Intrauterine growth restriction and fetal body composition




To assess the differences in fetal body compartments between fetuses with normal growth and those with reduced intrauterine growth, during the third trimester, through ultrasonographic determination of subcutaneous tissue thickness (SCTT).


Twenty-eight patients were enrolled into this case control study carried out at 30–31 weeks' gestation. Two study groups were matched for maternal age and pregestational body mass index: controls (n = 14) and intrauterine growth-restricted (IUGR) fetuses (n = 14). Routine ultrasound-derived biometric parameters (head circumference, abdominal circumference, femur length and humerus length) were measured. Additionally, the mid-arm fat mass and lean mass (MAFM and MALM), the mid-thigh fat mass and lean mass (MTFM and MTLM), the abdominal fat mass (AFM) and the subscapular fat mass (SSFM) were measured. The Mann–Whitney U-test and Student's t-test were used to compare the two groups.


The abdominal circumference and the humerus were significantly smaller in IUGR fetuses than in controls. Most of the SCTT values were different in the two groups. The SSFM (3.6 ± 1.1 vs. 2.6 ± 0.7 mm; P = 0.011), the AFM (5.1 ± 0.7 vs. 4 ± 1 mm; P = 0.01), the MAFM (3.5 ± 0.9 vs. 2.2 ± 0.8 cm2; P < 0.01) and MALM (2.1 ± 0.4 vs. 1.7 ± 0.5 cm2; P = 0.029) were all significantly greater in fetuses with normal development compared to those with growth restriction.


During the third trimester, SCTT (with the exception of MTFM and MTLM) is reduced in fetuses with IUGR. Furthermore, MALM is lower in growth-restricted fetuses, confirming that the parameters measured in this study are affected in IUGR fetuses. Our findings indicate that specific changes in fetal body compartments occur as a result of chronic metabolic impairment. Copyright © 2005 ISUOG. Published by John Wiley & Sons, Ltd.


Fetal weight can be measured ultrasonographically using estimated fetal anthropometric measurements and population-based growth charts. Estimated fetal weight (EFW) is commonly used as an index of fetal growth and is generally calculated through a combination of parameters that include, amongst others, abdominal circumference. Nevertheless, both the EFW and abdominal circumference show a wide range of variation that could potentially impact on clinical practice1, 2. Error in EFW may be as high as 25%3 and results from technical measurement errors, as well as the assumptions that fetal density is constant throughout gestation and is independent of the fetal pathological processes that alter normal muscle/fat ratios4, 5.

Fat content correlates directly with energy stores. Fat mass and lean body mass are often used in the nutritional assessment of the individual. Fat constitutes 12–14% of birth weight and has been shown to account for 46% of the variation noted in neonatal weight6. Consequently, ultrasound-generated estimates of fetal fat may be useful in the evaluation of fetal growth abnormalities. Several authors have previously used ultrasonography to assess anthropometric measurements of fetal body composition4, 5. For instance, Bernstein et al5 compared fat and lean body mass measurements in healthy fetuses across gestation and demonstrated significant correlations with birth weight and estimates of neonatal lean and fat mass.

Recently (2004), Deter stated the importance of assessing fetal growth not on the basis of single anatomical variables, such as birth weight or abdominal circumference, but rather using the novel concepts of individualized growth assessment and the Prenatal Growth Profiles, in which growth is assessed with each fetus serving as its own control7. This is a more precise approach, since fetal growth potential is in part linked to demographic-, age- and women-specific variables7. Utilization of fetal soft tissue measurements was discussed in this study7 and in a previous study by the same author8.

Reference values for fetal subcutaneous tissue thickness (SCTT) parameters have been recently reported in a 300-subject longitudinal ultrasound study. Fetuses at 22 weeks' gestation were studied from mothers who had healthy pregnancies compared to women with gestational diabetes. When the growth profiles from the two populations were compared, a ‘fat-free mass index’ using the mid-thigh lean mass (MTLM) was shown to be an effective means (even at 22 weeks' gestation) of differentiating between normal fetuses and those with diabetic mothers. Other conventional ultrasound measurements have been shown to be limited in their ability to differentiate between normal and affected fetuses at 22 weeks' gestation9.

