Social Environmental Correlates of Health Behaviors in a Faith-Based Policy and Environmental Change Intervention

Abstract

Diet and physical activity are behavioral risk factors for many chronic diseases, which are among the most common health conditions in the United States. Yet most Americans fall short of meeting established dietary and physical activity guidelines. Faith-based organizations as settings for health promotion interventions can affect members at multiple levels of the social ecological model. The present study investigated whether change in the church social environment was associated with healthier behavior at church and in general at 1-year follow-up. Six churches received mini-grants and technical assistance for 1 year to support policy and environmental changes for healthy eating (HE) and physical activity (PA). Socioenvironmental (social support and social norms) and behavioral (HE and PA at church and in general) outcomes were derived from baseline and 1-year follow-up church member surveys (n = 258). Three of six churches demonstrated significant improvements in all three socioenvironmental aspects of HE. Two of five churches exhibited significant socioenvironmental improvements for PA at follow-up. Church social environmental changes were related to health behaviors at church and in general (p < .05). Change in social support for HE, social support for PA, and social norms for PA were each associated with three church-based and general behavioral outcomes. Social norms for healthy eating were related to two general behavior outcomes and social norms for unhealthy eating to one general behavioral outcome. Study findings demonstrate that socioenvironmental characteristics are essential to multilevel interventions and merit consideration in designing policy and environmental change interventions.

Keywords

faith-based setting healthy eating physical activity policy and environment change social environment

Chronic diseases are among the most common health conditions in the United States, and the leading causes of death and disability (Centers for Disease Control and Prevention [CDC], 2015). Approximately half of all adults in the United States suffer from one or more chronic diseases or conditions (Ward, Schiller, & Goodman, 2014). Most major chronic diseases, including heart disease, stroke, cancer, and diabetes, are preventable and linked to poor diet and physical inactivity (CDC, 2015; U.S. Department of Agriculture & U.S. Department of Health and Human Services, 2010; U.S. Department of Health and Human Services, 2008). Multiple large prospective studies have found that normal weight, dietary patterns, and physical activity (PA) are strongly associated with reduced chronic disease risk (Chiuve et al., 2012; Colditz, Philpott, & Hankinson, 2016). Yet, in 2013, only 13.1% and 8.9% of U.S. adults consumed recommended daily amounts of fruits and vegetables (FV), respectively, per day (Moore & Thompson, 2015). Additionally, from 2005 to 2015, only 20% of adults met established guidelines for aerobic and muscle-strengthening activities (CDC, 2015; Johnson et al., 2014; U.S. Department of Health and Human Services, 2008).

Health behavior change is best achieved when policies and environments support healthy choices, when social norms and social support reinforce healthy choices and health-supportive behaviors, and when individuals are knowledgeable and motivated to make healthy choices (Frieden, 2010; Glanz, Rimer, & Viswanath, 2008). Faith-based organizations (FBOs) as settings for health promotion interventions have potential to affect members at multiple levels of the social ecological model (Bowen, Barrington, & Beresford, 2015; Campbell et al., 2007; Holt, Clark, Debnam, & Roth, 2014; Kegler et al., 2012). The model suggests interventions address several levels of influence of health behavior change. Additionally, interventions aimed at higher levels of the model can influence levels below (McLeroy, Bibeau, Steckler, & Glanz, 1988; Sallis, Owen, & Fisher, 2008). Often, FBOs implement policy and environmental change strategies because such approaches are intended to ultimately affect members’ behaviors (Arriola et al., 2017).

Health promotion programs are increasingly common in FBOs, especially African American churches (Campbell et al., 2007; Holt et al., 2014), yet little work has explored the relationship between church social environments and behavior within and outside of the church setting. The Faith, Activity, and Nutrition program investigators explored predictors and mediators of change in FV consumption and PA in an intervention implemented in African American churches in South Carolina. Results suggested an influence of social support from family, friends, coworkers, and fellow churchgoers for FV consumption (Condrasky, Baruth, Wilcox, & Carter, 2013) and PA (Baruth & Wilcox, 2015). Other investigators observed a direct effect of social support on FV consumption in a Body & Soul trial (Fuemmeler et al., 2006) and the Delta Body & Soul Effectiveness Trial (Thomson, Zoellner, & Tussing-Humphreys, 2014). Shaikh, Vinokur, Yaroch, Williams, and Resnicow (2011) found no direct relationship between social support and FV intake in the Healthy Body Healthy Spirit Trial. Studies exploring descriptive social norms related to healthy eating (HE) and PA in FBOs are lacking. However, there is evidence that social norms for HE (e.g., observing others eating healthy foods) are associated with healthier eating and that social norms for PA (e.g., having active friends, observing neighbors being active) can predict PA behaviors (Ball, Jeffery, Abbott, McNaughton, & Crawford, 2010; Firestone, Yi, Bartley, & Eisenhower, 2015).

