Strategies to improve the implementation of intensive lifestyle interventions for obesity

COPYRIGHT © 2023 Finn, Whang, Hong, Costa, Callahan and Huang. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Strategies to improve the implementation of intensive lifestyle interventions for obesity


. Introduction
Obesity is one of the most consequential diseases in the United States given its prevalence, long-term consequences, and high costs. More than 40% of U.S. adults and more than one-fifth (21.5%) of children aged 2-19 years have obesity, with worsening trends of inequities by race, ethnicity, socio-economic status, and geography (e.g., rural vs. urban) (1)(2)(3). People with obesity are at higher risks of multiple comorbidities, including cardiovascular disease, Type 2 diabetes, fatty liver disease, some forms of cancer, and depression (4,5). The social consequences of obesity include discrimination, stigma and adverse impacts on education and employment (6, 7). The U.S. annual economic burden of obesity, including direct (health-related) and indirect (lost productivity) costs, has been estimated to be as high as $1.72 trillion (8).
First-line clinical management of obesity typically consists of multi-component, intensive lifestyle interventions (ILIs) that combine nutrition, physical activity, and behavior change support (9). Based on recommendations from authoritative bodies such as the U.S Community Preventive Services Task Force and the American Academy of Pediatrics, these interventions are often based on social cognitive models, including health education, goal setting, and social support (e.g., coaching) and can include motivational incentives such as rewards and behavioral feedback mechanisms (10, 11). Although several ILIs have demonstrated efficacy in clinical trials, their real-world impact, especially over the long-term, has been more limited; post-ILI weight regain is not uncommon. One systematic review demonstrated that some patients begin to experience weight regain around 36 weeks postintervention, and that many participants regain all lost weight by 48 weeks (12)(13)(14)(15)(16)(17). The persistent increase in rates of obesity suggests that we need to critically examine and directly address challenges in ILI design and delivery.
The goal of this opinion paper is to identify shortcomings in current ILIs, or "pain points, " experienced by patients-particularly those in underserved or minoritized communities-affected by obesity in the context of ILI implementation, that have not been adequately addressed to date. This paper is not meant to be an exhaustive discussion but reflects insights from the authors' research and reflections on the literature.The limitations are organized into three domains: (1) implementation context, (2) intervention components, targets, and sequencing, and (3) delivery strategies.

. Mismatch between intervention and community readiness
Community readiness is defined by a community's preparedness to undertake change, which is influenced by its perceived importance of the health issue in question as well as its capacities, resources, and political climate (18). Low community readiness diminishes an intervention's impact potential. The Community Readiness Model (CRM), developed to guide the creation of community-level behavior change interventions, has been applied to a wide range of public health topics, including alcohol abuse, cardiovascular disease, and children's socialemotional development (19)(20)(21)(22). A community's readiness varies by issue; for obesity, community readiness is often low or middling, in part because people may discount negative consquences of obesity that will accrue in the distant future (23-25). In this context, ILIs that target weight-related behaviors (e.g., diet or physical activity) may fail if they are not aligned with the community's highest priorities. Moving a community through stages of readiness can itself be an intervention outcome (18, 26). For example, baseline CRM findings were used to tailor a youth advocacy intervention to increase the community readiness for obesity prevention (27, 28).

. . Overlooking the built and social environment in a community
One premise of community readiness is that some communities may be unable to focus on obesity and its longer-term consequences until more immediate threats to safety and health (e.g., housing insecurity, crime, etc.) are addressed. Research has shown the importance of the role of built (e.g., distribution of food outlets) and social (e.g., perceived safety) environments in obesity. Few interventions, however, have incorporated environmental change as a critical component of ILIs (29-31). Dietz (32) cites "social conditions", including safety, participant trauma, and housing insecurity, as a reason that seemingly promising ILIs have failed to improve body mass index (BMI) in low-income communities. For example, programs that focus on physical activity but do not account for neighborhood safety, park accessibility, quality of sidewalks or other environmental factors are unlikely to succeed (33). Environmental change often necessitates local policy intervention and cross-sectoral partnerships, which should be incorporated to a greater extent in the next generation of ILIs (34). The use of community health workers (CHWs) or patient navigators may also help patients address social determinants of health by linking patients to clinical and social services (35).

