Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) and suicidality: what do we know and future vistas

KEYWORDS:

• Glucagon-like peptide-1 receptor agonists (GLP-1 RAs)
• liraglutide
• semaglutide
• tirzepatide
• weight loss
• obesity
• overweight
• diabetes
• depression
• suicide

1. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs): suicidality reported to regulatory agencies and current indications

Within the last year, the European Medicines Agency (EMA), Food and Drug Administration (FDA) and the Medicines and Healthcare Products Regulatory Agency in the United Kingdom (MHRA) began investigations into whether glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are causally related to suicidality [1,2]. The impetus for investigation was provided by reports of self-injury and suicidal thoughts associated with exposure to GLP-1 RAs.

The possibility that bariatric treatment may be associated with suicidality and/or worsening of psychopathology has precedent. For example, bariatric surgery has been associated with, in some circumstances, emergence of psychopathology and suicidality [3]. In addition, Rimonabant, a cannabinoid receptor-1 antagonist/inverse agonist, was approved for obesity but subsequently discontinued from the European market due to reports of worsening depression, anxiety, and suicidality [4–6]. Notwithstanding, other approved bariatric treatments such as naltrexone-bupropion has not been shown to worsen psychopathology or suicidality [7].

Glucagon-like Peptide-1 Receptor Agonists are highly efficacious in the treatment of Type 2 Diabetes Mellitus (T2DM) and as part of integrated weight management in persons who are obese (body mass index; BMI = 30 kg/m2) or overweight (BMI = >27 kg/m2) with an associated medical comorbidity [8,9]. In addition, preliminary evidence suggests that GLP-1 RAs may be of potential therapeutic value in other medical and psychiatric disorders (e.g. depressive disorders, alcohol use disorders, major neurocognitive disorders, Parkinson’s Disease) [10–18]. The aforementioned scientific and clinical developments along with extraordinary media and public interest in GLP-1 RAs has contributed to a significant increase in prescription of these agents in the general population globally.

1.1. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs): review of pharmacovigilance and pharmacoepidemiology studies

An analysis by our group of the FDA Adverse Event Reporting System (FAERS) indicated that there was disproportionate reporting of suicidal ideation and depression with GLP-1 RAs, semaglutide and liraglutide [19]. Notwithstanding the aforementioned observation, we did not observe disproportionate reporting of any parameter of suicidality compared to insulin and metformin for any FDA-approved GLP-1 RAs [19]. We concluded that taking into consideration confounders, no causal link between GLP-1 RAs and suicidality exists.

We also noted that no causal link could be discerned due to the limitations of pharmacovigilance reporting in public databases. For example, the FAERS does not verify the validity of case reports in any systematic method and there is also the possibility of duplicate reporting. In addition, the hazard of suicidality would need to be comprehensively evaluated in a rigorous, adequately controlled comparator experimental study using validated measures of suicidality before any strong statements of causality could be made. Furthermore, aspects critical to cause-and-effect determination as listed below in the Bradford Hill criteria are largely unavailable in the FAERS database. Finally, the extraordinary attention given to this topic, not only from the regulatory and scientific community, but more notably from the media and global business community has the potential of increasing reporting of events that may not be causally related to a particular medication.

Our conclusion was chronologically followed by a statement by the FDA ‘Preliminary evaluation does not suggest a causal link’ between GLP-1 RA exposure and suicidality [20]. The position by the FDA that no causal link between GLP-1 RAs and suicidality exists comports with results from two pharmacoepidemiological studies from Europe and the United States that did not identify a causal link with increased depression and/or suicidality with these agents. For example, an analysis of data from National Danish registries (n = 116,699) of patients with T2DM compared to a reference group of individuals without diabetes (n  = 116,008) concluded that exposure to glucose-lowering agents, including GLP-1 RAs, was associated with a lower risk of depression in persons with diabetes when compared to non-users [21].

Separately, a retrospective cohort study using electronic health records from a commercial database evaluated incidence and recurrent suicidal ideation during six months of follow-up with GLP-1 RAs. The cohort included persons who were overweight or obese (n = 240,618) and persons with T2DM (n = 1,589,855) [22]. It was concluded that persons prescribed semaglutide for either overweight/obesity [(0.11% versus 0.43%; hazard ratio (HR) = 0.27, 95% confidence interval (CI) = 0.20–0.36) or T2DM (0.13% versus 0.36%; HR = 0.36, 95% CI = 0.25–0.53)] had a significantly lower rate of incident and recurrent suicidal ideation when compared to persons not receiving semaglutide. The authors also reported that the reduction in risk was greater in the overweight/obesity group than the group with T2DM, suggesting a dose-response relationship between exposure and risk reduction, as dosing in overweight/obesity are relatively higher.

2. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs): potential therapeutic benefits in psychiatric populations

During the past decade, GLP-1 RAs have been evaluated as potential treatments for psychotropic-drug related weight gain (PDWG) in persons being treated for serious mental illness [23]. It could be hypothesized that persons with serious mental illness would be at greater risk of experiencing suicidality when compared to persons in the general population who do not have a mental illness. The aforementioned studies evaluating GLP-1 RAs as antidotes for PDWG have reported benefits on anthropometrics and metabolics but have not observed an emergence or worsening of psychopathology or suicidality in what would be considered an at-risk population [24,25].

