Anda di halaman 1dari 5

Foot Length in Fetuses with Abnormal Growth

Natalie B. Meirowitz, MD, Cande V. Ananth, PhD, MPH, John C. Smulian, MD, MPH, David A. McLean, MD, Edwin R. Guzman, MD, Anthony M. Vintzileos, MD

Sonographic fetal foot length is highly predictive of gestational age. In order to assess the reliability of this parameter in predicting gestational age in cases of abnormal fetal growth, we examined fetal foot length in small- and large-for-gestational-age fetuses. A nomogram of foot length versus gestational age between 15 and 37 weeks was constructed using cross-sectional data obtained from 5372 singleton fetuses. Fetal foot lengths for small-for-gestational-age fetuses (estimated fetal weight below the 10th percentile) and large-for-gestational-age fetuses (above the >90th percentile) fetuses were plotted against the foot length nomogram in order to determine the number of small-for-gestational-age fetuses and large-for-

gestational-age fetuses with foot lengths below the 10th and above the 90th percentiles, respectively. Of the 586 small-for-gestational-age fetuses, 355 (60.6%) had foot lengths below the 10th percentile on the nomogram. When foot lengths from large-forgestational-age fetuses were plotted on the foot length nomogram, 29.4% (219 of 744) had measurements above the 90th percentile. Fetal foot length can be influenced by growth restriction as well as states of accelerated fetal growth. Our findings imply that there are limitations to the use of fetal foot length for gestational age assessment, particularly in fetuses with growth abnormalities. KEY WORDS: Fetus, foot length; Foot, length, fetal; Growth abnormalities, fetal.

ABBREVIATIONS SGA, Small-for-gestational-age; EFW, Estimated fetal weight; LGA, Large-for-gestational-age; H/A, Head to abdominal circumference

Received August 17, 1999, from the Division of Maternal-Fetal Medicine (N.B.M., J.C.S., D.A.M., E.R.G., A.M.V.), Division of Epidemiology and Biostatistics (C.V.A.), and Center for Perinatal Health Initiatives (C.V.A., J.C.S.), Department of Obstetrics, Gynecology, and Reproductive Sciences, UMDNJRobert Wood Johnson Medical School/St. Peters University Hospital, New Brunswick, New Jersey. Revised manuscript accepted for publication November 2, 1999. Address correspondence to Natalie B. Meirowitz, MD, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, St. Peters University Hospital, 254 Easton Avenue, MOB 4th Floor, New Brunswick, NJ 08903-0591. Reprints not available. The Center for Perinatal Health Initiatives is supported, in part, by a grant (#029553) from the Robert Wood Johnson Foundation, NJ. Opinions, views, and conclusions expressed in this paper are those of the authors and not those of Robert Wood Johnson Foundation.

umerous fetal structures measured sonographically correlate well with gestational age. Fetal biometry is useful for screening and diagnosing growth disturbances as well as structural and genetic abnormalities. One measurement that has been suggested to be useful is fetal foot length, but growth of the fetal foot has not been examined thoroughly. Nearly eight decades ago, Streeter1 reported a strong relationship between fetal foot length and menstrual age in a series of fetal pathologic specimens. That study and others2 have validated the practice of using foot length for accurate gestational age assessment in the postmortem fetus. More recently, Mercer and coworkers3 demonstrated that sonographic fetal foot lengths were highly correlated with gestational age beyond the first trimester. Existing literature has concentrated on fetal foot size in the normally grown fetus,36 while little is known regarding the effect of fetal growth disturbances on foot length. Given the fact that

2000 by the American Institute of Ultrasound in Medicine J Ultrasound Med 19:201205, 2000 0278-4297/00/$3.50

