Factors Associated with Use of Recreational Facilities and Physical Activity Among Low-Income Latino Adults

Abstract

Introduction.

The purpose of this study was to identify longitudinal correlates of low-income, urban, Latino adults’ use of recreational facilities and engagement in physical activity (PA).

Methods.

This secondary data analysis is from a placed-based initiative; the parent study recruited a cohort of parents of kindergarteners from schools in intervention and control communities. Using a self-administered questionnaire, we collected baseline correlates and 1-year follow-up recreational facilities use and PA outcomes. We conducted ordinal logistic regression to assess correlates for recreational facilities use and PA outcomes.

Results.

Our sample of 273 Latino adults was mostly female and married or living with partners. Those who considered physical activity to be important had higher odds of using trails (adjusted OR=2.36) and of moderate-intensity PA (adjusted OR=2.68). Reported perception that the neighborhood is very safe to walk was associated with higher odds of using trails (adjusted OR=3.45, 95% CI 1.39-8.53) and parks (adjusted OR=2.92, 95% CI 1.19-7.16). Meeting PA guidelines at baseline was associated with higher odds of using parks, walking, and moderate- and vigorous-intensity PA. Quality of recreational facilities was directly associated with walking. Speaking English was associated with higher odds of engaging in moderate-intensity PA.

Conclusion.

Results from this study of urban, low-income Latinos indicate that promotion is needed not only for sustained use of recreational facilities among current users, but also for concerted efforts to reach less acculturated Latinos and make recreational facilities appealing to currently inactive adults. Both individual-level and neighborhood-level interventions are needed to build positive attitudes towards physical activity, while also enhancing neighborhood safety and the quality of neighborhood recreational facilities.

Introduction

Although many adults in the United States do not meet physical activity guidelines [1], Latino adults are even less likely than non-Hispanic Whites to meet them [2]. Previous research studies have shown differences in correlates of physical activity by racial/ethnic group [3, 4], and some have identified physical activity correlates for Latino adults [5, 6]. Most studies identified demographic and individual-level physical activity correlates in Latinos: age [7, 8], gender [9], income [10], and physical activity self-efficacy [6, 7]. Results are mixed for some community and environmental correlates. There is evidence that greater U.S. acculturation or the level of acclimatization to the culture and social norms of the U.S. is associated with increased physical activity in Latinos [8, 11–15], though there are opposite findings [9, 16]. Strong social bonds and mutual trust, referred to as community cohesion, predict physical activity in certain Latina groups [10, 16, 17], but not in other Latino samples [7, 18]. The association of perceived safety, crime, and other physical indicators of social disorder, termed neighborhood incivilities, with physical activity are also controversial [6, 8, 19, 20]. Furthermore, there have been few longitudinal studies with Latino samples[17]. Studying correlates of physical activity in Latino adults remains important in order to plan successful interventions to increase physical activity in this community.

Public recreational facilities, such as walking trails, parks, and recreational centers, could provide a venue for physical activity promotion in Latino communities. However, access to public recreational facilities alone does not ensure their use [19], signifying a need for additional usage-enhancing promotion [21]. Identifying the factors associated with use of recreational facilities among Latino adults could help researchers and practitioners understand which Latino subgroups to target and how to intervene to promote use of recreational facilities. The purpose of this study was to identify correlates of use of recreational facilities and of physical activity in low-income Latino adults.

Methods

Study design and data collection

This study was a secondary data analysis from the evaluation of the ongoing placed-based initiative Go! Austin, Vamos! Austin (GAVA). GAVA is an ongoing project in a predominantly Latino, low-income East Austin (Texas) community that uses community coalition building and a multi-level intervention to improve the built and food environments and the policy surrounding physical activity and nutrition. The neighborhood was selected based on both needs, including limited infrastructure for healthy food and physical activity, as well as assets to build on, including identifiable community leaders, previous community organizing, and relationships with governmental and other community organizations. Additional details on the development of this project and the planned evaluation are published elsewhere [22]. Briefly, as part of GAVA and its evaluation, study staff first recruited five intervention elementary schools within the community, which were selected based on their feeder pattern to the middle school present in the neighborhood (for long-term follow-up). Then, five control elementary schools that were demographically similar but outside the intervention community were recruited. Data were collected yearly from a cohort of parent-child dyads to monitor changes in psychosocial and behavioral variables. Study staff recruited parents of kindergarteners from the intervention and control schools. Parents eligible for GAVA’s cohort study had a child in kindergarten (in 2013) attending one of the intervention or the control schools, lived in the geographically predefined area, and were proficient in either Spanish or English.

