Table 4.

Risk of bias ranking for the identified intervention studies.

Study author, year	Sample	Design	Outcome	Total score	Bias ranking
WEIGHT LOSS
Stolley et al. (89)	>50	Randomized	Objective	99	Low bias
Valle et al. (93)	< 50	Randomized	Objective	88	Moderate bias
Sheppard et al. (81)	< 30	Randomized	Objective	77	Moderate bias
Chung et al. (90)	< 30	Single group	Self-reported	33	High bias
Delgado-Cruzata et al. (94)	< 30	Randomized	Objective	77	Moderate bias
Greenlee et al. (92)	< 50	Randomized	Objective	88	Moderate bias
Djuric et al. (91)	< 30	Randomized	Objective	88	Moderate bias
Stolley et al. (52)	< 30	Single group	Objective	55	High bias
DIETARY INTAKE
Paxton et al. (99)	>50	Randomized	Self-reported	77	Moderate bias
Griffith et al. (98)	< 30	Single group	Objective	55	High bias
Paxton et al. (49)	>50	Randomized	Objective	99	Low bias
PHYSICAL ACTIVITY
Piacentine et al. (107)	< 30	Single group	Objective	55	High bias
Paxton et al. (99)	>50	Randomized	Self-reported	77	Moderate bias
Spector et al. (104)	< 30	Single group	Objective	55	High bias
Nock et al. (105)	< 30	Single group	Objective	55	High bias
Moadel et al. (108)	>50	Randomized	Self-reported	77	Moderate bias
Wilson et al. (106)	< 30	Single group	Objective	55	High bias

Sample sizes of <30, 30 to 49, and 50 or higher were characterized as low, moderate, and high quality, respectively. Randomized trials were characterized as high quality, while other designs were characterized as low quality. Objectively assessed anthropometric, dietary, and physical activity outcomes were characterized as high quality, whereas self-reported outcomes were characterized as low quality. Each rating was transformed to a numerical, whereby scores of 11, 22, and 33 characterized as low, moderate, and high quality. Scores were summed across categories. Total scores of < 60, 60 to < 99, and 99 were characterized as high, moderate, and low bias studies.