Abstract
Epidemiological evidence of the protective role of fruits and vegetables for a host of chronic health conditions is well-documented. However, there is a dearth of studies examining predictors of fruit and vegetable intake among African American men living with HIV. We report secondary analyses—multiple regression and logistic regression models fitted to examine the strength of the relationships between the reasoned action approach constructs; namely, attitudes, subjective norms, descriptive norms, self-efficacy and intention to consume fruits and vegetables, and self-reported adherence to 5-A-DAY guidelines. We used baseline data from a randomized controlled trial of a physical activity intervention trial with 302 African American men aged 40 years or older (M = 53.9; SD = 7.2) living with HIV. Attitudes, subjective norms, descriptive norms, and self-efficacy were positively associated with intention to meet 5-A-DAY guidelines. More positive attitudes toward 5-A-DAY guidelines were associated with higher odds of meeting 5-A-DAY guidelines. More positive attitudes and self-efficacy were also positively associated with meeting the guidelines for intake of vegetable servings and fruit-and-vegetable servings combined. To increase fruit and vegetable intake among African American men living with HIV, interventions should be tailored to address the perceived benefits of consumption.
The benefits of a healthy dietary pattern on a range of cardiometabolic and other disease outcomes are well-documented (Virani et al., 2021). According to the Dietary Guidelines for Americans (U.S. Department of Health and Human Services, 2015), a healthy diet is higher in vegetables, fruits, whole grains, and low-fat or nonfat dairy, and low in sugar-sweetened foods and drinks. Dietary recommendations specifically include fruits and vegetables due to their concentrations of vitamins, minerals, phytochemicals, and dietary fiber (Slavin & Lloyd, 2012). Consuming more fruits and vegetables in place of more energy-dense foods (i.e., foods high in fat with low water content) adds nutrients to the diet, reduces disease risk, and helps manage body weight (Moore & Thompson, 2015). Empirical evidence also suggests that diets high in fruits and vegetables are associated with a decreased odds of mortality from cardiovascular disease, cancer, and other chronic diseases (Sharma et al., 2014). Despite these known benefits, only 12% of adults meet the recommended daily intake of fruits, and just 9.3% meet the recommended daily intake for vegetables (Centers for Disease Control and Prevention, 2018b). Moreover, fruit and vegetable consumption was lower among African Americans, men, and adults living in poverty (Centers for Disease Control and Prevention, 2018b). A low frequency of fruit and vegetable consumption has also been noted among individuals living with HIV (Gastao, 2019).
A better understanding of fruit and vegetable intake among African American men in general, and particularly among older men living with HIV is needed. It has been shown that older people and African American men have a higher risk of chronic noncommunicable diseases (Quiñones et al., 2019). African American men account for 39% of new HIV diagnoses among men, a higher percent than Hispanics and non-Hispanic White males, 28%, and 27%, respectively (Centers for Disease Control and Prevention, 2018a). Also, compared with Mexican American and non-Hispanic White males, African American men display lower consumption patterns of fruits and vegetables (Virani et al., 2021). For instance, African American men accounted for the lowest proportion of men meeting fruit and vegetable recommendations (8.9% and 2.2%, respectively), compared with Mexican American men (16.5% and 3.2%, respectively) and non-Hispanic White men (13.7% and 6.4%, respectively; Virani et al., 2021). This pattern of low intake of fruits and vegetables among African Americans is due to a diet characterized by high-fat, processed meats, and fried foods (Judd et al., 2013). A study on the psychosocial correlates of fruit and vegetable intake among African American men found that fruit consumption appeared to be motivated by perceived benefits and normative beliefs of significant referents. In contrast, extrinsic rewards and preferences for high-calorie, fatty foods influenced vegetable consumption (Moser et al., 2005).
