Predictors of change in fruit and vegetable consumption in a faith-based intervention with African American adults

Abstract

A majority of African American adults do not eat the recommended daily amount of fruit and vegetables. This study examined baseline demographic, health-related and psychosocial variables as predictors of change in fruit and vegetable consumption from baseline to post-program in a sample of church members taking part in a 15-month intervention. Participants who had a greater waist circumference, greater baseline fruit and vegetable consumption, greater leisure-time physical activity, higher levels of social support, greater attendance at worship service, were obese, and did not have diabetes at baseline showed higher post-test fruit and vegetable consumption.

Introduction

Higher fruit and vegetable intake has been shown to be associated with a decreased risk of a number of chronic diseases such as cancer, cardiovascular disease, diabetes, and obesity ^1–5. African Americans have a higher prevalence of these conditions than the general population ^6–9. African Americans are less likely to consume the recommended daily 2.5 cup equivalents of fruit and 2.5 equivalents of vegetables ¹⁰ compared to whites or the population as a whole ^11,12. These differences may contribute to the health disparities among the African American population. In an effort to improve health and reduce health disparities, there is a need to increase fruit and vegetable consumption among African Americans.

Churches provide a unique opportunity for improving the health of their members, as many include health as part of their mission, are willing to partner with secular agencies, and have the ability to recruit underserved populations ^13–15. Furthermore, church attendance has been linked to positive health care practices. Aaron and colleagues ¹⁶ found that those who attended church more frequently were more likely to have a regular source of medical care, blood pressure measurements, and dental visits.

Faith-based interventions within African American church communities have been successful in increasing fruit and vegetable intake ^17–20. To date, four large-scale studies have shown significant increases in fruit and vegetable consumption among African American church members. Healthy Body Healthy Spirit resulted in a 1.13 increase in fruit and vegetable consumption at 1 year ¹⁹, the Eat for Life trial showed a 1.1 serving increase in fruit and vegetable consumption after 1 year ¹⁸, and the Black Churches United for Better Health project showed a 0.85 increase in fruit and vegetable consumption after 2 years ¹⁷. The Body and Soul intervention, which combined aspects the Eat for Life and Black Churches United for Better Health interventions, increased fruit and vegetable intake by 0.7–1.4 servings, depending on the measure used, at 6 months ²⁰.

Although faith-based studies targeting African Americans have shown meaningful increases in fruit and vegetable consumption, no faith-based studies have explored which variables predict changes in consumption. A better understanding of who is more or less likely to increase fruit and vegetable consumption can help inform more targeted interventions. For example, if individuals with higher self efficacy are more likely to increase consumption, interventions can incorporate additional, or more intense, intervention strategies targeting those with low self efficacy upon program entry. Perhaps closest to our targeted population was a study by Langenberg and colleagues ²¹ targeting a diverse sample of WIC mothers (>50% African American) in Maryland. They found that education and employment (borderline), and baseline attitudes, self efficacy, barriers, knowledge, and responsibility (for food shopping and preparation) were significant predictors of change in fruit and vegetable consumption at the 6-month follow-up.

Most of the literature to date (faith-based and/or targeting African Americans) has been cross-sectional studies examining correlates of fruit and vegetable consumption at one time point (i.e. at baseline)^22–26. Variables such as gender ^23,26, age ^24–26, employment status²⁵, income ^25,26, marital status ²⁴, comorbidities ²⁵, education ^23,24,26, self-rated health ^23,26,weight status ²⁶, cooking practices ²⁵, beliefs,²³ knowledge ²³, outcome expectations ²⁵, barriers ^22,25, rewards ²², benefits ²² self efficacy ^23,25, social support ^23,25, social norms ²², church attendance ²⁵, and exercise ^25,26, have been associated with fruit and vegetable intake in African Americans. Although cross-sectional studies can be useful, studies examining predictors of change in outcomes (i.e. fruit and vegetable consumption) offer additional insight which may assist in efforts aimed at improving the targeted health behaviors.

