Table 6.

Sugar-sweetened beverage consumption according to socio-economic or cultural characteristics of adolescents^* (four reports)

First author, year, reference	Population, design, time of collection, country	Age (years)	n	Diet collection method	Intake or frequency of consumption	Exposure variables (number of categories)	Association	Adjustments
Drouillet-Pinard (2017)(56)	INCA2 study, cross-sectional, 2006–2007, France	11–17	881	Record (7 d)	Intake (g/d)	Parental occup. (4)	NS	Age, gender and energy intake
						Parental educ. (3)	Higher when educ. lower
						Household income (tertiles)	NS
						Household wealth index (tertiles)	Higher when wealth lower
						Global SES index (all SES indicators combined, tertiles)	Higher when SES lower
Grosso (2013)(50)	Secondary schools of Sicily, cross-sectional, 2010–2011, Italy	13–16	1135	FFQ	Intake (g/d)	SES index (parental educ. and occup., 3)	Higher when SES lower	Age, gender, BMI, physical activity, place of living and SES
Harris (2015)(57)	GINIplus study, cohort at follow-up age 10 and 15 years, 2005–2008 and 2010–2013, Germany	10–15	1232	FFQ	Change v. tracking of intake over time (% of energy intake)	Parental educ. (2)	NS	Age at baseline, baseline energy intake, diet changes, study centre, study intervention arm, pubertal onset, BMI, screen time and all SES variables
						Family income (tertiles)	NS
Llull (2015)(58)	Balearic Islands, cross-sectional, 2007–2008, Spain	12–17	1231	FFQ	Daily consumption	Birthplace (4)	Higher for Latin America and other countries than Balearic Islands	Gender and age
						Birthplace (2)	Higher for non-Mediterranean
						Length of time living in Balearic Islands (4)	Higher when length of time lower

INCA2, second French national cross-sectional dietary survey; occup., occupation; educ., education; SES, socio-economic status.

^*Details on risk of bias assessment are not presented since only studies of good quality are tabulated.