  4. 4School of Social and Community Medicine, University of Bristol, Bristol, UK


Introduction Size at birth differs between South Asian and White European infants, but little is known about differences in growth in later infancy. We use multilevel linear spline models to describe ethnic differences in growth of infants from the Born in Bradford study (BIB).

Methods Data for 1378 singleton births from BIB with follow-up visits at 6, 12 and 18 months were used (614 White and 764 Pakistani infants). Weight and length data from clinic visits and routine measurements were used. Mulitlevel linear spline models with knot points at 4 and 10 months were fitted separately for weight and length including interactions with ethnicity and sex.

Results Models for weight and length with knot points at 4 and 10 months fitted the data well; the differences between actual and predicted measurements were small in each period. There were ethnic differences in weight and length at birth; Pakistani boys were on average 0.23 kg lighter (95% CI −0.31 to −0.14) and 1.05 cm shorter (−1.48 to −0.61) than White boys, while Pakistani girls were 0.16 kg lighter (−0.24 to −0.08) than White girls with no significant difference in length. The gains in weight in each time period were similar for both ethnicities. Pakistani boys and girls gained length faster than their White peers between 0 and 4 months.

Conclusions Differences in weight and length by ethnicity arise largely through differences at birth and growth in the early months of life. Further work will investigate relationships between exposures during pregnancy and differences in early life growth trajectories.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.