Dietary Assessment by Pattern Recognition: a Comparative Analysis

Area of Science:

• Nutrition science
• Dietary assessment methodology
• Health informatics

Background:

• Traditional dietary assessment tools face limitations in healthcare integration.
• Diet quality photo navigation (DQPN) was developed to overcome these limitations.
• Validating new tools against established methods is standard practice.

Purpose of the Study:

• To assess the validity of Diet ID (DQPN) against food records (FR) and food frequency questionnaires (FFQ) for diet quality, food groups, and nutrient intake.
• To evaluate the test-reproducibility and reliability of Diet ID (DQPN).

Main Methods:

• Recruited 90 participants via an online platform; 58 completed assessments.
• Participants completed DQPN, a 3-day FR (Automated Self-Administered 24-hour Dietary Assessment Tool), and an FFQ (Dietary History Questionnaire III).
• Estimated mean nutrient and food group intake, and calculated Pearson correlations between instruments.

Main Results:

• DQPN showed strong correlations for diet quality with FFQ (0.58) and FR (0.56).
• Moderate correlations were observed for selected nutrients and food groups.
• Test-retest reproducibility for DQPN diet quality assessment yielded a correlation of 0.70.

Conclusions:

• DQPN demonstrates comparable performance to traditional tools for estimating overall diet quality.
• The tool's ease-of-use and scalability support its potential role in universal clinical dietary assessment.
• DQPN may enhance the integration of dietary assessment into routine healthcare.