1. Background
Stillbirth is known as one of the most unpleasant consequences of pregnancy (1) that is defined by the World Health Organization (WHO) as a baby born without vital signs or as fetal death after 28 weeks of pregnancy with a birth weight of 1,000 grams or a body length of 35 cm (2). Most stillbirths occur at the end of pregnancy and their main cause is mostly unknown (3). The Center for Disease Control and Prevention (CDC) has classified stillbirth as early stillbirth (death in the 20th to 27th weeks of pregnancy), late stillbirth (death in the 28th to 36th weeks of pregnancy), and term stillbirth (death in the the37th or 38th weeks of pregnancy or at delivery) (4).
Epidemiological studies show that 3 million stillbirths occur around the world every year (5). The majority of stillbirths (nearly 98%) occur in low- and middle-income countries and more than half of them (55%) occur in rural sub-Saharan Africa (6). Although some developed countries have reported the rate of stillbirth as three per thousand live births (7), a ten-fold increase has been observed in sub-Saharan Africa and Southeast Asia compared to in high-income countries; thus the rate of stillbirth in these countries is 30 per thousand live births (8). In the Middle East and neighboring countries of Iran, the rate of stillbirth varies from eight per thousand live births in Qatar to 38 per thousand in Afghanistan (9). According to the reports of the WHO collaborative centers in Argentine, Egypt, India, Peru, South Africa, and Vietnam, the rate of stillbirth is 12.5 per thousand live births. In the past two decades, the rate of stillbirth has varied from 12.8 to 40.0 per thousand live births in different regions of Iran (10).
The risk factors of stillbirth consist of bacterial infection, congenital diseases, chromosomal abnormalities, congenital diabetes, high blood pressure (preeclampsia), mother’s drinking and smoking or chemical drug use, placental abruption, trauma, exposure to radiation, female genital mutilation (11), umbilical cord problems (12), mother’s age, multiple pregnancy, a previous history of stillbirth and miscarriage, low maternal education and premature rupture of the membranes in women over the age of 35 (13).
2. Objectives
The high rate of stillbirth is a major public health concern that demonstrates the rate of population growth. Given the lack of sufficient evidence and knowledge about stillbirth and its pre-pregnancy risk factors in Iran and the absence of comprehensive studies at the national level, as well as the ethnic differences between countries, the present study was conducted to determine some of the pre-pregnancy risk factors of stillbirth in mothers presenting to healthcare centers across different regions of the country.
3. Methods
The present population-based case-control study was conducted to review the pre-pregnancy risk factors of stillbirth in ten provinces/cities of Iran, including Fars, Hormozgan, Kermanshah, Hamadan, Kohgiluyeh and Boyer-Ahmad, Yazd, South Khorasan, and Golestan provinces and the cities of Mashhad and Zahedan. A total of 1,459 (47.3%) women who their last pregnancy ended in stillbirth and a total of 1,626 (7.52%) women who their last pregnancy ended in a live birth presenting to primary healthcare centers (PHC) across the country entered the study. Data were collected from different geographical regions and ethnic groups of Iran so as to enable the generalization of the results to the entire Iranian population.
As defined by the CDC, stillbirth is classified into three categories, including early stillbirth (death in the 20th to 27th weeks of pregnancy), late stillbirth (death in the 28th to 36th weeks of pregnancy) and term stillbirth (death in the 37th or 38th weeks of pregnancy or at delivery). This study collected the overall records of stillbirth from the PHCs of the mentioned provinces and cities. The intended urban and rural PHCs were taken as clusters from which the household health records and indexes and hospital death certificates were randomly selected. PHCs are established to follow up pregnant women who do not refer to the centers for care. Any woman reporting a stillbirth is asked specific questions to ensure her diagnosis and the recorded data is then announced to the health network and the health network receives records of stillbirth from each hospital and compares it against the statistics provided by the PHCs, and if a hospital reports a case of infant death that the center takes into account as stillbirth, the center is informed so that they can correct the statistics and record accurate data.
