Table 4. Multiple linear regression analysis for the associations of WC change with each parameter.

			β coefficient	95% CI	p-value
Male (n=1,699)
	Age		−0.021	−0.053 to 0.012	0.211
	Baseline WC		−0.116	−0.144 to −0.089	<0.001*
	Pre-intervention action history	Yes	−0.760	−1.168 to −0.351	<0.001*
		No	Ref.	Ref.	Ref.
	Number of days delayed		−0.0002	−0.005 to 0.005	0.949
	Videophone communications, n		−0.680	−1.051 to −0.308	<0.001*
	Messages, n		−0.585	−1.024 to −0.146	0.009*
	Smoking history	Yes	−0.113	−0.484 to 0.257	0.549
		No	Ref.	Ref.	Ref.
Female (n=295)
	Age		0.095	0.001–0.189	0.048*
	Baseline WC		−0.138	−0.211 to −0.066	<0.001*
	Pre-intervention action history	Yes	−1.906	−3.068 to −0.744	0.001*
		No	Ref.	Ref.	Ref.
	Number of days delayed		−0.008	−0.024 to 0.008	0.320
	Videophone communications, n		−0.035	−1.156 to 1.086	0.951
	Messages, n		0.731	−0.427 to 1.889	0.215
	Smoking history	Yes	0.131	−1.110 to 1.373	0.835
		No	Ref.	Ref.	Ref.

BMI, body mass index; BW, body weight; CI, confidence interval; WC, waist circumference.

Adjusted R²=0.057 (total), 0.047 (male), 0.086 (female) for WC change in multiple linear models. Reference (Ref.) indicates the reference for each value (β coefficient and corresponding 95% confidence intervals). Asterisk (*) indicates statistical significance (p<0.05).