Background
Excessive or inadequate gestational weight gain (GWG) has been associated with various adverse pregnancy outcomes, including gestational diabetes mellitus (GDM) [
1], preeclampsia [
2], cesarean delivery [
1,
3,
4], stillbirth [
5], preterm birth [
3,
4], macrosomia [
1,
3,
4], large for gestational age (LGA) [
3] and small for gestational age (SGA) [
3]. The Institute of Medicine (IOM) recommended the optimal GWG range over the course of pregnancy and GWG rate for the last two trimesters in 2009 [
6]. However, since the IOM guideline was mainly based on the Caucasian population and standards, its generalizability to Asian women was limited [
7‐
9]. Considering there were approximately 187 million Chinese women of childbearing age (20–39 years) in 2020 [
10], establishing an optimal weight gain range for Chinese women is needed.
Several analyses have tried to explore the optimal GWG in Chinese population, yet the limited sample size [
11‐
14] or retrospective design [
15,
16] confined the credibility and generalizability. In these investigations, the optimal GWG range is generally proposed either by calculating frequency [
12] or by estimating risk [
11,
14,
16] of multiple adverse pregnancy outcomes using binary logistic regression, which may lead to misspecification due to purely linear relationships [
17]. The generalized additive models (GAMs), as a nonlinear model, will capture the association between risk factors and health outcomes that is not revealed by the binary logistic models, and can avoid model misspecifications in these models [
17]. Recently, GAMs were applied to analyze the association between GWG and poor pregnancy outcomes in US [
18], and showed high interpretability and predictive accuracy, however, its practice among Chinese pregnant women is lacking. Furthermore, except for optimal total GWG range, the trimester-specific GWG rate is also an important parameter to monitor weight at different stages of pregnancy, but it has not been considered in existing evidence conducted in Chinese women [
11‐
16] or other Asian women [
7‐
9]. It should be noted that in 2021, the Chinese Nutrition Society (CNS) released its GWG guideline [
19] and the first report about its applicability among Chinese women has published recently [
20]. In this study, we attempted to propose the optimal total and trimester-specific GWG with a cohort of relatively large sample size based on modelling technique GAMs, which may provide a methodological reference for future studies to determine optimal GWG range in their countries, as well as providing evidence for the supplementation of the 2021 CNS guideline in the future.
Moreover, increasing rates of multi-parity and advanced maternal age (AMA) have been detected [
21] after the full implementation of a two-child policy of China since 2015 [
22]. With the announcement of a three-child policy in 2021 [
23], it is likely that more and more Chinese women are giving birth at older ages. However, the impact of GWG on pregnancy outcomes in AMA as well as the optimal GWG for AMA has not been much explored.
Using prospective data from a large cohort, we thus aimed 1) to investigate the association of GWG on multiple maternal and neonatal outcomes including GDM, preeclampsia, cesarean delivery, stillbirth, preterm birth, macrosomia, LGA, and SGA using the GAMs, and 2) to create an optimal total GWG range and trimester-specific GWG range based on Asian-specific body mass index (BMI) categories for Chinese women.
Discussion
In this cohort, we figured out a U-shaped prospective association between GWG and the risk of aggregated adverse pregnancy outcomes among 51,125 participants. Based on these associations, an optimal GWG range was suggested as 8.2–13.0 kg, 7.3–12.5 kg, and 2.0–9.4 kg respectively for underweight, normal weight and overweight/obese Chinese women, and a higher rate of GWG in the first two trimesters than the last trimester was suggested.
During gestation, pregnant women and their children are, to varying degree, simultaneously at risk of multiple adverse health outcomes. Our finding of the U-shaped association between GWG and total risk of combined adverse maternal and neonatal outcomes was in line with previous analyses [
7‐
9,
11], which implies the importance of proper GWG. Thus, proposing an optimal GWG range specific for Chinese women to keep the balance between maternal health and neonatal growth is crucial.