The Faith, Activity, and Nutrition (FAN) study was a 15-month, physical activity and dietary intervention targeting African Methodist Episcopal (AME) churches in South Carolina. Results from FAN showed that fruit and vegetable consumption was significantly higher at the post test follow-up in the intervention group compared to the control group ²⁷. This study addresses a current gap in the literature by examining baseline predictors of change in fruit and vegetable consumption in a sample of African American church members taking part in FAN. We examined whether demographic, health-related and psychosocial variables at baseline predicted increases in fruit and vegetable consumption from baseline to 15 months (i.e. post- program).

Methods

The methods of the Faith, Activity, and Nutrition (FAN) program have been described in more detail elsewhere ^27,28. In brief, FAN was a physical activity and nutrition intervention implemented in African Methodist Episcopal (AME) churches in South Carolina. FAN used a community-based participatory research approach in which a planning committee consisting of church leaders, lay members of the church, and university staff worked together at all stages of the research project to develop, implement, and evaluate the program. The primary goals of FAN were to increase moderate to vigorous intensity physical activity and fruit and vegetable consumption, and to improve blood pressure ²⁸.

Research Design

This study used a group randomized design and included three waves of implementation. Churches were randomized to receive the intervention immediately following baseline assessments (i.e. intervention group) or at the end of the 15-month intervention period, following post measurements (i.e. control group).

Church Recruitment

As reported in more detail elsewhere ²⁸, 131 pastors from 4 geographically-defined AME districts in South Carolina were sent letters from their presiding elder introducing the FAN program and inviting participation. FAN staff made follow-up telephone calls to the pastors to provide more detail about the FAN program and to answer any questions. Pastors from interested churches typically appointed a liaison to assist in scheduling and coordinating measurement sessions and church intervention trainings.

Procedures

Liaisons from interested churches were asked to recruit members of their congregation to take part in a measurement session at baseline (pre-intervention). Recruitment goals were a function of church size (13 members for small churches, 32 for medium, 63 for large). Study staff provided the churches with flyers and announcements that could be used to recruit participants. At each baseline measurement session, participants completed an informed consent form that was approved by the Institutional Review Board at the University of South Carolina and by the FAN planning committee. To be eligible, participants had to be at least 18 years of age, be free of serious medical conditions or disabilities that would make small changes in physical activity or diet difficult (self-identified), and attend worship services at least once a month (to ensure intervention exposure). Upon providing consent, trained FAN staff took physical assessments and participants completed a comprehensive survey.

The same measures were repeated 15 months later (post-program). Prior to the scheduled post-measurement session, participants were mailed a letter inviting them to take part in a post-test assessments and the survey to complete prior to the session, and churches were asked to make announcements at worship services. Participants also received a phone call from study staff, reminding them to attend the scheduled measurement session. Participants unable to attend their scheduled measurement session were invited to attend a future session at a nearby church. Repeated contacts were made with participants not attending a session, asking them to return their survey in a postage-paid envelope.

Intervention

The intervention targets, guided by the structural ecologic model ²⁹, were developed by the FAN planning committee during the first year using a CBPR approach. Churches were asked to implement intervention activities, focusing on physical activity and healthy eating, which targeted each of the four structural factors within the model. Although churches had a great deal of flexibility in how they addressed each of the factors, they were asked to implement a set of core activities targeting both physical activity and healthy eating: distribute bulletin inserts, share messages from the pulpit, pass out educational materials, create a FAN bulletin board, and suggest guidelines and practices that the pastor can set.

Each church formed a committee and attended a one-day training. Each committee developed a formal intervention plan that followed the structural ecological model ²⁹ and was in line with the overall FAN objectives. Upon submission of their plan and budget, FAN churches received a stipend to assist with FAN-related activities. More details of committee training can be found elsewhere ²⁸.

Each church also sent two individuals to attend a cooks training that focused on the Dietary Approaches to Stop Hypertension (DASH) diet plan. More details of the committee and cooks trainings can be found elsewhere ^28,30.

In addition to the trainings, committees (including cooks), and pastors received monthly mailings over the 15 month intervention period. Each mailing focused on either physical activity or healthy eating, and highlighted a health behavior change strategy consistent with the social cognitive theory ³¹ (e.g. social support, self monitoring, self-efficacy), and a health condition related to poor activity or dietary habits. Finally, study intervention staff made follow-up technical assistance calls to pastors, FAN coordinators, and cooks. The purpose of the calls was to learn what types of activities were being implemented, and to help problem-solve any challenges they were facing.

