Eleven studies were included: 9 retrospective clinical series (n=281), 1 prospective uncontrolled trial (n=51), and 1 prospective study of consecutive patients (n=150).
The percentage of children who became seizure free ranged from 7 to 33% (9 studies). The pooled percentage was 15.3% (95% confidence interval, CI: 11.0, 21.7).
The percentage of children with a greater than 90% reduction in seizures ranged from 22 to 56% (6 studies).The pooled percentage was 32.2% (95% CI: 25.3, 39.8).
The percentage of children with a greater than 50% reduction in seizures ranged from 29 to 100% (9 studies). The pooled percentage was 55.8% (95% CI: 41.2, 69.7).
A sensitivity analysis limited to prospective studies showed a slight reduction in the percentage response rates. This result was also found in a separate analysis limited to high-quality studies.
There was inconsistent reporting of compliance with the diet and adverse effects. The range of reported non-compliance was 3 to 32%. The adverse effects reported included mild gastrointestinal symptoms, which occurred in one-third to one-half of the treated children, and kidney stones and metabolic abnormalities, which occurred in less than 5% of the treated children.