Higher Risk Of Miscarriage Research Articles

Risks of miscarriage and early preterm birth in trichorionic triplet pregnancies with embryo reduction versus expectant management: new data and systematic review

Triplet pregnancies are associated with a high risk of miscarriage and early preterm birth. It is uncertain if the outcome is improved by embryo reduction (ER). We examined trichorionic triplet pregnancies with three live fetuses at 10-14 weeks of gestation that were managed expectantly or by ER. The two groups were compared for the rates of miscarriage, defined as pregnancy loss before 24 weeks, and preterm delivery prior to 32 weeks. In addition, systematic searches were performed to identify studies comparing outcomes in expectant management versus ER in triplet pregnancies. We combined data from 365 pregnancies managed in our centre with those of five previous studies. In total there were 893 pregnancies. In the ER group (n=482) compared to the expectantly managed group (n=411), the rate of miscarriage was higher [8.1 versus 4.4%; relative risk (RR)=1.83, 95% confidence interval (CI)=1.08-3.16, P=0.036] and the rate of early preterm delivery was lower (10.4 versus 26.7%, RR=0.37, 95% CI=0.27-0.51, P<0.0001). It was calculated that seven (95% CI=5-9) reductions needed to be performed to prevent one early preterm delivery, while the number of reductions that would cause one miscarriage was 26 (95% CI=14-193). In trichorionic triplets, ER to twins is associated with an increase in the risk of subsequent miscarriage and decrease in risk of early preterm birth.

Reproductive outcome after chromosome analysis in couples with two or more miscarriages: case-control study

Objective To compare reproductive outcomes in couples carrying a structural chromosome abnormality and non-carrier couples referred for chromosome analysis after two or more miscarriages.Design Case-control study.Setting Six centres for clinical...

Does rural or urban residence make a difference to neonatal outcome in premature birth? A regional study in Australia

Background: Patients living in rural areas may be at a disadvantage in accessing tertiary health care. Aim: To test the hypothesis that very premature infants born to mothers residing in...

IN FOCUS: Factor V Leiden and prothrombin G20210A polymorphisms as risk factors for miscarriage during a first intended pregnancy: the matched case–control ‘NOHA first’ study

Factor V Leiden (FVL) and prothrombin G20210A (FIIG20210A) mutations are associated with a higher risk of miscarriage: we sought to understand whether this association differs by clinical time of unexplained miscarriage, and by ethnic origin, among women with no previous thrombotic episode, during the first intended pregnancy. We performed a case-control study nested in a cohort of 32 683 women. We analyzed 3496 pairs of women matched for classical confounding factors. The FVL and FIIG20210A mutations were associated with an increased risk of miscarriage in Caucasian women [odds ratio (OR) 3.19, 95% confidence interval (CI) 2.37-4.30, P < 0.001 and OR 2.36, 95% CI, 1.72-3.24, P < 0.001, respectively]. Among non-Caucasian women, the mutations were rare and the associations with risk of miscarriage less clear. FVL and FIIG20210A mutations were independent risk factors for miscarriages only for women with related clinical signs occurring from the 10th week of gestation on (OR 3.46, 95% CI 2.53-4.72, P < 0.001 and OR 2.60, 95% CI 1.86-3.64, P < 0.001, respectively). These results indicate that FVL and FIIG20210A mutations are associated with a significant risk of spontaneous abortion which clinical signs occur from the 10th week on of the first intended pregnancy.

Long-term use of oral contraceptives increases the risk of miscarriage

In this prospective, case-control study, a step-wise logistic regression model was used to examine the relationship between the consumption of oral contraceptives before pregnancy and the potential increase in the risk of miscarriage. Consumption of oral contraceptives for more than 2 years before pregnancy is associated with a higher risk of miscarriage.

