https://orcid.org/0000-0001-5403-8168
Figures
Abstract
The geographical variation of maternal dietary patterns related to birth outcomes is important for improving the health of mothers and children; however, it is currently unknown. Thus, the objective of the study was to investigate geographical variations of maternal dietary pattern during pregnancy, and evaluate the spatial varying association of maternal dietary patterns in pregnancy with abnormal birth weight. A population-based cross-sectional study was conducted in Shaanxi province in Northwest China in 2013 to evaluate the relationship between abnormal birth weight and dietary pattern using the Geographically Weighted Logistic Regression (GWLR). Three dietary patterns during pregnancy were extracted through factor analysis, explaining approximately 45.8% of the variability of food intake. Approximately 81.6% of mothers with higher scores on the equilibrium pattern was more unlikely to have small for gestational age (SGA) infants, with the lower OR observed in Central and South Shaanxi. The snacks pattern was positively associated with low birth weight (LBW) for 23.2% of participants, with the highest OR in Central Shaanxi. Among about 80.0% of participants with higher scores on the snacks pattern living in South and Central Shaanxi, there was a higher risk for SGA. The OR values tend to descend from South to North Shaanxi. The OR values of the negative association between prudent pattern and LBW decreased from South to North Shaanxi among approximately 59.3% of participants. The prudent pattern was also negatively associated with the increasing risk of fetal macrosomia among 19.2% of participants living mainly in South Shaanxi. The association of maternal dietary patterns during pregnancy with abnormal birth weight varied geographically across Shaanxi province. The findings emphasize the importance of geographical distribution to improve the dietary patterns among disadvantaged pregnant women.
Citation: Liu Y, Zhang H, Zhao Y, Chen F, Mi B, Zhou J, et al. (2021) Geographical variations in maternal dietary patterns during pregnancy associated with birth weight in Shaanxi province, Northwestern China. PLoS ONE 16(7): e0254891. https://doi.org/10.1371/journal.pone.0254891
Editor: Frank T. Spradley, University of Mississippi Medical Center, UNITED STATES
Received: December 28, 2020; Accepted: July 6, 2021; Published: July 22, 2021
Copyright: © 2021 Liu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: The study was supported by National Natural Science of Foundation of China (No. 81602928, 81541069) and Natural Science Foundation of Shaanxi Province (No. 2017JM8102). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We certify that all authors have not financial interest related to the work of this paper. The authors have declared that no competing interests exist. Meanwhile, there are no prior publications or submissions with any overlapping information, including studies and patients. We have reviewed the final version of the manuscript and approve it for publication. To the best of our knowledge and belief, this manuscript has not been published in whole or in part nor is it being considered for publication elsewhere.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Dietary patterns are useful to examine the quality of a person’s overall diet related to health outcomes because it is acknowledged that consumption of individual food or nutrients does not occur in isolation but interacts with each other. Pregnancy is a critical period during which healthy diet characterized by consuming a variety of nutritious foods is essential for promoting and maintaining maternal and child health [1–4]. Because of socio-cultural and geographical reasons, maternal dietary patterns during pregnancy are different and population-specific in each area [5, 6]. Current studies regarding to the association of maternal dietary patterns during pregnancy with adverse birth outcomes have not always been consistent across regions [7, 8], which at least in part might have resulted from geographical heterogeneity. Geographical heterogeneity could cause biases for association estimates and their standard errors if they are not properly accounted for. Thus far, however, there was few evidence demonstrating the geographical difference in the relationship between maternal dietary patterns during pregnancy and birth outcomes of infants. Therefore, it is necessary to find an efficient way to uncover such geographically varying characteristics of maternal dietary patterns during pregnancy that are associated with birth outcomes, which would facilitate the identification of links between diet and adverse birth outcomes.
Shaanxi located in Northwest China, is one of the provinces with huge variations in geographical environment and dietary structure in different areas. Thus, geographical heterogeneity cannot be ignored when we study the spatial variation of maternal dietary patterns during pregnancy in relation to birth outcomes. Abnormal birth weight (e.g. low BW or high BW) representing an important adverse birth outcome is strongly correlated with various health outcomes including neonatal mortality and morbidity as well as developing chronic diseases throughout their life [9–12]. In previous studies, conventional global regression models were commonly used to explore the association of maternal dietary pattern during pregnancy with abnormal birth weight. The traditional global regression which ignores the impacts of local variations produces average parameters to reflect the geographical relationships equally over the whole study area. Due to the lack of consideration in geographical heterogeneity, evaluation by conventional global regression is inadequate. To address this issue, geographically weighted regression (GWR) model could be applied to explore how one dependent variable changes in response to one or more independent variables at the local scale. Through taking the samples within a defined neighborhood into calculation by giving more weights to nearby samples than those further away, GWR generates local coefficients to elucidate the relationships within the neighborhood. Using maps to show the geographical variation in the relative importance of each covariate, GWR provides a visually powerful way of conveying the observations to the public health practitioners.
