American Journal of Epidemiology
Published by the Johns Hopkins Bloomberg School of Public Health 2008.
Vol. 168, No. 4
DOI: 10.1093/aje/kwn140
Advance Access publication June 16, 2008
Original Contribution
Interactions between Smoking and Weight in Pregnancies Complicated by
Preeclampsia and Small-for-Gestational-Age Birth
Roberta B. Ness1, Jun Zhang2, Debra Bass1, and Mark A. Klebanoff2
1
Department of Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, PA.
Division of Epidemiology, Statistics and Prevention Research, National Institute of Child Health and Human Development,
National Institutes of Health, Department of Health and Human Services, Bethesda, MD.
2
Received for publication February 20, 2008; accepted for publication May 1, 2008.
Cigarette smoking protects against preeclampsia but increases the risk of small-for-gestational-age birth (SGA).
Regarding body weight, the converse is true: obesity elevates rates of preeclampsia but reduces rates of SGA. The
authors assessed the combined effects of smoking and weight among US women developing preeclampsia or
SGA, studying 7,757 healthy, primigravid women with singleton pregnancies in 1959–1965. Smoking (never, light,
heavy), stratified by prepregnancy body mass index (BMI (weight (kg)/height (m)2); underweight, overweight,
obese), was examined in relation to preeclampsia and SGA. Among underweight (BMI <18.5) and normal-weight
(BMI 18.5–24.9) women, smoking decreased the risk of preeclampsia (for heavy smoking, light smoking, nonsmoking, test for trend p ¼ 0.002 for underweight and p ¼ 0.009 for normal weight) after adjustment for age, race,
and socioeconomic status. However, among overweight/obese women (BMI 25), this trend was not apparent
(p ¼ 0.4). Among both underweight and overweight women, smoking significantly increased SGA risk (trend p <
0.001 for underweight and p ¼ 0.02 for overweight/obese). Obesity eliminated the inverse association between
smoking and preeclampsia but did not substantially alter the positive association between smoking and SGA.
A possible unifying biologic explanation is discussed in this paper.
body weight; infant, small for gestational age; overweight; pre-eclampsia; pregnancy; smoking
Abbreviations: BMI, body mass index; OR, odds ratio; SGA, small-for-gestational-age birth.
One of the most perplexing, yet consistent findings in
reproductive epidemiology is that smoking, while elevating
small-for-gestational-age birth (SGA) risk (1, 2), protects
against preeclampsia (3, 4). In contrast, maternal overweight and obesity increase risk of preeclampsia and decrease risk of SGA (5–11). The interaction between these
two counteracting factors on these two adverse perinatal
outcomes has not been well examined. Two previous studies
of low birth weight showed that smoking was a risk factor
for that outcome among both normal-weight and obese
women (12, 13). Two studies of preeclampsia reported discordant results: one found no differential smoking effect by
weight group, whereas the other found that, among obese
women, the protective effect of smoking was magnified (14,
15). The small number of studies and their methodological
limitations, including nonprospective designs and small
sample sizes, leaves in question the interaction between
smoking and body size.
To further assess the joint effects of smoking and weight
on both preeclampsia and SGA, we used data from the
Collaborative Perinatal Project, a large, US multicenter cohort of pregnant women, conducted at a time when women
commonly smoked during pregnancy. Examining the joint
effects of obesity, a driver of metabolic abnormalities, and
smoking, a mediator of endothelial dysfunction, may shed
light on the pathophysiology of both conditions (16).
Correspondence to Dr. Roberta B. Ness, University of Pittsburgh, Graduate School of Public Health, Room A527 Crabtree Hall, 130 DeSoto
Street, Pittsburgh, PA 15261 (e-mail: [email protected]).
427
Am J Epidemiol 2008;168:427–433
428 Ness et al.
MATERIALS AND METHODS
Subjects
Details of subject selection and recruitment, data collection, and follow-up for the Collaborative Perinatal Project
have been described elsewhere (17). Briefly, women who
received prenatal care at 12 hospitals from 1959 to 1965
were invited to participate. Mothers enrolled were serially
followed through labor and delivery. A total of 55,908 pregnancies were included in the project.
