Simulation model to assess the validity of the clinical portfolio diet score used in the PortfolioDiet.app for dietary self-tracking: a secondary analysis of a randomized controlled trial in hyperlipidemic adults

Abstract

IntroductionThe Portfolio Diet combines cholesterol-lowering plant foods for the management of cardiovascular disease risk. However, the translation of this dietary approach into clinical practice necessitates a user-friendly method for patients to autonomously monitor their adherence.ObjectiveThis study aimed to develop and validate the clinical-Portfolio Diet Score (c-PDS) as a food-based metric to facilitate self-tracking of the Portfolio Diet.MethodsUsing a simulation model to estimate the c-PDS, the validity was assessed in a secondary analysis of a completed trial of the Portfolio Diet in 98 participants with hyperlipidemia over 6 months. Concurrent and predictive validity of the estimated c-PDS were assessed against the reference measure (weighed 7-day diet records) and concomitant changes in LDL-C from baseline to 6 months. Bland–Altman analysis was used to assess the limits of agreement between the two methods.ResultsThe c-PDS was positively correlated with dietary adherence as measured using the 7-day diet records (r = 0.94, p < 0.001). The c-PDS was negatively correlated with change in LDL-C (r = −0.43, p < 0.001) with a 1-point increase in the c-PDS being associated with a − 0.04 mmol/L (CI:−0.06,−0.03; p < 0.001) or a 1.09% reduction in LDL-C. Visual evaluation of the Bland–Altman plots showed reasonable agreement.ConclusionThese findings indicate good validity of the c-PDS for primary prevention in adults with hyperlipidemia. The predictive validity findings have informed the goals and messaging within the PortfolioDiet.app, a digital health application for delivering the Portfolio Diet. Future research will assess the effectiveness of the intended combination of the c-PDS and the PortfolioDiet.app in supporting behavior change.