In Shaanxi province of Northwest China, to the best of knowledge, how to evaluate the geographical variation of maternal dietary patterns that are related to birth outcomes are not known but are important for improving the health of mothers and children. Thus, to address the aforementioned research questions, the objectives of this study were twofold: (1) to investigate geographical variations of maternal dietary pattern during pregnancy; (2) use GWR to evaluate the geographically varying association of maternal dietary patterns during pregnancy with abnormal birth weight.
Materials and methods
Study design and participants
From August to November 2013, a large-scale cross-sectional survey was conducted in Shaanxi province, in Northwest China with the purpose to assess the birth outcomes. The subjects of the study were those live births born from January 2010 to November 2013 and their mothers. Participants were obtained by means of multi-stage random sampling method, that is, district random sampling was carried out after urban and rural areas were stratified to control for the impact of uneven distribution of dietary patterns on the risk of birth outcomes in different regions. In China, the administrative structure is composed of three layers, county, township and village in rural areas and district, street and community in urban areas, respectively. Based on the proportion of urban and rural areas, and considering the population density and fertility level, 20 counties in rural areas and 10 districts in urban areas were randomly selected by proportionate probability size (PPS) method. From each selected counties, 6 villages each from 6 townships were randomly selected. In urban areas, however, from each selected districts, 6 communities from each of 3 selected streets were sampled randomly. Finally, 10 babies born during 2010–2013 and their mothers in each chosen villages and 20 in each community were selected by a completely random sampling method. A total of 7,934 eligible women who completely fulfilled the diet survey were included into the study. The study was conducted in accordance with the Helsinki Declaration and was approved and reviewed by the Ethics Committee of Xi’an Jiaotong University Health Science Center (approval code: 2012008) [13].
Data collection
A structured questionnaire was used to obtain all information through face-to-face interview of mothers of the neonates. Firstly, socio-demographic characteristics were obtained by a standard family questionnaire. Secondly, dietary intakes of mothers during the entire pregnancy was assessed by a semi-quantitative food frequency questionnaire (FFQ). It has been demonstrated that FFQ in the survey has satisfied reliability and validity for pregnant women in rural western China [14, 15]. In the FFQ, there are 102 food items, including fourteen food groups (cereals, potatoes, legumes, vegetables, fungi, fruits, nuts, meats, dairy, eggs, fish, ethnic snacks, fast foods, and beverages) based on the China Food Composition Table [16]. The frequency of 102 food items was divided into eight categories ranging from "never" to "≥ 2 times/day". With the assistance of food portion images, the portion sizes for the 102 food items were recorded [14, 15]. The information of neonates including birth weight, birth date, and so on, was obtained from medical records.
Quality control
Field surveys and data collection were performed by the trained field staff from Xi’an Jiaotong University Health Science Center. To control the quality of the data, all questionnaires and procedures were pre-tested in field prior to the formal survey, and then detailed interview guidelines were formulated. A total of 10 teams from Xi’an Jiaotong University Health Science Center performed the surveys in the selected counties or districts. Each team was composed of 10–12 qualified investigators, a supervisor from, and a pediatrician from the local maternal and child health hospital. During the field survey, the supervisors monitored all field work and reviewed the questionnaire. When errors or missing items were identified, participants were re-interviewed immediately. With sustained support from local hospitals, health administration departments and the Provincial Ministry of Health, we collected all information within the local village clinics and community health service centers [13].
Study variables
Abnormal birth weight including low birth weight (LBW, birth weight < 2500 g), macrosomia (birth weight > 4000 g), small for gestational age (SGA, birth weight was below the 10th percentile of the average gestational age) and large for gestational age (LGA, birth weight was above the 90th percentile of the average gestational age) was considered as final outcome variables in the multivariate analysis [17, 18]. As the independent variable, dietary pattern had two dummy variables: T1 (1 = yes, 0 = no) and T3 (1 = yes, 0 = no), with T2 as a reference group. The socio-demographic variables include child gender (male, female), fetal number (singleton, twin and multi-fetal), infant parity (1, 2, or ≥ 3), childbearing age (18–24, 25–29, ≥ 30 years), mother’s education (primary school and below, junior high school, senior high school, college and above), mother’s residence during pregnancy (floating, permanent), Household wealth Index (poor, middle-income, rich), altitude of residence (< 500, 500–1000, ≥ 1000), and area (South Shaanxi, Central Shaanxi, North Shaanxi).
Statistical analysis
Epidata 3.1 software with double entry was used to establish the database. Maternal dietary patterns (equilibrium, snack, and prudent) were derived by principal component analysis (PCA) and varimax rotation method using SPSS software. Before deriving the dietary patterns, the reliability of PCA was verified by Kaiser–Meyer–Olkin value and Bartlett’s test of sphericity. The number of dietary patterns was calculated by eigenvalues (>1.2), variance contribution rate (>5%), the scree plot and the interpretability of the factors [19, 20]. Food groups with absolute factor loadings >0.3 were considered to be highly correlated to the identified factors. Factor scores were calculated for each mother to indicate if mother’s diet was consistent with the respective dietary patterns.
The vector map of China available at the National Platform for Common Geospatial Information Services (http://www.tianditu.com/) was used to obtain the geographic information (including geographic coordinates and altitude) for each township/street in Shaanxi province. The basic geographic unit in the study was townships and streets. Geographic coordinates (i.e., latitude/longitude) of the townships and streets was used to define their geographic location.
