Preterm birth risk in women with skeletal dysplasias and short stature

1 Introduction

Skeletal dysplasias refer to a heterogeneous group of over 400 disorders that affect an individual’s bone development [1]. One of the most prominent features of SD is short stature, defined as a height less than 148 centimeters at maturity [2]. Additional key clinical features include bony deformities and recurrent fractures. The determination of a specific type of skeletal dysplasia can be made with a combination of clinical exam, radiologic evaluation, as well as biochemical and molecular testing. The overall estimated incidence of skeletal dysplasias is 30-45 per 100,000 live births [3]. Achondroplasia is the most common non-lethal skeletal dysplasia, and it is associated with a severe short stature as well as other morbidities [4]. Inherited in an autosomal dominant pattern, the genetic basis of achondroplasia is a mutation in the fibroblast growth factor receptor type 3 gene resulting in abnormal cartilaginous bone growth [3, 4]. Most other skeletal dysplasias are also the result of a genetic mutation.

Many women with skeletal dysplasias have a near normal life expectancy and enter menarche and pubarche at a similar age to the general population [4, 5]. Thus, many women with skeletal dysplasias enter reproductive years with the desire to create a family. To date, the literature on pregnancy outcomes of women with short stature is limited, consisting of case reports and very few larger contemporary studies [5–7].

Many clinical questions still remain regarding obstetrical outcomes in the population of women with short stature in the current era. Specifically, the actual risk of preterm birth has not been quantified in the literature to date. The primary objective of this study was to estimate the risk of preterm delivery in a population of women with short stature. The secondary outcome aimed to identify factors that increase the risk of preterm delivery in a population of women with short stature, including maternal height, worsening respiratory function during pregnancy, worsening pain during pregnancy, preeclampsia, preterm pre-labour rupture of membranes (PPROM), delivery of a small for gestational age (SGA) neonate defined as birthweight less than the tenth percentile for gestational age, and having a neonate with a diagnosis of short stature.

2 Methods

Research ethics board approval (142-2017) was obtained through the local teaching hospital site, Sunnybrook Health Sciences Centre in Toronto, Canada.

This study is based on a cross sectional survey asking detailed pregnancy and reproductive health questions aimed at women who were little people. The overarching study consisted of demographics as well as antenatal, delivery and postpartum experiences of little people, with an option to participate in a qualitative interview. This survey was developed with the assistance of Little People Canada, who provided feedback on the study in the design and survey question development stage, as well as assisted with distribution. This was to ensure the survey used appropriate language and addressed questions important to the little person community. Multiple rounds of testing were performed following programing of the survey to ensure cohesive logic and formatting by the authors. The survey was administered between July 2017 and July 2018 to a convenience sample through Little People of America, Little People United Kingdom, Little People Canada, and the World Dwarf Games. These groups were selected because they represent little people and provide health and social information for members of the little people community. Each organization had access to the link and was asked to distribute it to their constituents as they felt was appropriate. This could be in the form of a banner on a website or through social media, email lists, or bulletins. Three reminders were sent to encourage distribution. This survey was voluntary (open) to people who frequented the abovementioned groups. A standardized letter was included for potential subjects to read prior to participation. In this letter, subjects were informed that the survey was anonymous, there were no incentives to participate, and completion of the survey was voluntary. It also indicated that the survey had 30 questions and would take approximately 20 minutes to complete. Subjects were informed that they could choose to withdraw from the survey at any time by closing their web browser. Consent was implied by completion of the online survey.

The items were not randomized, and adaptive questions were used. The majority of questions had a “no answer” option.

Current pregnancy data in this rare population is very limited; therefore, a baseline population sample was not accessible in order to ensure that subjects in the study reflected those of the general pregnant population of individuals with skeletal disorders. However, by distributing the survey to multiple organizations from several different countries, a representative sample from the general population of short statured individuals was likely obtained.

The survey was administered and data initially stored through an online Canadian survey system that was compliant with the Personal Health Information Protection Act and Personal Information Protection and Electronic Documents Act. All data was stored in Canada with password protected user authentication.

The system utilized bi-hourly back-up systems, firewall restrictions, and ongoing logging auditing to capture failed login attempts. IP addresses from individual respondents were not obtained to confirm individual responses. Downloaded data was stored on a password protected institutional computer in a locked office. Data was anonymous, except for circumstances when participants chose to provide their email address to be contacted for an optional interview.

Inclusion criteria were unadjusted height less than or equal to 148 cm and a completed survey for pregnancy-related questions and outcomes. Exclusion criteria were height greater than 148 cm, unknown height, or incomplete survey completion for pregnancy-related questions and outcomes.

