Abstract MP92: Sedentary Behavior: Do Television-viewing and Total Time Sitting Measure the Same Component?

Abstract

Background: Sedentary behaviors (SB) are activities associated with prolonged time periods of sitting, reclining, or laying down during waking hours. While the relation between SB and physical activity is complex, the common consensus is that SB is not the absence of physical activity and consists of its own determinants posing distinct health outcomes. These behaviors are of significant public health importance as the majority of Americans spend much of their days in SB and due to the increased risks of morbidity and mortality associated with SB. Adverse health outcomes associated with SB include cardiovascular disease, obesity, metabolic syndrome, hypertension and mortality. Television-viewing time and total sitting time have both been used widely to assess time spent in SB and therefore we hypothesize that TV-viewing time and total hours sitting will have high concordance and can be used interchangeably to represent sedentary behaviors. Methods: Using a sample (n = 2858) from the Images of a Healthy Worksite study, a group-randomized control trial involving nutrition and physical activity, the current study assessed how two different tools measured time spent in SB. Tertiles were created based upon the distribution of hours sitting and hours spent TV-vewing. Weighted Kappa statistics were used to measure concordance between hours of TV-viewing and total hours of time spent sitting for the entire sample and for subgroup analyses. Results: Weighted Kappa statistics for tertiles of hours sitting and tv hours were 0.0046, indicating little agreement on the television and the sitting items. Kappa w statistics for BMI categories also showed poor agreement (obese Kappa w = 0.02, overweight Kappa w = 0.002, and healthy subjects Kappa w = 0.006. The Kappa w statistics for males and females were -0.006 and 0.02, respectively. Kappa w statistics for the intervention group (Kappa w = 0.007) and for the control group (Kappa w = 0.0005) also showed little agreement. Conclusions: These results suggest that although commonly used, using television viewing time and total time spent sitting as interchangeable markers of SB, is not a valid assumption. We propose that total time spent sitting and hours spent television-viewing represent different domains within the construct of sedentary behavior. It is important for future researchers to use measures of sedentary behavior that capture the numerous domains involved in measuring SB to allow for the most sensitive measurement of this high-risk behavior.