J Neurol Neurosurg Psychiatry doi:10.1136/jnnp-2012-303688
  • Epilepsy
  • Short report

Weight change associated with antiepileptic drugs

  1. Mark I Rees1,2
  1. 1Neurology and Molecular Neuroscience Research Group, Institute of Life Science, College of Medicine, Swansea University, Swansea, UK
  2. 2Wales Epilepsy Research Network (WERN), Institute of Life Science, Swansea University, Swansea, UK
  3. 3Health Informatics Research Unit (HIRU), Institute of Life Science, Swansea University, Swansea, UK
  4. 4Neurology Department, University Hospital Wales, Cardiff, UK
  1. Correspondence to Dr William Owen Pickrell, Neurology and Molecular Neuroscience Research Group, Institute of Life Science, College of Medicine, Swansea University, 3rd Floor, ILS, Swansea SA2 8PP, UK; w.o.pickrell{at}
  • Received 16 July 2012
  • Revised 8 November 2012
  • Accepted 11 November 2012
  • Published Online First 12 December 2012


Aim To investigate antiepileptic drug (AED)-related weight changes in patients with epilepsy through a retrospective observational study.

Method We analysed the anonymised electronic primary care records of 1.1 million adult patients in Wales. We included patients aged 18 years and over with a diagnosis of epilepsy, whose body weight had been measured up to 12 months before starting, and between 3 and 12 months after starting, one of five AEDs. We calculated the weight difference after starting the AED for each patient.

Results 1423 patients were identified in total. The mean difference between body weight after and before starting each AED (together with 95% CI and p values for no difference) were: carbamazepine (CBZ) 0.43 (−0.19 to 1.05) p=0.17; lamotrigine (LTG) 0.31 (−0.38 to 1.00) p=0.38; levetiracetam (LEV) 1.00 (0.16 to 1.84) p=0.02; sodium valproate (VPA) 0.74 (0.10 to 1.38) p=0.02; topiramate (TPM) −2.30 (−4.27 to −0.33) p=0.02.

Conclusions LEV and VPA were associated with significant weight gain, TPM was associated with significant weight loss, and LTG and CBZ were not associated with significant weight change.


Around 6 00 000 people (∼1%) in the UK have a diagnosis of epilepsy and are prescribed antiepileptic drugs (AEDs).1 Most of these people will need lifelong treatment with AEDs. It is important for clinicians to have an accurate understanding of the side-effect profile of AEDs in order to select the most suitable AED and be able to provide appropriate information to the patient. Both clinicians and patients identify AED side effects as a priority area of research in epilepsy.2

There are several reasons why patients with epilepsy may experience weight change, one of these being AED therapy.3 Weight gain is associated with an increased risk of comorbidities (eg, type 2 diabetes mellitus, heart disease and some types of cancer) and impairs quality of life and self-esteem. Weight loss is also associated with comorbidity (eg, osteoporosis, immune and metabolic problems)—of particular importance in the young and elderly populations. Several AEDs have previously been linked to both weight gain and weight loss.3 If the risks of weight change are not adequately explained to patients then: (1) they may lose the opportunity to ameliorate the risk via prospectively monitoring their weight and executing early lifestyle change; (2) subsequent weight change may contribute to discontinuation of the associated AED with the associated risk of increased seizure frequency; (3) or the physician may risk stopping another medication not implicated in weight change by failing to identify the most likely drug.

Most of the limited literature currently available on weight change and AEDs is derived from data from clinical trials. Our aim was to investigate and quantify weight change associated with five of the most commonly prescribed AEDs in the UK in a primary care setting. Having been routinely collected in the community, these prospectively acquired data are free from pharmaceutical industry bias.


General practice (GP) electronic health records are stored at the health information research unit (HIRU) at Swansea University as part of the secure anonymised information linkage (SAIL) system.4 ,5 At the time of analysis, approximately 40% of the Welsh population's GP records were available on the SAIL system (around 1.1 million people).

We electronically searched all SAIL records for adult patients with a coded diagnosis of epilepsy (note, that it was not possible to record the exact type of epilepsy from the GP records). We then refined our search by including only those people who

  • had their weight recorded up to 12 months before starting (carbamazepine (CBZ); lamotrigine (LTG); levetiracetam (LEV); sodium valproate (VPA) or topiramate (TPM)) for the first time and

  • had their weight recorded within 3–12 months of starting and remaining on the same AED and

  • were 18 years of age or older at the time of starting the AED. (Significant weight gain can occur during various stages of child development, and so, children were not included in this study).