The GWLR model was employed to determine the relationship between abnormal birth weight and individual factors using the formula below [21]:
The equation hypothesizes that yi (dependent variables) was any abnormal birth weight for each individual i, xij represented a set of independent variables(j = 1, …, k) for the individual i, (ui, vi) was regarded as the x-y coordinates of the individual i; βij (regression coefficient) was deemed as the estimated effect of independent variable j for the individual i. The GWLR model is an extension and deformation of the typical linear regression model. The geographical data was incorporated into the model to create conditions for analyzing the geographical characteristics of the regression relationship.
We selected the GWR 4.0 software in this study (https://geodacenter.asu.edu/software/downloads/gwr_downloads) and the iterative reweighted least squares method was used to establish the GWLR model. During the process of modeling, a distance-based weighting scheme was used to assign weights to each township/street and the GWLR equation was estimated for each township/street. The kernel type and function for geographic weighting was used to estimate local coefficients for each township/street and bandwidth size was adaptive bisquare. Corrected Akaike Information Criterion (AICc) was used as the golden section search method to select the bandwidth of the adaptive kernel. The variability of each coefficient in geographic location was tested by comparing the difference [22]. In the GWLR model, OR corresponding to the unit change in the variables was obtained by exponentiation of the explanatory variable coefficients [22]. The variability of the influence of abnormal birth weight on individuals in different regions was evaluated by comparing the variability of each coefficient in each geographic location [23].
The analysis was completed by the following steps. Firstly, the effects of each dietary pattern on abnormal birth weight were determined by non-spatial logistic regression analysis using IBM SPSS Statistics 25.0. For bilateral test, the test level was set at the level of α = 0.05. Subsequently, all explanatory variables were entered into the GWLR model as described above. The variables with the largest p-value were gradually eliminated from the model until only those with local p-value <0.1 were obtained. Finally, ArcGIS 10.0 software (Environmental Systems Research Institute, Inc., Redlands, CA, US) was used to show the local average estimates (OR) and p-values of individuals on the Shaanxi map. These values were used to intuitively reflect the geographical differences in the relationship between abnormal birth weight outcome and individual factors.
Results
Baseline characteristics of the participants
In this study, a total of 173 towns/streets were chosen within 30 counties in Shaanxi province, and 7,934 live infants were included in the final analysis. Among the mothers included in the study, the average age was 26.88±4.62 years and the average parity was 1.44 ± 0.56. The proportions of study participants from northern, central and southern Shaanxi were 18.1, 51.1 and 30.7%, respectively. From south to north, Shaanxi is divided into three physiographic regions by the North Mountains and the Qin Mountains. These three parts are North Shaanxi, Central Shaanxi, and South Shaanxi. North Shaanxi is a plateau, central Shaanxi is a plain and south Shaanxi is a mountainous region with an elevation of about 1500–3000 meters. Of the infants, the average age was 8.92 ±6.30 months (range 0–45 months), 53.6% were male and 98.8% were singletons.
Characteristics of dietary pattern distribution during pregnancy among mothers
Through principal component analysis (PCA), three dietary patterns during pregnancy were extracted in Shaanxi province, northwest China, which explained approximately 45.8% of the variability in food intake (28.4%, 9.5% and 7.9% for three dietary patterns, respectively). According to the high positive loadings from various foods, including vegetables, homonemeae, beans meats, meat or flesh, fruit, fish and shrimp, potato, nuts, dairy, fast-food, sweetmeat, cereal, eggs and beverages, this first pattern was labelled as the ‘equilibrium pattern’. The highest and lowest scores of the equilibrium pattern were in the Central Shaanxi and Northern Shaanxi, respectively. (Fig 1). The second pattern labelled as ‘snacks pattern’ had high loading from beverages, sweetmeat, fast-food, dairy and eggs. The scores of the snacks pattern were the higher in North Shaanxi, followed by South Shaanxi and Central Shaanxi (Fig 2). The third pattern with high intakes of dairy products and eggs was labelled as the ‘prudent pattern’ (Table 1). The scores of the prudent pattern had decreasing trend of from south to north. (Fig 3).
Table 2 showed the sociodemographic characteristics and median daily intakes of energy and nutrients according to tertiles of dietary patterns scores among pregnant women in Shaanxi province. In terms of equilibrium pattern, pregnant women with high scores were more likely to live in south or central Shaanxi, areas with lower altitudes (<500 meters) and be better educated and wealthier. Besides, mothers with high scores on equilibrium pattern tend to have higher childbearing age and lower parity. The sociodemographic characteristics of the mothers with prudent pattern was similar to those with equilibrium pattern, except for maternal childbearing age which did not differ between tertiles of prudent pattern scores. On the contrary, mothers with high scores on the snacks pattern were more likely to live in north Shaanxi, areas with higher altitudes (>1000 meters), be less educated and poor, and be aged 18–24 years at delivery. Women with higher scores on the equilibrium pattern had higher intakes of energy and all selected nutrients, while those with lower scores on the prudent pattern and the snacks pattern had higher energy intake, carbohydrate intake and higher intakes of other selected nutrients.
