RESULTS
In this study, we finally included eligible 12 044 reproductive-age women with singleton birth. We excluded 5414 women without successful pregnancy within six months. Women who had a singleton birth were eligible for this study (N = 13 380). Participants with history of chronic disease and stillbirth were excluded, yielding an analytic sample of 12 044 women (Figure 1).
The median of time from baseline examination to pregnancy was 2.2 months (interquartile range (IQR):1.07-3.67). The mean maternal age was 29.16 ± 4.73 years old, and 17.18 % was aged 30 and over. The mean birthweight was 3163.12 ± 417.20 g, and the frequency of macrosomia was 1.9%. In terms of maternal FPG level before pregnancy, 1.8% women were in a hyperglycemic state. Around 17.5% women had a history of adverse pregnancy outcomes. Most of the women were Han (99.3%) and college or higher (77.2%). The percentage of history of passive smoking and parity over three times were 31.1 and 17.1, respectively.
Table 1 showed the baseline characteristics of study subjects according to maternal age. Compared with younger mothers, women with advanced age were more likely to be greater birthweight and higher preconception FPG levels. And, the frequency of macrosomia and hyperglycemia was significantly increased among older mothers. Women with advanced age showed to be higher BMI and college or higher, had more history of adverse pregnancy outcomes and parity.
In adjusted logistic regression model (Table 2), maternal age and BMI was associated with hyperglycemia. Women with advanced age significantly had a higher adjusted OR for hyperglycemia by 1.82 (95% CI, 1.31 to 2.52). Every one unit increased in BMI was associated with a 4% increase per 1 kg/m2 in the risk of hyperglycemia (OR: 1.04; 95% CI: 1.02 to 1.06).
When maternal age was modeled as a continuous variable and fitted in the multivariate logistic regression model using cubic restricted splines, continuously increasing trend was found between maternal age and macrosomia (Figure 2a, P linear< 0.001). It showed that the age of 30 years were likely to be reasonable cutoff values for advanced maternal age regarding the macrosomia. Similarly, an obvious linear association was also observed for preconception FPG levels in macrosomia (Figure 2b, Plinear < 0.001). Based on this age cutoff values, in multivariate logistic regression model, women with advanced age had a significant increase of macrosomia (OR: 1.66; 95% CI: 1.22 to 2.26, Table 3). Similarly, a positive association was found between the risk of macrosomia and preconception FPG level before pregnancy (OR : 1.30; 95% CI: 1.14 to 1.47, Table 3).
Finally, we explored the contribution of preconception FPG levels to the association of advanced maternal age with offspring birthweight by a causal mediation analysis (Table 4). Overall, the total influence of maternal age on offspring birthweight (the β coefficient), which includes independent and mediated effects, was 4.89 (95% CI: 2.92 to 6.81). The presence of preconception FPG level mediated significantly 8% (95%CI: 4% to 16%) of the total influence of advanced maternal age on offspring birthweight. Additionally, the mediated effect was increased among women with history of adverse pregnancy outcomes by mediating 18%.