All parents of children in kindergarten classes at the five intervention schools were sent letters inviting them to participate in the study. Interested parents came to the school for the assessments, where study staff measured their height and weight, after obtaining written consent. Parents then completed a self-administered survey, and each received $50 cash as compensation.

For this study, we used self-reported baseline data collected in 2013 for all covariates (intervention group, sex, age, and income) and correlates (meeting physical activity guidelines, language acculturation, attitude, community cohesion, neighborhood safety, quality, and neighborhood incivilities) and follow-up data 1 year later (2014) for the behavioral outcomes (use of trails, parks and recreational centers, walking, moderate- and vigorous-intensity activity). During the study period, GAVA activities primarily involved coalition building; thus, no intervention activities related to changes in neighborhood recreational facilities or direct PA promotion took place in the intervention communities. This project was reviewed and approved by the University of Texas Health Science Center at Houston Committee for Protection of Human Subjects [HSC-SPH-13-0108].

Measures

Parents of kindergarteners completed self-administered surveys at baseline (2013) and 1 year later (2014). Baseline data were used for covariates and correlates, and data from 1-year follow-up were used for the behavioral outcomes.

Covariates

The GAVA Parent questionnaire assessed basic demographics, including age, sex, and annual combined family income. At baseline, a self-administered version of the Modifiable Activity Questionnaire was used to measure physical activity [23]. Baseline physical activity was assessed by adding the total number of minutes per day in the last 7 days that participants reported doing one or more of 30 activities of moderate- or vigorous-intensity (with additional space to indicate other activities) and then calculating total minutes of moderate-to-vigorous intensity physical activity over the past 7 days.

Independent variables

Language acculturation was assessed by asking participants which language they normally spoke at home—only or mostly English, only or mostly Spanish, or both English and Spanish. Attitude towards physical activity was determined with the statement “I feel it is important for my family to be physically active”; individuals responded with one of the following options: Not at all, a little, somewhat, or a lot. Community cohesion was assessed with items from the Project on Human Development in Chicago Neighborhoods [24, 25]. Our data had good internal consistency for this scale (Cronbach’s alpha = 0.82), but only when the positively worded items were retained. Thus, the resulting scale consisted of 3 items averaged together, which included identifying if people in the neighborhood are willing to help each other, their neighborhood is close-knit, and people can be trusted. Response options were on a 5-point Likert scale from “Strongly disagree” to “Strongly agree” (range 0 to 4). Perceived safety of the neighborhood was assessed using one item adapted from a previous study on perceptions of neighborhood safety [26]. This item asked “In general, how safe is it for teens/adults to bike or walk in the neighborhood?” with individuals responding not at all safe, mostly safe, or very safe. Quality of the neighborhood’s public recreational facilities for physical activity was measured with two questions that assessed the condition of the neighborhood’s public recreational facilities and the quality of its sidewalks, streets, and open places. Response options were on a 4-point scale, ranging from “Poor” to “Excellent”. The two items were moderately to highly correlated (Pearson’s correlation coefficient-0.54) and were averaged (range 0 to 3). Neighborhood incivilities were determined by 6 questions from the Project on Human Development in Chicago Neighborhoods [24, 25] and assessed to what degree each of the following was an issue in the neighborhood: theft, assault, gangs, trash and litter, traffic, and drugs. Response options were on a 4-point scale ranging from “Not a problem” to “Serious Problem”. The six items were averaged (Cronbach’s alpha=0.92, range 0 to 3) to provide a measure of neighborhood incivilities.

Behavioral Outcomes

Use of recreational facilities was measured with three questions that assessed how often in the previous year the individual had used neighborhood trails, parks, or recreational centers for physical activity. The items were retained as three separate outcomes, with ordinal responses (<1 time monthly, 1-2 times monthly, 1-2 times weekly, or ≥3 times weekly). Physical activity was measured with three questions that captured how many times over the previous seven days the participant exercised or took part in, for at least 20 minutes, any 1) moderate or 2) vigorous physical activity, or 3) walking. The response options were ordinal (never, 1-2 times, 3-4 times, 5-6 times, ≥7 times).

