Risk Factors for Miscarriage


This is a presentation for Risk Factors of Miscarriage From A Prevention Perspective: A Nationwide Follow-up Study, done in Denmark. Introduction. Miscarriage is the most common adverse pregnancy outcome, accounting for 5% to 15% of all pregnancies. Unfortunately, prevention is the only way we can intervene. Studies have shown that chromosomal abnormality of fetus accounts for almost 50% of miscarriages. It has been shown to be associated with advanced maternal age, as well. Studies of potentially modifiable risk factors– for example, smoking and alcohol– have show inconsistent results. Aim. The aim of this study is to identify modifiable risk factors for miscarriage, and also to estimate the preventable proportion of miscarriages that could have been attributed to these risk factors. Methods. All pregnancies registered in the Danish National Birth Cohort were included. However, there were some exclusion criteria, such as pregnancies enrolled at 20 weeks or later, pregnancies with no information of gestational age, hydatidiform mole and ectopic pregnancies. So women were recruited at their first antenatal visit at a general practitioner. The presented at a mean gestational age of 10.6 weeks. Following informed consent and permission from the Danish Data Protection Board and a scientific ethics committee, computer-assisted telephone interview was carried out at 12 weeks of gestation or as soon as possible after that. Women who miscarried prior to the schedule were offered an interview with similar questions. Established risk factors that are potentially modifiable and possible to evaluate were identified from literature review. A total of nine risk factors were identified, and they are classified according to before pregnancy and during pregnancy. So before pregnancy criteria include maternal age at conception, pre-pregnant weight status, previously diagnosed genital disease. During pregnancy, risk factors include alcohol consumption, coffee consumption, smoking, amount of exercise, daily lifting of heavy burdens of more than 20 kilograms, and lastly, work schedule. The outcome of interest in this study is miscarriage. Miscarriage is defined as spontaneous fetal death and/or expulsion before 22 weeks of gestation. And this is Denmark’s definition. Information on pregnancy outcome was obtained from the Danish registers or they were self-reported information from the subjects themselves. Multiple Cox regression analysis using gestational age as the time variable was performed to investigate if any of the nine variables affects the risk of miscarriage. A pregnancy was considered to be at risk of miscarriage from the time of enrollment to the first of the following events, namely miscarriage or induced abortion, or loss to follow-up, or survive more than or equal to 22 weeks. In order to distinguish between early and late abortion, a separate analysis was done for miscarriage in the first and second trimester. Differences in effect were tested using a standard Wald test. The risk of miscarriage for each individual woman was estimated from the Cox regression model, and this provided a background for estimating the population attributable fractions, which is the proportion assessed risk of miscarriage associated with exposure to a risk factor, i.e., a PAF represents proportion of cases that would be prevented if the risk factors were eliminated. In this analysis parity and highest household occupational status were included as non-modifiable confounders. In this study, paternal age is being excluded. Studies have shown that it is unrelated to the risk of miscarriage after adjusting for maternal age. In addition, for this study, many pregnancies have missing values on paternal age, as well. A separate subcohort analysis was done for two categories– number one, pregnancies for which information were collected before miscarriage. This is to test for recall bias. Number two, primigravid women who became pregnant within five months of trying. This is to test for behavior modification bias. A total of 100,418 pregnancies were enrolled in the Danish National Birth Cohort from 1996 to 2002. 7,701 pregnancies were excluded, as they have no interview done. After removing pregnancies that fulfill the exclusion criteria and pregnancies that have missing values on risk factors or confounders, a remaining of 88,373 pregnancies were eligible for the multiple Cox regression analysis. In this figure, it includes 3,079 miscarriages for analysis. Results. This graph shows the gestational age at which the interviews were done for the pregnancies. The dark blue lines are interviews done during pregnancy, whereas the light blue lines are the ones done after miscarriage has occurred. As you can see from this graph, most of the interviews were done after miscarriage has occurred. This table shows the results from the interviews. First, let’s look at the potentially modifiable risk factors. As you can see from the table, women who were more than 40 years of age at conception had two to three times higher risk of miscarriage as compared to women who were 25 to 39 years of age at conception. The amount of exercise and the amount of alcohol consumption during pregnancy appears to have a dose-response increase in risk of miscarriage. In other words, if you did more than 300 minutes of exercise or drank more than four drinks of alcohol per week, you are more likely