Table 1

 Quality assessment of included studies using Newcastle-Ottawa scale

Representativeness of exposed cohortSelection of non-exposed cohortAscertainment of exposureAssessed outcome of interest was not present at startTotalRisk of biasComparability of cohorts on basis of design or analysisTotalRisk of biasAssessment of outcomeFollow-up adequate for outcomes to occurAdequate follow-up of cohortsTotalRisk of bias
White 198711***3Medium0High*1High
Pereira 19961***3Medium0High*1High
Fenton 20038***3Medium**2Low*1High
Chilopora 200710***3Medium0High**2Medium
Hounton 20093***3Medium*1Medium*1High
McCord 200912***3Medium0High*1High