Summary of the main findings
| Exposure | Preterm delivery | Small–for-gestational age | Pre-eclampsia/hypertension |
|---|---|---|---|
| ns, not significant at the 5% level; PE, pre-eclampsia; PIH, gestational hypertension. | |||
| The numbers in brackets represent the number of studies with the stated RR as compared with the number of all studies for the exposure–outcome combination. | |||
| Working hours | RR ⩽1.34 (11/16) | RR ∼1.0 (5/7) | PIH: RR ⩽1.1 (2/2) |
| RR⩾2.0 (0/16) | RR ⩾2.0 (1/7) | PE: no studies | |
| Four biggest studies, RR = 0.6–1.3 | Largest study, RR = 1.6 | Limited evidence | |
| Meta-analysis RR = 1.31 Extensive evidence | Reasonable body of evidence (but only two cohort studies) | Consistent Favours no effect | |
| Reasonably consistent Balance of evidence suggests a small effect and makes a large effect unlikely | Generally consistent Balance of evidence tends to favour a no more than moderate effect More research would be prudent | More research would be prudent | |
| Shift work | RR ⩽1.5 (9/14) | RR ∼1.0 (4/6) | PIH: RR <1.0 (1/1) |
| RR ⩾2.0 (2/14) | RR 1.5 (1/6) | PE: RR = 1.3 (1/1) | |
| Four biggest studies, RR ∼1.0 | Largest study, RR ∼1.0 | Very limited evidence (only two studies) | |
| Meta-RR = 1.2 Extensive evidence | Reasonable body of evidence (but only two cohort studies) | No more than a moderate effect found More research would be prudent | |
| Reasonably consistent Balance of evidence suggests a small effect and makes a large effect unlikely | Generally consistent Balance of evidence tends to favour no effect, or an effect that is no more than a moderate | ||
| Lifting | RR ⩽1.35 (11/12) | RR ⩽1.2 (5/5) | PIH: RR = 1.1 (1/1) |
| RR ⩾1.5 (0/12) | Two large studies, three cohort | PE: RR = 0.7–1.7 (two studies) | |
| Extensive evidence | Limited evidence | Very limited evidence (only three studies) | |
| Consistent | Consistent | More research would be prudent | |
| Balance of evidence tends to rule out a more than moderate effect | Balance of evidence tends to favour no effect However, more research would be prudent | ||
| Standing | RR ⩽1.5 (12/20) | RR ⩽1.4, 7/8 | PIH: RR ⩽1.26 (3/3) |
| RR ⩾2.0 (2/20) | RR ⩾2.0 1/8 (ns) | PE: RR = <1.0 (3/3) | |
| Five biggest studies, RR 1.07–1.56 | Three biggest studies, RR = 1.2–1.4 | Limited evidence | |
| Meta-RR = 1.28 | Reasonable amount of evidence | Consistent | |
| Extensive evidence Moderately consistent Balance of evidence suggests a small effect and makes a large effect unlikely | Generally consistent Balance of evidence tends to favour an effect that is no more than moderate | Tends to favours a no more than moderate effect More research would be prudent | |
| Physical activity | RR ⩽1.4 (15/21) | RR ⩽1.4 (6/7); <1 (4/7) | PIH: RR ⩽1.6 (3/4); RR >3 (1/4) |
| RR ⩾1.7 (6/21); ⩾2 (3/21) | RR ⩾2.0 (1/7) | Biggest study, RR = 1.2, cohort = 0.6 | |
| Three biggest studies, RR ∼1.0 | Two biggest studies, RR = 1.3–1.4 | PE: RR <1.0 (2/3); RR >2 (1/3) | |
| Extensive evidence | Reasonable amount of evidence | Biggest study, RR = 0.8, cohort = 0.7 | |
| Less consistent than for other exposures and prematurity Harder to assess—outcome measure not robust or clear, and more prone to bias | Generally consistent Favours a no more than moderate effect More research unhelpful unless better targeted | Limited evidence Mixed findings More research would be prudent | |
| More research unhelpful unless better targeted | |||