George L, Mills JL, Johansson ALV, Nordmark A, Olander B, Granath F, Cnattingius S. Plasma Folate Levels and Risk of Spontaneous Abortion. JAMA. 2002;288(15):1867-1873. doi:10.1001/jama.288.15.1867
Author Affiliations: Department of Medical Epidemiology (Drs George, Granath, and Cnattingius, and Ms Johansson) and Department of Medical Laboratory Sciences and Technology (Ms Nordmark), Huddinge University Hospital, Karolinska Institutet, Stockholm, Sweden; Pediatric Epidemiology Section, Division of Epidemiology, Statistics and Prevention Research, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, Md (Dr Mills); and Department of Clinical Chemistry, Karolinska Laboratoriet, Danderyds Hospital, Stockholm (Dr Olander).
Context Both folate deficiency and folic acid supplements have been reported
to increase the risk of spontaneous abortion. The results are inconclusive,
however, and measurements of folate have not been available in all studies.
Objective To study the association between plasma folate levels and the risk of
Design, Setting, and Population Population-based, matched, case-control study of case women with spontaneous
abortion and control women from January 1996 through December 1998 in Uppsala
County, Sweden. Plasma folate measurements were available for 468 cases and
921 controls at 6 to 12 gestational weeks.
Main Outcome Measure Risk of spontaneous abortion vs maternal plasma folate level.
Results Compared with women with plasma folate levels between 2.20 and 3.95
ng/mL (5.0 and 8.9 nmol/L), women with low (≤2.19 ng/mL [≤4.9 nmol/L])
folate levels were at increased risk of spontaneous abortion (adjusted odds
ratio [OR], 1.47; 95% confidence interval [CI], 1.01-2.14), whereas women
with higher folate levels (3.96-6.16 ng/mL [9.0-13.9 nmol/L] and ≥6.17
ng/mL [≥14.0 nmol/L]) showed no increased risk of spontaneous abortion
(OR, 0.84; 95% CI, 0.59-1.20; and OR, 0.74; 95% CI, 0.47-1.16, respectively).
Low folate levels were associated with a significantly increased risk when
the fetal karyotype was abnormal (OR, 1.95; 95% CI, 1.09-3.48) but not when
the fetal karyotype was normal (OR, 1.11; 95% CI, 0.55-2.24) or unknown (OR,
1.45; 95% CI, 0.90-2.33).
Conclusion Low plasma folate levels were associated with an increased risk of early
Some studies,1- 5 but
not all,6- 9 have
found that folate deficiency is a risk factor for spontaneous abortion. These
conflicting results could be due to small sample size, highly selected populations,
or lack of control for potential confounders, such as age, smoking, and alcohol
consumption. Moreover, folate levels change during pregnancy,10 and
most studies have included nonpregnant women as controls.
Conversely, it has also been suggested that folic acid supplementation,
by increasing folate levels, may increase the rate of or shift the timing
of spontaneous abortion.11- 17 This
has been highly controversial, with other studies challenging these findings.18- 21 None
of these studies have included measurements of folate levels.
Folic acid supplementation has been reported to prevent neural tube
defects.22- 24 Since
1998, the US Food and Drug Administration has required food fortification
with folic acid,25 whereas in Sweden, folic
acid fortification has not been introduced. The possibility of studying the
relationship between folate deficiency and spontaneous abortion is hampered
in populations with folic acid food fortification,26 whereas
the ability to study high exposure is enhanced. In this large, population-based,
case-control study in Sweden, we investigated the association between levels
of plasma folate and risk of early spontaneous abortion after controlling
for possible confounders.
The methods of this matched, case-control study are described in detail
elsewhere.27 The study was conducted in Uppsala
County, Sweden, from January 1996 through December 1998. Women who had spontaneous
abortion of a fetus with a gestational age of 6 to 12 completed weeks and
whose pregnancies had been confirmed by a positive human chorionic gonadotropin
test result were identified at the Department of Obstetrics and Gynecology
of Uppsala University Hospital, which is the only place in the county for
the care of women with spontaneous abortions. During this period, we identified
652 women as potential case patients, of whom 562 (86%) agreed to participate.
