Hispanic adults’ physical activity and sedentary behavior profiles: examining existing data to drive prospective research

Abstract

Background

Because physical activity (PA) and sedentary behavior (SB) are two distinct, interdependent behaviors, increases in PA may influence SB. As a limited number of SB interventions have been tested in Latino/Hispanic samples,. The purpose of this study is to assess if there was more PA and less SB in a Hispanic border community where there had been a PA-targeted community-wide campaign since 2005.

Methods

This cross-sectional study included Hispanic adults (N = 676) sampled from both intervention and comparison Texas-Mexico border communities in 2010. Our dependent variable was four-categories based on meeting PA guidelines and excessive SB (≥540 mins/day) cut-points. We conducted adjusted multivariable analysis to assess the association of intervention group with the PA/SB groupings.

Results

In 2010, most adults were in the Low PA/Low SB group. Compared to the comparison group, the intervention group had 6.45 (p < 0.001) times the adjusted odds of being in the High PA/Low SB vs. Low PA/High SB group.

Conclusions

Five years into the campaign, more PA and less SB were more likely in the intervention community, indicating the association of some PA interventions with SB. PA-targeted interventions should capture effects on SB to expand the literature on effective SB interventions for Hispanic adults.

health promotion, individual behaviour, physical activity

Introduction

Physical activity (PA) has been recently reconceptualized as a multidimensional behavior,^1,2 whereby PA and sedentary behavior (SB) should be considered two distinct behaviors, although both part of an individual’s overall daily energy expenditure profile. More specifically, physical activity is characterized as bodily movement that results in energy expenditure,³ whereas SB involves a sitting or reclining posture, with energy expenditure ≤1.5 metabolic equivalents (METs) and is not equivalent to physical inactivity.⁴ Just as PA has been linked to many health benefits,^5,6 conversely, excessive SB has been independently associated with several negative health outcomes, even when controlling for PA^7,8 or when meeting moderate-to-vigorous intensity PA (MVPA) guidelines.^8,9 However, given that it is possible for increases in MVPA to result in increases in SB so that individuals maintain a consistent level of daily energy expenditure,¹⁰ it is important to ensure that interventions targeting increases in MVPA do not inadvertently result in higher levels of SB. Therefore, the two behaviors should be investigated together.

Using data from existing evidence-based PA interventions as a starting point to expand the limited evidence for SB interventions in adults is a novel approach.^11,12 Despite calls for interventions to target SB specifically and in addition to MVPA,¹³ simply measuring and evaluating the potential effect of existing PA interventions on SB¹⁴ could help expand the SB literature more rapidly and lead to additional insights on the relationship between the two behaviors within the course of an intervention.

Deciphering the PA/SB profile is particularly relevant to Hispanic/Latino (heretofore just Hispanic) populations as substantial documentation indicates that Hispanics obtain less leisure-time MVPA than non-Hispanic Whites.¹⁵ While there have been some PA interventions with Hispanic adults,¹⁶ to our knowledge, no SB intervention studies have been published with this population. Given the low levels of MVPA, paired with the growing Hispanic population in the U.S.,¹⁷ there is a need to specifically study the relationship between MVPA and SB in Hispanics, as well as identify strategies associated with reduced excessive SB in this population. Existing data on Hispanic populations where PA and SB have been measured can be used to spur new research questions and directions.

The purpose of this study was two-fold. Through a cross-sectional assessment, we examined if there was greater adherence to MVPA guidelines and less excessive SB in a Hispanic border community where there had been a 5-year PA-targeted community-wide campaign (an evidence-based PA intervention), as compared to a comparison community that had not received the community-wide campaign. Second, we provide a snapshot of the PA/SB profile of a Hispanic, Texas-Mexico border community without a PA intervention by examining data from a comparison community. While these two behaviors have been examined separately in the past, we have never looked at them concurrently and aim here to use existing data to answer new research questions that have emerged in the field.^18,19

