Abstract

The poorer health status experienced by Hispanics, the largest US minority group, is a concern. Since participating in moderate to vigorous physical activity (MVPA) can lead to health benefits, assessing the participation of specific physical activities (i.e., a list of MVPA, e.g., basketball, walking, etc.) has been widely used in national surveillance systems to facilitate MVPA assessment at a population level. However, it is unknown if the MVPA items employed are biased to Hispanics. By controlling overall PA participation, an item is a biased, or DIF (differential item functioning), item if it favors a subgroup. PURPOSE: The purpose of this study was to use the Mantel-Haenszel (MH) DIF method to determine if there are DIF PA items in the National Health and Nutrition Examination Survey's (NHANES) self-report PA questionnaire across Non-Hispanic White (NHW) and Hispanic (H) subgroups. METHODS: A NHANES 2003-2004 adult data subset (unweighted N =2,214, NHW = 1,688 and H = 526; including demographic, PA questionnaire, PA individual activity and PA activity monitor information) was analyzed. There were 38 specific MVPA items and participation in a specific activity was a "1=Yes" or "0=No." Ability levels of the reference (NHW) and focal (H) groups were matched using both internal (total MVPA MET*Minutes over 30 days) and external matching criterion (average physical activity counts during a usual ActiGraph data collection week). RESULTS: Nine of the specific activity items functioned differently across NHW and H subgroups, with significantly large DIF sizes (MH D-DIF>1.5 and p<0.01). The analyses based on both internal and external matching criterion confirmed the observations. NHW were more like to do the activities of bicycling (DIF =2.13), fishing (2.82), golf (3.05), swimming (2.54), walking (1.64) and hunting (5.85) while H favored the activities of dance (1.87), hiking (2.08) and soccer (2.12). Soccer was the most likely engaged in activity for H males while hiking was the preferred activity for H females. CONCLUSIONS: There were some DIF MVPA questions used in the NHANES PA self-report questionnaire, which could potentially affect the comparison between Non- Hispanic white and Hispanic subgroups. The identification of Hispanic favored PAs allows for PA promotion targeting this subpopulation.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call