«T. Bettina Cornwella a Lundquist College of Business 1208 University of Oregon University of Oregon Eugene, OR 97403-1208 (541) 346-8241 USA Anna R. ...»
Contingent Choice: Exploring the Relationship Between Sweetened Beverages and Vegetable
T. Bettina Cornwella
Lundquist College of Business
1208 University of Oregon
University of Oregon
Eugene, OR 97403-1208
Anna R. McAlisterb
Department of Advertising, Public Relations, and Retailing
309 Communication Arts and Sciences Building
Michigan State University
East Lansing, MI 48824
Forthcoming in Appetite (http://www.journals.elsevier.com/appetite/) APPET1525
10.1016/j.appet.2012.05.001 Corresponding author: firstname.lastname@example.org (Bettina Cornwell)
Not for publication:
Phone: +1 541 346 8241 Fax: +1 541 346 3341 Financial support from the University of Michigan (OVPR grant) to the first author is gratefully acknowledged, as is research assistance from Seung-Pil Lee, Christopher Swift, Arielle Mellen, and Mandie Maurer Abstract Adults and children are repeatedly exposed to the pairing of food and drink as found in meal deals and “combos.” There may arise from this indoctrination, a contingent relationship between drink context and food preference. Our multi-method research examines food and drink combining. A survey-based study examines the food and drink pairing preferences of adults (N = 60), while a laboratory study with young children (N = 75, aged three to five) examines the role of drink context on vegetable consumption. The adult survey finds strong food and drink combining preferences. The pairing of soft drinks with calorie dense foods is regarded favorably, while the pairing of soft drinks with vegetables is not. In child food trials, vegetable consumption is not influenced by the child’s fussiness but is influenced by the drink accompaniment. In limited contexts, these findings demonstrate the contingent relationship between drink context and food consumption. Both palate preference and associative learning may be mechanisms driving the effects of drink context on food consumption. The findings suggest simple consumer strategies that might be employed to change dietary patterns (e.g., drink water with meals), and hold straightforward policy implications (e.g., increase water as the default option in meal deals).
Keywords: food choice, marketing, children, vegetables, soft drink, soda, palate It has recently been suggested that one’s developed palate may play a role in the obesity epidemic faced in many societies (Cornwell & McAlister, 2011). The thinking is that, despite being malleable over time, the taste preferences for sugar, salt and fat that are developed early in life set the stage for diets high in calories and low in nutrients. Although not addressing food and drink combining, Cornwell and McAlister’s study of young children did find that knowledge of fast-food and soda brands was linked to development of a preference for sugar, salt and fat.
This raises the question of how one comes to hold food and drink preferences and the role they play in dietary intake. We suggest in this research that marketing may play a role in establishing expected food and drink combinations. Empirically, we examine the relationship between beverage context and food preference and choice.
Beverage consumption and the relationship of drink to food consumption is an area seeing increasing researcher attention and with good reasons. First, individuals do not appear to compensate for calories in their total energy intake when those calories are consumed as a beverage (Stookey, 2010). Second, the eating rate for liquids allows rapid ingestion and this is positively related to energy intake (Viskaal-van Dongen, Kok & Graaf, 2011). Third, modern diets tend to involve high intake of sugar-sweetened beverages (Lasater, Piernas & Pokin, 2011) and at the same time not enough water (Armstrong, 2010; Patel & Hampton, 2011). Importantly, children’s intake of sweetened beverages is positively associated with BMI, waist circumference, and poor dietary choices (Collison et al., 2010) while the intake of water is associated with preferable dietary profiles in children (Stahl et al., 2007) and weight management (specifically reduced excess weight gain, Stookey, 2010).
Early learning about food and drink is influenced by availability within a child’s environment (Aldridge, Dovey & Halford, 2009) and so in many cultures, commercially available food and drink combinations may influence preference and choice. International research conducted in Cardiff, Wales gives some insight to the practical and social challenges of food and drink choice. This qualitative study of the dental health of thirty-three parents/caregivers of children aged three years and under found many barriers to giving young children water to drink, including child rejection and negative adult feelings that it was “cruel” to offer water instead of a sweet drink and drinking water was a signal of poverty (Chestnutt, Murdoch & Robson, 2003).
Meal deal “combos” and “bundled” food products are popular and have behavioral consequences. Generally speaking, food and drink items are valued more highly (Venkatesh & Kamakura, 2003) and consumed in greater amounts (Stremersch & Tellis, 2002) when offered as part of a bundle. Although not focused on food combining per se, research on young adults’ restaurant choice shows correlates with dietary patterns (Larson et al., 2011) and particular restaurants are known for their use of combo meals. In Larson et al.’s study of over two thousand young adults, frequent use of fast-food restaurants serving burgers and fries was associated with high risk of overweight/obesity, high caloric intake and, interestingly, high intake of sugar-sweetened beverages and fat. “Those who reported burger-and-fries restaurant use on three or more occasions per week consumed nearly one additional sugar-sweetened beverage per day as compared to those who reported burger-and-fries restaurant use on less than one occasion per week” (Larson et al., 2011, p. 1696). By contrast, Larson and colleagues also found that full service (sit down and order) restaurant use was unrelated to weight status but positively related to higher intake of vegetables. Researchers have expressed concern that combo meals (i.e., soda and fries paired with an entrée) not only inflate caloric intake but also threaten the potential efficacy of proposed policy interventions. For example, Sharpe and Staelin (2010) argue that proposed soda taxes may have little effect on reducing overall caloric intake when bundles incorporating soda remain available.
