Many lifestyle intervention trials depend on collecting self-reported outcomes, such as dietary intake, to assess the intervention’s effectiveness. Self-reported measures are subject to measurement error, which may impact treatment effect estimation. Methods exist to correct for measurement error using external validation studies, which measure both the self-reported outcome and accompanying biomarker, to model the measurement error structure. However, there is growing concern over the performance of these methods when the validation study differs greatly from the intervention study on pre-treatment covariates that relate to treatment effect. In this paper, we evaluate the impact of such covariate imbalance on measurement error correction methods through simulation, and propose an improvement upon the methods by weighting the validation study using propensity score-type techniques, followed by the implementation of the measurement error correction. We apply the methods to the PREMIER study, a multi-arm lifestyle intervention trial aimed at reducing self-reported sodium intake, and OPEN, a validation study measuring both self-reported diet and urinary biomarkers.