Preventive Medicine 63 (2014) 58–62
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Ethnic disparities in objectively measured physical activity may be due to occupational activity Jennifer L. Gay a,⁎, David M. Buchner b a b
Department of Health Promotion and Behavior, University of Georgia, Athens, GA, USA Department of Kinesiology and Community Health, University of Illinois, Urbana–Champaign, IL, USA
a r t i c l e
i n f o
Available online 28 February 2014 Keywords: Exercise Health disparities Accelerometer Mexican-American
a b s t r a c t Objectives. We examined whether Mexican American adults report occupations that involve higher levels of objectively assessed physical activity compared with Non-Hispanic White and Black adults, and if the differences were independent of income. Methods. Data from the 2003–2004 National Health and Nutrition Examination Survey (NHANES; N = 2707) were analyzed in 2012–2013. An existing classification scheme was used to classify self-reported occupation as sedentary, low-active, or moderately active. From NHANES accelerometer data, proportion of wear time was stratified by intensity. Results. A dose–response relationship was found such that workers in more active occupations spent more time in light-intensity activity and less time engaged in sedentary activities. The findings did not suggest a compensation effect for moderate-to-vigorous intensity physical activity. Mexican American adults engaged in more activity than Non-Hispanic Black or White adults for incomes between $10,000 and $64,999. Conclusions. Mexican American adults may have higher total physical activity levels in NHANES because of occupational activity, particularly among lower income households. To the extent that light-intensity activity may provide health benefits, occupational activity may partly explain the Hispanic paradox. © 2014 Elsevier Inc. All rights reserved.
Introduction In Healthy People 2020 physical activity is a leading health indicator (U.S. Department of Health and Human Services, 2012). Like most other important health indicators, it demonstrates disparities in health by ethnic group. In nationally representative surveys of adults such as the Behavioral Risk Factor Surveillance System (BRFSS), self-reported levels of physical activity are consistently higher in Non-Hispanic Whites than in African-Americans and Hispanics (Carlson et al., 2008). Disparities in self-report activity persist between Non-Hispanic Whites and Hispanics after adjustment for socioeconomic status, though levels of activity vary considerably among Hispanic subgroups (Neighbors et al., 2008). However, the first nationally representative survey to use an objective measure of physical activity (accelerometers) reported a different pattern of behavior: Mexican American adults (≥ 20 years) had higher mean activity counts than Whites and African-Americans and more minutes per week of moderate-to-vigorous physical activity (Troiano et al., 2008).
⁎ Corresponding author at: Department of Health Promotion and Behavior, 311 Ramsey Center, University of Georgia, Athens, GA 30602, USA. Fax: +1 706 542 4956. E-mail address: [email protected] (J.L. Gay).
http://dx.doi.org/10.1016/j.ypmed.2014.02.015 0091-7435/© 2014 Elsevier Inc. All rights reserved.
The finding of higher levels of objectively measured physical activity in Mexican American adults is of great interest, as it may help explain why Hispanic Americans have a life expectancy advantage over Whites and African-Americans. The advantage is about 7.7 years over nonHispanic blacks and about 2.5 years over Non-Hispanic Whites (Arias, 2010). This advantage is supported by data that US Mexican Americans have lower rates of mortality from cardiovascular disease and certain forms of cancer, and has been referred to as the “Hispanic Paradox” (Markides and Eschbach, 2005). The advantage is regarded as paradoxical because lower socioeconomic status is associated with higher mortality, and on average Hispanics have lower socioeconomic status (SES) than the Non-Hispanic White population (CDC, 2011). Because higher levels of physical activity reduce risk of cardiovascular disease and cancer often independent of weight status (Gay et al., 2013; Wei et al., 1999), and mortality risk (Lee and Skerrett, 2001), understanding ethnic and racial differences in physical activity behavior may help explain the “Hispanic Paradox.” One possible explanation for higher levels of physical activity in Mexican American adults is that they have higher levels of occupational activity. Currently, BRFSS questions assess three domains via selfreport: recreational, transport, and household activity; the questions do not assess occupational activity (Centers for Disease Control and Prevention, 2007). The National Health and Nutrition Examination Survey (NHANES) accelerometer protocol asked participants to wear the
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
accelerometer for a week during all waking hours (Troiano et al., 2008). That is, the accelerometers measured all domains of activity, including occupational physical activity. Recent literature highlights the need to better understand the ethnic disparities in physical activity (Ham and Ainsworth, 2010), particularly in relation to socioeconomic class (Clarke et al., 2009). Therefore, the first aim of this study was to analyze NHANES data to determine if Mexican American adults report occupations that involve higher levels of physical activity. Occupations with higher levels of activity often involve manual labor (i.e., physical work that may be repetitious) (U.S. Department of Labor, 2008). Wages for these jobs are typically lower than white collar positions that are usually salary-based (Jones, 2009; U.S. Department of Labor, 2008, 2011). One possibility is that higher occupational activity has nothing to do with ethnicity and culture, and that any group with lower household incomes would have higher levels of occupational activity. But this does not consider households on a single income, or those with an adult with disability who cannot work. And on average Mexican Americans have lower household incomes than Whites (Centers for Disease Control and Prevention, 2007). Furthermore, African-Americans also have lower household incomes and yet are not as active as Mexican Americans in NHANES (Troiano et al., 2008). So the second study aim was to analyze total physical activity levels stratified by income, to see if the effect of ethnicity is independent of income. Reasoning that Mexican Americans may work in physically active occupations more often than non-Hispanic adults and most of these jobs should be lower paid, the study hypothesized that Mexican Americans with lower incomes should have higher total activity than Whites and Blacks. However, Mexican Americans with higher incomes might not be as physically active, as so few higher income jobs nowadays in the United States require much physical activity. Methods Study population NHANES assesses a variety of health conditions and behaviors among children and adults in the United States. NHANES employs a multi-stage probability sampling frame, and consists of interviews, laboratory tests and a physical examination. In the 2003–2004 wave of data collection, an objective measure of physical activity was included in the examination. During this year 10,122 adults and children participated in NHANES. Demographics Study participants were asked to self-report age and self-identify with racial and ethnic groups: Mexican American, Other Hispanic, Non-Hispanic White, Non-Hispanic Black, and Other Race including multi-racial. In concordance with the NHANES Analytic Notes (Centers for Disease Control and Prevention, 2011), participants reporting Hispanic ethnicity (n = 211) other than Mexican American were also excluded. To be consistent with other published NHANES accelerometer data, adults aged 20 years and older were included in the analyses. Household income was reported at the family level and was collapsed into 11 categories ranging from b $5000 to N$74,999. Participants who did not provide income information (n = 170) were excluded from the analyses. Physical activity The Actigraph AM-7164 accelerometer (formerly the CSA/MTI AM-7164; Fort Walton Beach, FL) was used to objectively measure physical activity behavior. Participants wore the monitor on the right hip. Accelerometers were programmed to count activity in one-minute epochs. Data were summarized into one observation per participant using adapted SAS code provided by the National Cancer Institute (http://riskfactor.cancer.gov/tools/nhanes_pam/). Participants were included if they had a minimum of 4 days with at least 10 h of valid wear time. The SAS program provided by the National Cancer Institute summarizes data by intensity for moderate- and vigorous-intensity physical activity (MVPA) using established cut points (Freedson et al., 1998). The program