Moreover, Marconi et al10 have shown that in intrauterine growth-restricted (IUGR) fetuses there is a rapid breakdown of proteins, indicating that the ‘fat-free mass’ is affected in IUGR fetuses. Furthermore, these authors have shown that maternal leucine infusions result in a lower fetal–maternal leucine enrichment ratio when compared to normal fetuses. This finding is further verified in animal experiments in which fetal amino acid infusions are able to prevent fetal growth restriction11. Conversely, Galan et al4 have reported that the reduced birth weight of a subset of North American newborns was due to a reduction in fetal subcutaneous fat tissue and not lean mass. Despite this apparent contrast, this work highlights the potential value of the ultrasonographic evaluation of fetal skinfold thickness in the detection of differences between specific populations.

Ambiguity in these results has been the motivation for our investigations on various tissue compartments of the fetus. The main aim of the present study was to assess the differences in the fetal compartments between normal and growth-restricted fetuses during the third trimester, through the use of ultrasonographic measurements of various SCTT parameters.



The study was approved by the Institutional Review Boards at Tor Vergata University of Rome. Patients were enrolled from the authors' outpatient high-risk pregnancy clinic at 30–31 weeks' gestation during the period January–December 2002. Two study groups, the control and growth-restricted groups, were matched for age and pregestational body mass index (BMI). The control group consisted of primigravida women with a healthy singleton pregnancy, while the growth-restricted group consisted of primigravida women with a singleton pregnancy with the ultrasound-determined diagnosis of IUGR12, as having a fetal abdominal circumference < 5th centile for gestational age by local reference values, an estimated fetal weight < 10th centile for gestational age and an umbilical artery pulsatility index of more than 2 SDs above the gestational mean compared to local reference values. Additional inclusion criteria for fetuses in this study were specific gestational age, absence of fetal chromosomal abnormalities and normal fetal anatomy.

Gestational age was calculated from the first day of the last menstrual period and confirmed by either a first- or second-trimester ultrasound scan. When the ultrasound-determined gestational age differed from that calculated from the last menstrual period by > 7 days in the first trimester, or by > 10 days in the second trimester, the ultrasound-determined gestational age was used.


Patients underwent a conventional ultrasound scan with the commercially available unmodified Teknos Esaote Ultrasound Machine (ESAOTE S.p.a. Headquarters, Genova, Italy), at the Fatebenefratelli Hospital, Tor Vergata University in Rome, using a 3.5-MHz probe. Routinely obtained biometric parameters included the head and abdominal circumferences and femur and humerus lengths.

To obtain fat mass and lean mass, several measurements were assessed as previously described by Valensise et al13. We used a similar technique to that of Bernstein et al5 to measure the fat and lean body mass areas on axial ultrasound images in the mid-upper arm and mid-upper leg regions and the cross-sectional images of the abdomen and the subscapular field5, 14, 15. Briefly, the mid-arm fat mass (MAFM), mid-arm lean mass (MALM), mid-thigh fat mass (MTFM) and mid-thigh lean mass (MTLM) (all measurements in cm2) were obtained using a longitudinal view of the long bone and extremities at an angle of 0° to the transducer. The transducer was then rotated 90° at the middle of the long bone to obtain an axial view of the extremity. The fat mass (MAFM, MTFM) was measured by taking the total cross-sectional limb area and subtracting the central lean area (MALM, MTLM) that consisted of muscle and bone (Figure 1a, 1b).

Figure 1.

Axial views of the extremities showing (a) the mid-arm fat mass (MAFM) and mid-arm lean mass (MALM) and (b) the mid-thigh fat mass (MTFM) and mid-thigh lean mass (MTLM). Ultrasound scans illustrating how (c) abdominal fat mass (AFM) and (d) subscapular fat mass (SSFM) measurements are made.

The abdominal fat mass (AFM, in mm) was determined by measuring the thickness of the anterior abdominal subcutaneous tissue on the same axial image from which the abdominal circumference is obtained (Figure 1c) as previously described by Gardeil et al16.

The subscapular fat mass (SSFM, in mm) was evaluated (Figure 1d) by measuring the SCTT perpendicularly at the lower apex of the flat bone, in a sagittal section, as described by Valensise et al13.

