Eric Stice, Ph.D. - US grants

The investigator's goal is to pursue a research career in an academic setting with a substantive focus on the etiology of adolescent psychopathology. In particular, he is interested in understanding the factors that promote bulimic pathology. Although the applicant's past training has provided him with a foundation in conducting longitudinal studies, none of his prior mentors have been eating disorder researchers. The central aims of the proposed Research Career Award (RCA) will be to (1) provide advanced training in research methods and quantitative techniques necessary for him to develop into an independent investigator and (2) to conduct two investigations that will advance understanding of the etiology of bulimic pathology. His training mentors will be Drs. Agras, Taylor, and Kreamer from Stanford University School of Medicine. The applicant would like to develop a unique system of inquiry that pairs prospective community studies with naturalistic randomized experiments in an effort to produce findings with maximal external and internal validity. Dr. Agras will provide training in conducting randomized experiments, Dr. Taylor in conducting longitudinal studies, and Dr. Kraemer in the use of quantitative techniques. Dr. Agras will serve as Primary Mentor and will coordinate the career development activities. Under the guidance of these mentors, the investigator will conduct a four year longitudinal community study of adolescents. This investigation will attempt to validate six putative risk factors for bulimic pathology that have emerged from the literature within a study that addresses some of the limitations of past research and employs advanced analytic techniques. This longitudinal study will be paired with a randomized experiment that attempts to corroborate the findings from prospective research. Specifically, this investigation will experimentally decrease body dissatisfaction (with a validated intervention), and assess the effect on consequent bulimic attitudes and behaviors. Furthermore, because past eating disorder research has focused almost exclusively on main effects, these two studies will also examine factors that may potentiate or buffer (moderate) the main effects of these six risk factors. He will also (1) participate in ongoing projects in Drs. Agras' and Taylor's labs, (2) learn state-of- the-art treatments for eating pathology and other adolescent psychopathology, (3) take courses in quantitative techniques, methodology, and ethics, and (4) receive training in grant writing.

DESCRIPTION (provided by applicant): Bulimic pathology is one of the most common psychiatric problems faced by adolescent girls, shows a persistent course, can result in serious medical complications, and is associated with psychiatric comorbidity and functional impairment. Although longitudinal studies have identified risk factors that increase the likelihood of bulimic pathology, these factors show modest predictive power and little is known about resource factors that decrease risk for bulimic pathology, protective and potentiating factors that mitigate or amplify the effects of risk factors, and how these factors "work together" in a multivariate sense. There have also been few studies on maintenance factors that predict bulimic pathology persistence or the role of psychiatric comorbidity in the etiology and maintenance of this disorder. Further, past studies suffered from certain methodological limitations (e.g., small samples, short follow-up periods, high attrition, mono-reporter data, and sole reliance on survey data). The primary aims of this study are to examine putative risk and resource factors for bulimic pathology and the protective and potentiating factors that moderate these effects, and to test multivariate etiologic models. Secondary aims are to examine putative maintenance factors for bulimic pathology and the role of psychiatric comorbidity in the onset and persistence of this pathology. A tertiary aim is to further document the impact of bulimic pathology on subjective distress, functional impairment, and health care utilization. Aims will be achieved by following a large representative cohort of girls ( N = 496), which has been assessed annually for 4 years since early adolescence (modal age = 13 at T1), for an additional 4 years. When complete, this cohort will have been followed annually for eight years, making it the longest prospective study of adolescent bulimic pathology to date. Key limitations of past studies will be addressed (e.g., by collecting multiple-reporter data and using structured interviews) and advanced quantitative methods will be utilized. Thus, the proposed continuation of this longitudinal study should advance our knowledge of the etiologic and maintenance processes for bulimic pathology and provide direction for the design of prevention and treatment programs for this serious psychiatric problem.