In the present study, data were collected using a researcher-made questionnaire, household health records from public healthcare centers and interviews with the women. The interviews inquired about the parents’ personal information (mother’s age, parents’ education, parents’ occupation, place of residence, ethnicity, family relationship, etc.) and pregnancy information (number of pregnancies, outcome of previous pregnancies, birth spacing, planned or unplanned pregnancy, method of contraception, menstrual cycle regularity, etc.). The validity of the questionnaire was approved by experts and its Cronbach’s alpha reliability was estimated as 0.66 through a pilot study of 50 mothers in Kermanshah province. Stillbirth was the dependent variable and the independent variables included the mother working on night shifts, the mother’s place of residence, ethnicity, birth order, menstrual cycle regularity, parents’ age at delivery, parents’ nationality, parents’ education, parents’ occupation, consanguineous marriage, mother’s pre-pregnancy weight, mother’s height, outcome of previous pregnancies, pregnancy spacing (time since last pregnancy), planned or unplanned pregnancy, type of delivery, reason for cesarean section, method of contraception, the smoking and drinking history of the mother and her relatives, a previous history of miscarriage and stillbirth and mother’s history of chronic diseases such as diabetes and mental disorder.
Ten provinces/cities of Iran, including Fars, Hormozgan, Kermanshah, Hamadan, Kohgiluyeh and Boyer-Ahmad, Yazd, South Khorasan and Golestan provinces and the cities of Mashhad and Zahedan, were selected through cluster sampling and based on their geographical location. The provincial capital and four cities of each province (including cities from the North, South, East, and West of each province) were selected as the clusters. One urban and one rural PHC were selected from each cluster and the women who presented to the PHCs were given the questionnaires to fill out. Ten cases of stillbirth filled out the questionnaires in each center. If the cases of stillbirth were less than ten in any of the centers, the closest center was selected so as to complete the remaining questionnaires. If, however, the cases of stillbirth were more than ten in the centers, ten of them were randomly selected to fill out the questionnaires. The controls were also randomly selected from the centers on the same day as the cases. Data were collected from these centers by trained interviewers based on a common guideline. A prerequisite of the study was to have an equal number of samples from each center in the case and control groups. The sample size was calculated as 800 per group with the design effect estimated at two. Given that 24 independent variables were under study, the overall sample size was increased to 1,040 per group, making for a total of 2,080 samples. During data collection, this sample size further increased due to the good cooperation of some of the provinces and in order to increase the accuracy of the findings. The data collected were then entered the Excel and analyzed by SPSS software. P values less than 0.05 were considered statistically significant.
3.1. Data Analysis
The univariate analysis examined the relationship between each independent variable and stillbirth. If P value was less than 0.25 in the analysis, the independent variable was deemed suitable for undergoing multivariate analysis, which is a type of analysis that examines the effects of several independent variables on the dependent variable simultaneously in order to control the different confounding variables. The chi-square test and a binary logistic regression model were, therefore, used to describe and analyze the univariates and the multivariates. Logistic regression models help describe the relationship between a dependent binary variable and one or several continuous or discrete independent variables. The dependent variable examined in this study was “having experienced a stillbirth or not”. The binary model of logistic regression was used in this study to control the variables that are likely to have confounding effects (14).
The modeling was performed using selected variables, including the parents’ demographic information (mother’s age, education, occupation, working on night shifts, place of residence and ethnicity), previous history of miscarriage and stillbirth, birth spacing, outcome of previous pregnancy, birth order, and menstrual cycle regularity. The variables were selected for the final model using a forward logistic regression and included the mother’s education, age and disease history, a previous history of stillbirth, menstrual cycle regularity, a history of miscarriage and pregnancy spacing.