Consumption of combo meals often begins early in life, with most children having consumed a combo meal (typically soda, fries, and burger) by age three or four (Schlosser, 2001). In addition to restaurants, grocery stores provide parents with options such as Lunchables (pre-packaged combinations of food and drink) targeted for children. Lunchables offered in the US include, for example, the Turkey + Cheddar Sub (turkey and cheddar on a wholegrain sub, packaged with sweet wafers, and spring water with Kool-Aid sweetening singles) or as an alternative, the Lower Fat Turkey and Cheddar Cracker Stackers (turkey and cheddar packaged with crackers and including a Capri Sun sweetened drink and fat free chocolate pudding). In response to concerns over obesity, Lunchables in the UK, as of 2004, no longer contain Capri Sun.
Whether it is through palate training or through other psychological mechanisms, individuals may come to anticipate certain items in combination and purchase them together even after a promotional effort has ended. As an example, increased consumption of Coke with breakfast has been credited to a late 1980s campaign that suggested “I need a Coke in the morning” (Schmeltzer, 2007). Of interest in the present research, is the notion that the seemingly pervasive practice of bundling particular “types” of food and drink may habituate individuals to the complementary tastes of the paired items.
Taste habituation, if it were observed, would hold important implications not only in the marketplace but also in consumer-planned consumption. For example, experience with a burger/cola combo in a fast-food restaurant may result in an individual seeking a similar entrée when consuming a glass of cola at home. The development of taste preferences that favor food and drink combining might drive decision-making but an alternative explanation might be found in associative learning (De Houwer, Thomas & Baeyens, 2001) or even cueing (Gall, Barnett, Lew & Selmants, 1987). Associative learning would suggest that past experiences lead to expectations about liked combinations, whereas cueing theory would suggest that the presence of one cue such as cola makes the person think of its savory accompaniment or visa versa. Before further examination, the stereotype of combining behaviors needs to be documented as a pervasive tendency.
The present research includes a survey study with young adults and lab work with young children. The purpose of the Study 1 survey is to establish the extent to which stereotypical food combinations are prevalent. As indicated above, we anticipate that marketplace offerings will have instilled in consumers an expectation that particular food and drink items belong together.
Hence, we predict that adults will agree that certain combinations make sense, while other food/drink pairings will be perceived as “wrong.”
Survey Method A short survey was designed to explore expectations regarding acceptable food/drink combinations. Following University of Michigan Institutional Review Board approval, undergraduate students were asked to participate. Data were collected in the Fall of 2009. As part of a larger study, 60 participants responded to questions about their food/drink preferences.
Students were invited to provide demographic information, however, this was a soft requirement and not all participants complied. The available data indicate that the age range was roughly 19 to 23 years (only undergraduate students were recruited and this was the age range obtained from those who provided their demographic information), and the gender split was roughly even.
Students were recruited via an invitation in class across four programs in kinesiology:
movement science, athletic training, physical education and sport management. Participants were not compensated for participation. Eight items regarding the participants’ agreement with various food and drink combinations were posed. For example, participants were asked if “Pizza goes well with tap water or plain bottled water.” A five-point scale was used to record responses (response anchors were 1 = disagree; 5 = agree).
Study 1 Results Survey findings are detailed in Table 1. Participants’ strongest feelings pertained to foods paired with soda. Highest agreement was observed when soda was paired with French fries (M = 4.20, SD =.93) or with pizza (M = 4.17, SD = 1.06). By contrast, participants disagreed with soda being a suitable complement to raw or cooked vegetables (Ms = 1.83 and 1.93, respectively). The difference between average ratings of soda paired with the aforementioned energy dense foods (M = 4.19) and vegetables (M = 1.85) was significant, t(1, 59) = 20.16, p.001. On the other hand, plain water received mostly average ratings, indicating that it was a reasonable complement to most foods.
Study 1 Discussion The finding that participants were more accepting of pairings of high-calorie foods and soda is consistent with our expectations. One could argue that calorie dense foods “naturally” pair well with soda but this would seem to vary across cultures, for example, green tea is commonly paired with calorie dense foods in Japan (McDonald’s, 2011). From an associative learning perspective, we know that many taste related perceptions and decisions are decidedly complex and culturally based. For example, food consumption habits influence the perception of taste. In a cross-cultural study, French and Vietnamese participants tasted mixed solutions of sweet lemon, sweet vanilla, sour lemon and sour vanilla (Valentin, Chrea & Nguyen, 2006).
Interaction comparisons showed that French participants experienced sweet enhancement in the presence of vanilla with the logic being that vanilla is used in France to flavor sweet dishes but this is not the case in Vietnam. The cultural influence of regular food and drink pairings may represent the same type of subtle, largely unexamined learning. Thus, important to our second study is the finding that sweet beverages are not perceived to combine well with vegetables in this sample of young US adults. When we consider this in light of the previously mentioned resistance on the part of parents and caregivers to serve young children water, the result is a drink context for the child that may negatively impact vegetable consumption.
The lab study examines children’s acceptance of raw vegetables as a “snack” paired with either a sweetened beverage or with plain water. It builds on the survey study findings by using measures of actual physical consumption, rather than a survey measure of acceptance of various combinations. Children are the population of interest because there is a need to understand how preferences develop at a young age. Interventions targeting dietary change may be more effective with younger participants (McAlister & Cornwell, 2010). Hence, we are interested in learning about conditions that might influence food consumption among very young children.
Though beverage context (sweet drink vs. water) is central to our interests, our study addresses the possibility that fussiness may explain vegetable consumption (or lack thereof) among preschool children.
Importantly for our research, established food preferences in childhood influence food choice over the lifespan and have both short- and long-term health consequences (Kemm, 1987).