Cognitive Effects of Anticonvulsants; a Review

Fernando Ivanovic-Zuvic, Carla Iriarte-Guggiana,


Introduction: Anti-epileptic drugs (AED) are used in multiple mental disorders beyond epilepsy. Mainly, they are also used in Bipolar Disorder (BD). It is known that bipolar patients present neurocognitive deficits in all their phases and that AED can also affect cognition. Therefore, establishing a profile of cognitive adverse effects is essential for patients.

Objective: Using the available evidence, we compared the neurocognitive effects of AED, to establish preferences of use. 

Method: We reviewed the PubMed database and Google Scholar through a list of keywords defined by the authors. Completed studies realized in the last ten years were included, excluding isolated clinical cases in the analysis.

Discussion: Despite the methodological differences and heterogeneity of the studies, it seems that in terms of neurocognitive effects, lamotrigine has the least adverse effects, followed in order by oxcarbazepine, topiramate, valproic acid and then carbamazepine.

Keywords: Cognitive effects, anticonvulsants, bipolarity


Anticonvulsant drugs are currently used to treat several mental disorders in addition to epilepsy. They are also used to treat mood disorders because some of them can help with mood stabilization (1).

When speaking of anticonvulsants, the reference is to an ample family of drugs with different characteristics among them, but they all have in common the ability to reduce neuronal excitability. It was first found that drugs that work on the sodium channels (lamotrigine, etc.) have fewer cognitive effects than those that work through GABA. The latter (for instance, topiramate, valproate, gabapentin) affect particularly on watchfulness and attention, acting more as sedatives (2,3).

The effect of drug treatments on cognition is known as a "double-edged sword" because they improve cognition by acting on psychotic and mood symptoms, but at the same time, can have deleterious cognitive effects, depending on the drug, the dose, and its adverse effect profile (4).

Bipolar disorders (BP) are described as having cognitive deficits throughout all its phases, including alterations in verbal memory, sustained attention, working memory, and executive functions (4). Sometimes it is difficult to distinguish whether these deficits are caused by the chronicity of the disease, comorbidities, aging, polypharmacy, substance dependence, or a combination of these factors (5).  Since patients with fewer neurocognition alterations have a better prognosis, neurocognitive tests are recommended at the beginning of the treatment and for long-term monitoring (5, 6).

To have a better understanding of these results, the specific properties of each one of these five drugs will be revised: (carbamazepine, oxcarbazepine, topiramate, valproate, and lamotrigine).

Valproate: Discovered in 1882, as an organic solvent, but it was not until 1964 that the first clinical research for epilepsy was done, and in 1966 its efficiency to treat bipolar disorder was proved. Finally, in 1995 it was approved by the FDA for BD.

It is a short-chain fatty acid, with a mainly hepatic metabolism (95% of it), so it must not be used for hepatopathies.  It is metabolized through UDP-glucuronyl transferase (UGT), beta-oxidation, Cytochrome YP2C9, and CYP2C19.  Its bioavailability is higher than 80%, with a protein fixation of more than 90%. Its maximum peak concentration is between 1 and 4 hours after administration, and its mean half-life is between 9 and 16 hours (7).

There are at least three possible action mechanisms of valproate:

1) Inhibition of voltage-sensitive sodium channels. 2) Enhancement of the actions of the GABA neurotransmitter. 3) Regulation of signaling cascades.          
 Another theory affirms that it could reduce the excess of neurotransmission (in mania) through a decrease in the flow of ions through voltage-sensitive sodium channels (8)

Lamotrigine: is a diaminotrazine, antagonist to glutamic acid.  The action mechanisms are mainly through the blockage of the type-2 sodium channels, calcium channels, and the decrease in the release of neuroexcitatory amino acids, such as glutamate. Moreover, it blocks the reuptake of serotonin and it has serotoninergic effects (8,9). When used as a mood stabilizer, the daily dose is between 100 and 200 mg/day.

It is wholly and rapidly absorbed, having a plasmatic peak in 2-3 hours after administration. It has a mean elimination half-life of 33 hours, which can increase to 59 hours when used along with valproate (10).

Its metabolism occurs mainly through hepatic glucuronidation, and not through cytochrome p450.  Its metabolite is inactive, and it is excreted through the kidney.
Carbamazepine: it is a  lipophilic substance synthesized in 1953. In 1965 it began to be used as an antiepileptic, and in 1971 it was replaced by antipsychotics as a treatment for mania. It is interesting to note that the mood-stabilizing effect of carbamazepine seems not to depend on its anticonvulsant effect, as the former requires regular administration for more than six weeks, while the latter appears during 24 to 48 hours after administration(9).

