Effective behaviour change techniques for physical activity and healthy eating in overweight and obese adults; systematic review and meta-regression analyses

Table 3 Results from meta-regression analysis of 32 long term outcome reports of PA and diet interventions

	Simple meta-regression^a				Multiple meta-regression^b
Study characteristics	b	95% CI	P value	Adj. R² %	b	95% CI	P value
BCT 1.1 Goal setting behaviour^c	0.228	(0.056, 0.400)	0.011	38.5	0.175	0.043, 0.307	0.011
BCT 1.2 Problem solving^c	0.161	(−0.005, 0.327)	0.057	25.1
BCT 1.3 Goal setting outcome^c	0.256	(0.095, 0.416)	0.003	53.2
BCT 1.5 Review behaviour goals^c	−0.319	(−0.678, 0.040)	0.078	19.8
BCT 2.3 Self-monitoring of behaviour^c	0.184	(0.009, 0.360)	0.040	30.8
BCT 2.7 Feedback on outcome of behaviour^c	0.249	(0.085, 0.412)	0.004	43.8	0.145	0.021, 0.269	0.024
BCT 3.1 Social support (unspecified)^c	0.192	(−0.011, 0.394)	0.063	21.6
BCT 8.7 Graded tasks^c	0.203	(0.043, 0.363)	0.014	37.1
BCT 12.5 Adding objects to the environment^c	0.182	(0.010, 0.354)	0.039	12.7
Method based^d
MI/SDT	0.000	reference
ACT/CT/HAES/Mindful/other	−0.303	(−0.500, −0.105)	0.004
Unclear	−0.199	(−0.372, −0.026)	0.026	57.5	−0.170	−0.294, −0.045^g	0.009
Number of BCTs unique to the intervention group^e	0.028	(0.012, 0.044)	0.001	54.3
Total number of BCTs in intervention + control group^f	0.030	(0.014, 0.046)	0.001	61.3
Adj. R² %					100.0

Abbreviations and symbols: BCT Behaviour change technique, PA physical activity, ß estimated meta-regression coefficient, CI confidence interval, Adj. R ² adjusted proportion of between study variance explained by predictors
^aSimple linear meta-regression of pooled estimates of 17 physical activity and 15 diet intervention’s outcome reports. Only predictors with significant or borderline significant effects are reported; ^bMultiple linear meta-regression: results after stepwise backwards elimination from model with all significant predictors included. Only effects with p < 0.05 are retained in the model; ^cThe difference of BCTs between intervention and control group contains this BCT, compared to studies not having this difference. ^dMethod-based interventions comparing MI or SDT based interventions with Acceptance and commitment therapy (ACT), Cognitive therapy (CT), Health-at-every-size (HAES) approach, Mindful based intervention or other methods, versus no method mentioned; ^eThe number of unique BCTs in the intervention groups as compared with the control group; ^fThe total number of BCTs in intervention and control group; ^gThe variable is dichotomized in the multiple meta-regression analysis to MI/SDT versus all others