The dose of behavioral interventions to prevent and treat childhood obesity: a systematic review and meta-regression

Table 1 Descriptive Statistics of Included and Excluded Studies

	Analytic Sample (n = 133)		Excluded Sample (n = 125)
	Median or Frequency	IQR or Percent	Median or Frequency	IQR or Percent
Hedges’ g effect size^a	−0.25	[−0.83, 0.33]	−0.40^b	[−1.09, 0.29]
Contact hours	18.0	[10.0, 36.0]	12.5^c	[7.6, 27.0]
Duration (in weeks)	17.3	[12.0, 26.0]	25.5^d	[12.0, 35.0]
Study Year	2010	[2012, 2014]	2012	[2008, 2014]
Intention to treat
Yes	89	67%	84	67%
No	44	33%	41	33%
Participant age group
2–11 only	52	39%	49	39%
12–18 only	22	17%	23	18%
Other combination	58	44%	53	42%
Intervention mode
In-person only	57	43%	42	34%
In-person plus other	76	57%	65	52%
No in-person dose	0	0%	18	14%
Setting^e
School/community only	27	22%	34	29%
Clinic/university only	55	46%	26	22%
All other combinations	38	32%	59	50%
Format^f
Individual only	26	20%	56	45%
Group only or both	106	80%	68	55%
Participants
Child only	24	18%	28	22%
Parent only, or parent and child, or parent and child and other family	109	82%	97	78%
Weight Status
Normal weight and overweight or obese	28	21%	45	36%
Overweight or obese only	105	79%	77	62%
Normal weight only	0	0%	3	2%

^aThe pooled effect size was estimated using the DerSimonian and Laird methodology for random-effects meta-analysis. The variance of the effect sizes is described by the interval from ±2*τ around the random-effects pooled estimate, which is an approximate 95% range of the effect sizes
^bEffect size was missing in 98 of the excluded studies (n = 27)
^cContact hours was missing in 63 of the excluded studies (n = 62)
^dDuration was missing in 25 of the excluded studies (n = 100)
^eSetting was missing in 13 of the analyzed studies (n = 120) and 6 of the excluded studies (n = 119)
^fFormat was missing for 1 of the analyzed studies (n = 132) and 1 of the excluded studies (n = 124)

ISSN: 1479-5868