Sie sind auf Seite 1von 10

Published by Oxford University Press on behalf of the International Epidemiological Association The Author 2005; all rights reserved.

. Advance Access publication 31 October 2005

International Journal of Epidemiology 2006;35:121130 doi:10.1093/ije/dyi218

Maternal smoking during pregnancy in relation to child overweight: follow-up to age 8 years
Aimin Chen,1* Michael L Pennell,2 Mark A Klebanoff,3 Walter J Rogan1 and Matthew P Longnecker1

Accepted

24 May 2005

Background Data from several studies indicate that children of mothers who smoked during pregnancy may be at a risk of overweight compared with children of non-smoking mothers. The size of this relation, however, is unclear, as is the age at which it becomes detectable. Methods Prospective data for 34 866 children enrolled in the US Collaborative Perinatal Project were analysed to examine maternal pregnancy smoking in relation to weight, height, and body mass index (BMI) in offspring at ages 1, 3, 4, 7, and 8 years. Compared with offspring of non-smokers, children of smokers had (i) weight that was lower at birth but then quickly equalled or exceeded that of non-smokers, (ii) consistently decreased height, and (iii) increased risk of overweight, particularly in girls. For example, at age 7 years, the adjusted odds ratio of BMI >85th percentile in boys of mothers who smoked on an average >20 cigarettes per day while pregnant was 1.22 (95% confidence interval 1.031.46), and in girls it was 1.30 (1.081.56).

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

Results

Conclusions In these data, maternal smoking during pregnancy was associated with a modest increase in risk of overweight in children before the age of 8 years. Keywords Smoking, pregnancy, child, obesity, body mass index

Introduction
Maternal smoking during pregnancy reduces birth weight and crown-to-heel length.1,2 Children of smoking mothers often show significant catch-up growth in weight before the age of 1 year,3,4 and then have weight similar to47 or slightly higher than810 children of non-smokers at age >5 years. In contrast to weight, the maternal smoking-related height deficit at birth does not disappear after birth. Data are fairly consistent in showing
1

Epidemiology Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, NC, USA. Biostatistics Branch, National Institute of Environmental Health Sciences, National Institutes of Health, Department of Health and Human Services, Research Triangle Park, NC, USA. 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, USA

* Corresponding author: Epidemiology Branch, National Institute of Environmental Health Sciences, Mail Drop A3-05, PO Box 12233, Research Triangle Park, NC 27709, USA. E-mail: chen17@niehs.nih.gov

continuous shorter stature for children of smoking mothers up to age 57 years or even older.6,8,1117 The potential implications of a tiny change in weight and height in offspring of smoking mothers initially received relatively little attention. Recent studies examining body mass index (BMI), however, indicate increased risk of overweight in children of smoking mothers, with odds ratios (ORs) ranging from 1.1 to 2.5.1826 In general, as the amount of cigarettes smoked increased, so did the prevalence of child overweight.20,21,24 Nonetheless, several questions about the prenatal smoking offspring BMI relation remain. Although the timing of catch-up growth is well-characterized, Ong et al.4 recently reported in a large study that prenatal smoking was not related to offspring weight and height at age 5 years, suggesting that the age at which the increased BMI becomes evident is not well understood. Furthermore, women who quit smoking early in pregnancy have babies with birth weights similar to babies of non-smokers, but for the smokingBMI relation, data on smoking effect by the trimester are inconsistent.20,23,24 Finally, previous analyses of US Collaborative Perinatal Project (CPP) data on maternal

121

122

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

pregnancy smoking in relation to child growth have been limited to fewer than 10 000 subjects,5,6 gave conflicting results, and did not explore the effect on BMI. For these reasons, we examined the association of maternal smoking in pregnancy and offspring overweight in the complete CPP data to address the questions relating to timing of effects, exposure by trimester, and dose response. The CPP was a large prospective study of pregnant women and their offspring. The CPP data were especially powerful because about half of the mothers in the CPP smoked during pregnancy. We examined the weight, height, and BMI of children of smoking and non-smoking mothers at ages 1, 3, 4, 7, and 8 years, and the risk of BMI being >85th or >95th percentile at ages 3, 4, 7, and 8 years.

trimesters (n 5 2568). Analogously, to study the effects of smoking during the third trimester, current smokers were divided into smoking information in the first and second trimesters missing due to registration in the third trimester (n 5 4659), smoker in the first or second trimester but non-smoker in the third trimester (n 5 1428), non-smoker in the first or second trimester but smoker in the third trimester (n 5 844); and smoker in both (first or second) and third trimesters (n 5 11 006).

Outcome
Most children had their weight and height measurements close to the nominal at the time of scheduled follow-up visits: 4 and 8 months, and 1, 3, 4, 7, and 8 years of age. Because this report was focused on overweight in childhood, measures from 4 to 8 months were not included in the analysis. Measurements between the ages of 1, 3, 4, 7, and 8 were sparse, and so we were unable to fit a smoothed model of the growth trajectory. Instead, we selected 1 year (1014 months), 3 years (3438 months), 4 years (4650 months), 7 years (8286 months), and 8 years (9498 months) as the time points to examine smoking effects. Subjects with observations outside a given age band were considered as not having an observation at that age (see exclusions, above). An expert panel defined child overweight as an age- and sexspecific BMI >95th percentile and at risk for overweight as a BMI >85th but ,95th percentile, according to the 2000 BMI-for-age Centers for Disease Control and Prevention (CDC) growth charts.30 In our analysis, we examined the risk of BMI >95th percentile and of BMI >85th percentile in relation to maternal smoking.

Methods
The methods of the CPP have been described in detail previously.27 In brief, the CPP was a prospective study of pregnant women and their children enrolled at 12 study centres from 1959 to 1965 in the United States. At registration and subsequent antenatal care, women were interviewed regarding the number of cigarettes smoked per day. Weight and height of offspring were measured at multiple follow-up visits. Among the 58 760 CPP pregnancies, 53 647 ended with a singleton live birth to 44 213 women. For each woman, we included only the first singleton live birth that occurred during enrollment in the study. Of the 44 213 eligible births, 34 866 children had at least one measurement of both weight and height ascertained at or near the 1, 3, 4, 7, and 8 year follow-ups (5645 exclusions), with valid maternal smoking information (196 exclusions) and without a major congenital malformation28 diagnosed before the age of 1 year (3506 exclusions).

