2 Hoyt et al. Brown,1981).Thus,the me eac d bd dcx (BMD. individuals-the very people these public-health messages We expected obese individuals to be more concemed are targeting (Teixeira,Mata,Williams,Gorin,&Lemieux, with their health and weight than individuals of average 2012 be (Neumark 1997 edicted that the disease message,relative es dise obesity is not a disease,would undercut the importance Study 2.we also tested whether reduced concern for that obese individuals place on dieting to improve their weight predicted higher-calorie food choices. health and undermine their concem lor the food).We also investigated whether the disease message and health problems (questions adapted from Blokstra with its focus on physiological explanations,improve Burns,Seidell,1999).We asked participants to indicate erate bod bese individ t 18074 tomorrow or were currently In study 2.participants indicated instead how con that such body-image concerns are a stronger motivating cemed they were with watching their weight on a 7-point force than health c emns for engaging in healthy behav ors (e.g.,M a me that oh ease alters crucial psychological pr cesses Ge concern order from the following choices:The Italian (980 calo about weight,body image)underlying obese individuals' ries). ed Turkey Breast (360 calories),Roast Bee motivation to engage in healthy behaviors (i.e.,choosing and Chedda (700 calories). BLT (634 calories),and lower-calorie food light(23 ed the number Norman,2010,for using parametric statistics with this Study 1 and Study 2 type of scale). Method edure Results Ordinary least squares(OLS)regt ession equations revealed (Buhrmester,Kwang,Gosling,2011).One hundred an interaction between participants'BMI (Study 1 o山 1M 261kem”6o6nmge1g50.sdy2 27.19kgm range oded)and me in Study 2.We randomly assigned participants to tion had an effect on both health either a recent New York Times article (Pollack,2013) focused dieting (b=-0.04,B=-0.13.p=.082;see Fig.1) discussing the decision of the American Medical Associa and concem for weight (b= -0.04,B ity as a d ora control articl or con effects across public-health message about weight-In the Time tive to the rol mpeop cle,the author summarizes some of the major benefits of decreases in health-focused dieting in Study 1(=-0.21. this decision,such as compensation for obesity-related B-0.12.p25)and significant decreases in concern surgery.counseling.an or weignt in st y 2 (b: 4,P 31,p .002 In the info ion based c ntrol article the author high dicted nonsignific lights standard tips and tools for managing weight,such health-focused dieting (=025 B 014=18)and as keeping exercise activities interesting and monitoring weight weekly. 3,201 2 Hoyt et al. attempts to regulate behavior (Carver & Scheier, 2011; Finlay-Jones & Brown, 1981). Thus, the message that obesity is a disease may undermine important psycho￾logical determinants of salubrious behaviors for obese individuals—the very people these public-health messages are targeting (Teixeira, Mata, Williams, Gorin, & Lemieux, 2012). We predicted that the disease message, relative to a standard weight-management message or a message that obesity is not a disease, would undercut the importance that obese individuals place on dieting to improve their health and would undermine their concern for their weight. In addition, this reduced concern would predict less healthy food choices (i.e., selection of higher calorie food). We also investigated whether the disease message, with its focus on physiological explanations, improves body satisfaction (e.g., Crandall, 1994) but whether such increases have costs. For obese individuals, mild or mod￾erate body dissatisfaction can serve as a motivator to reduce consumption. Indeed, some research illustrates that such body-image concerns are a stronger motivating force than health concerns for engaging in healthy behav￾iors (e.g., McDonald & Thompson, 1992). In summary, we examined whether the message that obesity is a dis￾ease alters crucial psychological processes (i.e., concern about weight, body image) underlying obese individuals’ motivation to engage in healthy behaviors (i.e., choosing lower-calorie foods). Study 1 and Study 2 Method Participants and procedure. For Studies 1 and 2, we recruited participants from Amazon’s Mechanical Turk (Buhrmester, Kwang, & Gosling, 2011). One hundred eighty-two participants (44% female, 56% male; median age = 30 years) took part in Study 1, and 185 participants (38% female, 62% male; median age = 27 years) took part in Study 2.1 We randomly assigned participants to read either a recent New York Times article (Pollack, 2013) discussing the decision of the American Medical Associa￾tion to categorize obesity as a disease or a control article (Luedtke, 2011) offering a standard information-based public-health message about weight.2 In the Times arti￾cle, the author summarizes some of the major benefits of this decision, such as compensation for obesity-related drugs, surgery, counseling, and reduced stigma against people who are obese, as well as some of the drawbacks. In the information-based control article, the author high￾lights standard tips and tools for managing weight, such as keeping exercise activities interesting and monitoring weight weekly. After reading their respective article, participants in each condition responded to measures, which were fol￾lowed by demographic questions including height and weight so we could compute body mass index (BMI).3 We expected obese individuals to be more concerned with their health and weight than individuals of average weight would be (Neumark-Sztainer et al., 1997). However, notably, we expected these self-regulatory pro￾cesses to be undermined by the disease message. In Study 2, we also tested whether reduced concern for weight predicted higher-calorie food choices. Measures. In Study 1, we presented participants with two health-focused reasons for dieting: medical advice and health problems (questions adapted from Blokstra, Burns, & Seidell, 1999). We asked participants to indicate, on a 9-point scale, how important each reason would be to them if they started a diet tomorrow or were currently dieting, r(180) = .74, p < .001. In Study 2, participants indicated instead how con￾cerned they were with watching their weight on a 7-point scale ranging from very unconcerned to very concerned. We also presented participants in Study 2 with a menu and asked them to indicate which sandwich they would order from the following choices: The Italian (980 calo￾ries), Smoked Turkey Breast (360 calories), Roast Beef and Cheddar (700 calories), BLT (634 calories), and Vegetable Delight (230 calories). We used the number of calories each participant ordered for analyses (see Norman, 2010, for using parametric statistics with this type of scale). Results Ordinary least squares (OLS) regression equations revealed an interaction between participants’ BMI (Study 1: M = 26.71 kg/m2 , SD = 6.06, range = 18–50; Study 2: M = 27.19 kg/m2 , SD = 7.07, range = 16–51) and message type (disease vs. control; effects coded 1 and −1, respec￾tively), and this interaction had an effect on both health￾focused dieting (b = −0.04, β = −0.13, p = .082; see Fig. 1) and concern for weight (b = −0.04, β = −0.31, p = .005; see Fig. 2).4 Tests of conditional effects across BMI revealed that for obese people, the disease message, rela￾tive to the control message, predicted nonsignificant decreases in health-focused dieting in Study 1 (b = −0.21, β = −0.12, p = .25) and significant decreases in concern for weight in Study 2 (b = −0.47, β = −0.31, p = .002). The opposite pattern emerged for people of average weight; the disease message predicted nonsignificant increases in health-focused dieting (b = 0.25, β = 0.14, p = .18) and concern for weight (b = 0.15, β = 0.10, p = .31). BMI significantly predicted health-focused dieting (b = 0.06, Downloaded from pss.sagepub.com by Cai Xing on February 13, 2014