. . Lack of attention to cultural nuance
Evidence-based ILIs are often assumed to be equally effective across different populations. In reality, systems, cultural phenomena, and lived experiences shape unique pathways to obesity, which warrants culturally responsive design for evidencebased interventions (36). For example, among Chinese-American residents in Manhattan's Chinatown, grandparents are key caretakers and strongly influence young children's dietary intake. Many Chinese grandparents believe that chubby babies are healthy babies, partly due to historical experiences of food insecurity; this perception can lead to over-feeding. It is also common for newly immigrated parents, who often work multiple jobs, to send their newborns back to China to be raised by grandparents until children reach school age. Given these cultural nuances, ILIs developed in predominantly white populations may not be as effective with newly immigrant Chinese American families if interventions fail to engage grandparents in the U.S. or China (37). The use of culturally and linguistically concordant CHWs may help to bridge this cultural divide (35). In addition, the need for ILI adaptations is increasingly recognized in implementation science; however, how to optimally effect these adaptations is the next frontier of research (38).
. Intervention components, targets, and sequencing . . Discounting di erent individual stages of change for di erent behaviors As with community readiness, many interventions fail to account for differences in participants' stages of change for various behaviors. The Transtheoretical Model, which maps individual behavior change progressing through six stages of change (SOC) (39), has been widely applied to obesity prevention and weight management efforts (40-43). However, individual SOC is behavior-specific: one may eat fruits and vegetables daily (action or maintenance stage) but not exercise (pre-contemplation or contemplation stage). As such, ILIs, which are designed to simultaneously address multiple aspects of obesity (e.g., physical activity, nutrition, sleep, etc.), must consider participant SOC relative to each behavior. Training for interventionists should include how to recognize participant SOC for each behavior and to tailor the intervention to each SOC.
. . Overlooking participants' mental health, cognitive load, and executive functioning Mental health influences individual engagement with and response to ILIs. Weight stigma, low self-esteem, and mental health disorders (e.g., depression, anxiety) can all impede motivation, which is critical to individual movement through the SOC (44-46). Chronic stress interferes with self-regulation by decreasing physical activity, impairing sleep, and facilitating unhealthy eating (47,48). Compared to other factors, poor mental health is a stronger predictor of attrition in obesity interventions, and mental health and weight loss have been strongly correlated, in some cases up to 12 months-and likely longer-after beginning ILIs (49,50).
Cognitive load refers to the amount of information the brain can simultaneously process. It is not only negatively associated with mental health, but also limits ILI reach and uptake as people face competing demands that can become barriers to participation in obesity interventions (51, 52). Cognitive overload can result in decreased cognitive functioning. As such, participants who are cognitively overloaded before an ILI is introduced may be less likely to process the intervention effectively (53, 54). Teaching strategies developed specifically to avoid contributing to cognitive overload can be applied during the intervention design phase (55).
ILIs will be most effective when they address mental health and minimize cognitive load; more research is needed on this topic. The few pilot studies that have assessed the extent to which supporting participant stress management affected weight loss had small samples, short durations, and low generalizability (56)(57)(58). Studies with more robust statistical power as well as repeated longer-term measures of stress and mental health, cognitive functioning, health behaviors, and body composition would inform more effective ILI development. Qualitative studies are also needed to better contextualize the life experiences of intervention participants, so that desired behaviors such as healthy eating and physical activity can be considered as part of other life demands and stresses rather than be treated in isolation.

. . Knowledge gaps around the optimal sequencing of intervention components
Despite significant research around prevention and management of obesity, the optimal ways to combine ILI components remain under-explored (59). Most weight interventions target only one or two of the many factors associated with obesity; diet and physical activity are the most common but the importance of other behaviors, such as sleep, is increasingly recognized (60). The complexity of obesity demands broad-based interventions, but more research is needed on how to optimize the combination and sequencing of multiple intervention components, taking into account that such optimization may vary across populations. Several methodologies are available to explore optimal sequencing. Multiphase optimization strategy trials enable rigorous exploration of sequencing intervention components for obesity (61)(62)(63)(64). The Fogg Behavior Model (FBM), which suggests that a behavior happens when motivation, ability, and prompt occur simultaneously, has been effectively applied to interventions for multiple health challenges. The FBM may enable obesity interventionists to more effectively support desired behavior changes, including by considering the specific sequencing of small behavioral steps before or after a current habit, leading to greater uptake of the target behavior (65)(66)(67)(68)(69). Adaptive interventions can be tailored to individual participant preferences and needs, and can be used to address early nonresponders by modifying intervention intensity or form (70)(71)(72). Lastly, the use of simulation models in systems science has been applied to support optimization of intervention sequencing across several public health areas and could be similarly applied to obesity (73-75).

