The Role of Attitude, Control and Intention to Explain Fruit and Vegetable Intake among Racial/Ethnic Minority Women with Low Socioeconomic Status

Special Issue – Eating Behavior Changes

Int J Nutr Sci. 2017; 2(1): 1015.

The Role of Attitude, Control and Intention to Explain Fruit and Vegetable Intake among Racial/Ethnic Minority Women with Low Socioeconomic Status

DeBiasse MA¹*, Bowen DJ², Pagoto SL³, Massaro JM4, Istfan N5 and Quintiliani LM6

¹Department of Health Sciences, Boston University, College of Health & Rehabilitation Sciences: Sargent College, USA

²Department of Bioethics and Humanities, University of Washington School of Medicine, USA

³Department of Medicine, Division of Preventive and Behavioral Medicine, University of Massachusetts Medical School, USA

4Department of Biostatistics, Boston University, School of Public Health, USA

5Department of Medicine, Boston University School of Medicine, USA

6Department of Medicine, Medical Information Systems Unit, Boston University School of Medicine, USA

*Corresponding author: DeBiasse Michele Ann, Department of Health Sciences, Boston University/ College of Health and Rehabilitation Sciences: Sargent College, Boston, USAa

Received: February 08, 2017; Accepted: February 21, 2017; Published: February 27, 2017

Abstract

Objective: Fruit and Vegetable (FV) intake-a modifiable risk factor for chronic diseases-is lower among racial/ethnic minorities and low Socio- Economic Status (SES) groups when compared to other populations. The Theory of Planned Behavior (TPB) is one theoretical model studied to explain and influence individual health behaviors, including FV intake, in middle class populations, but not exclusively in diverse, low SES groups. This cross-sectional study evaluated the utility of select TPB variables to explain intention to consume and intake of FV in this population.

Design: Demographics, BMI, select TPB variables, and FV intake were measured via survey. Bivariate analyses were conducted to explore relationships between variables. Hierarchical regression analyses were used to fit two models: one to explain intention and one to explain behavior with regard to FV vegetable intake.

Results: Participants (n=114) age 25-69 years and were mostly African American/Black and Hispanic (21.9% and 73%, respectively). The TPB variable perceived behavioral control was the only significant predictor of intention to consume FV (OR=2.55, 95% CI OR: 1.23, 5.27), and with BMI, FV intake (R2=0.08; F [2,130] =5.72, p=0.0042).

Conclusion: Perceived behavioral control and BMI are the most significant predictors of FV intake but explain only 8% of the variability in intake in our cohort. Our results support prior research which suggests an attenuation of the intention-behavior relationship by SES, and may question the utility of the TPB as it is currently operationalized as a foundational model for future health behavior change research and programs in low SES racial/ethnic minorities.

Keywords: Theory of planned behavior; Fruit and vegetable intake; Racial/ ethnic minorities; Low socioeconomic status; Public housing

Introduction

Increased consumption of fruit and vegetables is associated with decreased risk of cardiovascular disease, some cancers, and obesity [1,2]. Although the US government has suggested that American’s increase their intake of fruit and vegetables to five servings per day, epidemiologic data suggest that most Americans have not attained intake levels which meet these guidelines [3,4]. The strongest disparity in intake is seen in racial/ethnic minorities of low Socio-Economic Status (SES) compared to those of higher levels of income with 32.2% versus 32.9% consuming = 2 servings per day of fruit and 22.0% versus 29.4% consuming =3 servings of vegetables per day, respectively [5- 7]. Differences in fruit and vegetable intake between racial/ethnic and socioeconomic groups are thought to contribute to the higher incidence of cardiovascular disease and obesity that is seen in these vulnerable populations [8-11].

Under-consumption of fruit and vegetables is thought to be the result of both individual and environmental factors [12-14]. Theoretical models of behavior seek to understand individual-level factors which shape health behavior including diet [15,16]. One theoretical model that was developed and is being studied to explain health behaviors is Ajzen’s Theory of Planned Behavior (TPB) [17- 20]. The TPB is a model which extends the Theory of Reasoned Action (TRA) [21]. The TPB posits that intention (and ultimately behavior) can be predicted effectively by a person’s attitude (personal judgement about the behavior), subjective norms (other’s judgement about the behavior), and perceived behavioral control (the individual’s perceptions about whether or not he/she has the skills and means necessary to bring about successful performance of a goal intention). Perceived behavioral control has influence on both goal intention and behavior and its influence depends upon the type of behavior and the nature of the goal intention. In instances where attitudes are strong, or where subjective norms are powerful, perceived behavioral control may be less predictive of goal intentions. But, when these factors are reduced, perceived behavioral control may play a more powerful role in goal intention strength, and performance of behavior.

The TPB has been studied as a model to explain health behaviors in a number of groups, but not exclusively in racial/ethnic minority women of low SES. If the TPB could be used to explain health behaviors in this vulnerable population, we could then use the TPB as a foundation to develop interventions to improve health behaviors (like fruit and vegetable intake) which we know are performed with less frequency in racial/ethnic minorities and low SES groups. Although the TPB has been used to understand fruit and vegetable intake specifically in multi-ethnic and overweight/obese groups, to date we know of no study investigating how TPB variables explain fruit and vegetable intake exclusively in racial/ethnic minority women of low SES [22,23]. The goal of this cross-sectional study was to evaluate whether the TPB variables attitude, perceived behavioral control, and goal intention strength explain fruit and vegetable intake in a group of racial/ethnic minority women of low SES. We hypothesized that attitude and perceived behavioral control would be significantly associated with goal intention strength, and goal intention strength would be significantly associated with fruit and vegetable intake, consistent with the TPB. Testing these hypotheses will provide insight into some of the constructs of the Theory of Planned Behavior in racial/ethnic minority women of low SES; information we hope can then be used to guide the development of effective strategies to increase fruit and vegetable intake in these often-marginalized groups.

