|Home | About | Journals | Submit | Contact Us | Français|
Video games provide extensive player involvement for large numbers of children and adults, and thereby provide a channel for delivering health behavior change experiences and messages in an engaging and entertaining format.
Twenty-seven articles were identified on 25 video games that promoted health-related behavior change through December 2006.
Most of the articles demonstrated positive health-related changes from playing the video games. Variability in what was reported about the games and measures employed precluded systematically relating characteristics of the games to outcomes. Many of these games merged the immersive, attention-maintaining properties of stories and fantasy, the engaging properties of interactivity, and behavior-change technology (e.g., tailored messages, goal setting). Stories in video games allow for modeling, vicarious identifying experiences, and learning a story’s “moral,” among other change possibilities.
Research is needed on the optimal use of game-based stories, fantasy, interactivity, and behavior change technology in promoting health-related behavior change.
Usual school health curricular and other behavior-change interventions targeted at children have had limited effectiveness.1,2 New channels are needed to reach children that offer promise of promoting substantial health-related behavior changes. One such new channel is the video game, since many children spend numerous hours playing them.3 Using video games to promote behavior change could capitalize on the children’s pre-existing attention to and enjoyment of them. No review has appeared of health-related behavior-change video games. A common component of games is “story.”4 For those not familiar with games and stories, a simple glossary of terms appears in Table 1. This paper emphasizes the use of theory to enhance the possibilities for behavior change in the design and creation of stories and video games. The focus is on behavior change, because creating knowledge structures, while laudable in educational venues, is not sufficient to induce behavior change.5
What is a game? Children and adults have played games since prior to written history,6 suggesting that playing games meets enduring psychological needs.7 A game is a physical or mental contest with a goal or objective, played according to a framework, or rules, that determines what a player can and cannot do inside a game world.8 A video game is any game played on a digital device and encompasses a wide range of games played at arcades, over the Internet on personal computers, or on dedicated game consoles (e.g., Nintendo GameCube, Sony PlayStation, or Microsoft Xbox) or handheld units (e.g., Nintendo Game Boy, Sony PSP).
Games are played primarily for entertainment or “fun,”9 but what constitutes “fun” is not well understood. Typical measures of enjoyment (or fun) have used synonyms of fun (e.g., enjoy, like, interested, pleasurable, energizing),10 which do not elucidate the concept. In one study, statements of what constituted fun while being physically active (e.g., playing with friends, talking with friends, doing something daring, being really good at something) did not lead to separate factors in a principal components analysis (R. Jago, personal communication). In another study, six factors of fun in action video games included: novelty and powerfulness, appealing presentation, interactivity, challenging, sense of control, and rewarding.11 Other aspects of a game that children likely find enjoyable are fantasy (e.g., imaginary characters, virtual worlds)12 and interactivity.13 Games satisfy the player’s needs for autonomy, connectedness, and control.7 To win the game, video games challenge players to use the information they obtain as they navigate the game world,14,15 thereby providing an important education and training modality.16 This has given rise to the emerging genre of “serious video games” that employ the medium’s rich, role-playing, story-based environments to teach, train, and change knowledge, attitudes, and behavior.17
Today’s children and young adults are extensive users of digital devices,3 and video games are a big part of their digital experience. In 2004, the average video game player was aged 30 years and had played computer games for almost 10 years.18 The average child aged 8–10 years spent 65 minutes per day playing video games; 52 minutes/day among youth aged 10–14 years and 33 minutes/day among teenagers aged 15–18 years.3 Thus, video games reach a large and diverse audience who expect extended contact, suggesting games can attract and maintain attention, a key component for effective behavior change.19
Theory provides the foundation for promoting behavior change.7 A comprehensive model of learning for behavior change in video games is based on social cognitive theory (SCT) and the elaboration likelihood model,20,21 and includes the following steps: attention, retention, production, and motivation. The elaboration likelihood model proposes that gaining and maintaining a person’s attention is the first step in getting a person to process the information in a message to promote behavior change.21 SCT proposes that behavior change is a function of enhanced skills and confidence (self-efficacy) in doing the new behavior,19 while modeling19 and feedback22 are keystones for learning skills. Self-control procedures such as goal setting mobilize a person’s personal resources and focus attention on making specific changes.23 Games add an element of fun, an aspect of intrinsic motivation, thereby enhancing behavior change through enhanced motivation.7 Use of electronic or video games for health-related behavior change is in the earliest stages of development, but incorporating theory-based change procedures provides reason to believe that they can be effective.