The reproducibility and precision of SCTT measurement have been previously reported9. All measurements were carried out by the same investigator.

Statistical analysis

The study compared two groups, where the anticipated difference in means is 1 and the anticipated SD is 0.5. The sample size required for an α = 0.05 and a power of 95% was eight cases per group. The sample size required for an α = 0.01 and a power of 95% was 11 cases per group. Power analysis was performed according to Sokel and Rolf17.

We compared the conventional ultrasound parameters and the SCTT parameters of the two study groups using the non-parametric Mann–Whitney U-test since, to the best of our knowledge, it is more powerful than its parametric equivalent.


The characteristics of the study groups are summarized in Table 1. The gestational age at delivery was significantly different for the two groups: 39 ± 1 weeks' gestation for normal fetuses vs. 35 ± 3 weeks' gestation for growth-restricted fetuses (P < 0.01). The birth weights and their percentiles were lower in growth-restricted fetuses than in normal fetuses (P < 0.01) (Table 1).

Table 1. Characteristics of the study groups
CharacteristicNormal fetuses (n = 14)IUGR fetuses (n = 14)Mann–Whitney P-value
  1. Values represent mean ± SD. BMI, body mass index (weight/height2); GA, gestational age; IUGR, intrauterine growth restriction.

Age (years)28.4 ± 3.529.2 ± 1.30.330
Pregestational BMI26.8 ± 2.127.1 ± 2.30.642
GA at study time (weeks)31 ± 230 ± 20.513
GA at delivery (weeks)39 ± 135 ± 3< 0.01
Birth weight (g)3425 ± 3011797 ± 559< 0.01
Birth weight percentile63.8 ± 18.69.4 ± 0.7< 0.01

Nine of 14 (64%) patients with growth-restricted fetuses developed gestational hypertension, two (14.3%) had pre-eclampsia and three (21.4%) women had normal blood pressure values. Gestational hypertensive and pre-eclamptic patients were classified using the International Society for the Study of Hypertension in Pregnancy criteria18.

The abdominal circumference was significantly smaller in growth-restricted fetuses than in normal fetuses (P < 0.01) as was humerus length (P = 0.036). Head circumference and femur length were not significantly different between the two study groups (Table 2).

Table 2. Differences in the conventional ultrasound parameters in normal and growth-restricted fetuses
Conventional ultrasound parametersNormal fetuses (n = 14)IUGR fetuses (n = 14)Mann–Whitney P-value
  1. Values represent mean ± SD. IUGR, intrauterine growth restriction.

Femur (cm)5.8 ± 0.65.4 ± 0.40.055
Humerus (cm)5.1 ± 0.54.8 ± 0.30.036
Abdominal circumference (cm)26.8 ± 2.323.0 ± 2.6< 0.01
Head circumference (cm)28.1 ± 2.027.3 ± 1.80.175

Most of the SCTT values were significantly smaller in the growth-restricted fetuses at the study time (30–31 weeks' gestation) when compared to the normal fetuses. Table 3 summarizes these soft tissues parameters. The results show that the SSFM, AFM, MAFM and MALM were statistically lower in growth-restricted fetuses than in fetuses with normal intrauterine growth.

Table 3. Differences in the fetal body compartments in normal and growth-restricted fetuses
SCTT parametersNormal fetuses (n = 14)IUGR fetuses (n = 14)Mann–Whitney P-value
  1. Values represent mean ± SD. AFM, abdominal fat mass; IUGR, intrauterine growth restriction; MAFM, mid-arm fat mass; MALM, mid-arm lean mass; MTFM, mid-thigh fat mass; MTLM, mid-thigh lean mass; SCTT, subcutaneous tissue thickness; SSFM, subscapular fat mass.

AFM (mm)5.1 ± 0.74.0 ± 1.00.010
SSFM (mm)3.6 ± 1.12.6 ± 0.70.011
MAFM (cm2)3.5 ± 0.92.2 ± 0.8< 0.01
MALM (cm2)2.1 ± 0.41.7 ± 0.50.029
MTFM (cm2)5.1 ± 1.64.2 ± 1.00.278
MTLM (cm2)4.2 ± 1.13.7 ± 1.80.100


The importance of the evaluation of fetal SCTT has been the subject of an increasing number of studies in recent years. Bernstein et al5 have previously shown that ultrasound can be used to measure subcutaneous fetal fat in the extremities. The variations noted when using ultrasonographic estimates of subcutaneous fat in fetuses do compare relatively reasonably with the variations observed in the measurement of skinfold thickness in neonates19.

