Obesity and Diabetes as Main Risk Factors for Stillbirth in a Population from Northeastern Mexico

Purpose Stillbirth is an important health problem in obstetrics practice. In Mexico, half of the stillbirths have an explainable cause. The aim of this study was to detect potential stillbirth risk factors and etiology by implementing a multidisciplinary workup at the Hospital Universitario “Dr. José Eleuterio González”. Methods This is a prospective, descriptive, observational study that included stillbirths from the Obstetrics Service from October 1st, 2019 to May 25, 2020. Evaluation strategies included a complete maternal medical history, physical examination of the fetus, and a photographic medical record. A stillbirth needed to have either a prenatal ultrasound, or a postnatal x-ray, and if possible, a fetal autopsy. Multiplex Ligation Probe Amplication (MLPA) was performed with an umbilical cord sample. Results Thirty-three stillbirths were reported; 21 were included in the analysis. Eleven mothers (52.3%) had known risk factors for stillbirth, mainly elevated body mass index and diabetes. On physical examination, external birth defects were found in 8 fetuses (38%). X-ray was performed in 14 cases (66%). Alterations were detected as a probable etiologic cause just in one. All cases underwent MLPA, which were reported negative. Three cases had criteria for autopsy. Findings were inconclusive to determine etiology. Conclusion Diabetes and obesity were the most frequent risk factors associated with stillbirth in this Mexican population. These factors are preventable by implementing strategies that lead to better prenatal care.


Introduction
Stillbirth is de ned by the World Health Organization (WHO) as "death prior to the complete expulsion or extraction from its mother of a product of conception, irrespective of the duration of pregnancy." According to gestational age, it is considered a fetus of 22 weeks or that has reached a weight of 0.5 kilograms [1]. It is a devastating problem in obstetrics for families, society, and public health. It accounts for two-thirds of perinatal mortality [2]. Stillbirth occurs in 3 to 5 of every 1000 births [3]; however, these numbers might be underestimated because of inadequate registry in some countries and a lack of consensus in its classi cation and de nition [4]. Developing countries report a higher incidence. Regions of Africa and Asia report an incidence of 28.7 and 25.5 for every 1000 births, respectively [5].
Stillbirth causes have been classi ed into different categories: maternal factors, fetal factors, or a combination of both. Suspected causes of stillbirth include genetic abnormalities, infection, fetal-maternal hemorrhage, and various medical conditions in the mother such as obesity, advanced maternal age, smoking, diabetes, hypertension, lupus, and thyroid diseases. Birth defects of chromosomal, monogenic, or multifactorial etiology have been described among the fetal factors. Combined factors include placental and umbilical cord pathologies [3,6,7].
It is di cult to be certain of the stillbirth etiology for several reasons. More than one cause may be involved, or some conditions might be associated but not directly causal. It is important to distinguish conditions that are mainly associations rather than causes [8]. An adequate workup for stillbirth has not been implemented. A lack of registry, consensus, comprehension about etiology, and ethical and emotional implications of some procedures might account for the absence of global stillbirth guidelines [8].
In general, a complete evaluation should have as a goal effective procedures to determine the cause and identify conditions that might be prevented. The use of a systematic evaluation, including postmortem examination, placental pathology, medical record abstraction, and maternal interview, has a higher potential of determining a true cause, as has been reported in some centers with an e cacy of 25 to 90% [2,8].
In Mexico, as in many other countries, half of the stillbirths have no explicable cause. Implementing a multidisciplinary workup that includes a complete clinical record, physical examination, autopsy, plain x-ray, cytogenetic or molecular studies permits adequate evaluation and the possibility of reaching an etiologic diagnosis, allowing the development of prevention strategies, and providing families with adequate genetic counseling [3,7,9,10].
This study aimed to implement a multidisciplinary workup to identify the etiology and potential risk factors for stillbirth at a tertiary-care university hospital in Northeast Mexico.

