This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

Clinical and Practical Psychopharmacology

Memantine as an Augmentation Treatment for Schizophrenia: Limitations of Meta-Analysis for Evidence-Based Evaluation of Research

Practical Psychopharmacology

Vertical divider


The action of memantine on N-methyl-d-aspartate (NMDA) glutamatergic receptors and the other pharmacodynamic actions of this drug suggest that it may benefit patients with schizophrenia. Many randomized controlled trials (RCTs) have examined this possibility. These RCTs have been meta-analyzed by at least 2 groups of authors. In one meta-analysis (8 RCTs, pooled N = 448), memantine (20 mg/d for 6-12 weeks) augmentation of antipsychotic drug therapy was found to attenuate the severity of negative symptoms and improve cognitive functioning; in both regards, the effect was large. Memantine was also associated with a small but statistically significant reduction in general psychopathology. Whereas memantine did not significantly attenuate positive symptom, depression, total psychopathology, and global illness ratings, it was also not associated with an increased risk of individual adverse events, discontinuation due to adverse events, or all-cause discontinuation. The findings of the other meta-analysis, which examined much the same body of literature, were largely similar. On the surface, these results suggest that memantine may be considered for the reduction of negative symptoms and cognitive impairment in schizophrenia. However, an examination of the individual RCTs and a careful look at the findings of the meta-analyses identify so many important concerns that it is probably premature to draw conclusions about the usefulness of memantine as an augmentation strategy in schizophrenia. At best, it may be stated that there is a signal that supports the study of memantine in schizophrenia patients who are specifically impaired by negative symptoms and cognitive complaints. This is a good example of a situation in which meta-analysis did not provide a trustable evidence-based interpretation of literature.

J Clin Psychiatry 2017;78(9):e1307-e1309


Many patients with schizophrenia do not respond completely to treatment with antipsychotic medications and continue to experience positive symptoms, negative symptoms, mood disturbance, cognitive impairment, and other difficulties. A wide range of pharmaceutical agents has been trialed as augmentation therapy for such patients; these agents include antidepressants, anticonvulsants, antihistaminics, antihypertensives, anti-inflammatory drugs, benzodiazepines, β-blockers, cannabis derivatives, cholinesterase inhibitors, glutamatergic agents, hormones such as oxytocin and erythropoietin, mood stabilizers, neuropeptides, omega-3 fatty acids, serotonin 5-HT3 and 5-HT6 receptor antagonists, sigma receptor ligands, statins, and xanthine oxidase inhibitors, among many others. Memantine is one among these.

Why Memantine?

Memantine is a low affinity, fast-off, voltage-dependent, noncompetitive N-methyl-d-aspartate (NMDA) receptor antagonist that has been approved for the treatment of moderate to severe Alzheimer’s disease.1 Speculatively, there are at least 4 reasons why memantine may benefit patients with schizophrenia:

  1. Glutamatergic excitotoxicity has been suggested as a pathological mechanism in schizophrenia, particularly in early schizophrenia, and as an explanation for positive symptoms and neuroanatomical changes.2,3 Memantine interferes little with physiological glutamatergic signaling but inhibits the mechanisms underlying glutamate-driven neuronal excitotoxicity.4
  2. Memantine has an inhibitory effect on the 5-HT3 receptor,5 and 5-HT3 antagonists have been suggested to improve negative symptoms in schizophrenia.6
  3. Memantine has complex actions on α4β2, α7, and other nicotinic cholinergic receptors,7 and these actions may be relevant for the attenuation of cognitive deficits in schizophrenia.
  4. Memantine may have D2 receptor agonistic effects8; whereas this could be of concern for positive symptoms of schizophrenia, it may contribute to the attenuation of negative symptoms.

These speculations could support the consideration of memantine to treat the positive, negative, and/or cognitive symptoms of schizophrenia.

Memantine for Schizophrenia: Clinical Evidence

No matter how elegant pharmacodynamic speculations are, they remain speculations until supported by clinical evidence. In this context, Kishi et al9 described a systematic review and meta-analysis of the clinical benefits of memantine augmentation of antipsychotic drugs in patients with schizophrenia.

These authors9 searched electronic databases, clinical trial registries, reference lists, and other sources and identified 8 randomized, double-blind, placebo-controlled trials (pooled N = 448) that met their study selection criteria. One study was a crossover trial; in this study, only data from the first treatment phase were examined because a crossover effect was identified despite a 2-week washout period.

