411?
American Journal of Epidemiology
Copyright © 1998 by The Johns Hopkins University School of Hygiene and Public Health
All rights reserved
Vol. 148, No. 2
Printed in U.S.A.
Maternal Cigarette Smoking, Regular Use of Multivitamin/Mineral
Supplements, and Risk of Fetal Death
The 1988 National Maternal and Infant Health Survey
Tiejian Wu, Germaine Buck, and Pauline Mendola
Data from the 1988 National Maternal and Infant Health Survey were used to examine whether regular use
of multivitamin/mineral supplements could modify the relation between maternal smoking and fetal death.
Maternal smoking was defined as the self-reported average number of cigarettes smoked after recognition of
pregnancy. Regular supplement use was defined as use of multivitamin/mineral supplements for at least 3
days per week during the 3 months before and/or after recognition of pregnancy. The sample comprises
12,465 singleton pregnancies, including 9,402 livebirths and 3,063 fetal deaths. Odds ratios were derived from
logistic regression analyses after adjustment for a number of demographic and reproductive variables. Major
findings are that 1) smoking increased the risk of fetal death; 2) regular supplement use either before or after
recognition of pregnancy did not affect the risk of fetal death in the absence of maternal smoking; 3) odds
ratios for fetal death among smoking women who regularly used supplements were generally smaller than
those for women who did not regularly use supplements but who smoked a comparable number of cigarettes;
and 4) a significant negative excess risk due to interaction was observed among women who regularly used
supplements before recognition of pregnancy and smoked 20 or more cigarettes a day. These findings
suggested that regular multivitamin/mineral supplement use might reduce the risk of fetal death associated
with maternal smoking. Am J Epidemiol 1998;148:215-21.
effect modifiers (epidemiology); fetal death; minerals; smoking; vitamins
Maternal smoking has been linked to numerous adverse pregnancy outcomes, including intrauterine
growth retardation and fetal death (1-9). Several pathways have been postulated to explain this smoking
"effect." Cigarette smoke contains an abundant number of free radicals that might attack biologic molecules such as protein and DNA (10). Intrauterine
exposure to those radicals may cause fetal free-radical
oxidative damage. In addition, exposure to tobacco
smoke can decrease the availability of certain nutrients, which may become depleted during the neutralization of select tobacco toxicants (11). These postulated mechanisms underlying the smoking effect
suggest that multivitamin/mineral supplementation in
pregnancy may reduce the risk of fetal death associated with maternal cigarette smoking. However, this
hypothesis has not been tested to date and serves as the
impetus for study. The purpose of this paper is to
examine whether regular multivitamin/mineral supplementation before and/or after recognition of pregnancy
could modify the risk of fetal death associated with
maternal smoking.
MATERIALS AND METHODS
The National Maternal and Infant Health Survey
(NMIHS), conducted by the National Center for
Health Statistics, is a survey of livebirths, fetal deaths,
and infant deaths in the United States based on a
probability sample of the vital events occurring in
1988. Livebirths were selected within six strata defined by race (black, nonblack) and birth weight
(< 1,500, 1,500-2,499, >2,500 g); black, very low
birth weight (< 1,500 g) and moderately low birth
weight (1,500-2,499 g) infants were purposely oversampled. Blacks also were oversampled for fetal and
infant deaths. Mothers identified from 13,417 livebirth
certificates, 4,772 fetal death certificates, and 8,166
infant death certificates comprised the study popula-
Received for publication July 7,1997, and accepted for publication December 30, 1997.
Abbreviations: Cl, confidence interval; NMIHS, National Maternal
and Infant Health Survey; OR, odds ratio; RERI, relative excess risk
due to interaction.
From the Department of Social and Preventive Medicine, School
of Medicine and Biomedical Science, State University of New York
at Buffalo, Buffalo, NY.
Reprint requests to Dr. Tiejian Wu, Department of Social and
Preventive Medicine, School of Medicine and Biomedical Science,
State University of New York at Buffalo, 270 Farber Hall, Buffalo, NY
14214.
