The beneficial effects of adhering to the Centers for Disease Control or the American College of Sports Medicine physical activity recommendations of accumulating 30 min or more of moderate-intensity physical activity on most, preferably all, days of the week (22) and to the updated American College of Sports Medicine or the American Heart Association guidelines specifying moderate-intensity aerobic (endurance) physical activity (10) are well established (36). Unfortunately, Latinos and non-Latino blacks (NLB) have significantly more individuals with largely preventable metabolic diseases that have been associated with sedentary lifestyles compared with non-Latino whites (NLW) (5,30,37). Multiple surveys examining leisure time physical activity (LTPA) have found that Latinos and NLB are markedly more sedentary than NLW (8,14). It is unclear whether these differences are due to social-cultural factors or to other factors such as lower levels of human capital (e.g., wealth, education, or neighborhood resources) (2,21). Furthermore, the number of ethnic or racial minorities in the United States continues to grow rapidly. For example, the U.S. Census Bureau projections indicate that by the year 2050, 24% of U.S. citizens will be Latino (35). In response, the U.S. Department of Health and Human Services set the increase of LTPA in ethnic minority groups as a major public health priority (33).
Although previous studies have demonstrated disparities in LTPA, few have examined disparities in occupational physical activity (OPA). One study found that ethnic or racial minorities have a greater likelihood of employment in physically demanding occupations (11). A strong relationship has not been documented between OPA and fitness (13,23,26), and research on the association between OPA and health has been limited and equivocal (20,27). The lack of demonstrated association is likely because OPA is often not done in bouts long enough to increase aerobic fitness or to achieve health benefits (i.e., bouts each lasting 10 or more minutes as specified in the American College of Sports Medicine or the American Heart Association guidelines).
Interestingly, until recently the extent to which OPA influences LTPA had been given little attention. It is commonly believed that manual laborers are less likely to be active when they are not working (16) on the basis of the assumption that people who engage in high levels of OPA may be too exhausted to be active during their leisure time (24). Similarly, it is possible that individuals participating in high levels of OPA carry the perception that they do not need to participate in LTPA because they do sufficient physical activity at work to get health benefits (9). Alternatively, it has been argued that people who are physically active in their occupations may be more active outside of work than the general population because of a training effect of greater fitness and high functional capacity resulting from their heavy labor (e.g., lifting, carrying, digging, sawing) (16).
To date, the research examining the effect of OPA on LTPA is equivocal, with studies producing varying results as to the direction and strength of the relationship. Some studies found that increased OPA had a negative relationship with LTPA (4,15,25), whereas others found that high OPA had a positive relationship with LTPA (16), and still others found no effect at all (11,18). Variations and limitations in measurement of key variables (e.g., OPA and LTPA), time span of data collection, sample sizes, and statistical analyses as well as samples representative of the population or use of non-U.S. samples have all contributed to discord in findings.
The purpose of this study was to determine levels of OPA and LTPA among employed adults and to examine the relationship between OPA and LTPA across racial or ethnic groups and Latino subgroups using data from the National Health Interview Survey (NHIS).
National Health Interview Survey (NHIS).
We combined data from four years (2000-2003) of the NHIS, an annual, nationally representative survey of the civilian noninstitutionalized population of the United States (3). The NHIS uses a stratified, multistage probability sampling procedure in which households are randomly selected within randomly selected geographical areas. NLB and Latinos are oversampled within the survey to allow stable estimates for these subgroups (3).
As a subcomponent of the family interview, an in-depth interview is conducted with a randomly selected adult member (age 18 yr and older) within the household. The adult subinterview focuses on health status and health risk behaviors. For the four years combined, 127,596 adults participated in the survey of 152,076 approached (response rate = 83.9%). After adjusting for households that were not available, the final adult response rate was 73.6%. Postsurvey sampling weights are added to the data 1) to adjust the sample to be representative of the population and 2) to account for the response rate. Although the full sample's data are used in computations to estimate variances, analyses in the current study are only on those individuals employed in the week before the NHIS interview.
Race or ethnicity.