A search was conducted for publications on video games for health-related behavior change by searching the authors’ personal files, contacting colleagues at professional meetings, and searching the following terms in PubMed: games, video games, and interactive multimedia, as well as combinations of these terms. Inclusionary criteria included using the word game to describe their software, with the goal of attempting to modify lifestyle behavior change. Exclusionary criteria were interactive multimedia programs that were not games, or that were games but did not target lifestyle behavior change. Games that were not video games24 were excluded. Twenty-seven articles about 25 different such video games were found and met inclusionary criteria through December 2006. These games addressed diet alone, physical activity alone, physical activity alone among the physically challenged, diet and physical activity combined, or other health-related behavior changes, as shown in Tables 2–6, respectively. Two authors abstracted all articles, and information in tables reflects consensus.
There was substantial variability across studies in their design, the targets for change, and the characteristics (and reporting of characteristics) of the games. The two-group (treatment versus control) randomized clinical trial (RCT) was the most common outcome evaluation design, but several were single-group-only. Four of the physical-activity video game studies tested primarily whether the game offered the possibility of enhancing fitness,24–29 with only two assessing whether the game changed behavior.30,31 Several had ample samples for the evaluation,32,34,35 while others were small pilot studies.33 The primary outcome measure varied from knowledge32,36; to psychosocial variables36; to behaviors33; to anthropometric,30 physiologic,37 or health-outcome38 variables. Many of the reports provided little data on the game’s characteristics. Lack of comparability across game descriptions and outcomes precluded a meta-analysis or any systematic comparison relating games to outcomes. Most reported some positive outcomes, with only one reporting no evidence of change.37
All of the diet-alone change games employed (n=3) a story of some kind, but it was difficult to discern characteristics of the story in one report.32 While one game was predicated on an operant form of learning theory,32 two were predicated on social cognitive theories.20,34 All the diet-alone change games resulted in some type of dietary change.
Few of the physical-activity-alone promoting games employed story, emphasizing, instead, simply being physically active.25,26,29,30 The decline in their use over time and the perception of their being boring within just 4 weeks of use30 might be attributed to their lack of story. The two physical-activity-alone games that used story39,40 used it in different ways, and targeted very different age groups. Three of the studies of the physical-activity-alone games assessed only the extent to which a single session attained a fitness-enhancing level of physical activity25,26,29 with all raising concerns about the modest intensity of physical activity that was elicited. The fourth study30 suggested that the duration of participation further limited what body composition outcomes could be expected from these games. One of the physical-activity-alone games obtained a decline in BMI among girls, but not boys,39 while the other conducted with boys alone attained no BMI change.40 The game that changed BMI paradoxically attained a small decline in moderate-intensity activity in boys and girls,39 while the other attained a moderate increase in light physical activity, but no change in BMI,41 and all objectively measured physical activity by accelerometry.
Four studies by one group provided innovative user interfaces (wheels or arm ergometry) to enable wheelchair users to be physically active and in doing that control action in a video game (see Table 4).27,28,31,40 In each case, the interface enabled the participant to be active at a fitness-enhancing level in a one-time trial with the expectation that being able to play the video game would motivate physical activity at an acceptable level of intensity outside the laboratory. One study demonstrated that seven of eight spina bifida patients were sufficiently motivated to attain a fitness effect after 16 weeks of use of the armergometer interface.31
Four games proposed to change diet and physical activity (see Table 5). Three were developed by one group1,20,41–43 and used story as an organizing framework with minigames inserted to deliver behavior change and self-regulation-related activities. The behavioral theory underpinning the first game, “Fun, Food and Fitness,” has been presented,20 and pilot study results have demonstrated some diet and physical activity change.33 A second pilot study tested as a stand-alone electronic-health program resulted in both diet and physical activity change.43 Preliminary 1-week outcomes of MetaKenkoh33 demonstrated some diet and physical activity change. Two games benefited from alpha testing,42 which enhanced several aspects of their functioning, but outcomes have not been published.