Descriptive statistics, including frequency, means, standard deviation, and median, were reported for variables of interest. To examine the relationship between gestational age at delivery and different outcomes, comparisons were made on gestational age at delivery between pregnancies with and without the outcomes. P-values of the comparisons were from Generalized Estimation Equations, a statistical method for analyzing longitudinal or nested data, which were used to account for multiple pregnancies from the same subject. All statistical analyses were conducted using SAS 9.4.

3 Results

The survey was accessed 158 times, with the first page being completed 117 times. This equates to a response rate of 74% (117/158). Three subjects were excluded as their height was above the predetermined threshold. One subject was excluded because no height was provided. It was determined that one subject completed two surveys based on identical answers throughout all questions, and so only one of their surveys was included for this study. Therefore, there was a total of 112 eligible surveys. Eighty-eight participants completed to the last page, making the overall completion rate 78.6% (88/112).

The total number of eligible subjects that responded ‘yes’ or ‘no’ to the question, “Have you ever been pregnant?” was 84. Of those, 57 responded ‘yes,’ indicating they had a history of at least one pregnancy. Two subjects were excluded as they did not provide any further details regarding their pregnancies; therefore, a total of 55 subjects were eligible for analysis. Among the 55 subjects, there were 113 pregnancies. Five pregnancies were excluded due to incomplete information provided within the survey. Therefore, 108 pregnancies were included in the analysis.

Achondroplasia was the most common diagnosis amongst subjects, as the etiology of short stature in 34/55 (62%). Other diagnoses included the following: pseudoachondroplasia (3/55), hypochondroplasia (3/55), spondyloepiphyseal dysplasia (3/55), unknown diagnosis (3/55), diastrophic dysplasia (2/55), familial short stature (2/55), acromesomelic dysplasia (1/55), multiple epiphyseal dysplasia (1/55), Ellis Van Creveld syndrome (1/55), spondylometaphyseal dysplasia (1/55), and growth hormone deficiency (1/55).

Out of 108 pregnancies, there were 72 live births, three stillbirths, 23 spontaneous miscarriages (defined as delivery prior to 20 weeks gestation) and nine therapeutic abortions. One subject had a pregnancy end at 14 weeks but did not indicate whether this was a spontaneous loss or a therapeutic abortion. Of the 72 live births, 53 occurred at 37 weeks or greater, while 19 occurred prior to 37 weeks. This equates to an overall preterm birth rate of 26.4%. Sixty-nine of the 72 live births were via a caesarean section, 1/72 had a forceps delivery, and 2/72 had a spontaneous vaginal delivery. The forceps and vaginal deliveries were all at term and to people without a specific diagnosis of skeletal dysplasia; however, they met the height inclusion criteria. Of those who delivered prior to 37 weeks, the average gestational age at time of delivery was 34.2 weeks. For those who delivered at 37 weeks or later, the average gestational age at time of delivery was 38.3 weeks.

Preterm delivery risk was also calculated using only the first pregnancy delivered after 20 weeks for each subject. Two subjects delivered a stillborn beyond 20 weeks as their first delivery and were excluded. There were 45 first-time live births. Eleven delivered prior to 37 weeks, for a preterm birth rate of 24.4%.

Within the achondroplasia group, there were a total of 40 live births. The rate of delivery prior to 37 weeks was 25% (10/40) and after 37 weeks was 75% (30/40). For those who delivered preterm, the average gestational age at delivery was 35.4 weeks. For those who delivered at term, the average gestational age at delivery was 38.3 weeks. Demographic information of those that delivered at term and preterm are shown in Table 1.

Sub-analyses were performed to investigate whether individual variables conferred an increased risk of preterm delivery (Table 2). This included worsening respiratory difficulty during pregnancy, worsening pain during pregnancy, a composite outcome of respiratory difficulty and worsening pain, having a baby with a postnatal diagnosis of short stature, gestational diabetes, preeclampsia, and PPROM. No significant associations were found between these variables and a risk for preterm delivery.

Table 3 provides a qualitative description of notable pregnancy events and complications of all subjects who delivered prior to 37 weeks, including an assessment of whether the birth was spontaneous (contractions started) or iatrogenic (delivery occurred in the absence of spontaneous contractions.)

Table 4 compares complications of pregnancy in this population of women with short stature to those of the general population quoted from various sources within the literature. The risk for preeclampsia, gestational diabetes, PPROM, polyhydramnios, and delivery of an SGA neonate are all elevated compared to general population estimates in this small sample size.

The results of this study demonstrated that the risk of preterm birth in women with skeletal dysplasias and short stature may be elevated compared to the general population. Although this is one of the largest studies to date of this population, it was, overall, of very limited size. In the study population, the risk of delivery prior to 37 weeks was 26.4% for all births and 24.4% for first-time births. The preterm birth risk for the general Canadian population was estimated at 7.7% for the years 2000 to 2013 [8], 10% in the United States for the year 2017 [9], and 9.6% worldwide in the year 2005 [10]. When preterm delivery occurred in the study population, it was at an average gestational age of 34.2 weeks.

