Confirmatory study on the retention rate of ketogenic diet therapy and related factors in children with drug-resistant epilepsy: A single-center cohort study.

OBJECTIVE

To confirm the retention rate of ketogenic diet therapy (KDT) in children with drug-resistant epilepsy and identify its related factors through a cohort analysis.

METHODS

This was a single-center cohort study. Baseline data were collected, and regular follow-up was conducted after KDT treatment. The retention rate of KDT at different time points was determined. Cox regression analyses and other statistical methods were used to identify the factors related to the retention rate.

RESULTS

A total of 337 patients were included. 43.3% (146/337) of the patients had only spasms, 16.9% (57/337) had a genetic etiology, and 14.0% (47/337) had a structural etiology. Overall, 51.9% (175/337) of the patients were classified as having infantile epileptic spasm syndrome. The median age at KDT initiation was 33 months. 38.3% (129/337) of the patients responded to KDT. The retention rate of KDT at 3, 6, 12 and 24 months were 86.6% (292/337), 62.6% (211/337), 35.0% (118/337), and 20.2% (68/337), respectively. 9.9% (56/282) stopped the treatment due to unsatisfactory seizure control and 54.6% (154/282) stopped due to a non-response. The retention rate in patients who responded to KDT was relatively greater than that in patients who did not respond to KDT [HR: 4.714 (95% CI: 3.493, 6.362), p < 0.001]. The retention rate of KDT at 3 months was associated with seizure type and anti-seizure efficacy. The retention rate of KDT at 6 months was associated with seizure type, status epilepticus and anti-seizure efficacy. The retention rate of KDT at 12 and 24 months were both associated with anti-seizure efficacy.

SIGNIFICANCE

Anti-seizure efficacy was the sole independent predictor of long-term retention. This large-scale cohort study advances the field by demonstrating the predominant role of efficacy in long-term adherence. The cohort size (N = 337), one of the largest reported, strengthens evidence that the long-term retention rate (>1 year) can be used as an important indicator to evaluate the efficacy of KDT, which is simple and reliable.