We restricted our analyses to women with singleton pregnancies of 20 weeks’ gestation (n ¼ 54,492) with their first
(primigravid) pregnancy (n ¼ 15,192) registered in the first
or second trimester by menstrual dating (n ¼ 11,711), with
three or more prenatal visits (n ¼ 11,390), and without any
of the following conditions before or during pregnancy:
thrombosis, phlebitis, history of coagulation defects, diabetes mellitus, glomerulonephritis, and hydramnios (remaining n ¼ 10,834). Moreover, we excluded women with
chronic hypertension or renal disease and women for whom
these or smoking data were missing (remaining n ¼ 9,651).
Our inclusion criteria were designed to maximize the validity of categorizing preeclampsia, which requires a finding
of hypertension and proteinuria in pregnancy in a woman
without preexisting elevations. It also enabled a fair assessment of whether uncomplicated overweight/obesity interacts with smoking in relation to preeclampsia and SGA.
We restricted the analysis to women who had information
on main exposures and outcomes of interest including
smoking, weight, height, birth weight, and gestational age
at delivery (calculated as the number of days between the
reported date of the last menstrual period and the birth date).
Finally, we compared only those women who smoked during pregnancy with never smokers, excluding past smokers
from analyses because this group was too small to provide
stable estimates for interaction effects. This process resulted
in 7,757 women who provided the basis for this analysis.
Measurement of smoking and weight
At study entry, all enrolled women underwent a standardized interview that collected detailed demographic, socioeconomic, behavioral, and medical history information.
A study physical examination was also conducted and blood
and urine specimens obtained. At each subsequent prenatal
visit, women were interviewed and examined.
Self-report of smoking was obtained at the entry interview
and then at each prenatal visit. Women who reported smoking
a total of five or more packs during their lifetime were asked
about smoking duration, cessation, and dose. On the basis of
this information, we categorized women as current smokers,
including those who smoked at study entry, those who initiated smoking after enrollment, and those who quit during the
first trimester. These women were compared with those who
denied ever smoking at least five packs of cigarettes. Among
current smokers, women who reported smoking 10 or more
cigarettes per day at baseline were categorized as heavy
smokers and women who reported smoking fewer than 10
cigarettes per day or who later in pregnancy quit or initiated
smoking were categorized as light smokers.
Prepregnancy weight was recalled by participants, and height
was measured at enrollment in a standardized way. Body mass
index (BMI) was calculated as weight (kilograms)/height
(meters)2. Standard cutoffs, recommended by the Centers for
Disease Control and Prevention (Atlanta, Georgia), were used
to group BMI into underweight (<18.5), normal weight (18.5–
24.9), and overweight/obese (25) categories (18). The combined overweight/obese category was necessitated because
there were too few obese (BMI 30) women with adverse
pregnancy outcomes to provide stable estimates.
Measurement of preeclampsia and SGA
Preeclampsia was defined as an elevation in blood pressure plus proteinuria. Blood pressure was recorded at study
enrollment, during each prenatal visit, during labor and delivery, and postpartum. Random urine samples were tested
for albumin at each prenatal visit. The blood pressure criterion used to define preeclampsia was diastolic blood pressure of 90 mmHg in the absence of hypertension before 24
weeks’ gestation or after 2 weeks postpartum (19). Elevation in blood pressure was assessed during each of at least
three measurements and had to be present on at least two
occasions from 24 weeks’ gestation to 2 weeks postpartum.
The proteinuria criteria were 1þ on random urine dipstick
testing on two occasions or 3þ on urine dipstick testing on
one occasion in the absence of chronic renal disease and in
the absence of proteinuria before 24 weeks’ gestation or
after 2 weeks postpartum. We derived the diagnosis of preeclampsia from the actual measurements at each prenatal
visit and during the peripartum period rather than from
summary forms completed by study obstetricians.