59
also classifies episodes of MVPA into both 1-minute and 10-minute bouts, where 1-minute bouts include all minutes of MVPA and 10-minute bouts count only MVPA periods long enough to meet US physical activity guidelines (U.S. Department of Health and Human Services, 2008). The program was adapted so that light-intensity physical activity was distinguishable from sedentary time by counting the number of minutes N 100 counts/ minute but less than the moderate-intensity threshold. The abbreviation LSBPA signifies Light-intensity + Short Bouts of MVPA; physical activity that is either of light-intensity (1.5 to 3.0 METs) or is a bout of MVPA less than 10 min in length. LSBPA represents physical activity that is not counted toward meeting current US guidelines (U.S. Department of Health and Human Services, 2008). Occupation Participants self-reported occupation and type of business. NHANES staff recoded the occupations according to the U.S. Census Bureau's Census 2000 Indexes of Industry and Occupations. The recoded variable for occupation available for analysis includes 41 occupational categories ranging from Management Related Occupations to Teachers to Construction Laborers. Two raters associated occupations with MET values (Tudor-Locke et al., 2011) based on job functions. MET values were categorized as sedentary (b2 METs), low (2.0–2.9 METs) and moderate (3.0–5.9 METs). Discrepancies between raters were resolved by examining related job descriptions and discussion with the study PI until consensus was reached. Participants indicating that they were currently unemployed were classified as Non-Workers. Data analyses Participants were excluded from analyses for being younger than 20 years (n = 2755). Of the 4075 participants aged 20 years and older who wore the accelerometer, 987 were excluded for invalid wear time, 211 for reporting a race or ethnicity excluded according to the NHANES analytic guidelines (Centers for Disease Control and Prevention, 2011) and 170 for missing household income. The final sample was analyzed in 2012 and 2013 using the sampling weights for the accelerometer sample with at least 4 valid wear days. A total of 2707 participants were included. Sample characteristics were calculated for age in years, BMI, accelerometer wear time, minutes spent in sedentary time, LSBPA and MVPA stratified by gender. The proportion of accelerometer wear time spent in sedentary behavior, LSBPA and MVPA was also stratified by occupational activity classification (Non-Worker, Sedentary occupation, Low Active or Moderately Active occupation) and accounted for sampling weights. Proportion of wear time was used because even after setting minimum requirements for accelerometer wear time, minutes of activity was confounded by total minutes of wear time. Using the proportion of wear time accounts for this intra-individual variation. To address the first study aim examining whether Mexican Americans report occupations that involve higher levels of physical activity, the chi-square test of proportions for occupational classification was calculated by race/ethnicity, stratified by gender. Ordinary least squares linear regression analyses using PROC SURVEYREG were conducted to assess if Mexican Americans worked more often in physically active occupations, independent of income, compared with their nonHispanic peers. The regression model for mean accelerometer activity counts controlled for accelerometer wear time, age, gender, BMI, occupational activity classification, and whether the participant was born in the US or abroad. Household size was also included as a covariate because the income variable represented household rather than individual income. An interaction term was included for race/ethnicity with household income to address the second aim of whether the effect of ethnicity is independent of income. Analyses accounted for the weighting, stratification and clustering of the NHANES survey design.
Results Participant characteristics stratified by gender and race/ethnicity are shown in Table 1. Of participants reporting their country of birth, 17.4% were foreign-born, with nearly 70.1% reporting being born in Mexico. The sample (n = 2707) was 51.8% female with a mean age of 46.81 years (SE = 0.56). Of the participants reporting an occupation (n = 1484) 37.2% reported occupations that were sedentary,
60
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
40.5% reported Low Active and 22.3% Moderately Active occupations. In the complete sample, the mean minutes per week of LSBPA were 388.74 (SE = 3.39) and mean minutes per week of MVPA were 25.82 (SE = 0.90). There were significant differences in the age of employed (M = 43.85) versus non-working participants (M = 63.52, P b 0.0001). However, much of this difference may be attributed to 52.4% of the non-working participants being retired. No differences by BMI were detected between Workers and Non-Workers (P = 0.542). There was a dose–response trend for mean accelerometer counts per minute of activity by occupational classification (F = 118.02; P b 0.0001) such that participants in Moderately Active occupations (M = 423.89) recorded more activity counts than workers in Low Active occupations (M = 355.27), Sedentary occupations (M = 325.57) and participants who were unemployed (M = 240.56). Figs. 1a and b illustrate the proportion of accelerometer wear time spent in sedentary behavior, LSBPA and MVPA stratified by occupational activity classification (Non-Worker, Sedentary occupation, Low Active or Moderately Active occupation). There is a dose–response relationship whereby workers in increasingly active occupations spent more time in LSBPA (F = 88.63, P b 0.0001) and less time in sedentary behavior (F = 95.46, P b 0.0001). The data also showed a dose–response relationship for MVPA (F = 145.37, P b 0.0001), but it was not as steep. Results of the chi-square test of proportions for occupational classification by race/ethnicity and stratified by gender are shown in Table 2. Among both men and women, Mexican Americans reported Moderately Active occupations more frequently than their White or Black counterparts (P b 0.0001 for men and P = 0.05 for women). White and Black men and women reported more sedentary occupations compared to Mexican American adults. Furthermore, fewer foreignborn Mexican Americans were employed in sedentary jobs (8.8%) than their US-born peers (25.3%, P b 0.0001). Multiple regression analyses were conducted to examine the interaction of race and ethnicity with household income (Fig. 2). The interaction term for household income and race/ethnicity was significant (P b 0.0001). Inspection of Fig. 2 revealed the interaction effect was significant because Mexican American adults consistently recorded higher mean accelerometer activity counts than White or Black adults when household income is between $10,000 and $64,999. The regression model accounted for 37.0% of the variation in mean accelerometer activity counts.
