Article Text

Download PDFPDF

Original research
Association of prepregnancy body mass index and gestational weight gain trajectory with adverse pregnancy outcomes—a prospective cohort study in Shanghai
  1. Ziwen Ma,
  2. Liming Chu,
  3. Zhiping Zhang,
  4. Yifan Hu,
  5. Yun Zhu,
  6. Fei Wu,
  7. Yan Zhang
  1. Shanghai Pudong New Area Health Care Hospital for Women and Children, Shanghai, China
  1. Correspondence to Ms Yan Zhang; 2571036100{at}qq.com

Abstract

Objectives The objective was to investigate the associations of maternal prepregnancy body mass index (BMI) and gestational weight gain (GWG) trajectories with adverse pregnancy outcomes (APOs).

Design This was a prospective cohort study.

Setting This study was conducted in Shanghai Pudong New Area Health Care Hospital for Women and Children, Shanghai, China.

Primary and secondary outcome measures A cohort study involving a total of 2174 pregnant women was conducted. Each participant was followed to record weekly weight gain and pregnancy outcomes. The Institute of Medicine classification was used to categorise prepregnancy BMI, and four GWG trajectories were identified using a latent class growth model.

Results The adjusted ORs for the risks of large for gestational age (LGA), macrosomia, gestational diabetes mellitus (GDM) and hypertensive disorders of pregnancy (HDP) were significantly greater for women with prepregnancy overweight/obesity (OR=1.77, 2.13, 1.95 and 4.24; 95% CI 1.3 to 2.42, 1.32 to 3.46, 1.43 to 2.66 and 2.01 to 8.93, respectively) and lower for those who were underweight than for those with normal weight (excluding HDP) (OR=0.35, 0.27 and 0.59; 95% CI 0.22 to 0.53, 0.11 to 0.66 and 0.36 to 0.89, respectively). The risk of small for gestational age (SGA) and low birth weight (LBW) was significantly increased in the underweight group (OR=3.11, 2.20; 95% CI 1.63 to 5.92, 1.10 to 4.41; respectively) compared with the normal-weight group; however, the risk did not decrease in the overweight/obese group (p=0.942, 0.697, respectively). GWG was divided into four trajectories, accounting for 16.6%, 41.4%, 31.7% and 10.3% of the participants, respectively. After adjustment for confounding factors, the risk of LGA was 1.54 times greater for women in the slow GWG trajectory group than for those in the extremely slow GWG trajectory group (95% CI 1.07 to 2.21); the risk of SGA and LBW was 0.37 times and 0.46 times lower for women in the moderate GWG trajectory group and 0.14 times and 0.15 times lower for women in the rapid GWG trajectory group, respectively; the risk of macrosomia and LGA was 2.65 times and 2.70 times greater for women in the moderate GWG trajectory group and 3.53 times and 4.36 times greater for women in the rapid GWG trajectory group, respectively; and the women in the other three trajectory groups had a lower risk of GDM than did those in the extremely slow GWG trajectory group, but there was not much variation in the ORs. Notably, different GWG trajectories did not affect the risk of HDP.

Conclusions As independent risk factors, excessively high and low prepregnancy BMI and GWG can increase the risk of APOs.

  • PERINATOLOGY
  • OBSTETRICS
  • Diabetes in pregnancy
  • Obesity
  • NUTRITION & DIETETICS

Data availability statement

Data are available upon reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • Compared with total gestational weight gain (GWG), the GWG trajectory was more refined and accurate.

  • The participants were from the same hospital, which could have limited the generalisability of the results.

  • Few lifestyle factor and family history data were collected in this study, which may have limited adjustment for confounders.

  • This study failed to further analyse the interaction effect of prepregnancy body mass index and GWG trajectory on adverse pregnancy outcomes.