Further lines of evidence suggesting GLP-1 RAs are not causally linked to suicidality is provided by studies that are evaluating these agents to treat psychopathology. For example, a small pilot study from our group revealed that liraglutide may benefit measures of general cognitive functions in persons with mood disorders and not worsen suicidality [26]. Finally, analysis of safety data related to liraglutide extracted from phase 2 and 3 trials also did not identify risk of suicidality [27].

3. Association versus causation: the Bradford Hill criteria

A framework for establishing causal links between drug exposure and a serious adverse event or safety concern is the Bradford-Hill Criteria [28]. The Bradford-Hill Criteria is comprised of nine variables: (1) strength, (2) consistency, (3) specificity, (4) temporality, (5) biological gradient, (6) plausibility, (7) coherence, (8) experimental, and (9) analogy. Some of the aforementioned Bradford-Hill criteria may apply to GLP-1 RAs and suicidality.

For example, plausibility is suggested insofar as it is known that GLP-1 RAs attenuate food, alcohol, and substance craving by modulating hedonic tone [29]. It is also observed that anhedonia, especially attenuated anticipatory hedonism, is highly associated with suicidality [30]. It could be hypothesized that susceptible people receiving GLP-1 RAs may, in some circumstances, experience a worsening of hedonic tone, predisposing psychopathology and/or self harm ideation and/or other aspects of suicidality. In contrast, however, it could be also hypothesized that GLP-1 RAs, rather than reducing hedonic tone, modulates maladaptive anticipatory hedonic tone yet favorably influence consummatory hedonistic processing [30].

Notwithstanding this possibility, there is also data to suggest that GLP-1 RAs improves hedonic tone as well as cognitive processes which are known to subserve impulsivity [12,31]. In addition, compelling evidence for cause and effect does not exist for the remaining Bradford Hill criteria. It remains a testable hypothesis that the risk of suicidality in persons prescribed GLP-1 RAs could be comprehensively and parsimoniously explained by an underlying mental disorder known to be overrepresented in persons with obesity and/or T2DM [32–36]. The association between obesity and suicidality is partially mediated by underlying mental illness but may include aspects of reward processing disturbance and/or impulsivity in the obese population independent of an underlying mental disorder [31].

4. Glucagon-like peptide-1 receptor agonists (GLP-1 RAs): future clinical and research vistas

When integrating extant datasets and available lines of evidence with respect to GLP-1 RAs and suicidality, no cause and effect can be established at this time. Persons prescribed GLP-1 RAs for on-label indication are differentially affected by underlying mental illness which provides potential explanation for observed reports of suicidality. Extant evidence provides the basis to hypothesize that select GLP-1 RAs may instead have beneficial effects on some domains of psychopathology, and separately suicidality. It is estimated that more than 800,000 people each year die by suicide globally and at least 10–20 times more persons attempt suicide [37]. More than 80% of persons who attempt or die by suicide have a diagnosable mental disorder, with mood disorders the most common [38]. It is a critical public health priority to identify all factors, including iatrogenic, that may be associated (i.e., positively or negatively) with suicidality.

In the interim, pharmacovigilance as it relates to this matter needs to continue and reported cases require verification and comprehensive characterization. Practitioners providing care to persons prescribed with GLP-1 RAs should be aware of the higher rate of mental illness and suicidality in persons with overweight/obesity and T2DM, which should be equally prioritized as part of comprehensive integrated care [39]. All persons prescribed GLP-1 RAs, presumably for weight and/or metabolic management, should be screened for commonly encountered mental disorders in clinical practice, more likely to affect this population [e.g 9-item Patient Health Questionnaire (PHQ-9), Rapid Mood Screener (RMS)] [40,41].

Until further evidence is available on GLP-1 RAs and suicidality, clinicians should be aware of the association between GLP-1 RAs and suicidality, but should not proscribe the use of these agents on the basis of concluding cause and effect as that has not hitherto been established. Finally, the introduction of tirzepatide, a GLP-1 receptor and glucose-dependent insulinotropic polypeptide (GIP) co-agonist, has an overlapping but different mechanism with extant GLP-1 RAs. Preliminary evidence suggests that this agent may also have potential brain therapeutic properties [42]. Research efforts that endeavor to ascertain potential worsening of psychopathology and/or suicidality with GLP-1 RAs should consider the agents separately, especially with the introduction of mechanistically overlapping but not identical GLP-1 RAs.

Declaration of interests

RS McIntyre has received research grant support from CIHR/GACD/National Natural Science Foundation of China (NSFC) and the Milken Institute; speaker/consultation fees from Lundbeck, Janssen, Alkermes, Neumora Therapeutics, Boehringer Ingelheim, Sage, Biogen, Mitsubishi Tanabe, Purdue, Pfizer, Otsuka, Takeda, Neurocrine, Neurawell, Sunovion, Bausch Health, Axsome, Novo Nordisk, Kris, Sanofi, Eisai, Intra-Cellular, NewBridge Pharmaceuticals, Viatris, Abbvie and Atai Life Sciences. RS McIntyre is a CEO of Braxia Scientific Corp. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Additional information

Funding

This paper was not funded.