Psychometric evaluation of the DAILY EATS questionnaire in individuals living with obesity

Journal of Patient-Reported Outcomes

Table 6 Range of Potential Meaningful Within-Person Change Thresholds Characterizing Improvement on DAILY EATS item and EDI Scores (Responder Definitions)

DAILY EATS/Anchor	Thresholds characterizing improvement
DAILY EATS/Anchor	Study 1 Obesity without diabetes	Study 2 Obesity with T2DM
Item 2. Worst hunger
PGIS-Hunger (1-point): mean (median), n	− 1.9 (− 1.6), 21	− 1.6 (− 1.5), 52
PGIC-Hunger (“Moderately”): mean (median), n	−1.0 (− 1.4), 17	− 1.1 (− 0.9), 29
Half-SD (SEM ^a)	−0.87 (− 0.80)	− 0.93 (− 1.00)
Item 3. Appetite
PGIS-Appetite (1-point): mean (median), n	−1.9 (− 1.8), 28	− 1.5 (− 1.3), 66
PGIC-Hunger ^b (“Moderately”): Mean (median), n	−1.0 (− 1.3), 17	− 1.1 (− 1.4), 29
Half-SD (SEM ^a)	−0.82 (− 0.80)	− 0.88 (− 1.12)
Item 4. Cravings
PGIS-Cravings (1-point): mean (median), n	−2.0 (−1.6), 26	−1.2 (− 1.1), 44
PGIC-Cravings (“Moderately”): mean (median), n	−0.9 (−1.3), 15	−1.2 (− 1.1), 39
Half-SD (SEM ^a)	−0.97 (− 0.78)	−1.05 (− 1.11)
EDI
PGIS-Hunger (1-point): Mean (median), n	−2.1 (− 2.1), 21	−1.5 (− 1.6), 52
PGIC-Hunger ^b (“Moderately”): Mean (median), n	−1.3 (− 1.6), 17	− 1.2 (− 1.3), 29
Half-SD (SEM ^a)	−0.81 (− 0.56)	− 0.88 (− 0.88)

Study 1 uses change from baseline to Week 26/EOT; Study 2 uses change from baseline to Week 12/EOT
The sample sizes for deterioration (i.e., by 1-point in PGIS or moderately worsen in PGIC) were too small for reliable estimates of thresholds (generally, n ≤ 5)
EDI Eating Drivers Index, EOT End of treatment, PGIC Patient Global Impression of Change, PGIS Patient Global Impression of Status, SD Standard deviation at baseline, SEM Standard error of measurement, T2DM Type 2 diabetes mellitus
^aComputed using SD at baseline and the intraclass correlation coefficients in 0
^bPGIC-Hunger was selected as the anchor for this concept a priori
Bolded values indicate measures with corresponding concepts. Green highlight is indicative of patterns that support the construct validity hypotheses and blue highlight is indicative of patterns that were not supportive
BMI Body mass index, EDI Eating Drivers Index, EOT End of treatment, EWM Ease of weight management, IWQOL-Lite 31-item Impact of Weight on Quality of Life, PGIS Patient Global Impression of Status, PGIS-PF PGIS-Physical Functioning, PROMIS PF SF Patient-Reported Outcomes Measurement Information System Physical Function Short Form, T2DM Type 2 diabetes mellitus
^aOnly Study 1 includes EWM, PROMIS PF SF 8b, and Week 26/EOT
^bStudy 2 uses PROMIS PF SF 10a and Week 12/EOT
*P < 0.01 for H0: ρ = 0