Bed rest during pregnancy for preventing miscarriage

Miscarriage is pregnancy loss before 23 weeks of gestational age and it happens in 10% to 15% of pregnancies depending on maternal age and parity. It is associated with chromosomal defects in about a half or two thirds of cases. Many interventions have been used to prevent miscarriage but bed rest is probably the most commonly prescribed especially in cases of threatened miscarriage and history of previous miscarriage. Since the etiology of miscarriage in most of the cases is not related to an excess of activity, it is unlikely that bed rest could be an effective strategy to reduce spontaneous miscarriage. To evaluate the effect of prescription of bed rest during pregnancy to prevent miscarriage in women at high risk of miscarriage. We searched the Cochrane Pregnancy and Childbirth Group trials register (July 2004). In addition, we searched The Cochrane Central Register of Controlled Trials (The Cochrane Library), MEDLINE, POPLINE, LILACS and EMBASE. We included all published, unpublished and ongoing randomized trials with reported data which compare clinical outcomes in pregnant women who were prescribed bed rest in hospital or at home for preventing miscarriage compared with alternative care or no intervention. Two authors independently assessed the methodological quality of included trials using the methods described in the Cochrane Reviewers' Handbook. Studies were included irrespective of their methodological quality. Only two studies including 84 women were identified. There was no statistically significant difference in the risk of miscarriage in the bed rest group versus the no bed rest group (placebo or other treatment) (relative risk (RR) 1.54, 95% confidence interval (CI) 0.92 to 2.58). Neither bed rest in hospital nor bed rest at home showed a significant difference in the prevention of miscarriage. There was a higher risk of miscarriage in those women in the bed rest group than in those in the human chorionic gonadotrophin therapy group with no bed rest (RR 2.50, 95% CI 1.22 to 5.11). It seems that the small number of participants included in these studies is a main factor to make this analysis inconclusive. There is insufficient evidence of high quality that supports a policy of bed rest in order to prevent miscarriage in women with confirmed fetal viability and vaginal bleeding in first half of pregnancy.

Influence of Thyroid Autoimmunity and Maternal Age on the Risk of Miscarriage

Recently, studies have shown an association between antiperoxidase for the detection of thyroid autoimmunity (TAI) and abortion. Another point to be considered is the association of high risk of abortion and maternal age. The aim of the present study was to evaluate if the association between TAI and miscarriage could also be verified whether a population of unselected pregnant young women who normally present a low risk of miscarriage. We studied 534 pregnant women, by determining their serum thyroid antiperoxidase antibodies (TPO-Abs), thyrotropin (TSH) and free thyroxine (fT4) levels. Our end point was the pregnancy loss or live birth. Age ranged from 12 to 49 years (mean +/- S.D.; 23.5 +/- 5.9). Of 534 women, 29 (5.4%) were TPO-Ab positive. TSH levels were significantly higher in TPO-Ab-positive women compared with TPO-Ab negative women (median; 1.9 versus 1.1; P = 0.001). Elevated TSH levels were found in 13.8% (4 of 29) of the TPO-Ab-positive women compared with only 2.4% (12 of 505) in the TPO-Ab-negative women. There were no significant differences in fT4 levels in relation with autoimmunity and risk of miscarriage. The overall risk of miscarriage was 2.4% (13 of 534). Risk of miscarriage was significantly higher among women older than 35 years (7.7%), TPO-Ab positive (10.3%) and presenting high levels of TSH (12.5%). These factors remained independently associated with the risk of miscarriage in full multivariate analysis. We conclude that TAI is independently associated with is a higher risk of miscarriage in a population of unselected pregnant presenting a low risk of miscarriage.

Perinatal outcomes of teenage pregnancies according to gravidity and obstetric history

Purpose Teenage pregnancy is a significant public health issue. The aim of this study was to compare the perinatal outcomes of teenage primigravidae and secundigravidae who previously had an induced abortion or a live-birth. Methods In this population-based retrospective cohort study, we used routinely collected perinatal data from 1990 to 1999 from Lower Saxony, Germany. The main outcomes were perinatal and neonatal mortality, stillbirth, preterm birth, and very low birth weight (VLBW). There were 10,354 births among teenagers (12–18 years). After exclusion of girls with multiple births, two or more previous pregnancies, a previous spontaneous abortion, or a previous stillbirth, the remaining 8669 subjects were stratified into primigravidae (n = 7695) and secundigravidae one previous live-birth (n = 767) or one previous induced abortion (n = 207). Bivariate and multivariable logistic regression analyses (adjusted for smoking status, nationality, partner status, and maternal height and weight at first and last antenatal visit) were conducted to calculate odds ratios (OR) with 95% confidence intervals (CI). Results Compared with primigravidae (referent group), teenagers with a second live-birth had higher risks of perinatal mortality (OR = 2.11, 95% CI = 1.13, 3.95) and of neonatal mortality (OR = 4.30, 95% CI = 1.77, 10.49) and teens with a previous induced abortion had higher risk of stillbirth (OR = 3.41, 95% CI = 1.04, 11.19) and preterm birth (OR = 2.21, 95% CI = 1.07, 4.58) in crude analyses. Adjusted for all confounders, teens with a second live-birth were more likely to experience perinatal mortality (OR = 2.32, 95% CI = 1.12, 4.81) and neonatal mortality (OR = 4.67, 95% CI = 1.60, 13.66), and teenagers with a previous induced abortion had a higher risk of VLBW (OR = 2.90, 95% CI = 1.10, 7.69). Conclusion Among teens, the second pregnancy bears greater risks than the first one. The prevention of the second pregnancy during adolescence should be paramount.