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

Statistical analysis
We compared weight, height, and BMI between never, ex- and current smokers using all available measurements at 0, 1, 3, 4, 7, and 8 years of age. The use of Z-scores of weight, height, and BMI instead gave similar results, so only results on the original scale are given in this paper. We also compared the occurrence of BMI >95th or >85th percentile across smoking categories at ages 3, 4, 7, and 8 years. Because the data involved multiple measurements for each study subject, generalized estimating equations (GEE) were used for the final analyses (with measurements at different ages in an overall regression model).31,32 The results of models obtained using GEE were very similar to age-specific comparisons (partly due to large sample size), so only the GEE results are reported here. Covariates considered for all models were the study centre (Boston, Buffalo, New Orleans, New York Columbia, Baltimore, Richmond, Minneapolis, New York Metropolitan, Portland, Philadelphia, Providence, Memphis); mothers age at recruitment (,20, 2024, 2529, 3034, >35 years); mothers race (white, black, others); mothers marital status (single, married, separated, others); mothers education (,9 years, 911 years, 12 years, >13 years); socioeconomic index (03.9, 45.9, 69.5);33 gestational age at recruitment (first, second, third trimester); mothers prepregnancy weight, height or BMI (corresponding to the outcome variable); childs sex; childs live birth order (first, second, third or more); breastfed in nursery (yes or no); and age at measurement (in days). Centre, race, and socioeconomic index were included

Exposure
The smoking status during pregnancy was based on the selfreport at registration and at up to 32 subsequent prenatal visits. Klebanoff et al.29 have shown that in the CPP smoking during pregnancy was reported fairly accurately, based on measures of maternal serum cotinine. The median gestational age at registration was 21 weeks. Never smokers (n 5 13 306) reported never smoking. Ex-smokers (n 5 3623) reported smoking before registration, but not at registration or thereafter. Current smokers (n 5 17 937) reported smoking at registration or at subsequent prenatal visits. Smoking-dose (,10, 1019, and >20 cigarettes/day; n 5 9575, 4731, and 3631, respectively) was calculated based on mean number of cigarettes smoked per day reported at registration and subsequent prenatal visits. For the analysis of smoking effects during specific trimesters, we defined the first trimester as ,14 weeks of gestation, the second trimester as 1427 weeks, and the third trimester as >28 weeks. Because registration took place in all the three trimesters of pregnancy, to study the effects of maternal smoking during the first trimester, current smokers were divided into four subcategories: smoking information in the 1st trimester missing due to registration in the second or third trimester (n 5 14 547), smoker in the first trimester but non-smoker in the second or third (n 5 212), nonsmoker in the first trimester but smoker in the second or third (n 5 610), and smoker in both the first and the (second or third)

MATERNAL SMOKING IN RELATION TO CHILD OVERWEIGHT

123

in all models. We used models of weight and height at age 7 years to determine which covariates to include in the final models. A covariate was included if its addition to the model changed the regression coefficient of current smoker (relative to never smokers) more than 10%. The final models for all outcomes included smoking category, nominal age (0, 1, 3, 4, 7, 8 years), centre, race (and its interaction with nominal age), socioeconomic index, maternal age at recruitment, maternal prepregnancy weight/height/BMI (corresponding to the outcome variable), childs birth order, breastfeeding, and age at measurement (the difference between exact age and nominal age and its interaction with nominal age). We checked for interactions between smoking category and study centre, race, ever breastfed, sex, and birth order. We detected an interaction between sex and smoking for weight and BMI at ages 7 and 8 years, so the main analyses of weight, height, and BMI were stratified by sex. We tested whether the effects increased from age 3 to 8 years by treating a childs exact age as a continuous variable and examining its interaction with a measure of amount smoked (doseresponse, see below). To examine doseresponse using 1 degree of freedom trend tests, for each smoking category we calculated the median number of cigarettes smoked (0 for never smokers, 3.4 for ,10 cigarettes/day, 14.0 for 1019 cigarettes/day, and 21.3 for >20 cigarettes/day),34 and fit a coefficient to this variable. Ex-smokers were excluded from the trend tests. Effects of

maternal smoking during the first trimester only and third trimester only were examined separately using trimester smoking subcategories and BMI >85th percentile at ages 7 and 8 years with both sexes combined (due to the limited numbers of subjects in some categories). We also conducted secondary analyses in which we repeated the linear and logistic models after including birth weight or length. Additional adjustment for childs gestational age at birth and childs birth weight at the same time gave results similar to the models with only additional adjustment for birth weight and is thus not discussed further. Variation in results across age could have been caused, in part, by inclusion of slightly different subjects at each point. We checked this possibility by conducting a sensitivity analysis with two subsets of children with more complete data. One subset included the 5405 children with all the measurements at 0, 1, 3, 4, 7, and 8 years, and the other included the 16 653 children with at least four measurements at 0, 1, 4, and 7 years.
Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

Results
The characteristics of the subjects according to the mothers smoking status are shown in Table 1. While the mean age of the mothers was 24 years in all three categories of smoking, among women who smoked during pregnancy a smaller proportion

Table 1 Characteristics of mothers and children (n 5 34 866) by smoking status during pregnancy (mean 6 SD or %), US Collaborative Perinatal a project, 195974 Characteristics Mothers age at recruitment (years) Mothers prepregnancy weight (kg) Mothers height (cm)
b b 2 b b b

Never smoker (n 5 13 306) 24.1 6 6.4 59.2 6 11.9 160.5 6 7.0 23.0 6 4.4

Ex-smoker (n 5 3623) 24.3 6 6.4 59.9 6 12.0 161.7 6 6.7 22.9 6 4.4

Current smoker (n 5 17 937) 24.0 6 6.0 58.4 6 11.3 161.2 6 6.9 22.5 6 4.2

Mothers prepregnancy BMI (kg/m ) Mothers race White Black Others Mothers education (years) ,9 911 12 13 or more Maternal socioeconomic index 0.03.9 4.05.9 6.09.5 Child live birth order First Second Third or more Male child Child ever breastfed
a b

33.6 54.8 11.6


b

51.3 46.2 2.4

49.4 44.8 5.8

20.7 36.1 30.6 12.7


b

14.4 33.9 32.2 19.6

17.2 43.1 29.6 10.1

43.7 29.9 26.4


b

35.2 27.3 37.5

39.3 32.1 28.6

37.9 19.9 42.2 49.8


b

40.3 19.7 40.0 49.0 21.9

34.7 21.6 43.7 49.9 13.3

15.3

Sum of percentages by category may not be 100 due to rounding. P , 0.01 based on the analysis of variance or chi-square test.