Methods

Intervention

Tu Salud ¡Si Cuenta! Your Health Matters! (TSSC) campaign is an ongoing community-driven, community-wide campaign that began in 2005, that used various coordinated components to convey theory-based nutrition and PA messaging in the Lower Rio Grande Valley, on the Texas-Mexico border. The intervention core components included a mix of media messages via local television stations, radio, social media and newsletters. TSSC also incorporated individual level interventions that included risk factor screenings, visits to the home by community health workers (CHWs) who delivered an educational and behavioral change intervention, and free local exercise classes delivered by CHWs. Efforts were coordinated by a community advisory board, where members from local health, academic, community and religious organizations were represented.²⁰ The campaign messaging and CHW intervention relied on behavior change techniques from both the Social Cognitive Theory and Transtheoretical Model. Thus, content involved creating awareness of the benefits of PA, using role models to present their personal coping strategies to being more physically active, encouraging individuals to set goals and monitor their progress, and building confidence to be physically active in spite of barriers, among other activities and techniques.²¹ Because data for the region indicated low levels of meeting MVPA guidelines, the messages for the campaign were designed with precontemplator and contemplator populations in mind. Early process evaluation of the messages showed that the content provided practical tips for incorporating the desired health behaviors into daily living, including how to move from being sedentary towards meeting MVPA guidelines. Additional details on the intervention content and design have been reported elsewhere.²⁰

Design and data collection

While Brownsville served as the intervention community, Laredo, 200 miles away on the Texas-Mexico border, served as the comparison community. Laredo was selected because it had similar demographic characteristics to Brownsville, including the population size, percent Hispanic of the population and similar income characteristics. For this study, data was collected in 2010, approximately 5 years after the TSSC campaign began, on randomly drawn samples from both the intervention and comparison communities. The sampling frame was neighborhood areas in each city, matching on size, percent Hispanic, and low-income status. From this sampling frame, we drew a sample using a 1-in-10 systematic sampling of homes. Randomly selected homes were approached by data collectors, to identify one adult, over the age of 18, who spoke either Spanish or English to complete the interviewer-administered questionnaire. Prior to completion of the questionnaire, data collectors obtained written informed consent. Completed data relevant to this analysis resulted in n = 332 individuals in the intervention group and n = 344 individuals in the comparison group. This study was reviewed and approved by the Institutional Review Board. More extensive details on data collection have been reported elsewhere.^21,22

Measures

We collected covariates for our analysis, including sex, marital status, education, employment, and ethnicity variables. Although only a fraction of the intervention community was actually exposed to any of the TSSC campaign components,²² our independent variable was still the groups (communities). We used the leisure-time PA and time spent sitting sections of the International Physical Activity Questionnaire (IPAQ) long-form to collect frequency and duration of MVPA and SB over the previous 7 days, respectively.^23,24

We defined high PA as meeting U.S. PA guidelines, identified as ≥600 MET adjusted minutes of MVPA in the past 7 days.²⁵ Low PA was defined as <600 MET adjusted minutes of MVPA in the past 7 days. We defined high SB as ≥540 minutes and low SB as <540 minutes of sitting time in the past 7 days, based on a newly identified cutpoint.²⁶ These definitions resulted in four PA/SB groupings for our dependent variable: Low PA/High SB (Reference group), Low PA/Low SB, High PA/High SB, High PA/Low SB.

Analysis

We used Chi-squared analyses to assess the association of the independent variable (intervention or comparison group) with the four PA/SB groupings of the dependent variable. Then, we conducted multivariable analysis using generalized logistic regression models to produce ORs for the three categories. For these models, we generated the odds of being in each PA/SB outcome group vs. Low PA/High SB group, comparing the intervention and comparison groups. We adjusted for covariates known to be associated with PA, including age, sex, marital status, education, ethnicity, and employment, using SAS 9.4 (SAS Institute Inc., Cary, NC).

Results

The analytic sample for the study from data collected was n = 676 (Table 1). From this analytic sample, 86% were female, 98% were of Mexican-origin, 68% were married, 31% were employed, and only 24% had a high school education or higher. There were significant differences between the four PA/SB groups for sex and education. Had fewer females and more individuals who had completed a high school education or higher.

Table 1

Exposure to TSSC by combined group/exposure variable and PA/SB groups (N = 676)

Variable		Total	Low PA/high SB	Low PA/low SB	High PA/high SB	High PA/low SB	P-value
		N = 676	N = 106	N = 465	N = 13	N = 92
Group	Intervention	332 (49.11)	33 (9.94)	227 (68.37)	4 (1.20)	68 (20.48)	<0.001
	Comparison	344 (50.89)	73 (21.22)	238 (69.19)	9 (2.62)	24 (6.98)
Female		578 (85.63)	87 (82.08)	404 (87.07)	8 (61.54)	79 (85.87)	<0.05
Graduated grade 12/GED or higher		162 (24.00)	39 (36.79)	94 (20.26)	10 (76.92)	19 (20.65)	<0.001
Married		451 (67.82)	72 (69.90)	315 (68.48)	7 (63.64)	57 (62.64)	0.68
Employed		208 (30.86)	36 (33.96)	140 (30.24)	5 (38.46)	27 (29.35)	0.80
Mexican-origin		646 (98.03)	103 (98.10)	444 (98.23)	11 (100.0)	88 (96.70)	0.77

Note: TSSC, Tu Salud ¡Si Cuenta! Your Health Matters!; PA, physical activity; SB, sedentary behavior. Missing data: sex n = 1, education n = 1, marital status n = 11, employment n = 2, ethnicity n = 17. Data collected in Brownsville (Intervention) and Laredo (Comparison) on the Texas-Mexico border.