SGA was defined as a birth weight for gestational age
below the fifth percentile. This cutoff was chosen to minimize the number of babies who were constitutively small
and to maximize the number with true growth restriction
(20). Population-based and temporally appropriate birthweight-for-gestational-age nomograms by race and offspring gender classified babies in the lowest 5 percentile
as an SGA (21, 22). Women categorized as having preeclampsia accompanied by SGA were excluded from primary data analyses but were later included with the
preeclampsia group in sensitivity analyses.
Covariates
Covariate information, also obtained by interview, included maternal age, race (White, Black, other), and socioeconomic status. The socioeconomic status index variable
was obtained by averaging the component scores for reported education, occupation, and family income to achieve
a continuous, composite numerical index (23).
Analysis
We first compared demographic and clinical features
among women who developed preeclampsia and women
who developed SGA. Baseline differences between groups
were analyzed with the chi-square test for categorical variables and test of trend for interval variables. Frequencies of
Am J Epidemiol 2008;168:427–433
Preeclampsia and SGA
429
TABLE 1. Descriptive characteristics of nulliparous US women enrolled in the Collaborative Perinatal
Project, by the pregnancy outcomes preeclampsia and SGA,* 1959–1965
No.
%
Normal
(n ¼ 6,514)
No.
%
Preeclampsia
(n ¼ 481)
No.
%
SGA
(n ¼ 679)
No.
%
Race
White
3,574
46.1
3,020
46.4
175
36.4
349
51.4
Black
3,422
44.1
2,878
44.2
243
50.5
263
38.7
Other
761
9.8
616
9.5
63
13.1
67
9.9
Age (years)
11–15
721
9.3
575
8.8
71
14.8
64
9.4
16–20
4,127
53.2
3,459
53.1
276
57.4
346
51.0
21–25
2,280
29.4
1,961
30.1
107
22.2
192
28.3
629
8.1
519
8.0
27
5.6
77
11.3
26
Socioeconomic status (quintile)
1 (lowest)
706
9.3
565
8.9
74
15.6
60
9.0
2
2,042
26.9
1,685
26.4
150
31.6
178
26.6
3
2,074
27.3
1,721
27.0
143
30.1
193
28.8
4
1,644
21.6
1,393
21.9
78
16.4
151
22.5
5 (highest)
1,135
14.9
1,009
15.8
30
6.3
88
13.1
Never smoker
3,516
45.3
2,998
46.0
249
51.8
222
32.7
Light smoker
2,981
38.4
2,500
38.4
187
38.9
268
39.5
Heavy smoker
1,260
16.2
1,016
15.6
45
9.4
189
27.8
Smoking status
Body mass indexy
Underweight (<18.5)
Normal weight (18.5–24.9)
Overweight/obese (25)
965
12.4
782
12.0
44
9.1
124
18.3
5,981
77.1
5,060
77.7
365
75.9
492
72.5
811
10.5
672
10.3
72
15.0
63
Mean gestational age (weeks)
7,748
Mean birth weight (g)
7,757
39.1
3,160
39.4
3,195
9.3
39.8
2,402
Gestational age (weeks)
<35
565
7.3
518
8.0
29
6.1
16
2.4
35–36
546
7.0
472
7.2
40
8.5
27
4.0
6,637
85.7
5,524
84.8
403
85.4
636
93.7
37
Birth weight (g)
<1,500
104
1.3
62
1.0
2
0.4
33
4.9
1,500–2,499
735
9.5
355
5.4
29
6.0
305
44.9
6,918
89.2
6,097
93.6
450
93.6
341
50.2
2,500
* SGA, small-for-gestational-age birth.
y Weight (kg)/height (m)2.
developing preeclampsia and of developing SGA were calculated for nonsmokers, light smokers, and heavy smokers,
stratified by weight (underweight, normal weight, and overweight/obese) categories. Within BMI categories, odds ratios (and 95 percent confidence intervals) were calculated by
analyzing the exposure of smoking and separately the outcomes of preeclampsia and of SGA. Logistic regression
models were then used to adjust for maternal age (continuous), race (Black, White, other), and socioeconomic status
(continuous). Statistical significance of the smoking 3 BMI
interaction term was tested in logistic models.