202

FOOT LENGTH IN FETUSES

J Ultrasound Med 19:201205, 2000

pathologists currently use the fetal foot for accurate gestational age assessment, it is reasonable to investigate whether this practice also is valid in the antepartum period, when a fetal growth abnormality is suspected. Therefore, this study was designed to evaluate the effect of growth disturbances on sonographic fetal foot lengths in both the SGA and LGA fetuses. MATERIAL AND METHODS The study population was identified by reviewing a computerized database of sonographic foot length measurements recorded prospectively at our institution from October 1994 to April 1998. All sonographic examinations were performed at St. Peters University Hospital, New Brunswick, NJ, using either an ATL Ultramark HDI 9 or 3000 (Advanced Technology Laboratories, Bothell, WA) or an Acuson 128 (Acuson, Mountain View, CA). Institutional review board approval was obtained for performing this study. Fetal Foot Length Nomogram A nomogram of fetal foot length versus gestational age for singleton fetuses was constructed for the period between 15 and 37 weeks gestation using available measurements. Inclusion of a fetus in the nomogram required concordance within 2 weeks between menstrual and ultrasonographic dating at the initial sonographic examination. Gestational age was recorded as completed weeks (based on menstrual dating). The last foot length measurement obtained before delivery was chosen for analysis, and this contributed a single observation for each fetus. Fetuses with abnormalities involving the extremities, such as micromelia, talipes, polydactyly, and arthrogryposis, were excluded. Foot length was measured in centimeters in the plantar view from the heel to the first or second toe (whichever was longer) using on-screen calipers. The relationship between foot length and gestational age was determined. Identification of Fetuses with Growth Abnormalities SGA fetuses were identified by first generating a nomogram of EFW versus gestational age for our population. This EFW nomogram was constructed using 18,628 obstetrical ultrasonographic examinations from 10,741 singleton fetuses obtained in our unit during the same study period. EFWs (in grams) were determined from 15 to 42 weeks gestation

using head circumference, biparietal diameter, femur length, and abdominal circumference measurements.7 For all fetuses analyzed, sonographic biometry at the initial examination was within 2 weeks of menstrual gestational age. Since multiple fetal weight estimations were available for many fetuses, the nomogram was constructed after accounting for this nonindependence using the generalized estimating equations procedure.8 SGA fetuses were identified as those with EFWs below the 10th percentile, and LGA fetuses were those having EFWs above the 90th percentile on our population nomogram. In order to distinguish fetuses with pathologic growth abnormalities from those who may have been constitutionally small or large for gestational age, we chose to further categorize the growth pattern of SGA and LGA fetuses using H/A ratios. First a nomogram of H/A ratio versus gestational age from 12 to 42 weeks was constructed based on the available 18,628 ultrasonographic examinations. Then both SGA and LGA fetuses were classified as symmetrically or asymmetrically on the basis of H/A ratios. SGA fetuses with H/A ratios above the 90th percentile for gestational age were defined as symmetrically small, and those with H/A ratios at or below the 90th percentile for gestational age were defined as asymmetrically small. Asymmetrically large fetuses were defined as those with H/A ratios below the 10th percentile, and symmetrically large fetuses were those with H/A ratios at or above the 10th percentile for gestational age. Fetal foot lengths of SGA and LGA fetuses were plotted against the foot length nomogram for fetuses in our general population in order to determine the affect of growth disturbances on this sonographic measurement. Only the foot length measurement obtained closest to delivery for each SGA and LGA fetus was used. The number of symmetric and asymmetric SGA fetuses with foot lengths less than the 5th, 10th, and 50th percentiles and the number of symmetric and asymmetric LGA fetuses with foot lengths above the 50th, 90th, and 95th percentiles were determined. A subgroup analysis was performed using only SGA and LGA fetuses beyond 24 weeks gestation. This cutoff was chosen as the earliest gestational age at which growth abnormalities are likely to be detected and are clinically relevant. Statistical Analysis Multivariable linear regression models were established in order to derive the smoothed curves for the 5th, 10th, 50th, 90th, and 95th percentiles for all nomograms, e.g., foot length versus gestational age, EFW versus gestational age, and H/A ratio versus