Statistical Analysis

Descriptive statistics were used to summarize sociodemographic characteristics of the sample. We then conducted ordinal logistic regression for 6 separate outcomes measured at 1-year follow-up: use of parks, use of trails, use of recreational centers, walking, moderate physical activity, and vigorous physical activity. While we used the outcomes from 1-year follow-up data, we used predictors and covariates, including intervention group, sex, age, and income, from the baseline assessment to maintain temporality. We ensured that the assumption of proportional odds was not violated by using the approximate likelihood-ratio test of proportionality of odds and the Brant test. Given the large amount of missing behavioral outcomes data (23-30%), we conducted multiple imputation by chained equations with 20 iterations [27]. We controlled for intervention group, sex, age, and income and ran all analyses using Stata/SE 14.2 (College Station, TX).

Results

At baseline, staff recruited 316 parents, 152 to the intervention group and 164 to the control group. For the purposes of this analysis, we limited the data to Latino adults who completed baseline assessments (n=273), resulting in 133 individuals from the intervention group and 140 from the control group. The mean age was 32 years; most were female, had a annual combined family income ≤$20,000 per year, were born outside the U.S., and were obese (Table 1). About half spoke only or mostly Spanish. At baseline, about 58% reported meeting moderate-to-vigorous intensity physical activity guidelines, 82% agreed physical activity was very important, and 20% reported that their neighborhoods were very safe to bike or walk. At baseline, mean community cohesion was 2.46 (range 0-4) and mean quality of neighborhood recreational facilities was 1.43 (range 0-3), with higher numbers indicating more cohesion and higher quality. Mean neighborhood incivilities was 1.05 (range 0-3), with higher numbers indicating incivilities were more of a problem. At follow-up, 45% reported using parks ≥1 time a week, 48% reported using trails ≥1 time a week, and 27% reported using recreational centers ≥1 time a week; 39% of participants reported walking ≥3 times a week, 35% of participants reported moderate activity ≥3 times a week, and 35% reported vigorous activity ≥3 times a week (data not shown).

Table 1.

Sociodemographic characteristics of study at baseline (n=273)

Characteristics	N or Mean	Proportion or SD
Age in years (Mean, SD)	32.1	7.4
Gender
Male	26	9.5
Female	242	88.6
Income
< $10,000	80	29.3
$10,000-$20,000	94	34.4
>$20,000	84	30.8
Marital status
Married or living with a partner	182	66.7
Divorced, separated or widowed	37	13.6
Never married	43	15.8
Language acculturation
Only/mostly English	83	30.4
Only/mostly Spanish	132	48.4
Both English and Spanish	48	17.6
Nativity
Born in U.S.	109	39.9
Born in another country	153	56.0
BMI
Normal weight	43	15.8
Overweight	82	30.0
Obese	147	53.9
Meeting MVPA guidelines at baseline	158	57.9
Believes physical activity very important at baseline	225	82.4
Community cohesion, (Mean, SD)	2.5	1.0
Neighborhood safety
Not at all safe	66	24.2
Mostly safe	144	52.8
Very safe	55	20.2
Quality of neighborhood recreational facilities, (Mean, SD)	1.4	0.7
Neighborhood incivilities, (Mean, SD)	1.1	0.9

Note: missing data includes n=15 for age, n=5 for gender, n=15 for income, n=11 for marital status, n=10 for language acculturation, n=11 for nativity, n=1 for BMI, n=7 for attitude, n=13 for community cohesion, n=8 for safety, n=6 for quality, and n=14 for neighborhood incivilities

Usage of recreational facilities

The overall models for trail and park use were statistically significant, while the overall model for recreational center use was not (Table 2). Individuals who reported physical activity as being very important at baseline had double the odds of using trails over the previous month at follow-up than those reporting physical activity as not important. Those who perceived the neighborhood to be very safe for walking at baseline had triple the odds of using trails and parks over the previous month at follow-up than those who perceived the neighborhood to be unsafe. At baseline, those who met the moderate-to-vigorous physical activity guidelines had double the odds of using parks over the previous month than individuals not meeting those guidelines.

Table 2.

Multivariable analysis of factors associated with use of recreational facilities in a low-income, Latino sample