Characteristics of abnormal birth weight
Among the live infants, the prevalence of abnormal birth weight was 3.7% for LBW, 6.2% for macrosomia, 10.9% for SGA, 10.0% for LGA, respectively. The prevalence of LBW and LGA were the highest in South Shaanxi (4.1% for LBW and 10.3% for SGA, respectively), followed by Central Shaanxi (3.7% and 10.2%) and North Shaanxi (3.3% and 10.1%) (S1 and S2 Figs). The incidence of macrosomia was 6.6% in North Shaanxi, 6.2% in Central Shaanxi and 6.1% in Southern Shaanxi respectively (S3 Fig). SGA incidence in Central Shaanxi (11.7%) was higher than that in South Shaanxi (10.9%) or Northern Shaanxi (10.0%). (S4 Fig). The prevalence of abnormal birth weight differed across the baseline characteristics, and are presented in Table 3. The incidence of LBW was significantly lower in T3 group of equilibrium pattern, T1 group of snacks pattern, and T1 or T3 group of prudent pattern, respectively (P<0.05). Likewise, there were lower prevalence of SGA in T3 group of equilibrium pattern and T1 group of snacks pattern (P<0.05).
Association between dietary patterns and abnormal birth weight using non-spatial logistic regression
Non-spatial logistic regression results showed that SGA was less likely to occur among the infants born by mothers in the T3 group of equilibrium pattern in comparison with those whose mothers were from T2 group (OR: 0.801, 95%CI: 0.665–0.965) (Table 4). Mothers with higher scores of snacks pattern were associated with increasing risk of macrosomia and SGA among the infants (OR: 1.265, 95%CI: 1.000–1.602; OR: 1.260, 95%CI: 1.056–1.505, respectively). Compared with those in the T2 group of prudent pattern, mothers in the T1 group of the prudent pattern had higher risk of LBW in their offspring (OR: 0.732, 95%CI: 0.537–0.997), and mothers in the T3 group had lower risk of fetal macrosomia (OR: 0.788, 95%CI: 0.623–0.997).
Association between dietary patterns and abnormal birth weight using spatial logistic regression
Based on geographically weighted regression (GWR) for LBW, macrosomia, SGA and LGA in the present study, the optimal bandwidth size were 7332.000, 6631.000, 7596.000, and 7573.000 (persons) respectively. Comparing the global regression and GWR, we concluded that the local regression has better fitness. Approximately 81.6% of mothers with higher scores on the equilibrium pattern was more unlikely to have SGA infants (P < 0.05), with the highest OR observed in North Shaanxi and the lowest OR observed in South Shaanxi. (Table 5 and S5 Fig). The higher scores of mothers with snacks pattern during pregnancy were found to be positively associated with LBW for 23.2% of participants (P< 0.05), with decreasing OR values from central Shaanxi to the north and south Shaanxi (Table 5 and S6 Fig). Similarly, infants born by women in the T3 group of snacks pattern was significantly correlated with macrosomia for 19.5% of participants (P< 0.05), with decreasing OR values from south to north (Table 5 and S7 Fig). Among about 80.0% of participants who lived in South and Central Shaanxi, there was a higher risk for SGA in the T3 group of snacks pattern than that in the T2 group (P < 0.05). The OR values tend to descend from South to North Shaanxi (Table 5 and S8 Fig). With T2 group of prudent pattern as the reference, a lower risk for LBW among the infants were observed among approximately 75.0% of participants in T1 group and 59.3% of those in T3 group, respectively (P< 0.05). For the OR values of LBW, there was a tendency to decrease from South to North Shaanxi (Table 5, S9 and S10 Figs). The mothers with higher scores of prudent pattern was also negatively associated with the increasing risk of fetal macrosomia among 19.2% of participants (P<0.05), with adjusted OR values ranging from 0.875 (South Shaanxi) to 0.939 (Central Shaanxi) (Table 5 and S11 Fig).
Discussion
In the present study, three dietary patterns (equilibrium pattern, prudent pattern and snacks pattern) were identified among the mothers during pregnancy in Shaanxi province of northwestern China. We used spatial regression models to explore the geographic variation of maternal dietary patterns in relation to the risk of abnormal birth weight after controlling for other baseline characteristics.
Previous studies suggested that the single nutrient analysis in nutritional epidemiology failed to account for interactions between food components or foods. Dietary patterns represent a broader picture of food and nutrient intakes and are able to account for cumulative and interactive effects [5]. The results suggested that pregnant women with high adherence to the equilibrium pattern tended to eat a diet with a balance of all of the food groups and foods that are rich in energy and various nutrients. Moreover, participants with high adherence to the equilibrium pattern were more likely to be better educated, wealthier, 25–29 years old at delivery, and living at lower altitudes. The characteristics of this dietary model identified in our study was homologous to the dietary patterns in Guangzhou, China [24]. The prudent pattern was characterized by high intakes of dairy products and eggs, and the snacks pattern had high loading of beverages, sweetmeat, and fast-food. Those with higher scores on the prudent pattern and the snacks pattern had lower energy and carbohydrate intake and lower intakes of some selected nutrients. The sociodemographic distribution of mothers with prudent pattern was similar to those with equilibrium pattern, while those with high scores on the snacks pattern were more likely to live in areas with higher altitude (>1000 meters), be poor and less educated, and be aged 18–24 years at delivery. Some studies reported that higher education level and better economic status might enable individuals to make healthy diet choices [25, 26]. Hence, people with higher education level and better economic status tend to be more inclined to adopt the equilibrium pattern [27–30]. Conversely, low socioeconomic groups with relatively low level of nutrition knowledge may make unhealthy diet choices and lead to a poor-quality snacks pattern [31].