Within the NHIS, ethnicity and race are self-reported by participants and recorded in accordance with the Office for Management and Budget 1997 directive for reporting in federal statistical documents (19). Individuals who self-reported as being Hispanic or Latino were then asked what group represented their Hispanic origin or ancestry and were categorized into the following groups: Puerto Rican, Mexican, Mexican American, Cuban or Cuban American, Dominican, Central or South American, other Latin American, other Spanish, and multiple Hispanic or Latino. For the current study, we analyzed data for individuals who self-identified as NLB, Hispanic or Latino, or NLW. Because one of our primary aims was to examine physical activity trends among Latino subgroups, we limited analyses to Latinos who reported a single country of origin from Latin American regions. Excluding individuals with origins outside Latin America or being of multiple origins led to a reduction of 4.1% of the Latino sample. Individuals who were employed in the prior week (N = 80,409) and met the racial or ethnic criteria were included in the analysis (N = 76,820).
Occupational physical activity.
Employment in the week before survey administration was used to determine employment status. Four raters categorized occupations using occupation descriptions provided with the NHIS and the compendium of physical activity (1) as a guide. Occupations were categorized as primarily sedentary (modal MET expenditure value of 3 or less) or primarily moderate or greater levels of physical activity (e.g., cleaning or construction trades). Initial agreement among independently coded occupations was modest: kappa = 0.54, Z = 8.48, P < 0.001. Raters discussed coding disagreements until consensus ratings were achieved for all occupations. Using the consensus codes, it was found that 14.1% of the employed population (95% confidence interval (CI) = 13.7%-14.5%) was in occupations that require moderate or greater levels of physical activity.
Leisure time physical activity.
Within the NHIS, questions about LTPA are asked in two intensity subcategories: light to moderate and vigorous. Participants are first asked about frequency of vigorous activities of a minimum duration of 10 min and then are asked for typical duration. Vigorous activity is defined as one that causes heavy sweating or large increases in breathing or heart rate. Participants are then asked the same frequency and duration questions regarding light to moderate activities. These are defined as activities that cause only light sweating or slight or moderate increase in breathing or heart rate.
For the current study, we defined a three-level physical activity variable: 1) no LTPA, 2) some LTPA but below recommended levels for health benefits (i.e., subthreshold), and 3) LTPA at recommended levels. We coded participants as meeting recommended levels of LTPA using algorithms consistent with recommendations by the American College of Sports Medicine or the American Heart Association (10) as well as the Healthy People 2010 objectives (34). Participants met recommendations if they participated in LTPA at least 5 d·wk−1 for a minimum of 30 accumulated minutes per day or participated in vigorous intensity LTPA at least 3 d·wk−1 for a minimum of 20 min·d−1.
To account for potential nonlinear trends in age, we entered age in discrete categories into the models. Proxy measures for socioeconomic status included education, employment status, and health insurance. Education was also defined in discrete categories: less than high school diploma, high school diploma or equivalent, some college, college, or advanced degree. Current work status was defined as whether the individual had paid employment in the previous week. Health insurance was defined in three categories: private or Medicare, Medicaid, and no health insurance. Family composition was defined by martial status (never married, separated or divorced or widowed, and married or living together), number of children (none, one child, and two or more children), and whether there was someone 65 yr or older living in the household.
Self-rated health was included using the five-item response general health question within the NHIS, which was redefined into two categories: health is fair or poor versus health is good or better. Physical limitations were measured as any difficulties with performing unaided activities, including physical (e.g., walking quarter mile, grasping small objects, moving large objects) and role functioning (e.g., go shopping or socializing). Psychological distress was defined as negative mood that interfered with functioning "some" or "a lot" during the past 30 d.
Behavioral risk factors.
Smoking status was categorized into three groups: never smoker (never smoked 100 cigarettes in lifetime), former smoker, and current smoker. Participating in weekly heavy drinking episodes was defined as consuming five or more beverages with alcohol in a single day at least 52 times during the past year.
We used two proxy measures for acculturation: language of interview and birthplace. The NHIS was translated into Spanish and administered by bilingual interviewers in areas with high concentrations of Latinos. Alternatively, with non-Spanish-speaking interviewers, the interview would have been administered with the assistance of bilingual family members. Language of interview has been used as a proxy measure for acculturation in other studies (17). We also used an indicator variable for individuals who were born outside the continental United States.