215
216
Wu et al.
tion. A more complete description of the NMIHS
survey has been published elsewhere (12). Mothers
who responded to the questionnaires included 9,953
(74 percent) women who had livebirths, 3,309 (69
percent) women who had late fetal deaths at least 20
weeks gestation, and 5,332 (65 percent) women who
had infant deaths. Our study sample (n = 12,465) was
restricted to mothers with singleton livebirths (n =
9,402) or singleton fetal deaths (n = 3,063). Thirtyfive fetal deaths with missing data on gender were
excluded.
Regular use of vitamin/mineral supplements both
before and after recognition of pregnancy was ascertained by asking mothers, "Which of the following
vitamin/mineral supplements did you take at least 3
' days a week during the 3 months before (after) you
found out you were pregnant?" The question was
followed by a series of responses and an instruction to
check all responses that applied, including multivitamins and/or minerals, vitamin A, vitamin C, folic acid,
calcium, iron, zinc, and none of above. In our study,
(regular) vitamin users are defined as women who
used multivitamins and/or minerals supplements at
least 3 days per week for 3 months.
Maternal cigarette smoking was ascertained by asking women "On average, how many cigarettes did you
smoke a day after you found out that you were pregnant?" Maternal smoking was categorized into 0, 1-9,
10-19, and 20 or more cigarettes per day, for assessment of potential effect modification at various levels
of smoking.
The following variables were considered as potential confounders: maternal race (black or nonblack),
age, marital status at pregnancy (married or unmarried), education, prepregnancy weight, wantedness of
the pregnancy before becoming pregnant (pregnancy
wanted, pregnancy wanted at a later time, pregnancy
not wanted now or later), vomiting during pregnancy,
alcohol use during pregnancy, previous history of miscarriage and/or stillbirth, infant's gender, and family
income during the 12 months before delivery.
Fetal death rates were estimated based on the number of fetal deaths per 100 pregnancies in the sample
and stratified by the level of maternal smoking and
vitamin use. With the complex sampling structure
taken into account, weighted rates of fetal death were
calculated with the application of the final weights
provided in the NMIHS data set (13). Crude rate ratios
were computed by comparing the weighted rates. Logistic regression models were constructed for the estimation of separate effects (the effects in the presence
of one factor) and joint effects (the effects in the
presence of both maternal smoking and vitamin use) at
various levels of maternal smoking. Relative excess
risk due to interaction (RERI) was used to summarize
the departure of the joint effect (vitamin use and
smoking) from the additivity of the separate effects
(14, 15). The 95 percent confidence interval for each
RERI also was computed (15). Tests for linear trends
for the smoking effect and effect modification at different smoking levels were performed by including 1)
the categorized ordinal smoking level as a continuous
variable; 2) a dummy variable for vitamin use; and 3)
a product term for maternal smoking level and vitamin
use in logistic regression. All potential confounders
were included in all logistic regressions. The final
weights were rescaled for livebirths such that the
weighted sample size was equal to the actual sample
size (16), while the final weights for fetal deaths were
rescaled to one. The rescaled weights were applied in
all logistic regression models. The SAS software package (17) was used for all statistical analyses.
RESULTS
Descriptive statistics for the study population are
shown in table 1. Among the 12,465 women in the
sample, 23.8 percent reported having smoked after
recognition of pregnancy, while 23.4 percent reported
using vitamins regularly before recognition of pregnancy, and 74.8 percent after recognition of pregnancy. Overall, heavier smokers were more likely to
have lower educational attainments, to drink more
alcohol during pregnancy, to have had previous adverse pregnancy outcomes, and to report less wantedness of pregnancies than were nonsmoking mothers.
Mothers who regularly used vitamins tended to be
older, to be married, to have higher educational attainments and family incomes, and to report more wantedness of pregnancies than did mothers who did not
use multivitamins regularly.
