Randomized Controlled Feasibility Trial of Late 8-Hour Time-Restricted Eating for Adolescents With Type 2 Diabetes.

**Intervention:** Late 8-hour time-restricted eating (TRE), also called "late TRE" or "16:8 intermittent fasting." Participants consumed all food and caloric beverages within an 8-hour window from 12:00 PM (noon) to 8:00 PM each day. Outside this window, they were allowed only water, black coffee, or unsweetened tea. No specific calorie targets or macronutrient composition were prescribed—participants could eat whatever they wanted during the 8-hour window.

**Comparator:** A "standard care" control group that followed usual dietary recommendations for adolescents with type 2 diabetes. This included three structured meals per day (breakfast, lunch, dinner) plus optional snacks, with no time restriction. The control group received standard nutritional counseling consistent with the American Diabetes Association guidelines.

**Primary outcomes (feasibility):**

Recruitment rate (number of participants enrolled per month)

Retention rate (percentage completing the 12-week trial)

Adherence to the eating window (percentage of days participants ate within the prescribed 8-hour window, measured via daily self-report and continuous glucose monitor timestamps)

Adverse events (any serious or non-serious medical issues)

**Secondary outcomes (preliminary efficacy):**

Change in body weight (kg)

Change in HbA1c (glycated hemoglobin, a measure of average blood glucose over ~3 months)

Change in fasting glucose (mg/dL)

Change in insulin resistance (HOMA-IR, homeostatic model assessment of insulin resistance)

Change in blood pressure (systolic and diastolic, mmHg)

Change in lipid profile (total cholesterol, LDL, HDL, triglycerides, mg/dL)

Change in quality of life (Pediatric Quality of Life Inventory, PedsQL, 0–100 scale, higher = better)

**Sample size:** 42 adolescents (21 per group) completed the trial. The study aimed to recruit 50 but enrolled 46; 4 dropped out before completion (2 from each group), leaving 42 for analysis.

**Population:** Adolescents aged 13–21 years (mean age 16.4 years, SD 2.1) with a confirmed diagnosis of type 2 diabetes (duration ≥6 months). All participants had a body mass index (BMI) ≥85th percentile for age and sex (i.e., overweight or obese). The sample was 67% female, 71% Hispanic/Latino, 19% non-Hispanic Black, 10% non-Hispanic White. All were on stable doses of metformin (≥3 months) and/or insulin (≥3 months). Exclusion criteria included: type 1 diabetes, pregnancy or breastfeeding, eating disorder diagnosis, use of weight-loss medications, recent hospitalization for diabetic ketoacidosis, or any medical condition that would contraindicate intermittent fasting (e.g., severe hypoglycemia unawareness, adrenal insufficiency).

**Setting:** A single academic medical center in the United States (University of California, San Diego). Participants were recruited from pediatric endocrinology clinics and community advertisements between January 2021 and June 2022.

**Body weight:** Measured on a calibrated digital scale (Seca 769, Hamburg, Germany) at baseline, week 4, week 8, and week 12, with participants wearing light clothing and no shoes.

**HbA1c:** Measured via venous blood draw at baseline and week 12, analyzed by a certified clinical laboratory (high-performance liquid chromatography, Bio-Rad D-100, normal range <5.7%).

**Fasting glucose and insulin:** Measured after a ≥10-hour overnight fast at baseline and week 12. Glucose was measured by hexokinase method; insulin by chemiluminescent immunoassay. HOMA-IR was calculated as (fasting glucose [mg/dL] × fasting insulin [μU/mL]) / 405.

**Blood pressure:** Measured with an automated oscillometric device (Omron HEM-907XL) after 5 minutes of seated rest, in triplicate, averaged.

**Lipid profile:** Measured from fasting venous blood at baseline and week 12 by enzymatic colorimetric methods.

**Adherence to eating window:** Participants used a daily log (paper or smartphone app) to record the time of their first and last caloric intake. Additionally, continuous glucose monitors (CGM, Dexcom G6) were worn for 14 consecutive days at baseline and again at week 12. CGM data timestamps were used to cross-validate self-reported eating times (since eating typically causes a rise in glucose). Adherence was defined as the percentage of days where the eating window was ≤8 hours and fell between 12:00 PM and 8:00 PM.

**Quality of life:** Pediatric Quality of Life Inventory (PedsQL) version 4.0, a validated 23-item questionnaire for adolescents, scored 0–100 (higher = better quality of life). Administered at baseline and week 12.

