|
Abstract (English)
:
| Background. To evaluate the risk of spontaneous abortion in relation to maternal pre-pregnant underweight.
Methods. The study was designed as a cohort study within the framework of the Danish National Birth Cohort (DNBC). The participants were a total of 23 821 women recruited consecutively to the DNBC from October 1, 1997 to March 31, 1999 and interviewed subsequently. The pregnant women were recruited in first half of pregnancy and interviewed about pre-pregnant body size, obstetric history, exposures in pregnancy, and socio-demographic factors. Pregnancies were followed-up regarding spontaneous abortion. Relative risk of spontaneous abortion was calculated as Hazard Ratios using Cox regression with delayed entry.
Results. The outcome measure was spontaneous abortion. The hazard ratio for spontaneous abortion in women with a pre-pregnant body mass index (BMI) below 18.5 was 1.24 (95% confidence limits 0.95–1.63) compared to women with pre-pregnant BMI 18.5–24.9. Women with a BMI of 25 or more had a smaller increase in risk of spontaneous abortion. Adjustment for maternal age, parity, previous miscarriages, and lifestyle factors did not affect the estimates substantially, neither did exclusion of women with metabolic or eating disorders.
Conclusion. These results indicate that pre-pregnant underweight may affect the risk of spontaneous abortion negatively.
hr>
Spontaneous abortion is a frequent event, affecting approximately 15% of recognized pregnancies, and impossible to cure. Identification of risk factors for spontaneous abortion through epidemiological studies might lead to potential preventive measures for the condition. Over the recent years a number of studies on overweight and spontaneous abortion have been published, but the literature about maternal underweight as a risk factor is sparse.
Recently, various authors have described an association between low serum leptin levels and spontaneous abortion (1,2), and it has also been shown that women with low body mass index (BMI) have low serum leptin levels (3).
Although the prevalence of overweight women is increasing in most countries, the proportion of underweight women stays the same or increases, and the proportion of underweight young women in Denmark is approximately 4% (4). The objective of this study was to investigate the association between maternal pre-pregnant underweight and spontaneous abortion in a prospective design.
Materials and methods
Participating women
The study was carried out within the framework of the Danish National Birth Cohort (DNBC), a nation-wide study of pregnant women and their offspring (5). The women were recruited to the DNBC at the first antenatal visit to their general practitioner, which is scheduled to take place as early as possible after recognition of the pregnancy, and were enrolled in the DNBC if they returned a signed informed consent form. Participation implied to provide information on exposures in pregnancy by means of computer-assisted telephone interviews. The first interview covering background information and exposures in first trimester of pregnancy was scheduled to take place in pregnancy week 12–16, and if the woman had experienced a spontaneous abortion prior to completion of this interview she was offered a similar case interview.
Data for this particular study was interview information from these two types of interview from all women recruited for the first time to the DNBC between October 1, 1997 and March 31, 1999. The data material has been described in details elsewhere (6).
The data included information on self-reported pre-pregnant weight and height, maternal age, obstetrical history, and life-style factors.
Outcome measure of interest was spontaneous abortion, which according to Danish national definitions is a non-deliberate fetal death of an intrauterine pregnancy before 28 completed weeks of gestation.
All pregnancies were followed-up to end of pregnancy to identify pregnancy outcome by record linkage to the Civil Registration System, the National Discharge Registry and to information on outcome from the participating mothers.
Permissions from the Danish Data Protection Board and the National Scientific Ethics Committees were obtained before initiation of the study.
Statistical analyses
Self-reported pre-pregnant weight and height were used to calculate BMI (weight in kilograms divided by the square of the height in meters). BMI was categorized as BMI < 18.5, BMI 18.5–24.9, and BMI 25 + 0. Underweight was defined as BMI < 18.5 kg/m2.
The relative risk of spontaneous abortion was estimated as hazard ratios by using the Cox regression model, which allowed for delayed entry. The underlying time variable was gestational days and entry time was gestational age at recruitment to the DNBC.
The estimates were adjusted for physical exercise (0, 1–119, 120–239, 240–419, and 420+ min per week), smoking during pregnancy (0, 1–10, and 11+ cigarettes per day), alcohol consumption during pregnancy (0, -1, 2–3, and 4+ drinks per week), coffee consumption during pregnancy (0, 1–6, and 7+ cups per day), previous spontaneous abortions (0,1, 2, 3+), maternal age (−19, 20–24, 25–29, 30–34, 35–39, and 40+ years), number of previous births (0 or 1+), eating disorders (never, ever), metabolic disorders during pregnancy (yes, no), and maternal occupational groups (professionals, skilled workers, unskilled workers, economically inactive, students, and unclassifiable) and likelihood ratio tests were used for testing the significance of the covariates.
Results
A total of 23 821 women, recruited consecutively between October 1, 1997 and March 31, 1999, provided interview information. Of these, more than 2300 were recruited before end of pregnancy week 8 and more than two-thirds of the women were recruited before the end of pregnancy week 12. Follow-up regarding outcome of pregnancy was possible on 99.9% of the women, and 1039 of the interviewed women experienced a spontaneous abortion.
Table I shows the distribution of selected risk factors for spontaneous abortion according to maternal BMI groups. A total of 1132 women corresponding to 4.8% had a BMI below 18.5. Underweight women were likely to be younger, smokers, coffee drinkers, and having a history of eating disorders than women with a higher BMI.