Due to the huge difference in geographical environment and socio-culture across various areas in China, it was well recognized that the dietary habits and food consumption have different geographical distribution [32, 33]. Similarly, dietary habits also varied largely among different areas in Shaanxi [34]. In our studies, the highest scores of the equilibrium pattern were observed in Central Shaanxi, while the lowest score of the equilibrium pattern was observed in North Shaanxi. The scores of the snacks pattern were higher in North Shaanxi, followed by South Shaanxi and Central Shaanxi. It appeared a decreasing trend of the scores of the prudent pattern from south to north. It is possible that the different geographical environment (e.g. the area altitudes) resulted in different geographical distribution of dietary patterns.
Healthy diet characterized by consuming a variety of nutritious foods during pregnancy is essential for promoting and maintaining fetal health [7, 28]. Global regression analysis revealed that mothers with high adherence to the equilibrium pattern during pregnancy were more unlikely to give birth to a baby with SGA. The findings suggested that equilibrium pattern with a balance of all of the food groups and foods may be beneficial to reduce the risk of SGA, which provided more evidence that healthy diets promote birth weight. Further, the GWLR model showed that equilibrium pattern might reduce more neonates with SGA in Central and South Shaanxi than those in North Shaanxi. Our survey data indicated that the scores of equilibrium pattern was higher in Central and South Shaanxi than those in North Shaanxi. Most parts of Central Shaanxi are developed urban areas (e.g., Xi ’an) where mothers have higher household economic levels and higher education. In contrast, the surveyed area in North Shaanxi comprises of small townships and villages where mothers have lower economic levels and are less educated. Pregnant women in villages with greater density are more likely to eat high-starch foods and tend to adopt more frugal dietary structure [35]. These results remind us that it is extremely urgent to advocate the equilibrium diet during pregnancy, especially in North Shaanxi. Additional studies will be required to corroborate these findings and to elucidate possible reasons for this geographical variations.
There is a high intake of snack food among the Chinese in recent years [36]. The snacks foods as an unhealthy food choice is unable to fulfill the increased key micronutrients requirements for normal fetal growth and may lead to a higher risk of some diseases [37]. According to the global regression, the higher scores of snacks pattern was associated with the increasing risk of macrosomia and SGA among the infants. The snacks pattern consists of foods with high concentrations of simple carbohydrates, lipids and low amounts of protein and micronutrients. Insufficient protein intake during pregnancy affects fetal organ development, leading to a variety of adverse pregnancy outcomes [38]. During pregnancy, fetal growth and development are very rapid, and inadequate nutrition intake can lead to fetal growth retardation and low birth weight [39]. However, some studies also found that this combination of substances in the snacks pattern may lead to rapid fetal weight gain and increase the incidence of macrosomia [40]. Therefore, the snacks pattern contains unbalanced nutrients, favoring abnormal birth weight [41, 42].
GWLR model further revealed that snacks pattern during pregnancy may increase the risk of LBW, Macrosomia and SGA in 23.2%, 19.5% and 80.0%, respectively of all participants. When we focus on the map drawn by OR values, it is clear that the highest OR values of LBW were in Central Shaanxi, and OR values of Macrosomia and SGA progressively decreased from South to North Shaanxi. Thus, it is essential to improve the maternal diet status during pregnancy according to regional characteristics.
The spatial regression model revealed a negative association between high propensity of prudent pattern and the risk of LBW in 59.3% of mothers residing in Central Shaanxi and parts of North Shaanxi. On the map used to visualize these results, prudent pattern might decrease more neonates with LBW in North Shaanxi, followed by Central Shaanxi. The negative association of prudent pattern with macrosomia were mainly present in South Shaanxi. Because the prudent pattern has high contents of dairy products and eggs, which included higher protein and micronutrients compared to snacks pattern, it might benefit maintaining appropriate birth weight. However, the geographical variations after controlling for the important risk factors provide potential policy implications, calling for further investigations.
However, we have to acknowledge some limitations that need to be addressed in the study. First, due to the nature of cross-sectional design, we cannot draw any cause-effect conclusion based on the study results. Secondly, it was difficult to isolate the variation trends in relevant factors based on limited number of areas sampled in our study. Thirdly, we adopted the mothers’ self-reported information on dietary intake during pregnancy. Thus, the possibility of information bias should be considered. Different degree of the reminder bias based on different birth dates is also one of the limitations of our survey even though we adopted systematic approaches to maximize power and minimize bias. Fourthly, the study may be subject to some potential confounders. For example, maternal nutrition status of pregnant (BMI), drug use, genetics, and environmental factors were not available and thus could not be controlled in multivariate analysis. Despite these limitations, this study has still provided important information on the geographic varying association of dietary patterns with abnormal birth weight in northwestern China.