This article presents findings from an outcome evaluation of a mini-grants program wherein six South Georgia churches implemented policy and environmental changes to support HE and PA. The purpose of the present study was to investigate whether member perceptions of church social environments for HE and PA changed from baseline to 1-year follow-up, and the extent to which perceived changes in church social environments for HE and PA were associated with healthier behavior at church and in general at 1-year follow-up.

Method

The “Prevention Strategies That Work” Mini-Grants Program

From 2012 to 2014, the Emory Prevention Research Center provided mini-grant funds and technical assistance to six rural South Georgia churches. Grantees selected organizational policy and environmental change strategies to promote HE and PA. Although not a focus of this study, sites could also choose a tobacco use prevention strategy. Grantees chose strategies from a list of evidence-based options (Figure 2; CDC, 2013; The Community Guide to Preventive Services, 2013). Grantees implemented as many or as few as they wanted, tailoring strategies for their site. Details about grantee selection and technical assistance methods for the program are described elsewhere (Arriola et al., 2016).

This program’s conceptual model illustrates hypothesized relationships between FBO environments and behavior (Figure 1). Pertinent to this study, model inputs and activities lead to FBOs creating policy/environmental changes in the FBO, which also promote change at interpersonal and individual levels, that is, perceived social support and social norms for healthy behaviors, and healthy behaviors at church and in general. The combined effect of these changes may contribute to a healthier community context in tandem with similar changes in other settings and lead to improved health behaviors and reduced incidence and prevalence of chronic disease. The analyses described here explored the relationship between selected social factors and related health behaviors, bolded in Figure 1, among FBO members.

Figure 1.

Conceptual model for promoting policy and environmental change in faith settings.

Figure 2.

Prevention strategies that work mini-grants policy and environmental change strategies (healthy eating and physical activity options).

Evaluation Design

The outcome evaluation used a pre–post single group design, with a convenience sample of participants completing baseline and 1-year follow-up surveys. This evaluation was exempt from Emory University’s Institutional Review Board review.

Population and Sample

At baseline, eligible participants were at least 18 years old and attended the grantee church at least monthly. Only one adult per household completed the survey.

Data Collection

Baseline surveys were self-administered by participants at each church site after Sunday service in 2013 (n = 319); follow-up surveys occurred approximately 1 year from baseline. Most follow-up surveys were collected in-person at church sites (n = 222, 86%); surveys were also collected by telephone (n = 26, 10%) and mail (n = 10, 4%) to reach participants unable to participate in-person. Participants received a gift card for completing each survey ($10 at baseline, $15 at follow-up).

Instrument and Measures

Data Collection Instrument

Surveys included a demographics module and, if churches chose to implement related strategies, HE and/or PA modules. Modules assessed church social environments (descriptive social norms and perceived social support for HE or PA), participants’ eating behavior or intention to be physically active while at church, and general eating behavior or physical activity level (Table 1).

Table 1.

Survey Measure Descriptives.

	Baseline			Follow-up
Variables	Mean (SD)	Range (actual)	Cronbach α	Mean (SD)	Range (actual)	Cronbach α
Church social environment factors
Social support for HE	16.86 (2.77)	9-24	.72	17.69 (2.81)	11-24	.74
Social norms for HE	14.0 (3.58)	5-20	^a	14.86 (3.31)	5-20	^a
Social norms for un-HE	11.18 (3.27)	4-16	^a	10.2 (3.19)	4-16	^a
Social support for PA	13.22 (5.4)	6-24	.93	13.92 (5.18)	6-24	.87
Social norms for PA	14.91 (6.1)	8-32	^a	17.1 (7.16)	8-32	^a
Behavior at church and in general
Healthy foods consumed at church	15.0 (3.6)	5-20	^a	16.0 (3.3)	7-20	^a
Unhealthy foods consumed at church	10.4 (3.2)	4-16	^a	9.3 (3.2)	4-16	^a
Intention to use PA facilities at church	14.5 (6.4)	8-32	^a	14.2 (5.9)	8-32	^a
HE in general^b	7.4 (2.7)	1-15	^a	5.9 (2.7)	0-13	^a
	Baseline			Follow-up
	n		%	n		%
PA in general
High	99		31.0	98		46.4
Moderate	66		20.7	52		24.6
Low	154		48.3	61		28.9

Note. SD = standard deviation; HE = healthy eating; un-HE = unhealthy eating; PA = physical activity.

Not computed for items not theorized to be correlated (Streiner, 2003). ^bLower scores suggest healthier eating habits.