4. Results
Table 1 presents the demographic characteristics of the participants using univariate analysis. The raw analysis of the data showed that the chance of stillbirth is higher in women over 35 than in those below the age of 35. The odds ratio of stillbirth was also higher in men over 37 than in those below the age of 37; however, after that, the mothers’ age was entered the model, the fathers’ age was removed, which may suggest extreme linearity between the two variables. The odds ratio of stillbirth was much higher in women with low levels of education compared to women from the other education groups. The analysis of the relationship between education and stillbirth using the chi-square test led to consistent results. A remarkable finding of the mother’s education is that this variable was significant in all models. The variable of occupation in both mothers and fathers had a significant relationship with stillbirth using the univariate analysis; however, this variable did not remain in any of the models. The odds ratio of stillbirth was higher in women with Kurd, Turk or Turkmen ethnicity compared to women with other ethnicities.
Table 1.Demographic Characteristics of the Cases and Controls and Univariate Analysis
| Variable | Case (%) | Control (%) | OR (95% CI) | P Value |
|---|---|---|---|---|
| Address | ||||
| Urban | 658 (46.4) | 761 (53.6) | 1.06 (0.92 - 1.23) | 0.387 |
| Rural | 760 (47.9) | 825 (52.1) | ||
| Age of mother | 2.14 (1.65 - 2.78) | 0.001 | ||
| Less than or equal 35 | 1257 (45.4) | 1514 (54.6) | ||
| Greater than 35 | 173 (64.1) | 97 (35.9) | ||
| Age of father | 1.42 (1.16 - 1.73) | 0.001 | ||
| Less than or equal 37 | 1166 (45.6) | 1390 (54.4) | ||
| Greater than 37 | 259 (54.4) | 217 (45.6) | ||
| Nationality | ||||
| Fars | 830 (44.9) | 1019 (55.1) | 1 | - |
| Lor | 93 (46.3) | 108 (53.7) | 1.05 (0.78 - 1.41) | 0.020 |
| Turk | 257 (50.7) | 250 (49.3) | 1.26 (1.03 - 1.53) | 0.002 |
| Kurd | 34 (68) | 16 (32) | 2.60 (1.43 - 4.76) | 0.002 |
| Arab | 19 (54.3) | 16 (45.7) | 1.42 (0.74 - 2.85) | 0.271 |
| Baloch | 99 (49.5) | 101 (50.5) | 1.20 (0.89 - 1.61) | 0.214 |
| Turkmen | 74 (54.4) | 62 (45.6) | 1.46 (1.03 - 2.07) | 0.032 |
| Other* | 15 (37.5) | 25 (62.5) | 0.73 (0.38 - 1.40) | 0.354 |
| Education level of a mother | ||||
| Illiterate | 108 (61.4) | 68 (38.6) | 3.39 (2.33 - 4.94) | 0.001 |
| Elementary | 395 (55.6) | 315 (44.4) | 2.68 (2.05 - 3.50) | 0.001 |
| Guidance | 363 (51.1) | 347 (48.9) | 2.23 (1.71 - 2.91) | 0.001 |
| High school | 475 (42.3) | 648 (57.7) | 1.56 (1.22 - 2.01) | 0.001 |
| Collegiate | 115 (31.9) | 246 (68.1) | 1 | Reference |
| Mother’s history of night shift | 0.51 (0.27 - 0.95) | 0.035 | ||
| Yes | 15 (32.6) | 31 (67.4) | ||
| No | 1315 (48.5) | 1394 (51.5) | ||
| Education level of father | ||||
| Illiterate | 69 (57) | 52 (43) | 2.29 (1.51 - 3.46) | 0.001 |
| Elementary | 290 (55.1) | 236 (44.9) | 2.12 (1.62 - 2.76) | 0.001 |
| Guidance | 471 (52) | 434 (48) | 1.87 (1.47 - 2.38) | 0.001 |
| High school | 475 (42.5) | 642 (57.5) | 1.27 (1.01 - 1.61) | 0.040 |
| Collegiate | 150 (36.7) | 259 (63.3) | Reference | |
| Mother’s occupation | ||||
| Housewife | 1339 (48.3) | 1433 (51.7) | Reference | |
| Employee | 65 (32.5) | 135 (67.5) | 0.51 (0.38 - 0.69) | 0.001 |
