Modified Atkins diet versus levetiracetam for non-surgical drug-resistant epilepsy in children: A randomized open-label study

•We compared efficacy of modified Atkins diet (MAD) among children with non-surgical drug-resistant epilepsy (DRE) to levetiracetam, when added to on-going anti-seizure medications.•Proportion of children with >50% seizure reduction at 12 weeks was significantly higher with MAD arm versus levetiracetam (27/51(52.9%) vs 11/50(22%); p < 0.001).•Constipation (41.1%) was the most frequent adverse effect with MAD and sedation/lethargy (18%) and anxiety and irritability (14%) with levetiracetam.•Addition of MAD was found to be superior to levetiracetam among children with non-surgical DRE in achieving seizure reduction at 12 weeks.•Both treatments were well tolerated though adverse effects were higher with MAD. This study was undertaken to compare the efficacy of modified Atkins diet (MAD) among children with non-surgical drug-resistant epilepsy (DRE) to levetiracetam, when added to on-going anti-seizure medications. An open-label, randomized controlled trial among children aged 2–12 years with non-surgical DRE was conducted. Eligible children were randomized in a 1:1 ratio to receive add-on MAD or levetiracetam. Baseline and post-intervention seizure frequency at 12 weeks was determined from seizure logs maintained by parents. The primary outcome was the proportion of responders, i.e., patients who achieved > 50% seizure reduction from baseline. Adverse events were compared. Analysis was intention-to-treat. (NCT04172311) One hundred and one children were enrolled (MAD-51, levetiracetam-50). The majority of the enrolled children had generalized seizures of mixed types secondary to structural brain lesions and Lennox-Gastaut syndrome was the most common electroclinical syndrome (46%). The proportion of children with >50% seizure reduction at 12 weeks was significantly higher in the MAD arm compared to the levetiracetam arm (27/51(52.9%) vs 11/50(22%); p < 0.001). At 12-weeks post-intervention, the change in mean seizure frequency compared to baseline was -47.33 ± 39.57% in the MAD arm and -31.15 ± 32.18% in the levetiracetam arm (p = 0.03). Constipation (41.1%) was the most frequent adverse effect with MAD. Sedation/lethargy (18%) and anxiety and irritability (14%) were the most frequent adverse effects in the levetiracetam group. Addition of MAD was found to be superior to levetiracetam among children with non-surgical DRE with predominant generalized seizures in achieving seizure reduction at 12 weeks. Both treatments were well tolerated. Adverse effects, altho