Three of the other health-related behavior-change games addressed behavioral issues in asthma. One demonstrated substantial changes resulting in fewer hospitalizations38; another demonstrated increased knowledge and more internal control, but no lung function improvements44; the third reported no demonstrable changes.37 It is not clear from the publications what differences across the games, or the samples, might have accounted for the different outcomes. Five other games addressed diabetes-related behavior.36,45 One of these games demonstrated substantial changes resulting in lower emergency room and urgent medical care use.45 The other games were very limited in focus (e.g., trying to enable children to understand the need and value to balance food intake with insulin administration to control circulating glucose levels),36 and demonstrated psychosocial changes at 3-months followup.36 An action–adventure game with the child shooting cancer-causing agents in the bloodstream targeted medication adherence among pediatric cancer patients46 and reported substantial psychosocial and regimen compliance change at 3 months after playing the game.46
Playing most of these behavior-change video games led to a broad spectrum of desirable outcomes from knowledge increases,32 to attitude changes,36 behavior changes,34 and other health-related38 changes. This bodes well for the future use of video games to promote health-related behavior changes and warrants an intensive analysis of aspects of video games that offer the most promise of promoting behavior change. There appear to be two primary methods by which video games can influence behavior. The first involves the insertion of behavior-change procedures (e.g., goal setting) into the process of playing the game. The second involves the use of story and inserting behavior-change concepts in the story.
Computers have been used to promote behavior change for some time47; however, most computerized behavior-change programs are not video games. SCT was the most commonly cited theory providing a foundation for behavior change.19 Some reported using a full range of behavior-change procedures,20 while others enhanced only knowledge.36 As an example, the SQ! game integrated SCT-specified behavior technology procedures (i.e., goal setting, decision making, goal review, social reward) into game play.34 SQ! goals included changing known environmental, personal, and behavioral factors regarding why children were not eating particular fruits and vegetables. The goals were tailored to whether these factors pertained to each child specifically: SQ! attempted to increase preferences for targeted fruits and vegetables and used self-regulation procedures, e.g., goal review and problem solving, when goals were not attained. Setting and attaining goals enhanced the fruit and vegetable intake among various student subgroups48 and overall the video game achieved a 0.9 serving per day increase in fourth-grade children’s fruit and vegetable consumption.34 How diverse behavior-change procedures can be inserted in video games has been addressed elsewhere.20
A story is a narrative of a series of events.4 Stories take place at a particular time (e.g., the winter of 1776, the week after college graduation); at a particular place (in a boat crossing the Delaware, on the bridge of a starship); and have characters. Ordinarily, one character, called the protagonist, takes the lead in the story. Protagonists can have external or internal conflicts. When protagonists oppose someone else, that character is the antagonist. The struggle between these two is the conflict, which is the motivating factor behind the story’s action and plot.4 Stories usually engage individuals by means of their empathy with the protagonist4 and are effective when the protagonist shows change in values in the story (e.g., cowardice to courage, betrayal to loyalty), exemplifying a lesson that can be learned (also called the story’s controlling idea4 or moral, e.g., eating fruits and vegetables and being physically active to provide strength and stamina to escape from oppressive situations). These changes in values are engineered by the writer in terms of events that pose conflict for the protagonist. The conflicts can be the commonly reported barriers to making behavior change, and the protagonist’s ways of overcoming the barriers can be the modeling of effective problem solving.
This very abbreviated exposition suggests that behavior change can be enhanced when stories address behavior-change issues, and the lesson to be taken from the story promotes health behavior changes. While stories have been used for health-related behavior changes (soap operas),49 the exposition of how this has been done, especially using story structure to promote behavior change, has not been addressed. An example of such a story was Squire’s Quest! (SQ!) a 10-session video game.34 The story’s controlling idea was that eating more fruits and vegetables gives strength to resist dangerous characters. SQ! used a variation of a common medieval story involving a king, queen, knights, invaders, and a struggle. Invaders were destroying a kingdom by destroying its fruits and vegetables, the source of energy. The king didn’t have enough knights to fight off the invaders, so each player was asked to become a squire, a person in training to become a knight. As per medieval lore, squires must face and overcome challenges to become knights. The challenges in this game just happened to require that the squires eat more fruits and vegetables, resulting in an increase of 0.9 serving/day in fruit and vegetable intake. Thus, SQ! used a story to engage the students, maintain their interest and model the desired behaviors by key characters.
