were to (a) examine the effect of a self-efficacy intervention on task, barrier, scheduling, goal- setting and relapse efficacy in relatives of colon cancer patients undertaking a 9-month home-based PA program, and (b) examine the self-efficacy and self-regulation model in predicting objective PA behavior. METHOD: Following a 12-week self- efficacy and structured exercise program, relatives of colon cancer patients (N 5 107; M 5 45.7 years, 78.7; 91.0% first-degree relatives; 67.3% female) participated in a 9-month home based PA pro- gram. Participants remained in the same rando- mized condition: self-efficacy (intervention- INT) or nutrition (attention control- AC). The interven- tion group received monthly emails that focused on self-efficacy information (e.g., task, scheduling, barrier, relapse), while the attention control group received nutrition information. Physical activity (i.e., energy expenditure; EE 5 daily kcals/min/kg) was objectively assessed using Actical s accelero- meters at 1 week following the structured exercise program, and at 1, 3, 6, and 9 months. Self-efficacy was assessed at all these time points except month 1. RESULTS: A repeated measure ANOVA revealed significant time effects for task (Z 2 5 .18), scheduling (Z 2 5 .28), goal-setting (Z 2 5 .39), and relapse (Z 2 5 .37), whereby all measures decreased across time. A significant group by time interaction effect was present for barrier efficacy (Z 2 5 .13), scores decreased across all time points in the AC condition, while barrier efficacy scores initially declined, followed by an increase and subsequent decrease across assessment points in the INT condition. Proxy self-efficacy variables that showed a relationship to EE were entered into a regression (i.e., week 1 self-efficacy predicting PA at month 1). Results revealed that week 1 self-efficacy variables predicted EE at month 6, and month 3 self-efficacy variables predicted EE at month 6, accounting for 11.1% and 8.8% of the response variance respec- tively. CONCLUSIONS: With the exception of barrier efficacy, the intervention was unsuccessful in preventing decreases in self-efficacious beliefs in a 9 month home-based exercise program. Self- efficacious beliefs are associated with objective measures of PA over a 9-month home-based exercise program. RESEARCH IMPLICATIONS: This trial is the first to explore the following issues in a population of relatives of colon cancer patients: (a) the effect of an intervention grounded in self-efficacy theory on self-efficacious beliefs (i.e., task and self- regulatory), and (b) whether self-efficacious beliefs predict objectively measured PA patterns. Results suggest that although self-efficacious beliefs fostered through initial experiences during an exercise pro- gram are associated with objective measures of PA behavior, these beliefs may be difficult to maintain. CLINICAL IMPLICATIONS: Alternative modes of delivering self-efficacy (i.e., motivational inter- viewing) based interventions need to be tested in order to find more effective ways to maintain efficacious beliefs over long-term home-based ex- ercise programs. ACKNOWLEDGEMENT OF FUNDING: The authors would like to thank the Social Sciences and Humanities Research Council of Canada (SSHRC) for funding this research study. IL 2.20.1 Theoretical and Evidence-Based Development of the Profile of Preferences for Cancer Information Scale Carmen Loiselle, Sylvie Lambert McGill University, Montreal, Canada BACKGROUND: Individuals diagnosed with cancer vary in terms of their preferences for the type, amount, and sources of cancer information they wish to be exposed to. Scales currently available to measure cancer information-seeking behaviour (ISB) mainly focus on the dichotomy of active seeking and information avoidance. However, recent evidence point to more nuanced cancer ISB patterns. In line with this evidence, we propose the Profile of Preferences for Cancer Information (PPCI) scale to comprehensively measure ISB. METHOD: Initially guided by Self-evaluation Theory, a set of items were developed to capture different cancer ISB. This initial tool was pilot tested among women newly diagnosed with breast cancer ( n 5 60). Preliminary findings suggested that the tool discriminated among various ISB patterns; however psychometric properties were questionable particularly for the avoidance subscale. A subsequent in-depth concept analysis of ISB followed by a grounded theory study among 62 individuals with various cancer diagnoses served to further delineate various patterns or profiles of cancer ISB. Based on this evidence, the initial tool was revised to optimally capture the more delineated ISB patterns that emerged. RESULTS: The resulting PPCI is a 20- item, self-report instrument that measures the follow- ing five information-seeking profiles: 1) intense, 2) complementary, 3) fortuitous, 4) minimal, and 5) guarded. These profiles are characterized by differing in terms of amount/reasons, type, source, and overall management of cancer information. Individuals in- dicate, from (1) not at all like me, to (5) totally like me, the extent to which each item, such as ‘I prefer the internet as a source of scientific cancer information’ (intense profile) and ‘I am mostly interested in seeking information from the perspective of others diagnosed with cancer’ (fortuitous profile), reflect their ISB. The PPCI is currently being tested among individuals diagnosed with melanoma and colorectal cancers. CONCLUSIONS: The PPCI is a promising, theore- tically, and empirically sound tool that can assist researchers and clinicians in capturing differential information-seeking profiles. RESEARCH IMPLI- CATIONS: Future studies will test the PPCI in samples differing in socioeconomic status, health literacy, culture, and with more varied cancer S74 Individual Lectures of the IPOS 11th World Congress r 2009 The Authors. Journal Compilation r 2009 John Wiley & Sons, Ltd. Psycho-Oncology 18(Suppl. 2):S1–S330 (2009) DOI: 10.1002/pon