Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

DietBet: A Web-Based Program that Uses Social Gaming and Financial Incentives to Promote Weight Loss

DietBet: A Web-Based Program that Uses Social Gaming and Financial Incentives to Promote Weight Loss Background: Web-based commercial weight loss programs are increasing in popularity. Despite their significant public health potential, there is limited research on the effectiveness of such programs. Objective: The objective of our study was to examine weight losses produced by DietBet and explore whether baseline and engagement variables predict weight outcomes. Methods: DietBet is a social gaming website that uses financial incentives and social influence to promote weight loss. Players bet money and join a game. All players have 4 weeks to lose 4% of their initial body weight. At enrollment, players can choose to share their participation on Facebook. During the game, players interact with one another and report their weight loss on the DietBet platform. At week 4, all players within each game who lose at least 4% of initial body weight are declared winners and split the pool of money bet at the start of the game. Official weigh-in procedures are used to verify weights at the start of the game and at the end. Results: From December 2012 to July 2013, 39,387 players (84.04% female, 33,101/39,387; mean weight 87.8kg, SD 22.6kg) competed in 1934 games. The average amount bet was US $27 (SD US $22). A total of 65.63% (25,849/39,387) provided a verified weight at the end of the 4-week competition. The average intention-to-treat weight loss was 2.6% (SD 2.3%). Winners (n=17,171) won an average of US $59 (SD US $35) and lost 4.9% (SD 1.0%) of initial body weight, with 30.68% (5268/17,171) losing 5% or more of their initial weight. Betting more money at game entry, sharing on Facebook, completing more weigh-ins, and having more social interactions during the game predicted greater weight loss and greater likelihood of winning (Ps<.001). In addition, weight loss clustered within games (P<.001), suggesting that players influenced each others’ weight outcomes. Conclusions: DietBet, a social gaming website, reached nearly 40,000 individuals in just 7 months and produced excellent 4-week weight loss results. Given its reach and potential public health impact, future research may consider examining whether a longer program promotes additional weight loss. (JMIR Serious Games 2014;2(1):e2) doi: 10.2196/games.2987 KEYWORDS commercial weight loss; Web-based intervention; social gaming; financial incentives http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen This study examined the effectiveness of DietBet, which is a Introduction commercially available Web-based program that uses social gaming and financial incentives to promote weight loss. Upon Obesity Prevalence and Treatment Availability enrollment, players join a game and enter money into a pool. Over 45% of the world population and approximately 66% of All players have 4 weeks to lose 4% of their initial body weight. American adults are overweight or obese [1,2]. Excess adiposity During the game, players report their weight and interact on the is associated with serious health risks, including diabetes, DietBet platform. At the end of the game, all players within cardiovascular disease, and some types of cancer [2,3]. each game who lose at least 4% in 4 weeks are declared Behavioral weight loss programs consistently yield weight loss “winners” and split the initial pool of money bet at enrollment. of 8 to 10kg, which are associated with significant health The primary aim of this study was to conduct a naturalistic improvements, including reduced risk for diabetes [4]. However, examination of engagement, retention, and weight loss outcomes behavioral programs are largely university-based and are, in DietBet. Previous findings from financial incentive weight therefore, only available to a small portion of overweight and loss trials have shown that the possibility of losing large (vs obese individuals in need. small) amounts of money for not meeting weight goals motivates better overall weight loss [10]. In addition, more frequent Accessibility of Internet Commercial Weight Loss self-weighing and greater social influence for weight loss have Programs been found to be associated with better weight outcomes In contrast, commercially available weight loss programs have [12-15]. Given these findings, we explored whether: (1) betting wide reach, particularly Web-based interventions. In fact, a more money at enrollment, (2) completing more weigh-ins, and large percentage of individuals who attempt weight loss report (3) having greater social engagement/influence predict greater using commercial programs [5], and, in recent years, enrollment percent weight loss and greater likelihood of winning. Finally, in Web-based interventions has increased substantially [6]. The given evidence that weight loss clusters in social networks and appeal of Internet interventions is likely due to the reduced that group characteristics impact weight outcomes in participant burden associated with frequent in-person visits (eg, group-based weight loss competitions [12], we also explored time, transportation). Another important benefit of Web-based whether weight loss clusters within games (ie, players in the interventions is their inherent reach; given that Internet access same game achieve similar weight loss) and whether game has increased exponentially over the past decade [7], Web-based characteristics (eg, number of players) are associated with programs are widely accessible to individuals who may not weight outcomes. otherwise have access to clinical weight management interventions. Similarly, such programs are accessible 24-hours Methods a day in a variety of locations (home, work, public libraries). Given their appeal and ability to reach large numbers of Procedures overweight and obese individuals in need, commercially DietBet is a social gaming website that uses financial incentives available Web-based interventions have significant public health and social influence to promote weight loss. Players are recruited potential. via press coverage (eg, Today Show, CNN, New York Times, Evidence for Commercial Weight Loss Programs Wall Street Journal), business development efforts (popular wellness experts with social capital are asked to host games and Despite their potential to improve public health, the scientific encourage fans/followers to participate), direct virality (players literature on the effectiveness of Web-based commercial weight recruit other players), and indirect virality (players share DietBet loss programs is sparse [5]. Only two programs have been information on Facebook). At enrollment, players bet money rigorously evaluated in randomized trials: eDiets [8,9] and The and join a game. Players are given the option to join an existing Biggest Loser Club [10]. Results of the eDiets trials showed game that has not yet started or create their own game. If they that participants assigned to eDiets achieved significantly less create their own game, it can be either open (anyone can join) weight loss than those assigned to a self-help condition or an or closed (invite only; ie, all players know each other). Internet behavioral weight loss program [8,11]. Similarly, while the Biggest Loser Club produced greater weight loss than a no Players are prompted to submit their official start weight two treatment control condition, given that the program was 3 days prior to the start of the game, (see below for weigh-in months in length, the weight loss was modest (-2.1kg) [9]. These procedures). Once the game begins, players have 28 days to randomized trials have clear benefits and are essential to lose 4% of their initial body weight. DietBet does not promote demonstrate efficacy. However, an important shortcoming is a specific diet or weight loss strategy; instead, players are that, given the nature of rigorous, randomized trials (screenings, allowed to choose any strategy to lose weight (eg, low-fat diets, run-ins, retention efforts, etc), the results may overinflate true low-carbohydrate diets, etc). During the game, players post engagement, retention, and weight losses outcomes of photos, comments, and weight loss tips (Figure 1 shows a commercially available programs. Thus, to complement the screenshot). They are also encouraged to stay accountable to clinical trial literature, ecologically valid studies are needed that one another by posting their weight loss. Players can view their examine real-world enrollees in naturally occurring Web-based weight loss relative to the weight loss of others in the game (see commercial weight loss programs and, thus, ascertain true Multimedia Appendix 1). To facilitate game communication program engagement, retention, and outcome data. Results from and sharing, players have access to the DietBet app, which is such studies may be used to inform consumer decision making designed for all smartphone mobile devices and includes all and public policy. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen aspects of the gaming platform. At the end of 28 days, players people won, after DietBet’s cut (see below), they would each have 48 hours to send in their final weight. All players who lose win US $50. If no one loses 4% of their starting weight, the at least 4% of their initial body weight in 28 days split the pool player who lost the most weight wins the pool of money. of money that was bet at the start of the game. For example, if Winners are notified by email of their payout, which they can the game consisted of 10 players who each bet US $25 and 4 either apply to their next DietBet game or cash out. Figure 1. Intervention screenshot. DietBet keeps a portion of the initial pool of money to cover identify players who have won more than US $300 in DietBets financial transaction costs and staff time associated with weight and identify players who have unusual weight gain between verification, customer service, and tech support. DietBet’s share DietBet games (ie, gained more than 1% per day). Those is taken out of each game’s starting pool of money before individuals are flagged and DietBet staff completes a detailed winners are paid. Thus, players who do not win do not incur a review of their weight, game activity, and profile information. charge to play. DietBet retains 15%-25% of the payout Using these algorithms and the photo-based system within 48 depending on the amount of each individual bet (US hours of the start and end of a game, if a weight is deemed $1-$99–25% fee; US $100-$249–20% fee; US $250 or questionable, players are required to submit extra proof by more–15% fee). completing a live Skype weigh-in with one of DietBet's referees, a video weigh-in, or an in-person weigh-in at a DietBet approved Measures location (eg, Walgreens drugstores or Customer Value and Service drugstores). Gender and Weight Participants report their gender at the beginning of the game. Weigh-Ins, Social Engagement, and Facebook Shares Within 48 hours prior to the start of each game, participants are All weigh-in, social engagement, and Facebook data were required to complete an official weigh-in. Similarly, within 48 obtained directly from the DietBet website. Specifically, during hours after the end of a game, participants are to complete an the game, players reported their weight on the DietBet platform official weigh-in. Official weigh-in procedures involve the and weigh-ins were summed for each player. Players interacted submission of two photos, one of the player on a scale in light with one another by cheering, posting pictures, posting status indoor clothing without shoes and another that captures the updates, “liking” others’ posts, and commenting on and replying scale’s readout and a piece of paper that includes the player’s to posts; to create an overall social engagement variable, the weigh-in password. DietBet staff reviews each photograph for number of social interactions for each variable (cheers, pictures, player-password correspondence. In addition, DietBet uses updates, etc) were summed for each player. DietBet collected prespecified algorithms to detect any unusual weight outcomes data on whether players posted their game participation on and game activity; specifically, across games DietBet will Facebook; data were coded 1 for “Yes” or 0 for “No.” The http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen option to share on Facebook was not available in early games. predominantly female (84.04%, 33,101/39,387) with a mean Thus, the sample size for this variable is smaller than the overall baseline weight of 87.8kg (SD 22.6kg). The average amount of sample (n=20,059 vs N=39,387). money bet at game entry was US $26.84 (SD US $21.93). Upon enrolling, 50.03% of players chose to share their DietBet Money Bet and Won participation on Facebook (note–the option to share on Facebook The amount of money bet and the amount of money won was was not available in earlier games, thus the total sample size obtained objectively via payment transaction data. for this variable was n=20,059, of those n=10,036 shared their DietBet participation). During the 4-week game, players Game Characteristics completed an average of 5.3 (SD 3.9) weigh-ins and engaged The number of players on each game and whether all players in 9.3 (SD 78.4) social interactions (eg, cheers, posts, likes, etc). knew each other (closed game/invite only vs open game) were A total of 71.71% of participants (28,244/39,387) self-reported obtained objectively by extracting data from the platform. their weight into the DietBet platform during week 4, and Statistical Analyses 65.63% (25,849/39,387) completed an official weigh-in at the Player characteristics, engagement, and completion outcomes end of the game (ie, completed the photo-based weigh-in process were examined using simple descriptive statistics, including immediately following the game). A greater proportion of men means, standard deviations, and percentages. Completers (ie, completed a photo-verified weigh-in than women (Men–68.39%, those who completed an official weigh-in procedure at the end 4292 out of 6275; Women–65.10%, 21,550 out of 33,101; of the game) and noncompleters were compared using analyses P<.001). Compared to noncompleters, completers weighed less of variance or chi-square tests for continuous or categorical at baseline (87.4kg, SD 22.2kg vs 88.5kg, SD 23.3kg, P<.001), variables, respectively. To examine the effects of DietBet on bet more money at program entry (US $27.79, SD US $24.99 weight loss, a conservative baseline carried forward vs US $25.01, SD US $14.21, P<.001), and had more weigh-ins intention-to-treat approach was used; players who did not finish (6.5, SD 4.0 vs 3.0, SD 2.5, P<.001) and more social interactions the game (ie, did not complete a final, verified weigh-in) were with their teammates during the game (eg, cheers, comments, assumed to have remained at baseline/entry weight and within likes, etc; 12.0, SD 95.3 vs 4.3, SD 22.6, P<.001). A greater subjects t tests were conducted. Predictors of weight loss and percentage of completers versus noncompleters shared their whether a player won their game were examined with regression game participation on Facebook (50.96%, 7074 out of 13,882 analyses. In addition, to determine which variables explained completers vs 47.74%, 2949 out of 6177; P<.001). unique variance in weight loss, a multivariate analysis that Weight Loss included all variables was conducted. The effects of game and Intention-to-treat analyses (assuming that individuals who did game characteristics on weight outcomes were also explored. not complete an official, photo-based weigh-in at the end of the To examine whether weight loss clustered within games, an game did not lose any weight) showed that players lost a unconditional multilevel model was conducted and an intraclass significant amount of weight from baseline to the end of the correlation coefficient (ICC) was calculated using the resulting game (P<.001; Table 1). The average weight loss was 2.6% between and within group variance components, (SD 2.3%). Out of the 39,387 players who enrolled, 43.60% (ICC=U /U +R). 0 0 (n=17,171) were winners (ie, lost at least 4% of initial body Ethics weight or, if no one lost 4%, lost the most weight in their game). The average amount won was US $58.79 (SD US $34.90) (net The Miriam Hospital’s Institutional review board approved this earnings–US $29.00, SD US $16.43). Game winners lost an study. average of 4.9% (SD 1.0%) of initial body weight. A total of 30.68% of winners (n=5268) achieved a 5%, or clinically Results meaningful, weight loss. Engagement and Completion From December 2012 to July 2013, 39,387 players participated in 1934 games on the DietBet platform. Players were Table 1. Player weight loss. Total sample All winners Winners who lost ≥ 4% Winners who lost most weight in game (N=39,387) (n=17,171) (n=16,687) (n=484) % weight loss, mean (SD) 2.6 (2.3) 4.9 (1.0) 4.9 (0.9) 3.4 (0.9) money at game entry (P<.001), completing more weigh-ins Participant Characteristics Associated With Weight during the game (P<.001), sharing game participation on Loss and Winning the Game Facebook (P<.001), and having more social interactions with Participant baseline characteristics and game variables predicted other players (P<.001). These same variables were also weight loss outcomes and whether a player won their game. significant predictors of whether a player won their game Significant predictors of greater percent weight loss were male (Ps<.001; Table 2). gender (P<.001), lower baseline weight (P=.03), betting more http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen Table 2. Winners’ versus nonwinners’ baseline characteristics and engagement. Winners Nonwinners Winners versus nonwinners (n=17,171) (n=8678) (P value) Male, n (%) 3489 (20.32) 803 (9.25) <.001 Baseline weight kg, mean (SD) 87.1 (22.1) 88.1 (22.4) <.001 Amount bet US $, mean (SD) $29.79 ($29.18) $23.83 ($12.30) <.001 Shared on Facebook, n (%) 4837 (53.19) 2241 (46.79) <.001 Weigh-ins, mean (SD) 7.0 (4.2) 5.4 (3.1) <.001 Social engagement, mean (SD) 13.6 (113.6) 8.8 (38.4) <.001 Given that completion was associated with baseline characteristics and engagement, to control for potential confounding of completion, only game completers were included in these analyses, n=25,849. Option to share on Facebook was not available in early games. Thus, the sample size for this variable is smaller than the overall sample of completers. Specifically, of those who were