Am J Epidemiol 2008;168:427–433
RESULTS
Of 7,757 healthy, primigravid women who were enrolled
in the first or second trimester of pregnancy and were followed to a singleton gestation, 481 (6.2 percent) developed
preeclampsia and 679 (8.8 percent) developed SGA (table 1);
83 (1.1 percent) developed both preeclampsia and SGA and
were included in secondary analyses only. Most women at
baseline were 20 years of age or younger. About 50 percent
were White, 40 percent were Black, and 10 percent were of
other races (mostly Puerto Rican). Over half of the women
430 Ness et al.
FIGURE 1. Frequency of developing preeclampsia or small-for-gestational-age birth among nonsmokers, light smokers, and heavy smokers
within categories of underweight, normal weight, and overweight, United States, 1959–1965. BMI, body mass index (weight (kg)/height (m)2).
currently smoked versus never smoked. Prepregnancy underweight and overweight/obesity each affected just over
one tenth of the women studied. Preeclampsia was less common and SGA more common among current smokers than
nonsmokers. Preeclampsia was more common and SGA less
common among overweight/obese than underweight women
(figure 1).
The reduced occurrence of preeclampsia and elevated occurrence of SGA among smokers was modified by weight
group. Among underweight women, smoking intensity was
associated with a pronounced stepwise reduction in preeclampsia, whereas, among overweight/obese women, heavy
smokers and light smokers compared with nonsmokers had
modest or no decreases in preeclampsia. Regarding SGA,
weight group only modestly modified the increased risk associated with smoking.
After adjustment for age, race, and socioeconomic status,
among underweight women (BMI <18.5), smoking decreased the risk of preeclampsia (light smokers, adjusted odds
ratio (OR) ¼ 0.54; heavy smokers, adjusted OR ¼ 0.24: test
for trend p ¼ 0.002) (table 2). However, among overweight/
obese women (BMI 25), we found no evidence for a
smoking-related reduction in risk (light smokers, adjusted
OR ¼ 1.04; heavy smokers, adjusted OR ¼ 0.80: test for
trend p ¼ 0.4). Among underweight women, smoking increased SGA risk (light smokers, adjusted OR ¼ 2.15; heavy
smokers, adjusted OR ¼ 3.52: test for trend p < 0.001); for
overweight/obese women, this trend remained significant
(light smokers, adjusted OR ¼ 1.42; heavy smokers, adjusted OR ¼ 2.22: test for trend p ¼ 0.02). Interaction terms
for weight by current smoking were tested after adjustment
for age, race, and socioeconomic status. For preeclampsia,
the weight 3 smoking interaction term was significant (p ¼
0.04); for SGA, it was not (p ¼ 0.11).
In sensitivity analyses, we added to the preeclampsia group
those women who had preeclampsia complicated by SGA.
Too few women had preeclampsia with SGA for meaningful
stratified comparisons (only 10 were heavy smokers; only
four were obese); thus, not surprisingly, adding women with
preeclampsia and SGA to the preeclampsia group had no
substantial impact on the odds ratios or trend tests.
DISCUSSION
The novel observation from this large, US multisite cohort study was that overweight and obesity, a strong driver
of insulin resistance, hyperlipidemia, inflammation, and impaired endothelial function (24), eliminated the protective
effect of smoking on preeclampsia risk. In contrast, regardless of weight group, smoking elevated the risk of SGA. The
overall trends for smoking and body size were consistent with
the previous literature: smoking was associated with a dosedependent elevation in SGA and a reduction in preeclampsia,
whereas obesity was more common in preeclampsia (3, 4)
and less common in pregnancies complicated by SGA than in
uncomplicated pregnancies (1, 2).