A healthy diet is crucial for people living with HIV. Maintaining good nutrition plays a vital role in enhancing the immune system, absorbing HIV medications, and improving quality of life (U.S. Department of Health and Human Services, 2020). There is an abundance of evidence on the association between diet quality and quality of health outcomes for people living with HIV, including being overweight (Duran et al., 2008), disease progression (Silva et al., 2010) and low CD4 count and anemia (Rawat et al., 2013). However, there is limited research on the diet quality of African American men living with HIV. Evidence from a recent study suggests that African Americans living with HIV consume fewer than the recommended number of servings of fruits and vegetables (Hernandez et al., 2017; Segal-Isaacson et al., 2006). Therefore, it is crucial to identify predictors of fruit and vegetable intake, which can be used to inform interventions tailored to improve the health status of African American men living with HIV.
The RAA (Fishbein, 2008; Fishbein & Ajzen, 2010) has been widely used to understand a variety of health behaviors, including determinants of fruit and vegetable consumption (Hackman & Knowlden, 2014; Sheats et al., 2013; Sheats & Middlestadt, 2013). According to the RAA, intentions are the foundation of behaviors, and these intentions are determined by attitudes, subjective norms, descriptive norms, and perceived behavioral control regarding the behavior. While several studies (Barnidge et al., 2015; Condrasky et al., 2013; Griffith et al., 2016; Ong et al., 2017) exist on adult fruit and vegetable behavior, there is a paucity of studies on fruit and vegetable intake specific to African American men living with HIV. Moreover, a review of the literature identified only one (Kelly et al., 2021) study utilizing the RAA constructs to examine health-promoting behaviors among low-income urban African American men living with HIV. To fill this gap in the literature, the purpose of this study was to examine the utility of the RAA for predicting 5-A-DAYdiet intentions and behavior among older African American men living with HIV.
Method
This current study was reviewed and approved by the University of Pennsylvania Institutional Review Board. Participants volunteered for a randomized controlled trial testing the efficacy of a health promotion intervention. The intervention was designed to increase physical activity and healthy diet behaviors, among African American men aged 40 years or older living with HIV. Participants were recruited from the University of Pennsylvania’s CFAR Clinical Core Adult/Adolescent Database, which includes data on people living with HIV who agreed to be contacted about participating in research studies. Recruitment was initiated with referrals of providers at the Penn Presbyterian HIV Clinic at the University of Pennsylvania, Drexel University/Partnership Clinics, and AIDS service organizations in Philadelphia. Recruitment ads were placed on social media sites, including Facebook, Twitter, and Craig’s list, and a local newspaper.
Men who were 40 years of age or older, identified themselves as Black or African American, and were receiving ART (Antiretroviral Therapy) for HIV, were eligible to participate in this study. We used the referral of an HIV care provider or proof of a prescription for ART to determine whether participants fit the inclusion criterion of receiving ART for HIV. We excluded men who had a blood pressure of 180/110 mm Hg or higher or who participated in any intervention trials in the past 12 months targeting physical activity, diet, or prostate or colon cancer screening. Due to potential difficulties in follow-up data collection, participants were excluded from the study if they were planning to relocate during the next 18 months or did not have a mailing address.
Measures
Self-report assessments via audio computer-assisted self-interviewing (ACASI) were administered. ACASI ensures the integrity of the data by integrating data quality-control measures into the data-collection process. It is a uniform and consistent data-collection method, which improves the quality of low literacy respondents’ self-report data. People may be reluctant to report engaging in socially sanctioned behaviors, or to express unpopular opinions, because they believe they will be judged negatively. Evidence indicates people feel that ACASI provides more privacy and say it elicits more honest responses than face-to-face interviews (Metzger et al., 2000). Several studies have found people report more socially undesirable behaviors when questioned with ACASI than face-to-face interviews and paper-and-pencil surveys (Ghanem et al., 2005; Metzger et al., 2000; Turner et al., 1998).