We chose five commonly prescribed AEDs and did not include older AEDs which are not as commonly prescribed in the UK. We included generic and all recognised trade names for the AEDs in the search. We considered that most weight change associated with AED treatment occurs during the first 3 months after initiation,6 and most people have reached an initial target dose in this period, and so we chose a time window of 3–12 months after starting the AED for the second weight measurement. We excluded people who had a biologically implausible (>40% of their baseline body weight) weight gain or loss within the study period.

For each patient meeting the above criteria, we recorded age, sex, weight before starting the AED and weight after starting the AED. It was not possible to record an exact defined daily dose of the prescribed AED using the GP dispensing information available on SAIL.

We recorded whether each patient was on AED monotherapy, was coprescribed a common drug which may affect weight (corticosteroids, hypoglycaemic agents, antidepressants or antipsychotics), or had a common comorbidity which may affect weight (diabetes, thyroid disease, depression, psychosis). We then calculated the time between the weight measurements and the weight difference (the difference between the weights measured before and after starting the AED) for each person. We checked that the distribution of the absolute and relative weight changes were approximately normally distributed by plotting frequency and normal plots. Using paired t tests, we calculated the significance levels and confidence limits for the mean weight difference and mean percentage weight difference for each drug group. We used χ2 tests to compare the proportions of each group on AED monotherapy, coprescribed common drugs which may cause weight change, or with common potentially weight-changing comorbidities.

This project was approved by the HIRU information governance panel (project 0193). The National Research Ethics Service has confirmed that projects using the anonymised data held by HIRU do not need specific National Health Service (NHS) research ethics committee approval.


We identified 19 622 adult patients with a diagnosis of epilepsy (approximately 1.5% of the population covered by the SAIL database). Of these, 1423 patients met our inclusion criteria. These patients had their weights recorded between 18 November 1988 and 4 April 2011. Table 1 and figure 1 give the main results.

Table 1


Figure 1

Plot of percentage weight change after confidence intervals starting each of the five antiepileptic drugs. The error bars represent 95% confidence intervals. CBZ, carbamazepine; LTG, lamotrigine; LEV, levetiracetam; VPA, sodium valproate; TPM, topiramate.

Although broadly similar, there were some differences between groups in terms of patient demographics. There were significantly fewer patients in the TPM group (n=67) compared with the other patient groups. Each patient group had 50% or more women, with LTG (66%) and TPM (73%) in particular having the highest proportion of women. The CBZ and VPA groups had a slightly older mean age in years (±1SD) of 50.3 (±18.1) and 50.7 (±19.7). The TPM group had a younger mean age of 40.4 (±13.2). The time between the weight measurements was similar for all drug groups. The mean time in days between the first weight measurement and starting the AED were CBZ 119, LTG 119, LEV 127, VPA 125 and TPM 119. The mean time in days between the second weight measurement and starting the AED were CBZ 208, LTG 200, LEV 193, VPA 211 and TPM 203.

Excessive weight changes are more important for the individual and can lead to discontinuing AED therapy. The proportion of people with a weight gain of >10% (95% CI) for each drug group were: CBZ 7.5% (4.6% to 10.3%); LTG 6.0% (3.2% to 8.9%); LEV 10.0% (5.6% to 14.3%); VPA 11.9% (8.9% to 14.8%) and TPM 4.5% (0% to 10.2%). The proportion of people with a weight loss of >10% (95% CI) for each drug group were: CBZ 7.7% (4.8% to 10.6%); LTG 3.7% (1.4% to 6.0%); LEV 4.0% (1.0% to 6.9%); VPA 5.8% (3.7% to 8.0%); TPM 13.4% (4.5% to 22.3%).


Our data show that patients taking TPM had a significant absolute and relative weight loss, and that the LEV and VPA groups had significant absolute and relative mean weight gain. Patients taking CBZ and LTG had non-significant absolute and relative mean weight gain. We believe that this is the first study to associate significant weight gain with LEV.