Conclusions
In summary, three dietary patterns (equilibrium pattern, prudent pattern and snacks pattern) were extracted among the mothers during pregnancy, which had different geographical distribution in Shaanxi province of northwestern China. The equilibrium pattern, which is rich in energy and various nutrients, was negatively associated with SGA. The prudent pattern were more likely to decrease the risk of LBW and macrosomia, respectively mainly in North Shaanxi and South Shaanxi. The snacks pattern during pregnancy may cause abnormal birth weight (e.g. LBW, macrosomia, SGA) among infants, especially in Central and South Shaanxi. The findings emphasize the importance of geographical distribution to improve the dietary patterns among disadvantaged pregnant women.
Supporting information
S1 Table. The association of diet pattern with abnormal birth outcomes using non-spatial logistical regression in South Shaanxi.
https://doi.org/10.1371/journal.pone.0254891.s001
(DOCX)
S2 Table. The association of diet pattern with abnormal birth outcomes using non-spatial logistical regression in Central Shaanxi.
https://doi.org/10.1371/journal.pone.0254891.s002
(DOCX)
S3 Table. The association of diet pattern with abnormal birth outcomes using non-spatial logistical regression in North Shaanxi.
https://doi.org/10.1371/journal.pone.0254891.s003
(DOCX)
S4 Table. The characteristics of dietary patterns among pregnant women (n 7934) in Shaanxi province, Northwest China, 2013.
https://doi.org/10.1371/journal.pone.0254891.s004
(DOCX)
S5 Table. The association of diet pattern with abnormal birth outcomes using non-spatial logistical regression.
https://doi.org/10.1371/journal.pone.0254891.s005
(DOCX)
S1 Fig. Geographical distribution of the incidence of LBW among the live infants in Shaanxi province, 2013.
https://doi.org/10.1371/journal.pone.0254891.s006
(TIF)
S2 Fig. Geographical distribution of the incidence of macrosomia among the live infants in Shaanxi province, 2013.
https://doi.org/10.1371/journal.pone.0254891.s007
(TIF)
S3 Fig. Geographical distribution of the incidence of SGA among the live infants in Shaanxi province, 2013.
https://doi.org/10.1371/journal.pone.0254891.s008
(TIF)
S4 Fig. Geographical distribution of the incidence of LGA among the live infants in Shaanxi province, 2013.
https://doi.org/10.1371/journal.pone.0254891.s009
(TIF)
S5 Fig. Geographical distribution of adjusted ORs of equilibrium pattern of T3, as associated with prevalence of SGA.
https://doi.org/10.1371/journal.pone.0254891.s010
(TIF)
S6 Fig. Geographical distribution of adjusted ORs of snacks pattern of T3, as associated with prevalence of LBW.
https://doi.org/10.1371/journal.pone.0254891.s011
(TIF)
S7 Fig. Geographical distribution of adjusted ORs of snacks pattern of T3, as associated with prevalence of macrosomia.
https://doi.org/10.1371/journal.pone.0254891.s012
(TIF)
S8 Fig. Geographical distribution of adjusted ORs of snacks pattern of T3, as associated with prevalence of SGA.
https://doi.org/10.1371/journal.pone.0254891.s013
(TIF)
S9 Fig. Geographical distribution of adjusted ORs of prudent pattern of T1, as associated with prevalence of LBW.
https://doi.org/10.1371/journal.pone.0254891.s014
(TIF)
S10 Fig. Geographical distribution of adjusted ORs of prudent pattern of T3, as associated with prevalence of LBW.
https://doi.org/10.1371/journal.pone.0254891.s015
(TIF)
S11 Fig. Geographical distribution of adjusted ORs of prudent pattern of T3, as associated with prevalence of macrosomia.
https://doi.org/10.1371/journal.pone.0254891.s016
(TIF)
Acknowledgments
We thank the women and their children participated in our study. We thank local hospitals and health bureau as well as the Ministry of Health in Shaanxi province.
References
- 1. Stephens TV, Woo H, Innis SM, Elango R. Healthy pregnant women in Canada are consuming more dietary protein at 16- and 36-week gestation than currently recommended by the Dietary Reference Intakes, primarily from dairy food sources. Nutr Res. 2014 Jul;34(7):569–576. Epub 2014 Jul 8. pmid:25150115.
- 2. Cetin I, Berti C, Calabrese S. Role of micronutrients in the periconceptional period. Hum Reprod Update. 2010 Jan-Feb;16(1):80–95. pmid:19567449.
- 3. Wrottesley SV, Lamper C, Pisa PT. Review of the importance of nutrition during the first 1000 days: maternal nutritional status and its associations with fetal growth and birth, neonatal and infant outcomes among African women. J Dev Orig Health Dis. 2016 Apr;7(2):144–162. Epub 2015 Aug 17. pmid:26279311.
- 4. Paknahad Z, Fallah A, Moravejolahkami AR. Maternal Dietary Patterns and Their Association with Pregnancy Outcomes. Clin Nutr Res. 2019 Jan 25;8(1):64–73. pmid:30746349.
- 5. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol. 2002 Feb;13(1):3–9. pmid:11790957.
