SUMMARY
Type one diabetes mellitus (T1D) is a debilitating disease with no cure. After an initial partial remission with
improved residual β-cell function, the “honeymoon period”, less than 17% of children achieve the glycemic
targets recommended by the American Diabetes Association (ADA), placing millions at risk for complications
and early mortality. Conversely, tight glycemic control - as achieved with greater insulin doses to match
carbohydrates consumed - has also been linked to complications, namely weight-gain, insulin resistance, and
metabolic syndrome. Preliminary evidence suggests that a very low-carbohydrate, high-fat, (i.e., ketogenic diet,
KD) in T1D may (1) improve glycemic control by mitigating postprandial glycemic excursions and (2) reduce
insulin exposure and associated adverse effects on peripheral tissues. Children with incident T1D may
experience additional benefits, as a KD may also (3) prolong the honeymoon period - as seen in case reports -
via immune and/or metabolic effects on β-cell function. Specifically, improved glycemia and insulinemia may
promote β-cell rest, and the physiologically elevated β-hydroxybutyrate (βOHB) levels on a KD have been linked
to decreased inflammation and gut microbiome changes that may reduce ß-cell autoimmunity.
We propose to test the hypothesis that a KD vs. standard diet (SD) will prolong diabetes remission and improve
diabetes control in children with incident T1D. In a 9-months parallel randomized controlled trial (RCT), fifty-two
children (26 per arm) aged 5-12 years will receive a family-based intervention with food deliveries and intensive
nutrition and diabetes education to promote a KD vs SD. Continuous glucose monitoring (CGM) and Bluetooth
enabled insulin pens will be used for cloud-based data collection. Anthropometrics, fasting biomarkers and
stimulated C-peptide area under the curve following a mixed meal tolerance test will will be assessed at baseline,
1, 5 and 9 months. The primary endpoint will be percent change in stimulated C-peptide between 1 and 9 months.
Secondary endpoints will include percent children with clinical diabetes remission (insulin dose adjusted HbA1c
[IADD1c] <9) at 9 months, indices of glycemic control from continuous glucose monitoring, and markers of
metabolic health (BMI, indices of insulin resistance, and the ratio of triglycerides to HDL cholesterol). To explore
pathways related to improved ß-cell function, we will also evaluate gut microbiome, metabolome, and
immunologic biomarker responses to a KD vs SD and test interactions of targeted biomarker groups with
changes in β-cell function and glycemia.
Compared with technological and pharmacological treatments, dietary intervention is inexpensive, relatively free
of major side-effects and directly translatable. A KD may have benefits on ß-cell function, glycemia and
insulinemia, and would thereby provide a major therapeutic advance for children living with T1D. Regardless of
outcome, our research will close an important knowledge gap on the safety and efficacy of a KD for children with
T1D, an approach with increasing patient popularity despite lack of high-quality research.