The mean total MET-adjusted minutes of MVPA in the past 7 days was 501 for the intervention and 174 for the comparison group, while the mean total number of minutes spent sitting in the past 7 days was 1914 (~4.5 hours/day) in the intervention and 2826 (~6.7 hours/day) for the comparison group (data not shown in Tables). We investigated the four combined PA/SB outcomes by group (Table 1). Most individuals in both intervention and comparison groups were in the Low PA/Low SB category, capturing a similar percentage of their respective samples (68% in Intervention and 69% in Comparison). Very few individuals in either group were in the High PA/High SB group. For the intervention group, there were more individuals (20%) in the High PA/Low SB group and less in the Low PA/High SB group (10%); the opposite pattern was seen in the comparison group, with more individuals in the Low PA/High SB group (21%) and fewer in the High PA/Low SB group (7%).

For our main analyses, we looked at the four combined PA/SB outcomes by group only (intervention vs. comparison). As compared to the comparison group, the intervention group had six times (adj. OR = 6.45; 95% CI = [3.35,12.43]) the odds of reporting High PA/Low SB than Low PA/High SB (Table 2). Furthermore, as compared to the comparison group, the intervention group had twice the odds (adj. OR = 2.10; 95% CI = [1.30,3.41]) of being in Low PA/Low SB than Low PA/High SB (Table 2). There were no significant differences seen between the groups for High PA/High SB.

Table 2

Adjusted^a odds ratio (OR)^b of denoted PA/SB outcome by intervention group and combined group/exposure variable

PA/SB outcome group	Low PA/high SB	Low PA/low SB	High PA/high SB	High PA/low SB
		Adjusted OR (95% CI)	Adjusted OR (95% CI)	Adjusted OR (95% CI)
Intervention vs. comparison	Ref	2.10 (1.30,3.41)**	1.76 (0.42,7.45)	6.45 (3.35,12.43)***

Note: PA, physical activity; SB, sedentary behavior; Ref, reference; CI, confidence interval. *P < 0.05, **P < 0.01, ***P < 0.001. ^aAdjusting for age, sex, marital status, education, ethnicity, and employment. ^bOdds of being in the denoted PA/SB outcome group vs. the Low PA/High SB group. Data collected in Brownsville (Intervention) and Laredo (Comparison) on the Texas-Mexico border.

Discussion

Main findings of this study

In this study, we aimed to examine a new and important research question, how SB and PA may be related to one another in Hispanic populations. We found that our PA-targeted intervention, TSSC, was associated with meeting MVPA guidelines as well as less excessive SB. When comparing the intervention group to the comparison, there were higher odds of being in the High PA/Low SB group as compared to the Low PA/High SB group, and higher odds of being in the Low PA/Low SB group as compared to the Low PA/High SB group. We also sought to provide a profile of PA/SB in Hispanics on the Texas-Mexico border through the examination of the comparison community. Most individuals in the comparison group (69%) were in the Low PA/Low SB group, indicating that most adults of Mexican-American descent on the Texas-Mexico border are generally not meeting MVPA guidelines, but they also do not have high levels of excessive SB.

What is already known on this topic

A profile of Low PA/Low SB has been previously found in the Lower Rio Grande Valley; Hispanics living on the Texas-Mexico border have been found to be less likely to meet MVPA guidelines than Hispanics nationally.²⁷ Furthermore, Hispanics in the U.S. report the lowest levels of sitting time compared to other racial/ethnic groups.²⁸ Our findings are consistent with these patterns. Our previous study found that TSSC was separately associated with higher odds of meeting MVPA guidelines and less excessive SB.²¹ However, from this previous study we were unable to deduce how TSSC was associated with both behaviors simultaneously, which could further elucidate the relation between the two behaviors in this sample. Lastly, a recent review concluded that PA-targeted interventions do not have beneficial effects on SB,¹¹ which is inconsistent with our findings indicating that less excessive SB was identified in the community where there had been a theory-driven, multi-component, coalition-driven, community-wide campaign targeting PA.

