Associations of COVID-19 lockdown exposure with gestational length
The exposed group had a shorter gestational length than the unexposed group (38.66±1.46 weeks vs 38.74±1.46 weeks). The Level I response (vs no exposure) was significantly associated with a 0.06 (95%CI: 0.05, 0.07) week decrease in gestational length in the total study sample after adjusting for confounders (Table 2). Subgroup analyses showed significant associations between lockdown exposure and decreased gestational length only among pregnant women whose gestational ages were <24 GWs or 28th-31st GWs on the first day of lockdown (1/23/2020). The mean difference varied between -0.11 and -0.04 weeks.
We observed a negative association between cumulative lockdown exposure dose and gestational length (Table 2). Each 100 unit increase in the cumulative exposure dose during the first 22 GWs was associated with a 0.05 (95%CI: 0.04, 0.06) week decrease in gestational length, after adjusting for confounders. In addition, compared to the unexposed group, the Q1, Q2, Q3 and Q4 quantiles of cumulative exposure were associated with 0.09 (0.07, 0.11), 0.13 (0.11, 0.16), 0.14 (0.11, 0.16), and 0.09 (0.07, 0.11) weeks decrease in gestational length, respectively.