Fetal death rates tended to be higher for mothers
who smoked in comparison with nonsmokers (tables 2
and 3). Regardless of vitamin use, the weighted fetal
death rate was 0.36 percent for nonsmoking mothers,
0.41 percent for mothers who smoked 1-9 cigarettes/
day, 0.42 for those who smoked 10-19 cigarettes/day,
and 0.47 percent for those who smoked 20 or more
cigarettes/day. When vitamin use either before or after
recognition of pregnancy was categorized as yes/no,
increased fetal death rates tended to be associated with
increased levels of maternal smoking, particularly in
the absence of vitamin use.
Logistic regression results for the effects of maternal smoking and vitamin use before recognition of
pregnancy are shown in table 4. In the absence of
regular vitamin use before recognition of pregnancy,
maternal smoking was significantly associated with an
increased risk of fetal death. Vitamin use before recAm J Epidemiol
Vol. 148, No. 2, 1998
Vitamin/Mineral Use, Smoking, and Fetal Death
217
TABLE 1. Characteristics of the study population by maternal smoking and regular use of multivitamin/minerai supplements
(n = 12,465)*, National Maternal and Infant Health Survey, 1988
Maternal smoking (cigarettes/day)
Characteristics
0
(n= 9,500)
1-9
(n= 1,194)
10-19
(n = 1,034)
Regular use of supplements
S20
(n=737)
Both before
and after
pregnancy
(n = 2,618)
None
(n= 2,846)
Before
pregnancy
(n = 293)
After
pregnancy
(n = 6,708)
64.8
38.7
51.1
46.5
14.9
83.9
1.7
49.2
2.8
11.6
60.1
50.2
50.2
38.6
11.6
86.3
1.7
50.5
1.4
11.3
43.4
58.4
51.3
39.6
8.4
83.3
0.8
46.7
2.3
15.4
35.0
63.7
52.3
33.7
7.9
79.8
0.7
41.3
2.8
19.1
Percentage
Black
Married
Male infant
Pregnancy wanted later
Pregnancy not wanted
NondrinMng during pregnancy
6 or more drinks/ week
Vomiting during pregnancy
Prior history of stilbirth
Prior history of miscarriage
47.8
57.4
51.5
38.9
8.9
87.5
0.2
46.3
2.3
14.9
56.3
41.3
50.5
45.7
10.5
69.6
1.8
48.7
2.8
14.0
Maternal age at pregnancy
Prepregnancy weight
(pounds)t
Maternal education}
Family income§
25.7 (6.0)
25.3 (5.4)
25.8 (5.5)
26.8 (5.6)
24.2 (6.0)
25.7 (5.9)
25.7 (5.9)
27.6 (5.6)
137.89 (30.0)
12.7 (2.5)
11.9(5.6)
136.2 (32.1)
12.0 (2.0)
10.2 (5.3)
135.9 (30.1)
11.7(1.8)
10.5 (5.3)
136.6 (32.1)
11.6(1.9)
10.3 (5.4)
138.3 (30.9)
11.7(2.4)
9.3 (5.5)
133.3 (29.6)
12.4 (2.5)
10.3 (5.6)
137.1 (29.6)
12.5 (2.5)
12.0 (5.6)
137.9 (29.6)
13.1 (2.4)
13.1 (5.4)
37.2
50.4
52.1
41.4
11.8
69.1
3.5
44.5
3.4
16.0
33.6
49.4
50.3
41.7
18.7
63.1
6.4
44.2
3.3
19.8
Mean (standard deviation)
* Restricted to singleton infants. All percentages and means are unweighted.
11 pound = 454 g.
t Maternal education was coded as 0-12 for 0 to 12th grade, 13-15 for 1-3 years of college, 16 for 4-5 years of college, 17 for 1 year of graduate school,
and 18 for more than 2 years of graduate school.
§ Level of family Income was coded as one of 20 levels for family income ranging from less than $1,000 to over $60,000.