Healthy Eating

Social support for HE was assessed by asking how often in the past 6 months people at church did six actions supportive or unsupportive of healthy eating, for example, discuss your eating habits with you, offer you food you’re trying not to eat (Sallis, Grossman, Pinski, Patterson, & Nader, 1987). Responses were never, rarely, sometimes, or almost always. Negative items were reverse scored and responses summed to create a Church Social Support for Healthy Eating score. To assess descriptive social norms related to HE at church (i.e., perceptions of what others eat and drink), participants indicated how often they saw others consume nine specific foods and drinks at church events where food was served, using the same 4-point scale as above. These nine items were each grouped as healthy (e.g., baked/broiled/grilled chicken/fish; fresh fruit) or unhealthy (e.g., fried chicken/fish; chips or other salty snacks) items (St. Louis University, 2004). Using these healthy and unhealthy groups, social norms had two subscales: Church Social Norms Related to Healthy Eating, the sum of responses to five healthy food and drink items, and Church Social Norms Related to Unhealthy Eating, the sum of responses to four unhealthy items.

Types of food and drink consumed at church were measured by asking participants to report how often they ate or drank the same nine foods and drinks used in the measures of social norms, using the same response options. Measures of food and drink consumed at church had two subscales: Healthy Foods Consumed, the sum of responses to five healthy food and drink items consumed at church events, and Unhealthy Foods Consumed, the sum of four unhealthy food and drink items consumed at church events.

The Starting the Conversation Diet Tool is a validated instrument assessing dietary patterns (Paxton, Strycker, Toobert, Ammerman, & Glasgow, 2011). This tool asks about consumption of healthy (e.g., FV, chicken, fish, and beans) and less healthy foods (e.g., fast food meals, desserts) over the past few months. Consistent with Paxton et al. (2011), responses representing a range of frequencies (e.g., <1 time/week, 1-3 times/week, ≥4 times/week) to eight individual scale items were scored with integers ranging from 0 to 2. Individual item scores were combined to create a summary score, with lower scores reflecting healthier eating.

Physical Activity

Social support for PA among members was measured by asking how often in the past 6 months people at church did six actions that could be supportive or unsupportive of physical activity, for example, offer to exercise with you, encourage you to stick to your exercise program (Sallis et al., 1987). Responses used the same 4-point scale as described above. Responses were summed to create an overall Church Social Support for Physical Activity score. To assess descriptive social norms related to PA at church (i.e., perceptions of other members’ PA habits), participants indicated how often they saw others using eight types of PA equipment or resources at church, for example, exercise room/equipment, walking groups or clubs (St. Louis University, 2004). Responses were summed to create Church Social Norms for Physical Activity.

To assess Intentions to Use Physical Activity Facilities at church, participants used the same 4-point scale described above and indicated how often they would use the same eight PA resources described above, if available, at their church. We did not measure actual use of physical activity facilities at church due to limited availability of such facilities and resources across sites at baseline.

The widely used and validated International Physical Activity Questionnaire (IPAQ) was used to measure overall PA (Craig et al., 2003). The IPAQ uses self-reported vigorous and moderate physical activities, and walking during the past 7 days to estimate total PA. Following IPAQ data processing guidelines (Patterson, 2010), participants were classified into one of three PA level categories (low, moderate, or high) based on total metabolic equivalent (MET) minutes per week.

Data Analysis

Preliminary Analysis

Frequency distributions of all relevant categorical variables and descriptive statistics on all continuous variables were examined (Table 1). Demographic variables were examined for relationships to independent and dependent variables using correlation, t test, analysis of variance, or χ² statistics as appropriate. Any demographic variables associated with independent or dependent variables were included in subsequent multivariate analyses.

Main Outcome Analysis

To explore change in perceived church social environments, within-church analyses compared baseline to follow-up means using paired samples t test or its nonparametric equivalent when sample sizes were less than 30. Next, we pooled data across sites to assess relationships between perceived church social environments and behavioral outcomes (HE and PA at church and in general). Analyses were conducted exploring relationships within and across behavioral domains (e.g., relationship between social support for healthy eating and physical activity behavior at church and in general). Generalized estimating equations accounted for clustered data and controlled for the covariates (i.e., demographic variables) described above. A change variable was calculated by subtracting baseline from follow-up values for each church social environment variable and entered into each regression model along with relevant covariates. Continuous outcomes measuring church behavior or intentions and the general eating behavior outcome used multiple regression models. An ordinal logistic regression model assessed the three-category PA outcome measure. Main outcome analyses were first conducted including all sites. To minimize the possibility of type II errors, we then repeated these analyses using the three sites that saw significant changes in perceived social environments at follow-up. Analyses were conducted using SPSS 23.0. An α level of .05 was used to determine statistical significance.

Results

Of 319 study participants who completed baseline surveys, 258 (81%) completed follow-up surveys. Table 2 shows follow-up participants’ baseline demographic characteristics. The follow-up sample was predominantly Black/African American (84.1%) and female (82.2%). Participants were approximately equally likely to be single (48%) or married (44.6%). A majority of participants were working full- or part-time (69.1%), had at least some college education (71.7%), and reported an annual household income above $25,000 (54.7%). About one third of all study participants (32.6%) belonged to one church; the remaining churches made up between 9.3% and 19.0% of the sample (Table 2). Dropout analyses comparing baseline characteristics of the 258 participants retained in the study to the 61 lost to follow-up revealed that those lost to follow-up tended to be younger, male, and demonstrated traits suggesting lower socioeconomic status (Arriola et al., 2016). This pattern of attrition was observed in a similar study (Wilcox et al., 2007). Analyses reported here are based on the 258 participants who completed follow-up surveys.

Table 2.

Demographic Characteristics of Follow-up Participants (N = 258).

Demographic characteristic	Follow-up participants, n (%)
Gender^a
Female	212 (82.2)
Male	45 (17.4)
Race^a
Black/African American	217 (84.1)
White	39 (15.1)
Marital status^a
Single	123 (48.0)
Married	115 (44.6)
Widowed	18 (7.0)
Employment status^a
Working full-time or part-time	177 (69.1)
Retired	44 (17.1)
Not employed	35 (13.6)
Education^a
Less than high school graduate	22 (8.5)
High school graduate or GED	48 (18.6)
Some college	87 (33.7)
College or graduate degree	98 (38.0)
Income^a
≤$10,000	28 (10.9)
$10,001-$25,000	50 (19.4)
$25,001-$50,000	73 (28.3)
>$50,000	68 (26.4)
Church participation^b
Church 1	49 (19.0)
Church 2	42 (16.3)
Church 3	20 (7.8)
Church 4	39 (15.1)
Church 5	24 (9.3)
Church 6	84 (32.6)

Percentages may not total 100% due to missing data. ^bChurch participation is the number of church members who completed the follow-up survey at each church.

Within-church analyses compared change in church social environments from baseline to follow-up regarding HE and PA (Table 3). Three churches (Churches 2, 4, and 6) demonstrated significant improvements in all three aspects of the social environment related to HE. Of the five churches that sought to change the PA environment, two (Churches 4 and 6) exhibited significant improvements in the church social environment for PA from baseline to follow-up. The three churches that experienced no significant change in the church social environment for HE or PA also reported fewer changes to their churches’ physical environment, as reported elsewhere (Arriola et al., 2016).

Table 3.

Significant Mean Changes in Church Environments From Baseline to Follow-Up.

Church (n^a)	Variable	Baseline	Follow-up	Difference	p value
2 (n = 39-42)	Social norms—Healthy food	13.2	15.1	1.8	.006
	Social norms—Unhealthy food	11.3	9.3	−2.0	.000
	Social support—Healthy eating	18.2	19.3	1.1	.027
	Social norms—Physical activity	—	—	—	—
	Social support—Physical activity	—	—	—	—
4 (n = 28-37)	Social norms—Healthy food	13.1	14.6	1.6	.021
	Social norms—Unhealthy food	13.9	12.8	−1.1	.031
	Social support—Healthy eating	14.3	15.6	1.3	.009
	Social norms—Physical activity	14.7	19.4	4.7	.001
	Social support—Physical activity	11.3	13.1	1.7	.030
6 (n = 73-81)	Social norms—Healthy food	14.6	15.5	0.9	.030
	Social norms—Unhealthy food	10.1	9.3	−0.8	.039
	Social support—Healthy eating	17.0	17.9	0.9	.014
	Social norms—Physical activity	17.2	19.9	2.7	.007
	Social support—Physical activity	14.2	15.4	1.2	.035

Note. “—” indicates that a church did not seek to change the environment in that area. Churches 1, 3, and 5 did not see any changes and are not included in table. All significant changes were in expected/healthy direction. Wilcoxon signed rank test used for pairs <30.

Sample sizes vary by study variable due to missing data.

Main outcome analyses explored change in perceived church social environments relative to church eating behavior and intention to use church PA resources as well as diet and physical activity behaviors generally. Results of the main outcome analyses including all sites are shown in Table 4. Main outcome findings among the three sites that saw significant changes in perceived social environments at follow-up are shown in Table 5. Among the subset of churches, we found that significant relationships observed among all sites were strengthened in most cases, and additional relationships became significant. The following findings are from analyses conducted with the subset of three churches (Table 5). An increase in social support for HE was associated with greater levels of intended use of PA facilities (Β = .44, p < .001), and greater PA levels (adjusted odds ratio [aOR] = 1.083; 95% confidence interval [CI] = [1.062, 1.105]), as well as healthier eating habits generally (Β = −.14, p < .05). Similarly, increased social support for PA was associated with greater levels of intended use of PA facilities (Β = .27, p < .001), greater PA levels (aOR = 1.073; 95% CI = [1.011, 1.137]), and healthier eating habits generally (Β = −.10, p < .001). Increases in social norms related to HE were associated with increased HE (Β = −.07, p < .01) but decreased PA at follow-up (aOR = 0.958; 95% CI = [0.954, 0.962]). Increased social norms related to unhealthy eating was associated with a decrease in PA in general at follow-up (aOR = 0.844; 95% CI = [0.795, 0.895]). An increase in social norms related to PA was associated with a greater level of healthy foods consumed at church (Β = .06, p < .05), intended use of PA facilities (Β = .14, p < .001), and PA generally (aOR = 1.041; 95% CI = [1.001, 1.082]).

Table 4.

Associations Between Change in Perceived Social Environment and Members’ Eating Behavior, Physical Activity, and Intentions, All Sites (N = 258).

	Church behavior			General behavior
Change (increase) in perceived church social environment	Healthy food consumed, B (SE)	Unhealthy food consumed, B (SE)	Intended use of physical activity facilities, B (SE)	Eating behavior^a, B (SE)	Physical activity, OR (95% CI)
Social support—Healthy eating	—	—	.36 (.11)**	−.14 (.03)***	1.062 (1.021, 1.104)**
Social support—Physical activity	—	−.09 (.04)**	.27 (.02)***	−.07 (.02)***	—
Social norms—Healthy eating	—	—	—	—	—
Social norms—Unhealthy eating	—	—	—	—	0.878 (0.828, 0.931)***
Social norms—Physical activity	.05 (.02)*	—	.18 (.03)***	—	1.030 (1.010, 1.051)***

Note. B = unstandardized regression coefficient; SE = standard error; OR = odds ratio; CI = confidence interval. All models utilized generalized estimating equations due to the clustered nature of the data. Dashes represent nonsignificant findings.

Lower scores reflect healthier eating habits.

***

p < .001. **p < .01. *p < .05.

Table 5.

Associations Between Change in Perceived Church Social Environment and Members’ Eating Behavior, Physical Activity and Intentions, Among Sites With Significant Changes in Perceived Social Environments at Follow-up (N = 165).

	Church behavior			General behavior
Change (increase) in perceived church social environment	Healthy food consumed, B (SE)	Unhealthy food consumed, B (SE)	Intended use of physical activity facilities, B (SE)	Eating behavior^a, B(SE)	Physical activity, OR (95% CI)
Social support—Healthy eating	—	—	0.44 (0.08)***	−0.14 (0.05)*	1.083 (1.062, 1.105)*
Social support—Physical activity	—	—	0.27 (0.02)***	−0.10 (0.01)***	1.073 (1.011, 1.137)*
Social norms—Healthy eating	—	—	—	−0.07 (0.02)**	0.958 (0.954, 0.962)***
Social norms—Unhealthy eating	—	—	—	—	0.844 (0.795, 0.895)***
Social norms—Physical activity	0.06 (0.02)*	—	0.14 (0.02)***	—	1.041 (1.001, 1.082)**

Lower scores reflect healthier eating habits.

***

p < .001. **p < .01. *p < .05.

Discussion

The Prevention Strategies that Work program supported six rural South Georgia churches to implement organizational policy and environmental change strategies that promote HE and PA. Strategies sought to change both the church physical and social environments. These mechanisms may affect the church as an organization, relationships between members, and members as individuals, consistent with ecological models of health promotion encouraging intervention at multiple levels (McLeroy et al., 1988; Sallis et al., 2008). Faith-based organizations are important settings for health promotion (Bopp & Fallon, 2008; Bowen et al., 2015; Campbell et al., 2007; Holt et al., 2014; Kegler et al., 2012); however, health promotion efforts have traditionally used the FBO as a venue for reaching members, rather than seeking to change the FBO environment itself (National Cancer Institute, 2001; Resnicow et al., 2002). Notable exceptions include the Faith, Activity & Nutrition Program and Body & Soul (Resnicow et al., 2004; Wilcox et al., 2010), which both included policy and/or environmental change.

A significant achievement of this project is its identification of specific characteristics of social norms and social support linked to member’s HE and PA behaviors. These findings highlight potential points of intervention for future chronic disease prevention studies addressing policy and environmental change within faith-based settings. Although this program was primarily focused on changing the physical environment, we hypothesized that there may also be social environmental changes (i.e., social support, social norms), which in turn may contribute to healthy behaviors at church and in general (Figure 1). Members at three of six churches reported significant improvements to the church social environment for HE, and members at two of five churches that selected PA strategies reported significant improvements to the church social environment for PA.