Control patients were primarily selected from pregnant women seeking
prenatal care in Uppsala County and were frequency-matched to the cases by
gestational week. Among cases and controls, gestational age was calculated
from the first day of the last menstrual period. Of 1037 patients receiving
antenatal care and asked to participate as controls, 953 (92%) agreed. All
potential controls underwent vaginal ultrasonography before the interview
to verify the viability of the fetus. If a nonviable intrauterine pregnancy
was detected, the woman was recruited as a case patient. This occurred in
53 of the case patients. Of women recruited as controls, 5 women had a spontaneous
abortion before 13 completed weeks. Since these women now fulfilled the criteria
of being recruited as cases, they were thus also included as such and subjected
to a new interview.
In Uppsala County, there are approximately 3 legally induced abortions
for every 10 completed pregnancies, and some of these terminated pregnancies
would have resulted in spontaneous abortion if the pregnancy had continued.
Women with induced abortions may differ from women continuing their pregnancies
in terms of factors associated with risk of spontaneous abortions, such as
age, smoking, and possibly other lifestyle-related factors. To avoid this
potential bias in the selection of control patients, women planning to have
induced abortions were added to the control group. In total, 310 women who
would later undergo induced abortion were asked to participate and 274 (88%)
agreed. Of these women, 75 were added to the control group, a number estimated
according to the gestational age distribution of induced abortions in Uppsala
County during the study period.
Three specially trained midwives conducted in-person interviews with
the cases and the controls recruited among women receiving antenatal care.
Two physicians conducted interviews with the control patients who would undergo
induced abortions. Ninety percent of the cases were interviewed within 2 weeks
after their miscarriage diagnosis, and the other 10% were interviewed within
7 weeks. All controls were interviewed within 6 days after their last completed
week of gestation used in matching. A structured questionnaire was used to
reduce bias because the interviewers could not be blinded to status (ie, pregnant
or aborted). Interviewers did not know folate status or the results of karyotyping.
Women were asked about possible risk factors for spontaneous abortion,
including age, previous reproductive history, smoking, caffeine and alcohol
use, shift work, and a number of other potential risk factors. Women were
also asked how many days each week they used prescription or nonprescription
drugs, starting 4 weeks before their last menstrual period and ending in their
last completed week of gestation. Daily folic acid supplementation was defined
as intake of at least 400 µg/d during this period.28 Women
were asked to quantify mean daily intake of various caffeine sources on a
weekly basis, which was used to calculate average daily caffeine intake during
pregnancy (ie, from conception through the last completed week of pregnancy).
Since pregnancy symptoms are associated with fetal viability and caffeine
consumption, we determined scores for the presence of nausea (0 [never], 1
[sometimes but not daily], 2 [daily but not all day], 3 [daily and all day]);
vomiting (0 [never], 1 [sometimes but not daily], 2 [daily]); and fatigue
(0 [no], 1 [yes, but unchanged sleeping habits], 2 [yes, slightly changed
sleeping habits], 3 [yes, pronounced change in sleeping habits]). The average
weekly score during pregnancy for each symptom was then calculated and averaged
to a mean score for each symptom.27
Oral informed consent was obtained from all the women, and the ethics
committee of the medical faculty at Uppsala University approved the study.
Patients were asked to provide blood samples. From the cases, blood
samples were obtained at the emergency ward at the time of miscarriage diagnosis
and from the controls at the time they were interviewed. Blood samples were
kept frozen at −80°C until assayed. A plasma blood sample for analysis
of folic acid and cotinine concentrations was available from 468 (83%) of
the cases, 852 (89%) of the controls from antenatal care, and 69 (92%) of
the controls with induced abortions. There were virtually no differences in
baseline maternal characteristics and exposures between the cases and the
controls who agreed to participate in the study and the cases and the controls
from whom blood samples for folic acid analysis were available (data available
on request from the authors).
Folate analyses were performed with an immunoassay analyzer (AxSYM,
Abbot Laboratories, Abbot Park, Ill), using ion capture reaction technology.