Measures

Sociodemographic and Health-related Variables.

Participants were asked to self-report their age, gender, race, marital status, total household income, highest grade or years of education completed, and rated their general health status on a scale from 1 (excellent) to 5 (poor). Participants also reported the number of times per month they attended worship services and other church activities or meetings, excluding worship service.

Fruit and Vegetable Intake.

The National Cancer Institute (NCI) fruit and vegetable all-day screener measured fruit and vegetable consumption (cups/day) ³². Nine of the original ten items were used (French fry consumption was excluded) ³³. Participants were asked about the types and quantity of fruits and vegetables consumed in the past month. This instrument correlates moderately with 24-hour recall measures (Men: r = 0.66; Women: r = 0.51) ³⁴, and is similar to one used in another faith-based intervention with African Americans ¹⁷ that also showed a moderate correlation with 3-day food records (r = .51).

Self efficacy for Fruit and Vegetable Consumption.

Self-efficacy for fruit and vegetable consumption was measured with a 10-item scale adopted from the Sallis et al. ³⁵ scale and used in two other faith-based projects ^19,20,36. On a scale of 1 (not at all confident) to 4 (very confident), participants were asked how confident, in the next 6 months, he/she could eat fruits and vegetables when faced with common barriers.

Social support for Fruit and Vegetable Consumption.

Social support for fruit and vegetable consumption over the past 12 months from family, friends or work colleagues, and people at church were each measured with a 3-item scale. On a scale from 1 (none) to 4 (a lot), participants were asked how much encouragement they got from family/friends or work colleagues/members of church to eat more fruits and vegetables. The items used to assess family and friend/colleague support were derived from a study by Eyler et al. ³⁷ involving minority women, which were adapted from the Sallis et al. ³⁸ scale. The items assessing support from church members were similar to those used in another faith-based project ¹⁹.

Church Support for Healthy Eating.

Because an existing church support scale was not available in the literature, we developed six items that assessed church support for healthy eating ³⁹ 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 ^40,41, input from church leaders and lay members, and the guiding theory for our intervention ²⁹. All items used a four point response scale ranging from 1 (rarely or never) to 4 (most or all of the time).

Physical Activity.

The Community Health Activities Model Program for Seniors (CHAMPS) questionnaire ⁴² assessed leisure-time moderate to vigorous physical activity. It assesses the frequency and duration of various physical activities completed “in a typical week during the past 4 weeks.” This measure has been shown to be valid ⁴³, have acceptable test-retest reliability ⁴³, and be sensitive to change ^42,44–47. A 36-item modified version, similar to the one described by Resnicow et al. ⁴⁸, was used in this study. Hours per week of leisure-time moderate to vigorous physical activity (≥ 3.0 METs, with the removal of household and related activities) was calculated.

Perceived Stress.

A 4-item version of the Perceived Stress Scale ^49,50 measured the degree to which situations in one’s life are appraised as stressful. On a scale from 1 (never) to 5 (very often), participants were asked how often, in the last month, he/she felt or thought a certain way.

Body Mass Index (BMI).

Height to the nearest quarter inch and weight to the nearest 1/10 kilogram were obtained by trained staff. Body mass index (BMI) was calculated as kg/m2 using standard procedures.

Diabetes.

Self-reported presence of diabetes was assessed by asking participants, “Have you ever been told by a doctor, nurse, or other health professional that you had diabetes ⁵¹?” Participants answering “yes” or “yes, but only during pregnancy” were considered to have diabetes.

Hypertension.

Resting blood pressure was taken three times on the right arm, after participants sat quietly for five minutes,⁵² with the automated DinaMap ProCare Monitor (DPC-100X-EN).⁵³ The average of the second and third measures was used. Because participants may have controlled hypertension, self-reported presence or absence of hypertension was also assessed by asking participants, “Have you ever been told by a doctor, nurse, or other health professional that you had high blood pressure ⁵¹ ?” Participants with a systolic blood pressure >140 mmHg, a diastolic blood pressure >90 mmHg, or answering “yes” to the self-report question were classified as hypertensive.