Stories have been told by every culture throughout human history,50 thereby also suggesting that they meet some enduring psychological need. Different cultures have evolved somewhat different ways of conveying stories, although many elements are common to all.51 The dominant format used in most Western society books, films, and video games is a three-act structure, a pattern Western audiences have come to expect: Act 1—exposition and the beginning of the conflict; Act 2—complications and climax; and Act 3—resolution and conclusion (denouement). Stories that do not adhere to these conventional storytelling rules risk alienating Western audiences. Why certain stories appeal to certain audiences is not clearly known.41
Patterns of story telling have evolved into story genre,52 e.g., western, murder-mystery, and comedy. Melodrama was originally developed for the stage to focus on plot and action and reduce story complexity.52 Most melodrama issues are reduced to struggles between good and evil, which appeal more to viewers’ emotions than cognitions,52 thereby presenting a potentially promising tool for behavioral intervention in serious games. Behavioral theory on the role of emotions in health-behavior change53 will need to advance rapidly to capitalize on these possibilities.
Serials are a single narrative developed through a number of individual, linear episodes. Each serial episode forms a separate unit of a larger story. Serial melodramas often end with a cliffhanger, unresolved tensions in a storyline designed to motivate viewers’ return to the next episode. Research on the Zeigarnik effect54 presages an understanding of the role of cliffhangers in story and game design, but more research on this point is necessary in advancing their use and effectiveness in behavior change video games.
Episodic stories are also short, complete narratives. Instead of linearly building to a conclusion, they loosely arc and intertwine substories, related by the same characters and settings. Soap operas are episodic, melodramatic stories involving a group of individuals. Each character has his or her own story, which link in unpredictable ways from episode to episode. Episodic stories may function primarily to maintain attention. In serious video games, interrelated components among the differing stories could be used to reinforce behavioral-change messages in diverse situations. For example, each component story could deal with different barriers to eating more fruits and vegetables (e.g., distaste, home availability, skills to prepare more, and tastier dishes).
Video games can encompass and capture a player’s full attention, as if the player were actually part of the game environment. This has been called immersion, or presence.7 Video game environments were confined initially to flight simulators and research laboratories55; realistic or otherwise complex virtual three-dimensional environments are now common in many video games. Game players become involved literally and emotionally in the story. While skilled game developers effectively create such immersive experiences regularly, the behavioral science of how this is achieved, or optimized, is not known. Immersion is believed to be a component of intrinsic motivation.7
Nine dimensions of interactivity in a website (i.e., accessibility, navigation, time, personalized content, delivery of message, data entry and use, entertainment, promotions, relationship) and 52 underlying variables have been delineated.56 Subsets of these appear relevant to interactivity in a video game. Directly participating in story situations and taking first-person control of events are key aspects of interactivity in video games.13 Just as in real life, players learn through planning, decision making, and personally witnessing cause-and-effect relationships. Role playing likely increases a player’s personal stake in a video game’s outcome. In part, role-playing performances combine the emotion of storytelling with the power of character immersion. The video game designer can structure the interactive options to provide meaningful feedback for player-made choices. The feedback can be in the form of reinforcement (e.g., statements of positive regard), information to make next choices, or experiential learning sequences (e.g., interactions with animated characters that provide theoretically specified experiences).
Fantasy, defined as the active use of imagination,57 has facilitated active engagement among youth,58 and is a primary source of intrinsic motivation.34, 57–59 Both youth and adults engage in fantasy, although the content of the fantasies vary in that they tend to reflect personal interests and concerns.57 For example, adolescent fantasies tend to include themes such as appeal to opposite gender, future vocation, sports, and achievement.57 Fantasy has been reported to peak between late adolescence and early adulthood.57 A review of text-based nutrition education materials found that fantasy was included in 16 of 30 of the materials reviewed.60 The materials that included fantasy were described as more creative and fun than those not including fantasy, and they were more likely to use characters and scenarios to foster engagement.60
Fantasy contexts have also been used in video games.58,59 Video games excel at “what if” scenarios, i.e., the ability to personally inhabit an improbable world, wander around on one’s own, and interact with fantastic characters or events. Fiction and action–adventure stories may be better suited to video games’ rich, immersive, fantasy role-playing environments than nonfiction genres. In a study of 7th- and 8th-grade children, focus groups to select video game storylines revealed that child participants preferred action–adventure over other genres (V. Thompson, personal communication). In another study with 3rd- and 4th-grade students, embedding educational information in fantasy contexts was found to result in significantly more learning and knowledge transfer than nonfantasy contexts.58 The effect was not mediated by choice, i.e., choosing the fantasy context was no more effective than having the fantasy context assigned. The fantasy context used in the study was created by embedding educational information in simple stories that required a child to solve problems.58 A video game program promoting asthma self-management to children aged 9–13 years59 included elements of fantasy. The video game had motivational appeal, and children who used the game had gains in knowledge, self efficacy, and attributions leading to self-management. SQ! also used a fantasy-based game format to effectively promote fruit and vegetable consumption to 4th graders.34
Video game cinematics (sometimes called cut scenes) advance a game’s storyline, but are relatively short movie clips that are watched passively by players intermittently throughout the game. Unlike traditional passive-only media, video games depend on the player’s action to move the story forward (via branching logic based on player selections). There is more than one ending in a video game (at a minimum, there are two: winning and losing), and sometimes multiple ways of getting to an ending. Players’ actions, often called game mechanics, are the way that players navigate story environments, interact with story events, and chose story paths. The mechanics of game play need to be inventoried and their role in involvement and individualizing messages more clearly understood to be able to create successful behavior-change video games.