TABLE 2. Fetal death rates by maternal smoking during
pregnancy and regular use of multivitamin/minerai
supplements before recognition of pregnancy, National
Maternal and Infant Health Survey, 1988
Vitamin use and
maternal smoking
(cigarettes/day)
No.
Fetal
deaths
Rate
Weighted
rate
TABLE 3. Fetal death rates by maternal smoking during
pregnancy and regular use of multivitamin/minerai
supplements after recognition of pregnancy, National
Maternal and Infant Health Survey, 1988
Crude
RRt
No
0
1-9
10-19
>20
Fetal
deaths
Rate
Weighted
rate
I "'
No
1,715
230
224
164
23.68
25.14
27.59
28.03
0.36
0.41
0.42
0.47
1.00
1.14
1.17
1.31
0
1-9
10-19
>20
2,333
320
277
209
538
81
82
62
23.08
25.31
29.60
29.67
0.38
0.45
0.53
0.54
1.00
1.18
1.39
1.42
2,258
279
222
152
571
68
56
35
25.29
24.37
25.23
23.03
0.35
0.40
0.39
0.26
0.97
1.11
1.08
0.72
Yes
0
1-9
10-19
>20
7,167
874
757
528
1,748
217
198
137
24.39
24.83
26.16
26.95
0.35
0.39
0.38
0.38
0.92
1.03
1.00
1.00
' Weighted rates are calculated using US vital events in 1988.
t Crude RR = rate ratio of weighted rates.
ognition of pregnancy in the absence of maternal
smoking tended to have little, if any, effect on the risk
of fetal death (odds ratio (OR) = 0.96, 95 percent
confidence interval (CI) 0.83-1.09).
Joint effects were compared with separate effects to
assess possible effect modification. The point estimates for odds ratios of fetal death tended to be lower
for mothers who used vitamins than for those who did
not but who smoked a comparable number of cigarettes. For example, compared with mothers who neither smoked nor used vitamins, the odds ratios for fetal
Am J Epidemiol
No.
7,242
915
812
585
Yes
0
1-9
10-19
>20
Vitamin use
and maternal
smoking
(cigarettes/day)
Vol. 148, No. 2, 1998
* Weighted rates are calculated using US vital events in 1988.
t Crude RR = rate ratio of weighted rates.
death for smoking 1-9, 10-19, and 20 or more cigarettes a day were 1.12, 1.36, and 1.59, respectively
(test for trend, p < 0.05). Among vitamin users, odds
ratios ranged from 0.84 to 1.09. Estimated RERIs were
all negative at various smoking levels, and a significant RERI was observed for mothers who smoked 20
or more cigarettes a day (i.e., RERI = 0.71, 95 percent
CI -1.24 to -0.18).
The effects of maternal smoking and vitamin use
after recognition of pregnancy are shown in table 5.
218
Wu et al.
TABLE 4. Risk of fetal death by maternal smoking during pregnancy and regular use of
multivitamin/mineral supplements before recognition of pregnancy, logistic regression analysis*
National Maternal and Infant Health Survey, 1988
Maternal
smoking
(cigarettes/
day)
Regular use ol supplements
before recognition ol pregnancy
No
Interaction between
smoking and
supplement use
Yes
ORt
95% Clt
OR
0
1-9
10-19
220
1.00
1.12
1.36
1.59
Referent
0.91 to 1.36
1.12 to 1.65
1.27 to 2.00
0.96
1.00
1.09
0.84
Trend*
p < 0.05
95% Cl
0.83 to
0.70 to
0.75 to
0.53 to
1.09
1.43
1.59
1.33
p >0.05
RERIt
95% Cl
-0.07
-0.22
-0.71
-0.50 to 0.36
-0.71 to 0.26
-1.24 to-0.18
p > 0.05
* Adjusted for maternal race, age, marital status at pregnancy, education, prepregnancy weight, wantedness
of the pregnancy, vomiting during pregnancy, alcohol drinking during pregnancy, previous history of miscarriage
and/or stillbirth, infant's gender, and family income during the 12 months before delivery.
t OR, odds ratio; Cl, confidence interval; RERI, relative excess risk due to interaction.
t Tested by including the ordinal smoking level as a continuous variable, a dummy variable for vitamin use, and
a product term for smoking and vitamin use.