124

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY


a

Table 2 Mean and adjusted difference (b and SE, relative to never smokers) of weight (grams) at specified ages by maternal smoking status, b US Collaborative Perinatal project, 195974 Sex Male Maternal smoking Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test Female
c

Birth Mean 3292 (ref) 63 (21) 126 (14) 84 (15) 149 (20) 220 (23) 100 (10) Mean 3161 (ref) 15 (21) 95 (14) 42 (15) 133 (20) 222 (24) 100 (10)

4 years Mean 16736 (ref) 97 (62) 66 (38) 53 (45) 70 (55) 101 (68) 43 (28) Mean 16195 (ref) 46 (63) 5 (41) 74 (47) 17 (61) 197 (67) 81 (29)

7 years Mean 23744 (ref) 7 (112) 21 (69) 2 ( 81) 95 (101) 17 (115) 12 (48) Mean 23167 (ref) 180 (116) 233 (74) 340 (87) 129 (108) 35 (124) 5 (52)

8 years Mean 26492 (ref) 86 (160) 87 (95) 27 (111) 181 (142) 157 (168) 98 (69) Mean 25881 (ref) 274 (160) 380 (105) 486 (121) 212 (155) 281 (191)

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

75 (77)

Adjusted for centre, race, socioeconomic status, maternal age at recruitment, prepregnancy weight, child live birth order, breastfeeding, and age at measurement. Current smoker includes those in all three smoking amount categories. The regression coefficients of ex-smoker and current smoker were from the model with never smokers, ex-smokers and current smokers. The regression coefficients of ,10 cigarettes/day, 1019 cigarettes/day, and >20 cigarettes/day were from the model with never smoker, ex-smoker, ,10 cigarettes/day, 1019 cigarettes/day, and >20 cigarettes/day. The trend test estimates were from the model with never smoker, ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day, using 0, 3.4, 14.0, and 21.3 cigarettes/day indicating each category; ex-smokers were excluded from the trend test. Number of subjects in the regression models was 16 859 (67 944 measurements: 16 593 at birth, 15 040 at age 1year, 6877 at age 3 years, 10 967 at age 4 years, 12 169 at age 7 years, 6298 at age 8 years) for males and 16 990 (68 813 measurements: 16 742 at birth, 15 203 at age 1 year, 7003 at age 3 years, 11 160 at age 4 years, 12 343 at age 7 years, 6362 at age 8 years) for females. c Trend test results are based on the regression estimate per 10 cigarettes smoked per day during pregnancy.
b

were >35 years of age. Current smokers had slightly lower prepregnancy weight, greater height, and lower prepregnancy BMI. A larger proportion of smokers and ex-smokers were white. Ex-smokers had the highest socioeconomic index while never smokers had the lowest. Women who smoked while pregnant were less likely to breastfeed. Due to missing values of mothers height (6% of 34 866 children) or weight (3% of 34 866 these children), the sample size in the multiple regression models varied slightly according to outcome (more subjects for weight comparisons, fewer for height and BMI comparisons). The number with both weight and height (or length) measures and complete data on all covariates was 31 713 (91%) at birth, 28 782 (83%) at age 1 year, 13 719 (39%) at age 3 years, 21 192 (61%) at age 4 years, 23 311 (67%) at age 7 years, and 12 416 (36%) at age 8 years. Tables 2, 3, and 4 show the adjusted differences in weight, height, and BMI for children of ex-smokers and current smokers (overall and by the amount smoked) compared with children of never smokers. For simplicity, only regression estimates at ages 0, 4, 7, and 8 years are given in the tables though all age points were included in the models fit with GEE. Among current smokers, decreased birth weight was evident at all levels of smoking. At ages 4, 7, and 8 years, boys had similar weights, whether their mothers smoked or not. The girls of smoking mothers, however, became heavier than those born to never smokers at ages 7 and 8 years. The trend test for smoking-dose showed that there was a 100 g [standard error (SE): 10] decrease per 10 cigarettes smoked per day for both boys and girls at birth, but no statistically significant trend was evident at ages 7 and 8 years. From age 3 to 8 years, the effect of amount smoked on weight did not change significantly in boys, but in girls the effects increased with age

(P , 0.05). Adding birth weight to the regression models of later weights made the estimates for current smokers larger, from age 1 to 8 years: for example, at age 8, there was a 246 g (SE 92) increase for boys and a 535 g (SE 101) increase for girls. In these models, birth weight was positively associated with childhood weight. Children of mothers who smoked during pregnancy were consistently shorter than children of never smokers, from birth through age 8 years. The trend test showed evidence of dose response at every age for both sexes. From age 3 to 8 years, the effect of amount smoked on height did not change significantly in girls, but in boys the effect increased with age (P , 0.05). Additional adjustment for birth length in the models made the difference between current smokers and never smokers smaller (and statistically non-significant); a 1 cm increase in birth length implied an expected 0.40.5 cm increase in height at ages 18 years. The BMI of children of current smokers was smaller than that of children of never smokers at birth; however, the direction of the difference was reversed for ages 18 years. From age 3 to 8 years, the effect of amount smoked on BMI increased significantly with age for both sexes (P , 0.05). With additional adjustment for birth weight, the difference in BMI between current and never smokers increased, for example, to 0.23 kg/m2 (SE: 0.04) for males and 0.34 kg/m2 (SE: 0.05) for females at age 8 years. In the weight, height, and BMI models, adjustment for socioeconomic index resulted in changes in the estimates for children of current smokers of 07%. Adjustment for prepregnancy weight, height, or BMI, however, changed these estimates more, with a range of 753%.