THE CONTRIBUTION OF MALARIA IN PREGNANCY TO PERINATAL MORTALITY

The link between malaria and perinatal mortality was explored by systematically reviewing 117 studies published between 1948 and 2002. The mean perinatal mortality rate was higher in malaria endemic countries (61.1/1,000, 95% confidence interval [CI] = 52.1-70.1) than in non-endemic countries (25.8/1,000, 95% CI = 21.1-30.6). Similarly, the fetal mortality rate was higher in endemic countries (40.1/1,000, 95% CI = 32.1-48.0) than in non-endemic countries (20.0/1,000, 95% CI = 13.2-26.8) countries. Considering that perinatal mortality is an important indicator of obstetric care quality and socioeconomic development, further analysis was restricted to countries with a human development index between 500 and 800. In this category, the perinatal mortality rate was also significantly higher in endemic countries (50.5/1,000, 95% CI = 35.5-65.5) than in non-endemic countries (30.0/1,000, 95% CI = 25.7-34.3). In some publications, the occurrence of placental malaria and stillbirth was available. Placental malaria was significantly associated with a higher risk for stillbirth, regardless of parity (odds ratio = 2.19, 95% CI = 1.49-3.22, P < 0.001). Despite the limitations involved in this kind of review, all information found indicates that in endemic countries, malaria is an important determinant of perinatal mortality. Preventive measures such as intermittent preventive treatment or insecticide-treated bed nets could substantially reduce perinatal mortality and fetal wastage.

Risks of Stillbirth and Early Neonatal Death by Day of Week

Unless labor is timed for high-risk pregnancies or there are differences in the quality of perinatal care, births on different days of the week should have similar outcomes. Not many such studies have been done, but all of them point to a higher risk of stillbirth and/or early neonatal death for infants born on weekend days rather than on weekdays. The investigators, using a Canadian database of 3,239,972 births taking place in the years 1985 through 1998, sought associations between these outcomes and the day of the week at birth. These women represented all births recorded in Canada during the period under review except for Ontario. Births were 24% less frequent on weekend days than during the week. Stillbirth rates were lowest on Mondays and highest on Saturdays. Neonatal deaths were considerably more frequent among infants born on Saturdays and Sundays (Fig. 1), particularly those caused by asphyxia. These patterns were consistent throughout the review period. Cause-specific excess risks for stillbirth varied widely. Stillbirths resulting from congenital anomalies were much less frequent on weekend days than on weekdays. Cause-specific excess risks for early neonatal death also varied widely, ranging from +28% for asphyxia and immaturity to -3% for congenital anomalies. On logistic regression analysis, controlling for completed weeks of gestational age, there was no apparent excess risk of either stillbirth or early neonatal death on weekends. For preterm infants, the gestational age-specific risks of early neonatal death were similar or slightly lower for infants born on weekend days. Any excess risk of stillbirth or early neonatal death on weekend days in this Canadian study was much smaller than has been reported from other countries. The present findings suggest that, in Canada, accessibility to quality obstetric and neonatal care is maintained on weekends.

Brucellosis

Brucellosis

Miscarriage rate is not increased in polycystic ovarian syndrome patients compared with other infertility diagnoses