It is clear from clinical practice that the routine use of ultrasonographically measured parameters is valuable in the assessment of gestational age. These parameters are reasonably accurate in part due to their resistance to environmental influence. However, it is this quality that makes them less well suited for the identification of fetal growth abnormalities. When compared to all the routinely measured parameters, abdominal circumference is the most sensitive when detecting growth abnormalities. Bernstein et al reported that fetal fat and lean body mass have unique growth profiles and that, as a result of an accelerated rate of growth in late gestation, the measurement of fetal fat may provide a more sensitive and specific means of identifying abnormal fetal growth when compared with index values of lean body mass5.

The evaluation of fetal growth should be made on an individualized basis, derived from the concept that fetal growth is a complex process that can be adversely affected in various ways, in different individuals7. Therefore, individualized growth assessment provides a comprehensive and integrated evaluation of fetal growth. It corrects for differences in age and growth potential, the two primary confounding variables of growth assessment. This new method takes into consideration the concept that soft tissues undergo early changes in abnormal growth conditions such as IUGR or macrosomia7.

Several studies have taken into account the assessment of the SCTT in pregnant women suffering from gestational diabetes and treated by dietary means or through the use of insulin20. Bernstein and Catalano21 showed that increased neonatal fat (independent of birth weight) was associated with a significant increase in the risk of birth by Cesarean section of infants of women with gestational diabetes. From both the scientific literature and clinical practice, the use of ‘fat-index’ as a predictor of morbidity has been widely used in neonates. Additionally, Whitelaw22 has demonstrated that subcutaneous fat is a more accurate indicator of maternal glucose control than is birth weight in infants of diabetic mothers.

We have previously provided reference ranges for soft tissue measurements in fetuses of normal mothers and those with gestational diabetes9, and the present study aimed to identify whether similar changes in the fetal SCTT measurements were observed in IUGR fetuses9, 13. Walther et al have shown that a low ponderal index (Rohrer's index of corpulence) is more strongly associated with neonatal growth restriction than is birth weight, emphasizing further the importance of studying the fetal body composition in pathological conditions23. In a longitudinal ultrasound assessment of fetal subcutaneous measurements, Galan et al4 reported a reduction in fat mass in growth-restricted fetuses; however, lean body mass measurements were not changed. In the present study we observed that most of the SCTT measurements (excluding the MTFM and MTLM) are decreased in growth-restricted fetuses at 30 weeks' gestation. Therefore, we conclude that ‘fat mass’ is significantly reduced in fetuses that develop IUGR. Interestingly, the MALM measurement was also found to be decreased in growth-restricted fetuses suggesting that the ‘lean mass’ may also be affected in IUGR.

The differences noted in the present study compared to those of Galan et al4 could be in part attributable to the different gestational age at which the measurements were performed in the studies and could also be due to population differences including the fact that the mothers in the Galan et al study had lived at high altitude.

We speculate that in adverse environmental conditions (i.e. chronic hypoxia, maternal hypertension or other pathological states) there may be a reduction in fetal energy stores (notably the fatty tissue), which may explain the reduction in fat mass noted in the present study. It is possible that the reduction in muscle mass observed may be due to muscle breakdown in order to provide a further source of energy in this adverse situation. In other words, even muscles and lean tissue undergo damage with loss of substrates. This finding could be due to diversion of incoming energy sources to direct energy production instead of energy storage as usually occurs in the late third trimester.

Conventional ultrasound-measured parameters such as the abdominal circumference do provide insight as to whether a fetus is growth-restricted or not, and the abdominal circumference has been shown to decrease with progression of IUGR. However, the present findings suggest that use of the SCTT may provide insight into the specific changes that occur in the various fetal compartments of the growth-restricted fetus. Furthermore, we propose that future use of SCTT measurements may provide an opportunity to quantify and locate the effects of various novel treatments such as administration of maternal amino acid infusions11 on the growth-restricted fetus.