Introduction
Maternal hyperglycemia during pregnancy is a well-known risk factor for adverse birth outcomes, such as macro-somia and neonatal hypoglycemia [
1]. Accumulating evidence suggests that early pregnancy is a critical period for the adverse effects of high maternal glucose concentrations on embryonic and placental development [
2,
3]. High maternal glucose concentrations from early pregnancy onwards may cause alterations in embryonic and placental development, and lead to an increased transfer of glucose to the developing fetus, predisposing to increased fetal growth and fat deposition and alterations in fetal metabolism. These fetal adaptations may, subsequently, predispose to increased risks of adverse birth outcomes [
1].
During pregnancy, most transfer of glucose across the placenta occurs in the postprandial state. These postprandial glucose concentrations are mainly determined by maternal dietary carbohydrate intake [
4]. The dietary glycemic index and glycemic load are measures that can be used to qualify and quantify the maternal postprandial glycemic response to the maternal dietary carbohydrate intake. These measures influence postprandial glucose available for maternal energy, storage, and transfer to the fetus [
5,
6]. Intervention studies suggested that a low-glycemic index diet during the second half of pregnancy may reduce birth weight and infant adiposity in women with gestational diabetes or an impaired glucose metabolism [
7,
8]. No increased risks of delivering a small-for-gestational-age infant were observed in these intervention studies. However, an observational study among pregnant women not at risk of an impaired glucose metabolism reported that a lower.
Maternal dietary glycemic index in the second half of pregnancy was associated with an increased risk of delivering a small-for-gestational-age infant [
9]. In pregnant women without an impaired glucose metabolism, not much is known about the effects of maternal dietary glycemic index and load during early pregnancy on directly measured fetal growth throughout pregnancy and the risks of adverse birth outcomes. We hypothesized that a lower maternal dietary glycemic index and load in early pregnancy might reduce the risks of fetal overgrowth and macro-somia, but might also lead to increased risks of fetal undergrowth and low birth weight, especially among a general, healthy population.
Therefore, in a population-based prospective cohort study among 3471 pregnant women without an impaired glucose metabolism, we examined the associations of maternal early pregnancy dietary glycemic index and load within a low-to-normal range with fetal growth throughout pregnancy and the risks of adverse birth outcomes.
Discussion
Among pregnant women without an impaired glucose metabolism, we observed that maternal early pregnancy dietary glycemic index across was not associated with fetal growth parameters, whereas a higher maternal early pregnancy dietary glycemic load was associated with a higher fetal abdominal circumference and estimated fetal weight in late-pregnancy only. A higher glycemic index, but not load, was associated with a lower risk of a large-for-gestational-age infant.
There is increasing interest in targeting maternal dietary glycemic index and load during pregnancy as a lifestyle intervention to improve pregnancy and birth outcomes. Small interventions studies among pregnant women with gestational diabetes, impaired glucose tolerance or obesity, have already shown that a lower glycemic index diet from the second half of pregnancy onwards improves maternal glucose concentrations and lowers the risk of delivering a large-for-gestational-age infant [
19,
34]. With dietary interventions, these studies achieved a median maternal dietary glycemic index around 50 or lower in their intervention groups and compared these effects to a normal or high maternal dietary glycemic index. Far less is known about the effects of maternal dietary glycemic index and load on birth outcomes among populations not at risk for an impaired glucose metabolism.
A few previous studies focused on the associations of maternal dietary glycemic index and load with birth characteristics and the risks of adverse birth outcomes among general, healthy populations, but no studies focused on directly measured fetal growth characteristics [
9,
17,
18,
30]. These studies differed strongly with regards to the methods used to calculate maternal dietary glycemic index and load, the timing of the dietary assessments, studied populations, and adjustment for maternal socio-demographic and lifestyle characteristics. An observational study among 47,003 Danish pregnant women reported that a higher maternal dietary glycemic load, but not index, in mid-pregnancy was associated with a higher birth weight and an increased risk of delivering a large-for-gestational-age infant [
30]. A study among 1,082 multi-ethnic non-diabetic pregnant women from USA showed that the lowest quintile of maternal mid-pregnancy dietary glycemic index, but not load, was associated with a lower birth weight and an increased risk of delivering a small-for-gestational-age infant. Using white bread instead of glucose as a reference, the glycemic index in this study varied < 71 for the lowest quintile to > 85 for the highest quintile. No associations of the highest quintile of maternal mid-pregnancy dietary glycemic index or load with a higher birth weight and increased risk of delivering a large-for gestational-age-infant were observed [
9]. Contrarily, a study among 842 low-risk Irish pregnant women reported no associations of maternal dietary glycemic index and load in early pregnancy continuously with birth weight or adverse birth outcomes, after adjusting for maternal age, pre-pregnancy BMI, and parity and considering multiple testing [
17]. Similarly, a study among 906 low-risk pregnant women from the UK showed no associations of maternal early or late-pregnancy dietary glycemic index and load continuously with fat and lean mass at birth [
18]. The mean and variability of the glycemic index in these two studies were comparable to ours.
In line with these previous studies focused on maternal early pregnancy dietary glycemic index and load, we observed that women within our study consumed diet with a relatively low mean dietary glycemic index. No consistent associations of maternal early pregnancy dietary glycemic index and load across the low-to-normal range with birth weight and the risks of adverse birth outcomes were observed. We did observe that a higher maternal early pregnancy dietary glycemic load, especially within the highest quartile, was associated with a higher late-pregnancy fetal abdominal circumference and estimated fetal weight, but findings were not consistent across pregnancy and may reflect a chance finding. However, fetal fat development mainly occurs in late-pregnancy and abdominal circumference is an important indicator of fetal fat deposition [
35]. This could suggest that a higher maternal early pregnancy dietary glycemic load may rather affect fetal body composition than growth, which is also suggested by the previous studies conducted in infants [
36,
37].