given the option to share on Facebook, n=20,059, a total of n=13,882 completed the game. Of those, n=9093 were winners and n=4789 were nonwinners. Sum of cheers, comments, replies, likes, photos, and updates. In a multivariate model, all variables remained significant, Comparison of DietBet Results to Other Internet independent predictors of weight loss (Ps<.001) with the Commericial Weight Loss Programs exception of number of social interactions and whether a player DietBet results compare favorably to other Web-based shared on Facebook (Ps>.38). commercial programs and to more intensive programs. In reports of the Biggest Loser Club, participants self-reported their weight Game Characteristics and Social Network Factors loss less than once a week [18], social engagement was low Associated With Weight Loss (median interactions=0, range 0-0) [18], and overall weight loss The 39,387 players represented a total of 1934 games. Weight ranged from 2% to 3% [9,18]. Similarly, reports on the eDiets loss tended to cluster within games (P<.001; ICC=.07, consistent program indicate low engagement and modest weight loss (1% with a small effect) [16], suggesting that a player’s weight loss to 3%), even when combined with in-person visits and when was influenced by other players’ weight loss. On average, there evaluated in a randomized trial that would presumably improve were 31.4 (SD 171.2) players in each game. Game size predicted adherence and outcomes [8,11]. Given that the Biggest Loser weight loss; larger games were associated with greater weight Club and eDiets are longer programs (3- and 6-months, loss (P<.001). Games in which players knew one another had respectively) and that weight loss steadily increases during the slightly more social interactions than games in which players initial 4 to 6 months of treatment [19], these earlier programs did not know one another (ie, invite only/closed games vs open would be expected to yield a greater magnitude of weight loss games; 4.9, SD 6.1 vs 4.1, SD 5.1, P=.001); however, weight than DietBet. Instead the one month weight losses produced by loss outcomes were not affected by whether players knew one DietBet are as good as, if not better than, those produced by 3- another at the start of a game (P=.74). and 6-month programs. In fact, they are comparable to weight losses achieved during the initial four weeks of university-based, Discussion intensive behavioral weight loss programs that involve weekly in-person meetings led by professional staff [19]. While these Summary of DietBet Engagement and Weight Loss results are promising, in order to draw more definitive DietBet, a 4-week commercial Web-based program for weight conclusions regarding the weight loss produced by DietBet loss, yields retention results that are comparable to other versus other commerical weight loss programs, the length of programs [9,17] and produces excellent engagement and weight DietBet needs to be extended so that it is more consistent with loss. On average, participants interacted with other players on the length of these longer, more intensive commerical weight the DietBet platform more than twice a week and reported their loss interventions. weight loss at least once a week. A total of 71.71% of Explaining DietBet Results: Behavioral Economics, participants (28,244/39,387) submitted a self-reported weight at week 4 and 65.63% (25,849/39,387) completed final, Social Influence, and Self-Monitoring photo-based weigh-in verification procedures. Average The favorable outcomes produced by DietBet may be attributed intention-to-treat weight loss was 2.6%, and 42.39% of players to its social gaming components-namely, the use of financial (16,696 out of 39,387) achieved the 4% weight loss goal. incentives and social influence. Results showed that those who Moreover, over 5000 participants (n=5268) achieved a 5%, or bet more money and had greater social engagement had a greater clinically meaningful, weight loss. magnitude of weight loss and were more likely to “win” their game. These findings are consistent with those from behavioral http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen economics and with findings in the behavioral weight loss familiarity created by “invite only” games (vs open games) did literature. Behavioral economics suggests that loss aversion (the not affect weight outcomes. Combined with the clustering effect, strong tendency to avoid losing something that is owned) is a these results suggest that games are able to effectively create significant motivator of human behavior and that the magnitude social influence and promote social interactions for weight loss of loss may moderate the effect, with greater potential loss even among strangers. yielding greater motivation [20]. Consistent with this theory, Study Limitations in several randomized trials Jeffery et al showed that behavioral Study limitations include a predominantly female sample, lack weight loss programs involving deposit contracts (participants of some player characteristic data (age, race, ethnicity, etc), the deposited money and got it back for meeting goals) yielded short program length, and lack of fully objective weight data. significantly greater weight loss relative to the same behavioral Majority female participation is common in weight loss trials interventions without such contracts [21,22]. Moreover, and commercial weight loss programs [8,9,18]; however, given participants who deposited more money at baseline, and could the prevalence of obesity in men [2], future DietBet games may have therefore lost more money, were more likely to reach consider using targeted advertisements to increase male weight loss goals [22]. There is also strong evidence that enrollment. Additional player information (eg, height, age, race, combining financial incentives with social influence further ethnicity) would help to better describe the large sample of improves outcomes. Specifically, randomized trials have shown individuals who enrolled in DietBet and make comparisons that group incentives for meeting weight goals, either between the DietBet population and populations of other Internet collaborative or competitive, are more effective than individual commercial weight loss programs; DietBet has begun to collect incentives [14,21,23]. Thus, the excellent weight losses, such information on new enrollees. The program length was 4 retention, and engagement produced by DietBet are likely due weeks; while players achieved excellent weight losses in this to its use of principles from behavioral economics and its short period of time, and there is evidence that initial weight inclusion of financial incentive and social influence strategies. loss is predictive of future success [27], a longer program is Given this success, future Internet interventions, commercial warranted. As such, DietBet has developed and launched a or otherwise, may consider harnessing financial incentives and longer, 6-month game, which will allow us to examine social influence for weight loss to promote optimal outcomes. magnitude of weight loss produced over a longer timeframe. Players who reported their weight loss more often also lost more Finally, while the DietBet weigh-in system is more rigorous weight. Not only is regular reporting indicative of better program than the self-report methods used in other commercial programs engagement, which alone is associated with better outcomes [19], a systematic validation study comparing the DietBet [13], but regular weighing is also linked to better weight loss photo-based weigh-in system to objective weights obtained results [24], likely via the process of self-regulation [25]. from unbiased assessors is warranted. Specifically, getting on the scale on a regular basis yields Study Strengths important information on whether weight loss efforts are working and, if not, communicates the need to reduce dietary This study has several strengths. It provides a reliable and intake and increase physical activity to reach weight goals. Thus, ecologically valid examination of the true engagement, retention, consistent with intervention recommendations from and weight loss achieved in DietBet, a wide-reaching, university-based behavioral weight loss programs, DietBet commercial, Web-based weight loss program that uses financial players weighed themselves an average of at least once a week, incentives and social influence to promote weight loss in large which likely contributed to the positive weight loss results numbers of individuals. Only a small number of studies have achieved. conducted a naturalistic examination of such programs. Moreover, these previous studies have obtained only self-report In addition to examining individual effects, we also explored weight [17,18,28]. In contrast, the weigh-in procedures used in the effects of game characteristics on outcomes. Consistent with DietBet were not solely reliant on self-report; instead, players previous findings showing that weight loss clusters in were required to provide photo-based weight information and team-based weight loss competitions [12], individuals in the rigorous weight verification procedures were employed, with same game tended to achieve similar weight losses, suggesting human referees reviewing multiple photos per player. Moreover, social influence for weight loss among players. In addition, as indicated by the Federal Trade Commission, the data provided games with more players achieved greater weight loss overall. herein are critical to consumers who are inundated with While previous research shows that group size does not affect Web-based weight loss options, but have limited information weight loss outcomes in group-based interventions [12], this on effectiveness [29]. Finally, and perhaps most importantly, earlier study did not involve incentives. However, this earlier given that DietBet was able to reach nearly 40,000 individuals study did not involve incentives. Consistent with the behavior in just 7 months and, with the use of financial incentives and economics principle that reinforcement size is positively social influence, engage players and produce promising weight associated with response strength [26], the larger pool of money losses, DietBet may have the potential to impact public health inherent in bigger games may have motivated players to lose and help address the epidemic of obesity. more weight to, thereby, “win” their game. Interestingly, player Acknowledgments The authors would like to acknowledge all DietBet staff and players who made this report possible. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen Conflicts of Interest Mr Rosen is the chief executive officer of DietBet, Inc and receives salary support from the company. Multimedia Appendix 1 Intervention screenshot of player progress. [PNG File, 66KB-Multimedia Appendix 1] References 1. Stevens GA, Singh GM, Lu Y, Danaei G, Lin JK, Finucane MM, Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Body Mass Index). National, regional, and global trends in adult overweight and obesity prevalences. Popul Health Metr 2012;10(1):22 [FREE Full text] [doi: 10.1186/1478-7954-10-22] [Medline: 23167948] 2. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999-2008. JAMA 2010 Jan 20;303(3):235-241. [doi: 10.1001/jama.2009.2014] [Medline: 20071471] 3. National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in Adults--The evidence report. Obes Res 1998 Sep;6 Suppl 2:51S-209S. [Medline: 9813653] 4. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002 Feb 7;346(6):393-403 [FREE Full text] [doi: 10.1056/NEJMoa012512] [Medline: 11832527] 5. Neve M, Morgan PJ, Jones PR, Collins CE. Effectiveness of Web-based interventions in achieving weight loss and weight loss maintenance in overweight and obese adults: A systematic review with meta-analysis. Obes Rev 2010 Apr;11(4):306-321. [doi: 10.1111/j.1467-789X.2009.00646.x] [Medline: 19754633] 6. Saperstein SL, Atkinson NL, Gold RS. The impact of Internet use for weight loss. Obes Rev 2007 Sep;8(5):459-465. [doi: 10.1111/j.1467-789X.2007.00374.x] [Medline: 17716303] 7. Research Center P. Pew Research Group. Washington, DC: Pew Internet and American Life Project Internet adoption 1995-2013 URL: http://www.pewinternet.org/Data-Tools/Get-The-Latest-Statistics.aspx [accessed 2014-01-22] [WebCite Cache ID 6MpCqMa5h] 8. Gold BC, Burke S, Pintauro S, Buzzell P, Harvey-Berino J. Weight loss on the Web: A pilot study comparing a structured behavioral intervention to a commercial program. Obesity (Silver Spring) 2007 Jan;15(1):155-164. [doi: 10.1038/oby.2007.520] [Medline: 17228043] 9. Collins CE, Morgan PJ, Jones P, Fletcher K, Martin J, Aguiar EJ, et al. A 12-week commercial Web-based weight-loss program for overweight and obese adults: Randomized controlled trial comparing basic versus enhanced features. J Med Internet Res 2012;14(2):e57 [FREE Full text] [doi: 10.2196/jmir.1980] [Medline: 22555246] 10. Jeffery RW, French SA. Preventing weight gain in adults: The pound of prevention study. Am J Public Health 1999 May;89(5):747-751. [Medline: 10224988] 11. Womble LG, Wadden TA, McGuckin BG, Sargent SL, Rothman RA, Krauthamer-Ewing ES. A randomized controlled trial of a commercial internet weight loss program. Obes Res 2004 Jun;12(6):1011-1018. [doi: 10.1038/oby.2004.124] [Medline: 15229342] 12. Leahey TM, Kumar R, Weinberg BM, Wing RR. Teammates and social influence affect weight loss outcomes in a team-based weight loss competition. Obesity (Silver Spring) 2012 Jul;20(7):1413-1418 [FREE Full text] [doi: 10.1038/oby.2012.18] [Medline: 22310234] 13. Linde JA, Jeffery RW, French SA, Pronk NP, Boyle RG. Self-weighing in weight gain prevention and weight loss trials. Ann Behav Med 2005 Dec;30(3):210-216. [doi: 10.1207/s15324796abm3003_5] [Medline: 16336072] 14. Wing RR, Jeffery RW. Benefits of recruiting participants with friends and increasing social support for weight loss and maintenance. J Consult Clin Psychol 1999 Feb;67(1):132-138. [Medline: 10028217] 15. Perri MG, McAdoo WG, McAllister DA, Lauer JB, Jordan RC, Yancey DZ, et al. Effects of peer support and therapist contact on long-term weight loss. J Consult Clin Psychol 1987 Aug;55(4):615-617. [Medline: 3624623] 16. Killip S, Mahfoud Z, Pearce K. What is an intracluster correlation coefficient? Crucial concepts for primary care researchers. Ann Fam Med 2004;2(3):204-208 [FREE Full text] [Medline: 15209195] 17. Longin R, Grasse M, Aspalter R, Waldherr K. Effectiveness of the online weight reduction program KiloCoach™ and comparison with other evaluated commercial direct intervention and online programs. Obes Facts 2012;5(3):372-383. [doi: 10.1159/000339726] [Medline: 22722385] 18. Neve M, Morgan PJ, Collins CE. Weight change in a commercial Web-based weight loss program and its association with website use: Cohort study. J Med Internet Res 2011;13(4):e83 [FREE Full text] [doi: 10.2196/jmir.1756] [Medline: 21993231] 19. Wing RR. Behavioral approaches to the treatment of obesity. In: Bray GA, Bouchard C, editors. Handbook of obesity: clinical applications. New York: Informa Healthcare; 2008:227-248. 20. Kahneman D, Knetsch J, Tahler R. Journal of Political Economy. 1990. Experimental tests of the endowment effect and coase theorem URL: http://www.jstor.org/discover/10.2307/ http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen 2937761?uid=3739920&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21103390662183[WebCite Cache ID 6MsNLQmkm] 21. Jeffery RW, Gerber WM, Rosenthal BS, Lindquist RA. Monetary contracts in weight control: Effectiveness of group and individual contracts of varying size. J Consult Clin Psychol 1983 Apr;51(2):242-248. [Medline: 6841768] 22. Jeffery RW, Bjornson-Benson WM, Rosenthal BS, Kurth CL, Dunn MM. Behavior Therapy. 1984. Effectiveness of monetary contracts with two repayment schedules on weight reduction in men and women from self-referred and population samples URL: http://www.sciencedirect.com/science/article/pii/S0005789484800295 [accessed 2014-01-24] [WebCite Cache ID 6MsNWuEKN] 23. Kullgren JT, Troxel AB, Loewenstein G, Asch DA, Norton LA, Wesby L, et al. Individual- versus group-based financial incentives for weight loss: A randomized, controlled trial. Ann Intern Med 2013 Apr 2;158(7):505-514. [doi: 10.7326/0003-4819-158-7-201304020-00002] [Medline: 23546562] 24. Vanwormer JJ, French SA, Pereira MA, Welsh EM. The impact of regular self-weighing on weight management: A systematic literature review. Int J Behav Nutr Phys Act 2008;5:54 [FREE Full text] [doi: 10.1186/1479-5868-5-54] [Medline: 18983667] 25. Wing RR, Tate DF, Gorin AA, Raynor HA, Fava JL. A self-regulation program for maintenance of weight loss. N Engl J Med 2006 Oct 12;355(15):1563-1571. [doi: 10.1056/NEJMoa061883] [Medline: 17035649] 26. Ghezzi PA, Lyons CA, Dixon MR. Gambling in socioeconomic perspective. In: Bickel WK, Vuchinich RE, editors. Reframing health behavior change with behavioral economics. Mahwah, N.J: Lawrence Erlbaum; 2000:313-338. 27. Nackers LM, Ross KM, Perri MG. The association between rate of initial weight loss and long-term success in obesity treatment: Does slow and steady win the race? Int J Behav Med 2010 Sep;17(3):161-167 [FREE Full text] [doi: 10.1007/s12529-010-9092-y] [Medline: 20443094] 28. Neve MJ, Collins CE, Morgan PJ. Dropout, nonusage attrition, and pretreatment predictors of nonusage attrition in a commercial Web-based weight loss program. J Med Internet Res 2010;12(4):e69 [FREE Full text] [doi: 10.2196/jmir.1640] [Medline: 21156470] 29. Federal Trade Commission. Commercial weight loss products and programs: What consumers stand to gain and lose. Washington, DC: Federal Trade Commission; 1997. URL: http://www.tandfonline.com/doi/abs/10.1080/ 20014091091733?journalCode=bfsn20#preview[WebCite Cache ID 6MsOlP4XY] Abbreviations ICC: intraclass correlation coefficient Edited by G Eysenbach; submitted 27.09.13; peer-reviewed by C Collins, J Nunez-Cordoba, K Hwang; comments to author 30.10.13; revised version received 11.12.13; accepted 11.01.14; published 07.02.14 Please cite as: Leahey T, Rosen J JMIR Serious Games 2014;2(1):e2 URL: http://games.jmir.org/2014/1/e2/ doi: 10.2196/games.2987 PMID: 25658966 ©Tricia Leahey, Jamie Rosen. Originally published in JMIR Serious Games (http://games.jmir.org), 07.02.2014. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org/, as well as this copyright and license information must be included. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 8 (page number not for citation purposes) XSL FO RenderX http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JMIR Serious Games JMIR Publications