The kinds of games appealing to young children should be quite different from those appealing to adolescents or young adults, because of substantial differences in cognitive or emotional development throughout this age range. Research, however, has not as yet led to development of a theory of developmentally appropriate games. Thus, development of an age-appropriate intervention requires substantial formative research (e.g., focus group discussions, intensive interviews, observations) with the targeted demographic group on story and character concept, story arc, personality and visual representation, and alpha testing on fun and functionality of the interactive components. No health-related behavior-change games were found with stories for adults. While in theory, games with stories should be able to be designed for this demographic, little guidance was found in the published literature.
To capitalize on the possibilities of video games for promoting health behavior changes, behavioral scientists need to collaborate with professionals who can write an engaging story and have knowledge and skills in game design, formative research, story boarding, producing, directing, music composition (for games with music scores), computer art, animation, and programming. In larger projects, each of these skills is offered by different professionals, whereas in smaller projects, one person may play several or even all roles. For a health behavior-change video game, additional expertise is required in regard to content expertise about the health problem and/or the health behavior(s) of concern, and behavior-change intervention design (theory and procedures). Initially there would be a vast divide in the understanding and even the languages used by these differing sets of professionals. Over time, they must learn to effectively communicate and respect each others’ contributions.61
There is no consensus model for serious video game development strategy or process. The process used in part will reflect the amount of funding available. Greater funding will have larger teams and larger incubation periods producing more fully developed products. Lesser funding likely imposes an incremental developmental period where smaller products are added to earlier products. The Spiral Technology Action Research (STAR) Model was propounded to capture the latter62 and capitalizes on action research methods.63
Developing a video game is a time-consuming process. Some commercial video games take 3 or more years to develop. (The authors spent 3.5 years developing a health-related behavior-change game.) The costs of developing a commercially viable video game have been estimated in the millions of dollars. Major components of such costs are the high-end graphics, animation, and interactivity needed to attract and maintain attention for a sophisticated audience. As the qualities of commercial video games increase in sophistication, similarly high expectations will be created for health-related behavior-change and other serious video games.
Perhaps economies can be achieved by reusing and adapting computer program code (sometimes called game engines) to create new video games. Perhaps certain characters will achieve a level of notoriety that their original art work transferred to new video games will assure attention in the market place. It is not clear the extent to which the market exists that will buy health-related behavior-change video games (e.g., parents of children suffering from certain health issues, schools for health education curriculum, or just concerned parents). The ideal purchasers of serious video games would be the children themselves, attracted by the prospect of a fun experience. Proof of markets would enable software publishers and investors to invest capital in the development of behavior-change video games. Until then, or absent such markets, governmental or charitable organizations will need to invest in this development.
To enable the field to advance, the outcome evaluations of health-related behavior-change video games need to employ state-of-the-art designs, adequately powered samples, valid and reliable measures of outcomes and mediators,64 and be reported in a way consistent with other randomized clinical trials (e.g., see the CONSORT statement)65 to ensure that data required for meta-analyses are available. Using the earlier framework for understanding behavior change (i.e., attention, retention, production, motivation), this section identifies priority research issues to better understand how change occurs in video games, and thereby how to better design them in the future.
