American Journal of Epidemiology Advance Access originally published online on October 31, 2007
American Journal of Epidemiology 2007 166(12):1359-1364; doi:10.1093/aje/kwm272
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ORIGINAL CONTRIBUTIONS |
Associations of Prepregnancy Cardiovascular Risk Factors with the Offspring's Birth Weight
1 Department of Public Health, Norwegian University of Science and Technology, Trondheim, Norway
2 Department of Social Medicine, University of Bristol, Bristol, United Kingdom
Correspondence to Dr. Pål R. Romundstad, Department of Public Health, Faculty of Medicine, NTNU, 7489 Trondheim, Norway (e-mail: paalr{at}ntnu.no).
Received for publication May 22, 2007. Accepted for publication August 22, 2007.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Low birth weight of offspring has been associated with increased risk of maternal cardiovascular mortality, and cardiovascular risk factors measured within pregnancy have been related to offspring birth weight. It is not clear whether cardiovascular risk factors assessed prior to pregnancy are associated with the offspring's birth weight. The authors combined baseline data from 3,461 women in the HUNT Study (1995–1997) and data on deliveries from the Medical Birth Registry of Norway up to 2005. They used linear regression to prospectively study associations between diastolic and systolic blood pressures, concentrations of triglycerides, serum total cholesterol, and high density lipoprotein cholesterol measured before conception and birth weight for gestational age of the offspring. Blood pressure measured before pregnancy was inversely associated with birth weight for gestational age, whereas unfavorable levels of serum total cholesterol, high density lipoprotein cholesterol, triglycerides, and glucose were positively associated with birth weight for gestational age. Thus, women with relatively high blood pressure tend to deliver small babies, whereas women with unfavorable lipid levels tend to give birth to large babies, suggesting reduced glucose tolerance. These findings suggest that low as well as high birth weight of the offspring may indicate increased cardiovascular risk for the mother.
birth weight; blood pressure; cardiovascular diseases; glucose; lipids; pregnancy complications
Abbreviations: BMI, body mass index; HDL, high density lipoprotein
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Low birth weight for gestational age in offspring has been related to increased risk of cardiovascular disease for mothers and, to a lesser extent, for fathers (1, 2). The findings for fathers can be interpreted as indicating that common genetic factors are related to both unfavorable birth outcomes and greater cardiovascular risk. The stronger associations for mothers, however, suggest that their higher risk could be generated during pregnancies that result in unfavorable outcomes, perhaps indicating that pregnancy serves as a metabolic stressor that provides early evidence for later metabolic dysregulation (3).
There is evidence that within-pregnancy cardiovascular risk factors such as high blood pressure, obesity, and high serum cholesterol and low high density lipoprotein (HDL) cholesterol predict birth outcomes such as relatively low birth weight for gestational age (4–6). However, it is also possible that adverse levels of these risk factors could be evident before pregnancy. If the latter is true, then it would be expected that prepregnancy cardiovascular risk factors predict birth weight, but no study has, to our knowledge, comprehensively examined this question. We therefore investigated the association of cardiovascular risk factors assessed prior to pregnancy with birth weight of the offspring in a population-based cohort study in Norway.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The present study was based on a linkage between data for women participating in the HUNT-2 study and in the Medical Birth Registry of Norway. The HUNT studies consist of two main cross-sectional waves of residents aged 20 years or more in the county of Nord-Trøndelag, Norway. The second wave (HUNT-2, 1995–1997) included 66,035 individuals and constituted 71.2 percent of the total adult population. Standardized measurements were made of height, weight, blood pressure, and nonfasting blood glucose and serum lipids, and information on several lifestyle factors, such as smoking history, socioeconomic position, and educational attainment, was collected by using questionnaires (7).
Since 1967, all deliveries of infants in Norway (>2 million births) have been recorded in the Medical Birth Registry of Norway. Registration is based on standardized forms completed by midwives at the delivery ward within 1 week of delivery. The reporting is mandatory and covers virtually all births in Norway of more than 16 weeks of gestation. The form gives information related to the mother's health before and during pregnancy, complications during pregnancy and at birth, as well as perinatal data for the child.
All women for whom a birth was registered (gestational age >22 weeks or birth weight above 500 g) in the Medical Birth Registry of Norway since their participation in the HUNT-2 study and up to 2005 were identified (n = 4,461). We excluded those with an ongoing pregnancy at the time of the HUNT-2 examinations (n = 685) and 105 for whom data were missing on essential factors from the HUNT-2 study, leaving 3,671 eligible women. Of these women, we excluded those with deliveries for which there were no data on gestational age or birth weight (n = 112), twin pregnancies (n = 83), and deliveries with recorded unrealistic combinations of gestational age and birth weight (n = 15). Altogether, we included 3,461 women with 4,859 singleton births in the final analyses.
