Schizophrenia is a chronic and severe psychiatric disorder characterized by positive and negative symptoms, neurocognitive dysfunction, and social dysfunction (1). Despite its relatively low lifetime prevalence of approximately 1% (2, 3), schizophrenia imposes substantial socioeconomic burdens and contributes significantly to disability worldwide (4). Long-term or lifelong administration of antipsychotics (APs), such as olanzapine and aripiprazole, is essential for preventing disease recurrence (5). However, these medications fail to address neurocognitive dysfunction (6) and may even exacerbate it (7, 8). Thus, there is an urgent need for effective therapeutic strategies targeting neurocognitive symptoms.

Guidelines from the European Psychiatric Association (9) recommend managing neurocognitive dysfunction in schizophrenia through pharmacological interventions, psychosocial strategies, and somatic treatments, which include non-invasive brain stimulation techniques such as transcranial direct current stimulation (tDCS) (10) and repetitive transcranial magnetic stimulation (rTMS) (11). However, these interventions provide only modest improvements in neurocognitive performance (9). Neurocognitive symptoms in schizophrenia often persist despite the resolution of positive symptoms, posing a significant therapeutic challenge (12). Therefore, it is urgent to improve neurocognitive function with effective pharmacological agents.

Metformin, a biguanide hypoglycaemic agent commonly prescribed for type 2 diabetes management (6, 13), has garnered attention for its potential neuroprotective effects. It readily crosses the blood-brain barrier and enhances neurocognitive function through anti-inflammatory mechanisms and improved cerebral energy metabolism (14). Concurrently, neuroimaging modalities such as magnetic resonance imaging and positron emission tomography have implicated neuroinflammation and cerebral metabolic stress in neurocognitive impairment (15, 16). Supporting these mechanisms, animal studies have demonstrated that metformin ameliorates clozapine-induced learning deficits (17) and dizocilpine-induced working memory impairments (18). Additionally, a longitudinal cohort study linked metformin use to improved performance in verbal learning, working memory, and executive function (19). However, randomized controlled trials (RCTs) (20–23) investigating metformin for neurocognitive dysfunction in individuals with schizophrenia and comorbid physical disease (e.g., type 2 diabetes) have reported inconsistent findings.

While previous systematic reviews and meta-analyses (6, 24–26) have focused on the efficacy of metformin in addressing antipsychotic-induced dyslipidemia and weight gain, no systematic review of RCTs has evaluated its effects on neurocognitive function in schizophrenia. To understand the current literature on the role of metformin for neurocognitive effects in schizophrenia, and to provide a more comprehensive and robust basis for clinical application, this systematic review systematically examines the neurocognitive effects of metformin in patients with schizophrenia.

This systematic review, which included four RCTs and 271 participants, is the first to examine the neurocognitive effects of metformin in patients with schizophrenia. The primary findings are as follows: 1) metformin demonstrated significant superiority over controls in improving neurocognitive function as measured by the MCCB, RBANS, and MMSE, but not the BACS; 2) metformin combined with TAU did not significantly improve total psychopathology; and 3) metformin is safe and well-tolerated for enhancing neurocognitive function, though further studies with larger sample sizes are warranted.

In this review, 75% of the included RCTs (20–22) reported that metformin outperformed controls in improving neurocognitive function. However, one study (23) did not observe significant improvements as assessed using the BACS. This inconsistency may stem from the differing assessment tools employed across studies. Evidence suggests that the MCCB has become the gold standard for evaluating neurocognitive impairments in schizophrenia due to its comprehensive scope and strong psychometric properties. Specifically developed for schizophrenia populations, the MCCB has been extensively validated across diverse settings and populations (44). In contrast, the BACS, while more concise, may not fully capture the breadth of neurocognitive deficits as effectively as the MCCB (45). Such inconsistencies have also been reported in previous studies examining neurocognitive function in schizophrenia using different assessment tools. For instance, Hei et al. (46) found that adjunctive sulforaphane significantly improved working memory and verbal learning in schizophrenia, as measured by the MCCB and Hopkins Verbal Learning Test, but showed no significant effects on other measures, including the BACS. Collectively, these findings highlight the importance of employing standardized neurocognitive test batteries, such as MCCB, to ensure consistent and robust assessments. Future research should focus on elucidating the neurocognitive effects of metformin in patients with schizophrenia using standardized tools and well-powered study designs (47).