There was no significant difference between the groups in terms of the proportion of patients losing or gaining more than 10% of their baseline weight, coprescribed common drugs, or with common comorbidities known to cause weight change. The CBZ group had significantly fewer people on AED monotherapy when compared with the LTG, LEV and VPA groups.

There are inherent biases with this retrospective observational study. In particular, only patients who had their weight measured during a GP consultation within the appropriate time period were included. Patients who did not visit their GP, or visited their GP but did not get their weight recorded were not included. However, this bias applies equally to each drug group.

The demographics of each group are also slightly different and may account for some of the differences in weight change between the groups; however, this would not explain the difference in weight changes between the LTG and LEV groups who were not significantly different in terms of age, sex or baseline weight. However, there might be some confounding factors present when comparing patients on the same AED who experienced weight change and those who did not.

Previous data, largely from clinical trials, have associated weight gain with VPA3 ,7 and weight loss with TPM.8 LTG is generally considered to be weight neutral.9 Both weight gain10 ,11 and weight neutrality12 have been associated with CBZ.

A study by Gidal et al13 using pooled data from four clinical trials showed that LEV did not cause weight gain. It is not entirely clear why this finding is not replicated in our study although patients in the Gidel et al study were monitored regularly as part of clinical trials as opposed to being followed-up in the community, and were already established on AEDs before starting LEV (76% of our patients were on LEV monotherapy).

Several mechanisms may explain how VPA causes weight gain, these include dysregulation of hypothalamic weight regulation mechanisms; altered adipokine (adipose tissue cytokine) transmission and altered insulin secretion/resistance.6 LEV has a different mode of action to that of VPA—its main molecular target is synaptic vesicle protein 2A (SV2A)—a protein which plays an important role in synaptic vesicle function. Although the exact function of SV2A is still largely unknown,14 it is expressed in insulin-secreting cells in rats and seems to have a role in modifying insulin secretion in response to glucose.15

TPM is not normally used as a first-line treatment for epilepsy—it is therefore less commonly prescribed than other AEDs, explaining the fewer patients in the TPM group. The TPM group contains the highest proportion of women—it may be that the fact that it is known to cause weight loss influences the number of women who chose to take this AED.

Epilepsy is considered to be slightly more common in males than females.16 The larger number of female patients in this study may be explained by the fact that women are more likely to visit their general practitioner,17 and may be more likely to consult about weight changes than men. VPA and CBZ are the oldest drugs in the group reflected in the slightly older mean age of the patients taking these drugs.

In conclusion, when patients with epilepsy are observed in a primary care setting, certain antiepileptic drugs are associated with weight change. LEV and VPA are associated with significant weight gain, and TPM with significant weight loss. CBZ and LTG are not associated with significant weight gain. We believe that this is the first time that LEV has been associated with weight gain in a cohort of patients. Patients and clinicians need to be aware of the potential for weight change when starting AED therapy.


This study makes use of anonymised data held in the Secure Anonymised Information Linkage (SAIL) system, which is part of the national e-health records research infrastructure for Wales. We would like to acknowledge all the data providers who make anonymised data available for research. (HIRU project number 0193). The Wales Epilepsy Research Network (WERN) provided the infrastructure and resources for this project. Thanks to C H Roberts for proofreading the text.


  • Contributors WOP designed the study, analysed the data and wrote the first draft of the paper. ASL wrote and executed the database queries and assisted in study design and data analysis. RHT assisted with study design and edited the paper. PEMS and MIR supervised the project and edited the paper.

  • Funding WERN is funded by the Welsh Government through the National Institute for Social Care and Health Research (NISCHR).

  • Competing interests WOP receives an unrestricted grant from UCB Pharma for his clinical research fellow salary. RHT has previously received an unrestricted grant from UCB Pharma for a clinical research fellow salary. PEMS and MIR obtained the above unrestricted UCB grant for clinical fellows salary through WERN.

  • Ethics approval This project was approved by the Health Informatic Research Unit (HIRU) at Swansea University information governance panel (project 0193). The National Research Ethics Service (NRES) has confirmed that projects using the anonymised data held by HIRU do not need specific NHS research ethics committee approval.

  • Provenance and peer review Not commissioned; externally peer reviewed.


Relevant Article

Visit the full archive of podcasts for JNNP here >>

Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of JNNP.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.