Methods
This was a descriptive, observational prospective study of stillbirths from the Obstetrics Service of the Hospital Universitario "Dr. José Eleuterio González" from October 1, 2019, to May 25, 2020. Cases were collected by a resident of the Genetics Department who veri ed the inclusion criteria and obtained informed consent from the mother. The Ethics Committee of the Facultad de Medicina and Hospital Universitario of the Universidad Autonoma de Nuevo Leon approved the study. Stillbirths of 22 weeks or more (early and late fetal deaths) with a fetus weight > 0.5 kilograms from adult mothers were included in the study. All participant mothers provided written informed consent for participation and use of biological material for further analysis. A Genetics department resident obtained the mother´s complete medical history. After birth, a physical examination was performed on the stillborn, and a photographic record was obtained as part of every patients´ le. If the parents consented and pathology requirements were ful lled, an autopsy was performed.
The autopsy requirements were a period less than six hours from fetus extraction to the autopsy, a copy of the stillbirth certi cate performed by the obstetrician, and a summary from the geneticist pinpointing the most important ndings of the medical history, physical examination, and suspected diagnosis. If any of these was not achieved, a plain x-ray was used as an alternate study.
Fetuses with any of the procedures mentioned before or maceration grades IV and V (stillbirth scale) [11] were excluded.
Inadequate or contaminated biological samples were eliminated.

Procedure
Cases needed to have an autopsy, a prenatal ultrasound, or a postnatal plain x-ray. Molecular studies were performed in all cases. A 1-cm tissue sample from the umbilical cord was obtained for DNA extraction using the Gentra Puregene DNA extraction kit. This sample was used to perform Multiplex Ligation Probe Ampli cation (MLPA) (SALSA MLPA Probemix P095 Aneuploidy, MRC Holland), which detects DNA sequences, and has probes for chromosomes X, Y, 13, 18, and 21.
The data obtained were analyzed using quantitative and qualitative statistics to calculate mean, median, and range. Probable associations were evaluated for each case. SPSS v21 was used for statistical analysis.

Results
Thirty-three stillbirth cases were reported at the "Dr. José Eleuterio González" University Hospital from October 2019 to May 2020, with a frequency of 7/1000 live births; of these, only 21 ful lled the inclusion criteria. Six were lost due to delayed noti cation, three mothers were less than 18-years-old, two decided not to participate, and one was eliminated due to extreme maceration.
The maternal demographic characteristics are shown in Table 1. The maternal age range was 18 to 38 years (median = 25 years); three were older than 35. Eleven mothers (52.3%) had known risk factors for stillbirth; some had more than one. Four mothers (19%) had pregestational diabetes with poor metabolic control before and during pregnancy. Overweight and obesity were found in 12 (57 %) women. Three (25%) were overweight (BMI = 25-29) and 9 (75%) were obese (BMI> 30). None of the mothers had preconceptional folic acid intake.
Two were twin pregnancies. In both, the second baby was born without complications. None of the families participating had the antecedent of stillbirth. One had two previous abortions at ve and eight weeks of gestation.
Of the stillbirths included in our study, 10 were a rst pregnancy (47.6%); 14 (66.6%) occurred between 29 and 37 weeks of gestation. Pregnancy ended by vaginal delivery in 18 cases (85.7%). Twelve of the stillbirths were female (57%), and 9 (42.8%) male. The stillbirths' weight was predominantly >1000 grams (62%), and none of the stillbirths had low weight for gestational age. The degree of maceration was important for inclusion and physical examination. This evaluation was adequately performed in most cases since 76.1% presented grade I maceration. The demographics of the stillbirths are shown in Table 2.
External birth defects were found in 8 fetuses (38%) on physical examination; 5 had multiple malformations, 1 anencephaly, and 2 pes valgus deformity (Table 3). At initial evaluation, chromosomal alterations were suspected in two patients, and in another, a monogenic syndrome was considered.
X-ray was performed in 14 cases (66%). Alterations were detected in one as a probable etiologic cause, nding pathognomonic signs of congenital syphilis.
All cases underwent MLPA, which was reported negative. Conventional cytogenetics was performed in 18 patients with an umbilical cord sample. Three had adequate cell growth and yielded normal results. Only three cases had criteria for autopsy; nevertheless, the ndings were inconclusive for determining the cause of stillbirth.

