Study on the correlation between maternal serum uric acid and foetal birth weight in Naqu, Tibet

Abstract In high-altitude regions, low birth weight is mainly caused by hypoxia. We aimed to determine whether maternal serum uric acid (SUC) level was associated with decreased foetal birth weight. The relevant data of individual pregnant women who delivered between 37 and 40 weeks in the People’s Hospital of Naqu City, Tibet were retrospectively collected. The correlation between maternal SUC and birth weight was examined using multivariate linear regression analysis and subgroup analysis. The results showed that there was a significant negative correlation between SUC and birth weight in pregnant women with proteinuria, female foetuses, and primiparas. Fitting smoothing curve analysis showed that there was a negative linear correlation between SUC and birth weight in primiparas and female foetuses. Maternal SUC is negatively associated with foetal birth weight in a single pregnancy with proteinuria, primipara, or female foetuses in the Naqu region of Tibet, China. IMPACT STATEMENT What is already known on this subject? Preeclampsia associated with hyperuricaemia can affect foetal birth weight, foetal birth weight in plains area is negatively correlated with maternal hyperuricaemia. What do the results of this study add? Maternal SUC was negatively correlated with foetal birth weight, especially in primipara, mothers with proteinuria, and pregnant girls. What are the implications of these findings for clinical practice and/or further research? The results suggest that attention should be paid to SUC in pregnant women, especially in primipara, mothers with proteinuria, and pregnant girls, in the prevention of low birth weight infants in Naqu Plateau area of Tibet.


Introduction
Low birth weight (LBW) is a serious public health concern (Tshotetsi et al. 2019).Infants with LBW have a higher mortality rate in the first month, and those who survive are at higher risk of developmental delay, low intelligence quotient, and chronic diseases, such as obesity, hypertension, and diabetes in adulthood (Xia et al. 2019).Social factors, maternal malnutrition (Buekens et al. 2013), younger mothers (Fraser et al. 1995), primipara (de Sanjose and Roman 1991), and smoking (van den Berg et al. 2012) are all known as major factors affecting the birth weight of newborns.
Altitude has an important effect on foetal growth, development, and birth weight (Julian et al. 2009).A large number of studies in different high-altitude populations (Tibet, the Andes, and Ethiopia) demonstrated a reduction in birth weight at high altitudes (Leonard et al. 1995).Compared with the populations at sea level, the distribution of birth weight shifted to the left (Zamudio andMoore 2000a, Moore 2017), and the incidence of LBW in newborns born at higher altitudes (above 2000 m) was increased by 2-3 fold, and was mainly related to the higher incidence of intrauterine growth retardation (IUGR) (Keyes et al. 2003).Although it has been reported that the ancestors who lived in the Tibet for a long time had some resistance to the retardation of foetal development caused by the plateau (Zamudio et al. 1993b), the average birth weight of newborns in the plateau area was still lower than that in the plains area (Giussani et al. 2001).Hypoxia is the most notable cause of LBW at high altitudes (Delgado-Rodr� ıguez et al. 1998), but it is difficult to correct altitude hypoxia under current conditions; thus, we examined other factors that may affect birth weight at high altitudes.Hypoxia is the main cause of hyperuricaemia (Ohuchi et al. 2015); therefore, hyperuricaemia is related to LBW, and at present, studies have only reported that high serum uric acid (SUC) in preeclampsia pregnant women is positively correlated with low foetal birth weight and poor prognosis (Koopmans et al. 2009, Bellomo et al. 2011).
Hyperuricaemia is extremely common in high-altitude populations; therefore, there is a correlation between maternal SUC and birth weight at high altitudes.The purpose of this study was to evaluate the association between maternal SUC levels and foetal birth weight.

Study population
Naqu is located in the north of the Tibet Autonomous Region, in the hinterland of the Qinghai-Tibet Plateau, with an average elevation of 4,100 metres, typical of plateau areas.We retrospectively collected the medical records of women who gave birth in Naqu People's Hospital (Naqu, Tibet, China) between January and June 2019.The inclusion criteria were primigravida or multipara, singleton pregnancy, and delivery at � 37 weeks.Patients with underlying chronic hypertension or pre-existing hypertension, gouty arthritis, chronic kidney disease stage 3 (creatinine clearance < 60 ml/min), on medication for uric acid reduction, smoking, and stillbirth were excluded.In total, we recruited 760 pregnant women aged 16 -47 years old at 37-40 gestational weeks (GA).

