Validation of a Ketogenic Diet Adherence Questionnaire

Introduction

The ketogenic diet (KD) has been adopted for a range of clinical and lifestyle purposes, from seizure control to weight management. Because the diets efficacy depends heavily on strict macronutrient ratios, researchers and clinicians need reliable tools to assess participants adherence. Selfreport questionnaires are attractive for largescale studies, yet their validity must be demonstrated against objective biomarkers.

Purpose of the Questionnaire

The primary aim of the Ketogenic Diet Adherence Questionnaire (KDAQ) is to provide a quick, noninvasive estimate of how closely an individual follows the core principles of a KD:

• Carbohydrate intake below 50g per day (or<10% of total energy).
• High proportion of calories from fat (70% of total energy).
• Moderate protein intake (20% of total energy).
• Consistent use of ketogenicfriendly foods and avoidance of highglycemic items.

By quantifying these components, the KDAQ can be used in epidemiologic surveys, clinical trials, and routine practice.

Development Process

Item Generation

Items were derived from dietary guidelines, expert consensus, and existing foodfrequency questionnaires. An initial pool of 25 statements was reduced to 12 after facevalidity testing with dietitians and individuals experienced with the KD.

Pretesting and Cognitive Interviews

Twenty participants completed the draft questionnaire while thinking aloud. Their feedback led to clearer wording, removal of ambiguous terms, and the addition of an item on exogenous ketone use.

Response Format

Each item uses a 5point Likert scale ranging from Never (1) to Always (5). Scores are summed, yielding a total possible range of 1260; higher scores indicate better adherence.

Validation Study Design

A multicenter validation study was conducted with 300 adults (age 1865) who had been following a KD for at least 4 weeks. The study adhered to the COSMIN guidelines for patientreported outcome measures.

Reference Standards

• Urinary ketone strips: Measured three times per week, reported as a semiquantitative score (04).
• Blood hydroxybutyrate (BHB): Fasted capillary samples collected weekly using a commercial meter.
• 24hour dietary recalls: Conducted on two nonconsecutive days and analyzed for macronutrient distribution with a validated nutrition software.

Statistical Analyses

Reliability was examined using Cronbachs alpha (internal consistency) and testretest intraclass correlation (ICC) over a 2week interval. Construct validity was evaluated through correlations with BHB levels (Pearson r), urinary ketone scores (Spearman rho), and macronutrient percentages from recalls (Spearman rho). Factor analysis (principal component with varimax rotation) tested the underlying dimensionality.

Results

Reliability

Overall Cronbachs alpha was 0.89, indicating excellent internal consistency. The ICC for the total score was 0.85 (95%CI0.800.90), supporting good testretest stability.

Factor Structure

Exploratory factor analysis revealed a single dominant factor accounting for 58% of variance. All items loaded above 0.60 on this factor, justifying the use of a totalscore metric.

Construct Validity

Significant positive correlations were observed between KDAQ scores and objective measures:

• Blood BHB: r = 0.62 (p<0.001)
• Urinary ketone score: = 0.55 (p<0.001)
• Carbohydrate intake (<50g/day): = -0.48 (p<0.001)
• Fat proportion (>65%): = 0.51 (p<0.001)

These findings confirm that higher questionnaire scores reflect greater biochemical ketosis and closer alignment with the prescribed macronutrient ratios.

Discriminant Validity

Scores did not correlate with unrelated variables such as total calorie intake (r = 0.08, p=0.32) or physical activity level ( = 0.04, p=0.71), indicating the instrument measures a specific construct.

Interpretation of Scores

Based on the distribution of scores in the validation sample, provisional cutpoints were suggested:

• 5260: Excellent adherence (consistent ketosis, carbohydrate<30g/day).
• 4051: Moderate adherence (occasional carbohydrate spikes, variable ketone levels).
• 39: Poor adherence (frequent carbohydrate intake, lack of sustained ketosis).

Future research should refine these thresholds across different KD protocols (e.g., therapeutic vs. weightloss).

Strengths and Limitations

Strengths

• Comprehensive development involving experts and target users.
• Robust validation against multiple objective biomarkers.
• Simple scoring system suitable for both research and clinical settings.

Limitations

• Validation sample was limited to adults without major metabolic disorders; external validity for pediatric or diabetic populations requires testing.
• Selfreport bias may still influence responses, especially among participants motivated to appear compliant.
• Shortterm validation (8weeks); longterm stability of the instrument remains unknown.

Practical Applications

The KDAQ can be incorporated into:

• Randomized controlled trials evaluating the therapeutic effects of ketosis.
• Clinical consultations where rapid adherence assessment is needed.
• Largescale epidemiological studies investigating correlations between KD adherence and health outcomes.

Because the questionnaire does not require laboratory testing, it reduces costs and participant burden while still offering a validated proxy for ketogenic compliance.

Future Directions

Further research should aim to:

• Validate the KDAQ in diverse populations (children, older adults, individuals with type2 diabetes).
• Assess sensitivity to change during diet transitions (e.g., initiation, reinduction after a break).
• Explore digital delivery (mobile app) with realtime scoring and feedback.
• Integrate the questionnaire with wearable ketone sensors for a multimodal adherence index.

Continued refinement will enhance the questionnaires utility for both scientific investigation and routine dietary counseling.

Conclusion

The Ketogenic Diet Adherence Questionnaire is a reliable, valid, and userfriendly instrument for quantifying compliance with a ketogenic nutritional regimen. Its strong correlations with blood and urinary ketone measures, together with excellent internal consistency, support its use as a surrogate marker of ketosis in research and clinical practice. Ongoing validation work will broaden its applicability across populations and settings.