Bias | Studies | Authors’ judgment | Support for judgment |
---|---|---|---|
Incomplete outcome data (attrition bias) | Gilliland et al., 2003[45]; Gold et al., 2003[46]; Mannino et al., 2006[48] | Unclear risk | Number of participants lost to follow-up not reported |
Mamun et al., 2007[47] | High risk | 46.5% of original cohort lost to follow-up by age 5Â years; 59.6% of original cohort has incomplete data at either age 5Â years or 14Â years and was excluded from analysis | |
Zhang et al., 2010[49] | Low risk | 89.6% of original cohort sampled at 6Â year visit; 82.4% of original cohort sampled at 8Â year visit | |
Ho et al., 2011[50] | Unclear risk | 82% of selected original cohort with complete data at 12Â month follow-up | |
Ascertainment bias | Gilliland et al., 2003[45]; Gold et al., 2003[46]; Mamun et al., 2007[47]; Mannino et al., 2006[48]; Zhang et al., 2010[49]; Ho et al., 2011[50] | Unclear risk | Unclear if children who were overweight or obese were more likely to be diagnosed with asthma |
Differential misclassification of exposure bias | Mannino et al., 2006[48] | High risk | Only 69% of height and 61% of the weights were measured |