Collection of data and specimens
Basic demographic data were collected from each recruited pregnant woman via medical records in the hospital information system, and included age, weight, height, blood pressure, serum triglyceride (TG), serum total cholesterol, blood urea nitrogen (BUN), GA, SUC, fasting blood glucose (FBG) and proteinuria.The height (cm) and weight (kg) of each subject were determined 1 week prior to delivery.GA was calculated according to the last menstrual period.After admission, blood pressure of the right upper limb was measured three times with a mercury sphygmomanometer in quiet state, and the mean value was taken as the patient's blood pressure.Urine protein was qualitatively detected in the middle of the first voiding in the morning.Laboratory indicators were determined using an automatic biochemical analyser (Olympus AU5800; Tokyo, Japan) at the Naqu People's Hospital.A range of birth outcomes were recorded for each foetus, including neonatal sex and birth weight.

Study outcome definition
According to the World Health Organisation definition of LBW, a LBW child is defined as weighing less than 2,500 grams [1].Hypertension was defined as a systolic blood pressure of 140 mmHg or greater and a diastolic blood pressure of 90 mmHg or greater.

Statistical analysis
Continuous variables were expressed as mean ± standard deviation, while categorical variables were expressed as percentages.The independent samples t-test and chi-squared test were used to detect statistical differences between groups.A univariate regression model was used to assess the association between clinical variables and birth weight.The relationship between SUC and birth weight was estimated using a multivariate linear regression model shown as odds ratio (OR) and 95% confidence interval (CI).The results from unadjusted, minimally adjusted, and fully adjusted analyses were shown simultaneously according to the recommendation of the Strengthening the Reporting of Observational Studies in Epidemiology statemen (von Elm et al. 2008).To ensure accuracy, we built three models: (1) unadjusted; (2) minimally adjusted model (adjusted for weight, proteinuria, BUN, TG, gravidity, and GA); and (3) fully adjusted model (adjusted for age, standing height, weight, proteinuria, BUN, TG, foetal sex, gravidity, GA, FBG, and systolic pressure(categorical variable)).Hierarchical multivariate regression analysis was used to analyse subgroups.We performed interaction and mediation analyses to determine whether the effect of SUC on birth weight was mediated through primiparas, female foetuses, and proteinuria.We further analysed the shape of the relationship between SUC and birth weight using the generalised additive model (GAM) and by fitting smoothing curves (penalty spline method).
All analyses were performed using R statistical software (http://www.R-project.org,The R Foundation) and Empower Stats (http://www.empowerstats.com,X&Y Solutions, Inc., Boston, MA).The level of significance of each test was set at P < 0.05.The regression coefficient, adjusted OR, 95% CI and p-value were reported.