Objective: To evaluate the risk of clinical pregnancy loss in patients with a diagnosis of polycystic ovary syndrome (PCOS) compared with other infertility diagnoses. Design: Retrospective weight- and age-matched cohort study in a tertiary REI teaching center. Materials and Methods: 486 consecutive infertility patients having a positive pregnancy test were evaluated. Patients diagnosed with PCOS were oligo-anovulatory, without ovulatory progesterone, and had evidence of either hyperandrogenism or hyperandrogenemia. All had polycystic appearing ovaries on endovaginal sonography. Comparison patients were ovulatory and had one or more diagnoses of infertility including unexplained infertility, mild male factor, pelvic factor, hypogonadotropic hypgonadism or mildly diminished ovarian reserve. All patients had at least one patent tube. Patients with unexplained infertility or total motile inseminates between 5 and 20 million/ml underwent washed intrauterine insemination following gradient separation. The most frequent intervention in both groups was gonadotropin therapy. Pregnancy was diagnosed by a positive serum βhCG 14 days following ovulation. Pregnancy outcome was determined by reviewing medical records. To control for the effects of weight and age, each PCOS miscarriage patient was matched with the next three consecutive PCOS delivered patients of the same age (within 1 year), and the matching process repeated for weight (within 5 pounds). SAS statistical analysis software was used. Chi-square evaluated classification versus outcome status and independent and matched t-tests were used for parametric data. Significance for p <0.05 was tested using two-tailed distributions. Results: Pregnancy results were available for all patients. The overall rate of early miscarriage in our non-PCOS infertility pregnancy group was 26.0%. In the 139 PCOS patient cohort, the miscarriage rate was 20.0% (p=n.s.). In our PCOS cohort, the mean age of the delivered group was 32.8 years vs. 35.2 years in the miscarriage group (p = 0.01). The mean weight in the PCOS delivered group did not differ from the PCOS miscarriage group (162.7 vs. 166.4 lbs, p=n.s.). However, among non-PCOS patients, a 6.4 lb. difference between miscarriage patients (151.7 lbs) and delivered patients (145.3 lbs) was significant (p < 0.05). When each PCOS patient with early miscarriage was age-matched with the next three consecutive PCOS delivered patients the difference in weight was 0.2 lbs (p = n.s.). In a similar weight-matched analysis of PCOS patients, the difference in age between early miscarriage and delivered patients was 1.2 years (p = n.s.). Examining separately those 15% of PCOS patients who underwent long protocol downregulation with subsequent luteal phase support for IVF, these similarities persisted. Serum total Testosterone showed no association with pregnancy outcome. Conclusion: Despite the perception that PCOS patients are at higher risk for miscarriage, these data suggest that treated PCOS patients when matched for age and weight with patients having other infertility diagnoses do not experience increased clinical pregnancy loss.

Outcome of twin pregnancies with complete hydatidiform mole and healthy co-twin

We assessed 77 twin pregnancies, comprising complete hydatidiform mole (CHM) and healthy co-twin, to ascertain the risks to the mother and baby of continuing the pregnancy, versus termination. 24 women with histologically confirmed CHM and healthy co-twin pregnancies decided to have a termination. 53 women continued with their pregnancies, though two had to have terminations because of severe pre-eclampsia, and 23 spontaneously aborted (<24 weeks' gestation). 28 pregnancies lasted 24 weeks or more, resulting in 20 livebirths. Chemotherapy to eliminate persistent gestational trophoblastic disease (pGTD) was required in three of 19 women (16%; 95% CI 3-39) who terminated their pregnancies in the first trimester, and in 12 of 58 (21%; 95% CI 11-33%) who continued their pregnancies. CHM and healthy co-twin pregnancies have a high risk of spontaneous abortion, but about 40% result in livebirths, without significantly increasing the risk of pGTD.

The differentials and determinants of perinatal mortality in rural Bangladesh

Objective: In Bangladesh, the perinatal mortality is very high. This study examined the differentials and determinants of perinatal mortality in rural Bangladesh.Methods: The study wasbased on the prospective data on maternal morbidity collected by the Bangladesh Institute of Research for Promotion of Essential and Reproductive Health and Technologies (BIRPERHT). This prospective studywas conducted during the period November 1992 to December 1993. In this study, the factors associated with perinatal deaths were examined, employing differential and multivariate analyses.Results:It was found that assisted delivery caused higher risks of stillbirth and death among live births during the first week of life. It was also observed that five or more pregnancies, prior to the index pregnancy,were positively associated with perinatal death.Conclusion: If the delivery is assisted or there are complications in delivery, then it is likely that the incidence of perinatal mortality willincrease sharply. If the newborn baby was given colostrum, then perinatal mortality decreased to a great extent.

Polycystic ovarian syndrome and the risk of spontaneous abortion following assisted reproductive technology treatment.

