The height of pregnant women, ultrasound parameters of bone mineral density status and pregnancy complications

Bone mineral density (BMD) reflects the mineral content and the structural and functional status of bone tissue. During pregnancy lactation involves changes in bone metabolism to facilitate increasing demand for calcium by the fetus requirements, changes in calcium-phosphorus metabolism, and hormone-induced maternal adaptations [1, 2]. According to modern observational data, changes in bone mineral density can be reversed in most women. However, in patients with baseline risk factors, decreased bone strength may have clinical significance [2–4].

The issue of association between BMD values and obstetric complications remains disputable. A large cohort study did not confirm a direct link between changes in BMD and the majority of adverse obstetric outcomes in healthy women [1]. At the same time, some studies indicate the importance of vitamin and mineral status, weight gain, and the musculoskeletal status during pregnancy and in the postpartum period [4–7].

Anthropometric characteristics of women can also influence the course of pregnancy and childbirth. Modern studies convincingly show an increased risk of some obstetric and perinatal complications in short height women, primarily due to a potential mismatch between the size of the fetus and the maternal pelvis [8]. However, there is insufficient data on the role of maternal height and ultrasonographic characteristics of bone tissue in the course of pregnancy.

Ultrasound bone densitometry (UBD) is a non-ionizing imaging modality for evaluation of bone tissue status and can be potentially used in pregnant women as a radiation-free screening tool [9, 10]. At the same time, the available data are insufficient to justify its use as a universal screening tool for all pregnant women. Further clarification is needed regarding the clinical significance of the detected changes and their association with pregnancy outcomes.

Currently, the data on the role of maternal height and ultrasound characteristics of bone tissue in the course of pregnancy are insufficient. This determined the objective of our study.

The objective of the study was to assess the possible association between the height of pregnant women, ultrasound parameters of bone tissue status, and the development of somatic diseases that led to complicated pregnancy, as well as obstetric complications.

Materials and methods

This observational and exploratory study was conducted from 2021 to 2025 and included 104 women with singleton term pregnancies who were hospitalized for delivery at two healthcare facilities. Quantitative US was performed in the third trimester, 3–7 days prior to childbirth. Information on current pregnancy complications was obtained from medical documentation. Associative and exploratory data analysis was performed.

To verify the threshold validity for 167 cm height, we additionally analyzed the data from our own obstetric database. In the sample from 2011–2012 (n=226) the average height was 166.2±5,5 cm, the median was 166.5 [163.0; 170.0] cm; in the sample from 2015–2020 (n=614) – 166.8±6.0 and 167.0 [164.0; 170.0] cm, respectively. The direction of the shift corresponds to the hypothesis of a trend toward increased height. However, inter-period differences in the local sample did not reach statistical significance (p=0.217). Therefore, the height of167 cm was not considered by us as a universal threshold for a tall height for all women in the Republic of Belarus. This was a pre-set threshold for the research program that coincides with the median height of the contemporary local obstetric cohort, that enabled to assess the differences within the study population.

Based on the height, all patients were divided into 2 groups (A and B):

Group A (N=52) included women whose height was <167 cm. Group B (N=52) included women with a height of ≥167 cm. 

Each of the groups A and B were divided into 2 subgroups (А1 and А2, В1 and В2) based on the results of UBD. Subgroups A1 (n=34) and В1 (n=30) included pregnant women who had UBD scores of ≥-1.0. Subgroups А2 (n=18) and В2 (n=22) included women who had UBD scores of ≤-1.1

This study used measurement of forearm bone mineral density as the exploratory non-ionizing modality.

BMD scores of ≥-1.0 are shown is the Tables as 1, and BDM scores of ≤-1.1 are shown as 2.

Inclusion criteria in the study were singleton pregnancy, pre-gestation body mass index 18.5–24.9 kg/m², weight gain during pregnancy no more than 14 kg.

Exclusion criteria were gestational age < 37weeks or ≥ 42 weeks, multiple pregnancy.

Limitations of the study. The data on a number of potential factors influencing BMD (25-OH vitamin D level, smoking, calcium and vitamin D intake) were not systematically registered.

Сode O60.0 in analysis denotes the registered episode of preterm labor without preterm delivery. The women who had preterm delivery were not included in the main cohort.

Anthropometric stratification was included in the initial design of the author’s doctoral dissertation topic “Pregnancy and childbirth in the context of acceleration”. The threshold of 167 cm was selected as an optimal threshold for the research program and was additionally verified using the author's own obstetric dataset, where the median height was 167.0 cm.

The study was conducted in accordance with the Declaration of Helsinki developed by World Medical Association and the Healthcare Law of the Republic of Belarus. Extract from the Protocol No. 11 of June 20.06.2025 of the Biomedical Ethics Committee, Belarusian State Medical University of the Ministry of Healthcare of the Republic of Belarus. All patients participating in the study signed informed consent.

The diagnoses were established according to the clinical protocol “Medical observation and delivery of medical care to women in obstetrics and gynecology” (approved by Resolution No. 17 of 19.02.2018 of the Ministry of Healthcare of the Republic of Belarus).