DESCRIPTION (provided by applicant): Bulimic pathology is one of the most common psychiatric problems faced by adolescent girls, shows a persistent course, can result in serious medical complications, and is associated with psychiatric comorbidity and functional impairment. Although longitudinal studies have identified risk factors that increase the likelihood of bulimic pathology, these factors show modest predictive power and little is known about resource factors that decrease risk for bulimic pathology, protective and potentiating factors that mitigate or amplify the effects of risk factors, and how these factors "work together" in a multivariate sense. There have also been few studies on maintenance factors that predict bulimic pathology persistence or the role of psychiatric comorbidity in the etiology and maintenance of this disorder. Further, past studies suffered from certain methodological limitations (e.g., small samples, short follow-up periods, high attrition, mono-reporter data, and sole reliance on survey data). The primary aims of this study are to examine putative risk and resource factors for bulimic pathology and the protective and potentiating factors that moderate these effects, and to test multivariate etiologic models. Secondary aims are to examine putative maintenance factors for bulimic pathology and the role of psychiatric comorbidity in the onset and persistence of this pathology. A tertiary aim is to further document the impact of bulimic pathology on subjective distress, functional impairment, and health care utilization. Aims will be achieved by following a large representative cohort of girls ( N = 496), which has been assessed annually for 4 years since early adolescence (modal age = 13 at T1), for an additional 4 years. When complete, this cohort will have been followed annually for eight years, making it the longest prospective study of adolescent bulimic pathology to date. Key limitations of past studies will be addressed (e.g., by collecting multiple-reporter data and using structured interviews) and advanced quantitative methods will be utilized. Thus, the proposed continuation of this longitudinal study should advance our knowledge of the etiologic and maintenance processes for bulimic pathology and provide direction for the design of prevention and treatment programs for this serious psychiatric problem.

DESCRIPTION (provided by applicant): Because obesity is very prevalent, results in marked morbidity and mortality, and is treatment resistant, it is vital to develop prevention programs for this public health problem. Unfortunately, virtually all prevention programs have not reduced risk for future weight gain. We developed a selected prevention program that targets the 50 percent of female adolescents with the highest body dissatisfaction because a preliminary study indicated that they have an 8-fold increased risk for obesity onset over a 4-year follow-up relative to those below this cut-off. In the Healthy Weight program, female adolescents with body image concerns are encouraged to reduce caloric intake and to increase physical activity as a way of improving body satisfaction. This intervention uses persuasion principles from social and clinical psychology (e.g., strategic self-presentation, motivational enhancement techniques) to increase the likelihood that participants will make lasting behavioral changes. It is very brief (3 hrs) relative to extant programs (M = 43 hrs), which should facilitate dissemination. A preliminary trial (N = 188) found that participants who completed the Healthy Weight program were at significantly lower risk for obesity onset over a 2-year follow-up (3 percent) than assessment-only controls (12 percent) and showed greater improvements in healthy eating and physical activity. This program also reduced eating disorder symptoms and negative affect, which is important because these disturbances result in functional impairment and adverse outcomes. We propose to conduct a large randomized trial of an enhanced version of this obesity prevention program. We will randomly assign 405 female adolescents (aged 17-19) with body image concerns from a local college to the Healthy Weight intervention or a psychoeducational control condition and follow them for 2 years. We will test whether the Healthy Weight program, relative to the control intervention, results in (1) decreased risk for obesity onset and increased weight, (2) decreased caloric intake and increased physical activity, and (3) decreased body dissatisfaction, bulimic symptoms, and depressive symptoms. We will also (4) test whether changes in caloric intake and activity level mediate intervention effects on weight gain prevention and (5) investigate potential moderators of intervention effects. The use of random assignment, an attention control condition, blinded diagnostic interviews, and a long-term follow-up make the proposed trial more rigorous than past trials. We will also validate key measures in a subgroup of participants with objective measures (e.g., doubly-labeled water and accelerometers). The development of an effective obesity prevention program that can be easily and inexpensively disseminated could substantially reduce the rate of obesity and resulting morbidity and mortality. The fact that this intervention also produces effects for eating disorder symptoms and depression further increases the public health relevance of the proposed trial.