. Delivery strategies
Interventionists must consider several aspects of ILI delivery. One is whether intervention materials are actively distributed ("pushed") or must be actively sought by the participant ("pulled"). While comparisons of "push" vs. "pull" modalities have been studied, further research is needed to better understand when and where to use each type (76). For instance, one study found that patients and clinicians had opposite preferences, with patients preferring "pulling" information on resources, and clinicians preferring "pushing" (77). However, pull and push strategies may also be used together, for example, using an opt-out approach for intervention enrollment to maximize reach while allowing patients to customize some of the content or its presentation in the intervention.
Another aspect is whether interventions are delivered virtually or in-person. The pivot of in-person ILIs to virtual delivery at the onset of the COVID-19 pandemic provided a natural experiment to compare these modes, and early evidence sugguests this virtual transition may work for some programs but perhaps not others if staff and/or participants lack the technological self-efficacy to engage digitally, or if staff do not have the bandwidth to gain the required technological know-how (78). Currently, the U.S. Community Preventive Services Task Force recommends digital health interventions to improve diet and physical activity in workplace, higher education and community settings (10). However, with the exception of adolescents for whom this recommendation is made specifically to reduce overweight or obesity, the other companion recommendations are made to improve lifestyle behaviors only.
The field of lifestyle medicine has put forth successful practices related to the delivery of ILIs for obesity in clinical settings, such as shared medical appointments (79). In addition, the American College of Lifestyle Medicine released a statement in April 2023 stating that lifestyle interventions to treat obesity are often inadequately dosed for success (as it relates to the six pillars of lifestyle medicine used to treat chronic conditions, including obesity), indicating that delivery strategies must carefully consider dose (80).
Last but not least, interventionists must consider new and creative ways to meet intended participants where they are, physically, emotionally, and socially. Manga comics, exergaming (i.e., video games with interactive physical activity), and weight management podcasts have been used to engage various groups with promising results (81)(82)(83). Similar creativity for in-person interventions could facilitate embedding interventions into participants' day-to-day life. For instance, nutrition components of ILIs could be offered through parent and child engagement during grocery trips. Exercise components for adults could rely on brief "exercise snacks" instead of longer bouts of physical activity (84). The current evidence base for such strategies is limited.

. Discussion
This opinion paper is focused on critical considerations for improving implementation and impact of obesity-related ILIs. The points raised have been understudied in the obesity field and warrant further research and development.
As the field moves toward the next generation of ILIs, further attention to the metrics used to gauge ILI success is important (85,86). For example, given intra-individual variations in SOC in the context of multi-component interventions, a participant could become more physically active without making dietary changes while another could experience the reverse-but both individuals may experience only minor changes in BMI, although the intervention has improved their health to some degree. The selection of intervention success metric(s) can be influenced by the target population, goal(s) of the intervention recipients, intervention type, and intervention duration, among other factors.
As obesity interventions scale, the expected magnitude of behavior or health changes needs to be considred in line with the reach of ILIs. For instance, many digital interventions (e.g., webor app-based) have had low or no statistically significant effect on participant weight despite high acceptability and feasibility (82,(87)(88)(89)(90). Considering near-universal cell phone ownership, these types of interventions may offer wider benefit on a population level than the literature acknowledges, especially when offered in addition to standard clinical care rather than a stand-alone intervention (82,87,(91)(92)(93).
The rise of implementation science (IS) has informed early efforts to improve the adoption, implementation, and sustainment of evidence-based interventions (94). IS offers a variety of frameworks and theories that can help design better interventions, improve outcomes or guide evaluations (95). Most IS frameworks are top-down and deterministic rather than bottom-up and human-centered. Although some recent advances have been made to address patient or community engagement in intervention implementation (96,97), "how to" solve for patient pain points is still a major inquiry. Note that patient pain points may be affected by other concomitant chronic diseases or competing life priorities, which may influence the motivation to engage in obesity interventions. Human-centered design (HCD) can help bridge this gap in establishing program-context fit. HCD is a process for innovation by which design of a product or system incorporates end-user needs, preferences, and usage to develop a product that solves the user's "pain point" (98). HCD has slowly entered public health practice as a way to solve implementation challenges identified by IS frameworks. For example, Haines et al. (99) employed a three-step process of "usability testing, " "ethnographic contextual inquiry, " and "iterative prototyping" to ensure that their intervention, including its delivery methods, was best designed to fit the implementation context. Though more research is needed, HCD may allow for more innovative, strategic, and contextually tailored intervention designs, which may increase participant adoption and maintenance of health behaviors.
For obesity interventions to be successful in producing longterm changes in greater numbers of people, more effective, customized, and engaging interventions are needed. We hope this opinion will help move the field toward this end.

Author contributions
TH conceived of this work, which was drafted by EF, PH, and CW, and substantially revised by TH, PH, CW, SC, and EC. All authors read and approved the submitted version of the manuscript.

Funding
The paper has been supported, in part, by funding from the National Institutes of Health (R01MD018209) and the Centers for Disease Control and Prevention (U48DP006396).

Conflict of interest
EC was employed by EAC Health and Nutrition. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.