Introduction

Overweight and obesity among pregnant women are becoming common health problems worldwide. From 1998 to 2003, the number of women who are overweight or obese increased from 29.8% to 38% worldwide.1 In the USA, 40% of pregnant women are overweight or obese, and more than 50% of pregnant women gain more weight during pregnancy than the recommended body mass index (BMI) established in the 2009 guidelines issued by the Institute of Medicine (IOM) for prepregnancy and gestational weight gain (GWG).2 In Germany, the prevalence rates of overweight and obesity among women aged 20–39 years were 20% and 9%–14%, respectively.3 In some Persian Gulf regions and Pacific islands, the prevalence of obesity has even reached 50%.4 China is not the only region where a growing number of women are affected by overweight and obesity during pregnancy, which can be ascribed to the poor understanding of nutrition during pregnancy in general, as indicated by the improper consumption of nutritional supplements, the decline in physical activity and excessive weight gain during pregnancy; therefore, as early as 2010, the rates of overweight and obesity among women in China were 30.6% and 12%, respectively,5 and the rates are still on the rise.

Previous studies have shown that prepregnancy obesity and excessive GWG are closely related to adverse maternal-infant outcomes, such as hypertensive disorders of pregnancy (HDP), macrosomia and low birth weight (LBW).6–9 However, the exact cause‒effect interaction is unclear, and the association between GWG and adverse pregnancy outcomes (APOs) remains controversial,10–15 especially regarding total pregnancy weight gain. GWG has been reported to have no correlation with gestational diabetes mellitus (GDM).16 On the other hand, GWG contributes to only 9% of the absolute change in insulin sensitivity during pregnancy,17 and excessive GWG in the first trimester has been shown to be associated with an increased risk of GDM regardless of prepregnancy BMI.6 Ashtree et al suggested that women with GDM or hypertensive disorders were more likely to gain weight outside these guidelines.13 Sámano et al reported that excessive GWG did not alter the risk of HDP, preterm labour, LBW or small for gestational age (SGA).16 Li et al reported the opposite results.10

It is well recognised that the GWG trajectory is more refined and accurate than total GWG; however, there are few data relating to this hypothesis, mostly to total GWG. The purpose of this study was to investigate the associations of prepregnancy BMI and GWG trajectories with pregnancy outcomes.

Patients and methods

Study design and participants

This prospective cohort study was conducted at Shanghai Pudong New Area Health Care Hospital for Women and Children from June 2020 to October 2021, including a total of 2174 participants (figure 1). Pregnant women were recruited at the first prenatal examination and followed up until delivery.

Figure 1

Flow chart of participants included in the study.

The inclusion criteria were as follows: (1) women with singleton pregnancies; (2) women without prepregnancy acute/chronic diseases, such as Sjogren’s syndrome or neurological disorders; (3) pregnant women at less than 11 gestational weeks; and (4) women of Han nationality.

The exclusion criteria were as follows: (1) prepregnancy acute or chronic diseases, such as prepregnancy diabetes, hypertension and heart disease; (2) multiple pregnancies; (3) failure to attend regular prenatal examinations during pregnancy; and (4) late abortion.

Patient and public involvement

Patients and the public were not involved in the design, conduct or dissemination of our research.

Measurements

We derived the pregnant women’s relevant information from their medical records and completed questionnaires pertaining to maternal age, height, BMI, delivery mode, gestational weeks, birth weight, annual income and so on. The women’s weight was measured at each time point using a standard weight scale (Huade, Shanghai, China). Every week, the body weight of each participant was measured twice, and the average value was used for subsequent analyses. During the measurement, each participant was supposed to wear light clothing and be barefoot.

According to the IOM-issued 2009 guidelines in conjunction with prepregnancy BMI,18 the participants were first divided into four groups: underweight, normal, overweight and obese. Because there were only a few participants in the obese group, we combined the overweight and obese groups for the final analysis. Four GWG trajectories were identified using a latent class growth model.