TABLE 5. Risk of fetal death by maternal smoking during pregnancy and regular use of
multivitamin/mineral supplements after recognition of pregnancy, logistic regression analysis*, National
Maternal and Infant Health Survey, 1988
Regular use of supplements
after recognition ot pregnancy
Maternal
smoking
(cigarettes/
day)
ORt
95% Clt
ORt
0
1-9
10-19
220
1.00
1.04
1.42
1.77
Referent
0.74 to 1.47
1.01 to 1.99
1.22 to 2.55
0.94
1.06
1.21
1.22
Trend*
p < 0.05
No
Interaction between
smoking and
supplement use
Yes
95% Clt
0.82 to
0.85 to
0.97 to
0.94 to
1.09
1.33
1.52
1.59
p < 0.05
RERIt
95% Clt
0.08
-0.15
-0.49
-0.34 to 0.49
-0.68 to 0.37
-1.18 to 0.21
p > 0.05
* Adjusted for maternal race, age, marital status at pregnancy, education, prepregnancy weight, wantedness
of the pregnancy, vomiting during pregnancy, alcohol drinking during pregnancy, previous history of miscarriage
and/or stillbirth, infant's gender, and family income during the 12 months before delivery.
t OR, odds ratio; Cl, confidence interval; RERI, relative excess risk due to interaction.
i Tested by including the ordinal smoking level as a continuous variable, a dummy variable for vitamin use, and
a product term for smoking and vitamin use.
Maternal smoking in the absence of vitamin use after
recognition of pregnancy is associated with an increased risk of fetal death in a dose-response fashion.
Test for the effect (i.e., odds ratio) of vitamin use after
recognition of pregnancy in the absence of maternal
smoking did not reach a statistically significant level
(OR = 0.94, 95 percent Cl 0.82-1.09). Odds ratios for
fetal death at maternal smoking 10-19 cigarettes or 20
or more cigarettes a day tended to be smaller for
mothers who used vitamins than for those who did not.
However, none of RERIs were significant.
Vitamin use before and after recognition of pregnancy and risk of fetal death was examined by logistic
regression analysis (table 6). Maternal smoking was
divided into two groups (smokers and nonsmokers) to
improve the stability of estimates. In the presence of
maternal smoking, point estimates (i.e., odds ratios)
tended to be lower for mothers who continuously used
vitamins in comparison with those who did not use
vitamins. All point estimates for RERIs were negative,
while their 95 percent confidence intervals included
zero.
DISCUSSION
Although fetal and perinatal deaths are relatively
rare events, previous studies with sufficient sample
sizes have consistently reported an increase in risk of
fetal death and perinatal mortality among mothers who
smoked during pregnancy compared with nonsmoking
mothers (6-8).
Despite the fact that maternal smoking is a risk
factor for numerous adverse pregnancy outcomes, the
prevalence of cigarette smoking during pregnancy remains high (at about 25 percent) in the United States
(18). Data from the 1988 NMIHS indicate that 83.7
Am J Epidemiol
Vol. 148, No. 2, 1998
Vitamin/Mineral Use, Smoking, and Fetal Death
219
TABLE 6. Risk of fetal death by maternal smoking during pregnancy and regular use of
multivitamin/mineral supplements before and/or after recognition of pregnancy, logistic regression
analysis*, National Maternal and Infant Health Survey, 1988
Supplement use
before and after
recognition of
pregnancy
None
Before
After
Before and after
Maternal smoking
during pregnancy
Yes
No
ORt
1.00
0.91
0.94
0.94
Interaction between
smoking and
supplement use
95% Clf
Referent
0.61 to 1.37
0.81 to 1.09
0.74 to 1.19
OR
1.38
0.86
1.22
0.95
95% Cl
1.09
0.41
1.01
0.73
to
to
to
to
1.74
1.81
1.46
1.24
RERIt
95% Cl
-0.43
-0.10
-0.28
-1.22 to 0.36
-0.46 to 0.25
-1.16 to 0.59
* Adjusted for maternal race, age, marital status at pregnancy, education, prepregnancy weight, wantedness
of the pregnancy, vomiting during pregnancy, alcohol drinking during pregnancy, previous history of miscarriage
and/or stillbirth, infant's gender, and family income during the 12 months before delivery.
