Suicide is a serious mental health problem with the number of global suicide mortality of over 800,000 per year and suicide attempts of around 16 million per year (1). Mental disorders are the most frequent diagnoses of suicidal behavior, notably depression (2). Currently, treatment options to improve aspects of suicidality are limited. Conventional interventions including antidepressants, cognitive behavioral therapy, and electroconvulsive therapy, are reported to reduce suicidal ideation and behavior in persons with depression, but this effect takes weeks, leaving a considerable proportion of patients suffering from suicide risk for a long time (3, 4). Moreover, only lithium has demonstrated the ability to lower suicide completion in persons with mood disorders (5).

Fortunately, in the past two decades, numerous RCTs have confirmed ketamine, a glutamate N-methyl-D-aspartate (NMDA) receptor antagonist, has rapid-onset antidepressant effects (6–10). Replicated evidence also indicates that ketamine can rapidly reduce aspects of suicidality (11–17). Suicidal ideation reductions typically manifest within a few hours, and some patients respond as soon as 40 min after the initial infusion (11–17). Currently, nasal esketamine, the S-enantiomer of ketamine has been approved by Food and Drug Administration for the treatment of depressive symptoms in adults with major depressive disorder (MDD) with acute suicidal ideation or behavior (18–20).

Most clinical studies showed a significant decrease in suicidal ideation and overall depressive symptom severity simultaneously following ketamine treatment, suggesting the ketamine’s antisuicide effects are, to a certain extent, mediated by the improvements in overall depressive symptom severity (11, 14). However, Ionescu et al. in an open-label ketamine trial observed a significant decrease in suicidal ideation following six ketamine infusions. The anti-suicide effects approached significance after controlling for total depressive symptom severity, indicating the ketamine’s anti-suicide effects were independent of the antidepressant effects (13). There is a need to determine whether ketamine’s anti-suicide effects are dissociable from its antidepressive effects.

In addition, cognitive dysfunction is recognized a core symptom in depression (21), and the potential pro-cognitive effects in the domains of working memory, visual learning memory, and processing speed with ketamine in treatment-resistant depression (TRD) is suggested by the recent reports (22–25). Our previous study assessed the cognitive effect of six infusions of ketamine in adults with TRD and/or suicidal ideation, and simple improvements in processing speed and verbal learning were observed 1 day following six infusions of ketamine (23). Currently, studies have suggested a close association between vulnerability to suicidal behavior and cognitive alterations (26–28). Deficits in executive function, working memory, decision-making, and impulsivity were associated with current or/and histories of suicidal ideation/behavior in patients with MDD (26–28). Altered value-based and cognitive control processes may be important factors of suicidal vulnerability (29). Available evidence suggests underlying the suicidal processing and help to identify the target of therapeutic interventions aimed at reducing the long-term risk of suicidal acts.

Relatively few studies have evaluated the mediational role of cognition in suicidality in persons with mood disorders. We observed previously that improvements in working memory were associated with anti-suicide outcomes after repeated ketamine infusions in depressed patients with baseline suicidal ideation (30). The foregoing observation supports the previous hypothesis that the anti-suicide effects of ketamine are mediated, in part via pro-cognitive effects (30).

However, our previous study takes no account of the effect of improvements in depressive symptoms on the anti-suicide effects following ketamine treatment and lacks the result of other cognitive domains except working memory. Herein, we sought to determine whether improvements in depressive symptoms and/or cognition mediated change in suicidal ideation in adult patients received six sub-anesthetic doses of ketamine, and examine predictors of ketamine’s anti-suicide effects.

Herein, we observed six infusions of ketamine in unipolar or bipolar depressive disorders with suicidal ideation were associated with improved mood, suicidal ideation, and cognitive function (i.e., processing speed, verbal learning). Mediational analysis indicated that the anti-suicide effects were partially mediated by improvements in depressive symptom and processing speed. The mediational effect of working memory, verbal learning, or visual learning on the change in SSI was not statistically significant. Herein, we speculate that for some individuals, ketamine’s anti-suicide effects may be subserved by improvements in processing speed. Baseline processing speed was a significant predictor of anti-suicide effect, indicating greater baseline processing speed performance was associated with an increase likelihood of achieving anti-suicide effect over six ketamine infusions.

Cognitive function in mood disorder can be broadly categorized as cold cognition and hot cognition. The former refers to information processing in the absence of any emotional influence, yet the latter is influenced by emotion (35). Usually, cold cognition is considered to include memory, attention, executive function, working memory while hot cognition includes catastrophic reactions to real and/or perceived slights, anhedonia, suicidal ideation, negativistic rumination, negative recall bias, and disproportionate attention to negative stimuli (36). Over the past years, an increasing number of literatures have established that cognitive deficits are closely related to suicidality (26–28). In subgroups of individuals with suicidal ideation and behavior represent deficits in cognitive function (i.e., executive function, impulsivity, attention) (26–28). A conceptual framework of cognitive function in mood disorder was provided by McIntyre and colleagues in a review indicating implicit suicidal ideation as a domain of hot cognition (35). These give us indication that improvements in cognitive function may be closed to the anti-suicide effect by the antidepressant treatment, with no exception to ketamine.

The hot and cold cognitive functions could benefit from subanesthetic dose of ketamine. Ketamine’s benefit on hot cognition is embodied in the improvement in suicidal ideation, anhedonia, and overall mood symptoms (7, 9, 14, 37). Furthermore, ketamine’s enhancement in positive bias and decline in negative emotional perception in MDD further supports its favorable effect on hot cognition (38). The cold cognition benefit is implicated by improvement in working memory, visual memory, and processing speed after ketamine treatment, although some studies reported it was mediated by depressive symptom improvement (22–25).

