Risk of recurrent adverse neonatal outcomes
While the rate of LGA or SGA was not different in index versus subsequent pregnancies as a group, the risk of these complications in their subsequent pregnancy was greatly increased in those women who had LGA or SGA in their index pregnancy compared to those who did not (Figure 1). For example, while the rate of LGA was similar in index and subsequent pregnancies (16.7% vs 15.8%, p=NS), the rate of LGA in the subsequent pregnancy was 45.1% in those women who had LGA in their index pregnancy, with a RR of 4.55 compared to women who did not. Likewise, while the rate of SGA was similar in index and subsequent pregnancies (7.8% vs 8.7%, p=NS), the rate of SGA in the subsequent pregnancy was 33.3% in women who had SGA in their index pregnancy, with a RR of 5.01 compared to women who did not (Figure 1). This greatly increased risk was also the case for the composite ANO (death/dystocia/LGA/SGA) (Table 2).
Conversely, having an SGA baby in the index GDM pregnancy was associated with a below average rate of LGA (6.1%, n=2/33), and having prior LGA was associated with an SGA rate of only 1.4% (n=1/71) in the subsequent GDM pregnancy. Women with no LGA or SGA history had a 7.8% rate of SGA and 10.3% rate of LGA in their subsequent GDM pregnancy (Table 2).
Factors associated with ANO in the subsequent GDM pregnancy
On univariate analysis, women with LGA in their subsequent GDM pregnancy were slightly younger with higher parity compared to those without LGA. They had a 17.5 kg greater median booking-in weight (84.5 (69.0-105.0) vs 67.0 (58.0-82.0) kg, p<0.001), higher booking-in BMI (31.3 (26.9- 37.4)) vs 26.5 (23.1-32.0) kg/m2, p<0.001) and a 2.5 kg greater interpregnancy weight gain than women without LGA (4.7±8.4 vs 2.2 ±5.4 kg, p=0.002), despite a similar interpregnancy interval. They had a higher fasting and two-hour glucose on the diagnostic GTT. 47.8% had LGA in their index pregnancy, whereas only 10.9% of women without LGA in the subsequent pregnancy had LGA in the index pregnancy (p<0.001) (Table 3).
On univariate analysis, women with SGA in their subsequent GDM pregnancy had a longer interpregnancy interval (3.9±2.1 vs 2.8±1.5 years, p<0.001) compared to women without SGA. 70.3% had SGA in their index pregnancy, versus 5.7% of women without SGA in their subsequent pregnancy (p<0.001). There was a trend to lower booking-in weight (65.0 (56.0-77.5) vs 71.0 (59.3-88.0) kg, p=0.07) but no differences in booking-in BMI or interpregnancy weight change (Table 3).
On univariate analysis, women who had the composite ANO in the subsequent GDM pregnancy had a higher parity compared to those who did not. They had a 19.5 kg greater booking-in weight, higher booking-in BMI (29.7 (24.9-35.4) vs 26.6 (23.1-32.1) kg/m-2, p=0.001) and a 1.7 kg greater weight gain (p=0.01) between pregnancies. These women had a higher fasting and 2-hour glucose on the diagnostic GTT. 43.5% had the composite ANO in their index pregnancy, compared to 21.2% of those without any ANO in their subsequent pregnancy (p<0.001) (Table 3).
Based on results of univariate analysis, potential predictors of LGA in the subsequent pregnancy were included in a binomial regression model (prior LGA, BMI at booking-in, interpregnancy weight gain, and fasting glucose at diagnostic OGTT). After backward stepwise removal, LGA in the index pregnancy remained the strongest predictor of subsequent LGA, with a RR of 3.13 (95%CI:2.20, 4.47, p<0.001) compared to women without prior LGA. Booking-in BMI showed a modest association with LGA outcome- RR 1.04 (95%CI:1.02, 1.07, p<0.001).
For the outcome of SGA in the subsequent pregnancy, prior SGA, interpregnancy interval and booking-in weight were included in the model. After adjustment, the RR of SGA in women with SGA in the index pregnancy was 4.71 (95%CI:2.66, 8.36, p<0.001). For every one-year increase in the interpregnancy interval, the RR of SGA was 1.51 (95%CI:1.19, 1.91, p<0.001)
For the composite ANO, the regression model included prior composite ANO, booking-in BMI, interpregnancy weight gain, fasting glucose at diagnostic OGTT and Europid/non-Europid ethnicity. After adjustment, the RR of ANO was 2.01 (95%CI:1.46, 2.78, p<0.001) in women with a prior history of ANO in the index pregnancy.