Weight change in control group participants in behavioural weight loss interventions: a systematic review and meta-regression study

Table 1 Bivariate associations between explanatory variables and control group weight change (kg)

Trial and participant characteristics:	Study n	Median group weight change (Inter-quartile range)		p-value
Trial Characteristics
Trial design
Randomised controlled trial	78	0.0	(−1.1, 0.8)
Quasi-experimental study design	7	0.5	(−0.2, 0.9)	0.200
Recruitment strategy
Searches of existing registers	36	0.4	(−0.8, 1.25)
Media appeal for volunteers	40	0.0	(−1.1, 0.6)	0.498
Intervention duration*^δ
≤6 months	54	0.1	(−0.8, 0.6)
>6 months	31	0.2	(−1.1, 1.9)	0.637
Total number of assessments*
2 (baseline and follow up)	59	0.2	(−0.5, 0.7)
>2 (at least one interim assessment)	26	−0.6	(−1.4, 1.3)	0.468
Control group treatment
No intervention control group	33	0.3	(−0.4, 0.6)
Wait list control group	14	0.4	(−0.4, 2.1)
Usual care group	38	−0.5	(−1.3, 0.8 )	0.139
Sample Characteristics
Age*
≤50 years	45	0.4	(−1.1, 0.9)
>50 years	36	0.0	(−0.7, 0.8)	0.427
Gender
Mostly female (>60 % female)	48	−0.1	(−1.1, 0.8)
Mixed gender or mostly male	37	0.3	(−0.8, 1.0)	0.346
Target population
Healthy or at risk of chronic disease	59	0.0	(−1.1, 0.6)
Chronic disease	26	0.3	(−0.8, 1.2)	0.247
Baseline mean BMI category
Healthy or overweight (≥ 18 - <25 kg/m ² )	33	0.3	(−0.5, 0.8)
Obese (> 30 kg/m2)	48	0.0	(−1.1, 0.9)	0.441

*Dichotomised based on the median value.
^δ When analysed as a continuous variable, there was no difference in trial duration for studies that had a mean control group weight loss and those that had a mean control group weight gain.

ISSN: 1471-2288