Reasoned Action Approach Constructs
Intention to consume at least five servings of fruits and vegetables was measured with three Likert items, rated on a 5-point scale from disagree strongly to agree strongly (α = .83). Attitude toward consuming at least five servings of fruits and vegetables consisted of five items, rated on a 5-point bipolar adjective scale (bad–good, foolish–wise, unpleasant–pleasant, dangerous–safe; α = .85). Subjective norms regarding consuming at least five servings of fruits and vegetables consisted of three items, rated on a 5-point scale from disagree strongly to agree strongly (α = .82). Descriptive norms regarding consuming at least five servings of fruits and vegetables was assessed by a single item, concerning the number of their five closest friends who ate at least five servings of fruits and vegetables each day, rated on a scale of 0 to 5. Self-efficacy to consume fruits and vegetables was assessed using 15 Likert items rated on a 3-point scale (α = .92).
Adherence to 5-A-DAY Diet Guidelines
We assessed adherence to 5-A-DAY guidelines with the seven-item food frequency questionnaire developed by the National Cancer Institute (Thompson & Byers, 1994). It included three items on fruit consumption and four on vegetable consumption (Campbell et al., 1999). A binary variable was calculated based on whether the participants met the 5-A-DAY guideline of consuming five or more servings of fruit and vegetables each day during the prior 30 days. We also examined the number of daily servings of fruits, vegetables, and fruits and vegetables combined. Additionally, participants reported their consumption of fatty or fried food during the previous 30 days.
Sociodemographic Characteristics
Participants also reported sociodemographic characteristics such as their age and marital status. To assess their education levels, we used the following question: “What is the highest grade of school you completed?” (El-Bassel et al., 2011). Possible options to the question were as follows: “no formal schooling,” “less than a high school diploma,” “a high school diploma (or GED),” “some college or a 2-year degree,” “4-year college degree,” or “postgraduate work.” To assess the participants’ income levels, we asked the following question: “What is your total monthly income (not including your spouse)?” (El-Bassel et al., 2011), which was followed by seven possible responses shown in Table 1. The question measuring employment status was, “Are you employed?”
Baseline Characteristics of African American Men Living With HIV, Philadelphia, PA.
Statistical Analysis Plan
We used descriptive statistics to assess the sociodemographic characteristics of the participants. Bivariate relations between RAA variables and diet outcomes were analyzed using Pearson product-moment correlation coefficients. Depending on whether the dependent variable was continuous or binary, multiple linear or logistic regression models were used to examine the relation of theoretical variables considered together as predictors. According to the RAA, the predictors of a behavior may vary from population to population, meaning external variables may moderate the relationships of the theoretical variables to intention and behavior. Accordingly, we also conducted moderation analyses to determine whether the predictors intention and behavior differed depending on three potential moderators, age, income, and education. Finding significant moderator effects would suggest different interventions might be needed depending on the level of the moderator. The moderating effect of age was examined by adding age and the interactions between age and each predictor to the model predicting intention or one of the diet behaviors. Moderation by income and moderation by education were also examined using the same method. SAS 9.4 was used for all analyses.
Results
The sociodemographic characteristics of the participants are shown in Table 1. The average age of participants was 53.88 years (SD = 7.24). About 76% had obtained at least a high school diploma, and 11% reported they were currently married. The majority (67%) earned less than $851 per month, and the proportion of unemployed participants was 83%. Participants reported that they ate 1.40 servings of fruits (SD = 1.40) and 1.11 servings of vegetables (SD = 1.11) per day on average, resulting in 2.51 servings of fruits and vegetables combined per day on average (SD = 3.14).
Consistent with RAA, attitudes, subjective norms, descriptive norms, and self-efficacy to 5-A-DAY diet were all positively correlated with the intention to consume fruits and vegetables, as shown in Table 2. There were significant correlations between the number of fruit and vegetable servings combined and three RAA constructs: 5-A-DAY diet intention, attitude to follow 5-A-DAY diet, and self-efficacy. The number of vegetable servings was also correlated with intentions, attitudes, and self-efficacy to 5-A-DAY diet. The number of fruit servings was significantly correlated with the 5-A-DAY diet attitudes. However, statistical significance was not detected for the correlations between meeting the 5-A-DAY diet guideline and the RAA constructs.
