A modified ketogenic diet combined with rehabilitation training produced greater improvements in autism-related behaviors than rehabilitation alone in a two-month study of 62 children. Research shows children on the special high-fat diet demonstrated larger reductions in autism behavior scores on both the Autism Behavior Checklist and Childhood Autism Rating Scale compared to children receiving standard therapy only.
Researchers tested whether a modified ketogenic diet, a high-fat eating plan, could help children with autism spectrum disorder improve their behavior and social skills. Sixty-two young children in China participated in a two-month study where half received the special diet along with therapy, while the other half received therapy alone. According to Gram Research analysis, children who followed the modified ketogenic diet showed greater improvements in autism-related behaviors compared to those who only received standard therapy. Both groups improved, but the diet group improved more, suggesting this eating approach combined with rehabilitation training may offer additional benefits for some children with autism.
Key Statistics
A 2026 randomized controlled trial of 62 children with autism found that a modified ketogenic diet combined with rehabilitation training produced significantly greater reductions in autism behavior scores than rehabilitation alone, with improvements measured on the Autism Behavior Checklist (P = 0.027) and Childhood Autism Rating Scale (P = 0.003).
In a two-month study published in BMC Pediatrics in 2026, children with autism who received a modified ketogenic diet alongside therapy showed larger behavioral improvements than children who received therapy with a regular diet, with statistical significance indicating the diet group’s extra benefit was not due to chance.
A 2026 study of young children with autism spectrum disorder found that both a modified ketogenic diet group and a control group improved over two months, but the diet group demonstrated notably greater reductions in autism-related symptoms as measured by validated assessment scales.
The Quick Take
- What they studied: Whether a special high-fat diet called a modified ketogenic diet could help children with autism improve their behavior, communication, and social skills when combined with therapy.
- Who participated: Sixty-two young children diagnosed with autism spectrum disorder in China, ages not specified in the abstract. Half received the special diet plus therapy, and half received regular food plus the same therapy.
- Key finding: After two months, children on the modified ketogenic diet showed bigger improvements in autism behavior scores than children who only received therapy. The diet group’s improvement was statistically significant and notably larger than the control group’s improvement.
- What it means for you: This research suggests a modified ketogenic diet combined with therapy may help some children with autism improve faster than therapy alone. However, this is early-stage research from one study, so families should discuss this approach with their child’s doctor before making dietary changes.
The Research Details
This was a randomized controlled trial, which is considered one of the strongest types of research studies. Researchers took 62 children with autism and randomly divided them into two groups of 31 each, like flipping a coin to decide who got what treatment. One group received a modified ketogenic diet (a high-fat, low-carbohydrate eating plan delivered as a special powder and structured meal plan) plus rehabilitation training. The other group received regular food plus the same rehabilitation training. Both groups attended therapy at the hospital on weekdays and did exercises at home on weekends for two months.
The modified ketogenic diet was not a strict home-managed plan. Instead, the hospital provided lunch on training days, while families managed breakfast, dinner, and weekend meals at home. This mixed approach made the diet more practical for families to follow. Researchers measured changes in autism-related behaviors using two standard assessment tools: the Autism Behavior Checklist (ABC) and the Childhood Autism Rating Scale (CARS). These tools measure things like social interaction, communication, and repetitive behaviors.
Using a randomized controlled trial design is important because it helps prove whether the diet itself caused the improvements, not just the therapy or other factors. By randomly assigning children to groups and having one group receive the diet while the other didn’t, researchers could compare the two groups fairly. The fact that both groups received the same therapy means any extra improvement in the diet group likely came from the diet, not from better therapy. This design is much stronger than simply watching children who chose to eat this way.
This study has several strengths: it used random assignment, had a control group for comparison, and measured outcomes using validated assessment tools used worldwide. However, the study lasted only two months, which is a short time to judge long-term effects. The sample size of 62 children is moderate but reasonable for this type of research. The study was conducted in one hospital in China, so results may not apply equally to children in other countries with different genetics or healthcare systems. The abstract does not provide detailed information about whether families and researchers knew which group received the diet, which could affect results if expectations influenced outcomes.
What the Results Show
Both groups of children showed significant improvements in autism-related behaviors after two months, meaning both the diet group and the regular food group got better. This improvement happened because both groups received rehabilitation training, which is known to help children with autism. However, the children who received the modified ketogenic diet showed larger improvements than those who only received therapy.
On the Autism Behavior Checklist, the diet group improved more than the control group (the difference was statistically significant with P = 0.027, meaning there’s only a 2.7% chance this difference happened by random chance). On the Childhood Autism Rating Scale, the diet group also showed greater improvement (P = 0.003, meaning less than a 0.3% chance this was random). These statistical measures indicate the diet group’s extra improvement was real and not due to chance.
The researchers concluded that combining a modified ketogenic diet with rehabilitation training produced better short-term results than rehabilitation training alone for this group of young children with autism. The diet approach was also feasible, meaning families could actually follow it without major problems.
