Table 4:
Summary of main meta-analysis results and quality of evidence rating conclusions
| Exposure Window | No. of studies | Effect estimate g per 10 μg/m3 (95% CI) | I2 (%) | Quality of evidence rating |
|---|---|---|---|---|
| PM2.5 | ||||
| 1st Trimester | 11 | −6.50 (−15.07, 2.07) | 87% | Very low (downgrades for imprecision and inconsistency) |
| 2nd Trimester | 12 | −5.69 (−10.58, −0.79) | 68% | Low (downgrade for imprecision) |
| 3rd Trimester | 12 | −10.67 (−20.91, −0.43) | 84% | Low (downgrade for inconsistency) |
| Full Pregnancy | 15 | −27.55 (−48.45, −6.65) | 94% | Low (downgrade for inconsistency) |
| PM10 | ||||
| 1st Trimester | 6 | 3.22 (−3.13, 9.58) | 14% | Low (downgrade for imprecision) |
| 2nd Trimester | 6 | −3.37 (−8.22, 1.48) | 0% | Low (downgrade for imprecision) |
| 3rd Trimester | 7 | −6.57 (−10.66, −2.48) | 0% | Moderate (no changes) |
| Full Pregnancy | 8 | −8.65 (−16.83, −0.48) | 84% | Moderate (heterogeneity explained by single study with inverse effect) |
| PM2.5–10 | ||||
| 1st Trimester | 3 | −2.70 (−3.90, −1.49) | 0% | Very low (downgrades for risk of bias and imprecision) |
| 2nd Trimester | 3 | −2.90 (−10.04, 4.23) | 70% | Very low (downgrades for risk of bias, imprecision, inconsistency) |
| 3rd Trimester | 4 | −4.93 (−10.82, 0.96) | 76% | Very low (downgrades for risk of bias, imprecision, inconsistency) |
| Full Pregnancy | 5 | −8.81 (−10.32, −7.31) | 0% | Low (downgrade for risk of bias) |
For PM2.5, we included 18 unique studies rated as “low” or “probably low” risk of bias. For PM10, we included 10 studies rated as “low” or “probably low” risk of bias. For coarse PM (PM2.5–10), there were no studies rated as “low” or “probably low” risk of bias, thus we included 5 studies rated as “high” or “probably high.”