Zero-Order Correlations (p Value) Among Reasoned Action Approach Variables and 5-A-DAY Diet Measures, African American Men Living With HIV, Philadelphia, PA (N = 302).
As multiple regression analyses results show in Table 3, all predictors of RAA—attitudes, subjective norms, descriptive norms, and self-efficacy—were positively associated with the intention to consume a 5-A-DAY diet. Odds ratios in Table 4 indicate that more positive attitudes were associated with greater odds of meeting the 5-A-DAY diet guidelines, adjusting for the other RAA constructs, but the associations of other RAA constructs to meeting the guideline were not statistically significant. Table 4 also presents regression analyses separately for the number of fruit servings, the number of vegetable servings, and the combined fruit and vegetable servings. Attitudes and self-efficacy were positively associated with the number of vegetable servings, while the associations between RAA constructs and fruit servings were not statistically significant. The 5-A-DAY diet attitude and self-efficacy were significantly correlated with the number of fruit and vegetable servings combined.
Multiple Regression Analysis of Reasoned Action Approach Predictors of Intention to 5-A-DAY Diet, African American Men Living With HIV, Philadelphia, PA (N = 302).
Odds Ratios and Regression Coefficients From Regression Analyses of Relation of Reasoned Action Approach Predictors to Meeting 5-A-DAY Diet Guidelines, Fruit Servings, Vegetable Servings, and Fruit and Vegetable Servings Combined, African American Men Living With HIV, Philadelphia, PA (N = 302).
Moderation
We examined the moderation of age, income, and education. First, we estimated a multiple regression model to test the moderation by age in the prediction of 5-A-DAY diet intention. There were no significant interactions between the RAA constructs and age. Also, age was not a significant moderator in the three multiple regression models predicting fruit servings, vegetable servings, and fruit and vegetable servings combined, respectively. No significant moderation by age was found in a model predicting meeting 5-A-DAY diet guidelines.
Second, we examined the moderation by income. Using a multiple regression model, we found that income significantly moderated the impact of 5-A-DAY diet attitude on the intention to adhere to 5-A-DAY diet. To be specific, the effect of attitude on intention was stronger for participants with low income (coefficient of the interaction between attitude and income: β = −.20, SEβ = .090, p = .029), although the effect of 5-A-DAY diet attitude on intention was positive irrespective of income. Interactions were not statistically significant between the other RAA constructs and the intention in this model. Moderation by income was not found in the multiple regression models predicting fruit servings, vegetable servings, and combined fruit and vegetable servings. No significant moderation by income was found in a model predicting meeting 5-A-DAY diet guidelines.
Last, we tested the moderation by education. Education significantly moderated the effect of self-efficacy on vegetable servings. Specifically, the effect of self-efficacy on vegetable servings was positive for participants with high education; in contrast, it was negative for participants with low education (coefficient of the interaction between self-efficacy and education: β = .45, SEβ = 0.20, p = .026). In this model, interactions were not statistically significant between the other RAA constructs and education. Also, education significantly moderated the effect of self-efficacy on fruit and vegetable servings combined. Participants with high education showed a positive effect of self-efficacy on fruit–vegetable combined servings. In contrast, the impact was negative for participants with low education (coefficient of the interaction between self-efficacy, and fruit and vegetable servings combined: β = .88, SEβ = .39, p = .027), while interactions between the other RAA constructs and education in this model were not significant. Moderation by education was not found in multiple regression models predicting 5-A-DAY diet intention, fruit servings, and meeting the guidelines, respectively.
Discussion
The results demonstrate the utility of the RAA to predict 5-A-DAY intentions and behavior among low-income urban African American men living with HIV, a population characterized by striking health disparities. Consistent with the RAA, attitudes, subjective norms, descriptive norms, and self-efficacy were positively related to intentions to meet the 5-A-DAY guidelines. Moreover, attitude and self-efficacy were positively associated with vegetable intake and combined fruit and vegetable intake. These results demonstrate that the RAA was an accurate and robust model for predicting 5-A-DAY intentions and fruit and vegetable consumption.