Methods

Participants and data collection

Participants for this ancillary study were the mothers from the evaluation cohort of a parent study (“Healthy Families”). Details of this study design have been published elsewhere [24]. For the current cross-sectional study, data were collected as part of the one year follow up survey for the Healthy Families study (May through November, 2014). In addition to a re-assessment of height and weight, attitude, perceived behavioral control, goal intention strength and fruit and vegetable intake were measured at this time. The study was approved by the Boston Medical Center Investigational Review Board.

Demographic measures

Age, education level, and race/ethnicity were collected via survey. Height was measured using a standard measuring tape, and weight was obtained using a digital scale.

Theory of planned behavior measures

Attitude, perceived behavioral control and goal intention strength were measured using 1 question each with a bipolar semantic differential scale for response as informed by the literature [25]. As a recent meta-analysis by Armitage and Connor (2001) as well as studies by Conner & Sparks (2005), Louis et al. (2009) and Emmanuel et al. (2012) indicate that subjective norms is the weakest predictor of intention in the TPB, and we were cognizant of the significant participant burden associated with our survey (over 125 questions) we decided to only included questions that we felt were essential to the study [26-29]. As such, we did not measure the TPB construct subjective norms.

Attitude was measured using the question: “For me, increasing my daily intake of fruit or vegetables by one serving every day over the next month is…” to which subjects could respond on a scale between 1 and 7 anchored with “foolish” or “wise”. Perceived behavioral control was measured using the question: “For me, increasing my daily intake of fruit or vegetables by one serving every day over the next month would be” to which subjects could respond on a scale between 1 and 7 anchored with “difficult” and “easy”. Goal intention strength was measured using the question: “I intend to increase my daily intake of fruit or vegetables by one serving every day for the next month.” Subjects could respond to this statement on a scale between 1 and 7 anchored with “disagree strongly” and “agree strongly”.

Fruit and vegetable intake measures

Fruit and vegetable intake was measured using the Behavioral Risk Factor Surveillance Survey Fruit and Vegetable Module (BRFSS FVM), a 6-item self-reported questionnaire measuring frequency of fruit and vegetables usually eaten during the past 30 days [30]. This measure has been validated in a similar population with a correlation between the BRFSS FVM and multiple 24-hour diet recalls of 0.46 [31]. The module includes an item designed to assess intake of potatoes excluding fried potatoes (e.g. French fries, potato chips) as well as an item to assess intake of green salad. A pictorial reference for serving size was added to the screener to improve its validity [31-33].

Analysis plan

Sample size was calculated based upon a medium effect size, 1-β=.80. α=.05, and 4 predictor variables to be 84 subjects. Bivariate correlational analyses were conducted to assess the strength and direction of relationships among demographic and TPB variables for addition into the multivariable models. Pearson product-moment correlation testing was used for normally-distributed continuous, and Spearmen’s rank correlation coefficient was used for categorical and ordinal measures. Cochran-Mantel-Haenszel testing was conducted to test for associations between education and TPB variable scores, and ANOVA was used to test for associations between education and mean fruit and vegetable intake. Strength of associations was assessed as suggested by Dancey & Reidy (2004) [34].

Following the bivariate analyses, hierarchical regression analyses were conducted to examine the utility of the TPB to explain both goal intention strength (intention) and intake of fruit and vegetables (behavior). Covariate additions to these models were guided by the aforementioned bivariate correlational analyses as well as the tenants of the TPB.

In the first series of hierarchical regression models, goal intention strength (dependent variable) was dichotomized into those who scored a “7” (most positive) for goal intention strength, and those who scored less positive (“1” through “6”). At step 1, demographic variables and BMI were entered. In addition, an indicator variable for housing development residence was entered to control for potential influence of the parent study. At step 2, attitude and perceived behavioral control were entered.

Prior to performing the second series of hierarchical regression analyses, model assumptions were checked. As it was determined that both the linearity and homoscedasticity assumptions were violated, fruit and vegetable intake was transformed using the natural log, and TPB predictor variables were dichotomized with score 1 through 6 assigned the value zero (0) and score 7 assigned the value one (1). At step 1, demographic variables, BMI and an indicator variable for housing development residence was entered. At step 2, goal intention strength and perceived behavioral control were entered.

For all tests, significance was set at α = 0.05. Data were analyzed using SAS 9.3© [35].

Results

Participant characteristics

Participants who completed the 1-year follow up survey for the parent study were included in this analysis (n=144; 68% of the total sample) (Table 1). Participants had a mean age of 38.7 (±7.85), and ranged from 25-69 years of age. Hispanic/Latina subjects comprised the majority race/ethnicity (73%), followed by Black/ African American (21.9%), White (5%) and Asian (1.5%). Most of the participants had a less than high school education (32%) or were high school/GED educated (33.3%). More than three quarters of subjects had a BMI which classified them as either overweight or obese (79%). The mean servings per day of fruit and vegetables for the group was 3.49 (SD=2.16).