A recent meta-analysis (6) highlighted that, beyond improving neurocognitive function, metformin also provides benefits for weight management and metabolic syndrome in patients with schizophrenia. Furthermore, several studies have demonstrated metformin's potential to enhance neurocognitive function in other neurological conditions, including Parkinson’s disease (48), Alzheimer’s disease (49), and pediatric brain tumors (50). In the broader context of pharmacological treatments for neurocognitive dysfunction, drugs such as sulforaphane (46), erythropoietin (51), and huperzine A (52) also show promise in improving neurocognitive outcomes. However, no head-to-head studies have compared the neurocognitive effects of metformin with sulforaphane or erythropoietin in adults with schizophrenia. As a result, the relative efficacy of these drugs for neurocognitive enhancement remains unclear.

The mechanisms underlying metformin’s role in improving neurocognitive function in schizophrenia are not yet fully understood. One plausible explanation involves its ability to improve insulin resistance (6, 53, 54), a key pathophysiological factor associated with neurocognitive impairment (55). Insulin resistance has also been linked to weight gain (25), and metformin’s weight-reduction effects may further contribute to neurocognitive improvements. Studies have demonstrated that weight loss is associated with better neurocognitive outcomes, particularly in individuals with obesity or metabolic syndrome (56, 57). Furthermore, metformin’s positive effects on other metabolic indicators, such as lipid levels and glycemic control (25, 58), may also play a role in indirectly enhancing neurocognitive function. Besides its impact on metabolism, metformin decreases inflammation by altering pro-inflammatory cytokines, potentially enhancing neurocognitive function. In an animal study, metformin reduced the levels of pro-inflammatory cytokines like interleukin (IL)-1β, potentially aiding in the enhancement of spatial memory in diabetic animals (59). Furthermore, elevated levels of peripheral cytokines such as IL-1β are found in some patients with schizophrenia and are associated with neurocognitive impairment (60). This multifaceted approach underscores the potential of metformin in the integrated management of neurocognitive impairment in patients with schizophrenia.

Two RCTs reported mixed findings regarding the effects of metformin on decreased appetite and diarrhea compared with controls (20, 23), while both groups exhibited similar rates of discontinuation and other adverse events. Consistent with prior research, daily metformin at recommended dosages (61), is considered notably safe for both short-term and long-term use, with no significant adverse effects reported (6, 20, 24). Gastrointestinal side effects, such as diarrhea and decreased appetite, were the most commonly reported adverse events in this systematic review. These effects, typically observed at the onset of therapy, can be mitigated by lowering the dose, implementing gradual dose titration, or taking the medication with meals (61). Metformin has also demonstrated safety and tolerability in patients with schizophrenia (25), Alzheimer’s disease (49), pediatric brain tumors (50), and bipolar depression (62). However, prolonged use of metformin has been associated with reduced vitamin B12 levels and, in some cases, biochemical B12 deficiency (63, 64). Therefore, it is advisable for patients with schizophrenia receiving metformin to undergo regular monitoring of blood lactate, serum B12, and folate levels (63).

Several limitations of this systematic review should be acknowledged. First, the included RCTs utilized diverse methodologies, including four distinct assessment tools to measure neurocognitive function. This heterogeneity prevented the conduct of a meta-analysis. Importantly, it is different to determine the superiority of metformin over controls in improving a specific neurocognitive function. Second, this review included only four RCTs with a small sample size (n=271, ranging from 30 to 100 participants), necessitating caution in interpreting the findings. Third, the four RCTs focused on the use of metformin in adult patients with schizophrenia, limiting the generalizability of the findings to broader populations. Fourth, this systematic review of metformin for neurocognitive dysfunction in schizophrenia is not registered. Finally, the RCTs included participants with varying conditions, such as schizophrenia combined with glucose metabolism disorder, type 2 diabetes or prediabetes, which may have influenced the outcomes.

Preliminary evidence suggests that metformin may have beneficial effects on neurocognitive function in schizophrenia. However, to validate these findings, future research should focus on conducting large-scale, double-blind, and high quality RCTs.