| Agricultural livestock | 25 (56.8) | 19 (43.2) | 1.40 (0.77 - 2.56) | 0.264 |
| Other | 18 (48.6) | 19 (51.4) | 1.01 (0.53 - 1.94) | 0.967 |
| Father’s occupation | ||||
| Businessman | 937 (48.5) | 993 (51.5) | Reference | |
| Governmental employee | 178 (37.4) | 298 (62.6) | 0.63 (0.51 - 0.077) | 0.001 |
| Farmer | 104 (49.1) | 108 (50.9) | 1.02 (0.76 - 1.35) | 0.888 |
| Rancher | 30 (53.6) | 26 (46.4) | 1.22 (0.71 - 2.08) | 0.459 |
| Other | 201 (52.1) | 185 (47.9) | 1.15 (0.92 - 1.43) | 0.206 |
Table 2 presents the pre-pregnancy and pregnancy risk factors of stillbirth in women. Based on the results, the odds ratio of stillbirth was higher in women who did not use any contraceptive methods compared to those who did. The odds ratio of stillbirth was also higher in women with a history of stillbirth and miscarriage compared to those with no such history (OR: 2.99 [2.12 - 4.22]). Finally, 11 of the 24 variables were entered the multivariate analysis. Table 3 shows the results.
Table 2.Univariate analysis of maternal factors before pregnancy related to stillbirth
| Variable | Case (%) | Control (%) | OR (95% CI) | P Value |
|---|---|---|---|---|
| The cause of cesarean | ||||
| Elective cesarean | 19 (30.6) | 43 (69.4) | Reference | Reference |
| History of cesarean | 88 (31.2) | 194 (68.8) | 1.02 (0.56 - 1.86) | 0.931 |
| Lack of labor progress | 62 (36.3) | 109 (63.7) | 0.28 (0.69 - 2.40) | 0.427 |
| Embryonic problems or abnormal fetal presentation | 148 (67.9) | 70 (32.1) | 4.78 (2.59 - 8.80) | 0.001 |
| Other | 51 (41.5) | 72 (58.5) | 1.60 (0.83 - 3.06) | 0.154 |
| Type of delivery | ||||
| Natural | 1063 (48.7) | 1120 (51.3) | Reference | Reference |
| Cesarean | 380 (43.2) | 500 (56.8) | 0.80 (0.68 - 0.93) | 0.006 |
| Distance from a previous pregnancy | ||||
| More than 3 years | 379 (41.5) | 534 (58.5) | Reference | Reference |
| 1 to 3 years | 412 (48.8) | 432 (51.2) | 1.34 (1.11 - 1.62) | 0.002 |
| Less than a year | 119 (68) | 56 (32) | 2.99 (2.12 - 4.22) | 0.001 |
| First pregnancy | 519 (47.6) | 572 (52.4) | 1.27 (1.07 - 1.52) | 0.007 |
| The use of contraceptive methods | ||||
| Yes | 583 (41.8) | 811 (58.2) | Reference | Reference |
| No | 862 (51.9) | 799 (48.1) | 1.50 (1.30 - 1.73) | 0.001 |
| Type of contraception | ||||
| Pill | 242 (46.5) | 278 (53.5) | Reference | |
| IUDs | 29 (36.3) | 51 (63.8) | 0.65 (0.40 - 1.06) | 0.08 |
| Ampoules of injection Depo-provera | 53 (43.1) | 70 ( 56.9) | 0.87 (0.58 - 1.29) | 0.49 |
| Condom | 178 (39) | 278 (61) | 0.73 (0.57 - 0.94) | 0.01 |
| Other | 80 (37.6) | 133 (62.4) | 0.69 (0.49 - 0.95) | 0.02 |
| Regular menstruation | ||||
| Yes | 1207 (46.6) | 1384 (53.4) | Reference | |
| No | 218 (53.4) | 190 (46.6) | 1.31 (1.06 - 1.62) | 0.010 |
| History of stillbirth | ||||
| Yes | 130 (73.4) | 47 (26.6) | 3.28 (2.33 - 4.62) | 0.001 |
| No | 1329 (45.7) | 1579 (54.3) | Reference | |
| History of abortion | ||||
| Yes | 225 (60.5) | 147 (39.5) | 1.83 (1.47 - 2.28) | 0.001 |
| No | 1234 (45.5) | 1479 (54.5) | Reference | |
| Mother’s background diseases | ||||
| Yes | 116 (8.1) | 72 (4.5) | 1.86 (1.38 - 2.53) | 0.001 |
| No | 1312 (91.9) | 1522 (95.5) | Reference |
Table 3.Modeling of the Maternal Factors Before and During Pregnancy and Demographics with Stillbirths