Statistical Analyses

A square root transformation corrected skewness in both baseline and post-program fruit and vegetable consumption scores. Differences in baseline demographic and health-related variables among those completing and not completing both pre and post-program fruit and vegetable consumption measures were assessed with chi-squares and t-tests.

Analysis of covariance (ANCOVA) examined baseline predictors of post-test fruit and vegetable consumption. SAS PROC MIXED was used to control for church clustering. Post-program fruit and vegetable consumption was the dependent variable in all analyses, and the baseline predictor variable of interest was the independent variable. A separate model was conducted for each baseline predictor examined. Baseline fruit and vegetable consumption, age, gender, education, church wave, church size, and intervention group were added as covariates to analyses that did not include these variables.

Results

Of the 1257 participants from 74 churches enrolled in FAN and included in the primary outcomes paper ²⁷, 1186 participants from 74 churches completed the fruit and vegetable consumption measure at baseline. Of these 1186 participants, 627 participants from 68 churches had complete post-program fruit and vegetable consumption data. Those with complete pre and post-program fruit and vegetable consumption data were older (p<0.0001), reported attending more church activities/meetings (p=0.01), participated in less leisure time physical activity (p=0.01), reported great self efficacy (p=0.003), were more likely to be married (p=0.04) and have hypertension (p=0.004) at baseline than those without complete data.

As shown in Table 1, the mean age of participants in this study was 57.4±12.3 years and the mean BMI was 32.6±7.3 kg/m². A majority of participants were female (76.2%), married (56.6%), had total household incomes less than $40,000 (57.7%), had at least some college education (57.1%), and were overweight or obese according to their BMI (88.1%). At baseline, participants consumed 3.9±3.7 cups of fruits and vegetables per day and engaged in 3.6±5.0 hours of leisure-time physical activity a week.

Table 1.

Demographic and Health-related Characteristics of Participants (n=627)

	N	Mean (SD) or %
Gender
Male	149	23.8
Female	478	76.2
Education
Less than high school	63	10.1
High school graduate	206	32.9
Some college	173	27.6
College graduate	185	29.5
Income
<$20,000	152	28.3
$20,000-$39,999	158	29.4
$40,000-$59,999	115	21.4
≥$60,000	112	20.9
Weight category
Normal weight (BMI <25)	73	11.9
Overweight (BMI ≥25>30)	175	28.4
Obese (BMI ≥30)	368	59.7
Marital Status
Married/ Member of unmarried couple	352	56.6
Not Married	270	43.4
Health Status
Excellent	37	6.0
Very good	142	22.9
Good	327	52.8
Fair	104	16.8
Poor	9	1.5
Hypertension
Yes	424	68.6
No	194	31.4
Diabetes
Yes	157	25.6
No	457	74.4
Age, years	627	57.4 (12.3)
Body Mass Index, m/kg2	616	32.6 (7.3)
Waist Circumference, cm	619	97.3 (14.8)
Fruit and Vegetable Consumption, cups/day	627	3.9 (3.7)
Leisure time Physical Activity, hours/week	624	3.6 (5.0)
Perceived stress1	615	2.3 (0.7)
Church Support for F&V2	608	2.2 0.7)
Social Support for F&V2	617	2.5 (0.9)
Self efficacy for F&V2	616	3.2 (0.7)
Attend worship service (times/month)	622	5.5 (3.8)
Attend church activities or meetings (time/month)	620	4.2 (3.7)

¹Range 1–5; lower scores indicate less stress

²Range 1–5; higher scores indicate more support

Mean adjusted post-test fruit and vegetable consumption (for categorical variables), estimates, standard errors, and p-values for each baseline predictor variable examined are shown in Table 2. Higher post-test fruit and vegetable consumption was significantly higher in obese participants compared to both overweight (p=0.04) and normal weight (p=0.03) participants, and in those without diabetes (p=0.04) at baseline. Participants with a greater waist circumference (p=0.04), higher fruit and vegetable consumption (p<0.0001), higher levels of leisure time physical activity (p=0.01), greater social support (p=0.01) and higher worship service attendance (p=0.04) at baseline also had greater fruit and vegetable consumption at post-test. There was a borderline significant relationship between post-test fruit and vegetable consumption and greater church support (p=0.06) and higher attendance at church activities/meetings (p=0.06). Gender, education, income, marital status, health status, hypertension, age, BMI, stress, and self-efficacy at baseline were not associated with post-test fruit and vegetable consumption (all p-values >0.05).

