Because diet quality (DQ) is associated with risk of chronic disease and is a common construct assessed in health-related research, validated tools to assess DQ are needed that have low respondent and researcher burden. Thus, content experts develop the Short Healthy Eating Index (sHEI) tool and an associated scoring system. The sHEI scoring system was then refined using a classification and regression tree (CRT) algorithm methodology with an iterative feedback process with expert review and input. The sHEI scoring system was then validated using a concurrent criterion validation process that included the sHEI DQ scores (calculated from responses from 50 participants) being compared to the participants’ Healthy Eating Index scores derived from 24 h recalls. The total HEI score from the CRT algorithm highly correlated with the 24 h recall HEI score (0.79). For individual food group items, the correlation between the CRT algorithm scoring and the 24 h recall data scoring ranged from 0.44 for refined grains to 0.64 for whole fruits. The sHEI appears to be a valid tool for estimating overall dietary quality and individual items (with correlations > 0.49) for fruits, vegetables, dairy, added sugar, sugar from sugar-sweetened beverages, and calcium.

Self-reported measures for body mass index (BMI) are considered a limitation in research design, especially when they are a primary outcome. Studies have found some populations to be quite accurate when self-reporting BMI; however, there is mixed research on the accuracy of self-reported measurements in adolescents. The aim of this study is to examine the accuracy of self-reported BMI by comparing it with measured BMI in a sample of U.S. adolescents and to understand gender differences. This cross-sectional study collected self-reported height and weight measurements of students from five high schools in four states (Tennessee, South Dakota, Kansas and Florida). Trained researchers took height and weight of students for an objective measurement. BMI was calculated from both sources and categorized (underweight, normal, overweight and obese) using the Centers for Disease Control and Prevention's BMI-for-age percentiles. Participants (n 425; 51⋅0 % female) had a mean age of 16⋅3 years old, and the majority were White (47⋅5 %). Limits of agreement (LOA) analysis revealed that BMI and weight were underreported, and height was overreported in the overall sample, in females, and in males. LOA analysis was fair for BMI in all three groups. Overall agreement in BMI categorisation was considered substantial (Κ 0⋅71, P < 0⋅001). As BMI increased, more height and weight inaccuracies led to decreased accuracy in BMI categorisation, and the specificity of obese participants was low (50⋅0 %). This study's findings suggest that using self-reported values to categorize BMI is more accurate than using continuous BMI values when self-reported measures are used in health-related interventions.

BACKGROUNDSocial media may be a useful tool in modifying health‐related behaviors; however, there is limited research on college students' use of health‐related social media and whether those behaviors vary by gender.METHODSIncoming first year students at eight universities were invited to complete screening surveys to determine if they were eligible to participate in Get Fruved, a study designed to decrease risk for unwanted weight gain. Eligible students (n=857) completed extensive health surveys and anthropometric measurements. Surveys included questions on social media usage and demographics. Pearson's Chi‐square tests were used to find differences in use of social media by gender.RESULTSMost participants (76.2% of females and 51.0% of males) reported they used social media to find health‐related information at least a couple of times a month, but only 10.2% of females and 4.4% of males reported using it more than four times per week. Increased time spent on social media was associated with increased use of social media to find health‐related information (p = .050) and reporting that social media impacted their view of nutrition (p = .017). Females were more likely than men to spend more time on social media (p = .003), use social media to find health‐related information (p < .001), and to report that social media impacted their view of nutrition (p < .001).CONCLUSIONResearchers need to determine if exposing college students to health‐related social media will influence behavior differently by gender.Support or Funding InformationThis abstract is based upon work that is supported by the National Institute of Food and Agriculture, U.S. Department of Agriculture, under award number 2014‐67001‐21851.

No abstract

No abstract