| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Atherosclerosis and Lipoproteins |
From the MRC Epidemiology Resource Centre (C.R.G., B.J., S.M.R., K.M.G., C.N.M.), University of Southampton; and Centre for Paediatric Epidemiology and Biostatistics (C.M.L.), Institute of Child Health, University College London, UK.
Correspondence to Dr Gale, MRC Epidemiology Resource Centre, Southampton General Hospital, Southampton, Hants SO16 6YD, UK. E-mail crg{at}mrc.soton.ac.uk
| Abstract |
|---|
|
|
|---|
Methods and Results We measured carotid intimamedia thickness (IMT) in 216 nine-year-old children whose mothers had participated in a study of nutrition during pregnancy. IMT was greater in boys, in children who were heavier, in those with higher systolic blood pressure, and in those who took less exercise. Increased IMT was associated with a lower maternal energy intake in early (P=0.029) or late (P=0.006) pregnancy, after adjustment for these factors. Mean IMT of children whose mothers were in the lowest quarter of the distribution of energy intake in late pregnancy was 0.027 mm (95% confidence interval, 0.004 to 0.049) greater than that of those whose mothers were in the highest quarter of the distribution.
Conclusion Lower maternal energy intake during pregnancy may increase the susceptibility to atherogenesis of the child.
A restricted maternal diet during gestation can lead to endothelial dysfunction in animals. We studied the relation between maternal diet during pregnancy and IMT in children. IMT was greater in children whose mothers had lower energy intakes. Lower maternal energy intake during pregnancy may increase the susceptibility to atherogenesis of children.
Key Words: carotid arteries atherosclerosis pregnancy diet children
| Introduction |
|---|
|
|
|---|
Animal models suggest that maternal diet during pregnancy influences susceptibility to atherosclerosis. Rats exposed during gestation to maternal energy- or protein-restricted diets have an increased incidence of endothelial dysfunction and hypertension.58 Similar outcomes have been observed in the offspring of rats fed a high-fat diet during pregnancy.9,10 Whether there is any relation in humans between characteristics of maternal diet during pregnancy and atherogenesis is not known, although the observation that maternal hypercholesterolemia during pregnancy is linked with faster progression of early atherosclerotic lesions in children suggests that the type and quantity of fat in the maternal diet may play a role in fetal atherogenesis.11
High-resolution B-mode ultrasound measurements of carotid intimamedia thickness (IMT) are widely used as a marker of early atherosclerosis in adults12,13 and predict vascular events across a wide age range.13 Fewer studies have investigated IMT in children, but children who are obese, hypercholesterolemic, or diabetic have an increased IMT.1416 Exposure to cardiovascular risk factors in early life has long-term influences on vascular function, predicting IMT and arterial elasticity in adulthood.17,18 These findings suggest that IMT measured in childhood may indicate the effect of early risk factor exposure on atherogenesis.19
We investigated the relation between maternal diet and IMT in children whose mothers had participated in a study of nutrition during pregnancy.
| Methods |
|---|
|
|
|---|
16 years with singleton pregnancies at <17 weeks of gestation were recruited at the Princess Anne Maternity Hospital in Southampton, UK; diabetics and those who had undergone hormonal treatment to conceive were excluded.20 In early (15 weeks of gestation) and late (32 weeks of gestation) pregnancy, a food frequency questionnaire (FFQ) that assessed frequency of consumption of 100 foods or food groups in the 3 preceding months was administered. The nutrient content of a standard portion of each food was multiplied by reported frequency of use to calculate average daily nutrient intake. Nutrient intakes in the first 3 months of pregnancy assessed with the FFQ were validated against prospective 4-day food diaries early in the second trimester.21 Correlation coefficients between nutrient intakes assessed by the FFQ and diaries ranged from 0.36 to 0.58 (all P<0.001). Agreement between the FFQ and diary was much stronger in women who were not nauseous in the first trimester, suggesting that the correlations for the whole group underestimate the true level of agreement.21 Women were asked about smoking and whether they took strenuous exercise, defined as sporting or other activity during which they became out of breath or sweaty. Five hundred fifty-nine children were followed-up at age 9 months, when data on weight and infant feeding were recorded.
When these children approached age 9 years, we wrote to the parents of those still living in Southampton, inviting the children to participate in a further study. Of 461 invited, 216 (47%) agreed to attend a clinic. They did not differ significantly from the remainder of the cohort in mean birthweight (3.32 versus 3.40 kg, P=0.10), proportion from manual social class (26% versus 34%, P=0.14), or maternal macronutrient intakes during pregnancy (all P>0.5).
