Comparative Study On The Weight Gain For Gestational Age Z Score And Traditional Gestational Weight Gain Messurements

BackgroundA large number of studies have shown that insufficient or excessive gestational weight gain(GWG)is associated with short-or long-term health of the mother and the fetus.Establishment of an appropriate range of weight gain is particularly important for the maintenance of maternal and child health.However,the international community has not yet reached a conclusion for the appropriate range of GWG.The standard for the average increase of 12.5 kg is still used in China,and the upper and lower limits have not been set according to BMI classification too.Taking into account the difference in weight gain during pregnancy between regions and nationalities,it is of great significance to formulate a suitable range for pregnant women in specific regions of China.To formulate a suitable recommendation range,the first task is to clarify the relationship between GWG and adverse pregnancy outcomes.Traditional GWG measurements have a significant correlation with gestational age,and will have serious bias when analyzing the relationship between GWG and gestational age related adverse pregnancy outcomes.Hutcheon et al.refer to the methods in the evaluation of children’s growth and development status,and introduced the z score into GWG measurements.By calculating the weight-gain-for-gestational-age z scores,researchers can express total GWG as an gestational-age-standardized z score,allowing epidemiologic analysis of the association between pregnancy weight gain and adverse pregnancy outcomes independent of gestational age.However,the current use of the z-score is limited to European and American countries.Whether it is generalizable to the Asian countries,especially in developing countries such as China,where pregnant women have lower prepregnancy BMI? Second,although Hutcheon had detected the relationship between GWG and adverse pregnancy outcomes measured by z score,she only limited to pregnant women with normalweight and overweight and adverse outcomes of preterm birth and small/large than gestational age infant.Third,since z-score is completely independent of gestational age,whether the appropriate GWG ranges based on z-score are more effective than traditional measurements in predicting gestational age-related adverse pregnancy outcomes? The above issues,to date,are still unclear.Objective By using the cohort of pregnant women from ethnic minority areas in Guangxi,the goal of the study was(1)to investigate the related factors of weight gain for gestational age z;(2)to determine the association of z score and gestational age-related adverse pregnancy outcomes and analyze the size and direction of bias caused by traditional GWG measurements(total GWG and rate of GWG);(3)to evaluate advantages of z score in formulating appropriate ranges of GWG comparing with traditional measurements.All together were aiming to evaluate the generalisability and application value of the z-score in China and to provide evidence for the development of GWG recommendation range for pregnant women in Guangxi and even our country.Methods(1)Related factors of z score and traditional GWG measurements were screened by univariate and multivariate linear regression methods,respectively,by using data from the prospective birth cohort study in Guangxi Zhuang population.Then pregnant women were classified into inadequate,adequate and excessive GWG groups based on IOM’s latest guideline and the z-score quartile values.And risk factors of inadequate and excessive GWG were analyzed by using multivariate logistic regression model.(2)Pregnant women cohort were included from registration data of the maternal and child information management system of Pingguo County in Guangxi in the past three years.Restricted cubic spline in logistic regression analysis was conducted to reevaluate the relationship between GWG and gestational age related adverse pregnancy outcomes including preterm birth,low birth weight,and macrosomia,by using the z score.Results of three different z score systems were compared to assess the stability of z score in evaluation of relationship between GWG and adverse pregnancy outcomes.The direction and size of bias caused by traditional measurements was analyzed in comparison to z score.(3)Pregnancy women with normalweight in early pregnancy in Part 2 were included as the study objects.The ranges of recommended values of z scores and traditional GWG measurements were produced using a percentile method and a noninferiority margin approach respectively.The consistency was compared and analyzed by kappa-values.The logistics regression model was used to analyze the relationship between those recommended ranges and risk of adverse pregnancy outcomes.The accuracy in predicting adverse pregnancy outcomes was compared by the area under the ROC curve,and the advantages of z scores in formulating the recommended ranges of GWG were analyzed.Results(1)Total GWG was positively while the rate of GWG and z score were not related with gestational age at delivery.It showed age,BMI,exposure to secondhand smoke,and calcium/iron supplementing during pregnancy were negatively while folic acid supplementing and pregnancy tests were positively correlated with GWG based on the three measurements.The strongest association factors were BMI and calcium/iron supplementing during pregnancy.(2)Risk factors analysis based on the z-score classification show that the factors increasing risk were prepregnancy overweight and vitamin supplementing during pregnancy while factors decreasing risk were <20 years old and folic acid supplementing for inadequate GWG.And factors increasing risk were prenatal visit times >8 while factors decreasing risk were senior pregnancies(≥35 years),multiparas and second-hand smoke exposure for excessive GWG.Those correlated factors were similar to that based on the IOM guideline classification.