Results
In total, we recruited 1010 pregnant females aged 16-47 years at a gestational age (GA) of 28-40 weeks.We deleted 103 cases without serum uric acid, 42 cases without gestational age and 105 cases preterm (Figure 1).Ultimately, 760 pregnant women, between the ages of 16 and 47 years and 37 to 40 GA, were enrolled.There were 55 cases (7.24%) of LBW infants.The results showed that the weight and height of pregnant women in the LBW group were lower than those in the normal birth weight group (p < 0.05).The incidence of diastolic blood pressure, BUN, proteinuria, and the proportion of female newborns in the LBW group were higher than those in the normal birth weight group (p < 0.05, Table 1).
To fully screen for factors related to birth weight, we first performed a univariate analysis.Variables with p < 0.1 in univariate analysis, were screened to determine if they may be related to the birth weight of newborns.The results showed that maternal weight, standing height, body mass index (BMI), BUN, SUC, FBG, proteinuria, primipara, systolic pressure (categorical variable), and foetal sex may be related to birth weight.Multifactor regression analysis was further applied to detect the correlation between the above factors and the birth weight of newborns.Without adjusting for confounders, SUC and foetal sex were correlated with the birth weight of newborns (Table 2).
After screening the covariates, unadjusted, minimally adjusted, and fully adjusted models were established.Multivariate linear regression analysis showed that SUC was negatively correlated with birth weight (b¼-0.52,95% CI: À 1.02, À 0.03).However, further analysis of SUC as a categorical variable showed that there was no correlation between SUC and birth weight in the fully adjusted model (Table 3); thus, we suspect that there may be special groups in the research subjects.
To find a special population, stratification analysis was performed according to proteinuria, foetal sex, and primipara.
The results showed that there was a significant negative correlation between SUC and birth weight in proteinuriapositive, female foetal, and primiparous mothers (Figure 2).To determine whether hyperuricaemia plays a role independent of proteinuria, foetal sex and primiparia, we performed interaction and mediation analyses.There was no interaction between proteinuria, foetal sex, primipara, and SUC on birth weight.The interaction analysis after partial and complete adjustment showed that proteinuria and SUC had no effect on birth weight (Supplementary Table 1).We further analysed the mediating effects of proteinuria, foetal sex, primipara, and SUC on birth weight, and found that proteinuria, foetal sex, and primipara did not play a significant mediating role (Supplementary Table 2).
To further investigate the relationship between SUC and birth weight in positive proteinuria, female newborns, or primiparas, we used a GAM and a fitted smoothing curve.The results showed that there were linear relationships between SUC and birth weight in primiparas (Figure 3(a)) or female foetuses (Figure 3(b)).In pregnant women with positive proteinuria, the relationship between SUC and birthweight was curved line (Supplementary Figure 1); but when four outliers of SUC greater than 500 lmol/L were deleted, further analysis showed that there was a linear relationship between SUC and birth weight in proteinuria-positive pregnant women (b ¼ À 2.38, CI:-3.78,À 0.98; p ¼ 0.0012): birth weight decreased by 2.38 g per 1 lmol/L of SUC (Figure 3(c)).Smooth curve fitting analysis was performed in pregnant women with positive proteinuria according to foetal gender stratification.This showed that there was still a linear relationship between SUC and birth weight during the delivery of female foetuses.However, there was an inverted U-shaped relationship between the SUC of pregnant women with positive proteinuria and the birth weight of male foetuses (Figure 3(d)), and the SUC cut-off point was 338 lmol/L (Supplementary Table 3).

