Table 5.

Multivariate linear regression analysis of associations of groups of tea and soda/sports/energy drink consumption in adolescents with anthropometric measures and blood biochemistry^*†

(β-Coefficients and 95 % confidence intervals)

Variable‡	LL (reference)	LH			HL			HH
	β	β	95 % CI	P	β	95 % CI	P	β	95 % CI	P
Log-BMI (kg/m2)	1·00	−0·004	−0·02, 0·01	0·569	−0·001	−0·01, 0·01	0·863	0·001	−0·02, 0·03	0·912
Log-WC (cm)	1·00	−0·011	−0·05, 0·03	0·597	0·001	−0·01, 0·01	0·806	0·008	−0·02, 0·04	0·552
Log-HDL (mg/dl)	1·00	−0·047	−0·17, 0·08	0·420	0·022	−0·02, 0·07	0·309	−0·012	−0·10, 0·07	0·753
LDL (mg/dl)	1·00	7·337	−1·22, 15·89	0·085	1·951	−3·26, 7·16	0·423	1·780	−3·88, 7·44	0·500
Log-TC (mg/dl)	1·00	0·030	−0·01, 0·07	0·098	0·017	−0·01, 0·04	0·194	0·015	−0·03, 0·06	0·479
Log-TAG (mg/dl)	1·00	0·035	−0·03, 0·10	0·257	0·011	−0·04, 0·06	0·602	0·024	−0·11, 0·15	0·683
Log-FBG (mg/dl)	1·00	0·005	−0·01, 0·01	0·318	0·002	−0·01, 0·02	0·766	−0·003	−0·03, 0·02	0·796
Log-SUA (mg/dl)	1·00	0·054	0·00, 0·10	0·034	0·022	−0·01, 0·06	0·181	0·048	−0·02, 0·11	0·133

LL, low tea (<7 times/week) and low soda/sports drink intake (<4·5 times/week); LH, low tea (<7 times/week) and high soda/sports drink intake (≥4·5 times/week); HL, high tea (≥7 times/week) and low soda/sports drink intake (<4·5 times/week); HH, high tea (≥7 times/week) and high soda/sports drink intake (≥4·5 times/week); WC, waist circumference; TC, total cholesterol; FBG, fasting blood glucose; SUA, serum uric acid.

^*β-Coefficients are linear regression coefficients derived from multivariable linear regression analyses. All models were adjusted for survey year, age, sex, metabolic equivalent per week (in kJ/kg per h) and total energy intake (in kJ).

^†This Table presents results without adjusting for total energy intake. The adjusted analyses were carried out and resulted in similar coefficients.

^‡Apart from LDL, other variables were all log transformed due to lack of a normal distribution.