J Ultrasound Med 19:201205, 2000

MEIROWITZ ET AL

203

gestational age. The smoothed percentiles were generated by fitting linear regression models based on the restricted cubic spline smoothing procedure.9 Prior to fitting the regression models, the dependent variables (foot length and H/A ratio) were checked for normality. In addition, the data were checked for the presence of extreme outliers (beyond four standard deviations from the mean), and coefficients of variation (expressed as a ratio of mean to the standard deviation) were computed. Chi-square test was used to compare the number of symmetric versus asymmetric SGA fetuses with foot lengths below the 5th, 10th, and 50th percentiles and the number of symmetric versus asymmetric LGA fetuses with foot lengths above the 50th, 90th, and 95th percentiles. A P value of < 0.05 was considered statistically significant. All statistical analyses were performed using the SAS system (SAS Institute, Cary, NC). RESULTS Our patient population was 85% Caucasian (other than Hispanic) and 10% African American, with the remaining patients being of Hispanic origin and other minorities. Approximately 45% of the population were nulliparous women and 31% were clinic patients.

Fetal Foot Length Nomogram Of the 10,741 singleton fetuses who underwent sonographic examinations during the study period, 5372 fetuses had foot length measurements between 15 and 37 weeks gestation. The median number (range) of foot length measurements available at each gestational week was 108 (15 to 1160). The relationship between foot length and gestational age (based on menstrual dating) was best described by a linear regression model (R2 = 0.93, P < 0.0001; Fig. 1) with gestational age modeled using cubic splines with five knots (located at 17, 19, 20, 22, and 30 weeks). Table 1 presents the predicted values for the 5th, 10th, 50th, 90th, and 95th percentiles of foot length, the number of fetuses examined, and the coefficient of variation in foot length measurements at each gestational week between 15 and 37 weeks. SGA Fetuses Overall, 586 (10.9%) of 5372 fetuses with foot length measurements had an EFW below the 10th percentile for gestational age and therefore were classified as SGA. The median gestational age (range) for this group was 21 (15 to 36) weeks. One hundred and twenty-nine (22.0%) of these 586 fetuses were classified as asymmetrically small, and 457 (78.0%) were

Figure 1 Fetal foot length percentiles (5th, 10th, 50th, 90th, and 95th) by gestational age between 15 and 37 weeks.

204

FOOT LENGTH IN FETUSES

J Ultrasound Med 19:201205, 2000

classified as symmetrically small. When foot lengths of all SGA fetuses were plotted against the foot length nomogram, 60.6% (355 of 586) fell below the 10th percentile (Table 2). The percentages of asymmetric and symmetric SGA fetuses with foot lengths below the 10th percentile (55.0% versus 62.1%; P = 0.28) were similar. As shown in Table 2, a subgroup analysis of SGA fetuses who were beyond 24 weeks gestation (median gestational age [range] of 27 [24 to 37] weeks) yielded similar results. Fifty-seven of 99 fetuses (57.6%) had measurements below the 10th percentile. LGA Fetuses Seven hundred and forty-four of 5372 (13.8%) fetuses with foot length measurements were classified as LGA based on an EFW above the 90th percentile for gestational age. The median gestational age (range) for this group was 19 (15 to 36) weeks. Of the 744 LGA fetuses, 25.3% (n = 188) were classified as asymmetric and 74.7% (n = 556) were classified as symmetric. When foot lengths of LGA fetuses were plotted against the foot length nomogram, 29.4% (219 of 744) fell above the 90th percentile (Table 3). The percentages of asymmetric and symmetric LGA fetuses with foot lengths above the 90th percentile were not significantly different regardless of whether the analysis was restricted to LGA fetuses of greater than 24 weeks gestation (median gestational age [range] of 26 [24 to 36] weeks). However, symmetric LGA fetuses were significantly more likely to have foot lengths above the 50th percentile in the analysis of both all LGA fetuses and those above 24 weeks gestation (Table 3). DISCUSSION Fetal foot length nomograms were originally established using pathologic specimens, using measurements obtained from both fresh and fixed specimens.1 Postmortem measurements of the fetus may be affected by the method of tissue fixation and the time interval between delivery and fixation.1 Formalin fixed tissues will swell to a variable degree, depending on the age of the specimen, and will gradually shrink to their original size after several months.1 Specimens that have not been fixed rapidly may be subject to drying artifacts. Nomograms established using postmortem specimens also may include a disproportionate number of abnormal (i.e., anomalous, growth restricted) fetuses. Despite these concerns, Mercer and coworkers3 demonstrated that prenatal ultrasonographic measurements of the fetal foot were in close agreement with Streeters 1920 postmortem data.