Table 2.

Baseline Predictors of Change in Fruit and Vegetable Consumption

	N	Adjusted Post-test Mean (cups/day)	Estimate (SE)	p-value
Gender	627
Male		1.77 (0.06)	0.03 (0.06)	0.62
Female		1.74 (0.04)	0.0 (Reference)
Education	627
Less than high school		1.84 (0.09)	0.0 (Reference)
High school graduate		1.68 (0.05)	−0.16 (0.10)	0.16
Some college		1.70 (0.06)	−0.14 (0.10)
College graduate		1.80 (0.06)	−0.04 (0.10)
Income	537
<$20,000		1.77 (0.06)	0.0 (Reference)
$20,000-$39,999		1.76 (0.06)	−0.003 (0.08)	0.82
$40,000-$59,999		1.70 (0.07)	−0.07 (0.09)
≥$60,000		1.76 (0.07)	−0.01 (0.10)
Weight category	616			0.03
Normal weight		1.63 (0.08)	−0.19 (0.09)
Overweight		1.70 (0.06)	−0.13 (0.06)
Obese		1.83 (0.05)	0 (Reference)
Marital Status	622
Married/ Member of unmarried couple		1.72 (0.05)	0.0 (Reference)	0.20
Not Married		1.79 (0.05)	0.07 (0.06)
Health Status	619			0.79
Excellent		1.77 (0.11)	0.11 (0.25)
Very good		1.81 (0.07)	0.15 (0.23)
Good		1.76 (0.05)	0.10 (0.23)
Fair		1.70 (0.07)	0.04 (0.24)
Poor		1.66 (0.23)	0.0 (Reference)
Hypertension	618
Yes		1.77 (0.05)	0.0 (Reference)	0.35
No		1.71 (0.06)	−0.06 (0.06)
Diabetes	614
Yes		1.66 (0.06)	0.0 (Reference)	0.04
No		1.79 (0.04)	0.13 (0.06)
Age, years	627	N/A	0.004 (0.002)	0.08
Body Mass Index. m/kg2	616	N/A	0.01 (0.004)	0.17
Waist Circumference, cm	619	N/A	0.004 (0.002)	0.04
Fruit and Vegetable Consumption, cups/day	627	N/A	0.41 (0.03)	<0.0001
Leisure time Physical Activity, hours/week	624	N/A	0.01 (0.01)	0.01
Stress	615	N/A	−0.02 (0.04)	0.65
Church Support for F&V	607	N/A	0.08 (0.04)	0.06
Social Support for F&V	617	N/A	0.08 (0.03)	0.01
Self efficacy for F&V	616	N/A	0.04 (0.04)	0.36
Attend worship service (times/month)	622	N/A	0.01 (0.01)	0.04
Attend church activities or meetings (time/month)	620	N/A	0.01 (0.01)	0.06

Note: The sample size is not 627 for all analyses due to missing data; baseline fruit and vegetable consumption, age, gender, education, church wave, church size, and intervention group were added as covariates to analyses that did not include these variables.

Discussion

A majority of African American adults do not eat the recommended daily amount of fruit and vegetables, perhaps contributing to the disparities in chronic disease the United States faces. Four major trials to date have focused on increasing fruit and vegetable consumption in African Americans through faith-based behavioral interventions: Eat for Life Trial¹⁸, Black Churches United for Better Health¹⁷, Healthy Body Healthy Spirit¹⁹, and Body and Soul²⁰. Although each intervention has successfully increased fruit and vegetable consumption, none of these studies have examined which (baseline) variables predict changes in intervention outcomes. The current study addresses a major gap in the current literature by examining which baseline variables predict change in fruit and vegetable consumption in a successful faith-based physical activity and dietary intervention targeting African Americans (FAN) ⁵⁴. The knowledge gained from this study should be considered when developing subsequent faith-based dietary interventions.

Predictor studies are valuable for helping to understand who is most and least likely to make targeted behavioral changes. Individuals lacking the characteristics associated with increased fruit and vegetable consumption can be identified from the outset, and steps to change the modifiable variables can be undertaken. Although non-modifiable predictors cannot be changed, different intervention approaches and/or more intense approaches can be used to counterbalance the unfavorable characteristics of the participant.

