Table 3.
Risk of Bias and Confounding of Included Studies
| First author (year) | Study design | Biasa | Confoundingb |
|---|---|---|---|
| Antoniou (2015)32 | Longitudinal | Medium | Medium |
| Brown (2015)33 | Cross sectional | Medium | Medium |
| Cao (2012)34 | Cross sectional | Low | Medium |
| Carruth (1998)25 | Longitudinal | Medium | Low |
| Carruth (2000)17 | Longitudinal | Medium | Medium |
| Carruth (2004)18 | Cross sectional | Low | Medium |
| Chatoor (2000)22 | Cross sectional | High | High |
| Dubois (2007)11 | Longitudinal | Low | Medium |
| Ekstein (2010)12 | Cross sectional | Medium | High |
| Equit (2013)35 | Cross sectional | Low | High |
| Faith (2013)26 | Cross sectional | Low | Medium |
| Ferreira (2008)27 | Cross sectional | Medium | Medium |
| Finistrella (2012)13 | Cross sectional | Low | Medium |
| Galloway (2005)28 | Cross sectional | Low | Medium |
| Gregory (2010)36 | Longitudinal | Low | Low |
| Hittner (2011)37 | Longitudinal | Low | Low |
| Jacobi (2003)19 | Longitudinal | Low | Medium |
| Jansen (2012)38 | Cross sectional | Low | Medium |
| Jiang (2014)39 | Cross sectional | Low | Medium |
| Johnson (2015)40 | Cross sectional | Low | Low |
| Laureati (2015)41 | Cross sectional | Low | High |
| Lewinsohn (2005)42 | Cross sectional | Low | High |
| Li (2008)53 | Cross sectional | Low | Medium |
| Loh (2013)43 | Cross sectional | Low | Medium |
| Mackenbach (2012)44 | Cross sectional | Low | Medium |
| Mascola (2010)20 | Longitudinal | Medium | Medium |
| Moroshko (2013)45 | Cross sectional | Medium | Low |
| Morrison (2013)29 | Cross sectional | Low | High |
| Mosli (2015)46 | Cross sectional | Low | Medium |
| Perry (2015)47 | Cross sectional | Medium | Medium |
| Rodenburg (2012)48 | Longitudinal | Low | Low |
| Sleddens (2008)49 | Cross sectional | Medium | High |
| Spence (2011)50 | Cross sectional | Medium | Medium |
| Svensson (2011)51 | Cross sectional | High | Medium |
| Tan (2012)30 | Cross sectional | Low | Medium |
| Tharner (2014)31 | Cross sectional | High | Medium |
| Webber (2009)52 | Cross sectional | Low | Medium |
| Werthmann (2015)16 | Cross sectional | Low | High |
| Wright (2007)21 | Cross sectional | Medium | Medium |
| Xue, Lee (2015)23 | Cross sectional | Low | Medium |
| Xue, Zhao (2015)24 | Cross sectional | Low | Medium |
As determined by the Agency for Healthcare Research and Quality RTI Item Bank,15 an item was low risk of no affirmative answers, medium risk if one affirmative answer, and high risk if more than one affirmative answer.
a
Bias = Q 1,3,7,8,9,11.
b
Confounding = Q 6,12,13.