Mexican Americans had higher mean activity counts even within the same income ranges as White and Black men and women. The higher mean activity counts were most evident for participants with incomes between $10,000 and $64,999. This study expands upon previous accelerometer data characterizations of NHANES participants' physical activity behavior (Hawkins et al., 2009; Troiano et al., 2008) by assessing the moderating effect of race and ethnicity on the association between income and physical activity. There was also a significant dose–response relationship for proportion of wear time spent in MVPA by occupational activity classification. This suggests that adults with more active occupations do not compensate by engaging in less MVPA than workers with sedentary or Low Active occupations. Another study of questionnaire assessments of physical activity in NHANES also reported that more occupational physical activity is associated with more (not less) recreational activity (Wolin and Bennett, 2008). Though Mexican Americans clearly have more counts per minute of activity, the extent that this activity provides substantial health benefits is not clear. Almost all of the difference among ethnic groups in counts per minute is due to differences in LSBPA. The health benefits of LSBPA are not yet proven, though there is growing evidence that more LBSPA (or roughly equivalently less sedentary time) provides health benefits. For example, in the Australian Diabetes, Obesity and Lifestyle study, men and women who engaged in activities at light-intensity had lower blood glucose compared to sedentary participants (Healy et al., 2007). Walking at a low intensity was linked to reduced risk of coronary heart disease in the Women's Health Study (Lee et al., 2001). It was recently estimated that reducing excessive sitting to less than 3 h per day would increase life expectancy by 2.0 years (Katzmarzyk and Lee, 2012). However, studies of occupational physical activity generally report that people in more active occupations have lower rates of chronic diseases such as cardiovascular disease and diabetes (Hu et al., 2007; van Uffelen et al., 2010). In particular, more active occupations have lower rates of obesity (King et al., 2001; Steeves et al., 2012). Because LSBPA does expend calories, it is quite plausible that occupational LSBPA does oppose weight gain and partly explains this finding. Church et al. argue that a 5 decade decline in levels of daily occupational energy expenditure can explain a significant portion of the increase in the BMI of US adults (Church et al., 2011). Strengths and limitations
Discussion The purpose of this study was to examine whether Mexican Americans more frequently reported occupations with higher levels of physical activity, and to assess if the association between race/ethnicity and occupational physical activity was due to income rather than differences by race or ethnicity. As hypothesized, Mexican Americans are employed in more active occupations compared to their non-Hispanic peers. Furthermore, this association was independent of income as
There are several limitations to this study. NHANES did not separately assess accelerometer counts while at work, and this study did not directly observe levels of occupational activity. The sampling frame for the 2003–2004 NHANES did not include a sufficient number of Hispanic participants with a country affiliation other than Mexico, and the results may not generalize to other Hispanic populations. However, Mexican Americans do comprise the largest proportion of Hispanics in the U.S. Wear time compliance differed by age and gender (Troiano et al.,
Table 1 Sample characteristics, NHANES 2003–2004. Male White (n = 793) Age BMI Accelerometer Wear time (mean min/day) Sedentary time (min/week) LSBPA (min/week) MVPA (min/week)
Female Black (n = 234)
Mexican American (n = 305)
White (n = 824)
Black (n = 245)
Mexican American (n = 302)
47.40 (0.52) 28.11 (0.27)
43.09 (0.61) 28.57 (0.65)
38.36 (0.70) 28.14 (0.27)
48.88 (0.78) 27.61 (0.33)
43.82 (1.40) 31.45 (0.47)
39.71 (1.07) 29.00 (0.56)
861.11 (3.55) 459.51 (4.46) 393.13 (6.66) 8.47 (0.75)
905.45 (12.34) 454.70 (8.62) 440.43 (7.81) 10.32 (1.66)
856.45 (7.51) 352.77 (7.13) 492.35 (9.60) 11.33 (0.69)
843.35 (3.44) 470.24 (4.61) 366.82 (4.98) 6.29 (0.61)
852.15 (8.48) 469.79 (10.05) 376.38 (7.49) 5.98 (1.11)
822.40 (4.81) 410.36 (5.35) 406.27 (2.71) 5.77 (0.51)
Means (standard errors) presented. BMI, body mass index; LSBPA, MVPA, moderate-to-vigorous intensity physical activity; light-intensity + short bouts of MVPA.
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
61
Fig. 1. a. Proportion of wear time by occupational activity category for sedentary and LSBPA wear time. b. Proportion of wear time by occupational activity category for MVPA wear time, NHANES 2003–2004.
2008), as well as BMI and race/ethnicity. Although patterns of missing data may limit generalizability of the findings, the NHANES dataset presents the opportunity for large-scale accelerometer analysis from a
national sample. The income values used were household income, not individual income. While less likely at the lower end of the income spectrum, the associations do not represent individual occupation tied to
Table 2 Proportion of participants in Sedentary, Low Active and Moderately Active occupations by race/ethnicity and gender, NHANES 2003–2004. Occupational activity classification, %
Non-Workers Sedentary occupations Low Active occupations Moderately Active occupations
Malea
Femaleb
White (n = 793)
Black (n = 234)
Mexican American (n = 305)
White (n = 824)
Black (n = 245)
Mexican American (n = 302)
26.1 23.4 26.9 23.6
28.1 15.9 36.4 19.7
15.5 10.4 33.1 41.1
37.1 30.2 26.8 6.0
37.5 27.5 30.1 4.8
43.1 21.2 22.7 12.9
Note: column percentages shown may not add to 100 due to rounding. Weighted frequencies shown. a χ2 = 37.81, P b 0.0001. b χ2 = 12.59, P = 0.05.
62
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
References
Fig. 2. Interaction of race and ethnicity with household income on mean accelerometer activity counts, NHANES 2003–2004.
individual income. Future studies incorporating occupation and income should assess individual level income. Conclusion In summary, it is plausible that the higher level of objectively measured physical activity in Mexican American adults reported in NHANES is due to higher levels of occupational physical activity in Mexican Americans. Using a standard method of classifying occupation, Mexican American men and women were most likely to have active occupations. As hypothesized, there is a dose response relationship between job classification and total counts per minute of activity. The dose–response relationship appeared to be due more to higher levels of LSBPA with no compensation effect for MVPA. Higher levels of occupational activity in Mexican Americans appear mainly due to more active occupations in those with household incomes in the range of $10,000 to $65,000. This finding highlights the need for more evidence of the health benefits of light-intensity physical activity or MVPA that is too short in duration to meet guidelines. Although LSBPA may not be of a great enough intensity to increase cardiorespiratory fitness, energy expenditure throughout the day may better regulate glucose. Employment in more active occupations may be the reason for the higher activity counts documented among Mexican Americans (Troiano et al., 2008). For first-generation Mexican Americans occupational physical activity may provide a partial explanation for the favorable cardiovascular mortality disparity. To the extent that occupational activity and LSBPA provide health benefits, higher levels of occupational activity may partly explain the so-called Hispanic paradox. There is a documented favorable disparity for Mexican Americans compared to Non-Hispanic Whites in mortality from chronic diseases (Markides and Eschbach, 2005). Mexican Americans also have a favorable disparity for physical activity participation, engaging in more minutes of activity compared with Non-Hispanic Whites and Blacks (Troiano et al., 2008). The improved rates of chronic disease and mortality may be partially explained by the increased participation in physical activity, despite greater risk based on weight status and socioeconomic status (Gay et al., 2013; Wei et al., 1999). This interaction of ethnicity and occupational physical activity behavior may contribute to better understanding health disparities and the Hispanic Paradox. This study also calls attention to the issue of how much the sustained, long-term decline in occupational activity is contributing to the epidemic of chronic diseases. Additional research is necessary to more completely understand the contribution of occupational physical activity to health, including studies that directly measure occupational activity with accelerometers and randomized trials to increase LSBPA at work. Conflict of interest statement The authors declare that there are no conflicts of interest.
