Incorporating automated digital interventions into coach-delivered weight loss treatment: A meta-analysis.

Abstract

Automated digital interventions (ADIs) represent a potentially promising approach to enhance the outcomes of human coach-delivered weight loss interventions. However, the extent to which ADIs provide additive benefit is unclear. This study represents the first systematic review and meta-analysis of the effectiveness of ADIs for improving the outcomes of human coach-delivered weight loss treatment. Electronic database searches were used to identify trials that compared differences in weight change between (a) weight loss interventions that were delivered exclusively by coaches and (b) interventions supplementing this same human coaching with an ADI. Subgroup and moderator analyses examined the influence of intervention duration, duration of human coach contact, presence of tailored coaching, modality of the ADI and demographic variables on ADI effectiveness outcomes. Thirteen studies met inclusion criteria (1,471 participants). Random-effects meta-analysis revealed a mean difference in weight change between conditions of 2.18 kg at postintervention, representing a medium effect size of .54 (95% CI [.13, .95]). Subgroup analyses suggested that lower duration of coach contact was associated with improved additive effectiveness of ADIs. No other subgroup differences were found. Publication bias appeared to be a potential concern, though high levels of heterogeneity and a small number of included studies likely limited the ability to infer its presence. Results support the use of ADIs to augment coach-delivered behavioral weight loss treatment, and also suggest that ADIs have the greatest impact when coaching is relatively low in frequency or duration. (PsycInfo Database Record (c) 2021 APA, all rights reserved).