MATERNAL SMOKING IN RELATION TO CHILD OVERWEIGHT


a

125

Table 3 Mean and adjusted difference (b and SE, relative to never smokers) of length/height (cm) at specified ages by maternal smoking status, b US Collaborative Perinatal project, 195974 Sex Male Maternal smoking Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test Female
c

Birth Mean 50.5 (ref) 0.23 (0.07) 0.74 (0.05) 0.53 (0.05) 0.93 (0.07) 1.12 (0.08) 0.53 (0.03) Mean 49.8 (ref) 0.10 (0.07) 0.72 (0.05) 0.43 (0.05) 1.06 (0.07) 1.25 (0.08) 0.62 (0.03)

4 years Mean 102.1 (ref) 0.05 (0.12) 0.50 (0.08) 0.35 (0.09) 0.57 (0.11) 0.86 (0.13) 0.37 (0.05) Mean 101.4 (ref) 0.31 (0.12) 0.38 (0.08) 0.13 (0.09) 0.54 (0.11) 1.01 (0.13) 0.45 (0.05)

7 years Mean 121.9 (ref) 0.04 (0.15) 0.57 (0.09) 0.35 (0.11) 0.72 (0.14) 1.04 (0.15) 0.46 (0.06) Mean 121.2 (ref) 0.33 (0.15) 0.27 (0.09) 0.03 (0.11) 0.62 (0.14) 0.83 (0.16) 0.42 (0.07)

8 years Mean 127.9 (ref) 0.02 (0.17) 0.52 (0.11) 0.31 (0.12) 0.61 (0.16) 1.08 (0.18) 0.47 (0.07) Mean 127.2 (ref) 0.43 (0.17) 0.29 (0.11) 0.02 (0.13) 0.58 (0.16) 0.83 (0.20)

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

0.41 (0.08)

Adjusted for centre, race, socioeconomic status, maternal age at recruitment, maternal height, child live birth order, breastfeeding, and age at measurement. Current smoker includes those in all three smoking amount categories. The regression coefficients of ex-smoker and current smoker were from the model with never smokers, ex-smokers, and current smokers. The regression coefficients of ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day were from the model with never smoker, ex-smoker, ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day. The trend test estimates were from the model with never smoker, ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day, using 0, 3.4, 14.0, and 21.3 cigarettes/day indicating each category; ex-smokers were excluded from the trend test. Number of subjects in the regression models was 16 277 (65 984 measurements: 16 016 at birth, 14 515 at age 1year, 6871 at age 3 years, 10 641 at age 4 years, 11 696 at age 7 years, 6245 at age 8 years) for males, and 16 380 (66 776 measurements: 16 133 at birth, 14 660 at age 1 year, 6998 at age 3 years, 10 812 at age 4 years, 11 875 at age 7 years, 6298 at age 8 years) for females. c Trend test based on the regression estimate per 10 cigarettes smoked per day during pregnancy.
b

Table 4 Mean and adjusted difference (b and SE, relative to never smokers) of BMI (kg/m ) at specified ages by maternal smoking status, b US Collaborative Perinatal project, 195974 Sex Male Maternal smoking Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test Female
z

Birth Mean 12.84 (ref) 0.06 (0.04) 0.22 (0.03) 0.09 (0.03) 0.35 (0.04) 0.45 (0.04) 0.23 (0.02) Mean 12.70 (ref) 0.12 (0.04) 0.23 (0.03) 0.09 (0.03) 0.37 (0.04) 0.50 (0.05) 0.24 (0.02)

4 years Mean 16.01 (ref) 0.08 (0.04) 0.06 (0.03) 0.04 (0.03) 0.07 (0.04) 0.10 (0.05) 0.05 (0.02) Mean 15.70 (ref) 0.08 (0.04) 0.09 (0.03) 0.09 (0.03) 0.10 (0.05) 0.04 (0.05) 0.03 (0.02)

7 years Mean 15.93 (ref) 0.001 (0.05) 0.10 (0.03) 0.06 (0.04) 0.08 (0.05) 0.22 (0.06) 0.09 (0.02) Mean 15.71 (ref) 0.03 (0.06) 0.20 (0.04) 0.21 (0.04) 0.22 (0.06) 0.16 (0.06) 0.08 (0.03)

8 years Mean 16.11 (ref) 0.03 (0.04) 0.14 (0.05) 0.06 (0.05) 0.22 (0.07) 0.30 (0.08) 0.15 (0.03) Mean 15.93 (ref) 0.03 (0.08) 0.27 (0.05) 0.28 (0.06) 0.23 (0.08) 0.28 (0.10) 0.11 (0.04)

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Adjusted for centre, race, socioeconomic status, maternal age at recruitment, prepregnancy BMI, child live birth order, breastfeeding, and age at measurement. Current smoker includes those in all three smoking amount categories. The regression coefficients of ex-smoker and current smoker were from the model with never smokers, ex-smokers, and current smokers. The regression coefficients of ,10 cigarettes/day, 1019 cigarettes/day,and >20 cigarettes/day were from the model with never smoker, ex-smoker, ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day. The trend test estimates were from the model with never smoker, ,10 cigarettes/day, 1019 cigarettes/day, and >20 cigarettes/day, using 0, 3.4, 14.0, and 21.3 cigarettes/day indicating each category; ex-smokers were excluded from the trend test. Number of subjects in the regression models was 16 066 (65 208 measurements: 15 809 at birth, 14 328 at age 1 year, 6798 at age 3 years, 10 516 at age 4 years, 11 576 at age 7 years, 6181 at age 8 years) for males and 16 146 (65 925 measurements: 15 904 at birth, 14 454 at age 1 year, 6921 at age 3 years, 10 676 at age 4 years, 11 735 at age 7 years, 6235 at age 8 years) for females. c Trend test results are based on the regression estimate per 10 cigarettes smoked per day during pregnancy.
b

The ORs for overweight are shown in Table 5. For BMI >95th percentile, the doseresponse was clear at age 8 years; for 10 cigarettes/day the OR (95% CI) was 1.18 (1.011.37) in boys and 1.17 (1.011.36) in girls. For BMI >85th percentile, offspring

of current smokers had higher risk. From age 3 to 8 years, the effect of amount smoked on risk of BMI >85th percentile did not change significantly in boys, but in girls the effect increased with borderline significance (P 5 0.06). Additional adjustment for

126

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY


th th a

Table 5 Proportion of children with BMI >95 or >85 percentile of age- and sex-specific CDC growth chart and adjusted ORs and 95% b CI (relative to never smokers) at different age points, US Collaborative Perinatal project, 195974 BMI >95th percentile Sex Male Maternal smoking Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test Female
c