Arias, E., 2010. United States life tables by Hispanic origin. National Center for Health Statistics. Vital Health Stat. 2 (152). Carlson, S.A., Fulton, J.E., Galuska, D.A., Kruger, J., Lobelo, F., Loustalot, F.V., 2008. Prevalence of self-reported physically active adults — United States, 2007. MMWR 57, 1297–1300. Centers for Disease Control and Prevention (CDC), 2007. Behavioral Risk Factor Surveillance System Survey Data. Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention (CDC), 2011. Analytic Note Regarding 2007– 2010 Survey Design Changes and Combining Data Across Other Survey Cycles [Online]. Available: http://www.cdc.gov/nchs/data/nhanes/analyticnote_2007-2010.pdf (Accessed 8/30/2012). Church, T.S., Thomas, D.M., Tudor-Locke, C., et al., 2011. Trends over 5 decades in U.S. occupation-related physical activity and their associations with obesity. PLoS One 6, e19657. Clarke, P.J., O'Malley, P.M., Johnston, L.D., Schulenberg, J.E., Lantz, P., 2009. Differential trends in weight-related health behaviors among American young adults by gender, race/ethnicity, and socioeconomic status: 1984–2006. Am. J. Public Health 99, 1893–1901. Freedson, P.S., Melanson, E., Sirard, J., 1998. Calibration of the Computer Science and Applications, Inc. accelerometer. Med. Sci. Sports Exerc. 30, 777–781. Gay, J., Salinas, J., Buchner, D., et al., 2013. Meeting physical activity guidelines is associated with lower allostatic load and inflammation in Mexican Americans. J. Immigr. Minor. Health 1–8. Ham, S.A., Ainsworth, B.E., 2010. Disparities in data on Healthy People 2010 physical activity objectives collected by accelerometry and self-report. Am. J. Public Health 100 (Suppl. 1), S263–S268. Hawkins, M.S., Storti, K.L., Richardson, C.R., et al., 2009. Objectively measured physical activity of USA adults by sex, age, and racial/ethnic groups: a cross-sectional study. Int. J. Behav. Nutr. Phys. Act. 6, 31. Healy, G.N., Dunstan, D.W., Salmon, J., et al., 2007. Objectively measured light-intensity physical activity is independently associated with 2-h plasma glucose. Diabetes Care 30, 1384–1389. Hu, G., Tuomilehto, J., Borodulin, K., Jousilahti, P., 2007. The joint associations of occupational, commuting, and leisure-time physical activity, and the Framingham risk score on the 10-year risk of coronary heart disease. Eur. Heart J. 28, 492–498. Jones, J.I., 2009. What Do OES Data Have to Say About Increasing Wage Inequality? [Online]. Division of Publication and Analysis, Office of Employment and Unemployment Statistics, Bureau of Labor Statistics, U.S. Department of Labor (Available: http://www.bls. gov/opub/mlr/2009/06/art3full.pdf). Katzmarzyk, P.T., Lee, I.M., 2012. Sedentary behaviour and life expectancy in the USA: a cause-deleted life table analysis. BMJ Open 2. King, G.A., Fitzhugh, E.C., Bassett Jr., D.R., et al., 2001. Relationship of leisure-time physical activity and occupational activity to the prevalence of obesity. Int. J. Obes. Relat. Metab. Disord. 25, 606–612. Lee, I.M., Skerrett, P.J., 2001. Physical activity and all-cause mortality: what is the doseresponse relation? Med. Sci. Sports Exerc. 33, S459–S471. Lee, I.M., Rexrode, K.M., Cook, N.R., Manson, J.E., Buring, J.E., 2001. Physical activity and coronary heart disease in women: is “no pain, no gain” passe? JAMA 285, 1447–1454. Markides, K.S., Eschbach, K., 2005. Aging, migration, and mortality: current status of research on the Hispanic paradox. J. Gerontol. B Psychol. Sci. Soc. Sci. 60 (Spec No 2), 68–75. Neighbors, C.J., Marquez, D.X., Marcus, B.H., 2008. Leisure-time physical activity disparities among Hispanic subgroups in the United States. Am. J. Public Health 98, 1460–1464. Steeves, J.A., Bassett Jr., D.R., Thompson, D.L., Fitzhugh, E.C., 2012. Relationships of occupational and non-occupational physical activity to abdominal obesity. Int. J. Obes. (Lond.) 36, 100–106. Troiano, R.P., Berrigan, D., Dodd, K.W., Masse, L.C., Tilert, T., Mcdowell, M., 2008. Physical activity in the United States measured by accelerometer. Med. Sci. Sports Exerc. 40, 181–188. Tudor-Locke, C., Ainsworth, B.E., Washington, T.L., Troiano, R., 2011. Assigning metabolic equivalent values to the 2002 census occupational classification system. J. Phys. Act. Health 8, 581–586. U.S. Department of Health and Human Services, 2008. Physical Activity Guidelines Advisory Committee Report. ODPHP Publication No. U0036. U.S. Department of Health and Human Services, 2012. Leading Health Indicators [Online]. Available: http://www.healthypeople.gov/2020/LHI/default.aspx (Accessed 1/17/2013). U.S. Department of Labor, 2008. Fact Sheet #17A: Exemption for Executive, Administrative, Professional, Computer & Outside Sales Employees Under the Fair Labor Standards Act (FLSA) [Online]. U.S. Department of Labor, Washington, DC (Available: http://www.dol.gov/whd/regs/compliance/fairpay/fs17a_overview.htm). U.S. Department of Labor, 2011. An Overview of U.S. Occupational Employment and Wages [Online]. Available: http://www.bls.gov/oes/highlight_2010.pdf. van Uffelen, J.G.Z., Wong, J., Chau, J.Y., et al., 2010. Occupational sitting and health risks: a systematic review. Am. J. Prev. Med. 39, 379–388. Wei, M., Kampert, J.B., Barlow, C.E., et al., 1999. Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA 282, 1547–1553. Wolin, K.Y., Bennett, G.G., 2008. Interrelations of socioeconomic position and occupational and leisure-time physical activity in the National Health and Nutrition Examination Survey. J. Phys. Act. Health 5, 229–241.
