- Top of page
Accurate estimation of fetal weight has an important role in routine antenatal care and in the detection of fetal growth abnormalities1, 2. For that reason, much effort has been invested in generating models that can accurately predict fetal weight. These regression models are based on various combinations of sonographically measured fetal biometric parameters, mainly abdominal circumference (AC), femur diaphysis length (FL), biparietal diameter (BPD) and head circumference (HC).
It has been previously shown that significant differences in intrauterine growth exist between male and female fetuses3, 4. These sex-specific growth patterns may result in differences in the correlation between the various biometric parameters and actual birth weight in male fetuses compared with female fetuses, and thus to differences in the accuracy of a given sonographic model in the prediction of birth weight for male fetuses compared with female fetuses. This information regarding sex-specific accuracy of weight estimation may be of great importance, especially in cases in which fetal growth abnormalities are suspected, and can provide support to the development of sex-specific sonographic models for fetal weight estimation.
Nevertheless, none of the widely accepted sonographic models for fetal weight estimation5–10 considers fetal sex in the equation. Furthermore, there are only limited data on whether the use of such sex-independent sonographic models results in different degrees of accuracy of weight estimation in male fetuses compared with female fetuses11.
Therefore, the aim of the present study was to test our hypothesis that the accuracy of sonographic fetal weight estimation may be related to fetal sex.
- Top of page
Our study has several key findings. First, for most models, the systematic error of the estimation of fetal weight was significantly lower for male fetuses than for female fetuses. Second, this finding was independent of the effects of potential confounders (including birth weight, cephalic index, parity, gestational age, maternal age, presentation, and the time interval between sonographic weight estimation and delivery). Third, it appears that the biometric parameters which contribute most to these sex-related differences are FL and AC, while models that include HC are associated with the lowest differences in the systematic error between male and female fetuses. Finally, the random error (a measure of precision rather than of accuracy) and correlation between estimated weight and birth weight are not affected by fetal sex.
There are significant differences in the intrauterine growth pattern between male and female fetuses3, 4. In addition, male and female fetuses have been shown to differ in body composition and in percentage of body fat24, and in the ratios between the various biometric parameters25. These differences may cause a given sonographic model to be associated with different degrees of accuracy for male and female fetuses. Nevertheless, this assumption has been tested in only a small number of studies. Heer et al.11, in a study of 820 singleton pregnancies, found no differences in the accuracy of the Hadlock equation (AC-FL-BPD; Model 6 in the current study) between male and female fetuses. However, their study included fetal weight estimations performed within the 14 days prior to delivery, a considerable period of time during which significant fetal growth may occur. This limitation probably accounts for the large systematic error in their study (9.4% for male fetuses and 9.0% for female fetuses, compared with 1.5% and 3.9%, respectively, in the current study (Table 3), which may mask any smaller sex-related differences in the systematic error. In contrast, in the current study, by using a large cohort of unselected women who underwent sonographic evaluation in a single tertiary center within 3 days prior to delivery, we were able to confirm our hypothesis and to demonstrate that the systematic error is consistently higher among female fetuses compared with male fetuses, even when different sonographic models are employed.
The reason for the persistently higher systematic error among female fetuses is unclear. One possible explanation is a systematic difference in the accuracy of the sonographic measurement of the different biometric parameters between male and female fetuses. Indeed, in a previous study in which we evaluated the accuracy of the sonographic measurement of HC, we found that the sonographic measurement of HC in female fetuses was significantly more similar to the actual postpartum HC measurement than for male fetuses26. A second explanation is that the regression models evaluated in the current study are associated with a more optimal fit to the actual birth weight of male fetuses compared with female fetuses. This preferential fit towards male fetuses may be explained as follows. First, there was a larger proportion of male fetuses compared with female fetuses in the cohort used to generate these regression models. Unfortunately, in the original papers of Hadlock et al.6, 7 and Warsof et al.8, no information is provided regarding the sex of the fetuses from which these models were derived. Second, there was a more significant influence of the male subgroup (compared with the female subgroup) in the cohorts from which the regression models were derived on the regression equation. Lastly, there were differences between the populations in the cohorts of our study and those from which the regression models were generated. For example, if the mean birth weight was lower in our population, then the regression models will fit better with the male fetuses in our cohort which, because of the higher birth weight in male infants compared with female infants, are more similar to the original cohort from which the regression models were derived.
We have found that the biometric parameters that contributed most to the sex-related differences in the accuracy of sonographic weight estimate were AC and FL. The reason for these findings is not clear but it may be a reflection of the selective differences in intrauterine growth patterns, between male and female fetuses, of the anatomic structures on which these parameters are based3, 25. Models that incorporate HC have been previously shown to be more accurate than models which are based on BPD, independently of fetal sex27. Considering the differences in the cephalic index between male and female fetuses in the current study, it is not surprising that sex-related differences in the accuracy of weight estimation were lower for models that incorporated HC (a measure that is independent of cephalic index) compared with models that are based on BPD.
In conclusion, there appear to be significant differences in the accuracy of sonographic weight estimation between male and female fetuses, and these differences are independent of other variables that are also related to fetal gender, such as birth weight and cephalic index. This information may provide support to the development of sex-specific sonographic models for fetal weight estimation, and such sex-specific models may be of importance, especially in cases in which fetal growth abnormality is suspected, as the integration of fetal sex may decrease the rate of false-negative or false-positive prediction of fetal growth abnormality. Indeed, Schild et al.28–30 have reported that the use of such a sex-specific model is more accurate than other widely accepted models. One issue that should be emphasized is that although the sex-related differences in the systematic error were statistically significant (as a result of the large sample size), the absolute magnitude of these differences was relatively small and the clinical significance of these differences is uncertain. Nevertheless, the development of sex-specific models may be one step towards the development of customized models that include corrections for additional factors which affect the accuracy of weight estimation and that may be associated with an even greater improvement in the accuracy of weight estimation. More studies are needed in order to provide a better understanding of the reasons for these sex-related differences and to determine whether the use of sex-specific models may improve the accuracy of sonographic weight estimation in male and female fetuses.