The analysis is quantified by measuring the amount of unoccupied folate-specific
binding sites bound to matrix using a conjugate of pteroic acid (a folate
analog) and alkaline phosphatase as the signal-generating molecule, and a
substrate, 4-methylumbelliferyl phosphate. Plasma folate levels were categorized
before data analyses. Low plasma folate level was defined as 2.19 ng/mL (4.9
nmol/L) or lower, which corresponds to the cutoff for recommending folic acid
supplements.29 Because there is no standard
definition of high serum folate levels in Sweden, the reference group was
defined as women with levels of 2.20 to 3.95 ng/mL (5.0-8.9 nmol/L) (including
most cases and controls), and high folate levels were stratified into 3.96
to 6.16 ng/mL (9.0-13.9 nmol/L) and 6.17 ng/mL (14.0 nmol/L) or more (including
approximately 20% and 10% of cases and controls, respectively).
Plasma cotinine was measured by gas chromatography with use of N-ethylnorcotinine as an internal standard.30 We
defined smokers as women who had a plasma cotinine concentration of higher
than 15 ng/mL (85.2 nmol/L).31 We allowed self-reported
daily smoking during all weeks of pregnancy to override missing cotinine values
for classification of smoking.
Curettage was performed in all cases with incomplete spontaneous abortion
at diagnosis. Intrauterine tissue was collected under sterile conditions and
routinely sent for karyotype analyses. Chorionic villi were identified in
tissue obtained by curettage in 243 of the 468 cases, allowing successful
karyotyping in 216 (89%). Cytogenetic analysis was performed using direct
preparation, and the chromosomes were banded with Giemsa stain. Eleven cells
in metaphase were routinely analyzed, and karyotyping was considered unsuccessful
if fewer than 3 cells in metaphase were obtained.
Data were analyzed with the use of conditional logistic regression analysis
using SAS PROC PHREG.32 The controls were frequency
matched to the cases by week of gestation. Odds ratios (ORs) with 95% confidence
intervals (CIs) were used to estimate the relative risk. The overall effect
of folate was tested by using a Wald χ2 test. The test considers
all strata in determining significance and not just pairwise comparisons with
the reference group. Since the study was frequency matched, all controls were
considered in the subanalyses of risks of spontaneous abortion according to
fetal karyotype and gestational age. Variables were included in the multivariate
analyses (maternal age, cigarette smoking, average caffeine intake during
pregnancy, previous spontaneous abortions, education, parity, country of birth,
body mass index, change of eating habits, and the pregnancy symptoms nausea,
vomiting, and fatigue) if they were judged a priori to be potential confounders
and if they were associated with risk of miscarriage, with low or high folic
acid levels, or with both (P<.05). A χ2 test was used to test homogeneity of folate level proportions and
folic acid supplement proportions.
The mean (SD) serum folate values among cases and controls were 3.76
(2.41) ng/mL (8.54 [5.48] nmol/L) and 3.84 (2.29) ng/mL (8.71 [5.20] nmol/L),
respectively (P = .56). Compared with folate levels
of 2.20 to 3.95 ng/mL (5.0-8.9 nmol/L), low (≤2.19 ng/mL [≤4.9 nmol/L])
folate levels were associated with a nonsignificant increase in risk of spontaneous
abortion in the univariate analysis. High folate level was not associated
with increased risk of spontaneous abortion (Table 1a). The women with spontaneous abortion were older than the
controls and were more likely to previously have had spontaneous abortions
and births, to be born outside the Nordic countries, to have medium to high
education, to be smokers, and to consume more coffee and alcohol. Pregnancy
symptoms, such as nausea, vomiting, tiredness, change in eating habits, and
aversion to food and coffee, were more prevalent and severe among the controls.
There were no significant differences between cases and controls with respect
to body mass index, shift work, or age at menarche.
Many possible risk factors for spontaneous abortion were also associated
with plasma folate levels. The proportion of women with low (≤2.19 ng/mL
[≤4.9 nmol/L]) folate levels was increased among women who were younger
than 30 years, had less than 12 years' education (P<.001
for all), were smokers, were obese (body mass index ≥30 kg/m2)
(P = .005 for all), and were born within the Nordic
countries (P = .027) and increased nonsignificantly
(P = .10) with caffeine consumption. The proportion
of women with high (≥6.17 ng/mL [≥14.0 nmol/L]) folate levels was increased
among women aged 30 years or older, who were nulliparous (P<.001 for all), had more than 12 years' education (P = .002), or were born outside the Nordic countries (P = .003). Other variables, including gestational age, did not influence
plasma folate levels (data available on request from the authors).