to have a miscarriage as compared to women who had none. Similar to exercise and alcohol consumption, the amount of coffee consumed per day had a dose-response increase in risk of miscarriage. Additionally, women who lifted weights of more than 20 kgs per day during pregnancy appears to have increased risk of miscarriage, as well. In this study, smoking during pregnancy does not appear to have any effect on miscarriages. However, women who are extremities of BMI– that is, those who are either underweight or those who are obese– appeared to have a higher risk of miscarriages. Women who were working night shifts also had an increased risk of miscarriages. On the other hand, looking at women with previously diagnosed genital diseases, only those who had dysplasia in the cervix had an increased risk, whereas those who have conization or other genital diseases did not appear to have any increased risk. Among the single risk factors, the highest proportion of miscarriages were preventable if all women conceived at 25 to 29 years of age. Alcohol consumption was the most important risk factor during the pregnancy itself. Combining all the potentially modifiable risk factors– namely, maternal age of 25 to 29 years at conception, no consumption of alcohol during pregnancy, pre-pregnant normal weight, no lifting of more than 20 kg daily during pregnancy, and daytime work during pregnancy– about 25% of miscarriages were preventable. In other words, one in four miscarriages are potentially preventable. In summary, 5.2% of miscarriages were preventable by modification of multiple risk factors to low-risk levels. And maternal age and alcohol consumption were the most important risk factors. The authors compared this study to other studies in the literature. Looking at risk factors before pregnancy, the increased risk of miscarriages with increased maternal age is consistent with literature. However, the results of pre-pregnant weight status in the literature appear to be varied. Some of the possible reasons could be that obese women are less fertile, or that fertility treatment is a risk factor for miscarriages itself. The risk of miscarriages for previously diagnosed genital diseases seemed to be consistent with literature only for dysplasia of cervix, whereas no association with so cervical conisation was found. For risk factors during pregnancy, the results for alcohol consumption, coffee consumption, daily lifting of heavy burdens, and night work schedule appears to be consistent with the literature. On the other hand, there are varied results for smoking as a risk factor for miscarriages. Some possible explanation would be that women with previous miscarriages have stopped smoking to increase their chances of successful childbearing. The amount of exercise showed no effect in other studies. The possible reason would be that most of the women in this study were interviewed only after they had a miscarriage, and these women may have had recall bias. As mentioned in the methods, a subcohort analysis was done for these two categories. In pregnancies for which information were collected before miscarriages, the dose-response effect of exercise on miscarriages disappeared. There was effect of smoking on miscarriages found. And dysplasia in cervix had a protective effect instead. Two, in primigravid women who became pregnant within five months of trying, the effects of alcohol consumption was even more pronounced. The protective effect of smoking disappeared. The increased risk of miscarriages associated with obesity also disappear. But there was an slightly increased risk associated with dysplasia in cervix. Here are some strengths and limitations of this study. The strengths include a large sample size, early recruitment allows risk estimates from six gestational weeks, and availability of detailed self-reported information on modifiable risk factors and confounders. The limitations include– majority of women were recruited after the gestational age where miscarriage is most common; compared to other pregnancy outcomes, an increased proportion of participants whose pregnancies ended in an early miscarriage decided not to participate; unable to conclude if the associations are causal; study population is not representative of Danish pregnant women; maternal co-morbidities were not taken into consideration; risk factors were examined only during pregnancy; and preconception and time from conception to confirmation pregnancy was not accounted for. Here we compared a study that was done in Singapore in 2013 for Asian women. These are the results. Here we did a comparison of the Danish and Singapore study. So, same results were found for advanced maternal age, extremities of BMI, and previously diagnosed genital disease, as well as alcohol consumption. There were varied results for coffee consumption and smoking. In addition, the Singapore study found that women who had a partner of more than 41 years old had eight times higher risk of miscarriage. In conclusion, miscarriages can be prevented. There is an impetus for public education to protect these at-risk pregnancies. This study needs further support from future prospective cohort studies, as well. These are the references we have used. We would like to thank Prof TC Tan and Ms. Mabel Yap for their assistance. Thank you for your attention.

Leave a Reply

Your email address will not be published. Required fields are marked *