Discussion
Although the stillbirth rate has declined in recent years, it is still an important obstetric health problem due to its association with increased perinatal mortality. Stillbirth rates remain high, especially in low and middle-income countries, where rates are 25 per 1000, ten times higher than in high-income countries.
In our study, the rate of fetal deaths was 7/1000, which is lower than the mean for Latin American countries (in 2015 rate was 8.1 per 1000 births, range 3.1-24.9) but is close to Mexico's average rate of 7.9/1000 [5]. However, problems in the noti cation of stillbirth and the study's duration could in uence our incidence. A long-term evaluation should be performed for more accurate data. Nevertheless, it represents a public health issue for which identifying risk factors is important to propose prevention strategies.
From the major causes of stillbirths, we can point out the following: childbirth complications, multiple pregnancies, post-term pregnancy, maternal infections (malaria, syphilis, and HIV), maternal disorders (especially hypertension, obesity, and diabetes), fetal growth restriction, and congenital abnormalities. Several algorithms have been proposed for the study of stillbirth to identify possible causes of this fatal outcome of pregnancy.
In this multidisciplinary workup, we used different tools. The maternal medical history, physical examination, photographs, medical record, prenatal ultrasound, x-ray, or autopsy were obtained. MLPA was performed in all cases.
The correlation between maternal medical history and the description of birth defects in the fetus gave us information for possible risk factors in 16 cases, representing a usefulness of 76%, higher than 26% reported by Pauli et al.
External physical examination is important for the detection of birth defects. This procedure must be performed by a medical professional who has been trained in dysmorphology. Nevertheless, the maceration grade should be between 0 and III to facilitate the examination [11]. This approach let us identify malformations in 8 cases. Macroscopic evaluation should complement an autopsy which is considered the most useful study for stillbirth since it allows identifying characteristics not observable by physical examination. A complete autopsy should include an analysis of the placenta and umbilical cord [8, 9,11].
It can identify stillbirth etiology in about 90% of cases [9]. However, to maximize the autopsy's e cacy, it is important to consider the time from fetal death to the autopsy. The longer it takes to perform the autopsy, the less e cient it will be [9,14]. In our workup, only three autopsies were accomplished. The rest could not be achieved due to delayed noti cation, and in two cases, the mothers did not consent to the autopsy. It was di cult to assess the e ciency of the autopsy in our study with the data obtained, but the shortage of information obtained in those three cases may relate to a prolonged time from fetal death to evaluation.
Approximately 3% of prenatal ultrasounds detect an abnormality. This can range from minor to multiple defects, which can put the development of the fetus at risk [12]. The detection of anomalies in prenatal ultrasound can affect the pregnancy outcome, so it should be considered a part of the prenatal care of all women [13]. In the present study, 17 (53 %) women had a prenatal ultrasound, of which 9 (53%) showed either fetal or placental ndings of relevance for diagnosis, such as holoprosencephaly, anencephaly, hydrops fetalis, and placenta previa. The e ciency of prenatal ultrasound for detecting birth defects has been reported as 3% [12], with our rate of detection being higher.
A plain X-ray is also considered an adequate study when an autopsy cannot be performed [9]. It can identify stillbirth etiology in about 16% [6]. However, a plain X-ray might be more e cient when skeletal pathology is suspected. We only detected suggestive ndings in one patient (4.8%) with congenital syphilis.
Genetic etiology of stillbirth can be found in 25-35% [2,9], and 6 to 12% are chromosomal alterations [8]. Conventional genetics was once considered the gold standard for stillbirth evaluation [3,11,15,16,17]. We decided to perform MLPA containing probes for chromosomes X, Y,13,18 and 21, to evaluate the most common chromosomal causes of stillbirth, which were not the case in our study; therefore, we cannot discard other chromosomal or genetic etiology. Chromosomal microarrays (CMA) are now considered the rst line of study [16,17].
Maternal obesity (BMI > 30 Kg/m 2 ), has been positively associated with stillbirth [18]; however, no BMI score has been reported as secure or protective for stillbirth. In a meta-analysis by Aune et al., 18 cohorts were included to analyze BMI and stillbirth risk, nding that this association is positive for obesity and an important higher risk with a BMI of 40 [19]. In our study, 43% of the mothers had a BMI >30, which can account for the outcome of their pregnancies.
