Impact of ketogenic diet therapy on growth in children with epilepsy

Ketogenic diet therapy (KDT) is an effective treatment for drug-resistant epilepsy in children. There is conflicting evidence regarding the impact of KDT on growth. We aimed to determine whether linear growth and weight were affected in children who followed KDT in the UK, and to explore potential associations with clinical or demographic factors with impacted growth.

A retrospective review of medical records of children with epilepsy following KDT at 3 UK centres was conducted. Height and weight measurements taken as part of routine clinical management were recorded at baseline, 1-8 years on diet, and 1-year post-diet. Measurements were converted into z-scores, and the differences from baseline analysed using Wilcoxon Signed Rank tests. Potential associations of impacted growth with feeding method, ambulatory status, diet type, age at diet onset and average daily protein intake were investigated using Mann-Whitney, Kruskal-Wallis tests or Spearman's Rank correlation.

265 individuals were included, of which 84 had post-diet data available. Median height z-score significantly decreased at 1- (n = 139, p = .018), 2- (n = 86, p < .0005) and 3 years (n = 27, p = .001) on diet. There was no significant change to height or weight z-score 1-year post-diet discontinuation. Median weight z-score significantly decreased from baseline at 4 years (n = 15, p = .020), and 6 years (n = 8, p = .025) on diet, but not at other time points. There was greater height z-score decrease in non-ambulatory children at 2 years (p = .009), in those following a classical diet compared with the modified ketogenic diet at 2 years (p = .006) and amongst younger children at 2 years (n = 86, p < .005) and 3 years (n = 27, p = .008) on diet. No significant differences were found in weight z-score change across any subgroup, following Bonferroni correction for multiple testing.

Median linear growth was significantly adversely affected for the first 3 years on KDT but catch-up growth post diet discontinuation was observed. Non-ambulatory children, younger children, and individuals following a classical diet may be more vulnerable to impacted growth when on KDT, although this was not consistent across all time points. The potential short-term impact on linear growth should be discussed with individuals considering KDT, and monitored closely.