It is almost 40 years since Borkovec et al. (1983) provided the definition of worry that has guided theory, research and treatment of Generalized Anxiety Disorder (GAD). This review first considers the relative paucity of research but the proliferation of models. It then considers nine models from 1994 to 2021 with the aim of understanding why so many models have been developed.

By extracting and coding the components of the models, it is possible to identify similarities and differences between them. While there are a number of unique features, the results indicate a high degree of similarity or overlap between models. The question of why we have so many models is considered in relation to the nature of GAD. Next, the treatment outcome literature is considered based on recent meta-analyses. This leads to the conclusion that while efficacy is established, the outcomes for the field as a whole leave room for improvement. While there may be scope to improve outcomes with existing treatments, it is argued that rather than continue in the same direction, an alternative is to simplify models and so simplify treatments.

Several approaches are considered that could lead to simplification of models resulting in simpler or single-strand treatments targeting specific processes. A requirement for these approaches is the development of brief assessments of key processes from different models. Finally, it is suggested that better outcomes at the group level may eventually be achieved by narrower treatments that target specific processes relevant to the individual.