Because the exclusion criteria included prepregnancy acute or chronic diseases, HDP exclusively included gestational hypertension, preeclampsia and eclampsia. According to the HDP diagnostic criteria, we used the standards recommended by the Gestational Hypertension Group of the Chinese Society of Obstetrics and Gynaecology in 202019: a systolic blood pressure ≥140 mm Hg or a diastolic blood pressure ≥90 mm Hg after 20 weeks gestation. According to the GDM diagnostic criteria, we used the standards recommended by the International Association of Diabetes and Pregnancy Study Group in 2010.20 The presence of at least one of the following abnormal values from an oral glucose tolerance test (OGTT) was sufficient to diagnose GDM: a plasma glucose level ≥5.1 mmol/L after an overnight fast; a plasma glucose level ≥10.0 mmol/L 1 hour after consuming 75 g of glucose; or a plasma glucose level ≥8.5 mmol/L 2 hours after consuming 75 g of glucose. The definition of macrosomia was a birth weight ≥4000 g. LBW was defined as <2500 g, SGA was defined as a birth weight less than the 10th percentile and large for gestational age (LGA) was defined as a birth weight greater than the 90th percentile for gestational age.

Statistical analysis

The data were double-entered by two qualified researchers into the Epi Data Database (V.3.1, the Epi Data Association, Odense, Denmark). Statistical analyses were performed using SPSS V.22.0, SAS V.9.4, Stata/SE V.15.1 and R V.4.1.2. Categorical variables are expressed as frequencies (rate; n (%)). Comparisons were made between the different groups using the likelihood ratio and Pearson tests. Group-based trajectory modelling was applied to the GWG trajectory at a time interval of 10–40 weeks.21 22 With 2–6 subgroups treated as an alternative, the optimal number of groups was determined based on the Bayesian information criterion, and the trajectories of each group were fitted with a polynomial of the highest order of 3, with a minimum group ratio no less than 5%. For the model performance evaluation metrics, we calculated the average posterior probability, the odds of correct classification, the expected probability of group membership and the consistency with the actual distribution ratio of each group. With the factors related to GDM and HDP that were selected, a multiple logistic regression analysis was performed to examine the associations of prepregnancy BMI and GWG trajectories with pregnancy outcomes.

Results

General characteristics of the study subjects

A total of 2174 pregnant women participated in this study (figure 1). According to their prepregnancy BMIs, they were classified into three groups: underweight, normal weight and overweight/obese. Significant differences were observed in age, gravidity, parity, annual income and GWG (all p<0.001). As indicated in table 1, there was a significant difference in GDM (p<0.001), HDP (p=0.001), macrosomia (p<0.001), LGA (p<0.001) and SGA (p=0.003) among the groups in terms of the effect of prepregnancy BMI on maternal and infant complications.

Table 1

Comparison of sample characteristics and outcomes among the three prepregnancy body mass index (BMI) groups

As indicated in figure 2, four GWG-based trajectories were defined: trajectory 1, extremely slow GWG; trajectory 2, slow GWG; trajectory 3, moderate GWG; and trajectory 4, rapid GWG. These trajectories accounted for 16.6%, 41.4%, 31.7% and 10.3% of the participants, respectively. Clearly, the slow GWG trajectory group had the highest proportion of pregnant women who were underweight and normal weight, followed by the moderate GWG trajectory group. Among pregnant women who were overweight/obese, the percentage of women in the extremely slow GWG trajectory group was significantly greater than that of women in the moderate and rapid GWG trajectory groups, although the percentage of women in the slow GWG trajectory group was the greatest. Significant differences were observed in age, parity, education level and prepregnancy BMI. In terms of neonatal complications, the incidence of LGA and macrosomia significantly differed among the different GWG trajectory groups; LGA had a significant effect on the incidence of GDM but did not affect the incidence of HDP. As indicated by the moderate GWG trajectory group, the incidence of GDM was the lowest (table 2).

Figure 2

Developmental trajectories of gestational weight gain.