Contents lists available at ScienceDirect
Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed
Ethnic disparities in objectively measured physical activity may be due to occupational activity Jennifer L. Gay a,⁎, David M. Buchner b a b
Department of Health Promotion and Behavior, University of Georgia, Athens, GA, USA Department of Kinesiology and Community Health, University of Illinois, Urbana–Champaign, IL, USA
a r t i c l e
i n f o
Available online 28 February 2014 Keywords: Exercise Health disparities Accelerometer Mexican-American
a b s t r a c t Objectives. We examined whether Mexican American adults report occupations that involve higher levels of objectively assessed physical activity compared with Non-Hispanic White and Black adults, and if the differences were independent of income. Methods. Data from the 2003–2004 National Health and Nutrition Examination Survey (NHANES; N = 2707) were analyzed in 2012–2013. An existing classification scheme was used to classify self-reported occupation as sedentary, low-active, or moderately active. From NHANES accelerometer data, proportion of wear time was stratified by intensity. Results. A dose–response relationship was found such that workers in more active occupations spent more time in light-intensity activity and less time engaged in sedentary activities. The findings did not suggest a compensation effect for moderate-to-vigorous intensity physical activity. Mexican American adults engaged in more activity than Non-Hispanic Black or White adults for incomes between $10,000 and $64,999. Conclusions. Mexican American adults may have higher total physical activity levels in NHANES because of occupational activity, particularly among lower income households. To the extent that light-intensity activity may provide health benefits, occupational activity may partly explain the Hispanic paradox. © 2014 Elsevier Inc. All rights reserved.
Introduction In Healthy People 2020 physical activity is a leading health indicator (U.S. Department of Health and Human Services, 2012). Like most other important health indicators, it demonstrates disparities in health by ethnic group. In nationally representative surveys of adults such as the Behavioral Risk Factor Surveillance System (BRFSS), self-reported levels of physical activity are consistently higher in Non-Hispanic Whites than in African-Americans and Hispanics (Carlson et al., 2008). Disparities in self-report activity persist between Non-Hispanic Whites and Hispanics after adjustment for socioeconomic status, though levels of activity vary considerably among Hispanic subgroups (Neighbors et al., 2008). However, the first nationally representative survey to use an objective measure of physical activity (accelerometers) reported a different pattern of behavior: Mexican American adults (≥ 20 years) had higher mean activity counts than Whites and African-Americans and more minutes per week of moderate-to-vigorous physical activity (Troiano et al., 2008).
⁎ Corresponding author at: Department of Health Promotion and Behavior, 311 Ramsey Center, University of Georgia, Athens, GA 30602, USA. Fax: +1 706 542 4956. E-mail address: [email protected] (J.L. Gay).
http://dx.doi.org/10.1016/j.ypmed.2014.02.015 0091-7435/© 2014 Elsevier Inc. All rights reserved.
The finding of higher levels of objectively measured physical activity in Mexican American adults is of great interest, as it may help explain why Hispanic Americans have a life expectancy advantage over Whites and African-Americans. The advantage is about 7.7 years over nonHispanic blacks and about 2.5 years over Non-Hispanic Whites (Arias, 2010). This advantage is supported by data that US Mexican Americans have lower rates of mortality from cardiovascular disease and certain forms of cancer, and has been referred to as the “Hispanic Paradox” (Markides and Eschbach, 2005). The advantage is regarded as paradoxical because lower socioeconomic status is associated with higher mortality, and on average Hispanics have lower socioeconomic status (SES) than the Non-Hispanic White population (CDC, 2011). Because higher levels of physical activity reduce risk of cardiovascular disease and cancer often independent of weight status (Gay et al., 2013; Wei et al., 1999), and mortality risk (Lee and Skerrett, 2001), understanding ethnic and racial differences in physical activity behavior may help explain the “Hispanic Paradox.” One possible explanation for higher levels of physical activity in Mexican American adults is that they have higher levels of occupational activity. Currently, BRFSS questions assess three domains via selfreport: recreational, transport, and household activity; the questions do not assess occupational activity (Centers for Disease Control and Prevention, 2007). The National Health and Nutrition Examination Survey (NHANES) accelerometer protocol asked participants to wear the
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
accelerometer for a week during all waking hours (Troiano et al., 2008). That is, the accelerometers measured all domains of activity, including occupational physical activity. Recent literature highlights the need to better understand the ethnic disparities in physical activity (Ham and Ainsworth, 2010), particularly in relation to socioeconomic class (Clarke et al., 2009). Therefore, the first aim of this study was to analyze NHANES data to determine if Mexican American adults report occupations that involve higher levels of physical activity. Occupations with higher levels of activity often involve manual labor (i.e., physical work that may be repetitious) (U.S. Department of Labor, 2008). Wages for these jobs are typically lower than white collar positions that are usually salary-based (Jones, 2009; U.S. Department of Labor, 2008, 2011). One possibility is that higher occupational activity has nothing to do with ethnicity and culture, and that any group with lower household incomes would have higher levels of occupational activity. But this does not consider households on a single income, or those with an adult with disability who cannot work. And on average Mexican Americans have lower household incomes than Whites (Centers for Disease Control and Prevention, 2007). Furthermore, African-Americans also have lower household incomes and yet are not as active as Mexican Americans in NHANES (Troiano et al., 2008). So the second study aim was to analyze total physical activity levels stratified by income, to see if the effect of ethnicity is independent of income. Reasoning that Mexican Americans may work in physically active occupations more often than non-Hispanic adults and most of these jobs should be lower paid, the study hypothesized that Mexican Americans with lower incomes should have higher total activity than Whites and Blacks. However, Mexican Americans with higher incomes might not be as physically active, as so few higher income jobs nowadays in the United States require much physical activity. Methods Study population NHANES assesses a variety of health conditions and behaviors among children and adults in the United States. NHANES employs a multi-stage probability sampling frame, and consists of interviews, laboratory tests and a physical examination. In the 2003–2004 wave of data collection, an objective measure of physical activity was included in the examination. During this year 10,122 adults and children participated in NHANES. Demographics Study participants were asked to self-report age and self-identify with racial and ethnic groups: Mexican American, Other Hispanic, Non-Hispanic White, Non-Hispanic Black, and Other Race including multi-racial. In concordance with the NHANES Analytic Notes (Centers for Disease Control and Prevention, 2011), participants reporting Hispanic ethnicity (n = 211) other than Mexican American were also excluded. To be consistent with other published NHANES accelerometer data, adults aged 20 years and older were included in the analyses. Household income was reported at the family level and was collapsed into 11 categories ranging from b $5000 to N$74,999. Participants who did not provide income information (n = 170) were excluded from the analyses. Physical activity The Actigraph AM-7164 accelerometer (formerly the CSA/MTI AM-7164; Fort Walton Beach, FL) was used to objectively measure physical activity behavior. Participants wore the monitor on the right hip. Accelerometers were programmed to count activity in one-minute epochs. Data were summarized into one observation per participant using adapted SAS code provided by the National Cancer Institute (http://riskfactor.cancer.gov/tools/nhanes_pam/). Participants were included if they had a minimum of 4 days with at least 10 h of valid wear time. The SAS program provided by the National Cancer Institute summarizes data by intensity for moderate- and vigorous-intensity physical activity (MVPA) using established cut points (Freedson et al., 1998). The program