Although other studies have examined the relationship of age, income, and education to a healthy diet, to our knowledge, this study is the first to consider whether these variables moderate the relation of theoretical constructs to a healthy diet among people living with HIV. An awareness of moderator effects is essential because such effects imply that the underlying mechanism of decision making and behavior differ depending on the level of the moderator. Therefore, different intervention strategies may be needed. We found that despite the broad age range in the sample, the prediction of intention and behavior did not vary by age. On the other hand, we found that attitudes were a stronger predictor of intention among low-income individuals. In contrast, self-efficacy was a stronger predictor of meeting 5-A-DAY guidelines among people with higher educational attainment. This pattern of findings suggests that interventions for economically marginalized African American men living with HIV should target improving attitudes about fruit and vegetable consumption, and interventions for those with higher educational status should target improving self-efficacy to adhere to the 5-A-DAY diet.
Consistent with findings from the study by Moser et al. (2005) on psychosocial correlates of fruit and vegetable consumption among African American men, perceived benefits, assessed as attitudes in the RAA, were a significant predictor for vegetable intake. But unlike Moser et al. (2005) norms were not a significant correlate of consumption of fruit or vegetable intake in the current study. This finding suggests that perhaps men in our study did not have salient referents influencing their beliefs about the important role of diet and/or did not perceive 5-A-DAY diet behaviors to be important among other men living with HIV. Therefore, we can surmise that interventions targeting men living with HIV need to intervene at the individual and peer-group levels. Furthermore, divergent from Moser et al. (2005), we found that self-efficacy was a significant vegetable intake predictor. Our findings on self-efficacy are consonant with those reported by Kelly et al. (2020) in their cross-sectional study about fruit and vegetable intake intentions among emerging adult Black women.
Several limitations should be considered when interpreting the results of this study. This study used a cross-sectional design; therefore, we cannot establish the directionality of the observed relationships. Longitudinal studies are necessary to establish if RAA variables predict adherence to 5-A-DAY guidelines. The use of a seven-item food frequency questionnaire is somewhat limited; therefore, future studies should use a more comprehensive measure of diet intake. Although our use of ACASI was a strength compared with using pencil-and-paper or interviewer-administered questionnaire, objective measures would have further reduced the possibility of self-report biases. Additionally, this study used a convenience sample of primarily low-income, unemployed, African American men aged 40 years and older living with HIV; therefore, the generalizability of the findings to different populations is limited. Despite these limitations, our study is one of the first to examine 5-A-DAY diet adherence among African American men living with HIV. The results of this study demonstrate the utility of the RAA in explaining 5-A-DAY intentions among African American men living with HIV and suggest the potential role of attitude and self-efficacy in interventions to increase fruit and vegetable intake. Numerous studies have revealed a strong link between high intakes of fruits and vegetables and a decreased risk of chronic diseases. Given that nutrition is a modifiable predictor of health, interventions to improve clinical outcomes for African American men living with HIV should employ strategies aimed at increasing fruit and vegetable intake. This study highlights the need for policy makers and public health professionals to develop diet-focused interventions to enhance attitudes and self-efficacy for different subgroups of African American men living with HIV to improve their nutritional status and quality of life.
Footnotes
Acknowledgements
The authors gratefully acknowledge the contributions of Scarlett Bellamy, PhD, Bruce Clifton, Mikia Croom, MEd, Ian Frank, MD, Janet Hsu, Ann O’Leary, PhD, Pandora Woods, and Brian Taylor. The authors also thank the men who participated in this research.
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: The research reported in this article was supported, in part, by grant R01 MD006232 (PI: John B. Jemmott, III) from the National Institute of Minority Health and Health Disparities.
