Ponderal index is a poor predictor of in utero growth retardation

Objective To evaluate the usefulness of ponderal index (PI) and related indices of weight and length in identifying asymmetric growth, body thinness and organ asymmetry associated with IUGR.

Design Cross sectional study.

Setting Aberdeen Maternity Hospital.

Population The population includes term (greater than or equal to37 weeks) singleton live births (n = 53,934) between 1986 and 1996, ultrasound measurements in 2522 pregnancies, 712 unselected term pregnancies in 1979/1980 and stillbirths (24-36 weeks) between 1986 and 1995 where the fetus was diagnosed as suffering from acute (n = 73) or. chronic (n = 30) anoxic death.

Methods The strength of association between direct measures of IUGR and various indices of weight and length was determined by linear and multiple stepwise linear regression.

Main outcome measures Weight, length, PI and skinfold thicknesses (triceps, biceps, flank thighs, back) were measured at birth. Abdominal circumference, biparietal diameter and femur length were measured by ultrasound at greater than or equal to37 weeks. Ratio of liver, heart and kidney to brain were measured in stillbirths.

Results Weight alone was a better predictor of skinfold thickness, abdominal circumference and the ratio of abdominal circumference to biparietal diameter than weight divided by length raised to the power 1, 2, 3 (PI), 4 or 5. The inclusion of gestational age made little difference to the predictive ability of weight for these full term births. Weight, but not PI, was significantly different between the two groups of stillborn fetuses (chronic and acute), which had significantly different (P < 0.001) organ ratios.

Conclusion Body weight alone was a better predictor of anthropometric ratios, organ asymmetry and measures of thinness at birth thought to be associated with IUGR than the PI. The inclusion of a length term generally reduced the predictive ability with the highest powers resulting in the poorest prediction.