Physical activity levels and cardiovascular disease risk among U.S. adults: comparison between self-reported and objectively measured physical activity
Background: In 2008, the US Department of Health and Human Services released the first set of official US physical activity guidelines (Physical Activity Guidelines for Americans - PAGA). These guidelines are based on numerous studies that have linked self-reported PA to substantial health benefits. However, levels of PA among US adults remain relatively low, and differ substantially depending upon whether measured subjectively or objectively. Therefore, a better understanding between these measures is needed. Additional evidence linking objective PA to health outcomes is also needed. Furthermore, no study has examined the proportion of Americans achieving the PAGA using an objective measure. Purpose: To evaluate PA assessment issues using data from the National Health and Nutrition Examination Survey (NHANES). The proportion of US adults meeting PAGA was examined using both self-report and accelerometry-based activity monitors. In addition, the relationship between PA and cardiometabolic risk was assessed using subjective and objective measures to allow for comparison between these tools. Methods: Levels of moderate plus vigorous PA were assessed using self-reported and accelerometer PA data from the NHANES 2005-2006. Self-reported PA in the NHANES was captured by a series of questions that assessed PA performed during leisure, transportation, and household tasks during the past 30 days. Accelerometry-based PA was captured using Actigraph GT1M accelerometers, which participants wore for 7 continuous days. Risk factors for cardiovascular disease (CVD) were assessed during the examination portion of the NHANES, and included blood pressure, blood lipids, blood glucose, and waist circumference. Metabolic syndrome and the Framingham risk score (FRS), used to estimate 10-year risk of coronary heart disease, were assessed as composite measures of CVD risk. Logistic regression was used to compare CVD risks for participants who met the PAGA vs. those who did not. Results: According to self-report, males and females obtained (mean y SE) 380.5 y 23.5 min/wk and 272.6 y 18.3 min/wk of MVPA, respectively. Weekly, accelerometry-based MVPA included 52.9 y 8.8 min/wk for males and 37.8 y 4.6 min/wk for females. In total, 62.0% of US adults achieved the PAGA according to self-report, but only 9.6% met the PAGA according to accelerometry. According to Spearman correlations, FRS showed a stronger, inverse relationship with accelerometer MVPA in men (r = -0.57) and women (r = -0.50) when compared to self-reported MVPA in men (r = -0.08) and women (r = -0.19). Logistic regression results showed that men who did not meet the PAGA according to accelerometry had significantly higher odds of having metabolic syndrome (OR = 2.10) when compared to men who met the PAGA after controlling for age, race, education, and smoking status. However, self-reported PA levels did not yield significantly different odds of having metabolic syndrome. For females, failing to meet the PAGA resulted in significantly greater odds of the metabolic syndrome when measured by both accelerometry (OR = 2.30) and self-report (OR = 1.66). Discussion: The proportion of adults meeting the PAGA is slightly higher when compared to studies that have examined previous guidelines. However, a large discrepancy still exists between self-reported and objective PA levels among US adults. According to the NHANES, accelerometry-based PA provides a stronger relationship with CVD risk factors when compared to self-reported PA, and therefore, should be encouraged for use in future epidemiological and health-related research whenever possible.