A Conceptual Model of Long-Term Weight Loss Maintenance: The Importance of Cognitive, Empirical and Computational Approaches

Abstract

Living with obesity is related to numerous negative health outcomes, including various cancers, type II diabetes, and cardiovascular disease. Although much is known about the factors associated with obesity, and a range of weight loss interventions have been established, changing health-related behaviours to positively affect obesity outcomes has proven difficult. In this paper, we first draw together major factors that have emerged within the literature on weight loss to describe a new conceptual framework of long-term weight loss maintenance. Key to this framework is the suggestion that increased positive social support influences a reduction in psychosocial stress, and that this has the effect of promoting better executive functioning which in turn facilitates the development of healthy habits and the breaking of unhealthy habits, leading to improved ongoing maintenance of weight loss. We then outline how the use of computational approaches are an essential next step, to more rigorously test conceptual frameworks, such as the one we propose, and the benefits that a mixture of conceptual, empirical and computational approaches offer to the field of health psychology.

One p-Factor for All? Exploring the Applicability of Structural Models of Psychopathology within Subgroups of a Population

“…Future research should look to use Monte Carlo simulations to attempt to validate and extend these findings and also should explore various different grouping methods to accommodate the nesting limitation mentioned above. However, even though simulation and computational methodologies are useful for providing a way to discretely and precisely examine a research question [ 30 ], research using large samples of human data is needed to validate these findings. Future research may also measure a larger range of correlates and assess the applicability of the higher-order factor model and the traditional bifactor model at the subgroup level.…”

Combining Persuasive System Design Principles and Behavior Change Techniques in Digital Interventions Supporting Long-term Weight Loss Maintenance: Design and Development of eCHANGE

“…Development of healthy habits and adjustment or breaking of less healthy habits are required to optimize and maintain the health benefits of weight loss, as well as maintain new weight, over time [ 103 , 104 ]. Therefore, self-regulation strategies and skills are essential to enhance when new physically active lifestyles or healthy eating patterns are not fully automated and are likely vital to maintain healthy behaviors and not relapse into previous habits [ 26 , 105 ].…”

“…Therefore, self-regulation strategies and skills are essential to enhance when new physically active lifestyles or healthy eating patterns are not fully automated and are likely vital to maintain healthy behaviors and not relapse into previous habits [ 26 , 105 ]. The findings from this study show that targeting holistic aspects (eg, cognitive, emotional, social, and behavioral) of behavior change [ 103 ] during the design of digital technologies may be essential when aiming to deal with the often multifaceted challenges and needs of weight loss maintenance [ 23 , 32 , 98 ]. Informed by the self-determination and self-regulation theories [ 91 , 92 , 95 ], some of the features identified in this study focus on the generation of habitual behaviors through recognized self-regulation and habit formation techniques.…”

Neurocognitive Artificial Neural Network Models Are Superior to Linear Models at Accounting for Dimensional Psychopathology

“…Artificial neural networks (ANN) have been successful used across levels, from psychological to genetic to help understand psychopathological and behavioural phenomena [ 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 ] and potentially provides an even richer methodology for studying the relations between neurocognitive abilities and psychopathology. For example, in a standard 3-layer feed-forward network, for which a problem may be specified and for which the desired outcome is known, input units are provided with a representation of the problem and the output of each unit is fed forward to all units it is connected to within a hidden layer (comprised of a number of processing units).…”