The crude hazard ratios for spontaneous abortion was slightly elevated for underweight women [hazard ratio = 1.22 {95% confidence interval (CI) 0.94–1.60}] compared to normal-weight women, while overweight did not affect the hazard ratios. Maternal age, number of previous spontaneous abortion, alcohol and coffee consumption, level of physical activity, and having had an eating disorder were significantly associated with the risk of spontaneous abortion. However, adjustment for potential confounders did not change the estimates for underweight women considerably (Table II). The analyses revealed a significantly elevated hazard ratio at 1.66 (95% CI 1.32–2.09) for women, who reported having a history of an eating disorder, i.e. anorexia nervosa or bulimia. To further address the effect of maternal underweight in healthy women an analysis where women with metabolic and eating disorders were excluded was performed, and the relation between maternal underweight adjusted for significant covariates showed a hazard ratio of 1.20 (95% CI 0.89–1.62).
Discussion
In this study, we found the indication of an increased risk of spontaneous abortion in underweight women.
Several studies have reported an association between maternal obesity and fetal death (7,8), and Cnattingius et al. estimated on basis of their findings that if all women with a BMI > 20 reduced this to BMI under 20, the number of late fetal deaths would be reduced with 27%. In contrast to this, others have reported that maternal pre-pregnant BMI below 20 was associated with an increased rate of stillbirths and neonatal death (9). Other reproductive outcomes have been associated with maternal underweight: Cliver and colleagues (10) reported that small maternal size, especially low weight, increased the risk of a growth-retarded child with a factor 3, low maternal BMI, and low weight gain during pregnancy have been demonstrated to result in a higher risk of preterm birth (11), and other studies have shown that underweight women experience amenorrhea and infertility more frequently than women with a normal BMI (12–14). Furthermore, maternal underweight as a result of hunger or anorexia nervosa and bulimia has been shown to be associated with fetal loss (15–17).
The discrepancy between our finding and the finding reported by Cnattingius et al. may be due to the suggestion that the mechanisms responsible for early and for late fetal death differ.
Few studies have specifically dealt with the question of maternal underweight and spontaneous abortion. Kline and Stein (18) reported from a case-control study of spontaneous abortions from New York in which most of the abortuses were karyotyped that pre-pregnant BMI was reduced in women with chromosomally aberrant abortion compared to controls. The authors concluded, however, that the magnitude of the finding and the fact that it was not found for specific chromosomally aberrations weakened the finding.
Likewise, the increased risk of spontaneous abortion for underweight women found in our study is small and does not reach statistical significance. The estimated risk is, however, higher than the risk of abortion among overweight women, and the lack of significance is likely to be due to the small number of underweight women.
The participation rate in the DNCB has been estimated to be about 60% (5). Consequently, the prevalence of underweight in our study may differ from the prevalence in a representative population sample. The risk estimates found should not, however, be affected because non-participation in cohort studies is not considered a threat to validity of the true relationship between the exposure under study and the outcome (19,20).
The strengths of this study are the prospective design and the size of the study. The setting provides some assurance that confounding by social factors does not account for the association, because underweight in a Danish context is not likely to be related to socio-economic deprivation. All data on height, weight, and confounding factors were self-reported. According to Spencer et al. (21), self-reported weight and height data are valid for identifying relationships in epidemiological studies. But it should be noted that history of eating disorders – bulimia and anorexia nervosa – are exclusively self-reported and the validity of these data is not known.
The early recruitment to the DNBC study and the use of survival analysis with left truncation made it possible to assess the risk of spontaneous abortion in early pregnancy. We have a sufficient number of pregnancies under risk from week 6 and onwards, however, the data does not allow us to estimate how maternal underweight affects the pregnancy in the very first weeks of pregnancy.
The possible biological mechanisms behind an association between underweight and early fetal loss may be via down regulation of hormones or a direct consequence of undernutrition. An animal study suggested that the mildly hyperglycaemic and aminoacid-depleted maternal environment generated by undernutrition in rats may act as an early mechanism of programming and initiate conditions of metabolic stress, restricting early embryonic proliferation, and the generation of appropriately sized stem-cell lineages (22), which may lead to fetal demise. Abecia et al. (23) concluded that maternal undernutrition in ewes may produce a reduction of the ability of the embryos to secrete trophoblastic interferon-τ and therefore, an increase in the production of prostaglandin from the endometrium, which can initiate luteolysis.
The hormone leptin, primarily considered to affect appetite and weight regulation, has also been shown to be involved in a variety of processes affecting the reproduction. In mice, leptin has been shown to have a positive effect on reproduction and ovarian function, and findings in humans suggest that leptin may regulate uterine angiogenesis (24).
The findings of significantly lower serum leptin in women suffering from a spontaneous abortion in the first trimester than in both pregnant and non-pregnant women (1,2) may provide support to the results of this study, and these findings indicate one possible mechanism behind spontaneous abortion.
Conclusion
From this study, we conclude that there seems to be an association between maternal pre-pregnant underweight and the risk of a spontaneous abortion. The association, albeit weak and not statistically significant, was stronger than that found for overweight women. The study needs replication before any public health conclusions can be drawn.
Acknowledgments
The Danish National Research Foundation has established the Danish Epidemiology Science Center that initiated and created the DNBC. The cohort is furthermore a result of a major grant from this foundation. Additional support for the DNBC is obtained from the Pharmacy Foundation, the Egmont Foundation, the March of Dimes Birth Defects Foundation, and the Augustinus Foundation. This particular study was supported by a grant from the Augustinus Foundation and The Wedell-Wedellborg Foundation. Conflicts of interest: none.
|
![[Valid RSS]](images/icons/valid-rss-rogers.png "Validate my RSS feed"){kind=icon}
Hard 