| Variable | OR (95% CI) | P Value |
|---|---|---|
| History of stillbirth | ||
| Yes | 2.64 (1.81 - 3.85) | 0.001 |
| No | 1 | |
| History of abortion | 1.57 (1.21 - 2.03) | 0.001 |
| Yes | 1 | |
| No | ||
| Education level of mother | ||
| Illiterate | 1 | |
| Elementary | 3.50 (2.30 - 5.33) | 0.001 |
| Guidance | 2.79 (2.07 - 3.76) | 0.001 |
| High school | 2.60 (1.63 - 3.51) | 0.001 |
| Collegiate | 1.81 (1.37 - 2.40) | 0.001 |
| Regular menstruation | ||
| Yes | 1 | 0.029 |
| No | 1.29 (1.02 - 1.64) | |
| Age of mother | ||
| ≥ 35 | 1 | |
| ≤ 35 | 1.58 (1.17 - 2.14) | 0.003 |
| Distance from previous pregnancy | ||
| Less than a year | 2.64 (1.80 - 3.87) | 0.001 |
| One to three years | 1.35 (1.09 - 1.66) | 0.005 |
| More than three years | 1.72 (1.40 - 2.10) | 0.001 |
| First pregnancy | 1 | - |
| Mother’s background diseases | ||
| Yes | 79.1 (52.2 - 28.1) | 0.001 |
| No | 1 | - |
According to Table 3, the results of the multivariate logistic regression showed that stillbirth was related only to seven of the 11 variables selected for entering the multivariate analysis model, which include, in the order of importance, mother’s education, history of stillbirth, pregnancy spacing, history of other diseases, age, history of miscarriage and menstrual cycle irregularity.
5. Discussion
The results of the study showed that the risk of stillbirth is higher in women over 35 compared to those younger than 35, which is consistent with the results of other studies (15). Hajian-Tilaki et al. stated that this higher risk may be associated with increased maternal age, the history of stillbirth, multiparity and placental abruption, the combination of which increases the risk of stillbirth in older women (16). Age may have no effects on the likelihood of stillbirth per se, and studies have shown that the increased risk of stillbirth in older women may be associated with their development of chronic diseases at this age (hypertension etc.) and pregnancy complications such as diabetes, gestational hypertension, multiparity, and previous history of miscarriage (17). Reddy et al. argued that although an increased maternal age is associated with an increased risk of stillbirth, maternal hypertension and diabetes are two common pregnancy complications at the age of 35 and above that are considered the confounding factors of this relationship and are responsible for the significant relationship between maternal age and stillbirth (18). Studies also have shown that increased maternal age is an independent risk factor for stillbirth even after controlling the effect of maternal morbidities (1). They did not find a significant relationship between higher maternal age and the risk of stillbirth, which is attributed to the elimination and exclusion of congenital anomalies and they explained since a higher maternal age increases the risk of congenital anomalies, the relationship between maternal age and stillbirth remains unidentified after eliminating and excluding these factors. Owing to the present study had a large sample size, it can be argued that the increased risk of stillbirth was associated with higher maternal age rather than with maternal factors. Preconception counseling and improving prenatal care can help reduce the risk of stillbirth in older women (18).