59
also classifies episodes of MVPA into both 1-minute and 10-minute bouts, where 1-minute bouts include all minutes of MVPA and 10-minute bouts count only MVPA periods long enough to meet US physical activity guidelines (U.S. Department of Health and Human Services, 2008). The program was adapted so that light-intensity physical activity was distinguishable from sedentary time by counting the number of minutes N 100 counts/ minute but less than the moderate-intensity threshold. The abbreviation LSBPA signifies Light-intensity + Short Bouts of MVPA; physical activity that is either of light-intensity (1.5 to 3.0 METs) or is a bout of MVPA less than 10 min in length. LSBPA represents physical activity that is not counted toward meeting current US guidelines (U.S. Department of Health and Human Services, 2008). Occupation Participants self-reported occupation and type of business. NHANES staff recoded the occupations according to the U.S. Census Bureau's Census 2000 Indexes of Industry and Occupations. The recoded variable for occupation available for analysis includes 41 occupational categories ranging from Management Related Occupations to Teachers to Construction Laborers. Two raters associated occupations with MET values (Tudor-Locke et al., 2011) based on job functions. MET values were categorized as sedentary (b2 METs), low (2.0–2.9 METs) and moderate (3.0–5.9 METs). Discrepancies between raters were resolved by examining related job descriptions and discussion with the study PI until consensus was reached. Participants indicating that they were currently unemployed were classified as Non-Workers. Data analyses Participants were excluded from analyses for being younger than 20 years (n = 2755). Of the 4075 participants aged 20 years and older who wore the accelerometer, 987 were excluded for invalid wear time, 211 for reporting a race or ethnicity excluded according to the NHANES analytic guidelines (Centers for Disease Control and Prevention, 2011) and 170 for missing household income. The final sample was analyzed in 2012 and 2013 using the sampling weights for the accelerometer sample with at least 4 valid wear days. A total of 2707 participants were included. Sample characteristics were calculated for age in years, BMI, accelerometer wear time, minutes spent in sedentary time, LSBPA and MVPA stratified by gender. The proportion of accelerometer wear time spent in sedentary behavior, LSBPA and MVPA was also stratified by occupational activity classification (Non-Worker, Sedentary occupation, Low Active or Moderately Active occupation) and accounted for sampling weights. Proportion of wear time was used because even after setting minimum requirements for accelerometer wear time, minutes of activity was confounded by total minutes of wear time. Using the proportion of wear time accounts for this intra-individual variation. To address the first study aim examining whether Mexican Americans report occupations that involve higher levels of physical activity, the chi-square test of proportions for occupational classification was calculated by race/ethnicity, stratified by gender. Ordinary least squares linear regression analyses using PROC SURVEYREG were conducted to assess if Mexican Americans worked more often in physically active occupations, independent of income, compared with their nonHispanic peers. The regression model for mean accelerometer activity counts controlled for accelerometer wear time, age, gender, BMI, occupational activity classification, and whether the participant was born in the US or abroad. Household size was also included as a covariate because the income variable represented household rather than individual income. An interaction term was included for race/ethnicity with household income to address the second aim of whether the effect of ethnicity is independent of income. Analyses accounted for the weighting, stratification and clustering of the NHANES survey design.
Results Participant characteristics stratified by gender and race/ethnicity are shown in Table 1. Of participants reporting their country of birth, 17.4% were foreign-born, with nearly 70.1% reporting being born in Mexico. The sample (n = 2707) was 51.8% female with a mean age of 46.81 years (SE = 0.56). Of the participants reporting an occupation (n = 1484) 37.2% reported occupations that were sedentary,
60
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
40.5% reported Low Active and 22.3% Moderately Active occupations. In the complete sample, the mean minutes per week of LSBPA were 388.74 (SE = 3.39) and mean minutes per week of MVPA were 25.82 (SE = 0.90). There were significant differences in the age of employed (M = 43.85) versus non-working participants (M = 63.52, P b 0.0001). However, much of this difference may be attributed to 52.4% of the non-working participants being retired. No differences by BMI were detected between Workers and Non-Workers (P = 0.542). There was a dose–response trend for mean accelerometer counts per minute of activity by occupational classification (F = 118.02; P b 0.0001) such that participants in Moderately Active occupations (M = 423.89) recorded more activity counts than workers in Low Active occupations (M = 355.27), Sedentary occupations (M = 325.57) and participants who were unemployed (M = 240.56). Figs. 1a and b illustrate the proportion of accelerometer wear time spent in sedentary behavior, LSBPA and MVPA stratified by occupational activity classification (Non-Worker, Sedentary occupation, Low Active or Moderately Active occupation). There is a dose–response relationship whereby workers in increasingly active occupations spent more time in LSBPA (F = 88.63, P b 0.0001) and less time in sedentary behavior (F = 95.46, P b 0.0001). The data also showed a dose–response relationship for MVPA (F = 145.37, P b 0.0001), but it was not as steep. Results of the chi-square test of proportions for occupational classification by race/ethnicity and stratified by gender are shown in Table 2. Among both men and women, Mexican Americans reported Moderately Active occupations more frequently than their White or Black counterparts (P b 0.0001 for men and P = 0.05 for women). White and Black men and women reported more sedentary occupations compared to Mexican American adults. Furthermore, fewer foreignborn Mexican Americans were employed in sedentary jobs (8.8%) than their US-born peers (25.3%, P b 0.0001). Multiple regression analyses were conducted to examine the interaction of race and ethnicity with household income (Fig. 2). The interaction term for household income and race/ethnicity was significant (P b 0.0001). Inspection of Fig. 2 revealed the interaction effect was significant because Mexican American adults consistently recorded higher mean accelerometer activity counts than White or Black adults when household income is between $10,000 and $64,999. The regression model accounted for 37.0% of the variation in mean accelerometer activity counts.