DietBet: A Web-Based Program that Uses Social Gaming and Financial Incentives to Promote Weight Loss

Leahey, Tricia; Rosen, Jamie
JMIR Serious Games , Volume 2 (1) – Feb 7, 2014
Download PDF
8 pages

Loading next page...
 
/lp/jmir-publications/dietbet-a-web-based-program-that-uses-social-gaming-and-financial-jKht4sVep2

References (33)

Publisher
JMIR Publications
Copyright
Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
ISSN
2291-9279
DOI
10.2196/games.2987
Publisher site
See Article on Publisher Site

Abstract

Background: Web-based commercial weight loss programs are increasing in popularity. Despite their significant public health potential, there is limited research on the effectiveness of such programs. Objective: The objective of our study was to examine weight losses produced by DietBet and explore whether baseline and engagement variables predict weight outcomes. Methods: DietBet is a social gaming website that uses financial incentives and social influence to promote weight loss. Players bet money and join a game. All players have 4 weeks to lose 4% of their initial body weight. At enrollment, players can choose to share their participation on Facebook. During the game, players interact with one another and report their weight loss on the DietBet platform. At week 4, all players within each game who lose at least 4% of initial body weight are declared winners and split the pool of money bet at the start of the game. Official weigh-in procedures are used to verify weights at the start of the game and at the end. Results: From December 2012 to July 2013, 39,387 players (84.04% female, 33,101/39,387; mean weight 87.8kg, SD 22.6kg) competed in 1934 games. The average amount bet was US $27 (SD US $22). A total of 65.63% (25,849/39,387) provided a verified weight at the end of the 4-week competition. The average intention-to-treat weight loss was 2.6% (SD 2.3%). Winners (n=17,171) won an average of US $59 (SD US $35) and lost 4.9% (SD 1.0%) of initial body weight, with 30.68% (5268/17,171) losing 5% or more of their initial weight. Betting more money at game entry, sharing on Facebook, completing more weigh-ins, and having more social interactions during the game predicted greater weight loss and greater likelihood of winning (Ps<.001). In addition, weight loss clustered within games (P<.001), suggesting that players influenced each others’ weight outcomes. Conclusions: DietBet, a social gaming website, reached nearly 40,000 individuals in just 7 months and produced excellent 4-week weight loss results. Given its reach and potential public health impact, future research may consider examining whether a longer program promotes additional weight loss. (JMIR Serious Games 2014;2(1):e2) doi: 10.2196/games.2987 KEYWORDS commercial weight loss; Web-based intervention; social gaming; financial incentives http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen This study examined the effectiveness of DietBet, which is a Introduction commercially available Web-based program that uses social gaming and financial incentives to promote weight loss. Upon Obesity Prevalence and Treatment Availability enrollment, players join a game and enter money into a pool. Over 45% of the world population and approximately 66% of All players have 4 weeks to lose 4% of their initial body weight. American adults are overweight or obese [1,2]. Excess adiposity During the game, players report their weight and interact on the is associated with serious health risks, including diabetes, DietBet platform. At the end of the game, all players within cardiovascular disease, and some types of cancer [2,3]. each game who lose at least 4% in 4 weeks are declared Behavioral weight loss programs consistently yield weight loss “winners” and split the initial pool of money bet at enrollment. of 8 to 10kg, which are associated with significant health The primary aim of this study was to conduct a naturalistic improvements, including reduced risk for diabetes [4]. However, examination of engagement, retention, and weight loss outcomes behavioral programs are largely university-based and are, in DietBet. Previous findings from financial incentive weight therefore, only available to a small portion of overweight and loss trials have shown that the possibility of losing large (vs obese individuals in need. small) amounts of money for not meeting weight goals motivates better overall weight loss [10]. In addition, more frequent Accessibility of Internet Commercial Weight Loss self-weighing and greater social influence for weight loss have Programs been found to be associated with better weight outcomes In contrast, commercially available weight loss programs have [12-15]. Given these findings, we explored whether: (1) betting wide reach, particularly Web-based interventions. In fact, a more money at enrollment, (2) completing more weigh-ins, and large percentage of individuals who attempt weight loss report (3) having greater social engagement/influence predict greater using commercial programs [5], and, in recent years, enrollment percent weight loss and greater likelihood of winning. Finally, in Web-based interventions has increased substantially [6]. The given evidence that weight loss clusters in social networks and appeal of Internet interventions is likely due to the reduced that group characteristics impact weight outcomes in participant burden associated with frequent in-person visits (eg, group-based weight loss competitions [12], we also explored time, transportation). Another important benefit of Web-based whether weight loss clusters within games (ie, players in the interventions is their inherent reach; given that Internet access same game achieve similar weight loss) and whether game has increased exponentially over the past decade [7], Web-based characteristics (eg, number of players) are associated with programs are widely accessible to individuals who may not weight outcomes. otherwise have access to clinical weight management interventions. Similarly, such programs are accessible 24-hours Methods a day in a variety of locations (home, work, public libraries). Given their appeal and ability to reach large numbers of Procedures overweight and obese individuals in need, commercially DietBet is a social gaming website that uses financial incentives available Web-based interventions have significant public health and social influence to promote weight loss. Players are recruited potential. via press coverage (eg, Today Show, CNN, New York Times, Evidence for Commercial Weight Loss Programs Wall Street Journal), business development efforts (popular wellness experts with social capital are asked to host games and Despite their potential to improve public health, the scientific encourage fans/followers to participate), direct virality (players literature on the effectiveness of Web-based commercial weight recruit other players), and indirect virality (players share DietBet loss programs is sparse [5]. Only two programs have been information on Facebook). At enrollment, players bet money rigorously evaluated in randomized trials: eDiets [8,9] and The and join a game. Players are given the option to join an existing Biggest Loser Club [10]. Results of the eDiets trials showed game that has not yet started or create their own game. If they that participants assigned to eDiets achieved significantly less create their own game, it can be either open (anyone can join) weight loss than those assigned to a self-help condition or an or closed (invite only; ie, all players know each other). Internet behavioral weight loss program [8,11]. Similarly, while the Biggest Loser Club produced greater weight loss than a no Players are prompted to submit their official start weight two treatment control condition, given that the program was 3 days prior to the start of the game, (see below for weigh-in months in length, the weight loss was modest (-2.1kg) [9]. These procedures). Once the game begins, players have 28 days to randomized trials have clear benefits and are essential to lose 4% of their initial body weight. DietBet does not promote demonstrate efficacy. However, an important shortcoming is a specific diet or weight loss strategy; instead, players are that, given the nature of rigorous, randomized trials (screenings, allowed to choose any strategy to lose weight (eg, low-fat diets, run-ins, retention efforts, etc), the results may overinflate true low-carbohydrate diets, etc). During the game, players post engagement, retention, and weight losses outcomes of photos, comments, and weight loss tips (Figure 1 shows a commercially available programs. Thus, to complement the screenshot). They are also encouraged to stay accountable to clinical trial literature, ecologically valid studies are needed that one another by posting their weight loss. Players can view their examine real-world enrollees in naturally occurring Web-based weight loss relative to the weight loss of others in the game (see commercial weight loss programs and, thus, ascertain true Multimedia Appendix 1). To facilitate game communication program engagement, retention, and outcome data. Results from and sharing, players have access to the DietBet app, which is such studies may be used to inform consumer decision making designed for all smartphone mobile devices and includes all and public policy. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen aspects of the gaming platform. At the end of 28 days, players people won, after DietBet’s cut (see below), they would each have 48 hours to send in their final weight. All players who lose win US $50. If no one loses 4% of their starting weight, the at least 4% of their initial body weight in 28 days split the pool player who lost the most weight wins the pool of money. of money that was bet at the start of the game. For example, if Winners are notified by email of their payout, which they can the game consisted of 10 players who each bet US $25 and 4 either apply to their next DietBet game or cash out. Figure 1. Intervention screenshot. DietBet keeps a portion of the initial pool of money to cover identify players who have won more than US $300 in DietBets financial transaction costs and staff time associated with weight and identify players who have unusual weight gain between verification, customer service, and tech support. DietBet’s share DietBet games (ie, gained more than 1% per day). Those is taken out of each game’s starting pool of money before individuals are flagged and DietBet staff completes a detailed winners are paid. Thus, players who do not win do not incur a review of their weight, game activity, and profile information. charge to play. DietBet retains 15%-25% of the payout Using these algorithms and the photo-based system within 48 depending on the amount of each individual bet (US hours of the start and end of a game, if a weight is deemed $1-$99–25% fee; US $100-$249–20% fee; US $250 or questionable, players are required to submit extra proof by more–15% fee). completing a live Skype weigh-in with one of DietBet's referees, a video weigh-in, or an in-person weigh-in at a DietBet approved Measures location (eg, Walgreens drugstores or Customer Value and Service drugstores). Gender and Weight Participants report their gender at the beginning of the game. Weigh-Ins, Social Engagement, and Facebook Shares Within 48 hours prior to the start of each game, participants are All weigh-in, social engagement, and Facebook data were required to complete an official weigh-in. Similarly, within 48 obtained directly from the DietBet website. Specifically, during hours after the end of a game, participants are to complete an the game, players reported their weight on the DietBet platform official weigh-in. Official weigh-in procedures involve the and weigh-ins were summed for each player. Players interacted submission of two photos, one of the player on a scale in light with one another by cheering, posting pictures, posting status indoor clothing without shoes and another that captures the updates, “liking” others’ posts, and commenting on and replying scale’s readout and a piece of paper that includes the player’s to posts; to create an overall social engagement variable, the weigh-in password. DietBet staff reviews each photograph for number of social interactions for each variable (cheers, pictures, player-password correspondence. In addition, DietBet uses updates, etc) were summed for each player. DietBet collected prespecified algorithms to detect any unusual weight outcomes data on whether players posted their game participation on and game activity; specifically, across games DietBet will Facebook; data were coded 1 for “Yes” or 0 for “No.” The http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen option to share on Facebook was not available in early games. predominantly female (84.04%, 33,101/39,387) with a mean Thus, the sample size for this variable is smaller than the overall baseline weight of 87.8kg (SD 22.6kg). The average amount of sample (n=20,059 vs N=39,387). money bet at game entry was US $26.84 (SD US $21.93). Upon enrolling, 50.03% of players chose to share their DietBet Money Bet and Won participation on Facebook (note–the option to share on Facebook The amount of money bet and the amount of money won was was not available in earlier games, thus the total sample size obtained objectively via payment transaction data. for this variable was n=20,059, of those n=10,036 shared their DietBet participation). During the 4-week game, players Game Characteristics completed an average of 5.3 (SD 3.9) weigh-ins and engaged The number of players on each game and whether all players in 9.3 (SD 78.4) social interactions (eg, cheers, posts, likes, etc). knew each other (closed game/invite only vs open game) were A total of 71.71% of participants (28,244/39,387) self-reported obtained objectively by extracting data from the platform. their weight into the DietBet platform during week 4, and Statistical Analyses 65.63% (25,849/39,387) completed an official weigh-in at the Player characteristics, engagement, and completion outcomes end of the game (ie, completed the photo-based weigh-in process were examined using simple descriptive statistics, including immediately following the game). A greater proportion of men means, standard deviations, and percentages. Completers (ie, completed a photo-verified weigh-in than women (Men–68.39%, those who completed an official weigh-in procedure at the end 4292 out of 6275; Women–65.10%, 21,550 out of 33,101; of the game) and noncompleters were compared using analyses P<.001). Compared to noncompleters, completers weighed less of variance or chi-square tests for continuous or categorical at baseline (87.4kg, SD 22.2kg vs 88.5kg, SD 23.3kg, P<.001), variables, respectively. To examine the effects of DietBet on bet more money at program entry (US $27.79, SD US $24.99 weight loss, a conservative baseline carried forward vs US $25.01, SD US $14.21, P<.001), and had more weigh-ins intention-to-treat approach was used; players who did not finish (6.5, SD 4.0 vs 3.0, SD 2.5, P<.001) and more social interactions the game (ie, did not complete a final, verified weigh-in) were with their teammates during the game (eg, cheers, comments, assumed to have remained at baseline/entry weight and within likes, etc; 12.0, SD 95.3 vs 4.3, SD 22.6, P<.001). A greater subjects t tests were conducted. Predictors of weight loss and percentage of completers versus noncompleters shared their whether a player won their game were examined with regression game participation on Facebook (50.96%, 7074 out of 13,882 analyses. In addition, to determine which variables explained completers vs 47.74%, 2949 out of 6177; P<.001). unique variance in weight loss, a multivariate analysis that Weight Loss included all variables was conducted. The effects of game and Intention-to-treat analyses (assuming that individuals who did game characteristics on weight outcomes were also explored. not complete an official, photo-based weigh-in at the end of the To examine whether weight loss clustered within games, an game did not lose any weight) showed that players lost a unconditional multilevel model was conducted and an intraclass significant amount of weight from baseline to the end of the correlation coefficient (ICC) was calculated using the resulting game (P<.001; Table 1). The average weight loss was 2.6% between and within group variance components, (SD 2.3%). Out of the 39,387 players who enrolled, 43.60% (ICC=U /U +R). 0 0 (n=17,171) were winners (ie, lost at least 4% of initial body Ethics weight or, if no one lost 4%, lost the most weight in their game). The average amount won was US $58.79 (SD US $34.90) (net The Miriam Hospital’s Institutional review board approved this earnings–US $29.00, SD US $16.43). Game winners lost an study. average of 4.9% (SD 1.0%) of initial body weight. A total of 30.68% of winners (n=5268) achieved a 5%, or clinically Results meaningful, weight loss. Engagement and Completion From December 2012 to July 2013, 39,387 players participated in 1934 games on the DietBet platform. Players were Table 1. Player weight loss. Total sample All winners Winners who lost ≥ 4% Winners who lost most weight in game (N=39,387) (n=17,171) (n=16,687) (n=484) % weight loss, mean (SD) 2.6 (2.3) 4.9 (1.0) 4.9 (0.9) 3.4 (0.9) money at game entry (P<.001), completing more weigh-ins Participant Characteristics Associated With Weight during the game (P<.001), sharing game participation on Loss and Winning the Game Facebook (P<.001), and having more social interactions with Participant baseline characteristics and game variables predicted other players (P<.001). These same variables were also weight loss outcomes and whether a player won their game. significant predictors of whether a player won their game Significant predictors of greater percent weight loss were male (Ps<.001; Table 2). gender (P<.001), lower baseline weight (P=.03), betting more http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen Table 2. Winners’ versus nonwinners’ baseline characteristics and engagement. Winners Nonwinners Winners versus nonwinners (n=17,171) (n=8678) (P value) Male, n (%) 3489 (20.32) 803 (9.25) <.001 Baseline weight kg, mean (SD) 87.1 (22.1) 88.1 (22.4) <.001 Amount bet US $, mean (SD) $29.79 ($29.18) $23.83 ($12.30) <.001 Shared on Facebook, n (%) 4837 (53.19) 2241 (46.79) <.001 Weigh-ins, mean (SD) 7.0 (4.2) 5.4 (3.1) <.001 Social engagement, mean (SD) 13.6 (113.6) 8.8 (38.4) <.001 Given that completion was associated with baseline characteristics and engagement, to control for potential confounding of completion, only game completers were included in these analyses, n=25,849. Option to share on Facebook was not available in early games. Thus, the sample size for this variable is smaller than the overall sample of completers. Specifically, of those who were given the option to share on