- 6. Völgyi E, Carroll KN, Hare ME, Ringwald-Smith K, Piyathilake C, Yoo W, et al. Dietary patterns in pregnancy and effects on nutrient intake in the Mid-South: the Conditions Affecting Neurocognitive Development and Learning in Early Childhood (CANDLE) study. Nutrients. 2013 May 3;5(5):1511–1530. pmid:23645026.
- 7. Thompson JM, Wall C, Becroft DM, Robinson E, Wild CJ, Mitchell EA. Maternal dietary patterns in pregnancy and the association with small-for-gestational-age infants. Br J Nutr. 2010 Jun;103(11):1665–1673. Epub 2010 Mar 9. pmid:20211035.
- 8. Northstone K, Emmett PM, Rogers I. Dietary patterns in pregnancy and associations with nutrient intakes. Br J Nutr. 2008 Feb;99(2):406–415. Epub 2007 Sep 3. pmid:17764600.
- 9. Bukowski R, Hansen NI, Willinger M, Reddy UM, Parker CB, Pinar H, et al. Eunice Kennedy Shriver National Institute of Child Health and Human Development Stillbirth Collaborative Research Network. Fetal growth and risk of stillbirth: a population-based case-control study. PLoS Med. 2014 Apr 22;11(4):e1001633. pmid:24755550.
- 10. Grissom NM, Reyes TM. Gestational overgrowth and undergrowth affect neurodevelopment: similarities and differences from behavior to epigenetics. Int J Dev Neurosci. 2013 Oct;31(6):406–414. Epub 2012 Nov 28. pmid:23201144.
- 11. Wennerström EC, Simonsen J, Melbye M. Long-Term Survival of Individuals Born Small and Large for Gestational Age. PLoS One. 2015 Sep 21;10(9):e0138594. pmid:26390219.
- 12. Francis JH, Permezel M, Davey MA. Perinatal mortality by birthweight centile. Aust N Z J Obstet Gynaecol. 2014 Aug;54(4):354–359. Epub 2014 Apr 15. pmid:24731210.
- 13. Pei L, Kang Y, Zhao Y, Yan H. Prevalence and risk factors of congenital heart defects among live births: a population-based cross-sectional survey in Shaanxi province, Northwestern China. BMC Pediatr. 2017 Jan 13;17(1):18. pmid:28086762.
- 14. Cheng Y, Dibley MJ, Zhang X, Zeng L, Yan H. Assessment of dietary intake among pregnant women in a rural area of western China. BMC Public Health. 2009 Jul 9;9:222. pmid:19589154.
- 15. Cheng Y, Yan H, Dibley MJ, Shen Y, Li Q, Zeng L. Validity and reproducibility of a semi-quantitative food frequency questionnaire for use among pregnant women in rural China. Asia Pac J Clin Nutr. 2008;17(1):166–177. pmid:18364342.
- 16.
Institute of Nutrition and Food Safety, China Center for Disease Control and Prevention (2009) China Food Composition Book 1, 2nd ed. Beijing: Peking University Medical Press.
- 17.
Jie L. Obstetrics and Gynecology [M]. 8 version. Beijing: People’s Medical Publishing House; 2013. pp.47–58.
- 18. Ikem E, Halldorsson TI, Birgisdóttir BE, Rasmussen MA, Olsen SF, Maslova E. Dietary patterns and the risk of pregnancy-associated hypertension in the Danish National Birth Cohort: a prospective longitudinal study. BJOG. 2019 Apr;126(5):663–673. Epub 2019 Jan 24. pmid:30675768.
- 19. Schulze MB, Hoffmann K, Kroke A, Boeing H. An approach to construct simplified measures of dietary patterns from exploratory factor analysis. Br J Nutr. 2003 Mar;89(3):409–419. pmid:12628035.
- 20. Shi Z, Hu X, Yuan B, Pan X, Dai Y, Holmboe-Ottesen G. Association between dietary patterns and anaemia in adults from Jiangsu Province in Eastern China. Br J Nutr. 2006 Nov;96(5):906–912. pmid:17092381.
- 21. Yang TC, Matthews SA. Understanding the non-stationary associations between distrust of the health care system, health conditions, and self-rated health in the elderly: a geographically weighted regression approach. Health Place. 2012 May;18(3):576–585. Epub 2012 Jan 28. pmid:22321903.
- 22. Zhou YB, Wang QX, Liang S, Gong YH, Yang MX, Chen Y, et al. Geographical variations in risk factors associated with HIV infection among drug users in a prefecture in Southwest China. Infect Dis Poverty. 2015 Sep 2;4:38. pmid:26329887.
- 23. Nakaya T, Fotheringham AS, Brunsdon C, Charlton M. Geographically weighted Poisson regression for disease association mapping. Stat Med. 2005 Sep 15;24(17):2695–2717. pmid:16118814.
- 24. He JR, Yuan MY, Chen NN, Lu JH, Hu CY, Mai WB, et al. Maternal dietary patterns and gestational diabetes mellitus: a large prospective cohort study in China. Br J Nutr. 2015 Apr 28;113(8):1292–1300. Epub 2015 Mar 30. pmid:25821944.
