Many centers still admit children for several days to start the ketogenic diet. The exact incidence of adverse effects during the admission and their potential later impact on seizure reduction has not been widely studied.

We performed a retrospective study of children with intractable epilepsy electively admitted for ketogenic diet initiation at our institution from 2011 to 2016. Charts were reviewed for adverse effects during the admission period and then examined for seizure reduction and compliance at three months. A rating scale (1 to 4) was created for severity of any adverse events.

A total of 158 children were included, with the mean age 4.6 years. Potentially attributable adverse effects occurred in 126 (80%) children, most commonly emesis, food refusal, and hypoglycemia. Seventy-three (46%) children received some form of intervention by the medical team, most commonly the administration of juice (24%). Younger age was correlated with an increased likelihood of moderate to severe adverse effects during admission, often repeated hypoglycemia (3.6 versus 4.9 years, P = 0.04). Fasting was more likely to result in lethargy and a single blood glucose in the 30 to 40 mg/dL range, but it was not correlated with emesis, repeated hypoglycemia, or higher adverse effect scores. There was no statistically significant correlation between the severity of adverse effects and the three-month seizure reduction.

Mild easily treated adverse effects occurred in most children admitted for the ketogenic diet. Younger children were at greater risk for significant difficulties and should be monitored closely. Because fasting led to more lethargy and hypoglycemia, it may be prudent to avoid this in younger children.