Our results suggested that repeated ketamine infusions have a pro-cognitive effect independent of decrease in depression severity. Mediational analysis indicated that the improvement in the processing speed was partly independent of decrease in the MADRS score, and the direct effect (coef = 2.732) was greater than the indirect effect (coef = –0.118 × –9.968 = 1.220). The foregoing is in accordance with the results of two other studies with repeated ketamine infusions. McIntyre et al. reported improvement in Trail Making Test-B, measurement processing speed and executive function, was fully independent of improvement in depressive symptom after four infusions in individuals with MDD and bipolar depression (25). Shiroma et al. (24) observed a mood independent pro-cognitive effect on processing speed, attentional set-shifting and spatial working memory among patients with TRD following six doses of ketamine treatment (24).

The results of mediation analysis suggested that the specific effect of ketamine on suicidal ideation also was positively related to pro-cognitive effect. In addition, the present results showed ketamine’s specific decrease in suicidal ideation was larger in individuals with greater processing speed at baseline. In a randomized controlled trial of a single infusion of ketamine, Price et al. reported patients with higher severity of baseline suicidal cognition manifest greater anti-suicide effects, which were partly mediated by a decrease in non-suicide-related depressive symptom (11).

In a naturalistic follow-up study of participants with suicidal ideation found that better cognitive flexibility, assessed by the Wisconsin Card Sorting Test which also reflected executive function, predicted less suicidal ideation after 6 months (39). Our previous clinical study found early increase in levels of kynurenic acid, an NMDA receptor agonist leading to neuroprotection by elevation of extracellular glutamate, could predict anti-suicide response following six infusions of ketamine treatment (40). This is partially supported that patients who have a higher intrinsic “restorative” capacity of neuronal plasticity may be more likely to benefit from ketamine.

In addition, the neuromechanism of ketamine on brain may be involved in the relationship between its pro-cognitive effects and anti-suicide as well as antidepressant effects. Ketamine as an NMDA receptor agonist has a unique mechanism of action involving glutamate modulation via increased NMDA and α-amino- 3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor throughput, leading to potentiation of BDNF and mechanistic target of rapamycin (mTOR) signaling pathways, and these finally contribute to the augmented synaptogenesis and synapse stabilization (41, 42). Ketamine’s glutamatergic modulation of numerous brain regions and neural circuits, as shown in neuroimaging studies, was considered to involve in its antidepressant and anti-suicide action, and also pro-cognitive effect (43–46). The prefrontal cortex (PFC) global connectivity could be normalized in responders and the extent of the PFC global connectivity increase was associated with response after a single dose of ketamine (43, 44). Another study found increased hippocampal volumes and decreased nucleus accumbens volumes 24 h post-single dose of ketamine (45). Our previous study also showed increased volumes of the left amygdale and the right hippocampus after six infusions of ketamine (46). These brain regions and neural circuits correlated to ketamine’s treating response also involve hot and cold cognitive functions in patients with MDD, as mentioned in the previous study (36). Herein, we propose that ketamine’s antidepressant effects are mediated, in part, by targeting neural circuits that subserve the relevant to cognitive processing and emotional processing (35). It stands to reason that given ketamine’s known effect on neural structures and functions, its anti-suicide effects may be in part via pro-cognitive effects. The results from this study provide some clues for future work aimed to understand the neural mechanism for the effect of ketamine on suicidal ideation, and these evidences augmented with neuroimaging would be an important pathway for future work.

The present findings should be considered the following limitations. First, the results were obtained from a single arm open label study without a control arm. This may introduce expectation bias both to assessments from the clinician and patients. Second, several lines of evidence suggested that executive function, particularly the decision-making was related to suicidality. Decision-making performance was measured usually using the Iowa Gambling Task (IGT), and poorly performed by patients with current or history of suicide attempt/ideation was observed compared with the healthy controls (28, 29). This study only assessed four cognitive domains, although the Category Fluency test, Trail Making A test, and BACS could well reflect the executive function, lack of decision-making tasks was a significant limitation to describe a complete association between cognition and suicidal ideation. This limitation led to an incomplete relationship between cognitive function and anti-suicide effects of ketamine. Third, patients with severe suicidal ideation or suicidal behavior were excluded for safety reason, therefore, making it difficult to generalize present results to that population. Thus, our findings should not be extrapolated to outcomes of suicidal behavior, particularly attempted or completed suicide. Fourth, all subjects were taking their prior prescribed medications, including benzodiazepine, lithium, and antipsychotics, which may moderate changes in cognition. While changes to medication were not permissible during the study, the effect of psychiatric medication could confuse the observed outcomes. Therefore, the current findings could not be extrapolated to medication-free patients. An additional limitation is we were looking at reduction in suicidal ideations future studies should be looking at reductions in suicide and we do not know if reductions in suicidal ideations would necessarily predict reduction in suicide.

The present results provide insights into potential moderators and psychological mechanisms of ketamine’s anti-suicide effects, which are associated with pro-cognitive effects. Further study is warranted to explore the mechanisms underpinning ketamine’s anti-suicide effects which might reflect glutamate-mediated synaptogenesis and neural circuits serving cognitive function. Future research should also endeavor to combine ketamine with psychosocial strategies (e.g., cognitive behavioral therapy) to determine whether there’s a more robust improvement in measures of suicidality and cognitive function.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Clinical Research Ethics Committee of The Affiliated Brain Hospital of Guangzhou Medical University. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

YZ, YN, and RM developed research hypothesis and study design. YZ conducted data analysis and wrote the final draft of the manuscript. CW, XL, WL, KW, and ZC involved with data collection. All authors contributed to the final manuscript proofreading, edits, and approval for submission.

The 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.

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.