DESCRIPTION (provided by applicant): Independent efficacy trials indicate that a selective eating disorder prevention program involving dissonance- inducing activities that decrease thin-ideal internalization in high-risk female high school and college students with body dissatisfaction reduces eating disorder risk factors, eating disorder symptoms, and risk for onset of future eating disorders and obesity relative to alternative interventions and assessment-only control conditions, with some effects persisting through 3-year follow-up. A large effectiveness trial found that this intervention reduced risk factors and eating disorder symptoms relative to a psychoeducational brochure control condition through 2-year follow-up when high school counselors and nurses recruited participants and delivered the intervention. Yet the effects from this latter trial were slightly smaller than those in the efficacy trial, potentially due to limited infrastructure, funds, and staff to support prevention programs in high schools. It is thus possible that this intervention will produce larger effects in an effectiveness trial involving colleges, as they typically have staff devoted to mental health programming with group intervention training and access to appropriate facilities. Although we conducted our initial effectiveness trial with high school students, there have been no effectiveness trials of empirically supported eating disorder prevention programs for college students; this is an important lacuna because the environment, resources, and endogenous providers are very different for colleges versus high schools and because there are over 1 million US female college students with threshold or subthreshold eating disorders. A large effectiveness trial involving multiple regions with a diverse sample would also permit an investigation of facilitator and participant factors that predict intervention effect size. We propose to conduct a 3-site effectiveness trial that tests whether this prevention program produces effects when college counselors, psychologists, and nurses recruit, screen, and deliver the intervention under ecologically valid conditions. We will randomize 432 high-risk young women with body dissatisfaction recruited from 9 colleges in Oregon, Pennsylvania, and Texas to the dissonance program or a psychoeducational brochure control condition. We will use an enhanced facilitator training procedure and newly published facilitator manuals and participant workbooks. Participants will complete a survey and a diagnostic interview at pretest, posttest, and 1- and 2-year follow-ups. We will test whether (1) the dissonance program results in larger reductions in risk factors, eating disorder symptoms, risk for future eating disorder and obesity onset, functional impairment, and mental health treatment relative to controls, (2) intervention effects are mediated by change in thin-ideal internalization, (3) certain factors moderate program effects (e.g., initial body dissatisfaction level), (4) process factors (e.g., intervention fidelity) predict larger effect sizes, (5) the savings produced by this intervention offset intervention delivery costs, and (6) colleges continue to offer the prevention program after they complete the 1-year intervention-delivery phase of this trial (i.e., assess sustainability). PUBLIC HEALTH RELEVANCE: Threshold and subthreshold eating disorders affect over 10% of adolescent girls and are associated with functional impairment, distress, medical complications, risk for obesity onset, and the highest mortality of any psychiatric condition; accordingly, a pressing public healthy priority is to develop effective prevention programs for eating pathology. The proposed project will be the first effectiveness trial to test whether an eating disorder prevention program with strong empirical support from efficacy trials produces effects under ecologically valid conditions among high-risk female college students, which is a vital step toward widespread dissemination of programs developed with NIH funding. The proposed cost-effectiveness analyses and examination of process factors that predict larger intervention effects will also represent novel contributions to the literature.

DESCRIPTION (provided by applicant): Few obesity prevention programs have reduced risk for weight gain over follow-up and those that have are very intensive (M duration = 52 hrs.), making dissemination difficult and costly. A brief 3-hr selective prevention program (Healthy Weight) involving participant-driven healthy dietary and physical activity lifestyle changes significantly reduced increases in BMI and obesity onset relative to alternative interventions and assessment- only controls through 3-yr follow-up. It targets young adults with body dissatisfaction as this is a prevalent risk factor for obesity and such individuals are motivated t make healthy lifestyle changes. To enhance efficacy, we added activities designed to induce dissonance regarding unhealthy dietary and activity practices, drawing from a highly efficacious dissonance-based eating disorder prevention program. Social psychology experiments show that people strive to maintain consistency between thoughts, words, and actions. Thus, we added verbal, written, and behavioral exercises wherein youth discuss costs of obesity, an unhealthy diet, and sedentary behaviors, benefits of leanness, a healthy diet, and physical activity, and institutions that contribute to obesity. Engaging in these activities putatively increases the likelihood that participants will seek to align their actions with the perspectives expressed in the sessions, resulting in healthier future lifestyle choices. A pilot