Table 2: Evaluating the role of bias in the included studies.

StudyAichbhaumik [20]Schönberger [21]Bisgaard [22]Gern [23]Bufford [24]Pohlabeln [25]Kerkhof [26]Nafstad [27]Remes [28]

Methods for selecting study participants

Appropriate source population176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face
Inclusion and exclusion criteria176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face

Methods for measuring exposure and outcome variables

Appropriate exposure measurement176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face
Appropriate outcome measurement176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face

Methods to deal with design-specific sources of bias

Recall biasNININANINI×NININI
Interviewer biasNINININININANANANA
Loss to followup176484.face176484.face(1)×176484.face×NA××176484.face
BlindingNANANINININANANANA

Methods to control confounding

Appropriate design176484.face××××××176484.facex
Appropriate analytical methods×176484.face×176484.face176484.face176484.face××176484.face

Statistical methods (excluding confounding)

Appropriate statistics××176484.face176484.face×176484.face×176484.face176484.face

Conflict of interests

Declarations of conflict of interests176484.face×176484.face××××××
Identification of funding sources176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face176484.face

NHMRC evidence grading (ref)

LevelsIII-2III-2III-2III-2III-2III-2III-2III-2III-2

NA: not applicable; NI: not described in article.
176484.face: adequate; ×: poor; ×: not done.