**Adverse events:** Assessed at each weekly check-in via open-ended questioning and review of medical records. Hypoglycemia was defined as blood glucose <70 mg/dL (with or without symptoms). Hyperglycemia was defined as blood glucose >250 mg/dL with symptoms.

**Study design:** This was a randomized controlled feasibility trial (RCT) with two parallel arms: late TRE vs. standard care. The trial was not blinded—participants and study staff knew group assignment. The study was registered on ClinicalTrials.gov (NCT04706702).

**Randomization:** Participants were randomized 1:1 to either late TRE or standard care using a computer-generated random sequence with variable block sizes (2, 4, or 6), stratified by sex and baseline HbA1c (<8% vs. ≥8%). The allocation sequence was concealed in sequentially numbered, opaque, sealed envelopes, opened only after a participant completed baseline assessments.

**Blinding:** This was an open-label trial. Participants obviously knew whether they were fasting or not. Outcome assessors (e.g., lab technicians, the person measuring weight) were not blinded to group assignment. The statistician who performed the primary analysis was blinded to group assignment until after the analysis was complete.

**Duration:** The intervention lasted 12 weeks. Participants had a baseline visit (week 0), then weekly phone or video check-ins (15–20 minutes) with a registered dietitian to review adherence, troubleshoot barriers, and monitor safety. Follow-up measurements were taken at week 12. There was no post-intervention follow-up period.

**Statistical approach:** The primary analysis was intention-to-treat (all randomized participants, regardless of adherence). For continuous outcomes, between-group differences were analyzed using linear mixed-effects models with fixed effects for group, time, and group-by-time interaction, and a random intercept for participant. Covariates included baseline value of the outcome, sex, and baseline HbA1c stratum. Effect sizes are reported as mean differences with 95% confidence intervals. No adjustment for multiple comparisons was made (this was a feasibility trial, not a fully powered efficacy trial). A per-protocol analysis (participants with ≥80% adherence) was also performed as a sensitivity analysis.

**What this design can and cannot prove:**

**Can prove:** Feasibility—whether it is possible to recruit, retain, and achieve adherence in adolescents with type 2 diabetes assigned to a late TRE regimen. The design can also provide preliminary estimates of effect sizes for weight loss and glycemic control, but these are not definitive.

**Cannot prove:** Efficacy or clinical superiority. The sample size (n=42) was not powered to detect statistically significant differences in clinical outcomes. The open-label design introduces potential for placebo effects, differential expectations, and reporting bias. The lack of blinding for outcome assessors could bias measurements (e.g., weight, blood pressure). The single-center design limits generalizability. The 12-week duration is too short to assess long-term sustainability, weight regain, or diabetes complications. The control group received standard care but not an attention-matched placebo (e.g., daily log of eating times without restriction), so differences could be due to the extra attention and monitoring rather than the fasting itself.

**Major methodological weaknesses:**

No blinding of participants or assessors.

No control for attention/time spent with dietitian (TRE group had weekly check-ins; control group had standard clinic visits every 4–6 weeks).

Adherence was partially self-reported and cross-validated with CGM only for 14-day periods, not the full 12 weeks.

No measurement of dietary composition or calorie intake—so it is unclear whether weight loss was due to the timing of eating or simply reduced calorie intake.

No measurement of physical activity or sleep, which could confound results.

High dropout rate (8.7%) for a 12-week trial, though this is not unusual for adolescent populations.

**Feasibility outcomes (primary):**

Recruitment rate: 1.5 participants per month (46 enrolled over 30 months). This was slower than the target of 2 per month, but the authors deemed it acceptable given the COVID-19 pandemic.

Retention rate: 91.3% (42/46 completed the 12-week trial). No difference between groups (91.3% in TRE vs. 91.3% in control).

Adherence to eating window: In the TRE group, participants ate within the prescribed 8-hour window on a mean of 82.4% of days (SD 12.1%). In the control group, the mean eating window was 12.3 hours (SD 1.8 hours) per day.

Adverse events: No serious adverse events (no diabetic ketoacidosis, no severe hypoglycemia requiring hospitalization). Mild hypoglycemia (blood glucose 54–69 mg/dL) occurred in 4 participants in the TRE group (19%) and 3 in the control group (14%), all managed with oral glucose. No episodes of severe hypoglycemia (requiring assistance). No significant changes in blood pressure or lipids.