trial found tha this new Project Health intervention significantly reduced increases in BMI relative to both the Healthy Weight intervention and an educational brochure condition from pre to post (d = .40 and .42 respectively), comparing favorably to effect sizes from the most efficacious intensive prevention programs. Further, other obesity prevention programs have not significantly outperformed credible alternative interventions. We propose to conduct a rigorous multisite effectiveness trial that will test whether adding the dissonance-induction elements to the originally efficacious Healthy Weight intervention improves weight gain prevention effects: 300 college students at risk for future unhealthy weight gain by virtue of body dissatisfaction will be randomized to a refined 8-hr dissonance-based Project Health intervention or an 8-hr Healthy Weight intervention, completing assessments of body fat and other outcomes at pre, post, and 6, 12, and 24 month follow-ups. Aim 1 will test whether Project Health reduces increases in % body fat relative to Healthy Weight; secondary outcomes are depressive and eating disorder symptoms. Aim 2 will test whether Project Health participants report cognitive dissonance regarding eating unhealthy foods and engaging in sedentary behaviors and whether changes in dissonance, as well as reported dietary intake and objectively measured activity level, mediate the effects of the intervention on change in % body fat. Aim 3 will investigate potential moderators that may amplify (pretest body dissatisfaction, preference for consistency) or mitigate (emotional eating, substance use) intervention effects. We will improve upon most obesity prevention trials by using a long-term follow-up, an active control condition, and direct measures of body composition and activity. PUBLIC HEALTH RELEVANCE: A brief effective obesity prevention program that can be easily and inexpensively disseminated could substantially reduce the incidence of obesity, which is vital because obesity is prevalent, causes marked morbidity and mortality, and is treatment resistant. The proposed project, which is based on promising preliminary data from rigorous prevention trials and a highly efficacious dissonance-based eating disorder prevention program, would be the first translational project to use persuasion principles from social psychology for obesity prevention.

DESCRIPTION (provided by applicant): Efficacy trials show that a dissonance-based selective prevention program that decreases pursuit of the unrealistic thin-ideal in high-risk young women with body image concerns reduces eating disorder risk factors (e.g., body dissatisfaction, negative affect), eating disorder symptoms, functional impairment, and future eating disorder onset relative to alternative interventions and assessment-only controls, with some effects lasting 3 yrs (Becker et al., 2005; Mitchell et al., 2007; Stice et al., 2000, 2006, 2008b, 2011c). An effectiveness trial found that when high school clinicians recruit students and deliver the program that it produces greater reductions in outcomes relative to educational brochure controls, including reduced eating disorder symptoms through 3-yr follow-up (Stice et al., 2009b). Larger effects are emerging from an effectiveness trial of an enhanced-dissonance version of this program with college clinicians (Prelim Studies). These trials confirm the effectiveness of this intervention, but reveal it can be difficult to identify and recruit clinicias at high schools and colleges with the time and expertise to capably deliver this prevention program. One solution to this dissemination barrier is to train college students in peer leader programs to recruit students and deliver the program, as peer-led interventions appear promising, sometimes producing larger effects than clinician-led versions of the intervention (Botvin et al., 1984; Rhee et al., 2011). Use of peer leaders could markedly extend the reach and sustainability of this program. In a pilot trial, peer-led groups of the enhanced-dissonance program produced significantly greater pre-post reductions in thin-ideal internalization, body dissatisfaction, dieting, negative effect, and eating disorder symptoms relative to educational brochure controls (M d = .76), producing effects similar to those from clinician-led groups. Effects compare favorably to those from clinician- led groups in our college effectiveness trial (M d = .66) and from uncontrolled studies of a peer-led version of the dissonance program adapted for sororities (M d = .46; Becker et al., 2010; Perez et al., 2010). We propose to conduct a large multi-site effectiveness trial of peer-led enhanced-dissonance groups, wherein 480 female college students with body image concerns will be randomized to peer-led groups, clinician-led groups, or an educational video control condition. Aim 1: test whether participants who complete peer-led groups show greater reductions in eating disorder risk factors, symptoms, eating disorder onset, and functional impairment over 3-yr follow-up than educational video controls and reductions similar to those who complete clinician-led groups. Aim 2: test whether fidelity and competence ratings are similar for peer-led vs. clinician-led groups. Aim 3: test whether reductions in thin-ideal internalization mediate the intervention effects on other downstream outcomes and examine moderators of intervention effects (e.g., elevated baseline thin-ideal internalization) for peer- and clinician-led groups. Aim 4: test whether cost savings and incremental cost-effectiveness will be greater for peer-led groups vs. clinician-led groups relative to the educational video control condition.