What this study adds

As mentioned, our previous study examined SB and PA separately, so it was unclear how TSSC was associated with both outcomes together (i.e., whether the intervention was associated with individuals meeting MVPA guidelines and performing less excessive SB, or only one or the other). Our current analyses are able to answer this question and indicate that there is an association with both behaviors, likely because of messaging specifically targeted at sedentary individuals to reallocate time spent in SB towards meeting PA guidelines. By examining profiles of both intervention and comparison groups, we observe that there are very few individuals in either group who meet High PA/High SB criteria. Therefore, in this group, with the typical pattern of Low PA/Low SB for this population, there may be independence of the two behaviors²⁹ or they may be weakly inversely related.³⁰ It is likely that there is a fair amount of light-intensity activity not currently being systematically captured by surveillance activities or intervention evaluations, which may explain what was found in this and other Hispanic samples. Given that even light intensity physical activity may be beneficial,³¹ there is a need to capture more of Hispanic individuals’ total energy expenditure profiles. There is also a need for more research in this area, particularly with designs that can elucidate causality, to determine if and how other evidence-based strategies and interventions tested for PA can be used to impact SB.³²

Our finding of an association between a community-wide campaign targeting PA with less excessive SB is interesting in light of recent reviews that found that PA-targeted interventions were not associated with less SB¹¹ or had small and inconsistent effects on SB.¹² Our study, which examined the cross-sectional association of a theory-driven, multi-component, coalition-drive, community-wide campaign with PA and SB, indicates that a well-developed PA-targeted intervention is associated with less excessive SB. However, further research could examine the association between PA and SB in the context of an intervention that compares intervention and control communities, utilizing both pre- and post-test data. Further studies should be conducted to clarify if other types of PA-targeted interventions with only some or none of the TSSC intervention characteristics (i.e., no theory, single-component, only individually-targeted, etc.) have an association with SB.

Limitations

There are several limitations to this study. We had a limited number of individuals in our High PA/High SB group, likely contributing to the insignificant adjusted ORs seen in Table 2. However, the models still converged and provided a valid estimate, so we retained this group. Furthermore, we used self-report data for SB and leisure-time MVPA, which can have issues with both over and under reporting. Future research should collect domain- and intensity- specific self-report measures for PA and SB, complementing it with accelerometer data, so that the latter can capture additional information on standing, sleep, and light-intensity PA that factor into energy expenditure during a 24-hour period. Our data is older, having been collected in 2010. However, lack of PA has been a persistent issue along the border,^27,33 in particular in the Lower Rio Grande Valley where this study took place,³⁴ and more recent data of PA along the border still suggest that interventions to address PA are necessary³⁵; therefore, these findings are still relevant to the population. Due to the initiation of TSSC based on community needs, and thus initially conducted with limited funds, we were not able to collect strong measures of PA and SB at baseline, and therefore, we are unable to assess change over the course of the TSSC campaign. Therefore, although individuals in the High PA/Low SB group may have reallocated time in SB to MVPA, this cannot be established because of the cross-sectional nature of the data. Furthermore, this limits our research to cross-sectional analyses and their inherent limitations, including the ability to determine the effect of the intervention. Future research should assess the interplay of these two behaviors within the context of PA-targeted, SB-targeted and combined PA and SB interventions.

Conclusion

In this study, we were able to use existing data to provide insights that can inform future novel research questions on PA and SB. Compared to a similar comparison community, meeting MVPA guidelines and lower excessive SB were more likely in a Hispanic border community where there was a PA-targeted intervention. Measuring SB within the context of PA-targeted interventions could be a step to more rapidly expand the sparse literature on effective SB interventions for adults. Furthermore, analyzing the two behaviors together can provide additional insights on the effect of these interventions and the relationship between these two energy balance variables.

Acknowledgments

The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.

Funding

This research was funded by National Cancer Institute/National Institutes of Health through the UTHealth School of Public Health Cancer Education and Career Development Program (R25CA57712) and The University of Texas MD Anderson Cancer Center Support Grant (CA016672), by the UTCO project from the University of Texas Medical Branch, the EXPORT Grant from the National Center on Minority Health and Health Disparities (P20 MD000170), by the National Center for Advancing Translational Sciences (UL1 TR000371), and by the Cancer Prevention and Research Institute of Texas (RP170259).