Table 2.
Summary OR for the relation between arsenic and the risk of adverse pregnancy/infant mortality and stratified/sensitivity analysis according to the study characteristics.
| Analysis | Spontaneous abortion | Stillbirth | Neonatal mortality | Infant mortality | ||||
|---|---|---|---|---|---|---|---|---|
| Random-effects model OR (95% CI) | Heterogeneity statistics Q (n)‑statistics, I2‑index (%), p‑value | Random-effects model OR (95% CI) | Heterogeneity statistics Q (n)‑statistics, I2‑index (%), p‑value | Random-effects model OR (95% CI) | Heterogeneity statistics Q (n)‑statistics, I2‑index (%), p‑value | Random-effects model OR (95% CI) | Heterogeneity statistics Q (n)‑statistics, I2‑index (%), p‑value | |
| Summary OR | 2.02 (1.40, 2.91) | 11.2 (6), 55.3, 0.048 | 1.84 (1.38, 2.45) | 38.40 (9), 79.2, 0.000 | 1.51 (1.28, 1.78) | 5.34 (5), 25.1, 0.254 | 1.35 (1.12, 1.62) | 8.31 (7), 30.4, 0.216 |
| Stratified analysis | ||||||||
| Assessment of arsenic exposure | ||||||||
| Individual data/biomarker | 2.20 (1.04, 3.46) | 7.96 (3), 74.9, 0.019 | 1.96 (1.17, 3.29) | 19.91 (5), 79.9, 0.001 | 1.30 (1.00, 1.67) | 4.78 (2), 16.4, 0.274 | 1.74 (0.92, 3.28) | 6.02 (3), 66.8, 0.049 |
| Group data | 1.51 (0.79, 2.87) | 1.59 (3), 0.0, 0.951 | 1.79 (1.29, 2.48) | 5.46 (4), 45.1, 0.141 | 1.59 (1.43, 1.77) | 1.75 (3), 0.0, 0.416 | 1.32 (1.08, 1.60) | 2.44 (4), 0.0, 0.486 |
| Sensitivity analysis | ||||||||
| Prospective cohort studiesa | 1.45 (0.99, 2.12) | 0.66 (2), 0.0, 0.951 | 1.13 (0.98, 1.30) | 0.67 (2), 0.0, 0.412 | 1.21 (0.98, 1.50)a | 2.12 (0.53, 8.42) | 4.84 (2), 79.3, 0.028 | |
| Studies adjusting for potential confounders | 1.72 (1.25, 2.37) | 4.43 (5), 9.7, 0.351 | 1.85 (1.22, 2.82) | 18.06 (7), 66.8, 0.005 | 1.53 (1.11, 2.10) | 4.63 (4), 35.2, 0.201 | 1.65 (1.01, 2.47) | 8.39 (5), 52.3, 0.078 |
| High-quality studies (> 7 on NOS) | 1.45 (0.99,1.12) | 1.65 (3), 0.0, 0.438 | 1.28 (0.98, 1.67) | 4.27 (4), 29.7, 0.234 | 1.49 (0.92, 2.31) | 2.57 (2), 61.1, 0.109 | 1.41 (1.04,1.92) | 7.55 (4), 60.2, 0.056 |
| Impact of missing studies on overall summary OR | ||||||||
| By trim and fill method | 2.02 (1.20, 2.91) | 55.3 (6), 55.3, 0.048 | 1.43 (1.11, 1.85) | 58.06 (13), 79.33, 0.000 | 1.47 (1.27, 1.71) | 7.15 (7), 2.15, 0.307 | 1.22 (0.98, 1.53) | 8.62 (10), 30.0, 0.017 |
| Abbreviations: NOS, Newcastle–Ottawa Scale; Q(n), n: number of studies.aOne prospective study reported on neonatal mortality. | ||||||||