ABSTRACT Only 3-20% of people with eating disorders (EDs) receive treatment, and they often do not receive evidence- based treatments because they are intensive and costly, and few clinicians deliver them. These factors have hindered broad implementation, implying that a brief frontline outpatient treatment for a range of EDs that could be easily, cheaply, and widely implemented would address a key public health problem. There are also limited experimental data regarding factors that maintain EDs and mechanisms of action for ED treatments. We hypothesize that EDs are maintained by (1) excessive valuation of the thin ideal, which prompts caloric restriction and other unhealthy weight control behaviors (vomiting, laxative/diuretic use, excessive exercise) that increase risk for binge eating, and (2) excessive valuation of high-calorie foods, which maintains binge eating. We propose to evaluate a group treatment that efficiently targets these 2 maintenance factors. We created a novel 8-session dissonance-based treatment (Counter Attitudinal Therapy; CAT) wherein women with any ED complete verbal, written, and behavioral activities in which they discuss costs of pursuing the thin ideal and the ED behaviors they use to purse this ideal, which putatively creates dissonance about engaging in these behaviors that reduces valuation of the thin ideal and high-calorie binge foods. In two pilot trials CAT reduced behaviorally-assessed valuation of thin models and binge foods, and blinded interviewer-assessed ED symptoms versus unstandardized (d=.94) and standardized usual care (d=.53), resulting in a 70% remission by 2-6 month follow-up across the two trials. Effect sizes compare favorably to those produced by 20-session individual therapies for EDs. Women who completed the dissonance-inducing activities regarding pursuit of the thin ideal showed a reduction in fMRI-assessed reward region response to thin models (Stice et al., 2015) providing further evidence of target engagement. The R61 randomized trial (N=60) will (1) validate whether CAT produces larger pre-post reductions in objective fMRI-assessed reward region response to thin models and binge foods (intervention target measures) versus waitlist controls in women with DSM-5 EDs; (2) generate preliminary data regarding the clinical effects of reducing the targets on ED symptom domains and functioning (outcomes); and (3) test whether reductions in targets and outcomes show a linear decrease over the 8 sessions or plateau earlier and correlate with greater session attendance and homework completion, to examine dose-response relations. If the R61 confirms that CAT sufficiently engages intervention targets and that the 8-session format produces optimal response, the fully-powered R33 randomized trial (N=120) will test whether (4) CAT produces greater reductions in the two fMRI-assessed targets than a usual care group treatment; (5) CAT produces greater reductions in ED symptoms and functional impairment than the usual care treatment through 6-month follow-up; and (6) change in targets correlates with change in ED symptoms and functioning, and mediates intervention effects on outcomes.

