In this sample of low- to moderate-income AA women residing in Marion County, Indiana (located in the mid-western region of the US), results indicated that most participants (80%) ate DGLV in the past 7 days and, on average, reported adequate intakes (ie, 1.5 cups in the past 3 days). Female sex and higher education level (ie, 57% completed at least 1 year of college) have been associated with general FV consumption in other samples.38,39
As previously noted, studies have identified DGLV to be commonly consumed among some, though not all, populations of AA. Although beyond the scope of the current study, regionally specific cultural practices should be assessed in the future, as they could contribute to a better understanding of DGLV intakes across different AA samples.16–18,40
Behavioral intention is an important determinant of nutrition-related behaviors,41
and participants in this study had strong intentions to buy and eat more DGLV each week over the subsequent 3 months. Yet correlations between reported consumption and intention to buy (r
= 0.20, P
< .001) and intention to eat (r
= 0.23, P
< .001) more DGLV were relatively low. Given that intention was assessed in terms of eating and buying “more” DGLV (vs a specific amount), intentions may differ if a more stringent behavioral criterion is applied. Additionally, the weakness of the behavior–intention associations not only supports the notion that one's intention to perform a behavior may not lead to action, but also highlights the need to consider other factors—some of which, such as local access or quality, may go beyond the individual.20,21,42
The current study's results also show that the stronger the intentions women had for each behavior, the stronger their attitudes and self-efficacy were about performing the behaviors of interest. Further, intention, attitude, and self-efficacy to buy were strongly associated with intention, attitude, and self-efficacy to eat. For both behaviors, attitude had the greatest weight in explaining intention, as the weight for self-efficacy was somewhat smaller. Although perceived norm was significantly associated with intention, it was not significant in the final regression analyses for either behavior. Study results for attitude and self-efficacy are consistent with previous research identifying them as important intrapersonal constructs for AA and other racial/ethnic groups' intentions to eat fruit, vegetables, and other healthful food.33,43–45
On the one hand, the significant correlation of perceived norm with intention is consistent with research showing that spouses and other family members influence food choices and purchasing among AA and other groups.12,46
On the other hand, the multivariate finding that perceived norm did not contribute when attitude and self-efficacy were considered is consistent with studies examining more general fruit and vegetable consumption47
and intention to eat healthier food.48
Clearly, the role of perceived norm deserves further attention.
The study results suggest that RAA-based interventions designed to promote increased buying and consumption of DGLV for AA women living in this region of the US may be more effective if they focus primarily on changing attitudes and increasing self-efficacy. For example, improving attitude may be addressed by identifying and addressing beliefs held by AA women about the advantages and disadvantages of buying and eating more DGLV. Self-efficacy may be addressed by improving skills related to buying and eating, such as cooking demonstrations and the provision of recipes showing healthy ways to prepare DGLV. Self-efficacy may also be addressed by altering features of the environment that make DGLV more accessible and easier to buy. In order to improve attitudes and self-efficacy about eating and buying more DGLV, a greater understanding of why these vegetables are healthful by the population of interest may be required, as well. For example, Lynch, Holmes, Keim, and Koneman found that among a sample of AA women, explanations of why and which food items are healthful differed within the sample.49
The use of qualitative methods to understand decisional processes around the behaviors of interest, such as those used by Reicks and colleagues, may also provide insight.50
Such strategies to understand and address attitudes and self-efficacy highlight the potential for integrative, multilevel public health interventions and/or campaigns designed to reach the target population. Finally, studies have shown strong positive associations between significant others/caregivers buying and preparing healthy food and family members (eg, men, adolescent girls) subsequently eating and purchasing healthy food.25,51
With over half of the sample reporting children living in their household, most participants perceived that all household members liked DGLV, and nearly two thirds of households ate their primary meal of the day together. These observations suggest that interventions that stimulate increased DGLV-related behaviors among this population may have particularly positive implications for other household members, relative to other populations. Given that family support has had a significant influence on eating behaviors of AA women, it may be beneficial to target and/or include household members in dietary intervention programs, as well.52
This study has several limitations. First, the study was cross-sectional, so causation cannot be inferred. Second, a sample of convenience was used, and study results should not be generalized beyond the characteristics of the group under study. Third, the relatively weaker reliability of the perceived norm measures may have had an impact on study the results. Following methods described by Fishbein and Azjen,29
for each behavior, there was an injunctive item about what other people think (eg, “Most people who are important to me think that I should buy more DGLV each week over the next 3 months”) and descriptive item about what other people do (eg, “Most people like me will buy more DGLV each week over the next 3 months”). Although they both make up perceived norm, they are slightly different facets. Therefore, to improve reliability, additional injunctive and descriptive items could be developed for each behavior. Fourth, data were collected during the summer months, which may have increased the number and variety of DGLV available. Finally, the dependent variables for the regression models were behavioral intentions to eat and to buy DGLV, and they may not be strongly linked with the actual behaviors of interest.
Limitations notwithstanding, the current study addressed a number of gaps in the literature by focusing on a less-often-studied subgroup of vegetables; examining factors potentially linked with buying and eating DGLV among AAs; and using a theory-based survey instrument. The RAA measures showed acceptable reliability,53,54
and research demonstrates that intention is often a reasonable predictor of behavior in many, though not all, cases.29
Although the authors acknowledge the contributions of social-ecological frameworks55
to the understanding of AA consumption of FV, this study, as far as it can be determined, is unique in its application of theory to identify intrapersonal determinants of buying and eating more DGLV among the study's target population. Thus, the primary implication of the study is that the RAA is appropriate for explaining intention to buy and intention to eat more DGLV among a sample of midwestern AA women. In an effort to build on the information obtained from the current results, future research should identify how the constellation of constructs assessed in this investigation could predict current and future DGLV buying and eating patterns in this population.