This study found that church members who had a greater waist circumference, greater baseline fruit and vegetable consumption, greater leisure-time physical activity, higher levels of social support, greater attendance at worship service, were obese, and did not have diabetes at baseline showed higher post-test fruit and vegetable consumption, after controlling for pre-test values. There were also borderline significant relationships between post-test fruit and vegetable consumption and higher baseline church support and attendance at church activities. These findings provide insight into who may benefit most from a faith-based intervention such as FAN, and has implications for future faith-based interventions targeting African American. Additional resources, activities, or materials may be necessary for individuals lacking suboptimal levels of the aforementioned indicators upon program entry. For example, individuals with suboptimal levels of social support or with low church attendance upon program entry can be identified early and intervened upon with additional, perhaps more intense, strategies. We were surprised to find the positive associations for waist circumference and obesity; it is possible that the total volume of food consumed by these individuals was higher, and therefore they inherently consumed greater amounts of fruits and vegetables.

No other faith-based studies have examined predictors of change in fruit and vegetable consumption; therefore caution should be taken when comparing our findings to other studies (i.e. different designs). Similar to our predictor findings, cross-sectional studies have found a positive association between fruit and vegetable consumption, social support ^23,25, and exercise ^25,26 in African Americans. There is cross-sectional evidence that risk behaviors, namely physical activity and fruit and vegetable consumption, may cluster in such a way that individuals with low physical activity also have low fruit and vegetable intake ⁵⁵. Perhaps engaging in one healthy behavior (i.e. physical activity) motivated improvements in another healthy behavior (i.e. fruit and vegetable consumption).

Participants with higher attendance at worship service and church activities (borderline) at baseline had greater increases in fruit and vegetable consumption. Worship attendance has been shown to be associated with healthy behaviors ⁵⁶. It is also likely that participants, at least in the intervention group, had greater exposure to FAN, which may have resulted in greater increases in fruit and vegetable consumption. This is in line with the findings of Resnicow and colleagues ²⁵ who found a positive association between church attendance and fruit and vegetable consumption and the findings of Campbell and colleagues ¹⁷ who found that church attendance over the study period was a strong predictor of increased fruit and vegetable consumption among individuals in the intervention group.

There was a borderline significant relationship between church support at baseline and fruit and vegetable consumption such that church members reporting higher support from his/her church for healthy eating had greater increases in consumption. Cross-sectional data from FAN also showed a positive relationship between church support and fruit and vegetable consumption ³⁹. Creating social and physical environments that promote good health for all is one goal of Healthy People 2020 ⁵⁷. A unique aspect of FAN was its focus on creating a church environment that supported healthy eating practices through getting the word out, providing opportunities, and developing policies and guidelines. Future faith-based studies may want to consider church support and its potential influence on changes in health behaviors such as fruit and vegetable consumption.

None of the demographic variables were associated with greater changes in fruit and vegetable consumption, and to our surprise, self efficacy was not associated with increases in fruit and vegetable consumption. Self efficacy has been shown to be one of the strongest predictors of fruit and vegetable consumption in adults ⁵⁸; cross sectional studies with African Americans have also found a positive relationship between self efficacy and fruit and vegetable consumption (cross-sectional) ^23,25. One possible explanation for the lack of relationship in this study was a restricted range, as overall, self-efficacy at baseline was fairly high in our sample.

Limitations, including the use of a self-report measure of fruit and vegetable intake, should be considered when interpreting the findings of this study. Although 24-hour recalls may be considered the gold standard for measuring dietary intake, this approach was not feasible given the size and scope of the study. In addition, the attrition rate in this study was higher than desirable, although in line with what other studies targeting African Americans have reported ⁵⁹.

This is the first faith-based study to examine baseline predictors of change in fruit and vegetables among African Americans and offer insight into who may be most (or least) likely to make changes. Non-significant predictors in this study should not be discounted, as this type of analysis in faith-based settings is in its infancy. Additional predictor studies will collectively help to identify individuals who may be more “at risk,” and subsequent interventions can incorporate more intense strategies and additional resources c to assist those who are less likely to change, ultimately leading to more effective interventions.