Its mechanism of action is to inhibit the voltage-dependent sodium channels, inhibiting the release of glutamate. Its pharmacological characteristics include protein binding between 75% and 80%, and a  hepatic metabolism (through cytochrome p450), producing active metabolites. It has a plasmatic peak between 4 and 8 hours after administration, and its mean half-life is widely variable, between 18 and 55 hours. This ample variability is because after 2 to 3 months of treatment, it undergoes enzymatic autoinduction (9).

Oxcarbazepine: it is a prodrug of the carbamazepine 10-hydroxy, and its mechanism of action is to block the voltage-sensitive sodium channels and to inhibit the release of glutamate. 

It metabolizes in the liver and is excreted through the kidney. It has a mean half-life between 8 and 12 hours. It inhibits the CYP450 2C19 and its a soft inductor of CYP450 3A4. The advantage concerning carbamazepine is that oxcarbazepine does not induce its metabolism, and has fewer pharmacological interactions as well as side effects (11).

Topiramate: it is an anticonvulsant that has multiple action mechanisms. Prominent among them are: blocking of the voltage-dependent sodium channels, inhibition of the release of glutamate, an increase of brain concentrations of GABA, and inhibition of carbonic anhydrase. 

Prominent among its adverse effects are those produced by the blocking of carbonic anhydrases, such as the formation of kidney stones, paresthesia and metabolic acidosis (12).


Both Pubmed and Google Scholar databases were reviewed.  In these reviews, the following terms were used: "anticonvulsants", "bipolar", "neurocognitive effects" and "adults", in Pubmed, having as searching filter studies done on humans during the last ten years. Thirty-four publications were found, two of which were relevant for our research. For this reason, the term "adult" was not used, obtaining 61 results; from these, eight that complied with the established criteria were selected. Four were considered not relevant to our research or did not have sufficient quality, thus not included. Moreover, from the suggested works in Pubmed, two reviews in 2001 and 2006 were included due to their relevance.

It has been noted that patients with monotherapy with lithium show similar results as healthy control patients in short-term auditive memory, long-term memory and attention. However, patients that took anticonvulsants as adjuvant therapy or are in monotherapy show worse results in these areas. Moreover, a subject who only use anticonvulsants showed worse results than controls in short-term visual memory, working memory and executive functions.
The limitations of these findings are probably that patients that achieve stabilization in monotherapy with lithium show a better evolution of the disease than those who need some type of adjuvant therapy, which could result in worse results in those tests (13).

In 2006, Gualtieri studied 159 bipolar patients, comparing the neurocognitive effects of 5 anticonvulsants vs. lithium (carbamazepine, lamotrigine, oxcarbazepine, topiramate and valproate). The requisites were that they had bipolar disorder as diagnosis, were between 18 and 70 years, and did not have neurological or cognitive comorbidities. Further, they had to remain in monotherapy with some of the medications mentioned, in regular dose at least for four weeks, and without using other drugs.

A 30-minute self-administered test called CNSVS was given to these patients, which is a set of neurocognitive questions that were to be answered in a computer. It consists of finger taping, verbal and visual memory, Stroop test, symbol digit coding, shifting attention test and continuing performance test, which, when combined, form five domains: reaction time, psychomotor speed, memory, cognitive flexibility and intricate attention (3) (Figure 1).

Figure No 1: Neurocognitive effects of 6 mood stabilizers.


This graphic allows us to conclude that lamotrigine has a lower level of adverse effects in the analyzed areas, and that valproate has worse results in the tests performed (3).

The average of the summation of the five domains is shown as the neurocognitive index (INC), which is similar to IQ in the WAIS scale. Despite its limitations, it could be used to compare neurocognitive alterations in patients using different medications in a range from 0 to 100, the latter being optimal neurocognitive functioning. Figure No 2 (3).


Figure No 2. Neurocognitive index.


In spite of the heterogeneity of the sample in this study, it seems that, in terms of the INC, lamotrigine is the one that shows the fewest adverse effects, followed by oxcarbazepine, topiramate, valproate, and carbamazepine (3).  Interestingly, carbamazepine has a lower index than valproate.