Independent Variable	Adjusted Odds Ratio (95% Confidence Interval)
	Trail Use	Park Use	Recreational Center Use
Intervention Group (Referent: Control)	1.05 (0.60, 1.85)	0.74 (0.43, 1.26)	0.98 (0.59, 1.64)
Age	0.99 (0.96, 1.02)	0.98 (0.95, 1.02)	1.01 (0.97, 1.05)
Male (Referent: Female)	0.84 (0.33, 2.15)	1.38 (0.47, 4.05)	1.23 (0.36, 4.22)
Income	0.81 (0.43, 1.51)	1.08 (0.58, 2.02)	0.82 (0.46, 1.47)
Meeting MVPA guidelines at baseline	1.34 (0.79-2.27)	2.12 (1.23-3.65)**	2.03 (1.18-3.51)*
Language acculturation (Referent: Primarily Spanish)	--	--	--
Mix of Spanish/English	0.55 (0.27-1.13)	1.02 (0.48-2.17)	1.29 (0.59-2.84)
Primarily English	0.77 (0.42-1.43)	0.60 (0.32-1.12)	0.97 (0.52-1.83)
Believes physical activity very important at baseline	2.36 (1.15-4.84)*	0.52 (0.24-1.01)	1.45 (0.67-3.11)
Community cohesion	0.95 (0.72-1.26)	0.98 (0.73-1.31)	1.14 (0.83-1.57)
Neighborhood safety (Referent: Not at all safe)	--	--	--
Mostly safe	1.08 (0.56-2.08)	1.42 (0.74-2.74)	1.17 (0.62-2.23)
Very safe	3.45 (1.39-8.53)**	2.92 (1.19-7.16)*	1.31 (0.55-3.17)
Quality of neighborhood recreational facilities	1.10 (0.69-1.75)	1.39 (0.88-2.19)	1.19 (0.75-1.89)
Neighborhood incivilities	1.04 (0.78-1.40)	1.04 (0.78-1.39)	0.82 (0.60-1.13)

Note:

^***p<.001,

^**p<.01,

^*p<.05;

MVPA, moderate-to-vigorous physical activity; correlates in bold are significant

Physical activity

Compared with individuals who did not meet the moderate-to-vigorous intensity physical activity guidelines at baseline, those who met the guidelines had higher odds of walking and undertaking moderate- and vigorous-intensity activity at follow-up (Table 3). Quality of neighborhood recreational facilities at baseline was associated with higher odds of walking over the previous seven days at follow-up. At baseline, being from an English-speaking household was associated with higher odds of engaging in moderate-intensity physical activity over the previous seven days at follow-up than those who spoke Spanish only or both Spanish and English. Those who considered physical activity to be important at baseline had nearly triple the odds of moderate-intensity activity over the previous seven days at follow-up than those who viewed physical activity as unimportant.

Table 3.

Multivariable analysis of factors associated with physical activity in a low-income, Latino sample

Independent Variable	Adjusted Odds Ratio (95% Confidence Interval)
	Walking	Moderate Activity	Vigorous Activity
Intervention Group (Referent: Control)	0.54 (0.32, 0.89)*	0.73 (0.42, 1.25)	0.97 (0.56, 1.69)
Age	1.01 (0.98, 1.04)	0.98 (0.95, 1.02)	0.98 (0.94, 1.01)
Male (Referent: Female)	0.67 (0.28, 1.59)	1.47 (0.56, 3.83)	1.94 (0.72, 5.19)
Income	0.97 (0.54, 1.74)	1.15 (0.61, 2.16)	0.86 (0.46, 1.61)
Meeting MVPA guidelines at baseline	1.79 (1.07-3.00)*	2.74 (1.58-4.73)***	3.00 (1.66-5.41)***
Language acculturation (Referent: Primarily Spanish)	--	--	--
Mix of Spanish/English	0.98 (0.48-1.97)	1.97 (0.92-4.21)	1.29(0.61-2.74)
Primarily English	1.80 (0.92-3.51)	3.28 (1.70-6.33)***	1.30 (0.66-2.55)
Believes physical activity very important at baseline	2.01 (0.96-4.24)	2.68 (1.21-5.91)*	1.72 (0.83-3.55)
Community cohesion	0.83 (0.62-1.11)	1.08 (0.82-1.44)	1.15 (0.86-1.53)
Neighborhood safety (Referent: Not at all safe)	--	--	--
Mostly safe	0.75 (0.37-1.52)	0.88 (0.43-1.79)	0.75 (0.38-1.48)
Very safe	1.48 (0.61-3.61)	1.57 (0.70-3.50)	1.66 (0.70-3.95)
Quality of neighborhood recreational facilities	1.61 (1.01-2.57)*	1.11 (0.72-1.73)	1.05 (0.69-1.61)
Neighborhood incivilities	1.04 (0.76-1.44)	0.85 (0.62-1.17)	0.88 (0.65-1.20)

Note:

^***p<.001,

^**p<.01,

^*p<.05;