Tobacco smoking causes maternal endothelial dysfunction and abnormal placentation, lesions common to preeclampsia and fetal growth restriction (25, 26). Maternal
Am J Epidemiol 2008;168:427–433
Preeclampsia and SGA
TABLE 2. Adjusted odds ratios (and 95% confidence
intervals) for preeclampsia or SGA,* by weight groupy and
current smoking statusz in the Collaborative Perinatal Study,
United States, 1959–1965
No.
Preeclampsia vs.
normal
OR*,§
95% CI*
SGA vs.
normal
OR§
95% CI
Underweight
Total
965
Nonsmoker
429
1.0
1.0
Light smoker
359
0.54
0.26, 1.08 2.15 1.32, 3.52
Heavy smoker
177
0.24
0.06, 1.08 3.52 2.00, 6.19
Trend p ¼ 0.002
Trend p < 0.001
1.0
1.0
Normal weight
Total
5,981
Nonsmoker
2,721
Light smoker
2,322
0.89
0.71, 1.13 1.41 1.14, 1.76
938
0.59
0.40, 0.88 2.30 1.77, 2.99
Heavy smoker
Trend p ¼ 0.009
Trend p < 0.001
Overweight/obese
Total
811
Nonsmoker
366
1.0
1.0
Light smoker
300
1.04
0.61, 1.77 1.42 0.75, 2.68
Heavy smoker
145
0.80
0.36, 1.78 2.22 1.11, 4.45
Trend p ¼ 0.4
Trend p ¼ 0.02
* SGA, small-for-gestational-age birth; OR, odds ratio; CI, confidence interval.
y Underweight: body mass index (BMI; weight (kg)/height (m)2)
<18.5; normal weight: BMI 18.5–24.9; overweight/obese: BMI 25.
z Light smoker: currently smoking <10 cigarettes/day; heavy
smoker: currently smoking 10 cigarettes/day.
§ Adjusted for maternal age, race, and socioeconomic status
indicator.
endothelial dysfunction is present before, during, and after
pregnancies complicated by both SGA and preeclampsia
(27–29). Why then does smoking have discordant effects
on these two perinatal morbidities?
Hypotheses regarding the protection for preeclampsia afforded by cigarettes have mostly centered on vascular explanations: hypotension from toxins such as thiocyanate,
inhibition of thromboxane-induced vasoconstriction, and inhibition of anti-angiogenesis, to name a few (30–32). However, they do not explain why cigarette smoking elevates the
risk of SGA.
We previously hypothesized that 1) preeclampsia develops in the context of both maternal endothelial function
and concomitant metabolic abnormalities (hyperglycemia/
hyperinsulinemia, inflammation, and hyperlipidemia) and
2) SGA develops when endothelial dysfunction is unaccompanied by metabolic abnormalities (16). In particular, markers
of endothelial dysfunction are common to both preeclampsia
and SGA (e.g., prepregnancy hypertension and abnormal vascular reactivity). However, factors associated with maternal
hyperglycemia/hyperinsulinemia, hyperlipidemia, and hyperAm J Epidemiol 2008;168:427–433
431
inflammation are discordant for preeclampsia and SGA.
Maternal obesity increases the rate of preeclampsia, and preeclamptic pregnancies demonstrate increases in all of these
metabolic abnormalities, starting early in pregnancy (5–9). In
contrast, underweight elevates the risk of SGA, and pregnancies resulting in SGA are characterized by lower lipid levels
than those found in uncomplicated pregnancies (10, 11).