At the clinic, the children sat in a temperature-controlled room (20°C±2°C) for at least 10 minutes. The ultrasonographer measured IMT in the distal portion of the right common carotid artery using an Acuson XP128 scanner and a 7-MHz linear-array transducer, while the child was recumbent. Three longitudinal views, including lateral, antero-posteral, and antero-oblique sections, were frozen at the end of the diastolic phase. IMT of the far wall was measured approximately 10 mm proximal to the beginning of bifurcation, using CVI software (National Instruments, Austin, Tex). Blood pressure was measured 3 times on the left arm placed at the level of the heart using a Dinamap 1846 (Critikon, UK), with cuff sizes based on middle upper-arm circumference.22 The means of the 3 measurements of IMT and blood pressure were used in the analysis. Height was measured using a stadiometer and weight using digital scales (SECA Model No. 835). Mothers were asked whether in the previous week, their child had played sport or exercised for at least 15 minutes, excluding school lessons.
Reproducibility of IMT measurements by the single ultrasonographer was assessed before the study started in 21 volunteers using 2 repeated measurements made a week apart. The intermeasurement coefficient of variation (the SD of repeated measurements as a percentage of the mean) was 1.8%.
The average maternal basal metabolic rate (BMR) was estimated from age, height, and prepregnancy weight.23
We used ANOVA, t, or
2 tests to examine the characteristics of the participants. Pearson correlation coefficients were used to examine the relations between characteristics and IMT. Linear regression was used to calculate mean differences in IMT according to quarters of the distribution of nutrient intakes and other characteristics. Where necessary, variables were transformed using logarithms to satisfy statistical assumptions of normality.
The Local Research Ethics Committee approved the study. The children and their parents gave written informed consent.
| Results |
|---|
|
|
|---|
|
|
Mean (SD) carotid IMT was 0.34 (0.06) mm (range 0.21 to 0.50 mm) and tended to be greater in boys than in girls (0.35 mm versus 0.34 mm, P=0.055). Subsequent analyses are presented for boys and girls combined, adjusting for sex. Tests of interaction were not significant (all P>0.2).
There were no associations between IMT and maternal social class, prepregnancy BMI, or smoking during pregnancy (Table 3). Mean IMT was slightly smaller in children whose mothers had undergone strenuous exercise in late pregnancy. Maternal age (P=0.29) and height (P=0.20) were not related to IMT (data not shown). IMT was greater in children who were heavier at age 9 years. There was a similar association with BMI (r=0.21, P=0.002). Children who were heavier at age 9 years tended to have been heavier at birth (r=0.34, P<0.001) and at 9 months (r=0.45, P<0.001), but neither measure was associated with IMT after adjustment for current weight (P>0.5). There was no difference in IMT according to duration of breastfeeding. IMT was greater in children with higher systolic blood pressure, although there was no relation with diastolic pressure. Children who participated in sport or exercise outside of school had a smaller mean IMT.
|
IMT was greater in children whose mothers had intakes of energy, protein, fat, and carbohydrate that were lower either in early or late pregnancy (Table 4), although the proportion of energy obtained from individual macronutrients was not related to IMT. The relation between lower maternal energy intake and increased IMT in the child was statistically significant when energy intake was expressed as a proportion of maternal basal metabolic rate in pregnancy: children whose mothers had a lower energy intake in relation to their basal metabolic rate had an increased IMT.
|
Twenty-four mothers reported severe nausea during pregnancy. These women had a lower mean energy intake. Severe nausea was not associated with IMT in the child (P>0.6), but excluding these women slightly strengthened the relation between maternal energy intake and IMT.
Mothers whose energy intakes were lower had children who tended to be heavier and to have a higher BMI at age 9 years. Correlation coefficients between energy intake in early pregnancy and child weight or BMI were r=0.137 (P=0.040) and r=0.201 (P=0.002), respectively; comparable values for energy intake in late pregnancy were r=0.104 (P=0.123) and r=0.133 (P=0.047). There was, however, no association between maternal energy intake in early or late pregnancy and the systolic blood pressure of the child.
Table 5 shows mean differences in IMT by quarters of the distribution of maternal intake of energy and other macronutrients relative to the highest quarter. Estimates are adjusted for sex, current weight, systolic blood pressure, and exercise and further adjusted for maternal social class, prepregnancy BMI, smoking, and strenuous exercise during pregnancy. Lower maternal intakes of energy, fat, protein, and carbohydrate in early and, more particularly, late pregnancy continued to be significantly associated with increased IMT in the child in the fully adjusted models. Inclusion of maternal age and height and use of current BMI, instead of weight, had little effect on the results (data not shown).
|
In the fully adjusted models, current weight, systolic blood pressure, and exercise were significantly associated with the IMT of the child. Mean IMT of children in the lowest quarter of the distribution of systolic blood pressure or weight was 0.030 mm (95% confidence interval [CI], 0.006 to 0.053) or 0.036 mm (0.011 to 0.061) smaller respectively than that of children in the highest quarter of the distribution of these variables; mean IMT of children who played sport or exercised outside school was 0.019 mm (95% CI, 0.001 to 0.037) smaller than those who did not.