MVPA, moderate-to-vigorous physical activity; correlates in bold are significant

Discussion

Several individual-level correlates were significantly associated with use of recreational facilities and behavioral outcomes. Language acculturation was significantly associated with moderate-intensity physical activity. This complements prior literature which found that more acculturated Latinos participate in more physical activity [5, 8, 11–15, 28]. Believing that physical activity is very important was significantly associated with use of trails and undertaking moderate-intensity physical activity. This finding is noteworthy, given that this construct has rarely been significantly correlated with physical activity [29, 30], though at least one other study among Latinos has demonstrated a significant association between a positive attitude toward physical activity and engaging in physical activity [31]. We found that past physical activity behavior was associated with current physical activity, consistent with previous research [5, 17], and it was associated with park use. This latter finding is illuminating, because as greater investment is made in infrastructure to promote physical activity through greater access and usage, it is not only important to ensure promotion of physical activity maintenance through the use of recreational facilities for those who are already active, but also attraction of inactive individuals. Future work should continue to investigate whether changes in access to parks and other public recreational facilities promote changes in physical activity among inactive individuals.

Certain perceived neighborhood-level variables were also associated with the outcomes. Quality of neighborhood recreational facilities was associated with walking; however, it was not associated with any use outcomes. This contrasts with a previous study among Latinos that identified quality of public recreational facilities being important for their use [32]. Community cohesion and neighborhood incivilities were not associated with any outcomes. Similar results were noted in other studies among Latino adults [7, 18, 33, 34], which may indicate that these specific factors do not play a major role in physical activity or use of recreational facilities. Nevertheless, perceived neighborhood safety was found to be associated with use of parks and trails in the previous month, although it was not associated with physical activity, consistent with prior studies on Latinos and across racial/ethnic groups [6, 8, 19, 20, 31, 33]. Further research is needed to examine the importance of neighborhood-level variables on physical activity in Latino communities.

Certain study limitations should be noted. Because these data were collected as part of a program evaluation of a multi-component study with multiple community partners, rather than a formal, controlled study, there were competing interests for items to include on the questionnaire. Therefore, several constructs were measured with single items. Additionally, other items, such as the physical activity instrument, were altered from a more comprehensive questionnaire at baseline (Modifiable Activity Questionnaire) [23] to a 3-item scale at follow-up in order to reduce respondent burden. Use of paper-and-pencil, self-administered questionnaires, which was the only practical way to collect data given the financial constraints of this program evaluation, likely contributed to substantial missing data, especially on behavioral outcomes. Participants may have found providing answers to the behavioral outcomes, such as physical activity, difficult without an interviewer to assist. For this study, we used multiple imputation to handle missing data and additional sensitivity analyses indicated that significant variables in those models were similar to those in the complete case analyses. As participants in this study were self-selected and may have had a more positive attitude towards physical activity (82% indicated that physical activity was very important at baseline), this may have influenced the results which indicated a significant association of thinking physical activity is important with both use of trails and moderate activity. Because this study was conducted in one region of Texas with a high proportion of Mexican Americans and consisted predominately of married women with elementary-age children, the results may not generalize to other regions of the country or other Latino subgroups.

Certain study strengths should be noted. Important factors associated with using recreational facilities and physical activity were examined in a seldom-studied demographic, i.e., low-income Latinos, who have some of the lowest levels of leisure-time physical activity in the U.S [2]. We were able to establish temporality between independent variables and the outcomes and thus add to the literature on longitudinal research in this population. Despite using intervention data from baseline and follow-up, no intervention activities related to physical activity or recreational facilities took place during this study period. The study results could contribute to intervention development intended to promote physical activity in green spaces, such as parks and trails, in urban, Latino neighborhoods, especially because the number of interventions promoting physical activity at public recreational facilities within Latino communities is limited.

Conclusions.

Results from this study of urban, low-income Latinos can help researchers and practitioners understand the types of individuals to target and how to intervene to promote use of neighborhood recreational facilities for physical activity. Our results indicate that Latino adults with high English language acculturation are more physically active, and those who are using neighborhood recreational facilities are already meeting physical activity guidelines. Promotional campaigns should ensure the promotion of neighborhood recreational facilities to sustain their use among current Latino users and encourage maintenance of physical activity in highly acculturated Latinos. However, concerted efforts are imperative to bring inactive Latinos to neighborhood recreational facilities to potentially promote physical activity. To ensure use of neighborhood recreational facilities and increase physical activity, interventions should include an individual-level component that emphasizes the importance of physical activity. Finally, neighborhood-level interventions are needed in Latino communities to improve actual (and perceived) neighborhood safety and maintain high quality recreational facilities, as both were associated with positive behavioral outcomes.