A high proportion of infertile patients have polycystic ovarian syndrome (PCOS) with a reportedly greater risk of spontaneous abortion. Because of the close link between PCOS and obesity and the independent association of obesity with poor pregnancy outcomes, it is important to distinguish the possible confounding effect of body mass index (BMI) or other variables from that of PCOS. This study aims to determine the effect of PCOS status on the risk of spontaneous abortion with adjustment for body mass and several other confounding factors in a large cohort of pregnant infertile women. The patients (n = 1018) were treated in a tertiary infertility centre. Their PCOS status was determined by standard criteria and their BMI had been taken less than 1 year before the pregnancy. Patients whose PCOS status or BMI measurements were not assessed were excluded. Student's t-test or chi2 test were used to test the difference between the PCOS and non-PCOS groups while a multivariate logistical regression model was used to assess the effect of PCOS, BMI and other confounding factors. Overall, the incidence of PCOS was 37% in this cohort. The overall incidence of spontaneous abortion in the study population was 21%. Univariate analysis showed that women with PCOS had a significantly greater risk of spontaneous abortion compared with non-PCOS women (25 versus 18%, P < 0.01). However, using multivariate logistic regression analysis this effect was reduced to a non-significant level [odds ratio (OR) = 1.10, 95% confidence interval (CI) 0.85-1.36] after adjusting for obesity and patients/treatment combination factor, and to nil after adjusting for all confounding factors considered in this study (OR = 0.98, 95% CI 0.75-1.28). The results of this study suggest that the higher risk of spontaneous abortion observed in women with PCOS is likely to be due to their high prevalence of obesity and the type of treatment they receive.

Maternal/Paternal Sharing of DQ-Alpha Type II Histocompatibility Antigens Not Associated With Pregnancy Outcome Following In Vitro Fertilization (IVF)-Embryo Transfer (ET).

The mechanism of how the fetal semi-allograft escapes maternal immune surveillance still remains a mystery. Absence of cytotoxic T-cell response against the fetus has been explained by the absence of major histocompatibility (MHC) type I antigens in the placenta (cytotoxic T-cell response requires presentation of the antigen by MHC-I antigens). However, some researchers believe that immune protective responses are induced by differences in some MHC-II antigens especially in the DQ-alpha region. One theory is that differences in DQ-alpha antigens result in induction of progesterone (P) receptors in gamma/delta T-cells, which, when exposed to high concentrations of P (concentrations only generated at the maternal-fetal interface) lead to the secretion by these cells of a protein called progesterone induced blocking factor (PIBF). PIBF, in turn, inhibits natural killer (NK) cell cytolysis and causes a shift from TH1 to TH2 cytokine dominance. One hypothesis is that by sharing DQ-alpha antigens, PIBF is not adequately secreted because of failure to induce adequate P receptors in gamma/delta T cells leading to a higher risk of spontaneous abortion, or may cause implantation failure following IVF-ET. Thirty-one couples undergoing in vitro fertilization-embryo transfer were evaluated for sharing of DQ-alpha antigens. There were 16 who shared a DQ-alpha antigen and 15 who did not. Clinical pregnancies were achieved in 7 (43.7%) of the ones sharing antigens and 6 (40.0%) of those who did not. Spontaneous abortions were found in 0 of the former and 1 of the latter. There were 7 patients sharing the 4.1 antigens and 3 conceived, and 8 shared a 1.1, 1.2, or 1.3 antigen and 4 conceived. These data do not show that sharing of DQ-alpha antigens is related to poor IVF outcome. Some recent data suggests that perhaps differences in an MHC-I antigen, HLA-E, may be the critical antigen. Similar studies will be needed to evaluate HLA-E differences and conception outcome.

ICSI outcome in patients of 40 years age and over: a retrospective analysis

Objective: To report and analyse our experience with ICSI treatment in infertile women ≥40 years of age, with the intention of contributing to current debates on the effect of aging on the reproductive potential. Study design: 107 infertile couples in which the female partner was aged ≥40 years and who received ICSI treatment between January 1996 and December 1998. Results: A total of 107 women underwent 171 treatment cycles during this period. Of 171 cycles initiated 33 were cancelled (cancellation rate=19.3%). In this way, 17 women did not have embryo transfer at all, while 90 patients had 138 cycles with oocyte retrieval and successful embryo transfer, with a mean number of embryos per transfer 2.36. Sixteen pregnancies occurred and eight of them ended in spontaneous abortion. The implantation rate was 4.9%, the pregnancy rate per initiated cycle was 9.35% and per transfer cycle 11.59%. The miscarriage rate was 50%. Moreover, 12 patients had supernumerary embryos, that were cryopreserved and transferred in 17 thawing cycles and resulted in two pregnancies ending in abortion. All pregnancies occurred when three embryos were available, except in two cases with two available embryos. The great majority of the total pregnancies (16 of 18) resulting in women aged between 40 and 42 years. Conclusion: Our data show that women 40 and older with existing ovarian function may benefit from ICSI treatment, even when the indication for treatment is male factor infertility. Supernumerary embryos, that are cryopreserved and transferred in subsequent cycles can improve the overall pregnancy rates per oocyte retrieval, although these women should be aware of the very high risk of miscarriage.