The secondary objective of this study was to determine if there were identifiable risk factors that, if present in women with short stature, would increase the likelihood for preterm delivery. It is well documented that women with short stature and skeletal dysplasias often experience significant worsening of respiratory function in pregnancy due to their underlying condition [4, 12]. Several case reports have cited respiratory dysfunction or respiratory failure as the sole indication for preterm delivery [7, 13]. Out of the 19 women who delivered prior to 37 weeks in this study, 10 reported worsening of their respiratory function. A statistical analysis comparing the gestational age at delivery in women with respiratory dysfunction to those without did not find an increased risk for preterm delivery (p = 0.172). Worsening of respiratory function is a very common occurrence in women with short stature, but it seems that in the majority of individuals it is not severe enough to warrant early delivery. It is recommend to have baseline pulmonary function testing at the beginning of pregnancy in order to monitor for this complication [4]. This is still an important investigation for the population of pregnant women with skeletal dysplasias, but it seems that decision for delivery based on this complication is highly individual and occurs in rare and extreme cases.

The majority of individuals with achondroplasia develop thoracolumbar kyphosis, which predisposes them to the development of spinal stenosis later in life [12]. During pregnancy, there is an exaggerated lumbar lordosis, which can lead to significant pain claudication [4]. Worsening of any type of pain was reported by many subjects in this study and present in nine of the 19 individuals who delivered preterm. When this variable was analyzed as a risk factor for preterm delivery, a statistically significant association was not found. Similar to worsening of respiratory function, worsening of pain is a very common occurrence in the population of women with short stature, but it did not appear to impact the risk of preterm birth.

Previous reports have speculated that the incidence of preeclampsia in individuals with skeletal dysplasia is higher than that of the general population [14]. In 1982, Lattanzi et al. reported on 16 cases of pregnancy in individuals with achondroplasia; four of these pregnancies were complicated by preeclampsia. Vance et al. (2012) reported on a case of pregnancy in an individual with primordial dwarfism who developed severe preeclampsia at 24 weeks, necessitating delivery. A proposed physiological explanation is that the placental mass is relatively large compared to the maternal body habitus, thus causing an alteration in the hormonal drivers of this disease process [14, 15]. In this study population, 8% (6/75) of subjects reported developing preeclampsia. The estimated incidence of preeclampsia in the general population is 1.15–3% [16, 17]. Therefore, this data supports the theory that preeclampsia risk may be elevated in this population. When sub-analysis was performed, there was not a significantly increased risk of preterm birth in subjects with this complication compared to those without. Thus, although the risk of preeclampsia seems to be higher, it does not appear to be a factor that can solely explain why the preterm birth risk is higher in subjects with skeletal dysplasias than those without. Potential explanations for this finding include that the preeclampsia risk is elevated only at later gestational ages or that the study population was too small to detect a difference.

This study also found in increased risk of PPROM and delivery of an SGA neonate when compared to the general population. Neither of these complications of pregnancy was solely associated with an increased risk of early delivery; however, increased chance of PPROM approached statistical significance in the preterm group, which is understandable as PPROM is associated with preterm birth. Underlying reasons as to why these complications seem to be higher in this population is an area of future study.

This study has a number of strengths. Due to the rarity of skeletal dysplasias, the literature to date consists of case reports and small series, thus limiting understanding of pregnancy complications in women with skeletal dysplasias. However, by means of a survey distributed to multiple organizations from multiple countries, accessing and including a large number of individuals with a rare condition was possible. Detailed information on both maternal and fetal outcomes were included due to the survey design. The main limitation of this study is that the results were susceptible to response bias. Individuals may have responded to the survey because of an either very positive or negative experience with their respective healthcare system. The survey was administered in English only and preferentially included those with access to the internet. The data was subject to recall bias since subjects may have remembered certain events of their health history in differing amounts of detail. Because the data was acquired anonymously and by means of survey, there was not the opportunity to seek out objective clinical assessments or clarification. As well, although this is the largest contemporary assessment of pregnancy outcomes in people with skeletal dysplasia, the numbers are still relatively small, which significantly limits the ability to assess less common outcomes and prevents adjustment for covariates. Further, a well-defined control group was not established given that the survey was administered to subjects in Canada, the United States, the United Kingdom, and worldwide.

In summary, this study demonstrated that women with skeletal dysplasias and short stature do appear to have an increased risk of preterm delivery compared to the general population. The reasons for preterm birth are likely to be multifactorial or due to unknown mechanisms. A formal registry for this population would help to clarify reasons for preterm delivery with the goal of developing strategies for surveillance and prevention.