All analyses were conducted using survey estimation commands available in Stata 9.0 (Stata Corporation, College Station, TX) to account for the complex sampling design (28). The analytical plan followed four steps. First, LTPA prevalence, OPA prevalence, and sociodemographic characteristics were compared between NLW and NLB and Latinos or Latino subgroups. Second, LTPA prevalence, OPA prevalence, and sociodemographic characteristics were compared between NLW and each of the Latino subgroups. Next, we used multivariate-adjusted modeling to test whether disparities in LTPA prevalence between NLW and NLB and Latinos persisted after adjusting for occupational and sociodemographic differences. Post hoc examinations of differences in odds ratios between NLB and Latinos were conducted using the adjusted Wald test for statistical significance (12,29). For the multivariate-adjusted modeling, we used ordinal logistic regression with the three-level LTPA variable as the outcome (12). Parameter estimates from these models indicate odds of being in a more active level of LTPA.
LTPA, OPA, sociodemographic, and other variable differences
Table 1 presents LTPA levels, OPA levels, sociodemographics, and other study variables by racial or ethnic groups. Among employed individuals, NLB and Latinos had significantly higher percentages reporting no LTPA compared with NLW. Moreover, Latinos had the greatest proportion of individuals reporting no activity during leisure time, with Latina women having the least LTPA participation of all groups. Similarly, Latinos, and NLB had fewer individuals meeting physical activity guidelines compared with NLW. Furthermore, it was found that significantly more Latinos had physically active occupations compared with NLB and NLB compared with NLW, respectively. More than one quarter of all employed Latinos had a physically active occupation.
Latinos and NLB reported lower levels of education, more children, poorer self-rated health, fewer physical limitations, were less likely to be current or former smokers, and were more likely to either receive Medicaid or not have health insurance than NLW (Table 1).
Table 2 presents LTPA levels, OPA levels, sociodemographics, and other study variables by Latino subgroup. Among employed Latinos, Cubans and Dominicans were most likely to report no LTPA and were most likely to not meet the American College of Sports Medicine or the American Heart Association recommendations. More than half of all Mexicans reported no LTPA and did not meet the American College of Sports Medicine or American Heart Association recommendations, but Mexicans also had the greatest percentage of workers with a physically active occupation by a wide margin. Mexicans were the Latino subgroup with the lowest education levels and were the most likely to not have health insurance and to engage in heavy drinking and were among the least acculturated. Puerto Ricans had the greatest number of individuals with functional limitations and who currently smoke but were the most likely to have health insurance. Cubans were the oldest subgroup, least likely to be single, and more likely to speak Spanish only.
Ordered logistic regression modeling
Table 3 presents results from the regression models, including odds ratios comparing LTPA among NLB and Latinos to NLW. Among employed individuals, LTPA is positively associated with greater education, having weekly heavy drinking episodes, and being a former smoker but negatively associated with age, being a woman, having a partner, having children, having an elder in the house, having poor health, being a current smoker, and receiving Medicaid or not having health insurance. Interestingly, LTPA was not significantly associated with having a physically active occupation, having physical limitations, or distress interfering with functioning. In the post hoc analysis, the odds ratios for Latinos and NLB were significantly different from one another (F(1,339) = 59.0, P < 0.001), indicating that Latinos had lower levels of LTPA than NLB after adjustment for sociodemographic characteristics.
Table 4 presents LTPA odds ratios for the Latino subgroups compared with NLW after adjustment for the same sociodemographic covariates shown in Table 3 as well as proxy measures for acculturation (language of interview and place of birth). For clarity in presentation, other covariates are not shown in Table 4, but the associations were the same as shown in Table 3. Three observations emerge from the data. First, OPA was not associated with LTPA. Second, the acculturation proxies were inversely associated with LTPA. Third, all Latino subgroups had lower levels of LTPA compared with NLW after adjusting for sociodemographic and acculturation measures. In post hoc analysis, only the differences in odds ratios between Cubans and Central or South Americans remained significant after a Bonferroni adjustment for the multiple comparisons (F(1,338) = 10.2, P < 0.01).
This study examined the prevalence of LTPA among employed NLW, NLB, and Latinos. Previous studies have shown that sociodemographic variables have an influence on LTPA (7,32). Despite controlling for sociodemographic factors, Latinos and NLB reported lower levels of LTPA participation than NLW. The findings suggest that factors other than the sociodemographic covariates we examined may be unique to racial or ethnic minorities and are affecting their physical activity levels (e.g., cultural factors). We also found differences in participation in LTPA and of meeting recommended LTPA levels among Latino subgroups. Cubans, Dominicans, and Mexicans had the fewest number of individuals participating in LTPA and were the least likely to meet the American College of Sports Medicine or the American Heart Association (10) guidelines for LTPA at levels that confer health benefits. Conversely, Mexican Americans had the greatest number of individuals participating in LTPA among the Latino subgroups. Importantly, less than 35% of Mexican Americans had individuals meeting physical activity recommendations.