Discussion
In highland areas, the decline in foetal growth rate begins in early gestation and is chronic and progressive, becoming more pronounced in the last GA (Revollo et al. 2017).The LBW of newborns in the plateau region has become one of the main factors affecting the survival rate of newborns in the plateau region (Zamudio andMoore 2000a, Moore 2017), which also increases the social burden; the data for this study was obtained from the only grade A hospital in Naqu, Tibet, which is representative of neonatal data in this region.Univariate analysis showed that maternal weight, standing height, BMI, BUN, SUC, FBG, proteinuria, primipara, systolic pressure (categorical variable), and foetal sex may be related to birth weight, which is consistent with the related factors affecting neonatal birth weight reported in the past (Sema et al. 2019).One of the mechanisms of preeclampsia hyperuricaemia is maternal renal dysfunction, which results in reduced uric acid clearance (Masoura et al. 2015).In the present study, the creatinine level of pregnant women in the LBW group did not increase, and only the serum urea  The incidence of diastolic blood pressure, BUN, proteinuria, and the proportion of female newborns in the LBW group were higher than those in the normal birth weight group (p < 0.05).HB: haemoglobin; Scr: serum creatinine; BUN, blood urea nitrogen; SUC: serum uric acid; TC: total cholesterol; TG: triglyceride; FBG: fasting blood glucose.
nitrogen level was slightly higher than that of the normal group.We speculated that the main reason for the increase in SUC in pregnant women in the plateau area was hypoxia (Ohuchi et al. 2015).Hypoxia can stimulate the conversion of xanthine dehydrogenase into xanthine oxidase (XO) (Harrison 2002), and upregulate XO; the last two steps of purine metabolism are the conversion of hypoxanthine to xanthine and the conversion of xanthine to uric acid, which are mainly dependent on XO.Recent studies have shown that environmental factors can affect foetal growth and development by altering the expression of microRNAs in placenta (Chiofalo et al. 2017).Whether the expression of microRNAs in pregnant women in plateau areas has its characteristics still needs further research.
Altitude hypoxia is known to be a major factor affecting the birth weight of newborns; however, it is difficult to directly control altitude hypoxia.In this study, we used multiple linear regression analysis to adjust for confounders and found a negative correlation between prenatal maternal blood uric acid and neonatal birth weight.This finding is consistent with studies conducted by Anna et al. (D'Anna et al. 2000) and Ayankunle et al. (Ayankunle et al. 2021).However, when we converted SUC into categorical variables for further analysis, no correlation was found in the fully adjusted model, suggesting that the above results are not stable in the whole study population, and there may be special groups in the study subjects.The univariate analysis results showed that maternal weight, standing height, body mass index (BMI), BUN, SUC, FBG, proteinuria, primipara, systolic pressure (categorical variable), and foetal sex may be related to birth weight.Multifactor regression analysis without adjusting for confounders, SUC and foetal sex were correlated with the birth weight of newborns.BMI: body mass index; HB: haemoglobin; Scr: serum creatinine; BUN, blood urea nitrogen; SUC: serum uric acid; TC: total cholesterol; TG: triglyceride; FBG: fasting blood glucose.Multivariate linear regression analysis showed that SUC was negatively correlated with birth weight.However, further analysis of SUC as a categorical variable showed that there was no correlation between SUC and birth weight in the fully adjusted model.Crude model: We did not adjust any covariants.Minimally adjusted model: We adjusted blood urea nitrogen.Fully adjusted model: We adjusted standing height, BUN, proteinuria (categorical variable), HCT, foetal gender, gravidity, fasting blood glucose, systolic pressure (categorical variable).SUC: serum uric acid; HCT: haematocrit; BUN: serum urea nitrogen.
Further stratified analysis showed that SUC levels remained negatively associated with birth weight in positive proteinuria, primiparas, or female neonates, and previous studies have shown that uric acid levels in pregnant women with preeclampsia during pregnancy are positively correlated with the incidence of LBW in newborns (Ryu et al. 2019), which also supports the results of this study.Roberts et al. (Roberts et al. 2005) suggested that uric acid is as important as proteinuria in determining pregnancy and foetal outcomes.However, our regression analysis showed no correlation between proteinuria and birth weight, and the interaction analysis showed no interaction between proteinuria, primipara, or female newborns and SUCs on birth weight.To determine whether SUC has a major impact on birth weight, we conducted mediation analysis, which showed that proteinuria, primiparas, or female newborns do not play a significant mediating role in the effect of SUC on birth weight, indicating that SUC mainly affects birth weight, which may be related to the chronic anoxic environment in the plateau.Previous studies have shown that hyperuricaemia is significantly correlated with proteinuria (Noone and Marks 2013), the treatment of asymptomatic hyperuricaemia can delay the progression of chronic kidney disease (Bonino et al. 2020), and SUC is correlated with proteinuria in preeclampsia pregnant women (Ryu et al. 2019), all of which indicate that SUC may be one of the factors contributing to proteinuria.
In primipara or women with female foetuses, we used the GAM and a fitted smoothing curve to analyse SUC and birth weight.The results show a negative linear correlation between the two factors.An inverted U-shaped relationship was found between SUC and male birth weight in pregnant women with proteinuria.It was previously reported that primiparas are a high-risk factor for the occurrence of LBW infants (Abubakari et al. 2015), which may be related to maternal immune dysfunction in the first pregnancy (Luo et al. 2007), and may also be one of the mechanisms behind the negative correlation between blood uric acid in primiparas and foetal birth weight.Previous studies on the birth weight of newborns also found differences between the sexes of newborns (Schild et al. 2004).In a study conducted in northwestern Ethiopia, the birth weight of female babies was found to be lower than that of males.Our results were similar to those of other studies in that the percentage of LBW infants was higher for female infants than that of male infants (Zeleke et al. 2012).The exact mechanism that affects the difference in birth weight is unclear, but it could be due to androgen activity or the Y chromosome, which carries genetic material for foetal growth (Das Gupta et al. 2019).Therefore, the intrauterine growth and birth weight of male infants may be higher than that of female infants, which also explains why females are more susceptible to maternal SUC.Low concentration of SUC has antioxidant effect, when the SUC is less than 338 lmol/L, synergistic androgens or Y chromosomes show positive correlation with male foetal birth weight.Low birth weight at this point may be more due to other factors.But when the SUC is larger than 338 lmol/L, the inhibition effect of SUC on weight gain is stronger than that of androgens or Y chromosomes.This may explain the inverted U-shaped relationship between SUC and male birth weight in pregnant women with proteinuria.
The mechanism by which maternal SUC affects foetal development is not fully understood.Studies have shown that uric acid can induce concentration-dependent attenuation of trophoblast invasion and integration into the monolayer of uterine microvascular endothelial cells (Bainbridge et al. 2009a); therefore, uric acid can induce placental dysfunction.An American study showed that uric acid inhibits the transport of amino acids in the placental system, leading to intrauterine growth restriction (Bainbridge et al. 2009b).As XO mediates uric acid production and produces reactive oxygen species (Landmesser et al. 2002), SUC may partly reflect the increase in oxidative stress.Other studies have also shown that uric acid can stimulate the basal expression of cyclooxygenase 2 both in vitro and in vivo, suggesting that inflammatory mechanisms may participate in the process of foetal growth restriction (Kanellis et al. 2003, Ohtsubo et al. 2004).This is the first study to investigate the correlation between maternal serum uric acid and foetal birth weight in the highlands of China.We found a negative linear relationship between maternal serum uric acid and neonatal birth weight in primiparas, urinary protein, and female foetuses.
However, this study has some limitations.First, this is a retrospective cross-sectional study and, thus, cannot conclude whether there is a causal relationship between SUC and birth weight.Furthermore, no laboratory data were available for the first and second trimesters.Therefore, we were unable to assess the early-and mid-stage effects of high uric acid levels in the mother on foetal birth weight.Finally, the study population included only pregnant women in the Naqu region of Tibet and may not be extrapolated to other regions.The smoothing curve fitting of SUC and birthweight in pregnant women with proteinuria positive was stratified by neonatal gender.We adjusted standing height, blood urea nitrogen, HCT, BUN, fasting blood glucose, systolic pressure(dichotomous variable), proteinuria(dichotomous variable), foetal gender and gravidity(dichotomous variable), but independent variables do not adjust themselves.SUC: serum uric acid; HCT: haematocrit; BUN: serum urea nitrogen.