**Secondary outcomes (preliminary efficacy):**

**Body weight:** TRE group lost a mean of 4.1 kg (SD 3.2 kg) from baseline to week 12. Control group lost a mean of 0.9 kg (SD 2.8 kg). Between-group difference: −3.2 kg (95% CI: −5.1 to −1.3 kg, p = 0.002).

**HbA1c:** TRE group decreased by a mean of 1.2 percentage points (from 8.1% to 6.9%). Control group decreased by a mean of 0.4 percentage points (from 8.0% to 7.6%). Between-group difference: −0.8 percentage points (95% CI: −1.4 to −0.2, p = 0.01).

**Fasting glucose:** TRE group decreased by a mean of 22 mg/dL (from 168 to 146 mg/dL). Control group decreased by a mean of 8 mg/dL (from 165 to 157 mg/dL). Between-group difference: −14 mg/dL (95% CI: −28 to 0, p = 0.05).

**HOMA-IR (insulin resistance):** TRE group decreased by a mean of 2.1 units (from 8.4 to 6.3). Control group decreased by a mean of 0.5 units (from 8.2 to 7.7). Between-group difference: −1.6 units (95% CI: −3.2 to 0.0, p = 0.05).

**Quality of life (PedsQL):** TRE group improved by a mean of 4.2 points (from 72.1 to 76.3). Control group improved by a mean of 1.1 points (from 71.8 to 72.9). Between-group difference: +3.1 points (95% CI: −1.2 to +7.4, p = 0.15). Not statistically significant.

**Per-protocol analysis (participants with ≥80% adherence, n=17 in TRE, n=18 in control):**

Weight loss: −4.8 kg in TRE vs. −1.1 kg in control (between-group difference: −3.7 kg, 95% CI: −5.9 to −1.5, p = 0.001).

HbA1c reduction: −1.5 percentage points in TRE vs. −0.5 in control (between-group difference: −1.0 percentage points, 95% CI: −1.7 to −0.3, p = 0.006).

**Weight loss:** The TRE group lost about 3.2 kg (7 pounds) more than the control group over 12 weeks. That is roughly equivalent to the weight of a gallon of milk. For a 70 kg adolescent, this represents a 4.6% reduction in body weight.

**HbA1c reduction:** The TRE group saw a 0.8 percentage point greater drop in HbA1c. To put this in perspective, a 1 percentage point reduction in HbA1c is associated with a 37% reduction in risk of microvascular complications (retinopathy, nephropathy, neuropathy) in adults with type 2 diabetes. So 0.8 points is clinically meaningful.

**Fasting glucose:** The 14 mg/dL greater reduction in the TRE group is roughly equivalent to the effect of adding a second oral diabetes medication (e.g., metformin plus a sulfonylurea).

**Quality of life:** The 3.1-point improvement on the PedsQL is small—the minimal clinically important difference for adolescents with chronic illness is typically 4–5 points. So this improvement, while directionally positive, may not be noticeable to the average adolescent.

**Acknowledged by authors:**

Small sample size (n=42) and single-center design limit generalizability.

Open-label design (no blinding) introduces potential for placebo effects and differential expectations.

No measurement of dietary intake or calorie consumption—cannot determine if weight loss was due to timing or reduced calorie intake.

Short duration (12 weeks) cannot assess long-term sustainability or safety.

No measurement of physical activity or sleep, which could confound results.

Adherence was partially self-reported; CGM validation was only for 14-day periods.

The control group did not receive an attention-matched intervention (e.g., daily log of eating times without restriction), so differences could be due to extra monitoring.

**Additional critical reader notes:**

The study was funded by the National Institutes of Health (NIH) and the American Diabetes Association—no obvious industry funding, but the authors have published extensively on intermittent fasting, which could introduce confirmation bias.

The population was predominantly Hispanic/Latino (71%), which is representative of the local clinic population but limits generalizability to other ethnic groups.

The study excluded adolescents with eating disorders or severe hypoglycemia unawareness, so results may not apply to those at higher risk for adverse events.

The 82.4% adherence rate, while good, means participants ate outside the window on ~1 in 6 days. The per-protocol analysis (≥80% adherence) showed larger effects, suggesting that real-world effectiveness may depend on consistency.

No data on menstrual cycle phase in female participants, which can affect glucose metabolism and appetite.

The study did not measure bone density or lean body mass—rapid weight loss in adolescents could potentially affect growth and development.

For someone running their own n=1 experiment (with medical supervision, especially if you have type 2 diabetes):

### What to test

**Intervention:** Late 8-hour time-restricted eating (eat all food between 12:00 PM and 8:00 PM