Mexican Americans had higher mean activity counts even within the same income ranges as White and Black men and women. The higher mean activity counts were most evident for participants with incomes between $10,000 and $64,999. This study expands upon previous accelerometer data characterizations of NHANES participants' physical activity behavior (Hawkins et al., 2009; Troiano et al., 2008) by assessing the moderating effect of race and ethnicity on the association between income and physical activity. There was also a significant dose–response relationship for proportion of wear time spent in MVPA by occupational activity classification. This suggests that adults with more active occupations do not compensate by engaging in less MVPA than workers with sedentary or Low Active occupations. Another study of questionnaire assessments of physical activity in NHANES also reported that more occupational physical activity is associated with more (not less) recreational activity (Wolin and Bennett, 2008). Though Mexican Americans clearly have more counts per minute of activity, the extent that this activity provides substantial health benefits is not clear. Almost all of the difference among ethnic groups in counts per minute is due to differences in LSBPA. The health benefits of LSBPA are not yet proven, though there is growing evidence that more LBSPA (or roughly equivalently less sedentary time) provides health benefits. For example, in the Australian Diabetes, Obesity and Lifestyle study, men and women who engaged in activities at light-intensity had lower blood glucose compared to sedentary participants (Healy et al., 2007). Walking at a low intensity was linked to reduced risk of coronary heart disease in the Women's Health Study (Lee et al., 2001). It was recently estimated that reducing excessive sitting to less than 3 h per day would increase life expectancy by 2.0 years (Katzmarzyk and Lee, 2012). However, studies of occupational physical activity generally report that people in more active occupations have lower rates of chronic diseases such as cardiovascular disease and diabetes (Hu et al., 2007; van Uffelen et al., 2010). In particular, more active occupations have lower rates of obesity (King et al., 2001; Steeves et al., 2012). Because LSBPA does expend calories, it is quite plausible that occupational LSBPA does oppose weight gain and partly explains this finding. Church et al. argue that a 5 decade decline in levels of daily occupational energy expenditure can explain a significant portion of the increase in the BMI of US adults (Church et al., 2011). Strengths and limitations
Discussion The purpose of this study was to examine whether Mexican Americans more frequently reported occupations with higher levels of physical activity, and to assess if the association between race/ethnicity and occupational physical activity was due to income rather than differences by race or ethnicity. As hypothesized, Mexican Americans are employed in more active occupations compared to their non-Hispanic peers. Furthermore, this association was independent of income as
There are several limitations to this study. NHANES did not separately assess accelerometer counts while at work, and this study did not directly observe levels of occupational activity. The sampling frame for the 2003–2004 NHANES did not include a sufficient number of Hispanic participants with a country affiliation other than Mexico, and the results may not generalize to other Hispanic populations. However, Mexican Americans do comprise the largest proportion of Hispanics in the U.S. Wear time compliance differed by age and gender (Troiano et al.,
Table 1 Sample characteristics, NHANES 2003–2004. Male White (n = 793) Age BMI Accelerometer Wear time (mean min/day) Sedentary time (min/week) LSBPA (min/week) MVPA (min/week)
Female Black (n = 234)
Mexican American (n = 305)
White (n = 824)
Black (n = 245)
Mexican American (n = 302)
47.40 (0.52) 28.11 (0.27)
43.09 (0.61) 28.57 (0.65)
38.36 (0.70) 28.14 (0.27)
48.88 (0.78) 27.61 (0.33)
43.82 (1.40) 31.45 (0.47)
39.71 (1.07) 29.00 (0.56)
861.11 (3.55) 459.51 (4.46) 393.13 (6.66) 8.47 (0.75)
905.45 (12.34) 454.70 (8.62) 440.43 (7.81) 10.32 (1.66)
856.45 (7.51) 352.77 (7.13) 492.35 (9.60) 11.33 (0.69)
843.35 (3.44) 470.24 (4.61) 366.82 (4.98) 6.29 (0.61)
852.15 (8.48) 469.79 (10.05) 376.38 (7.49) 5.98 (1.11)
822.40 (4.81) 410.36 (5.35) 406.27 (2.71) 5.77 (0.51)
Means (standard errors) presented. BMI, body mass index; LSBPA, MVPA, moderate-to-vigorous intensity physical activity; light-intensity + short bouts of MVPA.
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
61
Fig. 1. a. Proportion of wear time by occupational activity category for sedentary and LSBPA wear time. b. Proportion of wear time by occupational activity category for MVPA wear time, NHANES 2003–2004.
2008), as well as BMI and race/ethnicity. Although patterns of missing data may limit generalizability of the findings, the NHANES dataset presents the opportunity for large-scale accelerometer analysis from a
national sample. The income values used were household income, not individual income. While less likely at the lower end of the income spectrum, the associations do not represent individual occupation tied to
Table 2 Proportion of participants in Sedentary, Low Active and Moderately Active occupations by race/ethnicity and gender, NHANES 2003–2004. Occupational activity classification, %
Non-Workers Sedentary occupations Low Active occupations Moderately Active occupations
Malea
Femaleb
White (n = 793)
Black (n = 234)
Mexican American (n = 305)
White (n = 824)
Black (n = 245)
Mexican American (n = 302)
26.1 23.4 26.9 23.6
28.1 15.9 36.4 19.7
15.5 10.4 33.1 41.1
37.1 30.2 26.8 6.0
37.5 27.5 30.1 4.8
43.1 21.2 22.7 12.9
Note: column percentages shown may not add to 100 due to rounding. Weighted frequencies shown. a χ2 = 37.81, P b 0.0001. b χ2 = 12.59, P = 0.05.
62
J.L. Gay, D.M. Buchner / Preventive Medicine 63 (2014) 58–62
References
Fig. 2. Interaction of race and ethnicity with household income on mean accelerometer activity counts, NHANES 2003–2004.
individual income. Future studies incorporating occupation and income should assess individual level income. Conclusion In summary, it is plausible that the higher level of objectively measured physical activity in Mexican American adults reported in NHANES is due to higher levels of occupational physical activity in Mexican Americans. Using a standard method of classifying occupation, Mexican American men and women were most likely to have active occupations. As hypothesized, there is a dose response relationship between job classification and total counts per minute of activity. The dose–response relationship appeared to be due more to higher levels of LSBPA with no compensation effect for MVPA. Higher levels of occupational activity in Mexican Americans appear mainly due to more active occupations in those with household incomes in the range of $10,000 to $65,000. This finding highlights the need for more evidence of the health benefits of light-intensity physical activity or MVPA that is too short in duration to meet guidelines. Although LSBPA may not be of a great enough intensity to increase cardiorespiratory fitness, energy expenditure throughout the day may better regulate glucose. Employment in more active occupations may be the reason for the higher activity counts documented among Mexican Americans (Troiano et al., 2008). For first-generation Mexican Americans occupational physical activity may provide a partial explanation for the favorable cardiovascular mortality disparity. To the extent that occupational activity and LSBPA provide health benefits, higher levels of occupational activity may partly explain the so-called Hispanic paradox. There is a documented favorable disparity for Mexican Americans compared to Non-Hispanic Whites in mortality from chronic diseases (Markides and Eschbach, 2005). Mexican Americans also have a favorable disparity for physical activity participation, engaging in more minutes of activity compared with Non-Hispanic Whites and Blacks (Troiano et al., 2008). The improved rates of chronic disease and mortality may be partially explained by the increased participation in physical activity, despite greater risk based on weight status and socioeconomic status (Gay et al., 2013; Wei et al., 1999). This interaction of ethnicity and occupational physical activity behavior may contribute to better understanding health disparities and the Hispanic Paradox. This study also calls attention to the issue of how much the sustained, long-term decline in occupational activity is contributing to the epidemic of chronic diseases. Additional research is necessary to more completely understand the contribution of occupational physical activity to health, including studies that directly measure occupational activity with accelerometers and randomized trials to increase LSBPA at work. Conflict of interest statement The authors declare that there are no conflicts of interest.
