The NHANES 2005–6 survey incorporated new questions to evaluate nutrition awareness of the Dietary Guidelines for Americans, the Food Guide Pyramid, the 5-A-Day Program, and diet behavior such as use of the nutrition facts label and the ingredient list. The NHANES is a continuous national survey that represents a stratified multistage probability sample of the non-institutionalized US population. Health interviews are conducted in participant’s homes and measurements are performed in specially designed and equipped mobile centers . The interview team consists of a dietary and health interviewers and a physician. Most of the staff are bilingual in English and Spanish. An advanced computer system is used to enter respondents’ information using notebook computers. Respondents are able to enter their own responses to sensitive questions in privacy using a touch-sensitive computer screen . Transportation to and from the mobile center is provided if necessary and participants are provided with compensation for participation. Adolescents 16 and 17 years of age have to have a parent read and sign the consent form in addition to signing a Household Interview Consent Form .
This study investigated the relationship between responses to the diet behavior questions using NHANES 2005-6 data for adolescents between the ages of 16 and 19 (n = 1160 designed to represent a sample size of 16,204,982). Seventeen different questions were asked on nutrition awareness and diet behavior as part of the NHANES 2005–6, although only a sub-sample of the 1160 answered questions on label reading behavior as described below. The main outcomes of interest were (1) awareness of nutrition program, and (2) active use of nutrition panel information defined as ‘always' or ‘most of the time’ using the nutrition fact label information.
Specifically, awareness of nutrition guidance programs such as the Dietary Guidelines for Americans, the Food Guide Pyramid and the 5-A-Day for Better Health Program was assessed (3 questions) as was use of the nutrition facts panel and additional label reading behavior (ingredient list, serving size or health claims information) (4 questions). The questions were of the form, “Have you heard of The Dietary Guidelines (or other nutrition program)?” and was repeated in the same format for the other nutrition guidance programs . Participants could answer ‘yes’ or ‘no’ that they were aware of one of the nutrition programs.
For the question on the food label they were given a card with a food label on it and asked, “Here is an example of a food label. This part of the food label is called the “Nutrition Facts” panel. How often do you use the Nutrition Facts panel (or other part of the food label) when deciding to buy a food product?”  They could answer that they ‘never’, ‘rarely’, ‘sometimes’, ‘most of the time’ or ‘always’ made use of the nutrition facts panel information, ingredient list, serving size or health claims information. As part of the NHANES dietary behavior interview process, additional label reading behavior was assessed among the sub-sample of adolescent respondents who made use of the nutrition facts panel, the serving size information, the ingredient list or the health claims on the package (n = 742). In this subgroup, label reading behavior was assessed for use of total calories, calories from fat, total fat content, trans fat content, saturated fat, cholesterol, sodium, carbohydrates, fiber and sugars in making a product selection (10 questions). The questions were phrased in the survey as follows, “When you use the food label to decide about a food product, how often do you look for information about X?  Would you say always, most of the time, sometimes, rarely or never?”
Quality assurance and control measures included using trained interviewers with a Computer-Assisted Personal Interviewing (CAPI) system to interview each participant. The CAPI system has built-in consistency checks to reduce data entry errors. All data were reviewed by NHANES field staff for completeness and accuracy . NHANES documentation does not indicate that the diet behavior questions which specifically asked about awareness of federal nutrition programs and use of nutrition facts panel information were tested for validity or reliability.
As part of the NHANES protocol, participants were weighed and measured. To collect weight measurements, participants stood on a floor scale, equipped with a digital read-out. Standing height was measured using a wall-mounted stadiometer. The stadiometer was connected to an automated data electronic database and data were entered automatically . Body mass index (BMI) was calculated as weight (kg)/(height (m))  and CDC growth charts and classifications were were used to determine overweight and obesity . Adolescent overweight was defined as having a BMI percentile ≥85th or, for those 18 years or older, a BMI ≥25 and <30. Adolescent obesity was defined as having a BMI percentile ≥ 95th or a BMI ≥ 30.
Percentages and 95% confidence intervals were calculated for all means and frequencies. We applied chi-squared tests to evaluate differences in proportions and student’s t-tests to evaluate differences in means to compare how socio-demographic and health variables differed in adolescents who were aware of nutrition programs and made active use of the nutrition panel information in comparison with those with did not. Unadjusted logistic regression models were used to assess binary outcomes for awareness of nutrition programs and frequent use of the nutrition facts label (always or most of the time) in relation to adolescent BMI percentile or category. Interaction terms for race/ethnicity and different nutrition label reading behaviors and nutrition awareness were evaluated to determine possible interactions between these variables and risk for overweight or obesity. Multivariate logistic models were used to compute adjusted estimates for the relation between nutrition awareness and label reading behavior and risk for adolescent overweight. Models were adjusted for adolescent age, race/ethnicity, foreign born versus US born and living in poverty. NHANES supplied sampling weights and strata were used to analyze the population survey data. All analyses were done using Stata 11.0 using svy commands for survey data. Data are expressed as means ± standard errors (SE).
The NHANES 2005–6 surveys received approval from the National Center for Health Statistics Ethics Review Board (ERB) (Protocol # 2005–6) . The University of California, San Francisco (UCSF) Committee on Research (CHR) concluded that the study did not need ethical approval as all analyses involved de-identified data.