The study of Gualtieri and Johnson also suggests dividing anticonvulsants into three groups, according to the degree of toxicity (high, medium and low).  Lamotrigine and oxcarbazepine are part of a small toxicity group, while topiramate, valproate and carbamazepine are part of the group with the highest disruption. Lithium is included in the study for comparison (even though it is not an antiepileptic) and is in the category of medications with intermediate effects.  Differences among groups could be seen in psychomotor speed, cognitive flexibility and intricate attention, being relevant to the variations when comparing lamotrigine to the other four anticonvulsants. (carbamazepine, oxcarbazepine, topiramate, valproate). Figure No 3 (3).

Figure No 3. Cognitive effects of anticonvulsants.


The following table shows the results of the items used to obtain the neurocognitive index (consisting of 5 domains: reaction time, psychomotor speed, memory, cognitive flexibility and intricate attention). To study the group differences multiple variance analyses considering covariables such as age, race and sex. If a variance analysis showed relevant group differences, a t-test was performed to determine the cause of the difference (Table No 1).

Table No 1 Average scores of 159 bipolar patients on mood stabilizers.        


Group discrepancies (high, medium, and low neurocognitive toxicity) were seen in 3 domains of the neurocognitive index: psychomotor speed, cognitive flexibility, and complex attention. As in the following tests: visual memory, Stroop reaction time, codification, shifting attention test and continuing test performance.

Significant differences were seen in the neurocognitive indices of:
a) Lamotrigine and carbamazepine
b) Lamotrigine and lithium
c)  Lamotrigine and topiramate
d) Lamotrigine and valproate.

In the following table, this is translated to disease ranking of the analyzed areas, being one the best option and six the worst:

Table No 2 Neurocognitive disease ranking.


It is difficult to evaluate whether the positive cognitive results observed in some studies with lamotrigine are due to a positive cognitive effect of the medication or because the patients that react to this medication monotherapy are in general less severe cases. 

There is a study that compares the cognitive results of medicated vs. non-medicated bipolar patients, but, once again, we find the same evaluation problem, because a patient that may not be taking medication maybe has a more benign course of the disease, and is less affected in cognitive terms.

Although the study of Gaultieri and Johnson is very attractive in its way of presenting information, maybe it is lacking a more detailed statistical analysis divided by sections, because it is difficult to conclude when comparing too many variables in one single study.

However, some conclusions can be reached from the analyzed studies in which lamotrigine stands out as the medication with the fewest side effects, followed by oxcarbazepine, topiramate, valproate, and carbamazepine.

Studies that evaluate the cognitive effect of antiepileptic medications in healthy volunteers could help to understand the real impact they have on cognition. However, it would be better to have measurements of our patients before they begin treatment to have contextualized points of comparison.


  1. Ivanovic-Zuvic, F. Psicopatología en la Epilepsia. Santiago: Mediterraneo, 2010.
  2. Morales M. Efectos cognoscitivos de los anticonvulsivantes. Acta Neurologica Colombiana 2008; S3-S7.
  3. Gualtieri  C. T, Johnson  L. G. Comparative Neurocognitive Effects of 5 Psychotropic Anticonvulsants and Lithium. Med Gen Med 2006; 8: 1-19.
  4. Dias, B. Pharmacological approaches in bipolar disorder and the impact on cognition: a critical overview. Acta Psychiatr Scand 2012; 126: 315-331.
  5. Goldberg J.F  , Burdick KE. Cognitive side effects of anticonvulsants. J Clin Psychiatry 2001; 62: 27-33.
  6. Dols A, Sienaert P. The prevalence and management of side effects of lithium and anticonvulsants as mood stabilizers in bipolar disorder from a clinical perspective: a review. Int Clin Psychopharmacol 2013; 28: 287–296.
  7. Stahl, S. Prescriber´s Guide. In:  S. Stahl, Prescriber´s Guide. New York: Cambridge University Press, 2015.
  8. Stahl, S. Estabilizadores del Humor. Psicofarmacología Esencial. Madrid: Aula Médica. 2014.
  9. Silva, H. Mecanismos de acción de los estabilizadores del ánimo. Rev Chil Neuro-Psiquiatr 2001;39: 219-230.
  10. Stahl, S. Lamotrigine. Prescriber's Guide, sixth edition. New York: Cambridge University Press. 2015.
  11. Stahl, S. Oxcarbazepine. Prescriber's Guide, sixth edition. New York: Cambridge University Press, 2015.
  12. Stahl, S.  Topiramate. Prescriber's Guide, sixth edition. New York: Cambridge University Press, 2015.
  13. Sabater,  A. Comparative neurocognitive effects of lithium and anticonvulsants in long term stable bipolar patients. J Affect Disord 2016; 190: 34–40