ABSTRACT Only 3-20% of people with eating disorders (EDs) receive treatment, and they often do not receive evidence- based treatments because they are intensive and costly, and few clinicians deliver them. These factors have hindered broad implementation, implying that a brief frontline outpatient treatment for a range of EDs that could be easily, cheaply, and widely implemented would address a key public health problem. There are also limited experimental data regarding factors that maintain EDs and mechanisms of action for ED treatments. We hypothesize that EDs are maintained by (1) excessive valuation of the thin ideal, which prompts caloric restriction and other unhealthy weight control behaviors (vomiting, laxative/diuretic use, excessive exercise) that increase risk for binge eating, and (2) excessive valuation of high-calorie foods, which maintains binge eating. We propose to evaluate a group treatment that efficiently targets these 2 maintenance factors. We created a novel 8-session dissonance-based treatment (Counter Attitudinal Therapy; CAT) wherein women with any ED complete verbal, written, and behavioral activities in which they discuss costs of pursuing the thin ideal and the ED behaviors they use to purse this ideal, which putatively creates dissonance about engaging in these behaviors that reduces valuation of the thin ideal and high-calorie binge foods. In two pilot trials CAT reduced behaviorally-assessed valuation of thin models and binge foods, and blinded interviewer-assessed ED symptoms versus unstandardized (d=.94) and standardized usual care (d=.53), resulting in a 70% remission by 2-6 month follow-up across the two trials. Effect sizes compare favorably to those produced by 20-session individual therapies for EDs. Women who completed the dissonance-inducing activities regarding pursuit of the thin ideal showed a reduction in fMRI-assessed reward region response to thin models (Stice et al., 2015) providing further evidence of target engagement. The R61 randomized trial (N=60) will (1) validate whether CAT produces larger pre-post reductions in objective fMRI-assessed reward region response to thin models and binge foods (intervention target measures) versus waitlist controls in women with DSM-5 EDs; (2) generate preliminary data regarding the clinical effects of reducing the targets on ED symptom domains and functioning (outcomes); and (3) test whether reductions in targets and outcomes show a linear decrease over the 8 sessions or plateau earlier and correlate with greater session attendance and homework completion, to examine dose-response relations. If the R61 confirms that CAT sufficiently engages intervention targets and that the 8-session format produces optimal response, the fully-powered R33 randomized trial (N=120) will test whether (4) CAT produces greater reductions in the two fMRI-assessed targets than a usual care group treatment; (5) CAT produces greater reductions in ED symptoms and functional impairment than the usual care treatment through 6-month follow-up; and (6) change in targets correlates with change in ED symptoms and functioning, and mediates intervention effects on outcomes.

ABSTRACT Obesity causes 300,000 US deaths yearly, but most treatments do not result in lasting weight loss. People who show greater brain reward and attention region response, and less inhibitory region response, to high-calorie food images/cues show elevated future weight gain (Demos et al., 2012; Stice et al., 2015; Yokum et al., 2011, 2014), consistent with the theory that overeating results from a strong approach response to high-calorie food cues paired with a weak inhibitory response (Wiers et al., 2007). This implies that an intervention that reduces reward and attention region response to such food and increases inhibitory control region response should reduce overeating that is rooted in exposure to pervasive food cues. Computer-based response-inhibition training with high-calorie foods has decreased attentional bias for and intake of the training food, increased inhibitory control, and produced weight loss in overweight participants in 3 proof-of-concept trials, with effects persisting through 6-mo follow-up (Allom & Mullan, 2015; Kemps et al., 2014b; Lawrence et al., 2015b; Veling et al., 2014). A pilot trial found that overweight/obese adults who completed a multi-faceted 4-hr response- inhibition training with high-calorie food images and response-facilitation training with low-calorie food images showed reduced fMRI-assessed reward and attention region response to high-calorie training foods and greater body fat loss than controls who completed a rigorous 4-hr generic response-inhibition/response- facilitation training with non-food images (d=.95), producing a 7% reduction in excess body fat over the 4-wk period. We propose to evaluate a refined and extended version of this response-training intervention. Aim 1 is to randomize 180 overweight/obese adults to a 4-wk response training obesity treatment or a generic inhibition training control condition that both include bi-monthly Internet-delivered booster training for a year and a smart phone response training app that can be used when tempted by high-calorie foods, assessing outcomes at pre, post, and at 3-, 6-, and 12-month follow-ups (e.g., % body fat, the primary outcome). Aim 2 is to use fMRI to test whether reduced reward and attention region response, and increased inhibitory region response to high-calorie food images used and not used in the response training mediate the effects of the intervention on fat loss. We will also test whether during training participants show acute reductions in reward and attention region response, and increases in inhibitory response to high-calorie training food images to capture the learning process, assess generalizability of the intervention to food images not used in training, and collect behavioral data on mediators. Aim 3 is to test whether intervention effects will be stronger for those who show less inhibitory control in response to high-calorie food images, a genetic propensity for greater dopamine signaling in reward circuitry, and greater pretest reward and attention region response, and weaker inhibitory region response to high-calorie food images, based on the theory that response training is more efficacious for those with a strong pre-potent approach tendency to high-calorie foods.