We examined whether the relations between maternal energy intake in early and late pregnancy and the IMT of the children differed according to weight or BMI, but interaction tests were not significant (all P>0.5). The Figure shows mean IMT according to thirds of the distribution of current weight and maternal energy intake in late pregnancy.
|
To check the consistency of our findings, we repeated the analysis using energy intake data from the food diary completed in early pregnancy.21 After multivariate adjustment, mean IMT of children whose mothers had energy intake that was in the lowest quarter of the distribution was 0.028 mm (95% CI, 0.001 to 0.053) greater than those whose mothers were in highest quarter, similar to the difference obtained with FFQ-derived data.
| Discussion |
|---|
|
|
|---|
One potential explanation is that maternal energy intake in pregnancy affects the lipid levels of the child. Restricted energy intake in pregnant rats results in hypercholesterolemia in the offspring.26 Adults exposed to famine during early gestation in the Dutch Hunger Winter had a more atherogenic lipid profile than those whose mothers had not faced calorie restrictions,27 although no such effects were seen in adults whose mothers were exposed to famine during the Leningrad siege.28 As we lacked data on lipid concentrations, we were unable to explore this. Although animal studies suggest that restricted maternal energy intake might lead to increased IMT by promoting higher blood pressure,7 we found no evidence for this.
Hypercholesterolemia in pregnancy has been associated with atherosclerotic lesions in fetuses and children,3,11 but we had no data on maternal serum cholesterol. Animal studies suggest that a maternal diet rich in saturated fat leads to endothelial dysfunction in the offspring.9,10 We found that a lower maternal intake of macronutrients overall, including total and saturated fat, was associated with increased IMT in the child, but IMT was not influenced by the proportion of energy derived from total or saturated fat.
One limitation was our inability to follow up all children in the cohort. Some had moved away from the area, and some declined to participate. Mean maternal intake of macronutrients, however, was similar in the groups who did and did not participate in the follow-up. We think it unlikely that bias will have been introduced.
A further limitation was that information on childhood diet was restricted to breastfeeding. Evidence on the relation between breastfeeding and atherosclerosis is inconsistent.2931 We found no indication that IMT differed by duration of breastfeeding, but we had no data on diet after weaning. In an intervention study, boys who received a diet low in saturated fat from age 7 months had lower serum cholesterol concentrations at ages 3 and 7 years32 and enhanced endothelial function at age 11 years.33 This suggests that early childhood diet may alter susceptibility to atherosclerosis. Maternal demographic and social factors influence diets of children,34 but the relation between maternal diet during pregnancy and how a mother feeds the child is not known. Whether the association in our data between lower maternal energy intake and increased IMT in the child is attributable solely to the effect of diet in pregnancy or whether it reflects some influence of diet after weaning is unclear.
The FFQ is the dietary assessment method most often used in epidemiological studies. It is capable of ranking and grouping individuals according to nutrient intake.35 The FFQ used here was validated by comparing intake estimates with those obtained from a 4-day food diary.21 The validation results suggest that estimates of intake from the FFQ are acceptable for ranking individuals within the distribution. When we repeated our analysis using energy intake data from the food diary, results were similar to those obtained with the FFQ-derived data.
IMT is a strong predictor of vascular events in adults: in 5056 nineteen- to ninety-year olds, an increase of 1 SD (0.16 mm) in common carotid IMT was associated with increases in stroke and myocardial infarction risk of 47% and 43%, respectively.13 The range of IMT in our study, as in other studies of children,15 was narrower than in older subjects, although values at the top end of the distribution were similar to those observed in some adults.13 The predictive value of childhood IMT is unknown, but our finding that IMT was increased in children who were heavier, had higher systolic blood pressure, or were less physically active adds to the evidence that early exposure to cardiovascular risk factors may promote atherogenesis.14,15,36 The observation that lower maternal energy intake in pregnancy predicts increased IMT suggests that the prenatal environment is also important in atherogenesis. Further investigations are needed to replicate this finding and to clarify the underlying mechanisms.
| Acknowledgments |
|---|
Sources of Funding
The study was funded by the Medical Research Council and WellChild (previously Children Nationwide).
Disclosure(s)
None.
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
P. D. Gluckman, M. A. Hanson, C. Cooper, and K. L. Thornburg Effect of In Utero and Early-Life Conditions on Adult Health and Disease N. Engl. J. Med., July 3, 2008; 359(1): 61 - 73. [Full Text] [PDF] |
||||
![]() |
C. R. Gale, M. K. Javaid, S. M. Robinson, C. M. Law, K. M. Godfrey, and C. Cooper Maternal Size in Pregnancy and Body Composition in Children J. Clin. Endocrinol. Metab., October 1, 2007; 92(10): 3904 - 3911. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
ATVB Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 2006 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |