Perceived Environmental Church Support is Associated with Dietary Practices among African American Adults

Abstract

Background

A unique strength of the African American community is the importance of the church and faith. Interventions promoting health may want to build on these strengths by developing faith-based interventions that encourage churches to create an environment that supports behavior change.

Objective

To examine the relationship between perceived environmental church support for healthy eating and intake of fruit and vegetables and fat and fiber-related behaviors, and to examine if these relationships differ by gender.

Design

A cross-sectional study in which participants completed self-report dietary and perceived church support measures prior to the initiation of an intervention. Relationships between fruit and vegetable consumption, fat and fiber-related behaviors, and perceived church support (total, written informational, spoken informational, instrumental [fruit and vegetable consumption only]), along with Support x Gender interactions were examined.

Participants/Setting

Participants were 1136 African American church members from four geographically-defined districts in South Carolina.

Statistical Analyses

Multiple regression models controlling for gender, age, years of education, health rating, and body mass index using SAS PROC MIXED. A separate model was conducted for each measure of perceived church support and each type of healthy eating index.

Results

Perceived total church support and perceived written and spoken informational church support were related to significantly higher fruit and vegetable intake and more favorable fiber-related behaviors, whereas only perceived total and perceived written informational support were associated with more low-fat dietary behaviors. Perceived instrumental church support was not associated with fruit and vegetable consumption. No gender differences were found.

Conclusion

The social and physical church environment may be important factors influencing the dietary habits of its members. Future faith-based interventions should further explore the role of the church environment in improving the dietary practices of its members.

Introduction

Poor diet and physical inactivity are among the leading causes of death in the United States (1) and have been linked to chronic health conditions (2, 3). Despite the benefits of a diet high in fruits, vegetables, and whole grains and low in fat (4, 5), many Americans are not meeting federal dietary recommendations (6). African Americans appear to have poorer dietary practices than Caucasians (7–9), perhaps contributing to existing disparities in health (10).

The role of individual factors (e.g. attitudes, beliefs) continues to be a major focus of behavior change research (11). However, behavior is complex and influenced by multiple levels. Ecological models (11, 12) recognize the importance of personal characteristics, while also realizing that factors outside the individual (e.g., socio-cultural and environmental factors) influence behavior.

The church has traditionally been an important organizing structure and source of support in the African American community (13). Because church leaders have significant influence on their members, faith-based environments may be able to influence behaviors, such as healthy eating, in unique ways. For example, churches can make healthy eating a social norm, they can institute and enforce policies that encourage healthy eating (e.g. low-fat options must be available), and key leaders can encourage and model healthy eating. Further exploration of the role of the church environment in improving the dietary practices of its members is needed. The purpose of this paper is to examine the relationship between perceived environmental church support and eating practices and to examine whether relationships differed by gender.

Methods

The Faith, Activity, and Nutrition (FAN) program is a 5-year physical activity and nutrition intervention implemented in African Methodist Episcopal (AME) churches. A more detailed description of the study methods are described elsewhere (14). Briefly, all churches in four, pre-defined, geographical districts in South Carolina were invited to participate. Pastor-appointed liaisons from interested churches recruited congregants to take part in a measurement session where research staff administered physical assessments and participants completed a detailed survey. Inclusion criteria were: (a) 18+ years, (b) attend worship services ≥1 time per month, (c) free of serious medical conditions or disabilities that would make physical activity difficult. The University of South Carolina Institutional Review Board approved the study. All participants provided written informed consent. This study uses baseline data only (collected from 2007–2009).

Measures

Sociodemographics and Health

Questions from the Behavioral Risk Factor Surveillance System (BRFSS) (15), asked participants to report their age, gender, race, education, and general health. Objectively measured height and weight were obtained by trained staff. Body mass Index (BMI) was calculated as kg/m² using standard procedures and cut points (16).

Fruit and Vegetable Intake

Fruit and vegetable (F&V) consumption was measured with two instruments. The National Cancer Institute (NCI) F&V all-day screener assessed F&V consumption in the past three months (17). We used nine of the original ten items (French fry consumption was excluded) (18). This instrument correlates moderately with 24-hour recalls (Men: r = 0.66; Women: r = 0.51) (19), including in similar populations as ours (r = .51) (20).

Participants also completed a two-item measure. Participants were asked, “How many servings of fruit (vegetables) do you usually eat each day?” This summed measure, used in two other studies targeting African American churches (21, 22), correlates with total serum carotenoids (excluding lycopene) (r = .22) similar to longer measures (r=. 29–.35) (23).

Fat and Fiber Intake

The Fat and Fiber-Related Behavior Questionnaire (24, 25) assessed fat- (27 items) and fiber-related (14 items) dietary behaviors over the past three months. These summary scores have high test-retest correlations over 3 (r = .60 to .79) and 12 months (r = .53 to .74), are moderately correlated with a food frequency questionnaire (r = .53 and r = .50), and are more responsive to change in a dietary intervention than more detailed food frequency questionnaires (24, 26).

Perceived Environmental Church Support

Because an existing church support scale was not available in the literature, we developed six items that assessed perceived support for healthy eating over the past 12 months. Items that had face validity were developed to capture important types and sources of support in church settings based on experiences from a previous faith-based project (27, 28), input from church leaders and lay members, and the guiding theory for our intervention (11).

Exploratory factor analysis in our sample revealed acceptable factor loadings (factor loadings > 0.63) for scales that were named perceived written informational (3 items) and perceived spoken informational (2 items) support. Items also loaded heavily on the total perceived church support scale (6 items; factor loadings > 0.59), with the exception of one item that assessed how often F&Vs were served at church events that involved food, which had a lower factor loading (factor loading=0.24). Internal consistency was high for the total perceived church support (α=0.84) and perceived written informational support scales (α=0.91) and acceptable for perceived spoken informational support (α=0.69). Perceived instrumental support (1 item) assessed how often F&Vs were served at church events that involved food. A mean score for each type of perceived support was calculated.

Statistical Analyses

A square root transformation corrected skewness in the NCI F&V measure. Gender differences in support measures and dietary behaviors were examined using independent sample t-tests.

Multiple regression models examined associations between perceived environmental church support and healthy eating. A separate model was conducted for each support measure and each dietary measure, and all models controlled for gender, age, education, health rating, and BMI. Because participants represented 75 churches and thus share a unique social/physical environment, all statistical analyses controlled for the influence of church (i.e., church was treated as a random variable with participants nested in churches). Support x Gender interactions tested whether the association between perceived church support and diet differed by gender. If the Support x Gender interaction was not significant, the main effect of support was interpreted; if it was significant, simple effect analyses (regression models were ran separately for men and women) were conducted to facilitate interpretation of the interaction. Because perceived instrumental church support focused specifically on F&V consumption, this variable was only examined relative to F&V consumption. Statistical significance was accepted at p<0.05.

Results

Table 1 presents demographic, health, dietary, and church support measures for all participants (n=1136; 99.6% African American) and for men and women separately. There were no significant gender differences for any of the dietary measures. Men reported receiving significantly more perceived written informational church support (p=0.01) than women.

Table 1.

Demographic, Health, and Church Support Characteristics of AME Church Members (N=1136)^a

	Total Sample (n=1136)		Men (n= 275)		Women (n=861)
	n	% or Mean (SD)b	n	% or Mean (SD)b	n	% or Mean (SD)b
Age, years	1136	54.0 (14.0)	275	55.4 (14.7)	861	53.6 (13.7)
Education
Less than HS graduate	113	10.0	34	12.4	79	9.2
HS grad or GED	352	31.0	94	34.2	258	30.0
Some college (1–3 years)	336	29.6	74	26.9	262	30.4
College graduate (4+ years)	335	29.5	73	26.6	262	30.4
Marital Statush
Married/Unmarried couple	615	54.4	187	68.5	428	49.9
Not married	516	45.6	86	31.5	430	50.1
BMI, kg/m2ch	1136	32.9 (7.4)	275	30.6 (6.3)	861	33.6 (7.6)
Weight statush
Underweight (BMI<18.5)	6	0.5	2	0.7	4	0.5
Normal weight (BMI<25)	119	10.5	42	15.3	77	8.9
Overweight (25≥BMI<30)	311	27.4	96	34.9	215	25.0
Obese (BMI≥30)	700	61.6	135	49.1	565	65.6
Dietary Behaviors
Fruit and Vegetable (NCI)d	1132	3.9 (3.8)	274	4.2 (4.5)	858	3.7 (3.6)
Fruit and Vegetable (2-item)	1130	3.5 (1.8)	273	3.4 (2.0)	857	3.6 (1.8)
Fat-Related Behaviorse	1133	2.7 (0.4)	275	2.7 (0.4)	858	2.7 (0.5)
Fiber-Related Behaviorsf	1124	2.9 (0.5)	271	2.9 (0.5)	853	2.9 (0.5)
Environmental Church Support Measures
Total Church Supportg	1122	2.2 (0.7)	271	2.2 (0.7)	851	2.1 (0.7)
Written Informationalgh	1111	1.9 (0.9)	270	2.0 (0.8)	841	1.9 (0.9)
Spoken Informationalg	1113	2.2 (0.8)	269	2.3 (0.8)	844	2.2 (0.8)
Instrumentalg	1131	2.9 (1.0)	273	2.8 (0.9)	858	2.9 (1.0)

^aNote: not all sample sizes add to 1136 because of missing data by some participants

^bSD= Standard Deviation

^cBMI= Body Mass Index

^dNCI= National Cancer Institute Screener

^erange 1–5; lower score indicates less fat

^frange 1–5; lower score indicates more fiber

^grange 1–4; higher score indicates more support

^hsignificant gender difference (p<0.05)

Table 2 shows the estimates, standard errors, and p-values for each regression model examining the relationship between healthy eating and perceived environmental church support. None of the associations between perceived support and diet differed by gender. Total perceived church support was significantly associated with greater F&V consumption (both measures), more favorable fiber-related behaviors (e.g., eating more whole grains and vegetables), and more low-fat dietary behaviors. Perceived written informational support was significantly associated with greater F&V consumption (both measures), more favorable fiber-related behaviors, and more low-fat dietary behaviors. Perceived spoken informational support was significantly associated with greater F&V consumption (both measures) and more favorable fiber-related behaviors, but not with fat-related behaviors. There was no relationship between perceived instrumental church support and F&V consumption (either measure).

Table 2.

Relationship between Environmental Church Support and Dietary Practices in AME Church Members (N=1136)^a,^b

	NCI F&V		2-item F&V		Fat-related Behaviors		Fiber-related Behaviors
Support Variable	Estimate (SE)c	p-value	Estimate (SE)	p-value	Estimate (SE)	p-value	Estimate (SE)	p-value
Total Church Support
Church	0.05 (0.01)	0.002	0.43 (0.09)	0.003	−0.04 (0.02)	0.01	−0.10 (0.02)	0.0003
Church x Gender	−0.01 (0.03)	0.71	−0.28 (0.19)	0.14	−0.03 (0.04)	0.43	0.02 (0.05)	0.76
Written Informational Support
Written	0.05 (0.01)	0.001	0.34 (0.07)	0.003	−0.04 (0.02)	0.01	−0.08 (0.02)	<0.0001
Written x Gender	−0.01 (0.02)	0.57	−0.22 (0.15)	0.14	−0.03 (0.03)	0.43	−0.01 (0.04)	0.80
Spoken Informational Support
Spoken	0.03 (0.01)	0.004	0.29 (0.08)	0.03	−0.004 (0.02)	0.07	−0.07 (0.02)	0.004
Spoken x Gender	0.01 (0.02)	0.73	−0.21 (0.16)	0.19	−0.06 (0.04)	0.09	0.02 (0.04)	0.71
Instrumental Support
Instrumental	0.01 (0.01)	0.95	0.13 (0.07)	0.15	N/A		N/A
Instrumental x Gender	−0.02 (0.02)	0.46	−0.06 (0.14)	0.67	N/A		N/A

^aAll models controlled for gender, years of education and health rating (categorical variables); age and body mass index (continuous variables)

^bNote: sample sizes ranged from 1101–1127 because of missing dietary and/or support data by some participants

^cSE = Standard Error

Discussion

This study examined the relationship between perceived environmental church support and a number of dietary behaviors. Total church support was associated with all three dietary behaviors assessed (F&V, fiber and low fat practices). When examining the specific types of church support, both written and spoken informational church support were associated with higher F&V intake and more favorable fiber-related behaviors. Written informational support was also associated with more low-fat dietary practices. These findings point to the possible importance of the church and church leaders in creating an environment that promotes and encourages healthy eating, however additional research examining such relationships is needed. The lack of gender differences indicates that church support may be equally important for both men and women.

Examples of written informational support include distributing information about healthy eating in church bulletins, via handouts, or on church bulletin board. These types of support may educate individuals on the benefits of healthy eating or may serve as a simple reminder to choose healthy food options and food preparations. Examples of spoken informational support include the incorporation of healthy eating messages from the pulpit and in sermons. Pastors are the guiding force in African American churches (29) and tend to be trusted sources of information; support from these prominent leaders as well as positive role-modeling may encourage healthier eating habits among congregants.

Like our study, in Body and Soul, receiving educational materials was associated with increased F&V consumption and decreased fat consumption (30). In contrast, pastor support from the pulpit was not associated with either (30). Participants in Black Churches United for Health cited bulletins, pastor sermons, and print materials as having the most impact on increasing F&V (31), and perceptions were associated with actual behavior change. Caution should be taken when comparing our results to the intervention trials named above, as we used a cross-section design. Studies have not examined the relationship between church support and fiber-related behaviors. Our study suggests that the church may promote dietary behaviors beyond F&V and low-fat dietary consumption, however additional research is needed to further explore such relationships.

In contrast to Black Churches United for Better Health (31), where serving more F&Vs at church events was associated with increased consumption, we found no such association. Offering F&V at church may not be enough to effectively increase intake, perhaps because this opportunity accounts for a relatively small portion of total F&V consumption by our participants. Perhaps offering written and spoken information in conjunction with opportunities to consume F&V would be more effective; additional studies should explore this further.

Many of the AME churches in our sample have kitchens, cooks, and kitchen staff that prepare food for church functions, thus having significant influence over foods served to members (14). Encouraging church cooks to prepare healthy yet tasty foods with cultural traditions may be another way to improve eating practices. Because food is served at many church events, this approach has the potential to reach large numbers. Furthermore, support can be delivered in a manner that is culturally and spiritually appropriate.

This study had a number of strengths including the large sample size and the use of well-validated dietary measures. Furthermore, a unique aspect of this study was its examination of environmental church support for healthy eating.

We also recognize study limitations. We used a cross-sectional design, relied on self-report dietary screening measures, did not use 24-hour dietary recalls (considered the gold standard), and assessed perceived environmental church support verses objective support. Future studies are needed to develop valid support measure measures of environmental church support. Furthermore, those interested in healthy eating may be more likely to notice health-related information. Finally, instrumental church support was assessed with a single item that only asked about support for eating F&V.

Conclusion

Interventions promoting healthy eating may want to build on strengths within the African American community by encouraging the church to provide a supportive environment. Our findings support the use of ecological models in these endeavors, as factors beyond the individual may be associated with more favorable eating practices.