- 25. Mi W, Lian W, Yang M, Jiang Y, Yi W, Wang Y, et al. Analysis on dietary patterns and influencing factors among the elderly in Yantai City. Wei Sheng Yan Jiu. 2017 Jan;46(1):27–31. Chinese. pmid:29903147.
- 26. Luksiene DI, Baceviciene M, Tamosiunas A, Daugeliene E, Kranciukaite D. Health, alcohol and psychosocial factors in Eastern Europe study: dietary patterns and their association with socio-demographic factors in the Lithuanian urban population of Kaunas city. Int J Public Health. 2011 Apr;56(2):209–216. Epub 2010 Aug 12. pmid:20703510.
- 27. Carniel EF, Zanolli ML, Antonio MARGM, Morcillo AM. Determinantes do baixo peso ao nascer a partir das Declarações de Nascidos Vivos.[J]. Revista Brasileira de Epidemiologia. 2008;11(1):169–179.
- 28. Cucó G, Fernández-Ballart J, Sala J, Viladrich C, Iranzo R, Vila J, et al. Dietary patterns and associated lifestyles in preconception, pregnancy and postpartum. Eur J Clin Nutr. 2006 Mar;60(3):364–71. pmid:16340954.
- 29. Ferrer C, Garcia-Esteban R, Mendez M, Robieu I, Sunyer MTJ. Determinantes sociales de los patrones dietéticos durante el embarazo[J]. Gaceta Sanitaria. 2009;23(1):38–43. pmid:19231721
- 30. Northstone K, Emmett P, Rogers I. Dietary patterns in pregnancy and associations with socio-demographic and lifestyle factors. Eur J Clin Nutr. 2008 Apr;62(4):471–479. Epub 2007 Mar 21. pmid:17375108.
- 31. Taylor MK, Sullivan DK, Ellerbeck EF, Gajewski BJ, Gibbs HD. Nutrition literacy predicts adherence to healthy/unhealthy diet patterns in adults with a nutrition-related chronic condition. Public Health Nutr. 2019 Aug;22(12):2157–2169. Epub 2019 May 31. pmid:31146797.
- 32. Cheng C.K. The culture of Chinese diet: Regional differentiation and developing trends[J]. Acta Geographica Sinica, 1994, 49(3): 226–235.
- 33. Song F, Cho MS. Geography of Food Consumption Patterns between South and North China. Foods. 2017 May 5;6(5):34. pmid:28475146.
- 34. Wang J, Lin X, Bloomgarden ZT, Ning G. The Jiangnan diet, a healthy diet pattern for Chinese. J Diabetes. 2020 May;12(5):365–371. Epub 2020 Jan 26. pmid:31846221.
- 35. Hajianfar H, Esmaillzadeh A, Feizi A, Shahshahan Z, Azadbakht L. The Association Between Major Dietary Patterns and Pregnancy-related Complications. Arch Iran Med. 2018 Oct 1;21(10):443–451. pmid:30415552.
- 36. Wang D, He Y, Li Y, Luan D, Yang X, Zhai F, et al. Dietary patterns and hypertension among Chinese adults: a nationally representative cross-sectional study. BMC Public Health. 2011 Dec 14;11:925. pmid:22168909.
- 37. Shang X, Li Y, Liu A, Zhang Q, Hu X, Du S, et al. Dietary pattern and its association with the prevalence of obesity and related cardiometabolic risk factors among Chinese children. PLoS One. 2012;7(8):e43183. Epub 2012 Aug 14. pmid:22905228.
- 38. Flynn AC, Schneeberger C, Seed PT, Barr S, Poston L, Goff LM. The Effects of the UK Pregnancies Better Eating and Activity Trial Intervention on Dietary Patterns in Obese Pregnant Women Participating in a Pilot Randomized Controlled Trial. Nutr Metab Insights. 2016 Jun 27;8(Suppl 1):79–86. pmid:27385914.
- 39. Diemert A, Lezius S, Pagenkemper M, Hansen G, Drozdowska A, Hecher K, et al. Maternal nutrition, inadequate gestational weight gain and birth weight: results from a prospective birth cohort. BMC Pregnancy Childbirth. 2016 Aug 15;16:224. pmid:27528213.
- 40. Flynn AC, Seed PT, Patel N, Barr S, Bell R, Briley AL, et al. UPBEAT consortium. Dietary patterns in obese pregnant women; influence of a behavioral intervention of diet and physical activity in the UPBEAT randomized controlled trial. Int J Behav Nutr Phys Act. 2016 Nov 29;13(1):124. pmid:27894316.
- 41. Coelho Nde L, Cunha DB, Esteves AP, Lacerda EM, Theme Filha MM. Dietary patterns in pregnancy and birth weight. Rev Saude Publica. 2015;49:62. Epub 2015 Sep 18. pmid:26398873.
- 42. Uusitalo U, Arkkola T, Ovaskainen ML, Kronberg-Kippilä C, Kenward MG, Veijola R, et al. Unhealthy dietary patterns are associated with weight gain during pregnancy among Finnish women. Public Health Nutr. 2009 Dec;12(12):2392–2399. Epub 2009 Mar 27. pmid:19323867.