After adjustment, low folate levels were significantly associated with
an increase in risk of spontaneous abortion (OR, 1.47; 95% CI, 1.01-2.14)
(Table 2). When we only adjusted
for maternal age, the corresponding OR was 1.45 (95% CI, 1.04-2.01), suggesting
that the increase in risk was mainly due to elimination of confounding by
maternal age. Because there was no significant interaction between maternal
age and folate levels with regard to the risk of spontaneous abortion (P = .09), we did not perform analyses of plasma folate
values stratified by maternal age. The risk of spontaneous abortion was also
increased among women aged 35 years or older, among women with previous spontaneous
abortions, among smokers, and among women consuming at least 300 mg/d of caffeine.
Compared with women in the reference group, women with high folate levels
had a nonsignificant decrease in their risk of spontaneous abortion (Table 2).
In analyses stratified by gestational age (Table 3), low folate levels were not associated with a significant
increase in risk of spontaneous abortion at 6 to 8 gestational weeks, at 9
to 10 weeks, or at 11 to 12 weeks. In addition, there were no significant
interactions between plasma folate levels and maternal age, smoking, caffeine
intake, or gestational age (data available on request from the authors).
Karyotype analysis revealed that 83 fetuses (50 male and 33 female)
were chromosomally normal, and 133 were abnormal. Of spontaneous abortions
with abnormal fetal karyotype, 88 had autosomal aberrations (84 trisomies
and 4 double trisomies); 14 were 45,X; 8 were 47,XXY; 4 were 47,XXX; 16 were
triploidy; and 6 were tetraploidy. When we performed analyses stratified by
fetal karyotype, the overall tests of an association between folate levels
and karyotype-specific abortions were statistically nonsignificant (Table 3). Compared with women in the reference
group, women with low folate levels were at increased risk of spontaneous
abortion of fetuses with abnormal karyotype, whereas there were no significant
associations with risk of normal and unknown fetal karyotype abortion. To
further investigate the association between plasma folate levels and risk
of spontaneous abortion with abnormal fetal karyotype, we stratified the abnormal
fetal karyotypes into lethal (triploidy, tetraploidy, and autosomal except
trisomy 21) and others (trisomy 21; 45,X; 47,XXY; and 47,XXX). Among women
with low (≤2.19 ng/mL [≤4.9 nmol/L]) plasma folate levels who had spontaneous
abortions with abnormal fetal karyotype, 19 (79%) of 24 had a fetus with a
lethal chromosomal abnormality. Similarly, among women with plasma folate
levels of 2.21 ng/mL (5.0 nmol/L) or more, 86 (79%) of 109 had fetuses assessed
as having lethal chromosomal abnormality.
Since the pairwise comparison between the reference category and the
low folate level group is significant for spontaneous abortions with abnormal
fetal karyotype but not for those with normal fetal karyotype, we also performed
an analysis restricted to cases. This case-case study does not yield a significant
difference in folate levels between abnormal and normal cases (P = .38). Thus, we cannot show a significant difference in folate-associated
risk patterns for normal and abnormal karyotype abortions.
In all, 4.7% (30 cases and 35 controls) of the women in the study reported
daily intake of at least 400 µg of a folic acid supplement, beginning
at a minimum of 4 weeks before conception and continuing through the last
completed gestational week. Compared with women not taking a supplement, women
taking a daily supplement were significantly older (P =
.003) and had significantly less nausea (P = .04).
Among women taking a folic acid supplement, none had low (≤2.19 ng/mL [≤4.9
nmol/L]) plasma folate levels and 91% (n = 59) had values of 3.96 ng/mL (9.0
nmol/L) or higher, whereas among those not taking a folic acid supplement,
16% (n = 206) had low levels and 27% (n = 352) had values of 3.96 ng/mL (9.0
nmol/L) or more. Compared with women not taking a folic acid supplement, the
risk of spontaneous abortion among women using a folic acid supplement was
not significantly increased (adjusted OR, 1.3; 95% CI, 0.7-2.4).
