Capitalizing on the GLP-1 RA revolution for behavioral health through digital companions

Introduction

Psychiatric disorders represent a major healthcare burden with estimates from the world health organization that over 900 million persons are afflicted globally, ¹ contributing to 400+ million disability-adjusted life years and an economic impact exceeding $5 trillion. ² Of myriad causal factors that have been studied, there is increasing evidence that the gut–brain axis may play a pivotal role in regulating cognitive and emotional function through fluctuations in gut microbiota levels. Imbalance in levels of these diverse microbes (dysbiosis) has been linked to susceptibility to psychiatric illnesses and mental health disorders as they impact neurotransmitter production (serotonin, GABA, histamine, acetylcholine, dopamine and kynurenine) and pathways regulated by the vagus nerve. ³ One of the strategies being investigated to restore balance is the use of probiotics, which show promise in the treatment of dysbiosis-induced anxiety and depression.^4,5 Another approach, recognizing the fact that the gut represents the largest endocrine gland in humans, has been to target pathways involved in cross talk between intestinal microbiota and the immune system, specifically hormones which act as signaling molecules. Among these, glucagon-like peptide 1 (GLP-1) has been identified as playing a key role in preserving integrity of the intestinal barrier, metabolic homeostasis, and inflammatory processes. ⁶ Moreover, clinical development of agonists of this hormone receptor (GLP-1 RA's) for treatment of obesity and type 2 diabetes have revealed key alterations in microbiota specifically the Firmicutes to Bacteroidetes ratio and the decreased abundance of Akkermansia muciniphila (hallmarks of dysbiosis associated with type 2 diabetes metabolic syndrome) can also be reversed. ⁷ Several of these drugs have now been approved and are having a dramatic impact on weight management at population-level scale globally. ⁸ The widespread use of these agents has allowed the clinical community to also identify ancillary benefits for patients and alongside certain cardiovascular and sleep-related conditions, improvement in depression and anxiety rates have been reported. ⁹ In the latter case, the role of the GLP-1 receptor signaling in the brain to promote neurotransmitter release, improve energy homeostasis, and mitochondrial function has been suggested, as has the role of GLP-1 agonists modulating inflammatory processes in the brain. Significantly, research also suggests that GLP-1 agonists directly modulate microbiota levels, including increasing abundance of Akkermansia and this may play a role in observed improvements in depressive symptoms. ¹⁰ Although preliminary, these observations could have marked impact on clinical care if substantiated through appropriately designed large-scale clinical trials.

Discussion

The widespread use of GLP-1 agonists for therapeutic management of cardiometabolic disorders offers the potential for patients to derive additional benefit from reduction in frequency and severity of comorbid depressive episodes and anxiety. ¹¹ There are also known bidirectional relationships between metabolic disorders and psychiatric disease and improvement in control of the former might result in improvement in the latter independent of the mechanism although additional research is needed.^12,13 The use of GLP-1 agonists presents a conundrum however, in that on-label prescription guidelines for providers and payers are typically tied to rigid criteria, for example, in some circumstances, patient BMI in excess of >30 kg/m² and or A1c levels above 7.5%. ¹⁴ Many healthcare systems, confronted with the high cost of these brand-name medications have severely limited coverage. ¹⁵ Additionally, in response to supply chain shortages, a number of compounded versions of the drugs have begun to enter the markets and concern has been raised relating to their potential for off-label administration. ¹⁶ Through their use, as and when patients enter into and maintain target weight range (recent estimates suggest that the GLP-1 RA tirzepatide / Zepbound^® can deliver weight loss of up to 25% within 18 months), ¹⁷ it is assumed that the rationale for prescription (and coverage by insurers) will be revisited and potentially removed, particularly if a plateau is reached. Unrelated, some studies report that patient discontinuation rates exceed 35% within 12 months due to a variety of reasons ¹⁸ potentially including psychiatric comorbidities. ¹³ Under either scenario, the potential exists for health benefit to be reversed, both for primary symptom (weight) and secondary symptom (depression). In the case of comorbid depression, this presents the prospect for patients to enter decline, particularly if therapy is stopped abruptly, and could lead to spiraling effects.

Patient education is key

Whether through financial concerns (medical insurance coverage or patient copay requirements), drug side effects, or apathy on having attained target health goals, the prospect of patients ceasing prescribed GLP-1 RA's abruptly presents a major concern. It has been demonstrated clinically that weight regain and other detrimental cardiometabolic effects occur, ¹⁹ but the impact on mental health under these circumstances is less studied. It should also be noted that many of these drugs are prescribed as step-therapies with staged dose-escalations over periods of many months. For example, in the case of Zepbound^®, the increments from 2.5 mg/0.5 ml through 5, 7.5, 10.0, 12.5 and ultimately 15 mg/0.5 ml with intervals of at least one month. ²⁰ One option could be for prescribers to introduce a ramp-down process to lower dose prior to cessation, but this would of course require active engagement and careful planning with the patient. Another, potentially more enduring approach would be to introduce lifestyle changes with the patient while on maintenance doses, and of these physical activity and especially dietary modifications would seem warranted. In the case of comorbid psychiatric diseases, the known link between microbiota and depression and anxiety presents an excellent venue for dietary optimization, with a particular focus on probiotics. Such counseling, in tandem with a registered dietitian, and possibly informed by 16s rRNA assays on the patient, ²¹ could identify sustainable, nondrug solutions to achieve positive benefits both in terms of cardiometabolic and psychiatric symptoms. For example, several studies have highlighted the benefits of high berry content diets and probiotics from yoghurts to produce striking improvement in colony composition and could presumably be incorporated easily into patients lifestyle.^22,23 Such interventions, however, at scale would require high levels of patient motivation and engagement and it will be important to consider behavioral models. One pertinent example is the Fogg Behavioral Model which posits that in order for a desired behavior to be actioned, the three elements of Motivation, Ability, and a Prompt (often referred to as the B-MAP doctrine) must be simultaneously present. Incorporation of these tenets was deployed successfully to bolster vaccination rates during the COVID-19 pandemic, ²⁴ and patient friendly approaches where prompts are introduced using mobile health apps are now commonplace. ²⁵ Such an approach driven through smartphone apps could focus on providing daily prompts on diet and exercise recommendations in response to physical activity and sleep (derived from ballistographic and activity sensors in the device) and supply motivational messaging, educational content and live support through a chatGPT-trained digital companion (Figure 1). Integrating such a system within the patient’s own healthcare ecosystem could allow key vitals to be accessed in order for algorithms to drive desired behaviors, for example, blood pressure, A1c, BMI, and medication dosing schedules. Clearly, the development of such a digital companion would require careful consideration and appropriate provisions on data security and medical ethics but the benefits could be considerable and address a current unmet need in support of patients with psychiatric illness.

