ABSTRACT Reliable dietary assessment methods underpin the confidence in reported dietary outcome measures including a priori dietary pattern indexes such as the dietary inflammatory index (DII) score. The food frequency questionnaire (FFQ) is commonly used to gather dietary data from which the DII is calculated. The FFQ, however, requires recall of dietary intake over several months. This study aimed to (i) determine the comparability of DII scores calculated from multiple 24 h recalls versus a FFQ, and (ii) identify the number of 24 h recalls required for a comparable DII score. Dietary data were collected from n = 94 community‐dwelling older adults in Australia (73.1 ± 4.8 years, 70.2% female) by an accredited practicing dietitian using four 24 h recalls over a two‐week period and a FFQ. Convergent validity was assessed between each possible comparative product of calculated DII scores by Pearson correlation, paired t ‐test, absolute difference, and Bland Altman analysis. Most participants had a healthy body weight for their age (average BMI = 25.9 ± 4.0 kg/m 2 ), were physically active ( n = 91.5%), and were highly educated (Vocational education = 9.6%, Tertiary education = 64.9%). There were positive correlations between the FFQ and one ( r = 0.219, p = 0.034), two ( r = 0.205, p = 0.072), three ( r = 0.334, p = 0.003), and four ( r = 0.444, p ≤ 0.001) 24 h recalls. Bland–Altman plots demonstrated that four 24 h recalls exhibited the closest alignment with FFQ derived DII scores. As the number of recalls increased, the DII scores became more comparable to those from the FFQ. In community‐dwelling older adults, utilizing a minimum of four 24 h recalls to calculate DII scores are comparable to scores calculated from a FFQ.
Linton et al. (Thu,) studied this question.