Changes to the church social environment were related to both behavior at church and in general at 1-year follow-up. Social support for HE, social support for PA, and social norms for PA were each associated with three behavioral outcomes, both at church and in general. These findings are consistent with other studies that have found relationships between social support at church and diet and PA (Baruth & Wilcox, 2015; Condrasky et al., 2013; Fuemmeler et al., 2006). Social norms for healthy eating was related to both general behavior outcomes, and social norms for unhealthy eating was related to one behavioral outcome (PA in general). Few studies have explored the relationship between church social norms and healthy behaviors, possibly because social norms are difficult to define and measure (Ball et al., 2010). The unexpected inverse relationship that we observed between increased social norms for healthy eating and general PA behavior deserves further study.

Notably, several associations we found crossed behaviors—aspects of the social environment around food were associated with PA and vice versa. This may reflect that individuals see diet and PA as interrelated lifestyle behaviors that together support health promotion. While the relationship between healthy eating and physical activity has not been widely studied, it does have some basis in the published literature (Halliday et al., 2014). Future studies may benefit from mediation analyses to understand how these variables interact. Studies of this program have explored church policy and environmental changes in relation to church and general behavior as well as how social environmental factors relate to these behaviors (Arriola et al., 2016) but have not explored how these independent variables are related to each other, or as mediating factors for behavior. Previous studies that have explored socioenvironmental or sociobehavioral predictors of diet and PA change have had mixed results. Results from the Delta Body & Soul trial found that increased social support was a predictor of improved diet quality but not PA (Thomson et al., 2014). In contrast, Baruth and Wilcox (2015) found that increases in social support were associated with PA, but not FV intake in the Faith, Activity & Nutrition study. This study is likely the first to explore the role of descriptive social norms on eating and PA behavior or intentions before and after a policy and environmental change program.

This study has several limitations. Without a comparison group, causal assumptions are limited. Second, while self-reported data are useful for accurately capturing participant perspectives and awareness, an inherent weakness is a lack of objectivity. Consequently, whether changes in reported perceptions of social environment characteristics are due to members’ increased awareness of their existing social environment or actual changes that occurred during the intervention period remains unclear. Additionally, the variable used as a proxy for use of physical activity facilities at church, Intention to Use Physical Activity Facilities at Church, is not ideal and limits our ability to understand the relationship between social environment and actual behavior within the church context.

With this study’s focus on social environments, including other socioenvironmental factors (e.g., pastor support, other formal messages) is beyond the scope of this study. We previously found that HE messages were linked to significant decreases in unhealthy eating at church (Arriola et al., 2016). It is possible that by excluding health promotion message variables from this study, we have yielded a conservative estimate of the relationship between church social environments and member behavior.

Investigators were unable to examine potential pathways whereby perceived church-level characteristics influenced relationships between members’ perceived social environments and reported health behavior. Enrolled churches varied substantially in membership size as evidenced by the fact that one third of participants belonged to one church. Investigation of whether membership size or other varying church characteristics mediated relationships between social environments and member health behavior was beyond the scope of this study. Future studies involving a larger sample of churches may allow investigators to explore how church-level factors influence these relationships.

Implications for Practice

Study findings demonstrate that social-environmental characteristics are essential to multilevel interventions. Within the context of the broader mini-grants intervention project, and consistent with ecological models of health and similar faith-based studies, findings suggest that intervention components addressing social norms and social support should supplement intervention components addressing changes in physical environments. Taken together, this integrated, environmental-level approach appropriately reflects the complexity of addressing health behavior change.

Footnotes

Acknowledgements

The authors thank the members of the Emory Prevention Center’s Community Advisory Board for their many contributions to this project. They also thank the organizations that participated in the “Prevention Strategies that Work” mini-grants program.

Authors’ Note

The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the Centers for Disease Control and Prevention and the National Cancer Institute, the funding agencies.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was supported by Cooperative Agreement Numbers U48DP001909 from the Centers for Disease Control and Prevention and the National Cancer Institute.

References

Arriola

K. R. J.

Hermstad

Flemming

S. S.

Honeycutt

Carvalho

M. L.

Cherry

S. T.

. . . Kegler

M. C.

(2016). Promoting policy and environmental change in faith-based organizations: Outcome Evaluation of a mini-grants program. Health Promotion Practice, 17, 146-155. doi:10.1177/1524839915613027

Arriola

K. R. J.

Hermstad

Flemming

S. S.

Honeycutt

Carvalho

M. L.

Cherry

S. T.

. . . Kegler

M. C.

(2017). Promoting policy and environmental change in faith-based organizations: Description and findings from a mini-grants program. American Journal of Health Promotion, 31, 192-199. doi:10.4278/ajhp.150212-QUAN-724.

