Studies examining an association between consumption of added sugars (AS) and weight
measures in children are inconclusive. This study examined the association between intake of AS
and 5 measures of weight or adiposity using a nationally recent representative sample of children.
National Health and Nutrition Examination Surveys 2003–2006 24-hour recall data from children 6
to 18 years (n = 3136) were used. United States Department of Agriculture's definition of AS and
MyPyramid Equivalents Database were used to estimate daily intake. Multiple linear regression and
squared partial correlation coefficients were used to estimate the strength of association between
weight, body mass index (BMI), BMIZscores, waist circumference, and triceps and subscapular
skinfolds as dependent variables with AS as the independent variable. Covariates were age, sex, race/
ethnicity, poverty income ratio, total energy intake (kJ), and physical activity. Mean intake of AS
was 23 ± 0.55 teaspoons (tsp) (21 tsp for 6-11 years and 25 tsp for 12-18 years) accounting for 17%
of total energy intake for both groups. The percent variance explained in BMIZscores was 3.9% in
children 6 to 11 years, with AS contributing only 0.03%. In children 12 to 18 years, the percent
variance explained in the BMIZscores was 6.5% with 0.18% coming from AS. No significant
associations were observed between intake of AS and weight or adiposity measures. Consumption of
AS did not contribute significantly to BMIZscores in children 6 to 18 years. Longitudinal studies
using more robust and precise measures of dietary intake are needed to further investigate the role of
AS and weight in children.
© 2011 Elsevier Inc. All rights reserved.
The NHANES data were used to compare intakes of AS
and body weight/adiposity measures in children and
adolescents 6 to 18 years (n = 3136). The age groups, 6 to
11 and 12 to 18 years were used because they reflect the
MyPyramid age groups and AS is part of discretionary
energies outlined in MyPyramid[2]. Only individuals with
complete demographic and physical activity data, with body
measurements, and whose dietary interview data were
deemed reliable by the Food Surveys Research group staff
were included. Because this was a secondary data analysis
and the participants lacked personal identifiers, this study
was exempted by the institutional review board of the Baylor
College of Medicine.
2.2. Demographics and physical activity
Age, sex, race, ethnicity, and poverty income ratio were
collected on the participants during the interview in the
mobile examination centers. Physical activity was measured
using an accelerometer, and the mean duration (minutes) of
moderate and vigorous activity bouts (minimum 1-minute
bouts) per day were calculated using the programs from
National Center for Health Statistics[36].
2.3. Anthropometric measures