Our data, showing that obesity eliminates the protective
effect of smoking on preeclampsia, support our a priori hypothesis. Smoking can trigger elements of the metabolic
syndrome (33), but smoking also has antimetabolic and,
particularly, antiinflammatory effects (34, 35). Inflammation
may be the key to creating a positive feedback loop in preeclampsia; therefore, the antiinflammatory effects of smoking
may curtail the progression to that outcome (16). Underweight women have fewer metabolic abnormalities than overweight/obese women, and smokers may be doubly protected
from metabolic abnormalities in pregnancy. According to our
hypothesis, since preeclampsia is triggered by metabolic
abnormalities, underweight smokers are protected. Among
overweight/obese smokers, the antimetabolic effects of
smoking counter the prometabolic effects of obesity. Preeclampsia thus occurs in overweight/obese smokers at about
the same rate as it does in nonsmokers.
Previous studies examining the joint effect of smoking
and obesity on low birth weight found an increased risk of
cigarette use for both normal-weight and obese women (12,
13). These results are consistent with our own, although, to
our knowledge, no previous studies have examined SGA per
se. Inconsistent with our findings are the results of two previous studies examining joint effects of smoking and body
size on preeclampsia risk. In a study of birth certificate data
from Missouri, the risk of preeclampsia was lower for women
who smoked regardless of reported BMI (14). Smoking and
preeclampsia, in particular, have been shown to be poorly
captured on birth certificates. Therefore, likely there was a sizable degree of misclassification in the Missouri study. Differential misclassification may have occurred if overweight/
obese women were less likely than underweight women to
have smoking information recorded on birth certificates.
Misclassification may have also biased the interaction effect
toward the null if transient hypertension was admixed with
preeclampsia more in the underweight than in the overweight/obese. Moreover, the authors did not present the
smoking and preeclampsia associations stratified by BMI
in adjusted models. Because our data were 1) prospective,
2) based on direct self-reports prior to outcomes, and 3) presented in adjusted models, we believe that our results are
more valid than those in the prior publication.
A second study, comparing preeclampsia cases with controls from Japan, reported that, among overweight women
(BMI 24), smoking was markedly more common than in
nonoverweight women (15). This study was small (71 cases),
and the few overweight women provided unstable estimates.
Strengths of our study include its size, prospective design,
careful outcome categorization based on repeated measures
of blood pressure and urinary albumin, ability to base SGA
on a low (5 percentile) standard, recall of prepregnancy
weight early in pregnancy, standardized measurement of
height, and structured assessment of smoking.
432 Ness et al.
The potential weaknesses dominantly relate to misclassification of exposures and outcomes, but these likely had
little effect on our results. Smoking misclassification has
been shown to be minimal in the Collaborative Perinatal
Project (36). Self-reported prepregnancy weight is also a
relatively accurate measure of true weight (37, 38). Misclassification of preeclampsia was minimized by using an
accepted categorization method previously published from
this study (19) and by excluding women who were multiparous, multigravid, or had chronic hypertension and renal
disease. Reassuringly, the incidence of preeclampsia reported
here is virtually the same as that reported in two recent multicenter clinical trials (9, 39). Because SGA categorization
was based on gestational age by menstrual dating, it may
have been misclassified. At the same time, misclassification
was minimized by using a 5-percentile cutoff within raceand gender-specific, temporally appropriate, birth-weightfor-gestational-age nomograms. That these cutoffs categorized
more than 5 percent of neonates as SGA in the Collaborative
Perinatal Cohort (9.8 percent of Whites and 7.7 percent of
Blacks met criteria for SGA) suggests that neonates of participating women were of lower birth weight for gestational age
than would be expected in the general population of that time
but does not invalidate the specificity of using these nomograms to classify growth restriction (20, 22). A final study
weakness is that this study was conducted in the 1950s–
1960s, when obstetric practices and tobacco processing
were different than they are today.
In summary, we found that obesity eliminated the protective effect of smoking on preeclampsia, whereas obesity
did not considerably modify the association between smoking and SGA. Our data are consistent with, but cannot prove,
the plausible hypothesis that preeclampsia risk is elevated
by combined effects of endothelial dysfunction and metabolic abnormalities, whereas SGA risk is elevated by endothelial dysfunction alone.
