Customised birthweight centiles are useful for identifying small-for-gestational-age babies in women with type 2 diabetes

Background: Customised birthweight centiles identify small-for-gestational-age (SGA) babies at increased risk of morbidity more accurately than population centiles, but they have not been validated in obese populations.
Aims: To compare the rates of SGA by population and customised birthweight centiles in babies of women with type 2 diabetes and examine perinatal outcomes in customised SGA infants.
Methods: Data were from a previous retrospective cohort study detailing pregnancy outcomes in 212 women with type 2 diabetes. Customised and population birthweight centiles were calculated; pregnancy details and neonatal outcomes were compared between groups that delivered infants who were SGA (birthweight < 10th customised centile) and appropriate weight for gestational age (AGA) (birthweight 10–90th customised centile).
Results: Fifteen (7%) babies were SGA by population centiles and 32 (15%) by customised centiles. Two babies of Indian women were reclassified from SGA to AGA by customised centiles. Nineteen babies were reclassified from AGA to SGA by customised centiles; of these, 15 (79%) were born to Polynesian women, five (26%) were born less than 32 weeks and two (11%) were stillborn. Customised SGA infants, compared with AGA infants, were more likely to be born preterm (19 (59%) vs 20 (16%), P  < 0.001) and more likely to be stillborn (4 (13%) vs 0 P  = 0.001). After excluding still births, admission to the neonatal unit was also more common (19 of 28 (68%) vs 43 of 127 (34%), P  < 0.001).
Conclusions: In our population more babies were classified as SGA by customised compared with population centiles. These customised SGA babies have high rates of morbidity.