Table 2

The sample characteristics and outcomes compared in the four gestational weight gain trajectory groups

Correlations between different prepregnancy BMIs and pregnancy outcomes

For neonatal complications, after adjusting for age, gravidity, parity, educational level, annual income, GDM status and HDP status, the ORs for the risks of LGA and macrosomia were significantly greater for women who were overweight or obese before pregnancy and lower for women who were underweight than for those who were normal weight. Moreover, the risk of SGA and LBW increased significantly in underweight women compared with those in women with a normal BMI; however, the risk did not decrease in women who are overweight or obese. The evidence persisted even after further adjustment for GWG trajectory (table 3).

Table 3

Association of prepregnancy body mass index (BMI) with pregnancy outcomes

After controlling for the confounding factors related to maternal complications, the results of the multilevel logistic regression indicated that the risk of GDM and HDP was 2.16 and 4.02 times greater (95% CI 1.59 to 2.93 and 1.93 to 8.39, respectively) in women who are overweight/obese than in those with a normal BMI. Additionally, the risk of GDM in underweight women was half that in normal-weight women. After further adjustment for the GWG trajectories, being overweight/obese was still a risk factor for developing GDM and HDP (table 3).

Different GWG trajectories and their respective correlations with pregnancy outcomes

After adjusting for confounding factors, the slow GWG trajectory did not result in an increase in the development of adverse infant outcomes compared with the extremely slow GWG trajectory. After further adjustment for prepregnancy BMI, however, the risk of LGA was 1.54 times greater in the slow GWG trajectory group than in the extremely slow GWG trajectory group (95% CI 1.07 to 2.21). In addition, the risk of SGA and LBW decreased significantly, and the risk of macrosomia and LGA increased significantly in the moderate and rapid GWG trajectory groups compared with the extremely slow GWG trajectory group, which held true even after further controlling for prepregnancy BMI (table 4).

Table 4

Association of gestational weight gain (GWG) trajectory with pregnancy outcomes

Interestingly, a lower risk of GDM was observed in the slow, moderate and rapid GWG trajectory groups than in the extremely slow GWG trajectory group, with the lowest value in the moderate GWG trajectory group (OR: 0.40; 95% CI 0.28 to 0.56) but there was not much variation in the ORs. Additionally, different GWG trajectories were not associated with a risk of HDP. This relationship remained, even after further adjustment for prepregnancy BMI (table 4).

Discussion

In China, there have been significant changes in lifestyle, one of which can be referred to as the development of a sedentary lifestyle. The tendency towards a sedentary lifestyle could be accelerated by occasional lockdowns amid the COVID-19 pandemic,23–26 during which pregnant women may become more likely to be overweight and obese and are thus more likely to develop APOs and even long-term metabolic diseases.

Through the analysis of the participants’ general characteristics in the present investigation, we found that women in the prepregnancy obese group and women in the extremely slow GWG trajectory group accounted for the greatest proportion of older mothers and multiparous women and that women in the prepregnancy underweight group and women in the extremely slow GWG trajectory group had a greater proportion of highly educated women, which may be explained by the diverse cultures of different regions, such as eating habits, pregnancy-related nutrition beliefs and information channels. To date, there has been no consensus on this phenomenon because the factors contributing to GWG are complex and multifactorial.27–30 Additionally, the slow GWG trajectory group had the highest proportion of pregnant women who were underweight and normal weight, followed by the moderate GWG trajectory group. Pregnant women who were overweight/obese accounted for a larger proportion of the extremely slow and slow GWG trajectory groups, which was consistent with the IOM guidelines.