(3)Risk curves were varied depending on BMI categories for relationship of GWG and preterm birth,with a weak U-shaped association in normalweight women;When compared to z scores,total GWG may overestimate risk of preterm birth at both low and high weight gain,while rate of GWG may underestimate risk of preterm birth when at moderate and high weight gain in underweight women.Risk curves showed a weak L-shaped association of GWG and low birth weight in underweight and normalweight women;When compared to z scores,total GWG may underestimate risk of low birth weight at high weight gain in underweight and normalweight women,while rate of GWG may underestimate risk of low birth weight at low weight gain in underweight women.Risk curves showed a reversed L-shaped(climbing quickly after a smooth stage)association of GWG and macrosomia.Risk curves varied depending on BMI categories between GWG and CAO,showing a typical U-shaped association in normalweight women,a weak U-shaped association in underweight women,and a reversed L-shaped association in overweight women;when compared to the z score,total GWG at high weight gain and rate of GWG at low weight gain may overestimate risk of CAO in normalweight women;Total GWG at high weight gain may underestimate and at low weight gain overestimate risk of CAO,while rate of GWG at low weight gain may underestimate the risk of CAO in underweight women;Both of total GWG and rate of GWG may underestimate risk of CAO at high weight gain in overweight women.(4)The recommendation ranges of GWG established by the percentile method showed that both the z score and the total GWG covered the IOM recommendation range and the lower limits varied greatly,while the rate of GWG were greater than the IOM recommendation range.By using those ranges,women were divided into inadequate,adequate,and excessive GWG.The classifications of the five GWG measurements and the IOM showed a high agreement(Kappa-value > 0.7),and there was a high degree of consistency between z scores(Kappa-value > 0.95).When compared to women with adequate GWG,those with inadequate GWG of total GWG and excessive GWG of z-score had a significant increased risk of preterm birth,those with inadequate GWG of all measurement but rate of GWG had a significantly increased the risk of low birth weight,and those with excessive GWG of all measurements had a significantly increased the risk of macrosomia.For CAO,excessive GWG of all measurements significantly increased risk of incidence,while only inadequate GWG of total GWG significantly increased the risk of CAO.The results of ROC curve analysis showed that there was no difference in the ability to predict adverse pregnancy outcomes for various ranges developed from different measurements.(5)The recommendation ranges of GWG established by a noninferiority margin approach(△= 20%)showed that all the lower limits but the rate of GWG were lower than IOM,and that of z score was lower than traditional measurements.The upper limits of traditional measurement were higher than IOM,and that of the z score was slightly lower than IOM.By using those ranges,women were divided into inadequate,adequate,and excessive GWG.The classifications of traditional measurements and IOM were highly consistent,and classification of z values and IOM classification were less consistent.When compared to women with adequate GWG,those with inadequate GWG of all measurements but rate of GWG had a significantly increased risk of living low birth weight infants,those with excessive GWG of all measurements had a significantly increased risk of macrosomia,only those with inadequate GWG of total GWG had a significantly increased risk of preterm birth,and only those with inadequate GWG of total GWG had a significantly increased risk of CAO.The ROC curve analysis showed that the accuracy of the recommendation ranges developed from total GWG was significantly higher than other measurements in predicting occurrence of low birth weight and macrosomia.However,there was no difference in prediction ability when predicting preterm birth and CAO.(6)The ROC curve analysis indicated that there was a significant small difference only in the total GWG for the accuracy of recommendation ranges established by the percentile method and the noninferiority margin approach in predicting adverse concept outcomes.However,there was no difficulty difference in prediction of other adverse Pregnancy outcomes and CAO,which was independent of GWG measurements.Conclusion(1)Factors associated with GWG in Guangxi minority area are age,BMI,exposure to secondhand smoke during pregnancy,folic acid supplementing during pregnancy,calcium/iron supplementing during pregnancy,physical exercise during pregnancy,and the number of prenatal visits.As a new GWG measurement method completely independent of gestational age,z score can be used to screen the related factors of GWG effectively.(2)There is a nonlinear association between GWG and the gestational age-related adverse pregnancy outcomes.The traditional GWG measurements may be biased when assessing the relationship between GWG and gestational age-related adverse pregnancy outcomes,which were depended on BMI classification,measurement methods,and related outcomes.The z scores have a good generalisability in evaluating the relationship between GWG and the adverse pregnancy outcomes.(3)There is no difference in the ability to predict adverse pregnancy outcomes based on the appropriate value ranges developed from the z-scores and traditional measurements,however,the ranges established by the noninferiority margin approach for z-scores are much lower.