Facebook, n=20,059, a total of n=13,882 completed the game. Of those, n=9093 were winners and n=4789 were nonwinners. Sum of cheers, comments, replies, likes, photos, and updates. In a multivariate model, all variables remained significant, Comparison of DietBet Results to Other Internet independent predictors of weight loss (Ps<.001) with the Commericial Weight Loss Programs exception of number of social interactions and whether a player DietBet results compare favorably to other Web-based shared on Facebook (Ps>.38). commercial programs and to more intensive programs. In reports of the Biggest Loser Club, participants self-reported their weight Game Characteristics and Social Network Factors loss less than once a week [18], social engagement was low Associated With Weight Loss (median interactions=0, range 0-0) [18], and overall weight loss The 39,387 players represented a total of 1934 games. Weight ranged from 2% to 3% [9,18]. Similarly, reports on the eDiets loss tended to cluster within games (P<.001; ICC=.07, consistent program indicate low engagement and modest weight loss (1% with a small effect) [16], suggesting that a player’s weight loss to 3%), even when combined with in-person visits and when was influenced by other players’ weight loss. On average, there evaluated in a randomized trial that would presumably improve were 31.4 (SD 171.2) players in each game. Game size predicted adherence and outcomes [8,11]. Given that the Biggest Loser weight loss; larger games were associated with greater weight Club and eDiets are longer programs (3- and 6-months, loss (P<.001). Games in which players knew one another had respectively) and that weight loss steadily increases during the slightly more social interactions than games in which players initial 4 to 6 months of treatment [19], these earlier programs did not know one another (ie, invite only/closed games vs open would be expected to yield a greater magnitude of weight loss games; 4.9, SD 6.1 vs 4.1, SD 5.1, P=.001); however, weight than DietBet. Instead the one month weight losses produced by loss outcomes were not affected by whether players knew one DietBet are as good as, if not better than, those produced by 3- another at the start of a game (P=.74). and 6-month programs. In fact, they are comparable to weight losses achieved during the initial four weeks of university-based, Discussion intensive behavioral weight loss programs that involve weekly in-person meetings led by professional staff [19]. While these Summary of DietBet Engagement and Weight Loss results are promising, in order to draw more definitive DietBet, a 4-week commercial Web-based program for weight conclusions regarding the weight loss produced by DietBet loss, yields retention results that are comparable to other versus other commerical weight loss programs, the length of programs [9,17] and produces excellent engagement and weight DietBet needs to be extended so that it is more consistent with loss. On average, participants interacted with other players on the length of these longer, more intensive commerical weight the DietBet platform more than twice a week and reported their loss interventions. weight loss at least once a week. A total of 71.71% of Explaining DietBet Results: Behavioral Economics, participants (28,244/39,387) submitted a self-reported weight at week 4 and 65.63% (25,849/39,387) completed final, Social Influence, and Self-Monitoring photo-based weigh-in verification procedures. Average The favorable outcomes produced by DietBet may be attributed intention-to-treat weight loss was 2.6%, and 42.39% of players to its social gaming components-namely, the use of financial (16,696 out of 39,387) achieved the 4% weight loss goal. incentives and social influence. Results showed that those who Moreover, over 5000 participants (n=5268) achieved a 5%, or bet more money and had greater social engagement had a greater clinically meaningful, weight loss. magnitude of weight loss and were more likely to “win” their game. These findings are consistent with those from behavioral http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen economics and with findings in the behavioral weight loss familiarity created by “invite only” games (vs open games) did literature. Behavioral economics suggests that loss aversion (the not affect weight outcomes. Combined with the clustering effect, strong tendency to avoid losing something that is owned) is a these results suggest that games are able to effectively create significant motivator of human behavior and that the magnitude social influence and promote social interactions for weight loss of loss may moderate the effect, with greater potential loss even among strangers. yielding greater motivation [20]. Consistent with this theory, Study Limitations in several randomized trials Jeffery et al showed that behavioral Study limitations include a predominantly female sample, lack weight loss programs involving deposit contracts (participants of some player characteristic data (age, race, ethnicity, etc), the deposited money and got it back for meeting goals) yielded short program length, and lack of fully objective weight data. significantly greater weight loss relative to the same behavioral Majority female participation is common in weight loss trials interventions without such contracts [21,22]. Moreover, and commercial weight loss programs [8,9,18]; however, given participants who deposited more money at baseline, and could the prevalence of obesity in men [2], future DietBet games may have therefore lost more money, were more likely to reach consider using targeted advertisements to increase male weight loss goals [22]. There is also strong evidence that enrollment. Additional player information (eg, height, age, race, combining financial incentives with social influence further ethnicity) would help to better describe the large sample of improves outcomes. Specifically, randomized trials have shown individuals who enrolled in DietBet and make comparisons that group incentives for meeting weight goals, either between the DietBet population and populations of other Internet collaborative or competitive, are more effective than individual commercial weight loss programs; DietBet has begun to collect incentives [14,21,23]. Thus, the excellent weight losses, such information on new enrollees. The program length was 4 retention, and engagement produced by DietBet are likely due weeks; while players achieved excellent weight losses in this to its use of principles from behavioral economics and its short period of time, and there is evidence that initial weight inclusion of financial incentive and social influence strategies. loss is predictive of future success [27], a longer program is Given this success, future Internet interventions, commercial warranted. As such, DietBet has developed and launched a or otherwise, may consider harnessing financial incentives and longer, 6-month game, which will allow us to examine social influence for weight loss to promote optimal outcomes. magnitude of weight loss produced over a longer timeframe. Players who reported their weight loss more often also lost more Finally, while the DietBet weigh-in system is more rigorous weight. Not only is regular reporting indicative of better program than the self-report methods used in other commercial programs engagement, which alone is associated with better outcomes [19], a systematic validation study comparing the DietBet [13], but regular weighing is also linked to better weight loss photo-based weigh-in system to objective weights obtained results [24], likely via the process of self-regulation [25]. from unbiased assessors is warranted. Specifically, getting on the scale on a regular basis yields Study Strengths important information on whether weight loss efforts are working and, if not, communicates the need to reduce dietary This study has several strengths. It provides a reliable and intake and increase physical activity to reach weight goals. Thus, ecologically valid examination of the true engagement, retention, consistent with intervention recommendations from and weight loss achieved in DietBet, a wide-reaching, university-based behavioral weight loss programs, DietBet commercial, Web-based weight loss program that uses financial players weighed themselves an average of at least once a week, incentives and social influence to promote weight loss in large which likely contributed to the positive weight loss results numbers of individuals. Only a small number of studies have achieved. conducted a naturalistic examination of such programs. Moreover, these previous studies have obtained only self-report In addition to examining individual effects, we also explored weight [17,18,28]. In contrast, the weigh-in procedures used in the effects of game characteristics on outcomes. Consistent with DietBet were not solely reliant on self-report; instead, players previous findings showing that weight loss clusters in were required to provide photo-based weight information and team-based weight loss competitions [12], individuals in the rigorous weight verification procedures were employed, with same game tended to achieve similar weight losses, suggesting human referees reviewing multiple photos per player. Moreover, social influence for weight loss among players. In addition, as indicated by the Federal Trade Commission, the data provided games with more players achieved greater weight loss overall. herein are critical to consumers who are inundated with While previous research shows that group size does not affect Web-based weight loss options, but have limited information weight loss outcomes in group-based interventions [12], this on effectiveness [29]. Finally, and perhaps most importantly, earlier study did not involve incentives. However, this earlier given that DietBet was able to reach nearly 40,000 individuals study did not involve incentives. Consistent with the behavior in just 7 months and, with the use of financial incentives and economics principle that reinforcement size is positively social influence, engage players and produce promising weight associated with response strength [26], the larger pool of money losses, DietBet may have the potential to impact public health inherent in bigger games may have motivated players to lose and help address the epidemic of obesity. more weight to, thereby, “win” their game. Interestingly, player Acknowledgments The authors would like to acknowledge all DietBet staff and players who made this report possible. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen Conflicts of Interest Mr Rosen is the chief executive officer of DietBet, Inc and receives salary support from the company. Multimedia Appendix 1 Intervention screenshot of player progress. [PNG File, 66KB-Multimedia Appendix 1] References 1. Stevens GA, Singh GM, Lu Y, Danaei G, Lin JK, Finucane MM, Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group (Body Mass Index). National, regional, and global trends in adult overweight and obesity prevalences. Popul Health Metr 2012;10(1):22 [FREE Full text] [doi: 10.1186/1478-7954-10-22] [Medline: 23167948] 2. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999-2008. JAMA 2010 Jan 20;303(3):235-241. [doi: 10.1001/jama.2009.2014] [Medline: 20071471] 3. National Institutes of Health. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in Adults--The evidence report. Obes Res 1998 Sep;6 Suppl 2:51S-209S. [Medline: 9813653] 4. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002 Feb 7;346(6):393-403 [FREE Full text] [doi: 10.1056/NEJMoa012512] [Medline: 11832527] 5. Neve M, Morgan PJ, Jones PR, Collins CE. Effectiveness of Web-based interventions in achieving weight loss and weight loss maintenance in overweight and obese adults: A systematic review with meta-analysis. Obes Rev 2010 Apr;11(4):306-321. [doi: 10.1111/j.1467-789X.2009.00646.x] [Medline: 19754633] 6. Saperstein SL, Atkinson NL, Gold RS. The impact of Internet use for weight loss. Obes Rev 2007 Sep;8(5):459-465. [doi: 10.1111/j.1467-789X.2007.00374.x] [Medline: 17716303] 7. Research Center P. Pew Research Group. Washington, DC: Pew Internet and American Life Project Internet adoption 1995-2013 URL: http://www.pewinternet.org/Data-Tools/Get-The-Latest-Statistics.aspx [accessed 2014-01-22] [WebCite Cache ID 6MpCqMa5h] 8. Gold BC, Burke S, Pintauro S, Buzzell P, Harvey-Berino J. Weight loss on the Web: A pilot study comparing a structured behavioral intervention to a commercial program. Obesity (Silver Spring) 2007 Jan;15(1):155-164. [doi: 10.1038/oby.2007.520] [Medline: 17228043] 9. Collins CE, Morgan PJ, Jones P, Fletcher K, Martin J, Aguiar EJ, et al. A 12-week commercial Web-based weight-loss program for overweight and obese adults: Randomized controlled trial comparing basic versus enhanced features. J Med Internet Res 2012;14(2):e57 [FREE Full text] [doi: 10.2196/jmir.1980] [Medline: 22555246] 10. Jeffery RW, French SA. Preventing weight gain in adults: The pound of prevention study. Am J Public Health 1999 May;89(5):747-751. [Medline: 10224988] 11. Womble LG, Wadden TA, McGuckin BG, Sargent SL, Rothman RA, Krauthamer-Ewing ES. A randomized controlled trial of a commercial internet weight loss program. Obes Res 2004 Jun;12(6):1011-1018. [doi: 10.1038/oby.2004.124] [Medline: 15229342] 12. Leahey TM, Kumar R, Weinberg BM, Wing RR. Teammates and social influence affect weight loss outcomes in a team-based weight loss competition. Obesity (Silver Spring) 2012 Jul;20(7):1413-1418 [FREE Full text] [doi: 10.1038/oby.2012.18] [Medline: 22310234] 13. Linde JA, Jeffery RW, French SA, Pronk NP, Boyle RG. Self-weighing in weight gain prevention and weight loss trials. Ann Behav Med 2005 Dec;30(3):210-216. [doi: 10.1207/s15324796abm3003_5] [Medline: 16336072] 14. Wing RR, Jeffery RW. Benefits of recruiting participants with friends and increasing social support for weight loss and maintenance. J Consult Clin Psychol 1999 Feb;67(1):132-138. [Medline: 10028217] 15. Perri MG, McAdoo WG, McAllister DA, Lauer JB, Jordan RC, Yancey DZ, et al. Effects of peer support and therapist contact on long-term weight loss. J Consult Clin Psychol 1987 Aug;55(4):615-617. [Medline: 3624623] 16. Killip S, Mahfoud Z, Pearce K. What is an intracluster correlation coefficient? Crucial concepts for primary care researchers. Ann Fam Med 2004;2(3):204-208 [FREE Full text] [Medline: 15209195] 17. Longin R, Grasse M, Aspalter R, Waldherr K. Effectiveness of the online weight reduction program KiloCoach™ and comparison with other evaluated commercial direct intervention and online programs. Obes Facts 2012;5(3):372-383. [doi: 10.1159/000339726] [Medline: 22722385] 18. Neve M, Morgan PJ, Collins CE. Weight change in a commercial Web-based weight loss program and its association with website use: Cohort study. J Med Internet Res 2011;13(4):e83 [FREE Full text] [doi: 10.2196/jmir.1756] [Medline: 21993231] 19. Wing RR. Behavioral approaches to the treatment of obesity. In: Bray GA, Bouchard C, editors. Handbook of obesity: clinical applications. New York: Informa Healthcare; 2008:227-248. 20. Kahneman D, Knetsch J, Tahler R. Journal of Political Economy. 1990. Experimental tests of the endowment effect and coase theorem URL: http://www.jstor.org/discover/10.2307/ http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR SERIOUS GAMES Leahey & Rosen 2937761?uid=3739920&uid=2129&uid=2&uid=70&uid=4&uid=3739256&sid=21103390662183[WebCite Cache ID 6MsNLQmkm] 21. Jeffery RW, Gerber WM, Rosenthal BS, Lindquist RA. Monetary contracts in weight control: Effectiveness of group and individual contracts of varying size. J Consult Clin Psychol 1983 Apr;51(2):242-248. [Medline: 6841768] 22. Jeffery RW, Bjornson-Benson WM, Rosenthal BS, Kurth CL, Dunn MM. Behavior Therapy. 1984. Effectiveness of monetary contracts with two repayment schedules on weight reduction in men and women from self-referred and population samples URL: http://www.sciencedirect.com/science/article/pii/S0005789484800295 [accessed 2014-01-24] [WebCite Cache ID 6MsNWuEKN] 23. Kullgren JT, Troxel AB, Loewenstein G, Asch DA, Norton LA, Wesby L, et al. Individual- versus group-based financial incentives for weight loss: A randomized, controlled trial. Ann Intern Med 2013 Apr 2;158(7):505-514. [doi: 10.7326/0003-4819-158-7-201304020-00002] [Medline: 23546562] 24. Vanwormer JJ, French SA, Pereira MA, Welsh EM. The impact of regular self-weighing on weight management: A systematic literature review. Int J Behav Nutr Phys Act 2008;5:54 [FREE Full text] [doi: 10.1186/1479-5868-5-54] [Medline: 18983667] 25. Wing RR, Tate DF, Gorin AA, Raynor HA, Fava JL. A self-regulation program for maintenance of weight loss. N Engl J Med 2006 Oct 12;355(15):1563-1571. [doi: 10.1056/NEJMoa061883] [Medline: 17035649] 26. Ghezzi PA, Lyons CA, Dixon MR. Gambling in socioeconomic perspective. In: Bickel WK, Vuchinich RE, editors. Reframing health behavior change with behavioral economics. Mahwah, N.J: Lawrence Erlbaum; 2000:313-338. 27. Nackers LM, Ross KM, Perri MG. The association between rate of initial weight loss and long-term success in obesity treatment: Does slow and steady win the race? Int J Behav Med 2010 Sep;17(3):161-167 [FREE Full text] [doi: 10.1007/s12529-010-9092-y] [Medline: 20443094] 28. Neve MJ, Collins CE, Morgan PJ. Dropout, nonusage attrition, and pretreatment predictors of nonusage attrition in a commercial Web-based weight loss program. J Med Internet Res 2010;12(4):e69 [FREE Full text] [doi: 10.2196/jmir.1640] [Medline: 21156470] 29. Federal Trade Commission. Commercial weight loss products and programs: What consumers stand to gain and lose. Washington, DC: Federal Trade Commission; 1997. URL: http://www.tandfonline.com/doi/abs/10.1080/ 20014091091733?journalCode=bfsn20#preview[WebCite Cache ID 6MsOlP4XY] Abbreviations ICC: intraclass correlation coefficient Edited by G Eysenbach; submitted 27.09.13; peer-reviewed by C Collins, J Nunez-Cordoba, K Hwang; comments to author 30.10.13; revised version received 11.12.13; accepted 11.01.14; published 07.02.14 Please cite as: Leahey T, Rosen J JMIR Serious Games 2014;2(1):e2 URL: http://games.jmir.org/2014/1/e2/ doi: 10.2196/games.2987 PMID: 25658966 ©Tricia Leahey, Jamie Rosen. Originally published in JMIR Serious Games (http://games.jmir.org), 07.02.2014. This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Serious Games, is properly cited. The complete bibliographic information, a link to the original publication on http://games.jmir.org/, as well as this copyright and license information must be included. http://games.jmir.org/2014/1/e2/ JMIR Serious Games 2014 | vol. 2 | iss. 1 | e2 | p. 8 (page number not for citation purposes) XSL FO RenderX

Journal

JMIR Serious Games – JMIR Publications

Published: Feb 7, 2014

Keywords: commercial weight loss; Web-based intervention; social gaming; financial incentives

There are no references for this article.