Fifty years of fat: news coverage of trends that predate obesity prevalence.

BACKGROUND: Obesity prevalence has risen in fifty years. While people generally expect media mentions of health risks like obesity prevalence to follow health risk trends, food consumption trends may precede obesity prevalence trends. Therefore, this research investigates whether media mentions of food predate obesity prevalence.

METHODS: Fifty years of non-advertising articles in the New York Times (and 17 years for the London Times) are coded for the mention of less healthy (5 salty and 5 sweet snacks) and healthy (5 fruits and 5 vegetables) food items by year and then associated with annual obesity prevalence in subsequent years. Time-series generalized linear models test whether food-related mentions predate or postdate obesity prevalence in each country.

RESULTS: United States obesity prevalence is positively associated with New York Times mentions of sweet snacks (b = 55.2, CI = 42.4 to 68.1, p = .000) and negatively associated with mentions of fruits (b = -71.28, CI -91.5 to -51.1, p = .000) and vegetables (b = -13.6, CI = -17.5 to -9.6, p = .000). Similar results are found for the United Kingdom and The London Times. Importantly, the "obesity followed mentions" models are stronger than the "obesity preceded mentions" models.

CONCLUSIONS: It may be possible to estimate a nation's future obesity prevalence (e.g., three years from now) based on how frequently national media mention sweet snacks (positively related) and vegetables or fruits (negatively related) today. This may provide public health officials and epidemiologists with new tools to more quickly assess the effectiveness of current obesity interventions based on what is mentioned in the media today.