3 years 6.1% (ref) 1.17 (0.841.63) 1.16 (0.951.41) 1.04 (0.841.30) 1.19 (0.891.60) 1.28 (0.931.75) 1.14 (1.001.31) 5.7% (ref) 1.17 (0.831.64) 1.19 (0.961.47) 1.17 (0.931.48) 1.24 (0.901.72) 0.95 (0.631.43) 1.03 (0.881.20) 15.7% (ref) 0.93 (0.751.16) 1.23 (1.081.40) 1.16 (1.011.34) 1.31 (1.091.59) 1.44 (1.171.78) 1.17 (1.071.28) 15.5% (ref) 1.00 (0.811.23) 1.10 (0.961.25) 1.08 (0.931.25) 1.20 (0.991.46) 1.01 (0.801.28) 1.03 (0.941.15)

4 years 8.6% (ref) 0.90 (0.691.17) 1.07 (0.921.25) 1.02 (0.851.21) 1.04 (0.831.30) 1.19 (0.951.51) 1.07 (0.971.19) 6.7% (ref) 0.89 (0.661.18) 1.20 (1.011.42) 1.21 (1.001.47) 1.27 (1.001.60) 1.02 (0.781.33) 1.04 (0.941.16) 21.4% (ref) 0.84 (0.711.00) 1.08 (0.971.20) 1.01 (0.891.14) 1.13 (0.981.31) 1.24 (1.061.46) 1.10 (1.031.18) 19.0% (ref) 1.00 (0.841.19) 1.14 (1.021.27) 1.13 (1.001.28) 1.19 (1.031.40) 1.06 (0.891.26) 1.04 (0.971.12)

7 years 5.0% (ref) 1.02 (0.751.38) 1.06 (0.881.27) 1.05 (0.851.30) 0.93 (0.701.22) 1.22 (0.911.61) 1.05 (0.931.18) 5.1% (ref) 0.82 (0.601.11) 1.23 (1.031.46) 1.24 (1.021.51) 1.35 (1.061.72) 1.00 (0.751.34) 1.03 (0.931.15) 14.0% (ref) 1.03 (0.861.24) 1.09 (0.981.23) 1.03 (0.901.18) 1.13 (0.961.34) 1.22 (1.031.46) 1.10 (1.021.18) 12.9% (ref) 1.10 (0.921.33) 1.23 (1.101.38) 1.18 (1.031.35) 1.30 (1.101.53) 1.30 (1.081.56) 1.12 (1.041.21)

8 years 4.3% (ref) 1.10 (0.771.57) 1.21 (0.961.51) 1.09 (0.841.42) 1.32 (0.941.85) 1.41 (0.982.04) 1.18 (1.011.37) 4.4% (ref) 0.71 (0.471.06) 1.31 (1.061.61) 1.28 (1.011.62) 1.26 (0.921.73) 1.53 (1.072.20)

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

1.17 (1.011.36) 11.8% (ref) 1.06 (0.841.33) 1.21 (1.051.39) 1.08 (0.921.28) 1.42 (1.161.74) 1.32 (1.041.67) 1.17 (1.071.29) 11.7% (ref) 1.30 (1.041.62) 1.37 (1.191.58) 1.33 (1.131.57) 1.48 (1.211.81) 1.33 (1.041.70) 1.15 (1.051.27)

>85th percentile

Male

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Female

Never smoker Ex-smoker Current smoker ,10 cigarettes/day 1019 cigarettes/day >20 cigarettes/day Trend test
c

Adjusted for centre, race, socioeconomic status, maternal age at recruitment, prepregnancy BMI, child live birth order, breastfeeding, and age at measurement. b Current smoker includes those in all three smoking amount categories. The regression coefficients of ex-smoker and current smoker were from the model with never smokers, ex-smokers and current smokers. The regression coefficients of ,10 cigarettes/day, 1019 cigarettes/day, and >20 cigarettes/day were from the model with never smoker, ex-smoker, ,10 cigarettes/day, 1019 cigarettes/day, and >20 cigarettes/day. The trend test estimates were from the model with never smoker, ,10 cigarettes/day, 1019 cigarettes/day, >20 cigarettes/day, using 0, 3.4, 14.0, and 21.3 cigarettes/day indicating each category; ex-smokers were excluded from the trend test. Number of subjects in the regression models was 14 486 (35 071 measurements: 6798 at age 3 years, 10 516 at age 4 years, 11 576 at age 7 years, 6181 at age 8 years) for males and 14 612 (35 567 measurements: 6921 at age 3 years, 10 676 at age 4 years, 11 735 at age 7 years, 6235 at age 8 years) for females. c Trend test result s are based on the regression estimate per 10 cigarettes smoked per day during pregnancy.

birth weight in the logistic models slightly increased the ORs, by ~10%, at most age points. Higher birth weight was associated with increased risk of overweight. For example, at age 8 years, for each 1 kg increase in birth weight, the OR of BMI >85th percentile was 1.7 (95% CI: 1.61.8) for boys and 1.6 (95% CI: 1.51.8) for girls. The trimester-specific analyses showed that the greatest risk of overweight was associated with smoking later in pregnancy. For example, at age 8 years, the OR of BMI >85th percentile for offspring of women who smoked only in the third trimester was 1.38 (95% CI: 0.981.95), as compared with 1.04 (95% CI: 0.741.47) in offspring of women who smoked earlier in pregnancy but not in the third trimester. Among offspring of women who smoked throughout pregnancy, the OR was 1.41 (95% CI: 1.231.61).