Based on prospective data from 51,125 mother-child pairs, we suggested the optimal GWG ranges for Chinese pregnant women, which were generally lower than that recommended by the IOM, for all pre-gravid BMIs. Different from the WHO BMI categories applied in the IOM, the GWG range established in our analysis was based on the Asia-specific BMI categories, which are more proper recommendations for Chinese women. From the aspect of racial or ethnic differences, Asian populations have lower BMI levels but higher body fat levels than Caucasians [
33] leading to their different susceptibility to weight gain in pregnancy. Moreover, evidence has shown that Asians have increased risk of obesity-related diseases at lower BMI levels than Caucasians [
34,
35], therefore, lower BMI cut-off utilized and lower GWG suggested in the present study should be safer for Asian populations. Besides, various recommendations of GWG for different populations were proposed in Japan [
7], Korea [
8], Singapore [
9], German [
36], Belgian [
37] and US [
18], which implies that the optimal GWG may be population-specific, and each country should consider its own optimal GWG range. Furthermore, compared to several studies in China with small sample size [
11‐
14] or retrospective design [
15,
16], the present study was a large, prospective cohort, and the characteristics of the participants (e.g., mean maternal age, the prevalence of underweight or overweight/obese in women, and birth weight) were similar to the data from National Statistical Yearbook [
38,
39] and national-wide surveys [
32,
40]. Notably, the national GWG guideline for Chinese women has been published in 2021 [
19]. The suggested GWG range in our study was slightly lower than the one recommended by the CNS, possibly due to the fact that the methodology was varied.
The 2009 IOM guideline [
6] has recommended the weekly GWG rate over the course of the 2nd and 3rd trimesters, and so far, only one US study [
18] has tried to analyze it. Actually, only 10–13% of Chinese women gained their weight within the IOM guideline in the last two trimesters [
41], and the pattern of GWG would be influenced by maternal ethnicity [
6], thus the GWG rate for Caucasians was not appropriate for Chinese women. In contrast to IOM [
6] and the US study [
18], we suggested a higher GWG rate in the first two trimesters than that in the last trimester, except for pre-gravid overweight/obese women. Our findings may have important public health implications, for providing specific guidance for pregnant women to track their weight over trimesters and achieve the recommend total GWG. Maintaining appropriate weight gain at different trimester is a key to prevent pregnancy complications and improve well-being of mothers and their children, given the impacts of GWG on pregnancy outcomes during various trimesters were observed in the present study.
There is an increasing trend of delayed childbearing in China, given the fertility rate of women aged 30–34 years increased from 5.7‰ in 1995 to 18.6‰ in 2015, whereas the rate declined for younger women [
38,
39]. Furthermore, there was a 7.2% monthly increase in multiparous births of mothers older than 35 years [
21] since the implementation of the two-child policy in 2015. Considering the alarming rise of older mothers, and they were at higher risks of pregnancy complications as shown in our study, a first attempt on the maternal age-specific optimal GWG range was made. We suggested a lower overall GWG for mothers older than 35 years compared to younger mothers. This might be partially because the proportion of women being overweight/obese before pregnancy increased with maternal age, given the pre-gravid overweight/obesity is a risk factor for multiple adverse pregnancy outcomes [
24] and a lower amount of GWG was generally recommended for overweight/obese women [
7,
11,
12,
18,
36,
37]. In order to better understand the association between maternal overweight/obesity and AMA, further research on optimal GWG for AMA at different pre-gravid BMIs is warranted.
Our study has several strengths. Our participants and their children were representative of the general population in age, distribution of pre-gravid BMI and birth weight according to National Statistical Yearbooks [
38,
39] and national surveys [
32]. The prospective nature and large sample size, in conjunction with the ability to consider for a broad range of pre-gravid BMI and maternal age categories of participants represent substantial strengths. Another advantage lies in the use of the non-linear statistical model to provide a more credible reference of optimal GWG. Notably, the maternal age specific GWG reference was considered in our analysis, which might provide new idea on prenatal care of women of AMA, on account of the rising prevalence of this population. The results may also stimulate future investigations to supplement the 2021 CNS guideline by considering different characteristics of population. Limitations are still worth noted. The current study only presented the preliminary results on the optimal GWG for AMA due to the limited sample size of subpopulations, further details with respect to different pre-gravid BMI or parity are needed. GWG and GWG rate in the first two trimesters were combined in this study as women usually gain limited amount of weight during the first trimester. However, the combined GWG rate may mask the effect of individual trimester GWG rate, and future research on optimal weight gain at each trimester is warranted. While the study population was representative of the general population in age, pre-gravid BMI and birth weight, they were recruited from the Southwest China, and the generalizability of the findings to the Chinese population need further investigation.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.