To evaluate parameters of fetal breathing movements—displacement of the fetal abdominal wall during inspiration and expiration, time of inspiration and expiration and speed of inspiration and expiration—between 30 and 36 weeks' gestation in normal pregnancies, and in those complicated by gestational diabetes or maternal hypertension.
Three categories of pregnancy were investigated: 49 were normal, 16 had pregnancy-induced diabetes and 10 were hypertensive. According to their gestational age, the patients were divided into two groups: Group A between 30 and 32 weeks' gestation and Group B between 33 and 36 weeks. Using photogrammetry and a computer-operated algorithm, six parameters of fetal breathing movements were investigated.
There were significant differences in the various fetal parameters measured among the three categories of pregnant women. Up until 32 weeks of gestation, the displacements during inspiration and expiration were larger, the speeds of inspiration and expiration were higher, and the times for inspiration and expiration were shorter in the diabetic and hypertensive groups than in the normal group. In the later period, between 33 and 36 weeks, fetuses of pregnancy-induced diabetic patients showed the lowest inspiration and expiration times and the highest speeds of inspiration and expiration.
The mechanisms controlling lung development and function during human fetal life are complex and poorly understood. To date, most studies on lung development have been carried out on animal models, the majority in sheep1–3, and a few in monkeys4 and in pigs5. It has been known for some time that factors such as the intrathoracic space, the production/excretion of lung fluid, normal innervation and the amniotic fluid volume influence lung development6–8. Furthermore, unlike the heart and brain, which develop and begin to function early in fetal life, the lungs start their process of differentiation and work during the second half of pregnancy.
Human fetal breathing movements were first observed independently by Ahlfeld9 and Weber10 over 120 years ago. However, it was not until the work of Boddy and Mantell11 in 1972 and Dawes12 in 1974 that breathing movements were accepted as part of normal human fetal lung development. Since then, fetal lung development and fetal breathing, investigated using various techniques including traditional ultrasound and real-time Doppler ultrasound imaging, have become a source of increasing interest as newly identified markers seem to discriminate normal from abnormal lung development before birth13–17.
This study investigated variations in the fetal breathing pattern between 30 and 36 weeks of gestation, assessed by photogrammetry, of fetuses from normal pregnancies, from mothers with pregnancy-induced well controlled diabetes, and from hypertensive mothers.
Patients and Methods
This work was carried out at San Cecilio University Hospital, Granada, and was approved by the university ethical committee. The inclusion criteria were: having a known gestational age (as determined by last menstrual period and ultrasound examination) and being free from medication that may affect fetal breathing movements according to a previously published protocol18. All the patients were aware of the nature of the test and all signed a consent form. Seventy-five cases fulfilled the inclusion criteria and were selected from 144 patients. The selected cases were placed into two groups based on their gestational age: Group A included pregnant women between 30 and 32 weeks' gestation and Group B included pregnant women between 33 and 36 weeks' gestation. The women were also separated into three categories: 49 women were undergoing a normal pregnancy (25 in Group A and 24 in Group B). In this category there were 26 nulliparous and 23 parous women and the mean age was 29.3 years. All these subjects had normal systemic arterial pressure, and fetal growth was within the normal centiles. A negative O'Sullivan test (carried out at between 24 and 28 weeks' gestation in all the subjects) ruled out pregnancy-induced diabetes.
There were 16 women with pregnancy-induced diabetes included in the study (eight in Group A and eight in Group B). Eleven of these were in their first pregnancy and five in their second pregnancy. The mean age of the women in this category was 30.7 years. One case showing risk indicators of diabetes gave a positive O'Sullivan test in the first trimester. The remaining 15 patients gave a positive O'Sullivan test at the end of the second trimester (24–28 weeks' gestation). A glucose overload test confirmed the diagnosis in all cases. In four cases diabetes was diet-controlled, but the remaining 12 cases needed additional insulin therapy. In all cases the pregnancy-induced diabetes was ‘well controlled’ until the time of the fetal breathing test.
There were 10 hypertensive pregnant patients (four in Group A and six in Group B), whose mean age was 31.4 years. Nine women were in their first pregnancy and one woman was in her second pregnancy. In nine cases arterial hypertension was pregnancy induced and the remaining woman had arterial hypertension diagnosed before the start of her pregnancy. All patients were treated with methyldopa. At least 2 days prior to and at the time of the fetal breathing test, in all patients, the systolic arterial pressure was between 140 and 160 mmHg and/or the diastolic component was between 90 and 110 mmHg.
In all the subjects ultrasound examination was carried out using a Philips–4000 ultrasound machine (Philips Medical Systems, Solingen, Germany) at approximately the same time in the morning, 2 h after breakfast, with the woman in the semi-Fowler position. Examinations lasted for around 20 min. Images of the fetal spine, the sternum and the diaphragm in the sagittal plane, were obtained and used as reference points for measurements.
The methodology of digital image processing, gathering of real coordinates and graphic representation of fetal breathing movements has been published previously19. Briefly, the fetal body was imaged in a sagittal medial plane, taking care to obtain a clear view of the spine posteriorly and the sternum anteriorly. In this way, a good view of the umbilical cord attached to the abdominal wall was obtained. Images were taken 0.12 s apart (8.5 frames/s) and digitized. Images were video-recorded using a DVD recorder (LG Super-Multi) and stored on a digital video-tape for later analysis19.