Ball

Jeffery

R. W.

Abbott

McNaughton

S. A.

Crawford

(2010). Is healthy behavior contagious: Associations of social norms with physical activity and healthy eating. International Journal of Behavioral Nutrition and Physical Activity, 7, 86-94.

Baruth

Wilcox

(2015). Psychosocial mediators of physical activity and fruit and vegetable consumption in the Faith, Activity, and Nutrition Programme. Public Health Nutrition, 18, 2242-2250.

Bopp

Fallon

(2008). Community-based interventions to promote increased physical activity. Applied Health Economics and Health Policy, 6, 173-187.

Bowen

D. J.

Barrington

W. E.

Beresford

S. A. A.

(2015). Identifying the effects of environmental and policy change interventions on healthy eating. Annual Reviews of Public Health, 36, 289-306.

Campbell

M. K.

Hudson

M. A.

Resnicow

Blakeney

Paxton

Baskin

(2007). Church-based health promotion interventions: Evidence and lessons learned. Annual Reviews of Public Health, 28, 213-234. doi:10.1146/annurev.publhealth.28.021406.144016

Centers for Disease Control and Prevention. (2013). Healthy communities program. Retrieved from http://www.cdc.gov/healthycommunitiesprogram/

Centers for Disease Control and Prevention. (2015). Chronic disease overview. Retrieved from http://www.cdc.gov/chronicdisease/overview/

10.

Chiuve

S. E.

Fung

T. T.

Rimm

E. B.

F. B.

McCullough

M. L.

Wang

. . . Willett

W. C.

(2012). Alternative dietary indices both strongly predict risk of chronic disease. Journal of Nutrition, 142, 1009-1018. doi:10.3945/jn.111.157222

11.

Colditz

G. A.

Philpott

S. E.

Hankinson

S. E.

(2016). The impact of the nurses’ health study on population health: Prevention, translation, and control. American Journal of Public Health, 106, 1540-1545.

12.

The Community Guide to Preventive Services. (2013). Retrieved from http://www.thecommunityguide.org/index.html

13.

Condrasky

M. D.

Baruth

Wilcox

Carter

(2013). Predictors of change in fruit and vegetable consumption in a faith-based intervention with African American adults. Family & Community Health, 36, 236-247. doi:10.1097/FCH.0b013e318293159b

14.

Craig

C. L.

Marshall

A. L.

Sjöström

Bauman

A. E.

Booth

M. L.

Ainsworth

B. E.

. . . Oja

(2003). International Physical Activity Questionnaire: 12-country reliability and validity. Medicine and Science in Sports and Exercise, 35, 1381-1395.

15.

Firestone

M. J.

S. S.

Bartley

K. F.

Eisenhower

D. L.

(2015). Perceptions and the role of group exercise among New York City adults, 2010-2011: An examination of interpersonal factors and leisure-time physical activity. Preventive Medicine, 72, 50-55.

16.

Frieden

T. R.

(2010). A framework for public health action: The health impact pyramid. American Journal of Public Health, 100, 590-595. doi:10.2105/ajph.2009.185652

17.

Fuemmeler

B. F.

Mâsse

L. C.

Yaroch

A. L.

Resnicow

Campbell

M. K.

Carr

. . . Williams

(2006). Psychosocial mediation of fruit and vegetable consumption in the body and soul effectiveness trial. Health Psychology, 25, 474-483.

18.

Glanz

Rimer

B. K.

Viswanath

(2008). Health behavior and health education: Theory, research, and practice (4th ed.). San Francisco, CA: Jossey-Bass.

19.

Halliday

T. M.

Davy

B. M.

Clark

A. G.

Baugh

M. E.

Hedrick

V. E.

Marinik

E. L.

. . . Winett

R. A.

(2014). Dietary intake modification in response to participation in a resistance training program for sedentary older adults with prediabetes: findings from the resist diabetes study. Eating Behaviors, 15, 379-382. doi:10.1016/j.eatbeh.2014.04.004

20.

Holt

C. L.

Clark

E. M.

Debnam

K. J.

Roth

D. L.

(2014). Religion and health in African Americans: The role of religious coping. American Journal of Health Behavior, 38, 190-199.

21.

Johnson

N. B.

Hayes

L. D.

Brown

Hoo

E. C.

Ethier

K. A.

, & Centers for Disease Control and Prevention. (2014). CDC National Health Report: Leading causes of morbidity and mortality and associated behavioral risk and protective factors—United States, 2005-2013. Morbidity and Mortality Weekly Report, 63(Suppl. 4), 3-27.