Effect of prepregnancy BMI and GWG trajectory on maternal adverse outcomes

Gestational diabetes mellitus

GDM is a common complication of pregnancy. A meta-analysis of 11 studies showed that during the middle of pregnancy, the risk of GDM increased 1.64-fold with every 1–3 kg/m2 increase in prepregnancy BMI (95% CI 1.28 to 2.11) and 2.42-fold (95% CI 1.62 to 3.62) with every ≥3 kg/m2 increase in prepregnancy BMI.31 In our study, the risk of GDM increased with increasing prepregnancy BMI. With the assessment of age, education level, annual income, gravidity and parity, the risk of GDM increased 2.16 times (95% CI 1.95 to 2.93) among the pregnant women who are overweight/obese, and the evidence was based on a normal prepregnancy BMI as the reference. After further adjustment for GWG trajectories, overweight/obesity remained a risk factor for GDM (OR 1.95, 95% CI 1.43 to 2.66). Based on a meta-analysis of 364 668 pregnant women, a previously reported 31-cohort study showed that the risk of GDM was 3.76-fold greater in women with obesity than in normal-weight women (95% CI 3.31 to 4.28).32 Other studies have also shown a positive correlation between prepregnancy BMI and the risk of GDM.33–35

Intriguingly, after adjusting for confounders and taking the extremely slow GWG trajectory group as the reference, with the change in the GWG trajectory, the incidence of GDM began to increase, with the lowest incidence occurring in the moderate GWG trajectory group. Moreover, after further adjustment for prepregnancy BMI and OGTT results, the risk of GDM still showed the same increasing trend. These results were similar to those previously reported.36 We must emphasise that although the p value was less than 0.05, the change in the OR was not significant. There is a clear relationship between GWG and gestational diabetes, especially in regard to different trimesters. However, the association between excessive GWG and GDM has been confirmed by numerous studies and may be related to rapid fat accumulation leading to insulin resistance.12 37 Insufficient GWG was suggested to affect the expression of the 11β-hydroxysteroid dehydrogenase type II gene, causing damage to the placental barrier, weakening the protective effect of glucocorticoids and increasing the risk of GDM.38 GDM has also been suggested to be associated with insufficient GWG.16 However, this evidence could be related to factors such as dietary habits, postdiagnosis weight control and family history. In contrast, a retrospective cohort study among Malaysians conducted by Yong et al revealed that excessive GWG in the first or second trimester was not an independent risk factor for GDM.39 The associations of prepregnancy BMI and GWG trajectories with GDM need to be further investigated.

Hypertensive disorders of pregnancy

Our investigation indicated that the incidence of HDP increased with increasing prepregnancy BMI. Multiple logistic regression analysis revealed that when the general related factors were excluded, women who are overweight/obese had a 4.02-fold greater risk of HDP than did normal-weight women (95% CI 1.93 to 8.39), and after further adjusting for the GWG trajectories, the OR increased. Some previous studies have shown that being overweight or obese before pregnancy is closely related to HDP.15 40 41 According to a retrospective cohort study of 436 414 pregnant women, being overweight or obese before pregnancy was an independent risk factor for HDP after controlling for age, family income, social status, race and other factors.42

However, we did not find any significant difference in the incidence of HDP among the different GWG trajectory groups. After adjusting for confounding factors, GWG trajectory did not affect the risk of HDP, regardless of prepregnancy BMI. Although numerous studies have shown that excessive GWG is a risk factor for HDP,31 41 43 44 the correlation between GWG and HDP is still controversial, and the pathogenesis remains unclear. A cohort study in Sweden showed that early pregnancy weight gain in various BMI groups was not associated with the risk of HDP,44 while no correlation was observed between weight gain in early pregnancy and the risk of HDP in normal-weight or underweight women,45 as in the same case of the findings reported in Mexico.16 In view of this, we argue that maternal oedema, family history of hypertension, sample size and vomiting-related weight fluctuations may interfere with research results. Additionally, we doubt whether the IOM guidelines are applicable to Chinese populations.