ACKNOWLEDGMENTS
This study was funded by grant P01 HD30367 from the
National Institutes of Child Health and Disease and grant
5MO1 RR00056 from the National Institute of Research
Resources.
Conflict of interest: none declared.
REFERENCES
1. Magee BD, Hattis D, Kivel NM. Role of smoking in low birth
weight. J Reprod Med 2004;49:23–7.
2. England LJ, Kendrick JS, Gargiullo PM, et al. Measures of
maternal tobacco exposure and infant birth weight at term. Am
J Epidemiol 2001;153:954–60.
3. Newman MG, Lindsay MK, Graves W. Cigarette smoking and
pre-eclampsia: their association and effects on clinical outcomes. J Matern Fetal Med 2001;10:166–70.
4. Conde-Agudelo A, Althabe F, Belizan J, et al. Cigarette
smoking during pregnancy and risk of preeclampsia: a systematic review. Am J Obstet Gynecol 1999;181:1026–35.
5. Bodnar LM, Catov JM, Klebanoff MA, et al. Prepregnancy
body mass index and the occurrence of severe hypertensive
disorders of pregnancy. Epidemiology 2007;18:234–9.
6. Cedergren MI. Maternal morbid obesity and the risk of adverse
pregnancy outcomes. Obstet Gynecol 2004;103:219–24.
7. Lorentzen B, Endresen MJ, Clausen T, et al. Fasting serum free
fatty acids and triglycerides are increased before 20 weeks of
gestation in women who later develop preeclampsia. Hypertens Pregnancy 1994;13:103–9.
8. Lorentzen B, Drevon CA, Endresen MJ, et al. Fatty acid pattern of esterified and free fatty acids in sera of women with
normal and pre-eclamptic pregnancy. Br J Obstet Gynaecol
1995;102:530–7.
9. Sibai BM, Gordon T, Thom E, et al. Risk factors for preeclampsia in healthy nulliparous women: a prospective multicenter study. Am J Obstet Gynecol 1995;172:642–8.
10. Zeitlin JA, Ancel PY, Saurel-Cubizokes MJ, et al. Are risk
factors the same for small for gestational age versus other
preterm births? Am J Obstet Gynecol 2001;185:208–15.
11. Sattar N, Greer I, Galloway PJ, et al. Lipid and lipoprotein
concentrations in pregnancies complicated by intrauterine
growth restriction. J Clin Endocrinol Metab 1999;84:
128–30.
12. Haworth JC, Ellestad-Sayed JJ, King J, et al. Relation of
maternal cigarette smoking, obesity, and energy consumption
to infant size. Am J Obstet Gynecol 1980;138:1185–9.
13. Hellerstedt WL, Himes JH, Story M, et al. The effects of
cigarette smoking and gestational weight change on birth
outcomes in obese and normal-weight women. Am J Public
Health 1997;87:591–6.
14. Stone CD, Diallo O, Shyken J, et al. The combined effect of
maternal smoking and obesity on the risk of preeclampsia.
J Perinat Med 2007;35:28–31.
15. Kobashi G. An association between maternal smoking and
preeclampsia in Japanese women. Semin Thromb Hemost
2002;38:507–10.
16. Ness RB, Sibai BM. Shared and disparate components of the
pathophysiologies of fetal growth restriction and preeclampsia. Am J Obstet Gynecol 2006;195:40–9.
17. Niswander KR, Gordon M, eds. The collaborative perinatal
study of the National Institute of Neurological Diseases and
Stroke: the women and their pregnancies. Philadelphia, PA:
W B Saunders, 1972.
18. Centers for Disease Control and Prevention (CDC). BMI—
Body mass index. (http://www.cdc.gov/nccdphp/dnpa/bmi/
bmi-adult.htm). (Accessed January 18, 2005).
19. Zhang J, Klebanoff MA, Roberts JM. Prediction of adverse
outcomes by common definitions of hypertension in pregnancy. Obstet Gynecol 2001;97:261–7.
