www.nutritionj.com

Table 7
Regression model estimating change in energy intake associated with adding caloric beverages in place of non-caloric beverages
Model 1
	B	P value	95% Confidence interval for B
			Lower	Upper
(Constant)	1040	<0.0001	852	1227
Sex -female	−371	<0.0001	−414	−328
Age (years)	−1.6	0.044	−3.1	0.0
current smoker	−63	0.001	−100	−25
weight (kg)	7.2	<0.0001	6.0	8.4
Dieting	−67	0.005	−114	−20
Activity (MET)	147	<0.0001	104	190
Valid reporter	659	<0.0001	617	700
Total beverages (100 g/d)	11	<0.0001	10	12
4caloricbeverages (100 g)*	15	<0.0001	10	20
Model 2, Keeping Food Constant
	B	P value	95% Confidence Interval for B
			Lower	Upper
(Constant)	62	0.01	15	109
Sex -female	−20	0.001	−31	−8
Age (years)	0.4	0.023	0.6	0.8
current smoker	16	0.001	7	26
weight (kg)	0.2	0.241	−0.1	0.5
Dieting	1	0.826	−10	13
Activity (MET)	−4	0.513	−14	7
Valid reporter	40	<0.0001	28	53
Food (kcal)	0.97	<0.0001	0.96	0.98
Total beverages (100 g)	2	<0.0001	1	3
4caloricbeverages (100 g)*	34	<0.0001	33	36

* 4caloric beverages = sum of milk, fruit juice, caloric soft drinks and alcoholic drinks.

Gibson and Shirreffs Nutrition Journal 2013 12:9 doi:10.1186/1475-2891-12-9

Table 7