Effect of prepregnancy BMI and GWG trajectory on neonatal outcomes

Prepregnancy BMI and GWG are well known to be two important factors affecting fetal outcomes, as shown in our investigation, where the incidence of macrosomia and LGA increased with increasing prepregnancy BMI and GWG. When the other factors were controlled for, the risks of macrosomia and LGA were 1.87 and 1.52 times greater, respectively, in the prepregnancy overweight/obese group than in the normal-weight group; after further adjustment for GWG trajectories, the ORs were greater in the former group than in the latter group. Furthermore, the moderate and rapid GWG trajectories were risk factors for macrosomia and LGA compared with the extremely slow GWG trajectory, and the risk became more significant after further controlling for prepregnancy BMI; similar findings have been reported previously.10 Additionally, prepregnancy BMI and GWG were suggested to be independent risk factors for macrosomia and LGA, respectively.46

The effects of prepregnancy BMI and GWG trajectories on LBW and SGA were intriguing. This investigation showed that the incidence of SGA was significantly different among the prepregnancy BMI groups, underweight was a risk factor for SGA, and overweight/obesity was not a protective factor. As previously reported, overweight and obesity did not affect the risk of SGA or LBW,10 and prepregnancy overweight and obesity increased the risk of neonatal complications, but only in white pregnant women, which suggested that racial factors could be the reason for the correlation between prepregnancy BMI and neonatal complications.47 BMI classification could be another reason.7 In addition, the effects on the risk of SGA and LBW were significant enough to be independent. These findings were consistent with ample evidence that GWG is associated with the risk of adverse neonatal outcomes.29 35 48 49

Intriguingly, neither prepregnancy BMI nor GWG trajectory affected the risk of premature birth. The biological mechanism underlying the associations of prepregnancy BMI and GWG with PTB is incompletely understood. However, contradictory results have been reported on the effect of prepregnancy BMI and GWG on premature birth risk.30 48 50 51 In the present investigation, we were not in a position to determine whether premature birth was spontaneous or medically indicated, which could explain the null results. Maternal-infant inflammation may serve as another contributing factor, which we did not rule out.

To our knowledge, we are the first to investigate the impact of prepregnancy BMI and GWG trajectories on pregnancy outcomes in Shanghai, China, involving a large sample of participants. However, our investigation had several limitations. The participants were predominantly women who were citizens of Shanghai and were from the same hospital, which could have limited the generalisability of the results; thus, further parallel studies are necessary in other populations. Next, women with different grades of obesity may have diverse ranges of GWG, exerting different effects on pregnancy outcomes. Given that we had a small number of patients with obesity, we failed to perform an independent analysis. Furthermore, we did not cover indicators affecting preterm birth, such as inflammation and nitric oxide, in our investigation. Additionally, we did not include dietary habits, smoking history, history of alcohol consumption, or family history of diabetes or hypertension, which may have influenced our research results.

Conclusion

In summary, our findings highlight the evidence that inappropriate prepregnancy BMI and GWG trajectories can increase the risk of APOs. To achieve favourable maternal-infant outcomes, it is important that women with prepregnancy emaciation or obesity actively improve their prepregnancy BMI through lifestyle intervention and that women with a normal prepregnancy BMI maintain optimal GWG. Further multicentre, multiregional, multiethnic prospective investigations with larger samples are warranted to establish optimal and generalisable GWG trajectories.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by Medical Ethics Committee of Shanghai Pudong New Area Healthcare Hospital for Women and Children (2020pdw01). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

We thank Dr. Zhang Yan for assistance with the experiments, Dr. Chu Liming and Zhang Zhiping for the valuable discussions, and Hu Yifan for help with the statistics. We also acknowledge all the women who participated in this study.

References

Footnotes

  • Contributors ZZ and ZM designed and conducted the investigation, who also drafted the manuscript. YH performed the statistical analysis. LC and YZ contributed to the study design. YZ and FW accomplished the data collection. YZ was responsible for the overall content as a guarantor. All authors read the final manuscript, and with an approval.

  • Funding This study was funded by the Technology Development Fund of Shanghai Pudong New Area (#: PKJ2020-Y72) and the Medical Discipline Construction Project of Pudong Health Committee of Shanghai (#: PWYts2021-19).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.

  • Provenance and peer review Not commissioned; externally peer reviewed.