Using the mouse cursor, a reference point in each image was defined along the abdominal wall, midway between the xiphoid process and the point of entry of the umbilical vessels into the abdomen. Occasionally, the keyboard was used to correct the position of the reference point. A complete respiratory cycle was regarded as a succession of positions in which the distance of the sternum and anterior wall of the chest and abdomen of the fetus increased (inspiration) then decreased (expiration) from the reference point. The software used selected the coordinates corresponding to each defined point on the screen and then performed an algorithm to determine the presence or absence of a respiratory movement. If the algorithm recognized fetal respiration, it measured all components of the breathing cycle process.
Breathing parameters that were measured included displacement of the fetal abdominal wall during inspiration (mm), displacement during expiration (mm), inspiration time (s), expiration time (s), speed of inspiration (mm/s) and speed of expiration (mm/s). Differences between variables were tested with the Mann–Whitney U-test. A P-value of < 0.05 was considered statistically significant.
Group A: gestational age between 30 and 32 weeks
Figures 1a and 1b show graphically the displacement of the fetal abdominal wall (in mm) during inspiration and expiration, respectively, according to pregnancy type. The fetuses of women with gestational diabetes and those with hypertension showed significantly larger displacement during both inspiration and expiration than those in normal pregnancies. Diabetic pregnancies showed larger values of fetal abdominal wall displacement during both inspiration and expiration than hypertensive pregnancies. The fetuses of women with gestational diabetes and those with hypertension showed shorter inspiration and expiration times than did those in normal pregnancies (Figures 1c and 1d). The speeds of inspiration and of expiration were significantly higher in the fetuses of women with gestational diabetes and those with hypertension than they were in normal pregnancies (Figures 1e and 1f).
Group B: Gestational age between 33 and 36 weeks
The fetuses of all three pregnancy types showed similar values for the displacement of the fetal abdominal wall during inspiration (Figure 2a). However, in the category of hypertensive women the displacement during expiration was significantly smaller than that in normal pregnancies (Figure 2b). The fetuses of women with gestational diabetes showed markedly shorter inspiration and expiration times than the other two categories, the difference being statistically significant when tested against normal pregnancies (Figures 2c and 2d). The speeds of both inspiration and expiration were significantly higher in the fetuses of women with gestational diabetes than in those of normal pregnancies (Figures 2e and 2f).
Photogrammetry has been adapted as a precision measurement tool and it is widely used when studying movements in sports20. When assessing fetal well-being, photogrammetry can be used to measure movements of relatively small amplitude. More importantly, this new technical approach is capable of measuring fetal breathing movements using a simple and nearly error free method19. Firstly, the operator does not have to follow fetal movements, a potential source of error. All that needs to be done, using a mouse cursor, is to determine one point on a screen. Secondly, it is the computer software that performs the measurements and calculations of the variables of the respiratory cycle. Thus, other potential sources of error such as inter- and intraobserver variations are reduced to a negligible minimum.
There are other anatomo-physiological considerations that need to be taken into consideration when assessing fetal breathing movements, for example, it has been shown that progressive hypoxia and hypercapnia alter the ventilatory responsiveness in low-birth-weight lambs13. In humans, it has been shown that the fetal breathing pattern shows changes that are linked to gestational age21, 22. In one of these studies22 it was shown that there is a clear difference in the fetal breathing pattern between 30 and 32 weeks and 33 and 36 weeks' gestation in normal pregnancies. Therefore, the design of our research protocol divided the cases into two groups: Group A, which corresponds approximately to the late cannalicular/pre-alveolar stage, and Group B, which corresponds to the pre-alveolar/alveolar stage of lung development.
Previous Doppler ultrasound-based studies have investigated fetal breathing movements indirectly, by extrapolation of parameters23, 24. In our study, however, we investigated fetal breathing movements through direct measurements of the musculoskeletal components of fetal breathing and used an algorithm to determine the parameters of each breathing movement. Based on this approach, our work shows significant differences in the breathing pattern of fetuses in women with pregnancy-induced diabetes, even when thought to be ‘well controlled’, compared to those in normal pregnancies.
It has been shown that non-enzymatic glycation, a reaction between carbohydrates and the free amino groups of proteins leading to the formation of advanced glycation end-products, occurs commonly in diabetes mellitus25. If glycation takes place in ‘well-controlled’ pregnancy-induced diabetic cases, then it could explain, in part, changes in the echogenic properties of fetal tissues, such as increased fetal lung density26. If glycation does occur in ‘well-controlled’ gestational diabetes then we would also expect it to have an effect on fetal physiological and neurological function. Our present work demonstrates just that, i.e. that in fetuses in pregnancy-induced, ‘well-controlled’, diabetic women there is an abnormal breathing pattern. There are a number of other considerations on this subject that will not be addressed here as they are beyond the scope of this paper. However, it should be emphasized that diabetes, however mild or well controlled during pregnancy, may affect the fetal respiratory system in form and in function.
The differences found in the fetal breathing parameters in hypertensive patients before 32 weeks' gestation compared to those found in normal pregnancies seem to disappear as pregnancy progresses towards term (Group B). These differences may be due, either fully or in part, to the treatment of hypertension by methyldopa. However, to date there is no documented evidence that methyldopa specifically affects fetal breathing movements18. These findings were unexpected and merit further investigation.
The combined use of photogrammetry and a computer-operated algorithm can accurately measure fetal breathing movements. Using this approach, abnormal fetal breathing patterns have been identified in pregnancy-induced diabetes and in pregnancies affected by arterial hypertension.
This work was supported by the Fondo de Investigaciones Sanitarias, Spanish Ministry of Health, project: PI-04/1737.