Risk factors for small-for-gestational-age infants by customised birthweight centiles: data from an international prospective cohort study.

OBJECTIVE To identify clinical and ultrasound variables associated with the birth of small-for-gestational-age (SGA) infants by customised centiles, subclassified according to whether their mothers were normotensive or developed hypertensive complications. DESIGN Prospective, multicentre cohort study. SETTING Participating centres of the Screening for Pregnancy Endpoints (SCOPE) study in Auckland, New Zealand, Adelaide, Australia, Manchester and London, UK, and Cork, Ireland. POPULATION The 3513 nulliparous participants of the SCOPE study. METHODS Women were interviewed at 15 ± 1 weeks, and had ultrasound growth measurements and umbilical and uterine Doppler studies at 20 ± 1 weeks. Variables associated with SGA infants were identified using logistic regression. MAIN OUTCOME MEASURES Small for gestational age (i.e. a birthweight of less than the tenth customised centile), normotensive-SGA and hypertensive-SGA. Comparison groups for statistical analyses were non-SGA, normotensive non-SGA and hypertensive non-SGA. RESULTS Among 376 (10.7%) SGA infants, 281 (74.7%) were normotensive-SGA and 95 (25.3%) were hypertensive-SGA. Independent risk factors for normotensive-SGA were low maternal birthweight, low fruit intake pre-pregnancy, cigarette smoking, increasing maternal age, daily vigorous exercise, being a tertiary student, head and abdominal circumference of less than the tenth centile and increasing uterine artery Doppler indices at the 20-week scan. Protective factors were: high green leafy vegetable intake pre-pregnancy, and rhesus-negative blood group. Risk factors for hypertensive-SGA were conception by in vitro fertilisation, previous early pregnancy loss and femur length of less than tenth centile at the 20-week scan. CONCLUSIONS Risk factors for infants who are SGA by customised centiles have been identified in a cohort of healthy nulliparous women. A number of these factors are modifiable; however, further studies are needed to replicate these findings.