Racial or ethnic minorities in the current study were found to have more individuals with physically active occupations than NLW. Notably, almost 30% of employed Latinos have a physically active occupation. However, this study does not support the contention that lower prevalence of LTPA is due to being active at work.
Our results are consistent with those of Troiano et al. (31) in that low levels of LTPA among Latinos might mistakenly lead one to believe that Latinos are inactive. Research demonstrates that is not the case. Prior studies have found that Latinos engage in less LTPA than NLW (8,14) but that Latinos report more OPA than NLW (11). It is plausible that when overall physical activity is objectively measured with accelerometers, the differences in LTPA among groups are offset by the increased OPA being performed by Latinos. In fact, Troiano et al. (31) examined accelerometer data from the National Health and Nutrition Examination Survey and found higher levels of overall physical activity among Mexican American (Latino) men and women than NLW or NLB men and women despite self-reporting lower levels of LTPA. Our study, which included a broader cross-section of Latino subgroups, found that Latinos had occupations with greater activity intensity. However, as noted previously, the evidence is weak as to whether physically demanding occupations confer health benefits.
We also found that there is wide variability in sociodemographic factors and perceived health among Latino subgroups identified by their country of origin. Although all Latino subgroups had lower levels of LTPA and higher OPA than NLW, there were significant differences in LTPA and OPA prevalence between the Latino subgroups. For example, significantly more Mexicans had physically active occupations compared with the other Latino subgroups.
The current study adds to recent studies with national surveillance data, which suggest that OPA participation does not have a significant impact on LTPA participation (11,18). Given these findings, further research is needed on factors that are influencing lower levels of LTPA participation among racial or ethnic minorities. The present study adds to the existing understanding of disparities in LTPA by considering many potential confounders-specifically OPA-and by examining physical activity across Latino subgroups. The strengths of the study also include using recent, nationally representative data and statistical methods for simultaneously examining physical activity across the spectrum of activity: no LTPA, some LTPA, and meeting the American College of Sports Medicine or the American Heart Association (10) guidelines for health benefits.
The strengths of the study need to be considered in light of some limitations. We attempted to be comprehensive in including sociodemographic covariates, but we could not include environmental access for physical activity, and thus the included covariates are not an exhaustive list of potential confounds. Further research is needed to determine which neighborhood factors and other social determinants may account for the differences. An additional limitation is that the weighting of occupations by intensity is not an exact science. However, our methods were extensive and rigorous, and there is precedence for weighting occupations by intensity using nationally representative data (6).
In sum, the current study demonstrated that employed Latinos and NLB had less participation in LTPA than NLW but greater participation in OPA. Differences in LTPA and OPA participation were noted among Latino subgroups. However, OPA was not found to have an impact on participation in LTPA for any racial or ethnic group studied.
Special thanks are extended to Jane Wheeler and Jennifer Kraemer for their assistance with the study. This study did not receive any outside funding for this work.
Conflict of interest: There are no professional relationships with companies or manufacturers who will benefit from the results of the present study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc
. 2000;32(suppl 9):S498-516.
2. Barnes P, Schoenborn C. Physical Activity among Adults: United States, 2000
(Publication No. 333). Hyattsville (MD): National Center for Health Statistics; 2003.
3. Botman SL, Moore TF, Moriarity CL, Parsons VL. Design and estimation for the National Health Interview Survey, 1995-2004. Vital Health Stat
4. Caban-Martinez AJ, Lee DJ, Fleming LE, et al. Leisure-time physical activity levels of the US workforce. Prev Med
5. Carter JS, Pugh JA, Monterrosa A. Non-insulin-dependent diabetes mellitus in minorities in the United States. Ann Intern Med
6. Coogan PF, Aschengrau A. Occupational physical activity and breast cancer risk in the upper Cape Cod cancer incidence study. Am J Ind Med
7. Crespo CJ, Ainsworth BE, Keteyian SJ, Heath GW, Smit E. Prevalence of physical inactivity and its relation to social class in U.S. adults: results from the Third National Health and Nutrition Examination Survey, 1988-1994. Med Sci Sports Exerc
8. Crespo CJ, Smit E, Andersen RE, Carter-Pokras O, Ainsworth BE. Race/ethnicity, social class, and their relation to physical inactivity during leisure time: results from the Third National Health and Nutrition Examination Survey, 1988-1994. Am J Prev Med
9. Cromwell SL, Berg JA. Lifelong physical activity patterns of sedentary Mexican American women. Geriatr Nurs
10. Haskell WL, Lee IM, Pate RR, et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc
11. He XZ, Baker DW. Differences in leisure-time, household, and work-related physical activity by race, ethnicity, and education. J Gen Intern Med
12. Hosmer DW, Lemeshow S. Applied Logistic Regression
. New York: John Wiley & Sons; 2000. p. 373.