While effects of prenatal smoking on height and weight both contributed to the increased BMI at older ages in the CPP, we considered whether, in the absence of an effect on height, the exposed would still have had an appreciable increase in BMI at age 8. The adjusted difference in height due to smoking (both sexes combined) was 0.46 cm. When this amount was added to the height of the children of smokers, their BMI was 0.12 kg/m2 greater than that of children of non-smokers. In contrast, when the adjusted increase in weight among smokers was subtracted from each of these children, their BMI was only 0.05 kg/m2 greater than the value among non-smoker offspring. These results suggest that the weight effect contributed more to the increased risk of overweight. The results from the 5405 children with all the measurements at 0, 1, 3, 4, 7, and 8 years or the 16 653 children with at least four

MATERNAL SMOKING IN RELATION TO CHILD OVERWEIGHT

127

Male 1000 Weight (gram) Weight (gram) 1000

Female

1000 0 1 2 3 4 5 6 7 8 9 Age (year)


Male 1.5 Height (cm) Height (cm)

1000 0 1 2 3 4 5 6 7 8 9 Age (year)


Female 1.5

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

1.5 0 1 2 3 4 5 6 7 8 9

1.5
0 1 2 3 4 5 6 7 8 9

Age (year)
Male 0.5

Age (year)

Female 0.5

BMI

BMI
0 1 2 3 4 5 6 7 8 9

0.5

0.5 0 1 2 3 4 5 6 7 8 9
Age (year)
Male BMI 85th percentile

Age (year)
Female BMI 85th percentile 10

10 Odds ratio

Odds ratio 3 4 7 8

0.1

0.1 3 4 7 8

Age (year)

Age (year)

Figure 1 Adjusted weight, height, BMI difference, and ORs of BMI >85th percentile (current smoker in pregnancy compared with never smoker) and their 95% CIs of 5405 children having all measurements, US Collaborative Perinatal Project, 195974. Adjusted for centre; race; socioeconomic status; maternal age at recruitment; prepregnancy weight, height, and BMI; child live birth order; breastfeeding; and age at measurement

measurements at 0, 1, 4, and 7 years were very similar to the results presented above. Figure 1 shows the weight, height (or length), BMI difference, and the OR of BMI >85th percentile for those whose mothers smoked during pregnancy compared with those whose mothers never smoked (5405 children with all measurements available).

Discussion
The findings from this analysis of CPP data corroborate the results of most previous studies of smoking during pregnancy in relation

to BMI among offspring,2024 and show a modest positive association. Adjustment of the smoking effect for confounding factors had relatively little impact on the associations observed in the CPP, as seen in other studies.20,21,24 As in previous studies, evidence for a doseresponse relation with BMI was present in the CPP, especially at ages 7 and 8 years. The association of smoking with BMI varied by age of offspring, especially for girls. A greater effect at later age was also seen in a British birth cohort20; the biological mechanism for this remains uncertain. It could be a cumulative result of a small change in metabolism or behaviour that becomes detectable only as the child ages.

128

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

The effect of smoking was greatest later in pregnancy, which was consistent with results of two other studies,20,24 but not a third23 in which the prevalence of smoking was lower than in the CPP. Higher birth weight3537 and post-natal catch-up growth are both associated with child overweight and adult obesity. In these data, we saw evidence of both relations. We did not show data on catch-up growth (weight growth) because its relation with child overweight is tautological. The lower birth weight among infants of prenatal smokers was associated with greater catch-up growth in the CPP, as has been seen previously.38 Such growth acceleration has been associated with increased risk of diabetes, hypertension, and cardiovascular diseases.39 Several investigators have examined the associations of prenatal smoking with blood pressure but the results are mixed.40 The persistent decrease in height associated with prenatal smoking may be due to the effects of nicotine and carbon monoxide, which have been linked with foetal growth restriction, probably through vasoconstriction and hypoxemia.41 Maternal smoking during pregnancy was shown to decrease cord blood catecholamine (both epinephrine and norepinephrine) concentrations at birth.42 Animals exposed to nicotine prenatally have decreased responsiveness to adrenergic stimuli and weigh more than unexposed controls.43 Thus, prenatal exposure to nicotine in humans may decrease responsiveness to adrenergic stimuli that release lipids from adipose tissue.44,45 In addition, maternal smoking during pregnancy may be associated with poor appetite control in offspring. Withdrawal from in utero smoking exposure may account for more weight gain after birth, as seen in smoking cessation.46 Whether leptin (an adipose cellsecreted protein hormone regulating food intake, energy expenditure, and body weight) is involved in the pathway of smoking effect remains to be determined.4749 In the large 1958 British birth cohort,20 results were stratified by sex, and the size of the smoking effect was not consistently greater in either sex during the follow-up to age 33 years. In the present study, smoking effects on weight and BMI were somewhat larger in females than males, though our data were collected only up to age 8 years. In the smaller German and Scandinavian studies,21,24 results were not stratified by sex. In a previous analysis of CPP data on maternal pregnancy smoking in relation to body size of offspring, data for 8193 subjects were included.6 The association of smoking with decreased height was clear. The association with increased weight, although present, was not statistically significant or remarked on by the authors. Their failure to detect the association with weight was possibly due to the smaller sample compared with the present study, or because their approach stratified the data by various factors before statistics were applied, effectively reducing their sample size more. In addition, they did not consider BMI. The other previous related analyses of CPP data were afflicted by similar limitations and included fewer than 300 subjects.5,6 The possible effect of environmental tobacco smoke during childhood can not be separated from that of maternal smoking during pregnancy in these data. Most of the larger studies of environmental tobacco smoke exposure during childhood show no relation with height.5054 In a German study,22 maternal smoking after pregnancy had no effect on the risk of child overweight. Furthermore, lack of information on the diet and physical activity of the CPP children was a drawback. Children of

smoking mothers may snack more or be more sedentary, and these may be associated with child overweight.21 In the study where these were adjusted,21 it appeared that they did not materially change the estimates of the prenatal smoking offspring BMI association. Those lost to follow-up were similar to those who remained in the study with regard to demographic characteristics and smoking categories. Across all smoking categories, those lost to follow-up had slightly lower BMI, and so no differential loss of low BMI girls in current smokers was observed. Smoking among those with more complete data, i.e. tests at 0, 1, 4, and 7 years at least, was similar to those with fewer follow-ups. Analysis of two subsets with more complete follow-up yielded results similar to the primary analysis, so the possibility that the associations we observed were due to differential loss to follow-up seems unlikely. Cigarettes have changed since the 1960s,55 and thus the relation of smoking during pregnancy with childhood overweight may also have changed. Indeed, the association of maternal pregnancy smoking with overweight in offspring appears stronger in more recent studies21,24 than in the present study or the 1958 cohort.20 Whether effect size has truly increased over time or whether some constant effect size has been modified by variation in the level of other risk factors is unclear. The prevalence of smoking among pregnant women in the US was greater in the 1960s than at present, and the decline has been greater among those of higher social class.56 Whether the overall population-attributable risk for child overweight owing to maternal pregnancy smoking might have dropped over time is unclear given the possible trends in effect size noted above. Nonetheless, because smoking prevalence is now greater among those with lower socioeconomic status than for those with higher socioeconomic status, we might expect the population attributable risk to be greater in the former group. On a percentage basis, the effects of prenatal smoking on height and weight were both small (e.g. ,1% for current smoking at 8 years). For the individual child, a few hundred grams in weight or 0.5 cm in height at age 7 or 8 years may not be important; however, even a modest increase of 1040% in the chance of overweight, such as seen in the CPP, is important, given the health consequences of this condition. Although the prevalence of smoking during pregnancy in the US has dropped markedly from ~50% in the 1950s to 12% in 2000,57,58 it remains a major public health problem in the US given the current prevalence of smoking in pregnancy: ORs of 1.11.4 are not negligible in terms of absolute effects on childhood overweight. Although few of the children in CPP were followed past childhood, there is substantial evidence that childhood overweight and adult obesity are linked.59,60 Public health recommendations already discourage smoking among youth, pregnant women, and others.61 Nevertheless, the doseresponse relation in the present data and similar findings among the other large studies of this association support the notion that prenatal exposure to cigarette smoke may have lifelong adverse consequences on health via an effect on body habitus. Smoking during pregnancy increases spontaneous abortion, perinatal mortality, and neonatal and infant morbidity,62 and its association with overweight in offspring is a further reason not to smoke.