These randomized controlled trials (RCTs) were 6-12 (median, 12) weeks in duration. Patients in the studies were young (mean, 39 years). Risperidone and clozapine were the augmented antipsychotic drugs in 2 studies each. One study augmented only first-generation antipsychotic drugs, 1 study augmented olanzapine, and the remaining 2 studies augmented a mixture of second-generation antipsychotic drugs. All studies dosed memantine augmentation therapy at 20 mg/d.

Important findings from this meta-analysis9 are presented in Table 1. In summary, memantine was associated with significant improvement in negative symptoms, cognition, and general psychopathology, but not in depression, total psychopathology, or global ratings of illness severity. The magnitude of improvement was large for both negative symptoms and cognition. The magnitude of improvement in general psychopathology was clinically negligible. Memantine was not associated with an increased risk of specific adverse events or of dropout due to adverse events. There did not appear to be evidence of publication bias.

Table 1

Click figure to enlarge

In meta-regression, younger age was associated with better negative symptom outcomes, but this could have been a false-positive error arising from the many relationships tested, and anyway the number of RCTs was too small for the meta-regression results to be confidently interpreted.

A Second Meta-Analysis

During the same year, Zheng et al10 also presented a systematic review and meta-analysis of RCTs of memantine augmentation of antipsychotic drugs in patients with schizophrenia. Seven of the 8 RCTs in this meta-analysis were the same as those in the meta-analysis by Kishi et al.9 This meta-analysis10 had 1 Chinese RCT in place of 1 Iranian RCT in the previously described meta-analysis.9

The second meta-analysis10 found memantine superior to placebo for negative symptom ratings and Mini-Mental State Examination (MMSE) scores but not for general psychopathology ratings. Other outcomes were largely the same as in the previously described meta-analysis.9

Concerns About the Primary Data

There are potential biasing influences in the primary data in the meta-analysis that was described in detail9; most of these concerns apply to the other meta-analysis,10 as well. One RCT, from Brazil, randomized just 22 patients and obtained a standardized mean difference (SMD) of 3.33, favoring memantine over placebo for negative symptom outcomes. This effect size is so large as to be incredible in parallel-group, clinical trial research.

Four of the 8 RCTs were from a single country (Iran), and all 4 found memantine superior to placebo on at least some efficacy outcomes. In contrast, only 2 of the remaining 4 RCTs from other countries showed an efficacy advantage for memantine. With specific reference to the negative symptom data, the SMDs for the 3 Iranian RCTs (that were included in the analysis) were very large and lay in the 1.1 to 1.6 range; that of the Brazilian RCT, as already mentioned, was astronomically large at 3.33; and those of the remaining 3 RCTs lay in the near null, 0.01 to 0.05 range. The discrepancies are substantial and striking.

Memantine was superior to placebo for 3 outcomes: negative symptoms ratings (7 RCTs), MMSE scores (3 RCTs), and general psychopathology ratings (4 RCTs). One RCT in each of these 3 analyses provided observed cases data instead of intent-to-treat data. This RCT was associated with significant and large effect sizes for both negative symptom and MMSE outcomes. The scope for bias in the analyses is therefore large.

The advantage for memantine for general psychopathology ratings was driven by a single, small (N = 40) RCT with 90% weightage in the analysis. The high weightage appeared to be because of the precision of the findings in the study. The other 3 RCTs had large confidence intervals that were roughly equally spread across the null line. The validity of this analysis is therefore uncertain.

Although memantine was superior to placebo for MMSE outcomes, this result is hard to interpret because the MMSE is not an appropriate instrument for the measurement of cognitive dysfunction in schizophrenia.

Seven of the 8 RCTs had a sample size in the 22 to 64 (median, 40) subjects range. These were all industry-independent. Six of these had outcomes that favored memantine. The only industry-driven RCT was large (N = 138), and no outcome favored memantine. This is an unusual reversal of what is often seen—that small studies fail to find a statistically significant effect (unless there is evidence of publication bias) and that large, industry-driven studies find a significant effect favoring the study medication.

Heterogeneity in most analyses was so large as to render the findings hard to interpret. Sensitivity analyses did suggest sources of heterogeneity, but these were based on too few RCTs, or the outcomes were insufficiently logical for confident conclusions to be drawn. As an example, in the negative symptom analysis, there was no heterogeneity (I2 = 0%) in 2 risperidone RCTs and substantial heterogeneity (I2 = 90%) in the remaining 5 RCTs. Two RCTs is too small a number to allow a conclusion that memantine effectively attenuates negative symptoms in risperidone-treated patients, and if this were a true finding, it would beg the question of what is special about the risperidone-memantine interaction.