The capabilities and limitations of diverse video game platforms deserve careful investigation. For example, the depth and breadth of the visual and audio experience on a full computer screen is quite different from that on a small handheld game unit. Whether the depth and breadth of this experience influences the player’s immersion in the story is not known. The handheld unit may hold the same appeal for younger players (who are likely more prone to fantasy) as the larger units have for adults. Characteristics of the game that lead to intense involvement, and the effect of immersion on behavior change, need to be better understood. For example, players might come to a game expecting to be immersed in the story, thereby not requiring much effort on the part of the game (e.g., before children started a game, they expected it to be fun).66 There may be elements of story, visual effects, or ability to participate in what happens (the interactivity) that trigger or facilitate immersion. Players may have certain expectations for health behavior-change video games. These may vary from other video games. Health behavior-change video games may minimize possible difficulties caused by some expectations (e.g., “they can’t be much fun because they are not first-person shooter games”).8
Novelists and screenwriters may intuitively know how to tailor a story to maximize the involvement of an audience, but behavioral scientists likely do not. Research on how best to use the three-act structure to design games offers the possibility of enhancing effective behavior-change programming. Research on why audiences expect these story conventions offers the possibility of innovative approaches to the use of stories with new structures in behavior change gaming.
Emotion has been strongly linked to memory,67 or retention. Stories are designed in part to evoke emotion.4 Recent models predicting behavior have incorporated emotional variables.53 Reinforcement mechanisms in video games influence positive and negative emotions.68 How emotions influence health-related behaviors and its changes are not well known. An empiric literature needs to be generated on how aspects of story, components of games and stories within games, evoke emotional responses, which in turn enhance attention to and retention of messages and otherwise enhance (or inhibit) behavior change.
Games promoting physical activity alone, whether for the fully abled or for those physically challenged, have emphasized primarily just being physically active (i.e., just producing the behavior), apparently assuming that the aspects of being active will be rewarding, or using commercially available nonactivity-promoting games as incentives for doing the activity.
Research is needed on the extent to which different categories of children (e.g., the usually inactive, the obese, different ages, both genders) are willing to initiate activity in response to these video games, how much activity is elicited, and for how long the behavior will be maintained. Mechanisms whereby such behaviors are elicited and maintained need to be elucidated. Activity-promoting video games with and without story need to be compared.
A number of video games are single-entity games,32 while others use a series of smaller minigames that encapsulate and present behavior-change procedures.69 These represent different uses of games; understanding when each is most appropriate and effective need to be delineated.
Industry and the military have used simulation experiences to train employees in how to better perform work-related behaviors, especially in virtual high-risk settings, where making a mistake in real life could have substantial personal and social costs (e.g., how to search for a missing soldier in a war zone). Similar kinds of simulated virtual experiences could be used to train parents to provide better food parenting,70 or to train children how to “ask” for fruit and vegetables.71
The theory of self determination (TSD)72 has posited that intrinsic motivation, i.e., doing something because you want to do it, is a predictor of initial and continuing performance of a behavior.7 “Fun” is an aspect of intrinsic motivation. Research needs to explicitly determine what aspects of playing a video game and of story are fun. This might be done experimentally by systematically varying aspects of games and testing their perceived fun,66 by doing surveys including items usually considered fun, or by comparing games that vary in their enjoyment ratings.
Mastery is another aspect of intrinsic motivation. A nine-factor solution with three higher-order factors captured adult’s motivations to play massively multiplayer role playing games.73 Consistent with TSD, “achievement” (an aspect of mastery) was one of the higher order factors. Boys were more motivated by achievement, while girls were more motivated by affiliative or social factors.73 Similar measures need to be developed for health-related behavior-change video games.
Serious video game–based behavior change is an exciting form of media-based intervention. A peer-reviewed literature is emerging on this topic. The many desirable outcomes warrant moving forward in this area. A second generation of research requires models of pathways of effects, and testing theory-based propositions. This should include how to use story to affect variables on the pathways to change. Understanding how to harness the power of media-based video game interventions offers great promise to promote health-behavior change, but requires extensive research.
This research was primarily funded by a grant from the National Institute of Diabetes & Digestive and Kidney Diseases (5 U44 DK66724-01). This work is also a publication of the United States Department of Agriculture (USDA/ARS) Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, Texas, and had been funded in part with federal funds from the USDA/ARS under Cooperative Agreement No. 58-6250-6001. The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products, or organizations imply endorsement from the U.S. government.
No financial disclosures were reported by the authors of this paper.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.