Pre-gestational diabetes mellitus (DM) (either type 1 or 2) is associated with an increased risk of fetal macrosomia, birth defects, and stillbirth at 36 to 42 weeks of gestation. Similar results have been reported in gestational diabetes [20,21,22]. Fiftyseven percent of the mothers in this study had poorly controlled pre-gestational DM, which is a high proportion compared to the 5.2% reported in another study (20). The physiopathology that explains how hyperglycemia might contribute to the events that lead to fetal death has not been fully determined. Elevation of plasma lactate was reported, predominantly in the third trimester, as a possible cause [23]. This result suggests that a lactic acid elevation leads to anaerobic metabolism, causing hypoxia, acidosis, and subsequent fetal death [24]. Obesity and diabetes are serious health problems in our country, mostly in northern Mexico, and occur in 5.2% of pregnancies [20].
It is important to establish that even if diabetes or obesity is identi ed as an important risk factor in a woman, a complete study of stillbirth should be done because that does not eliminate the probability of other causes.
Amongst maternal factors, infections account for 20% of stillbirths in developed countries [25]. Infections like cytomegalovirus and syphilis are frequently reported in these patients; usually, infectious etiology is not directly looked for; suspicion based on prenatal ultrasound or physical examination leads to its evaluation [8,25]. We identi ed infection risk factors by medical record data and plain X-ray.
Other maternal diseases have been implicated in stillbirth like thyroid pathologies, cardiovascular, renal, rheumatic diseases, and cancer. These conditions can be associated with stillbirth in either acute or chronic states; however, most women with these conditions can achieve live births, so they are considered important for stillbirth assessment, but causality should be evaluated in every case [8].
In Mexico, one of every 90 pregnancies are twin pregnancies. These convey a higher risk of birth defects (10 times more than singleton pregnancies) [26], as well as an increased stillbirth risk, 13 times higher in monochorionic pregnancies, and 5 times higher in dichorionic pregnancies [27].
The global strategy to end stillbirth classi es twin pregnancies as an important risk factor, recommending early delivery between 34 and 37 weeks of gestation in monochorionic pregnancies and 37 to 29 in dichorionic pregnancies [27,28,29]. Prenatal care in this scenario should be stricter.
Placental causes are attributed in 64.9% of stillbirths. Identi cation of placental pathology requires a microscopical and macroscopical evaluation after birth, prenatal markers are not considered useful for evaluation [30] Placental causes may be associated with growth restriction, but not all patients present growth restriction, and the ones that do, do not all result in stillbirth [30]. We did not perform studies on the placenta, so we cannot con rm or exclude placental etiology.
Another risk factor identi ed in our study was tobacco consumption. Exposure to tobacco, either active or passive, has been positively identi ed as a stillbirth risk factor. However, when exposure is implicated, fetal growth restriction is reported, when that signi cant data is lacking, the risk of stillbirth has been proposed to be like women without tobacco consumption [18,31] Conclusions We found that well-known public health problems in Mexico are present as stillbirth risk factors in our population. Diabetes and obesity are identi ed as the most important associations. These factors are preventable by implementing strategies that lead toward better prenatal care.
The best tools for identifying stillbirth risk factors at our institution were the obtention of a medical history, physical examination, and prenatal ultrasound; however, valuable tools such as an autopsy and genetic studies should be pursued. To increase the e cacy of an autopsy, we must reduce the time from death to study. To achieve a genetic diagnosis, we should implement studies that help identify chromosomal etiologies, not only the most frequent, but also those that are not dependent on cell culture, ideally CMA. Next-Generation Sequencing (NGS) platforms are useful when a monogenic cause is suspected.
By modifying these limitations, a long-term prospective study would be much informative to continue the study of stillbirth etiology.  (25) Obese (>30 kg/m 2 ) 9/12 (75) † Two patients consumed more than one substance. † † One patient had two diseases Grade III 2 (9.5) Table 3. Summary of stillbirths with birth defects and twin pregnancies, with risk factors of importance