Social desirability bias is related to children's energy intake in a laboratory test meal paradigm

Dietary intake is notoriously difficult to measure in children. Laboratory test meals address some of the methodological concerns of self-report methods, but may also be susceptible to social desirability bias, referring to the tendency for individuals to adjust their behaviors in order to be perceived more positively. The aim of the current study was to evaluate whether social desirability bias was associated with children's energy intake during a laboratory test meal, and whether this association varied by food type (total caloric intake, snack food intake, fruit/vegetable intake) and sex. A total of 82 children (M age = 9.45 ± 0.85; 50 % girls; 84.1 % rural; 85.4 % White) completed several surveys, including the Children's Social Desirability Scale and had their body composition measured. At lunchtime, they were granted access to a multi-array test meal (>5000 kcal). After adjusting for lean mass, fat mass, depressive symptoms, and parental food restriction, children who reported higher social desirability bias consumed fewer calories from snack foods (B = -11.58, p = .009, semi-partial correlation = -0.28). Boys with higher social desirability bias consumed less calories from fruits and vegetables (B = -6.47, p = .010, semi-partial correlation = -0.411); this association was not significant in girls. The desire to be perceived in a positive manner may influence children's eating behaviors in experimental paradigms. Replication studies with larger, more diverse pediatric samples are needed, as are strategies to reduce the effects of social desirability bias on test meal intake in order to enhance the validity of this dietary assessment approach.

Selective screening of hospitalized patients for the presence of asymptomatic hepatitis B virus carriage: predictability based on census tract data

Review of all hepatitis B surface antigen-positive patients reported to the county health department over the past 3 years showed that 60% of patients had been diagnosed at our hospital, which serves an inner city, predominantly black, indigent population. Sera from 524 adult patients admitted to the hospital were prospectively collected and tested for hepatitis B surface antigen, eight (1.5%) of which were antigen-positive. Census tract data allowed us to perform selective screening of sera from 95 other patients who were subsequently hospitalized and deemed "high risk" for hepatitis B infection. Nine of 95 (9.5%) selectively screened samples were positive for hepatitis B surface antigen, which represented a significantly higher rate of antigen carriage (P less than 0.0001) in hospitalized patients from "high risk" areas than that found in the general population screen. The methodology presented herein may be useful in providing a means of identifying hepatitis B surface antigen-positive patients from adult, urban, indigent care populations.