The study did not report detailed breakdowns of which specific autism symptoms improved most (such as social skills versus repetitive behaviors). The researchers noted that the intervention was feasible in the short term, suggesting families could manage the diet plan with hospital support, though long-term adherence was not measured. The study did not examine whether certain children benefited more than others based on age, severity of autism, or other characteristics.
This research adds to growing interest in using metabolic interventions, particularly ketogenic diets, for autism. Previous research has suggested that diet may influence brain function and behavior in children with autism, but high-quality studies have been limited. This randomized controlled trial provides stronger evidence than observational studies that simply watched families who chose this diet. However, most previous research on ketogenic diets for autism has been small or involved case reports rather than controlled trials, so this study represents an important step forward in understanding whether this approach truly works.
The study lasted only two months, so we don’t know if improvements continue, stay the same, or fade over longer periods. The sample size of 62 children is moderate, and all participants were from one hospital in China, so results may not apply to children in other countries or healthcare settings. The study did not clearly describe whether researchers and families knew which group received the diet, which could influence results if expectations affected outcomes. The study did not measure whether children maintained the diet after the study ended or whether improvements lasted. Additionally, the abstract does not provide information about potential side effects, costs, or whether all families could realistically follow this diet plan at home.
The Bottom Line
Based on this single randomized controlled trial, a modified ketogenic diet combined with rehabilitation training appears to produce greater short-term improvements in autism-related behaviors than rehabilitation alone. However, this is early-stage evidence from one study in one location. Families interested in trying this approach should discuss it with their child’s doctor, neurologist, or autism specialist before making dietary changes. The diet should not replace standard therapies like behavioral training, speech therapy, or occupational therapy. More research is needed to understand long-term effects, which children benefit most, and whether improvements continue beyond two months.
Parents and caregivers of children with autism who are looking for additional approaches to support their child’s development may find this research interesting. Healthcare providers treating children with autism should be aware of this emerging evidence. However, this research is too early-stage to recommend the diet as a standard treatment. Children with certain medical conditions, metabolic disorders, or those taking specific medications may not be appropriate candidates for a ketogenic diet and should not attempt this without medical supervision.
This study measured improvements over two months, so that’s the timeframe for which we have evidence. Children showed measurable behavioral improvements within this short period. However, we don’t know if improvements continue, plateau, or decrease after two months. Families should expect to work with healthcare providers to monitor progress over time and adjust the approach as needed.
Frequently Asked Questions
Can a ketogenic diet help children with autism improve their behavior?
A 2026 study of 62 children found that a modified ketogenic diet combined with therapy produced greater behavioral improvements than therapy alone over two months. However, this is early research from one study, so families should consult their child’s doctor before trying this approach.
How long does it take to see improvements from a ketogenic diet for autism?
This study measured improvements over two months and found measurable behavioral changes within that timeframe. However, we don’t know if improvements continue longer or how sustainable they are beyond two months based on this research.
Is a ketogenic diet safe for children with autism?
This study reported the diet was feasible to follow, but the abstract doesn’t detail safety information or side effects. Any child considering a ketogenic diet should be medically supervised by their doctor, as this diet requires careful planning and monitoring.
Should I replace my child’s autism therapy with a ketogenic diet?
No. This research shows the diet worked best combined with rehabilitation training, not instead of it. Standard therapies like behavioral training and speech therapy should continue. The diet may be an additional tool to discuss with your child’s healthcare team.
What is a modified ketogenic diet for autism?
In this study, it was a high-fat, low-carbohydrate eating plan delivered as special nutritional powder and structured meals. The hospital provided lunch on training days while families managed other meals at home, making it more practical than a strict home-managed diet.
Want to Apply This Research?
- Track daily autism-related behaviors using a simple scoring system similar to the Autism Behavior Checklist, noting improvements in social interaction, communication attempts, and repetitive behaviors. Record this weekly to see trends over time.
- If a family and their doctor decide to try a modified ketogenic diet, use the app to log meals, track which foods the child eats, and monitor behavioral changes daily. This creates a record to share with healthcare providers and helps identify which foods may affect behavior.
- Set monthly check-ins to review behavior scores, diet adherence, and any side effects. Compare current scores to baseline to measure progress. Share this data with the child’s healthcare team to make informed decisions about continuing, adjusting, or stopping the diet.
This research describes a single two-month study and should not be considered definitive medical advice. A ketogenic diet is a significant dietary change that requires medical supervision, especially for children. Parents and caregivers should consult with their child’s pediatrician, neurologist, or autism specialist before making any dietary changes. This diet may not be appropriate for all children and could interact with medications or existing medical conditions. This information is for educational purposes and does not replace professional medical diagnosis, treatment, or advice. Long-term safety and effectiveness of ketogenic diets for autism remain unclear and require further research.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.