f OR, odds ratio; Cl, confidence interval; RERI, relative excess risk due to interaction.
percent of white mothers and 66.5 percent of black mothers reported using multivitamin/mineral supplements
regularly during pregnancy; mothers who smoked were
more likely to use multivitamin/mineral supplements
regularly during pregnancy than were nonsmoking mothers (19). This finding may suggest that women who
smoked perceive the taking of vitamins or are advised to
take vitamins during pregnancy to compensate for the
adverse effects of cigarette smoking on fetal growth and
development. Smoking during pregnancy has been found
to be associated with poor diet (20-22). However, the
possible beneficial effect(s) of vitamin use in preventing
fetal damage due to maternal smoking has not been
demonstrated clearly.
Our study used a large, national, representative sample and multivariate models to control for potential
confounding variables. The results of our study indicate that maternal smoking was associated with an
increased risk of fetal death. Mothers who smoked and
used vitamins tended to have a somewhat lower risk
for fetal death than did mothers who smoked and did
not use vitamins. Effect modification tended to be
more obvious among heavier smokers and when vitamins were used before the recognition of pregnancy.
This finding may suggest that vitamin use among
smokers reduces the risk of fetal death associated with
maternal smoking.
Observational and randomized clinical studies have
demonstrated that the use of multivitamins containing
folic acid during pregnancy has a protective effect on
the risk of birth defects, especially neural tube defects
(23-28). A more recent study by Shaw et al. (29) has
indicated that the protective effect of multivitamin
supplementation on the risk of neural tube defects is
greater among women who smoke cigarettes. Although these studies did not examine fetal death directly, their results may suggest a possible protective
effect of vitamin use on fetal development and, possibly, survival.
Am J Epidemiol
Vol. 148, No. 2, 1998
The results of our study are consistent with postulated biologic mechanisms. The postulated mechanisms of the smoking effect include free-radical oxidative damage and tobacco-induced nutritional
deficiency. Multivitamin/mineral supplements generally contain antioxidants such as vitamins C and E and
/3-carotene. Thus, vitamin use in pregnancy may counteract the adverse effects of smoking and benefit fetal
growth. Since the embryonic period (up to 8 weeks of
gestation) is a critical phase in which the embryo is
more vulnerable to teratogenic damage (30), it is important for vitamin status to be optimal during the
critical period of early embryogenesis. Hence, the
beneficial effect of minimizing the smoking effects on
the fetus may be more obvious if vitamins are used
early in pregnancy.
Our study is subject to several limitations. As described earner, mothers who used vitamins were quite
different from those who did not. This diversity may not
be fully adjusted by the potential confounders considered
in our analysis. As a result, our findings may be biased
and await confirmation by other investigators.
Maternal smoking prior to pregnancy was not taken
into account in our analysis. Of 3,807 mothers who
reported smoking prior to pregnancy, 855 (22.5 percent) stopped smoking after learning they were pregnant. The proportions of mothers who stopped smoking after learning they were pregnant were similar
regardless of whether or not they used vitamins regularly before recognition of pregnancy, i.e., 23.4 and
22.2 percent, respectively. However, the proportion of
mothers who stopped smoking after learning they were
pregnant was higher for mothers who used vitamins
after recognition of pregnancy (24.7 percent) than for
those who did not (15.8 percent). Since prepregnancy
maternal smoking may have residual effects, treating
mothers who stopped smoking due to pregnancy as
nonsmokers in our analysis may dilute the estimated
effect of vitamin use.