20. Goldenberg RI, Cliver SP. Small for gestational age and intrauterine growth restriction: definitions and standards. Clin
Obstet Gynecol 1997;40:704–14.
21. Zhang J, Harville E. Birth-weight-for-gestational-age patterns
by race, sex, and parity in the United States population: the 5th
percentile. Paediatr Perinat Epidemiol 1998;12:352–4.
22. Overpeck MD, Hediger ML, Zhang J, et al. Birth weight for
gestational age of infants of Mexican-American descent born
in the United States. Obstet Gynecol 1999;93:943–7.
23. Myrianthopoulos NC, French KS. An application of the US
Bureau of the Census socioeconomic index to a large, diversified patient population. Soc Sci Med 1968;2:283–99.
24. Ramsey JE, Ferrell WR, Crawford L, et al. Maternal obesity
is associated with dysregulation of metabolic, vascular, and
inflammatory pathways. J Clin Endocrinol Metab 2002;87:
4231–7.
Am J Epidemiol 2008;168:427–433
Preeclampsia and SGA
25. Kaufmann P, Black S, Huppertz B. Endovascular trophoblast
invasion: implications for the pathogenesis of intrauterine
growth retardation and preeclampsia. Biol Reprod 2003;69:
1–7.
26. Salafia C, Shiverick K. Cigarette smoking and pregnancy II:
vascular effects. Placenta 1999;20:273–9.
27. Sibai BM, Lindheimer M, Hauth J, et al. Risk factors for
preeclampsia, abruptio placentae, and adverse neonatal outcomes among women with chronic hypertension. N Engl J
Med 1998;339:667–71.
28. Ray JG, Vermeulen MJ, Schull MJ, et al. Cardiovascular
health after maternal placental syndromes (CHAMPS):
population-based retrospective cohort study. Lancet 2005;366:
1797–803.
29. Savvidou MD, Hingorani AD, Tsikas D, et al. Endothelial
dysfunction and raised plasma concentrations of asymmetric dimethylarginine in pregnant women who subsequently develop pre-eclampsia. Lancet 2003;361:
1511–17.
30. Klonoff-Cohen H, Edelstein S, Savitz D. Cigarette smoking
and preeclampsia. Obstet Gynecol 1993;81:541–4.
31. Ylikorkala O, Viinikka L, Lehtovirta P. Effect of nicotine
on fetal prostacyclin and thromboxane in humans. Obstet
Gynecol 1985;66:102–5.
Am J Epidemiol 2008;168:427–433
433
32. Levine RJ, Qian C, Yu K, et al. Circulating angiogenic factors
in smokers and non-smokers during normotensive pregnancy.
(Abstract). Am J Obstet Gynecol 2006;195(suppl):37S.
33. Reaven G, Tsao PS. Insulin resistance and compensatory hyperinsulinemia: the key player between cigarette smoking and
cardiovascular disease? J Am Coll Cardiol 2003;41:1044–7.
34. van der Vaart H, Postma DS, Timens W, et al. Acute effects
of cigarette smoke on inflammation and oxidative stress: a
review. Thorax 2004;59:713–21.
35. Frohlich M, Sund M, Lowel H, et al. Independent association
of various smoking characteristics with markers of systemic
inflammation in men. Results from a representative sample of
the general population (MONICA Augsburg Survey 1994/95).
Eur Heart J 2003;24:1365–72.
36. Klebanoff MA, Levine RJ, Clemens JD, et al. Serum cotinine
concentration and self-reported smoking during pregnancy.
Am J Epidemiol 1998;148:259–62.
37. Palta M, Prineas RJ, Berman R, et al. Comparison of selfreported and measured height and weight. Am J Epidemiol
1982;115:223–30.
38. Stewart AL. The reliability and validity of self-reported weight
and height. J Chronic Dis 1982;35:295–309.
39. Levine RJ, Hauth JC, Curet LB, et al. Trial of calcium to
prevent preeclampsia. N Engl J Med 1997;337:69–76.