13. Ilmarinen J, Louhevaara V, Korhonen O, Nygard CH, Hakola T, Suvanto S. Changes in maximal cardiorespiratory capacity among aging municipal employees. Scand J Work Environ Health
. 1991;17(suppl 1):99-109.
14. Jones DA, Ainsworth BE, Croft JB, Macera CA, Lloyd EE, Yusuf HR. Moderate leisure-time physical activity: who is meeting the public health recommendations? A national cross-sectional study. Arch Fam Med
15. Kaleta D, Jegier A. Occupational energy expenditure and leisure-time physical activity. Int J Occup Med Environ Health
16. Kruger J, Yore MM, Ainsworth BE, Macera CA. Is participation in occupational physical activity associated with lifestyle physical activity levels? J Occup Environ Med
17. Marín G, Marín BV. Research with Hispanic Populations
. Newbury Park (CA): Sage Publications; 1991. p. 130.
18. Marshall SJ, Jones DA, Ainsworth BE, Reis JP, Levy SS, Macera CA. Race/ethnicity, social class, and leisure-time physical inactivity. Med Sci Sports Exerc
19. Office of Management and Budget. Revisions to the standards for the classification of federal data on race and ethnicity. Fed Regist
20. Paffenbarger RS, Hale WE, Brand RJ, Hyde RT. Work-energy level, personal characteristics, and fatal heart attack: a birth-cohort effect. Am J Epidemiol
21. Parks SE, Housemann RA, Brownson RC. Differential correlates of physical activity in urban and rural adults of various socioeconomic backgrounds in the United States. J Epidemiol Community Health
22. Pate RR, Pratt M, Blair SN, et al. Physical activity and public health. A recommendation from the Centers for Disease Control and Prevention and the American College of Sports Medicine. JAMA
23. Rantanen T, Sipila S, Suominen H. Muscle strength and history of heavy manual work among elderly trained women and randomly chosen sample population. Eur J Appl Physiol Occup Physiol
24. Ruzic L, Heimer S, Misigoj-Durakovic M, Matkovic BR. Increased occupational physical activity does not improve physical fitness. Occup Environ Med
25. Salmon J, Owen N, Bauman A, Schmitz MK, Booth M. Leisure-time, occupational, and household physical activity among professional, skilled, and less-skilled workers and homemakers. Prev Med
26. Sobolski JC, Kolesar JJ, Kornitzer MD, et al. Physical fitness does not reflect physical activity patterns in middle-aged workers. Med Sci Sports Exerc
27. Sofi F, Capalbo A, Marcucci R, et al. Leisure time but not occupational physical activity significantly affects cardiovascular risk factors in an adult population. Eur J Clin Invest
28. StataCorp. Release 9 User's Guide
. College Station (TX): Stata Corporation; 2005. p. 370.
29. StataCorp. Stata Survey Data Reference Manual Release 9
. College Station (TX): Stata Corporation; 2005. p. 278.
30. Sundquist J, Winkleby MA. Cardiovascular risk factors in Mexican American adults: a transcultural analysis of NHANES III, 1988-1994. Am J Public Health
31. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc
32. Trost SG, Owen N, Bauman AE, Sallis JF, Brown W. Correlates of adults' participation in physical activity: review and update. Med Sci Sports Exerc
33. U.S. Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health
. Washington (DC): U.S. Department of Health and Human Services; 2000. p. 668.
34. U.S. Department of Health and Human Services. Tracking Healthy People 2010
. Washington (DC): U.S. Department of Health and Human Services; 2000. p. 996.
36. Warburton DE, Nicol CW, Bredin SS. Health benefits of physical activity: the evidence. CMAJ
37. Wei M, Valdez RA, Mitchell BD, Haffner SM, Stern MP, Hazuda HP. Migration status, socioeconomic status, and mortality rates in Mexican Americans and non-Hispanic whites: the San Antonio Heart Study. Ann Epidemiol