220
Wu et al.
Another limitation of this study pertains to the use
of self-reported exposure and outcome data. Both exposure variables—maternal smoking and regular use
of multivitamin/mineral supplements—were entirely
self-reported. Since the data were collected after delivery, recall bias is a potential limitation. For example, if mothers with fetal death were less likely to
recall vitamin/mineral use and more likely to recall
smoking during pregnancy (even though we have no
reason to expect so), the observed effect modification
could be overestimated.
An additional limitation of the study is the absence
of dietary information. An implicit assumption of this
study is that variation in dietary intake of nutrients
within groups (smokers or nonsmokers) is negligible
compared with that between groups. As mentioned
before, smokers tend to have poorer diets during pregnancies compared with nonsmokers. Smoking may
also cause depletion of certain nutrients. Therefore, the
nutritional status of smoking women may be quite
different from that of nonsmoking women. However,
dietary intake of nutrients may vary largely even
within groups, and the effect of vitamin use may be
highly dependent on dietary intake of nutrients. We
had no further dietary information available for study
purposes.
Lastly, the absence of a reference group, or women
who never used multivitamin/mineral supplements, is
another limiting factor. In our study, the effect(s) of
regular use of supplements was estimated by comparing regular users with nonregular users and not nonusers per se. This was the result of a very high prevalence of vitamin use during pregnancy. It remains
plausible that stronger effects for regular use of multivitamin/mineral supplements might be observed if a
nonuser reference group was available for study
purposes.
In summary, the results of this study suggest that
regular use of multivitamin/mineral supplements, particularly among heavier smokers, may reduce the risk
of fetal death associated with maternal cigarette smoking. These suggestive findings await confirmation by
future studies.
ACKNOWLEDGMENTS
The authors thank Dr. Michael Cummings, Dr. Jo
Freudenheim, and Kimberly Perez for their helpful comments on earlier versions of this paper and John Brasure for
his assistance with data management.
REFERENCES
1. Benjamin P. Sharing the cigarettes: the effects of smoking in
pregnancy. In: Rosenberg MJ, ed. Smoking and reproductive
health. Littleton, MA: PSG Publishing Co., Inc., 1985.
2. Berkowitz GS. Smoking and pregnancy. In: Niebyl JR, ed.
Drug use in pregnancy. 2nd ed. Philadelphia, PA: Lea &
Febiger, 1988.
3. US Department of Health and Human Services . The health
consequences of smoking for women: a report of the Surgeon
General. Rockville, MD: US Department of Health and Human
Services, 1980. (US DHHS publication no. 410-889/12840).
4. US Department of Health and Human Services. The health
benefit of smoking cessation: a report of the Surgeon General.
Rockville, MD: US Department of Health and Human Services, 1990. (DHHS publication no. (CDC) 90-8416).
5. US Department of Health, Education, and Welfare. Pregnancy
and infant health. In: Smoking and health. Rockville, MD: US
GPO, 1979.
6. Kleinman JC, Pierre MB, Madans JH, et al. The effects of
maternal smoking on fetal and infant mortality. Am J Epidemiol 1988;127:274-82.
7. Malloy MH, Kleinman JC, Land GH, et al. The association of
maternal age and cause of infant death. Am J Epidemiol
1988;128:46-55.
8. Rush D, Cassano P. Relationship of cigarette smoking and
social class to birth weight and perinatal mortality among all
births in Britain, 5-11 April 1970. J Epidemiol Community
Health 1983;37:249-55.
9. Abel EL. Smoking during pregnancy: a review of effects on
growth and development of offspring. Hum Biol 1980;52:
593-625.
10. Hallwell B. Cigarette smoking and health: a radical view. J R
Soc Health 1993;113:91-6.
