Demographic characteristics of the participants
After exclusion, a total of 98, 386 pregnant women were included. Among the 98386 included women, 780 had a history of HA management of IUAs before this pregnancy, and 97, 606 had no history of IUAs. As shown inTable 1 , the two groups differed in maternal age and parity, prior history of abortion, mode of conception and pre-pregnancy BMI. However, these differences were eliminated by the PSM procedure. After matching, 780 exposed pregnancies and 3010 matched control pregnancies were included in the final analysis.
Association between history of HA-treated IUAs and risk of adverse obstetric outcomes
As shown in Table 2 , the incidences of GDM (20.3% vs. 18.3%, P =0.19), GH (3.7% vs. 4.1%, P=0.61) and ICP (1.2% vs. 1.0%, P < 0.75) were generally similar between women with a history of HA-treated IUAs and women with no history of IUAs. However, the incidence of PE (5.9% vs. 3.5%, P < 0.01) in women with a history of HA-treated IUAs was significantly higher than that in women with no history of IUAs. Women with a history of HA-treated IUAs were at higher risk of PE, and the adjusted RR was 1.69 (RR=1.69, 95% CI, 1.23, 2.33).
For placental abnormalities, the incidences of placenta previa (5.8% vs. 1.4%, P < 0.01) and placenta accreta spectrum (25.0% vs. 5.3%, P < 0.01) were significantly higher in women with a history of HA-treated IUAs than in women with no history of IUAs. A history of HA-treated IUAs was associated with an increased risk of placenta previa (RR=4.23, 95% CI, 2.85, 6.30) and placenta accreta spectrum (RR=4.72, 95% CI, 3.90, 5.73). As placental abnormality is a known risk factor for PPH, we further examined the associations of a history of HA-treated IUAs with PPH. As shown in Table 2 , women with a history of HA-treated IUAs were at higher risk of PPH (RR=2.86, 95% CI, 1.94, 4.23), and those women were also more likely to receive hemostatic therapies (RR=2.17, 95% CI, 1.75, 2.69).
Regarding adverse birth outcomes, a history of HA-treated IUAs was significantly associated with an increased risk of PTB. Women with a history of HA-treated IUAs were more likely to receive cervical cerclage (RR=5.63, 95% CI, 3.95, 8.02) and had a higher risk of PPROM and iatrogenic PTB. The adjusted RRs were 3.02 (RR=3.02, 95% CI, 1.97, 4.64) for PPROM and 2.86 (RR=2.86, 95% CI, 2.14, 3.81) for iatrogenic PTB. However, no significantly association was found between history of HA-treated IUAs and SGA (RR=0.89, 95% CI, 0.53, 1.49).