The study was conducted in the third trimester of pregnancy, 3–7 days prior to childbirth. The structural and functional status of bone tissue was evaluated using ultrasound densitometer “Sunlight Omnisense 9000” (Israel). Model S was used for adult patients.

T-score is the difference between measurement of the speed of sound (SOS) in the patient and the averaged value of SOS for young population in the units of standard deviation.

Z-score is the difference between measurement of SOS in the patient and the mean SOS in the population of the same age or gender in the units of standard deviation.

SOS also gives information about the absolute value of SOS (m/s), which is used by the device to calculate T-scores and Z-scores based on the reference database.

Statistical analysis

Statistical processing of the results was performed using the Python programming language, Version 3.9, and Pandas, SciPy and Statsmodels libraries for data analysis.

The normality of distribution of the quantitative data (T-scores, the height) was tested using the Shapiro-Wilk test. Given that distribution of most characteristics differed from normal and the presence of binary variables (presence or absence of the disease or complication), the relationships were tested using non-parametric methods.

The frequencies of qualitative (binary) characteristics were compared between the four independent subgroups (A1, A2, B1, B2) using 4×2 contingency tables. Given the low number of certain outcomes and the presence of zero cells, Fisher–Freeman–Halton exact test was used for the primary analysis. The calculated value was denoted as the overall p-value.

Pairwise comparisons between the subgroups were performed only for the parameters with the overall p-value <0.05 obtained in overall comparison of the four subgroups. Fisher’s exact test with the Bonferroni correction for six comparisons was used for 2×2 pairwise comparisons. The differences were considered statistically significant at p<0.05. Given that the study was exploratory, the results were considered as exploratory and hypothesis-generating.

Results and discussion

The mean age of patients in group A was 29.10±5.90 years, and in group B – 29.79±5.00 years. No statistically significant differences were found between the groups (p=0.379), that indicates comparability of the groups by age.

The studied groups A and B were comparable by clinical and anamnestic characteristics, somatic, obstetric and gynecologic pathology (Table 1). Comparative characteristics of the diseases were performed using ICD-10.

The results of analysis of somatic diseases that complicated pregnancy course, and obstetric pathology in the subgroups А1, А2, В1 and В2 are represented in Table 2.

Analysis of baseline clinical and anamnestic characteristics found no statistically significant differences between the 4 subgroups (Table 1). Analysis of somatic diseases that complicated pregnancy, and obstetric pathology (Table 2) found that the overall p<0.05 was obtained for acute vaginitis during pregnancy (N76), the episode of preterm labor without delivery (O60.0), and maternal care for excessive fetal growth (O36.6).

Following pairwise comparisons with the Bonferroni correction, statistically robust difference between subgroups A2 и B2 (p__adj =0.028) persisted only for O60.0. Other differences should be considered as exploratory signals requiring verification in a larger sample, since a small number of events increases the risk of Type II error and restricts interpretation of negative results.

Our findings are consistent with literature data indicating that changes in bone tissue status during pregnancy are often adaptive and reversible, and are not always associated with adverse obstetric outcomes [1–4]. At the same time, the absence of statistically significant differences should not be interpreted as the evidence of the lack of clinical association between the explored factors and pregnancy outcomes. Given the small sample size and rare outcomes, statistical power of the study is low and can fail to detect clinically significant effects of moderate magnitude.

The most overt exploratory signal was obtained for the episode of preterm labor without preterm delivery (O60.0), where the difference between subgroups A2 and B2 persisted after Bonferroni correction. This result requires cautious interpretation, since the number of events was small, and the outcome reflected the registered episode of preterm labor rather than preterm birth. Nevertheless, it can be used for planning subsequent study with a larger sample size and preliminary statistical power calculation.

The selected threshold of 167 cm should be considered not as a universal clinical criterion for a tall height, but as an optimal cut-off value determined by the research program and corresponding to the median of the local obstetric cohort. This approach enables within-study comparisons in the study population, but limits direct extrapolation of the results to other populations.

Analysis of our own obstetric database showed anthropometric shift toward changes in the height: the mean height increased by 0.6 cm, and the median by 0.5 cm between the periods from 2011 to 2012 and from 2015 to 2020. This data confirm the validity of using 167 cm as an optimal threshold within this research program, but do not allow to consider it as a universal criterion for a tall height for all pregnant women in the Republic of Belarus.

Conclusion

In the pilot sample of 104 pregnant women, no statistically robust dependence was found between the majority of the explored somatic diseases, obstetric complications, and the height of women and ultrasound parameters. The overall comparison of the four subgroups showed the differences in the overall p-value<0.05 for acute vaginitis during pregnancy (N76), the episode of preterm labor without preterm delivery (O60.0), and maternal care for excessive fetal growth (O36.6). The difference between subgroups A2 and B2 persisted only for O60.0 after the Bonferroni correction.

The obtained findings are exploratory and do not make it possible to justify universal ultrasound screening for all pregnant women. The limited sample size, small number of events, and lack of systematic registration of certain factors influencing bone tissue status do not allow to exclude clinically significant associations.