Table 1: Fetal Foot Length Percentiles by Gestational Age*


Gestational Age (weeks) 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Fetal Foot Length Percentiles (Smoothed) CV (%) 5th 10th 50th 90th 95th 12.7 10.4 9.7 9.8 8.9 9.3 8.5 8.9 8.1 7.0 7.1 7.0 6.3 5.4 6.2 5.2 5.7 5.3 4.4 6.8 6.2 5.5 5.3 1.4 1.6 1.9 2.2 2.5 2.8 3.1 3.4 3.7 4.0 4.3 4.6 4.8 5.1 5.3 5.6 5.8 6.0 6.3 6.5 6.8 7.0 7.3 1.5 1.7 2.0 2.3 2.6 2.9 3.2 3.5 3.8 4.1 4.4 4.7 4.9 5.2 5.4 5.7 5.9 6.1 6.4 6.6 6.9 7.1 7.4 1.8 2.1 2.4 2.7 3.0 3.2 3.5 3.9 4.2 4.5 4.8 5.1 5.3 5.6 5.8 6.1 6.3 6.5 6.8 7.0 7.3 7.5 7.7 2.2 2.5 2.8 3.1 3.3 3.6 3.9 4.2 4.6 4.9 5.1 5.4 5.7 5.9 6.2 6.4 6.7 6.9 7.1 7.4 7.6 7.9 8.1 2.3 2.6 2.9 3.2 3.4 3.7 4.0 4.3 4.7 5.0 5.2 5.5 5.8 6.0 6.3 6.5 6.8 7.0 7.2 7.5 7.7 8.0 8.2

N 18 146 375 613 1160 929 552 360 222 177 125 123 108 74 66 65 62 65 39 37 24 15 17

*Values for percentiles are in centimeters. N, Number of fetuses; CV, coefficient of variation.

Using both pathologic and sonographic data, previous investigators16 have suggested that fetal foot length is a reliable predictor of gestational age. However, the results of this investigation illustrate that fetal foot length can be influenced by growth restriction as well as states of accelerated fetal growth. Our findings imply that there are limitations to the use of fetal foot length for gestational age assessment, particularly in fetuses with growth abnormalities. The effects on fetal foot length that are observed with variations in fetal growth are likely to be attributable to the fact that measurements of the fetal foot incorporate both soft tissue and bone. Soft tissue stores of subcutaneous fat are reduced in cases of growth restriction and increased in situations of accelerated growth. It therefore makes sense that fetal foot length is relatively smaller in the SGA fetus and larger in the LGA fetus. Since the foot is not entirely soft tissue, the effects of abnormal fetal growth on foot length are probably attenuated by the relatively unaffected foot bones. Thus, over 95% of SGA fetuses have foot lengths below the 50th percentile but only 60.6% are below the 10th percentile. Similarly, LGA fetuses are very likely (greater than

J Ultrasound Med 19:201205, 2000

MEIROWITZ ET AL

205

Table 2: Fetal Foot Length Percentiles for SGA Fetuses


FFL Percentile All SGA Asymmetric SGA Symmetric SGA P Value*

Table 3: Fetal Foot Length Percentiles for LGA Fetuses


FFL Percentile All LGA Asymmetric LGA Symmetric LGA P Value*

All Fetuses Number 586 <5th 210 (35.8%) <10th 355 (60.6%) <50th 559 (95.4%)

129 43 (33.3%) 71 (55.0%) 122 (94.6%)

457 167 (36.5%) 284 (62.1%) 437 (95.6%)

0.62 0.28 0.98

All Fetuses Number 744 >95th 128 (17.2%) >90th 219 (29.4%) >50th 680 (91.4%)

188 21 (11.2%) 39 (20.7%) 159 (84.6%)

556 107 (19.2%) 180 (32.4%) 521 (93.7%)