Figure 1.

How a digital companion might help mitigate discontinuation/rebound effects in patients transitioning from glucagon-like peptide-1 (GLP-1) RAs.

Existing approaches

A number of existing apps show promise addressing some of the motivational and behavioral goals envisioned and could be ideal candidates for further exploration in clinical trials. One such example is cal.ai, which allows users to assess dietary plans by scanning foodstuffs for caloric and nutrient information. ²⁶ The powerful AI-based recognition systems identify meal components from images and also estimate portion sizes to aid in decision-making. It also analyzes dietary habits of the user and provides recommendations and motivational prompts to help users track and achieve goals. These elements could form the basis of a powerful tool and might also allow integration with smart menus in restaurants and QR-based coding systems in grocery stores. ²⁷ In the management of depressive episodes, the Finch app has received considerable interest. ²⁸ Allowing users to control symptoms through gamification and motivational prompts-based around animal companion avatars, the app encourages users to manage large tasks through simplification and focus on positive habits. Elements include mood tracking and reflection to help users understand emotional states and develop coping mechanisms, and mindfulness, breathing exercises and soundscape elements are all introduced in order to reinforce a positive environment and support framework. Such components would appear to be well suited to integration with a motivational support platform tied to dietary tracking and decisioning potentially leading to lasting benefit (Figure 1). An important consideration in the design of and adherence to any digital/behavioral companion is to understand the behavioral profile of each prospective patient. This could include understanding individual motivations for change (weight and cardiovascular management, lifestyle factors) and their appreciation of strategies required for lasting change. This is particularly relevant to help avoid rebound and fallback effects and could form the basis of discussions with their (GLP-1 RA) prescribing physician or counsellor.

The path to implementation

As the GLP-1 RA market grows dramatically it will be imperative that prescribing physicians and healthcare providers begin to systematically track the behavioral health of prescribed patients. Beyond routine inquiry, emphasis should be placed encouraging aggressive reporting of off-label events related to behavioral health, however minor in nature. This will allow the community to develop statistically relevant data from which to design appropriate trials and interventions (e.g., off boarding from GLP-1 RA in declining dosage ramps). Additionally, there is opportunity for behavioral health support and focus groups to encourage reporting and systematic observations relating to behavior modification for persons currently or previously prescribed GLP-1 RA's. These elements may then provide the basis for technology developers to design appropriate apps and digital tools to achieve lasting benefit. Specific near term actions could be:

Raising awareness of the links between GLP-1 RA pathways and behavioral health in scientific forums (journals such as this, conference sessions).

Establishing working groups between GLP-1 RA drug developers and professional societies engaged with behavioral health (e.g., APA, AMHCA, NAMI, and ABCT).

Development of modified prescribing guidelines for GLP-1 RA's which expand inclusion of behavioral health criteria.

Increasing interaction between the medical community and app / technology developers to obtain patient voice and experiences relative to behavioral health associated with GLP-1 RA and microbiota-influenced pathways.

Developing and introducing such a digital support system for widespread use would naturally require careful consideration of numerous factors including the heterogeneity in patient access to GLP-1 RA medications and behavioral therapy options, digital literacy levels, and availability of smartphone technologies. Additionally, given the sensitivity of the health information involved, appropriate guardrails (privacy, governance, safety monitoring) would need to be established, ideally with guidance from expert working groups in the health technology community. We hope this commentary serves to help stimulate additional thought and ultimately progress in pursuit of these actions.

Conclusion

The widespread use of GLP-1 receptor agonists for cardiometabolic diseases is now highlighting the role of the gut–brain axis in psychiatric diseases such as depression and anxiety. While the precise benefits of GLP-1 agonists for improving symptoms of these conditions off-label and as comorbidities continues to evolve, the potential for patients to cease administration of these agents either voluntarily or through healthcare loss could result in adverse outcomes in terms of mental health. We advocate that counseling and digital support tools can play a role in coaching patients through these transitions and underscore the positive benefits of dietary modification with focus on the microbiome. These tools could be deployed by providers as patients ramp down from high-dose GLP-1 agonists and could be augmented by key medical data including 16s rRNA screening. For sustained benefit, this will require app developers and healthcare providers to join forces to embrace modeling of patient behaviors and focus on lifestyle optimization through nutrition and physical activity. These sentiments have been visited many times, however, the widespread popular media coverage of the so-called GLP-1 “revolution” and increased efforts to destigmatize mental health present ideal conditions for change to be enacted. We encourage this dialog to develop and evolve for the benefit of the many millions of patients who suffer from these illnesses.