The Feminism &amp; Psychology Undergraduate Prize 1999: Prizewinning Entry

Polycystic Ovarian Syndrome (PCOS),1 first reported by Stein and Leventhal (1935), is a heterogeneous condition (not to be confused with large cysts on the ovaries) characterized by infertility or higher risk of miscarriage, hirsutism (‘excess’ hair growth), amenorrhea or oligomenorrhea (irregular or no menstrual cycle), menorrhagia (excessive menstrual bleeding), anovulation, weight gain or obesity, acne vulgaris, androgenic alopecia (male pattern hair loss), insulin resistance and excess androgen production (Adams et al., 1986; Ginsburg and Havard, 1976). It is estimated that up to 20 percent of women (Polson et al., 1988) may be affected by PCOS, but not all display symptoms, present for diagnosis or are accurately diagnosed. Outside the realm of gynaecology, public knowledge of PCOS is limited, and despite a far-reaching concern with women’s health issues, feminist awareness of, or activism in relation to, the condition appears to be minimal. The New Our Bodies, Ourselves (Phillips and Rakusen, 1996) fails to mention it, and the Spare Rib Reader on women’s health (Balsdon and Kaluzynsha,1987) dedicates just seven lines to the condition and its symptoms. Treatment of PCOS is largely aimed at reducing weight, restoring fertility and ameliorating hirsutism (Bruce-Jones et al., 1993). Women with PCOS may experience considerable stress related to their symptoms. In a study of hirsute women, 68 percent avoided a number of social situations (Lobo et al., 1985). Hirsutism may also breed a sense of shame or freakishness and fear of exposure (Money and Clopper, 1974). The authors report the primary concern of girls and women experiencing hirsutism is restoration of a female (i.e. hair-free) appearance. From a feminist perspective, the problem is

Fetal Loss Rate After Second Trimester Amniocentesis at Different Gestational Age

Background Recent studies have debated the safety of early amniocentesis, one of which reported a significant increase in spontaneous fetal losses following amniocentesis performed at 10-13 weeks of gestation. The aim of this study was to determine the fetal loss rate following amniocentesis performed at different stages in the second trimester. Methods One thousand six hundred and fifty-one consecutive amniocenteses undertaken in a low risk population between 13 weeks+0 days and 20 weeks+6 days were evaluated. In 87% of the women the reason for testing was maternal age (mean 37.1 years). Data concerning patients, procedure details and pregnancy outcome were recorded. Results Spontaneous abortion before 28 weeks of gestation occurred in 1.1%, one third of which were within two weeks and the remaining within seven weeks after the procedure. The fetal loss rate was higher when the amniocentesis was performed earlier in pregnancy, at 13 weeks+0 days exceeding 3%. Abnormal color of the amniotic fluid and leakage of amniotic fluid were strong predictors of fetal loss. Transplacental needle insertion did not increase the risk of pregnancy loss. Conclusion Amniocentesis performed very early in the second trimester is followed by an increased fetal loss rate that cannot be explained solely by a higher risk of spontaneous abortion at this time of gestation.

Drinking water source and spontaneous abortion: A cross-sectional study in a rural Chinese population.

The authors examined the association between the risk of spontaneous abortion and the type of drinking water source in a rural Chinese population. Information about pregnancy outcomes and various exposures was collected by means of a detailed interview questionnaire. 2,876 pregnancies occurring in 1989-1993 among 2,201 nonsmoking and non-alcohol-drinking women were analyzed by logistic regression. Generalized estimating equations were used to adjust for correlations between multiple pregnancies in the same woman. The use of pond water as a drinking water source was associated with a higher risk of spontaneous abortion than the use of well or river water (odds ratio, 1.63; 95% confidence interval, 1.11 to 2.39), adjusting for potential confounders. More studies are needed to identify the agents responsible for the observed association. The results of this study emphasize the importance of monitoring potential drinking-water contamination in communities.