0.31 0.37 0.003

Fetuses >24 Weeks Gestation Number 99 33 <5th 32 (32.3%) 11 (33.3%) <10th 57 (57.6%) 17 (51.5%) <50th 93 (93.9%) 31 (94.0%)

66 21 (31.8%) 40 (60.6%) 62 (94.0%)

0.88 0.39 1.00

Fetuses >24 Weeks Gestation Number 99 33 >95th 21 (21.2%) 4 (12.1%) >90th 33 (33.3%) 9 (27.3%) >50th 88 (88.9%) 25 (75.8%)

66 17 (25.8%) 24 (36.4%) 63 (95.5%)

0.12 0.37 0.003

*P value refers to comparisons between asymmetric and symmetric categories. FFL, Fetal foot length.

*P value refers to comparisons between asymmetric and symmetric categories. FFL, Fetal foot length.

90%) to have foot lengths above the 50th percentile, but only 29.4% are above the 90th percentile. The methods we have used to provide normative data on sonographic fetal foot length have several advantages over previously published reports. Earlier studies3,4,6 employed polynomial terms for the variables (foot length and gestational age) in their models. Although this approach would depict nonlinear relationships between the variables, it imposes curvatures that may not be valid. The restricted cubic spline model used to generate our foot length nomogram allows flexible modeling by not restricting the relationship between the independent and dependent variables. Our nomogram was constructed using a patient population several orders of magnitude larger than those used by previous studies,36 and our population appears to have a broader ethnic composition. Although our values for the 50th percentile are almost identical to those of Mercer and colleagues,3 our outer percentiles are different, perhaps more reflective of population norms. Previously published studies have not presented information regarding coefficient of variation for ultrasonographic measurements of fetal foot length.36 Our data demonstrated larger coefficients of variation at early gestational ages compared with later gestational ages. This was an unexpected finding, because foot length is easier to measure during the second trimester than at later gestational ages. The most plausible explanation for this finding is that in early gestation foot length is so small that minor variations in measurements (for technical reasons) will result in a large coefficient of variation. The relatively consistent findings for fetuses analyzed beyond 24 weeks of gestation indicates that fetal foot length may be affected by abnormal growth even prior to the third trimester. Although we

observed that a statistically greater number of symmetric LGA fetuses were above the 50th percentile than asymmetric LGA fetuses, the magnitude of the difference suggests that fetal foot length would not be helpful for categorizing the type of LGA growth. Finally, our findings suggest that the practice of using fetal foot length to assess gestational age in the antepartum period should be tempered, especially in the presence of suspected fetal growth abnormalities. REFERENCES
1. Streeter GL: Weight, sitting height, head size, foot length, and menstrual age of the human embryo. Contrib Embryol Carnegie Inst 11:143, 1920 Huxley AK: Comparability of gestational age values derived from diaphyseal length and foot length from known forensic foetal remains. Med Sci Law 38:42, 1998 Mercer BM, Sklar S, Shariatmadar A, et al: Fetal foot length as a predictor of gestational age. Am J Obstet Gynecol 156:350, 1987 Goldstein I, Reece EA, Hobbins JC: Sonographic appearance of the fetal heel ossification centers and foot length measurements provide independent markers for gestational age estimation. Am J Obstet Gynecol 159:923, 1988 Hata T, Senoh D, Hata K, et al: Mathematical modeling of fetal foot growth: Use of the Rossavik growth model. Am J Perinatol 13:155, 1996 Platt LD, Medearis AL, DeVore GR, et al: Fetal foot length: The relationship to menstrual age and fetal measurements in the second trimester. Obstet Gynecol 71:526, 1988 Hadlock FP, Harrist RB, Carpenter RJ, et al: Sonographic estimation of fetal weight. Radiology 150:535, 1984 Liang K-Y, Zeger SL: Longitudinal data analysis using generalized linear models. Biometrika 73: 13, 1986 Durrelman S, Simon R: Flexible regression models with cubic splines. Stat Med 8:551, 1989

2.

3.

4.

5.

6.

7. 8. 9.