Table 3.

Criteria for methodological quality.

Criteria for methodological quality	All articles n = 25 [n (%)]	Good n = 1 [n (%)]	Fair n = 19 [n (%)]	Poor n = 5 [n (%)]
Study population
Description of population	20 (91)	1 (100)	17 (89)	4 (80)
Participation of eligible participants ≥50%	18 (82)	1 (100)	16 (84)	3 (60)
Inclusion criteria precise	21 (96)	1 (100)	19 (100)	4 (80)
Loss to follow-up ≤20%	7 (32)	0 (0)	7 (37)	1 (20)
Measured exposures
Exposures measured before outcome	22 (100)	1 (100)	19 (100)	5 (100)
Levels of exposure examined	13 (59)	1 (100)	12 (63)	3 (60)
Exposure measures valid	10 (45)	1 (100)	9 (47)	0 (0)
Exposures assessed more than once	10 (45)	1 (100)	8 (42)	1 (20)
Measured outcome
Sufficient timeframe to detect outcome	22 (100)	1 (100)	19 (100)	5 (100)
Outcome measures valid	8 (36)	1 (100)	7 (37)	1 (20)
Study characteristics
Research question clearly stated	19 (86)	1 (100)	18 (95)	3 (60)
Sample size justification	3 (14)	1 (100)	2 (11)	0 (0)
Outcome assessors blinded	1 (5)	0 (0)	1 (5)	0 (0)
Confounding variables adjusted	14 (64)	1 (100)	14 (74)	1 (20)