In the HUNT Study, blood pressure was measured by using an automatic oscillometric method (Dinamap 845XT; Criticon, Tampa, Florida). The pressure was measured after a minimum of 2 minutes' rest in the sitting position, and three consecutive standardized blood pressure measurements were recorded at 1-minute intervals. The mean of the second and third readings was used in the analysis. Blood sampling was performed nonfasting, and levels of serum total cholesterol, HDL cholesterol, triglycerides, and glucose were analyzed subsequent to sampling (7). Time (hours) since the last meal was reported prior to blood sampling. Glucose was measured at the testing site, and HDL cholesterol was measured after precipitation with phosphor tungsten and magnesium ions. Triglycerides were measured with an enzymatic colorimetric method. The day-to-day coefficients of variation were 1.3–2.0 percent for glucose, 1.3–1.9 percent for cholesterol, 2.4 percent for HDL cholesterol, and 0.7–1.3 percent for triglycerides.
The study was approved by the regional committee for medical research ethics and by the Norwegian Data Inspectorate.
We calculated the z score of birth weight according to sex and gestational week by using standards estimated from births registered in the Medical Birth Registry of Norway for the period 1988–1998 (8). The z score indicates the standard deviations of the offspring's birth weight above (positive values) or below (negative values) the expected mean, according to sex and length of gestation. In the present study, we used differences in z score of birth weight for gestational age as the outcome measure. For categorical variables, this difference is given in absolute terms compared with the reference category. A difference of one standard deviation (i.e., z score = 1) corresponds to about 450 g at term for both sexes.
First, we investigated associations of possible confounding factors with offspring birth weight for gestational age. We then analyzed potential associations of prepregnancy cardiovascular risk factors with the offspring's birth weight for gestational age. We adjusted for maternal age at participation in the HUNT-2 study, maternal height (continuous), parity (0, 1, 
2), offspring sex, time interval between participation in HUNT-2 and delivery (continuous), smoking status at HUNT-2 (never, former, current smoking), socioeconomic position (employed/student/housewife vs. unemployed/receives social security benefits), and maternal education (<10, 10–12, 13–15, 16 years). In the analyses of serum lipids, we also adjusted for time since the last meal. In additional analyses, we adjusted for prepregnancy body mass index (BMI). For the analyses, we used random-effects linear regression (Stata software, version 9, Stata Corporation, College Station, Texas) to account for multiple deliveries within the same mother (correlated outcome).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
At the HUNT-2 study, the mean age of the 3,461 eligible women was 26.6 years; at delivery, the mean age was 30.2 years. For some women (n = 105), data were missing from the HUNT-2 screening, but mean birth weight of the first offspring for this group (3,613 g) was similar to that for offspring of mothers for whom data were complete (3,595 g; p = 0.8).
As expected, lower birth weight for gestational age was associated with current smoking, shorter height, no previous birth, young age at delivery, low maternal BMI, and low level of education (table 1).
|
In crude analyses, higher levels of maternal systolic blood pressure were associated with lower birth weight for gestational age (table 2). After adjustment for age, smoking status, height, parity, socioeconomic position, and time interval from measurement at HUNT-2 until delivery, the associations were attenuated. However, after additional adjustment for prepregnancy BMI, the associations of high maternal systolic blood pressure with low birth weight for gestational age were further strengthened. Prepregnancy diastolic blood pressure showed similar associations, but the results were less pronounced than for systolic blood pressure.
|
In contrast to blood pressure, triglyceride levels were positively associated with birth weight for gestational age (table 2). This association was strengthened in the adjusted model but was attenuated toward the null after additional adjustment for maternal prepregnancy BMI. The crude analysis of total cholesterol showed a positive association with birth weight for gestational age. Similarly to that for triglycerides, the association between cholesterol and birth weight for gestational age was stronger in the adjusted analyses, and this association was less attenuated after additional adjustment for prepregnancy BMI (table 2). Low levels of HDL cholesterol were associated with high birth weight for gestational age in both crude and adjusted analyses. Additional adjustment for BMI led to attenuation of this association as well. Prepregnancy blood glucose was positively associated with birth weight for gestational age, and adjustment for potentially confounding factors did not substantially influence this association.