The distribution of gestational age at abortion among cases differed
between those taking and not taking a folic acid supplement. Among cases taking
daily folic acid supplements (n = 30), 6 (20%) had a spontaneous abortion
in weeks 6 to 8, 19 (63%) in weeks 9 to 10, and 5 (17%) in weeks 11 to 12,
whereas among cases not taking daily folate supplements (n = 438), 127 (29%)
had a spontaneous abortion in weeks 6 to 8, 163 (37%) in weeks 9 to 10, and
148 (34%) in weeks 11 to 12 (P = .02). Thus, women
taking supplements experienced spontaneous abortion earlier than did women
not taking supplements.
Our study found no excess in spontaneous abortion risk in women with
high levels of plasma folate. Women with low folate levels were found to be
at significantly increased risk of spontaneous abortion.
No mechanism by which low folate levels could cause spontaneous abortion
has yet been identified. Folate deficiency has been tentatively associated
with abruptio placentae and preeclampsia,33 and
early vascular effects related to folate deficiency might also increase the
risk of spontaneous abortion. Low folate levels increase the incidence of
neural tube defects, and fetuses affected with neural tube defects are more
commonly aborted spontaneously.34 Women who
give birth to infants affected by neural tube defects more often have a history
of previous miscarriages.35 However, neural
tube defects are rare conditions and could explain only a fraction of the
association between low folate levels and spontaneous abortions. It is also
known that folate plays a critical role in DNA synthesis and regulation, by
providing carbon groups for synthesis of both purine and pyrimidine and by
providing methyl groups required for DNA regulation by methylation. Thus,
it has been suggested that the rapidly developing cells in the embryo may
be damaged by lack of adequate folate.36 Failure
to produce sufficient DNA and to regulate DNA function could lead to spontaneous
When we stratified our analyses by fetal karyotype, we found that the
risk was primarily confined to spontaneous abortions with abnormal karyotypes.
Most chromosomally abnormal concepti are spontaneously aborted, and we found
no differences in the severity of fetal karyotype abnormality by plasma folate
levels. It is possible that low folate status causes a larger proportion of
these losses to occur in the 6- to 12-week period. Although our study has
far better data than previous studies of folate effects on chromosomal status
and timing of losses, future studies will be required to resolve these issues.
Our results show that high folate levels are not associated with an
increased risk of spontaneous abortion; in fact, there was a nonsignificant
trend toward a protective effect associated with high folate levels. Given
the concern that folic acid supplements, which produce high folate levels,
may increase abortion rates,13,17 this
is an important finding. It is reassuring because today women of childbearing
age in many countries are advised to take folic acid supplements daily to
prevent neural tube defects, and food fortification has been introduced in
other countries to raise folate levels in the population. We were able to
study actual folate levels rather than having to rely on self-reported folic
acid use as a surrogate for folate status. Few studies have been able to identify
fetal losses occurring as early as 6 weeks' gestation. Including cases of
early fetal loss in this study enabled us to show that high folate levels
were, if anything, associated with a reduced risk of spontaneous abortion
even in early pregnancy.
Our study has a number of strengths. Unlike most previous investigations,
we were able to examine folate status within a large population-based study,
where plasma folate levels were measured at the time of fetal loss in cases,
and in pregnant controls matched by gestational age. All patients were interviewed
in person using a structured questionnaire. Our results also clearly show
the necessity to control for potential confounding factors, especially maternal
age. Measurement of plasma cotinine allowed us to adjust for smoking as a
There are also limitations to our study. Only slightly less than 5%
of the patients in our study used folic acid supplements, probably because
there were no recommendations for pregnant women to use folic acid supplements
in Sweden during the study period. The low rate of supplement use resulted
in somewhat limited power to study this group separately. It is impossible
to identify all early fetal losses since many occur before pregnancy is recognized
and before 6 weeks' gestation.37 Karyotyping
was not possible for all samples, which is to be expected since all women
with spontaneous abortion did not undergo curettage, and among some cases
where curettage was performed, no chorionic villi were identified. The success
rate of karyotyping increased, as expected, by week of gestation.27
Our results have important public health and clinical implications.
High folate status, as increasingly seen in the United States and many other
Western countries because of food fortification and supplement use, is not
associated with an increased risk of spontaneous abortion, whereas low folate
levels are associated with an increased risk of spontaneous abortion. Countries
that have fortified their food supplies with folic acid and those considering
doing so can be reassured that fortification most likely will not increase
spontaneous abortion rates and might even decrease them.