Conclusion
In conclusion, maternal SUC is negatively correlated with foetal birth weight in the population of proteinuria-positive pregnant women, primiparas, or women with female foetuses in the Naqu region of Tibet.Thus, attention should be paid to the prevention of hyperuricaemia in clinically positive proteinuria pregnancies in Naqu regions.Further prospective cohort studies in the Tibetan Plateau region are necessary to determine whether maternal hyperuricaemia leads to decreased birth weight.

Figure 1 .
Figure 1.Flowchart of study population in the study.Data were collected from pregnant women aged 16-47 who delivered a single live birth at Naqu People's Hospital from January to June 2019.

Figure 2 .
Figure2.Correlations between the SUC and birth weight by stratified analysis.The results showed that there was a significant negative correlation between SUC and birth weight in proteinuria-positive, female foetal, and primiparous mothers.Note: We adjusted standing height, HCT, BUN, fasting blood glucose and systolic pressure(categorical variable).SUC: serum uric acid; HCT: haematocrit; BUN: serum urea nitrogen.

Figure 3 .
Figure 3. Smoothing curve fitting of SUCs and Birth weight in different population.(a) Smoothing curve fitting of SUC and birthweight in first pregnancy; (b) Smoothing curve fitting of SUC and birthweight in pregnant female foetuses; (c) Smoothing curve fitting of SUC and birthweight in proteinuria positive pregnant women; (d)The smoothing curve fitting of SUC and birthweight in pregnant women with proteinuria positive was stratified by neonatal gender.We adjusted standing height, blood urea nitrogen, HCT, BUN, fasting blood glucose, systolic pressure(dichotomous variable), proteinuria(dichotomous variable), foetal gender and gravidity(dichotomous variable), but independent variables do not adjust themselves.SUC: serum uric acid; HCT: haematocrit; BUN: serum urea nitrogen.

Table 1 .
Baseline characteristics of the total cohort and stratified according to birth weight.

Table 2 .
Correlation analysis between different indexes and birth weight.

Table 3 .
Correlation analysis between SUC and birth weight by multivariate analysis.