Pulsatile versus continuous growth in the human fetus

Abstract

DEBORAH PITTINARO, SHIRLEY EBERLY, University of Rochester, Department of Obstetrics and Gynecology, Rochester, New York, University of Rochester, Department of Biostatistics, Rochester, New York OBJECTIVE: To determine if fetal growth occurs in a pulsatile, rather than continuous, fashion. STUDY DESIGN: Nonsmoking women of normal body habitus, with medically uncomplicated, singleton pregnancies, were enrolled between 26 and 29 weeks gestation. Only data from pregnancies delivering normally grown fetuses at term, in women with normal weight gain, were included. Sonographic fetal biometric parameters were measured frequently, with a goal of 3 times per week, for a minimum of 4 weeks. For each fetus, six replicate measurements, obtained by a single examiner, were made on each of five parameters: biparietal diameter (BPD), head circumference, abdominal circumference, humerus length, and femur length. For each fetus and parameter, a 2nd degree polynomial regression on gestational age was generated. Under the assumption of pulsatile growth, a cyclical pattern of positive and negative residuals from a fitted curve would be expected. A runs test was used to test the randomness of the regression residuals. Under the assumption of continuous growth, serial growth increments would be expected to follow a Gaussian distribution. For each fetus and parameter, serial growth increments were tested for normality using the Shapiro-Wilk test. P values ! .05 were considered significant. RESULTS: 8 subjects meeting inclusion criteria, with a mean gestational age at entry of 28.6 weeks, were examined at an average interval of 2.3 days between observations. 10 to 21 observations, for an average of 15 observations per subject, were obtained. 1/8 BPDs demonstrated a nonrandom pattern in the residuals (P ! .05). Patterns supporting pulsatile growth were not observed for any other parameters in any fetus studied. CONCLUSION: These findings demonstrate minimal support for a pulsatile pattern of fetal growth. Increasing the number of observations per subject may improve the ability to detect a pulsatile pattern. OBJECTIVE: To describe the normal biometry of the cerebellar vermis by fetal magnetic resonance (MR) and its relationship to gestational age. STUDY DESIGN: Fetuses with normal intracranial anatomy underwent blinded retrospective measurement of total vermian length (TVL) and inferior vermian distance (IVD) in the midline sagittal plane. TVL was measured as the longest cranial-caudal dimension of the vermis. IVD was defined as the measurement between the most inferior vermian edge and the cerebral spinal fluid space of the 4th ventricle as an indicator of the angle between the vermis and midbrain structures. Images were obtained from sagittal MR slices with a thickness of 4 to 7 mm using a superfast MR protocol. Correlation of TVL and IVD to gestational age (GA) was established using linear regression and calculated confidence intervals. RESULTS: 53 MR studies were available with GA ranging from 13 to 38 weeks. TVL increased in a linear fashion throughout gestation, and this association was significant (P ! .001). IVD also increased with GA (P = .004) but measurements were at the threshold for MR resolution. IVD was always less than 5 mm independent of GA. CONCLUSION: Total vermian length measured in a sagittal plane on fetal MR increases linearly throughout gestation. An understanding of normal vermian development may be helpful in identifying pathologic conditions affecting the cerebellum and posterior fossa.