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

MATERNAL SMOKING IN RELATION TO CHILD OVERWEIGHT

129

Acknowledgement
The authors thank Dr Beth C Gladen at the National Institute of Environmental Health Sciences for her contribution to the statistical analysis and her comments on earlier versions of the paper.

KEY MESSAGES     In these data maternal smoking during pregnancy was a risk factor for child overweight. The maternal smokingchild overweight association was observed before the age of 8 years in offspring. Maternal smoking in the third trimester of pregnancy was more strongly associated with child overweight than maternal smoking in the first trimester of pregnancy. There was a general doseresponse trend in the maternal smokingchild overweight association.
16

References
1

US Department of Health and Human Services. The Health Consequences of Smoking: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2004. Floyd RL, Rimer BK, Giovino GA, Mullen PD, Sullivan SE. A review of smoking in pregnancy: effects on pregnancy outcomes and cessation efforts. Ann Rev Public Health 1993;14:379411. Conter V, Cortinovis I, Rogari P, Riva L. Weight growth in infants born to mothers who smoked during pregnancy. BMJ 1995;310:76871. Ong KK, Preece MA, Emmett PM, Ahmed ML, Dunger DB. Size at birth and early childhood growth in relation to maternal smoking, parity and infant breast-feeding: longitudinal birth cohort study and analysis. Pediatr Res 2002;52:86367. Hardy JB, Mellits ED. Does maternal smoking during pregnancy have a long-term effect on the child? Lancet 1972;2:133236. Naeye RL. Influence of maternal cigarette smoking during pregnancy on fetal and childhood growth. Obstet Gynecol 1981;57:1821. MacArthur C, Knox EG, Lancashire RJ. Effects at age nine of maternal smoking in pregnancy: experimental and observational findings. BJOG 2001;108:6773. Vik T, Jacobsen G, Vatten L, Bakketeig LS. Pre- and post-natal growth in children of women who smoked in pregnancy. Early Hum Dev 1996;45:24555. Fried PA, Watkinson B, Gray R. Growth from birth to early adolescence in offspring prenatally exposed to cigarettes and marijuana. Neurotoxicol Teratol 1999;21:51325. Fried PA, James DS, Watkinson B. Growth and pubertal milestones during adolescence in offspring prenatally exposed to cigarettes and marihuana. Neurotoxicol Teratol 2001;23:43136. Goldstein H. Factors influencing the height of seven year old children results from the National Child Development Study. Hum Biol 1971;43:92111. Rantakallio P. A follow-up study up to the age of 14 of children whose mothers smoked during pregnancy. Acta Paediatr Scand 1983;72: 74753. Elwood PC, Sweetnam PM, Gray OP, Davies DP, Wood PD. Growth of children from 0-5 years: with special reference to mothers smoking in pregnancy. Ann Hum Biol 1987;14:54357. Fogelman KR, Manor O. Smoking in pregnancy and development into early adulthood. BMJ 1988;297:123336. Fox NL, Sexton M, Hebel JR. Prenatal exposure to tobacco: I. Effects on physical growth at age three. Int J Epidemiol 1990;19:6671.

Jones G, Riley M, Dwyer T. Maternal smoking during pregnancy, growth, and bone mass in prepubertal children. J Bone Miner Res 1999;14:14651. Blake KV, Gurrin LC, Evans SF et al. Maternal cigarette smoking during pregnancy, low birth weight and subsequent blood pressure in early childhood. Early Hum Dev 2000;57:13747. von Kries R, Koletzko B, Sauerwald T et al. Breast feeding and obesity: cross sectional study. BMJ 1999;319:14750. Montgomery SM, Ekbom A. Smoking during pregnancy and diabetes mellitus in a British longitudinal birth cohort. BMJ 2002;324:2627. Power C, Jefferis BJ. Fetal environment and subsequent obesity: a study of maternal smoking. Int J Epidemiol 2002;31:41319. von Kries R, Toschke AM, Koletzko B, Slikker W Jr. Maternal smoking during pregnancy and childhood obesity. Am J Epidemiol 2002;156: 95461. Toschke AM, Koletzko B, Slikker W Jr, Hermann M, von Kries R. Childhood obesity is associated with maternal smoking in pregnancy. Eur J Pediatr 2002;161:44548. Toschke AM, Montgomery SM, Pfeiffer U, von Kries R. Early intrauterine exposure to tobacco-inhaled products and obesity. Am J Epidemiol 2003;158:106874. Wideroe M, Vik T, Jacobsen G, Bakketeig LS. Does maternal smoking during pregnancy cause childhood overweight? Paediatr Perinat Epidemiol 2003;17:17179. Toschke AM, Vignerova J, Lhotska L, Osancova K, Koletzko B, Von Kries R. Overweight and obesity in 6- to 14-year-old Czech children in 1991: protective effect of breast-feeding. J Pediatr 2002;141:76469. Bergmann KE, Bergmann RL, Von Kries R et al. Early determinants of childhood overweight and adiposity in a birth cohort study: role of breast-feeding. Int J Obes Relat Metab Disord 2003;27:16272. Niswander KR, Gordon M. The Women and Their Pregnancies: The Collaborative Perinatal Study of the National Institute of Neurological Diseases and Stroke. Washington, DC: US Goverment Print Office, 1972. Myrianthopoulos NC, Chung CS. Congenital malformations in singletons: epidemiologic survey. Report from the Collaborative Perinatal project. Birth Defects Orig Artic Ser 1974;10:158. Klebanoff MA, Levine RJ, Clemens JD, DerSimonian R, Wilkins DG. Serum cotinine concentration and self-reported smoking during pregnancy. Am J Epidemiol 1998;148:25962. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM et al. CDC growth charts: United States. Adv Data 2000;127. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika 1986;73:1322. Stokes ME, Davis CS, Koch GG. Categorical Data Analysis Using the SAS System. 2nd edn. Cary, NC: SAS Institute Inc., 2000.

