Evaluation of theory-based eating disorder prevention program: Triangulation of two methodologies to identify relationships between key theoretical constructs

Background: Oral Health Providers have an opportunity for early detection of disordered eating behaviors as they are often the first health professionals to observe overt signs. An intervention was developed, based on the Information Motivation Behavior (IMB) and Health Belief (HBM) Models, aimed at increasing secondary prevention (SP) behaviors among oral health students. As part of the larger RCT, this study's aim was to identify relationships between key theoretical mechanisms. Methods: Path analysis and qualitative comparative analysis (QCA) were employed to test interrelationships between IMB contingencies and HBM constructs. Path models were evaluated via fit indices, path coefficients, and variance explained in mediating variables and intervention outcome (self-efficacy). Fuzzy-set QCA necessity and truth table analyses determined exogenous theoretical constructs necessary and/or sufficient for high/low scores on mediating and outcome variables.

Results: Good model fit (CFI=.977; SRMR = 0.02) demonstrated and explained 38% of the variation in self-efficacy. Perceived barriers and benefits of SP, program, and year in program were observed with significant path coefficients to self-efficacy. QCA revealed high perceived severity of and susceptibility to disordered eating behaviors, and high perceived benefits of SP as necessary for high role beliefs. High role beliefs were deemed necessary but insufficient for low barriers to SP behaviors. Additionally, high perceived benefits and low barriers were sufficient for high self-efficacy among DH students (coverage = .57; consistency = .93).

Conclusion: Triangulation clarified relationships between key theoretical mechanisms. QCA provided unique insights regarding construct relationships that is critical for informing intervention adaptation.