The current study examined the longitudinal relationship between television viewing of adolescents in middle school and high school and their dietary intake five years later. Among participants in the older cohort, those who watched more than five hours of television per day when they were in high school reported less healthful eating habits (lower intakes of fruits, vegetables, whole grains and calcium-rich foods, and higher intakes of snack foods, fried foods, fast food, sugar-sweetened beverages, and trans fat). In contrast, television viewing in middle school only predicted lower fruit and greater sugar-sweetened beverage consumption five years later. Thus, television viewing during adolescence longitudinally predicted poorer dietary intake patterns five years later, with stronger and more consistent patterns seen during the transition from high school to young adulthood than during the transition from middle school to high school. Both time periods are critical developmental periods for adolescents, in which they are forming lifelong behaviors. However, behaviors exhibited in high school may more strongly influence behaviors reported in subsequent years than behaviors exhibited during middle school. Additionally, there may be environmental influences for which the current study did not have data, such as food and beverage television advertising, which may influence the eating behaviors of the cohorts differently.
Several cross-sectional studies have found a negative correlational association between television viewing and diet among students in preschool , elementary school [11, 29, 30], middle and high school [12, 31, 32], and college [33, 34]. However, only two studies, both using data from Planet Health, which engaged middle school students in an obesity prevention intervention, have reported longitudinal relationships between greater amounts of television viewing and lower fruit and vegetable intake  and higher sweet and salty snack foods, fast food, and sugar-sweetened beverages  – results similar to those found in the current study. However, there are substantive differences between the Planet Health studies and the current study. Planet Health was an intervention that targeted reducing television viewing; the current study explored the influence of previous television viewing behavior over a longer period of time (5 years compared to 19 months); and the current study examined the transition from high school to the post-high school years, which to the authors' knowledge, has not been examined by any other research.
Among both cohorts, categories of food and beverages (snack foods, sugar-sweetened beverages, and fast foods) commonly advertised on television  were associated with previous television viewing behavior in the current study. Healthy foods, such as fruits, vegetables, whole grains, milk, and low-fat items are rarely advertised on television . Repeated exposure to high calorie, low nutrient foods may increase desire for these foods and subsequently purchases and consumption of advertised products. Adolescents who watch too much television become adults who watch too much television , and thus continue to be exposed to advertisements for unhealthy foods. Although young people may be aware that many foods advertised on television are not healthy, they may chose to ignore or do not fully realize the consequences, because the actors they see advertising and eating the foods in the commercials are usually not overweight .
Post-hoc analyses in the older cohort showed there was an interaction with race/ethnicity in the observed relationship between television viewing and dietary intake. The stronger influence found in minority, especially black, participants may reflect the types of advertisements or program content most likely to reach these racial/ethnic groups. More food advertisements that focus on fast food, candy, and soda are shown during programs targeted to black Americans than programs targeted to general audiences .
There is also the issue of eating while viewing television. Studies have found that for some adolescents, a significant proportion of their total daily energy intake is consumed while viewing television [30, 38]. Viewing television while eating may cause a distraction resulting in a lack of awareness of actual food consumption or overlooking satiety cues [33, 39], which may lead to overconsumption. If these habits are formed at a younger age, they may continue as adolescents become older.
To our knowledge, this is the first study to examine the long-term longitudinal association between television viewing and dietary intake during the transition through the adolescent years into young adulthood. The strengths of this study are a prospective design; use of large, socioeconomically and racially diverse samples; and the examination of multiple dietary behaviors in two age cohorts. A limitation of the current study is that specific mechanisms through which television viewing may be influencing dietary behavior cannot be examined with these data. For example, the type and content of commercials watched by the participants are unknown. Future research is needed to explore these issues in more detail. Although the longitudinal nature of the current study allows for the assessment of temporality of the association between television viewing and dietary behavior, it does not allow for an assessment of causality. Television use may be overestimated in this study because the measure used to assess television use also included video use. It is unknown whether television viewing and video viewing influence dietary behaviors differently. However, among adolescents the proportion of total screen time spent watching videos is small compared to the proportion of time spent watching television .