We also performed stratified analyses by smoking status and by prepregnancy BMI. The associations were less pronounced for prepregnancy current smokers compared with nonsmokers and former smokers (table 3). When we stratified by BMI, the associations for diastolic blood pressure and triglycerides tended to be stronger for overweight women (BMI 25 kg/m2) compared with women with a prepregnancy BMI of less than 25 kg/m2 (table 4). However, the opposite was found for glucose and HDL cholesterol. For triglycerides, the association changed qualitatively from a negative association for women with a BMI of less than 25 kg/m2 to a positive association for overweight women (table 4). None of the interactions studied were significant.
|
|
In additional analyses, we excluded pregnancies complicated by preeclampsia (n = 153). Although the negative association between blood pressure and birth weight for gestational age was slightly attenuated after excluding pregnancies complicated by preeclampsia, the associations with lipids were stable or slightly strengthened (results not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
We found that smoking and higher blood pressure prior to pregnancy were associated with lower birth weight for gestational age in the offspring. However, unfavorable levels of prepregnancy serum total cholesterol, HDL cholesterol, triglycerides, and glucose were associated with higher birth weight for gestational age.
The prospective design of this population-based study makes bias an unlikely explanation for the findings. However, the results for serum lipids and especially for glucose may be attenuated because of nondifferential measurement error related to nonfasting blood sampling. Our findings are in line with those from studies of maternal factors measured within pregnancy, but the strength of the associations observed in the present study appears to be weaker than the associations observed in studies of cardiovascular risk factors measured within pregnancy (4–6).
The direction of the associations suggests that glucose and triglycerides levels, which are immediate indicators of impaired glucose tolerance, are positively associated with the offspring's birth weight for gestational age. This finding is compatible with substantial evidence that impaired glucose tolerance and diabetes are related to higher birth weight of the offspring. For blood pressure, the association was in the opposite direction. Elevated blood pressure before pregnancy was related to lower birth weight of the offspring, which would elevate cardiovascular disease risk for mothers of offspring of low birth weight for gestational age, whereas the association of impaired glucose tolerance with higher birth weight of the offspring would generate an association in the opposite direction. The overall direction of association would therefore depend upon the balance of these two countervailing trends and would lead to different expectations with respect to the overall association of offspring birth weight and cardiovascular disease risk among populations with different prevalences of impaired glucose tolerance.
|
ACKNOWLEDGMENTS |
|---|
This study was financially supported by the Norwegian University of Science and Technology and by a grant from the Norwegian Medical Research Council. The HUNT study is a collaborative effort of the Faculty of Medicine, The University of Science and Technology, The Norwegian Institute of Public Health, and the Nord-Trøndelag County Council.
Conflict of interest: none declared.
|
NOTES |
|---|
Editor's note: An invited commentary on this article is published on page 1365.
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
- Davey Smith G, Hart C, Ferrell C, et al. Birth weight of offspring and mortality in the Renfrew and Paisley study: prospective observational study. BMJ (1997) 315:1189–93.
[Abstract/Free Full Text] - Davey Smith G, Sterne J, Tynelius P, et al. Birth weight of offspring and subsequent cardiovascular mortality of the parents. Epidemiology (2005) 16:563–9.[CrossRef][Web of Science][Medline]
- Sattar N, Greer IA. Pregnancy complications and maternal cardiovascular risk: opportunities for intervention and screening? BMJ (2002) 325:157–60.
[Free Full Text] - Roberts CL, Algert CS, Morris JM, et al. Hypertensive disorders in pregnancy: a population-based study. Med J Aust (2005) 182:332–5.[Web of Science][Medline]
- Sattar N, Greer IA, Galloway PJ, et al. Lipid and lipoprotein concentrations in pregnancies complicated by intrauterine growth restriction. J Clin Endocrinol Metab (1999) 84:128–30.
[Abstract/Free Full Text] - Claussen T, Burski TK, Øyen N, et al. Maternal anthropometric and metabolic factors in the first half of pregnancy and risk of neonatal macrosomia in term pregnancies. A prospective study. Eur J Endocrinol (2005) 153:887–94.
[Abstract/Free Full Text] - Holmen J, Midthjell K, Krüger Ø, et al. The Nord-Trøndelag Health Study 1995 –97 (HUNT 2): objectives, contents, methods and participation. Norsk Epidemiologi (2003) 13:19–32.
- Skjærven R, Gjessing HK, Bakketeig LS. Birth weight by gestational age in Norway. Acta Obstet Gynecol Scand (2000) 79:440–9.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in Am. J. Epidemiol.:
- Invited Commentary: Timing and Types of Cardiovascular Risk Factors in Relation to Offspring Birth Weight
- Roberta B. Ness and Janet Catov
Am. J. Epidemiol. 2007 166: 1365-1367.[Abstract] [FREE Full Text]
This article has been cited by other articles:
|
|
 |
|
  R. B. Ness and J. Catov Invited Commentary: Timing and Types of Cardiovascular Risk Factors in Relation to Offspring Birth Weight Am. J. Epidemiol., December 15, 2007; 166(12): 1365 - 1367. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||