11. Institute of Medicine. Substance use and abuse during pregnancy. In: Institute of Medicine, ed. Nutrition during
pregnancy: weight gain and nutrient supplements. Washington, DC: National Academy Press, 1990.
12. Sanderson M, Placek PJ, Keppel KG. The 1988 National
Maternal and Infant Health Survey: design, content, and data
availability. Birth 1991 ;18:26-32.
13. National Center for Health Statistics. Public use data tape
documentation: 1988 National Maternal and Infant Health
Survey. US Department of Commerce, National Technical
Information Service. Hyattsville, MD: Centers for Disease
Control and Prevention, 1991.
14. Rothman KJ. Modern epidemiology. Boston, MA: Little,
Brown, & Co., 1986.
15. Hosmer DW, Lemeshow S. Confidence interval estimation of
interaction. Epidemiology 1992;3:452-6.
16. Lee ES, Forthofer RN, Lorimor RJ. Analysis of complex
sample survey data: problems and strategies. Soc Methods Res
1986;15:69-101.
17. SAS. SAS/STAT user's guide. Version 6, fourth ed. Cary,
NC: SAS Institute, Inc., 1990.
18. National Center for Health Statistics. Health promotion and disease prevention, United States, 1985. (Vital and health statistics,
Series 10, no. 163). Hyattsville, MD: US Public Health Service,
1988. (DHHS publication no. (PHS) 88-1591).
19. Yu MS, Keppel KG, Singh GK, et al. Preconceptional and
prenatal multivitamin-mineral supplementation in the 1988
National Maternal and Infant Health Survey. Am J Public
Health 1996;86:240-2.
20. Haste FM, Brooke OG, Anderson HR, et al. Nutritional intakes during pregnancy: observations on the influence of
smoking and social class. Am J Clin Nutr 1990;51:29-36.
21. Trygg K, Lund-Larsen K, Sandstad B, et al. Do pregnant
smokers eat differently from pregnant non-smokers? Pediatr
Perinat Epidemiol 1995;9:307-19.
22. Mckenzie-Pamell JM, Wilson PD, Parnell WR, et al. Nutrient
intake of Dunedin women during pregnancy. N Z Med J
1993;106:273-6.
Am J Epidemiol Vol. 148, No. 2, 1998
Vitamin/Mineral Use, Smoking, and Fetal Death
23. Medical Research Council Vitamin Study Research Group.
Prevention of neural tube defects: results of the Medical
Research Council vitamin study. Lancet 1991;339:131-7.
24. Shaw GM, Lammer El, Wasserman CR, et al. Risks of orofacial
clefts in children bom to women using multivitamins containing
folk acid periconceptionally. Lancet 1995;346:393-6.
25. Czeizel AE, Dudas I. Prevention of the first occurrence of
neural tube defects by periconceptional vitamin supplementation. N Engl J Med 1992;327:1832-5.
26. Mulinare J, Cordero JF, Erickson JD, et al. Periconceptional
use of multivitamins and the occurrence of neural tube defects.
JAMA 1988;260:3141-5.
Am J Epidemiol
Vol. 148, No. 2, 1998
221
27. Werler MM, Shapiro S, Mitchell AA. Periconceptional folic
acid exposure and risk of occurrent neural tube defects. JAMA
1993;269:1257-61.
28. Milunsky A, Jick H, Jick SS, et al. Multivitamin/folic acid
supplementation in early pregnancy reduces the prevalence of
neural tube defects. JAMA 1989;262:2847-52.
29. Shaw GM, Schaffer D, Velie EM, et al. Periconceptional
vitamin use, dietary folate, and the occurrence of neural tube
defects. Epidemiology 1995;6:219-26.
30. Dyball RE, Tale PA. Basic embryology and the embryological
basis of malformation syndromes. In: Roberton NRC, ed.
Textbook of neonatology. London, England: Churchill
Livingstone, 1986:109-19.