Arias, E., 2010. United States life tables by Hispanic origin. National Center for Health Statistics. Vital Health Stat. 2 (152). Carlson, S.A., Fulton, J.E., Galuska, D.A., Kruger, J., Lobelo, F., Loustalot, F.V., 2008. Prevalence of self-reported physically active adults — United States, 2007. MMWR 57, 1297–1300. Centers for Disease Control and Prevention (CDC), 2007. Behavioral Risk Factor Surveillance System Survey Data. Department of Health and Human Services, Centers for Disease Control and Prevention, Atlanta, GA. Centers for Disease Control and Prevention (CDC), 2011. Analytic Note Regarding 2007– 2010 Survey Design Changes and Combining Data Across Other Survey Cycles [Online]. Available: http://www.cdc.gov/nchs/data/nhanes/analyticnote_2007-2010.pdf (Accessed 8/30/2012). Church, T.S., Thomas, D.M., Tudor-Locke, C., et al., 2011. Trends over 5 decades in U.S. occupation-related physical activity and their associations with obesity. PLoS One 6, e19657. Clarke, P.J., O'Malley, P.M., Johnston, L.D., Schulenberg, J.E., Lantz, P., 2009. Differential trends in weight-related health behaviors among American young adults by gender, race/ethnicity, and socioeconomic status: 1984–2006. Am. J. Public Health 99, 1893–1901. Freedson, P.S., Melanson, E., Sirard, J., 1998. Calibration of the Computer Science and Applications, Inc. accelerometer. Med. Sci. Sports Exerc. 30, 777–781. Gay, J., Salinas, J., Buchner, D., et al., 2013. Meeting physical activity guidelines is associated with lower allostatic load and inflammation in Mexican Americans. J. Immigr. Minor. Health 1–8. Ham, S.A., Ainsworth, B.E., 2010. Disparities in data on Healthy People 2010 physical activity objectives collected by accelerometry and self-report. Am. J. Public Health 100 (Suppl. 1), S263–S268. Hawkins, M.S., Storti, K.L., Richardson, C.R., et al., 2009. Objectively measured physical activity of USA adults by sex, age, and racial/ethnic groups: a cross-sectional study. Int. J. Behav. Nutr. Phys. Act. 6, 31. Healy, G.N., Dunstan, D.W., Salmon, J., et al., 2007. Objectively measured light-intensity physical activity is independently associated with 2-h plasma glucose. Diabetes Care 30, 1384–1389. Hu, G., Tuomilehto, J., Borodulin, K., Jousilahti, P., 2007. The joint associations of occupational, commuting, and leisure-time physical activity, and the Framingham risk score on the 10-year risk of coronary heart disease. Eur. Heart J. 28, 492–498. Jones, J.I., 2009. What Do OES Data Have to Say About Increasing Wage Inequality? [Online]. Division of Publication and Analysis, Office of Employment and Unemployment Statistics, Bureau of Labor Statistics, U.S. Department of Labor (Available: http://www.bls. gov/opub/mlr/2009/06/art3full.pdf). Katzmarzyk, P.T., Lee, I.M., 2012. Sedentary behaviour and life expectancy in the USA: a cause-deleted life table analysis. BMJ Open 2. King, G.A., Fitzhugh, E.C., Bassett Jr., D.R., et al., 2001. Relationship of leisure-time physical activity and occupational activity to the prevalence of obesity. Int. J. Obes. Relat. Metab. Disord. 25, 606–612. Lee, I.M., Skerrett, P.J., 2001. Physical activity and all-cause mortality: what is the doseresponse relation? Med. Sci. Sports Exerc. 33, S459–S471. Lee, I.M., Rexrode, K.M., Cook, N.R., Manson, J.E., Buring, J.E., 2001. Physical activity and coronary heart disease in women: is “no pain, no gain” passe? JAMA 285, 1447–1454. Markides, K.S., Eschbach, K., 2005. Aging, migration, and mortality: current status of research on the Hispanic paradox. J. Gerontol. B Psychol. Sci. Soc. Sci. 60 (Spec No 2), 68–75. Neighbors, C.J., Marquez, D.X., Marcus, B.H., 2008. Leisure-time physical activity disparities among Hispanic subgroups in the United States. Am. J. Public Health 98, 1460–1464. Steeves, J.A., Bassett Jr., D.R., Thompson, D.L., Fitzhugh, E.C., 2012. Relationships of occupational and non-occupational physical activity to abdominal obesity. Int. J. Obes. (Lond.) 36, 100–106. Troiano, R.P., Berrigan, D., Dodd, K.W., Masse, L.C., Tilert, T., Mcdowell, M., 2008. Physical activity in the United States measured by accelerometer. Med. Sci. Sports Exerc. 40, 181–188. Tudor-Locke, C., Ainsworth, B.E., Washington, T.L., Troiano, R., 2011. Assigning metabolic equivalent values to the 2002 census occupational classification system. J. Phys. Act. Health 8, 581–586. U.S. Department of Health and Human Services, 2008. Physical Activity Guidelines Advisory Committee Report. ODPHP Publication No. U0036. U.S. Department of Health and Human Services, 2012. Leading Health Indicators [Online]. Available: http://www.healthypeople.gov/2020/LHI/default.aspx (Accessed 1/17/2013). U.S. Department of Labor, 2008. Fact Sheet #17A: Exemption for Executive, Administrative, Professional, Computer & Outside Sales Employees Under the Fair Labor Standards Act (FLSA) [Online]. U.S. Department of Labor, Washington, DC (Available: http://www.dol.gov/whd/regs/compliance/fairpay/fs17a_overview.htm). U.S. Department of Labor, 2011. An Overview of U.S. Occupational Employment and Wages [Online]. Available: http://www.bls.gov/oes/highlight_2010.pdf. van Uffelen, J.G.Z., Wong, J., Chau, J.Y., et al., 2010. Occupational sitting and health risks: a systematic review. Am. J. Prev. Med. 39, 379–388. Wei, M., Kampert, J.B., Barlow, C.E., et al., 1999. Relationship between low cardiorespiratory fitness and mortality in normal-weight, overweight, and obese men. JAMA 282, 1547–1553. Wolin, K.Y., Bennett, G.G., 2008. Interrelations of socioeconomic position and occupational and leisure-time physical activity in the National Health and Nutrition Examination Survey. J. Phys. Act. Health 5, 229–241.