One study was conducted in patients with negative symptoms; a few in patients with residual positive symptoms or refractory schizophrenia; and a few in unselected patients with schizophrenia. Furthermore, different studies were conducted with different objectives; reduction in negative symptoms, improvement in cognition, and reduction in total psychopathology were all primary objectives in different studies. Therefore, it is difficult to conclude to what kind of schizophrenia patient the results of the meta-analysis can be generalized.


On the surface, the findings of meta-analysis appear to suggest that the average antipsychotic-treated patient with schizophrenia can expect to experience a substantial reduction in negative symptoms, a substantial improvement in cognition, and a small improvement in general psychopathology after receiving augmentation treatment with memantine (20 mg/d) for 6-12 weeks. However, there are far too many concerns about the RCTs that were the source material for the meta-analysis, and far too many concerns in the findings of the meta-analysis, for confident clinical guidance to be possible. At best, it is suggested that there is a signal to encourage the study of memantine in schizophrenia patients who have persistent negative symptoms and/or clinically significant cognitive impairment. This conclusion could have been drawn from an inspection of the results of the original RCTs. Meta-analysis has not contributed to a better understanding of the literature in the field.

Published online: November 21, 2017.

Each month in his online column, Dr Andrade considers theoretical and practical ideas in clinical psychopharmacology with a view to update the knowledge and skills of medical practitioners who treat patients with psychiatric conditions.

Department of Clinical Psychopharmacology and Neurotoxicology, National Institute of Mental Health and Neurosciences, Bangalore, India (
Financial disclosure and more about Dr Andrade.


1. Alam S, Lingenfelter KS, Bender AM, et al. Classics in chemical neuroscience: memantine. ACS Chem Neurosci. 2017;8(9):1823-1829. PubMed CrossRef

2. Stone JM, Morrison PD, Pilowsky LS. Glutamate and dopamine dysregulation in schizophrenia—a synthesis and selective review. J Psychopharmacol. 2007;21(4):440-452. PubMed CrossRef

3. Plitman E, Nakajima S, de la Fuente-Sandoval C, et al. Glutamate-mediated excitotoxicity in schizophrenia: a review. Eur Neuropsychopharmacol. 2014;24(10):1591-1605. PubMed CrossRef

4. Parsons CG, Stvffler A, Danysz W. Memantine: a NMDA receptor antagonist that improves memory by restoration of homeostasis in the glutamatergic system—too little activation is bad, too much is even worse. Neuropharmacology. 2007;53(6):699-723. PubMed CrossRef

5. Rammes G, Rupprecht R, Ferrari U, et al. The N-methyl-D-aspartate receptor channel blockers memantine, MRZ 2/579 and other amino-alkyl-cyclohexanes antagonise 5-HT3 receptor currents in cultured HEK-293 and N1E-115 cell systems in a non-competitive manner. Neurosci Lett. 2001;306(1-2):81-84. PubMed CrossRef

6. Kishi T, Mukai T, Matsuda Y, et al. Selective serotonin 3 receptor antagonist treatment for schizophrenia: meta-analysis and systematic review. Neuromolecular Med. 2014;16(1):61-69. PubMed CrossRef

7. Rammes G, Danysz W, Parsons CG. Pharmacodynamics of memantine: an update. Curr Neuropharmacol. 2008;6(1):55-78. PubMed CrossRef

8. Seeman P, Caruso C, Lasaga M. Memantine agonist action at dopamine D2High receptors. Synapse. 2008;62(2):149-153. PubMed CrossRef

9. Kishi T, Matsuda Y, Iwata N. Memantine add-on to antipsychotic treatment for residual negative and cognitive symptoms of schizophrenia: a meta-analysis. Psychopharmacology (Berl). 2017;234(14):2113-2125. PubMed CrossRef

10. Zheng W, Li X-H, Yang X-H, et al. Adjunctive memantine for schizophrenia: a meta-analysis of randomized, double-blind, placebo-controlled trials. Psychol Med. 2017;1-10. PubMed CrossRef

Related Articles

Volume: 78

Quick Links: Schizophrenia and Schizoaffective Disorders