Table 4. Nine-year changes in body weight (BW), body mass index (BMI) and waist circumference (WC): adjusted estimates (mean, Δ, RR).

Outcomes	Analysis on dataset	Intervention group				Control group				Intention-to-treat analysis: intervention versus control
Continuous		N	Mean	95% CI	p	N	Mean	95% CI	p	N	Δ	95% CI	p
BW (kg)	available	125	+3.1	+1.3 to +4.8	0.001	134	+5.1	+3.5 to +6.8	<0.001	259	-2.1	-4.1 to -0.1	0.043
BW (kg)	imputed	-	+3.1	+1.5 to +4.7	<0.001	-	+5.3	+3.6 to +7.0	<0.001	-	-2.2	-4.6 to +0.2	0.073
BMI (kg/m²)	available	125	+1.11	+0.44 to +1.77	0.002	132	+1.90	+1.26 to +2.54	<0.001	257	-0.79	-1.57 to -0.01	0.046
BMI (kg/m²)	imputed	-	+1.19	+0.59 to +1.79	<0.001	-	+2.00	+1.37 to +2.63	<0.001	-	-0.81	-1.69 to +0.08	0.074
WC (cm)	available	124	+1.9	+0.1 to +3.7	0.043	134	+4.8	+3.0 to +6.5	<0.001	258	-2.9	-5.0 to -0.8	0.008
WC (cm)	imputed	-	+2.1	+0.3 to +3.9	0.022	-	+4.5	+2.8 to +6.1	<0.001	-	-2.4	-4.7 to -0.0	0.046
Binary	Dataset	n/N	Pr	95% CI	p	n/N	Pr	95% CI	p	N	RR	95% CI	p
BW loss	available	40/125	0.32	0.24 to 0.41	<0.001	28/134	0.21	0.14 to 0.29	<0.001	259	1.60	1.04 to 2.46	0.032
BW loss	imputed	-	0.35	0.27 to 0.43	<0.001	-	0.25	0.17 to 0.32	<0.001	-	1.50	1.02 to 2.22	0.040
BW loss ≥ 5%	available	21/125	0.17	0.11 to 0.25	<0.001	12/134	0.09	0.05 to 0.15	<0.001	259	1.83	0.90 to 3.71	0.096
BW loss ≥ 5%	imputed	-	0.19	0.13 to 0.26	<0.001	-	0.12	0.06 to 0.18	<0.001	-	1.66	0.91 to 3.02	0.100
BMI reduction	available	40/125	0.32	0.24 to 0.41	<0.001	27/132	0.20	0.14 to 0.28	<0.001	257	1.61	1.04 to 2.50	0.032
BMI reduction	imputed	-	0.35	0.27 to 0.43	<0.001	-	0.24	0.17 to 0.32	<0.001	-	1.49	1.01 to 2.20	0.046
WC reduction	available	55/124	0.44	0.35 to 0.54	<0.001	36/134	0.27	0.20 to 0.35	<0.001	258	1.60	1.13 to 2.25	0.007
WC reduction	imputed	-	0.44	0.36 to 0.52	<0.001	-	0.30	0.22 to 0.37	<0.001	-	1.46	1.06 to 2.00	0.019

95% CI = 95% confidence interval. Pr = univariate proportion (= n/N). RR = adjusted relative risk (intervention group as exposure category versus control group as reference category). BW loss ≥ 5% of the inclusion measurement. Continuous outcome within group = follow-up measurement minus inclusion measurement. Δ = intergroup difference in mean change (intervention minus control). p for test with (H₀: mean = 0) or (H₀: Δ = 0) or (H₀: Pr = 0) or (H₀: RR = 1) according to outcome and comparison. All results, except Pr, were adjusted on baseline characteristics (gender, age, BW or BMI or WC, HbA1c, occupation, stress). Missing data were imputed under MAR assumption with MICE method implemented in Stata 10 (ice and Micombine packages). Variables included in the imputation models: baseline characteristics, exposure-group, participation in the medical visit at trial completion (yes/no), and the shifted log transformation of BW and WC at follow-up, respectively. Forty imputed datasets were generated from an original dataset of 432 observations.