The risk of stillbirth was higher in illiterate women than in those with university education, which is consistent with the results of other studies (19) and can be attributed to the greater knowledge in educated mothers about the outcomes of pregnancy, their greater adherence to pregnancy health and their more frequent visits to healthcare centers for receiving prenatal care.
According to the findings, the risk of stillbirth was higher in women with a history of stillbirth or miscarriage compared to those with no such history, which is consistent with the findings of other studies (20). Akolekar et al. demonstrated that the risk of stillbirth increases in women with a history of miscarriage and depends on their gestational age at pregnancy loss. In this regard, the risk of stillbirth is 1.6 times higher in women with a history of miscarriage before 16th week of pregnancy, 6.3 times higher in those with miscarriage between 16th and 23th weeks and three times higher in those with miscarriage at later weeks. Akolekar et al. also showed that the risk of stillbirth is 50% lower in women with successful previous pregnancies than in women with a history of stillbirth or miscarriage in their previous pregnancy (21). In another study, Bhattacharya et al. showed that a history of stillbirth is associated with the risk of stillbirth in the current pregnancy as well as obstetric complications such as preterm delivery, low birth weight, and pre-eclampsia in subsequent pregnancies (22).
Samueloff et al. compared women with a previous history of stillbirth and women with their first experience of stillbirth and found hypertension and diabetes as probable reasons for stillbirth (23). Pariente et al. reported significantly higher rates of cardiovascular or renal diseases in women with a history of one or more stillbirths (24). History of stillbirth is known to be a risk factor of stillbirth in pregnant women; however, the underlying causes of previous stillbirths cannot be detected and identifying these factors can, therefore, help health service providers mark high-risk women for further supervision and pregnancy care. Promoting public health and increasing awareness about the risk factors of stillbirth can also help reduce the rate of stillbirth (25).
The risk of stillbirth was also higher in women with less than three years between their pregnancies compared to others, which is also in line with the findings of other studies (26). Different studies have found that short pregnancy intervals does not allow the mother’s body to compensate for the complications and pressures of the previous pregnancy, delivery, and breastfeeding and leads to negative outcomes such as low birth weight, preterm delivery, stillbirth, miscarriage, and bleeding in late pregnancy (27). Short pregnancy interval is also associated with more pregnancy bleeding, which may be caused by the disruption in the normal reconstruction of the endometrial blood vessels after delivery followed by a new pregnancy. In this regard, the placental blood flow decreases inside the uterine; thus the risk of bleeding increases in late pregnancy. Therefore, training women to change their poor reproductive behaviors and observe optimal pregnancy intervals by improving their access to contraceptive methods are essential.
According to the present findings, some pre-pregnancy risk factors are significantly associated with stillbirth. Educational interventions to improve the awareness of women of childbearing age and performing effective health measures can help reduce the risk factors of stillbirth and the prevalence of this complication. Considering that this study was a case-control research that interviewed patients about their history of pregnancy and diseases, a recall bias may have occurred for some of the independent variables, which is a common disadvantage of retrospective studies; nonetheless, the researcher made great efforts to reduce this bias via selecting objective variables that could have a higher recall and lower recall error.
5.1. Study Limitations and Ethical Issues
This study was designed in such a way to yield minimum limitations; however, no studies can evade limitations completely. The multiplicity of the interviewers may have led to some inter-interviewer variation. To fix this variation, the researcher prepared a common guideline for the interviewers and distributed it among all the centers.