Contrary to our prior hypothesis, we observed that a higher maternal early pregnancy dietary glycemic index within a low-to-normal range was associated with lower fetal length growth rates from late-pregnancy onwards and with a lower risk of delivering a large-for-gestational-age infant only. These association were not explained by maternal socio-demographic and lifestyle characteristics. It could reflect a chance finding. Our study population is a relatively healthy population not at high risk of an impaired glucose tolerance. We only included Dutch women without pre-gestational diabetes and we observed largely similar results for women with a normal weight, younger than 35 years old, and without gestational diabetes. Possibly, the range of maternal dietary glycemic index within our population reflects a relatively healthy range for women at a low risk of an impaired glucose metabolism in early pregnancy. Maternal dietary glycemic index within this range may be not related to increased risks of fetal undergrowth or overgrowth. The timing of dietary glycemic index assessment in early pregnancy may also be important. Maternal insulin sensitivity is much higher in early pregnancy as compared to mid- and late-pregnancy, which leads to smaller fluctuations in postprandial glycemic responses to carbohydrate containing foods in early pregnancy [
5]. Potential adverse effects of a higher maternal dietary glycemic index on fetal growth and the risk of macro-somia may be more pronounced in the second half of pregnancy, when pregnant women are physiologically more insulin resistant and the postprandial glycemic response shows larger fluctuations. Finally, postprandial peaks in maternal glucose concentrations and subsequent peak increases in fetal glucose concentrations may rather have an effect on fetal body composition and fetal metabolism than on skeletal growth, by affecting fetal development of adipocytes and the cardio-metabolic system [
18,
36,
37]. This hypothesis is supported by the associations which we observed of a higher maternal early pregnancy glycemic load with fetal abdominal circumference and estimated fetal weight in late-pregnancy when fetal fat accumulation occurs. Further studies using multiple assessments of dietary intake throughout pregnancy are needed to examine the detailed associations of maternal dietary glycemic index and load with both fetal and neonatal growth and body composition.
Importantly, in a secondary analysis, we observed no increased risks of preterm birth, small-for-gestational-age at birth, or caesarian delivery, as complication of abnormal fetal growth, among women consuming a low-glycemic index diet, as compared to women consuming a normal-glycemic index diet. The mean dietary glycemic index of women consuming a low-glycemic index diet within our study was largely similar to the mean dietary glycemic index reported in intervention studies stimulating a low-glycemic index diet through advising low-glycemic index food products [
19,
21]. This suggests that even among pregnant populations without an impaired glucose metabolism, a diet with lower glycemic index products in early pregnancy does not appear to be associated with fetal growth restriction and related adverse birth outcomes. These findings are important from a public health perspective, as there is an increasing interest in stimulating a diet with low-glycemic index products during pregnancy to improve birth and childhood outcomes. Our findings suggest that adhering to a diet with low-glycemic index products may be a safe intervention during pregnancy without adverse effects on fetal growth and birth outcomes in women without an impaired glucose metabolism. The beneficial effects of a lower dietary glycemic index and load within general, healthy populations on fetal growth and birth outcomes remain to be determined.
Strengths and limitations
Strengths of this study were the prospective study design, large sample size, and repeatedly measured fetal growth data from mid-pregnancy onwards available. Limitations of this study should also be taken into account when interpreting results. First, the response rate at baseline for participating in the Generation R study cohort was 61%. The non-response would have led to biased effect estimates if the associations were different between those included and not included in the analyses. However, this seems unlikely because biased estimates in large cohort studies often arise from loss to follow-up rather than from non-response at baseline [
38]. Second, we did not have information on previous gestational diabetes or polycystic ovarian syndrome, which are also associated with an increased risk of an impaired glucose metabolism. Although we expect the number of cases of previous gestational diabetes and polycystic ovarian syndrome to be low, as we had a relatively healthy population, this may have affected our results. Further studies excluding these women should replicate our findings. The selection towards a relatively healthy Dutch population may affect the generalizability of our findings and might have led to reduced statistical power. Most women had a dietary glycemic index and load within the normal range and the number of adverse birth outcomes was also relatively low. Further studies are needed among multi-ethnic populations with a more diverse dietary intake to replicate our findings. Third, even though the FFQ is widely used for dietary assessment in observational studies, measurement of food intake by an FFQ may be affected by measurement error, recall bias, and reporting bias. Subsequent calculation of the dietary glycemic index and load from the FFQ may further be affected by uncertainty induced by preparation of foods, mixed dishes, variations of food products of time, or unavailability of specific food products [
20]. Fourth, we obtained information on maternal dietary intake only once during pregnancy. Further studies from preconception onwards are needed using repeated assessments of maternal dietary intake prior and throughout pregnancy to obtain further insight into critical periods for the influence of maternal dietary glycemic quality and quantity on embryonic and fetal development and adverse birth outcomes. Finally, although were able to adjust for multiple confounding factors, there might still be residual confounding as in any observational study.
Acknowledgements
The Generation R Study is conducted by the Erasmus Medical Center in close collaboration with the School of Law and Faculty of Social Sciences of the Erasmus University Rotterdam, the Municipal Health Service Rotterdam area, Rotterdam, the Rotterdam Homecare Foundation, Rotterdam, and the Stichting Trombosedienst and Artsenlaboratorium Rijnmond (STAR), Rotterdam. We gratefully acknowledge the contribution of participating mothers, general practitioners, hospitals, midwives, and pharmacies in Rotterdam.