Table 1.

Covariates considered in meta-regression based on outcome definition.

	Meta-regression by asthma outcome
Covariate	Current	Ever	Incident
Geographic region	Yes	Yes	Yes
Age category	Yes	Yes	Yes
Type of SHS source	Yes	Yes	No
Timing of exposure (postnatal only)	Yes	Yes	Yes
Source of study subjects	Yes	Yes	Yes
Type of study	Yes	Yes	No
Not controlled for age or sex	Yes	No	Yes
Not controlled for race	Yes	Yes	No
Not controlled for child’s own smoking	Yes	Yes	No
Not controlled for child history of atopy	Yes	Yes	Yes
Not controlled for family history of atopy	Yes	Yes	No
Not controlled for family and child atopy	Yes	Yes	No
Asthma not identified by doctor diagnosis	Yes	Yes	Yes

“Yes” indicates that the variable was evaluated in meta-regression; “No” indicates that the variable was not evaluated in meta-regression because no observations differed for this covariate or the variable was redundant to another variable.