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

17

18

19

20

21

22

23

24

25

26

10

27

11

28

12

29

13

30

14

31

15

32

130
33

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY


48

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;2:28399. Greenland S. Avoiding power loss associated with categorization and ordinal scores in doseresponse and trend analysis. Epidemiology 1995;6:45054. Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal gestational diabetes, birth weight, and adolescent obesity. Pediatrics 2003;111:e22126. Curhan GC, Willett WC, Rimm EB, Spiegelman D, Ascherio AL, Stampfer MJ. Birth weight and adult hypertension, diabetes mellitus, and obesity in US men. Circulation 1996;94:324650. Curhan GC, Chertow GM, Willett WC et al. Birth weight and adult hypertension and obesity in women. Circulation 1996;94:131015. Ong KK, Ahmed ML, Emmett PM, Preece MA, Dunger DB. Association between postnatal catch-up growth and obesity in childhood: prospective cohort study. BMJ 2000;320:96771. Singhal A, Lucas A. Early origins of cardiovascular disease: is there a unifying hypothesis? Lancet 2004;363:164245. Pausova Z, Paus T, Sedova L, Berube J. Prenatal exposure to nicotine modifies kidney weight and blood pressure in genetically susceptible rats: a case of gene-environment interaction. Kidney Int 2003;64: 82935. Lambers DS, Clark KE. The maternal and fetal physiologic effects of nicotine. Semin Perinatol 1996;20:11526. Oncken CA, Henry KM, Campbell WA, Kuhn CM, Slotkin TA, Kranzler HR. Effect of maternal smoking on fetal catecholamine concentrations at birth. Pediatr Res 2003;53:11924. Levin ED. Fetal nicotine overload, blunted sympathetic responsivity, and obesity. Birth Defects Res A Clin Mol Teratol 2005;73:48184. Levin ED, Slotkin TA. Developmental neurotoxicity of nicotine. In: Slikker W Jr, Chang LW (eds). Handbook of Developmental Neurotoxicology. San Diego: Academic Press, 1998, pp. 587615. Schiffelers SL, Saris WH, Boomsma F, van Baak MA. beta(1)and beta(2)-Adrenoceptor-mediated thermogenesis and lipid utilization in obese and lean men. J Clin Endocrinol Metab 2001;86:219199. Toschke AM, Ehlin AG, von Kries R, Ekbom A, Montgomery SM. Maternal smoking during pregnancy and appetite control in offspring. J Perinat Med 2003;31:25156. Pardo I, Geloneze B, Tambascia M, Barros-Filho A. Does maternal smoking influence leptin levels in term, appropriate-for-gestationalage newborns? J Matern Fetal Neonatal Med 2004;15:40810.

Helland IB, Reseland JE, Saugstad OD, Drevon CA. Smoking related to plasma leptin concentration in pregnant women and their newborn infants. Acta Paediatr 2001;90:28287. Mantzoros CS, Varvarigou A, Kaklamani VG, Beratis NG, Flier JS. Effect of birth weight and maternal smoking on cord blood leptin concentrations of full-term and preterm newborns. J Clin Endocrinol Metab 1997;82:285661. Berkey CS, Ware JH, Speizer FE, Ferris BG Jr. Passive smoking and height growth of preadolescent children. Int J Epidemiol 1984; 13:45458. Ferris BG Jr, Ware JH, Berkey CS, Dockery DW, Spiro A III, Speizer FE. Effects of passive smoking on health of children. Environ Health Perspect 1985;62:28995. Rona RJ, Chinn S, Florey CD. Exposure to cigarette smoking and childrens growth. Int J Epidemiol 1985;14:40209. Chinn S, Rona RJ. Quantifying health aspects of passive smoking in British children aged 511 years. J Epidemiol Community Health 1991;45:18894. Eskenazi B, Bergmann JJ. Passive and active maternal smoking during pregnancy, as measured by serum cotinine, and postnatal smoke exposure. I. Effects on physical growth at age 5 years. Am J Epidemiol 1995;142:S1018. Hoffmann D, Hoffmann I, El-Bayoumy K. The less harmful cigarette: a controversial issue. a tribute to Ernst L. Wynder. Chem Res Toxicol 2001;14:76790. Ventura SJ, Martin JA, Curtin SC, Menacker F, Hamilton BE. Births: final data for 1999. Natl Vital Stat Rep 2001;49:1100. Underwood PB, Kesler KF, OLane JM, Callagan DA. Parental smoking empirically related to pregnancy outcome. Obstet Gynecol 1967;29:18. Ventura SJ, Hamilton BE, Mathews TJ, Chandra A. Trends and variations in smoking during pregnancy and low birth weight: evidence from the birth certificate, 19902000. Pediatrics 2003;111:117680. Guo SS, Roche AF, Chumlea WC, Gardner JD, Siervogel RM. The predictive value of childhood body mass index values for overweight at age 35 y. Am J Clin Nutr 1994;59:81019. Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T. Do obese children become obese adults? A review of the literature. Prev Med 1993;22:16777. US Department of Health and Human services. Healthy People 2010: Understanding and Improving Health. 2nd edn. Washington, DC: US Government Printing Office, 2000. Higgins S. Smoking in pregnancy. Curr Opin Obstet Gynecol 2002; 14:14551.

34

49

35

50

36

51

37

52 53

38

39

54

40

Downloaded from http://ije.oxfordjournals.org/ by guest on March 27, 2013

55

41

56 57 58

42

43

44

59

45

60

46

61

47

62