Evaluating the Robustness of Targeted Maximum Likelihood Estimators via Realistic Simulations in Nutrition Intervention Trials

Several recently developed methods have the potential to harness machine learning in the pursuit of target quantities inspired by causal inference, including inverse weighting, doubly robust estimating equations and substitution estimators like targeted maximum likelihood estimation. There are even more recent augmentations of these procedures that can increase robustness, by adding a layer of cross-validation (cross-validated targeted maximum likelihood estimation and double machine learning, as applied to substitution and estimating equation approaches, respectively). While these methods have been evaluated individually on simulated and experimental data sets, a comprehensive analysis of their performance across ``real-world'' simulations have yet to be conducted. In this work, we benchmark multiple widely used methods for estimation of the average treatment effect using ten different nutrition intervention studies data. A realistic set of simulations, based on a novel method, highly adaptive lasso, for estimating the data-generating distribution that guarantees a certain level of complexity (undersmoothing) is used to better mimic the complexity of the true data-generating distribution. We have applied this novel method for estimating the data-generating distribution by individual study and to subsequently use these fits to simulate data and estimate treatment effects parameters as well as their standard errors and resulting confidence intervals. Based on the analytic results, a general recommendation is put forth for use of the cross-validated variants of both substitution and estimating equation estimators. We conclude that the additional layer of cross-validation helps in avoiding unintentional over-fitting of nuisance parameter functionals and leads to more robust inferences.