22.

Kegler

M. C.

Escoffery

Alcantara

I. C.

Hinman

Addison

Glanz

(2012). Perceptions of social and environmental support for healthy eating and physical activity in rural southern churches. Journal of Religion and Health, 51, 799-811. doi:10.1007/s10943-010-9394-z

23.

McLeroy

K. R.

Bibeau

Steckler

Glanz

(1988). An ecological perspective on health promotion programs. Health Education Quarterly, 15, 351-377.

24.

Moore

L. V.

Thompson

F. E.

(2015). Adults meeting fruit and vegetable intake recommendations—United States, 2013. Morbidity and Mortality Weekly Report, 64, 709-713.

25.

National Cancer Institute, Research Tested Intervention Programs. (2001). Eat for life. Retrieved from http://rtips.cancer.gov/rtips/programDetails.do?programId=224488

26.

Patterson

(2010). IPAQ scoring protocol. Retrieved from https://sites.google.com/site/theipaq/scoring-protocol

27.

Paxton

A. E.

Strycker

L. A.

Toobert

D. J.

Ammerman

A. S.

Glasgow

R. E.

(2011). Starting the conversation: Performance of a brief dietary assessment and intervention tool for health professionals. American Journal of Preventive Medicine, 40, 67-71. doi:10.1016/j.amepre.2010.10.009

28.

Resnicow

Campbell

M. K.

Carr

McCarty

Wang

Periasamy

. . . Stables

(2004). Body and soul: A dietary intervention conducted through African-American churches. American Journal of Preventive Medicine, 27, 97-105. doi:10.1016/j.amepre.2004.04.009S0749379704000856

29.

Resnicow

Jackson

Braithwaite

DiIorio

Blisset

Rahotep

Periasamy

(2002). Healthy Body/Healthy Spirit: A church-based nutrition and physical activity intervention. Health Education Research, 17, 562-573.

30.

Sallis

J. F.

Grossman

R. M.

Pinski

R. B.

Patterson

T. L.

Nader

P. R.

(1987). The development of scales to measure social support for diet and exercise behaviors. Preventive Medicine, 16, 825-836.

31.

Sallis

J. F.

Owen

Fisher

E. B.

(2008). Ecological models of health behavior. In Glanz

Rimer

B. K.

Viswanath

(Ed.), Health behavior and health education: Theory, research and practice (4th ed., pp. 465-485). San Francisco, CA: Jossey-Bass.

32.

Shaikh

A. R.

Vinokur

A. D.

Yaroch

A. L.

Williams

G. C.

Resnicow

(2011). Direct and mediated effects of two theoretically based interventions to increase consumption of fruits and vegetables in the healthy body healthy spirit trial. Health Education & Behavior, 38, 492-501.

33.

St. Louis University, Center for Advanced Studies in Nutrition and Social Marketing. (2004). St. Louis Church Health Survey. Davis, CA: Author.

34.

Streiner

D. L.

(2003). Being inconsistent about consistency: When coefficient alpha does and doesn’t matter. Journal of Personality Assessment, 80, 217-222. doi:10.1207/s15327752jpa8003_01

35.

Thomson

J. L.

Zoellner

J. M.

Tussing-Humphreys

L. M.

(2014). Psychosocial predictors of diet and physical activity in African-Americans: Results from the delta body and soul effectiveness trial, 2010-2011. American Journal of Health Promotion, 28(3), e81-e91.

36.

U.S. Department of Agriculture, & U.S. Department of Health and Human Services. (2010). Dietary guidelines for Americans 2010. Retrieved from http://www.health.gov/dietaryguidelines/dga2010/DietaryGuidelines2010.pdf

37.

U.S. Department of Health and Human Services. (2008). 2008 Physical activity guidelines for Americans. Retrieved from http://health.gov/paguidelines/pdf/paguide.pdf

38.

Ward

B. W.

Schiller

J. S.

Goodman

R. A.

(2014). Multiple chronic conditions among US adults: A 2012 update. Preventing Chronic Disease, 11, 130389.

39.

Wilcox

Klaken

Bopp

Gethers

Huang

McClorin

. . . Yancey

(2007). Increasing physical activity among church members: Community-based participatory research. American Journal of Preventive Medicine, 32, 131-138.

40.

Wilcox

Laken

Parrott

A. W.

Condrasky

Saunders

Addy

C. L.

. . . Samuel

(2010). The Faith, Activity, and Nutrition (FAN) program: Design of a participatory research intervention to increase physical activity and improve dietary habits in African American churches. Contemporary Clinical Trials, 31, 323-335.