Skip to main content

Table 3 Inter-method agreement, limits of agreement (LOA) and correlation coefficient (r) between RFD app and 24 h dietary recalls (n = 50)

From: Feasibility of a commercial smartphone application for dietary assessment in epidemiological research and comparison with 24-h dietary recalls

  Mean agreement (95% CI) LOA Correlation
Coefficient
(95% CI)
Energy intake (kJ/d) −268 (−895, 358)a −4699, 4162a 0.52 (0.28, 0.70)
Protein (% E) 0.4 (−0.4, 1.2)a −5.5, 6.3a 0.79 (0.66, 0.88)
Total fat (% E) −0.8 (−2.3, 0.7)a −11.7, 10.1a 0.63 (0.43, 0.77)
Carbohydrate (% E) −0.5 (−2.2, 1.1)a −12.1, 11.0a 0.72 (0.56, 0.83)
Saturated fat (% E) 0.5 (−0.2, 1.2)a −4.6, 5.5a 0.60 (0.39, 0.76)
Calcium density (mg/MJ) 5.9 (−4.0, 15.8)a −64.4, 76.1a 0.45 (0.20, 0.65)
Iron density (mg/MJ) 0.05 (−0.08, 0.18)a −0.89, 0.98a 0.42 (0.16, 0.62)
Polyunsaturated fat (% E) 100 (92, 108)b 57, 174b 0.64 (0.44, 0.78)c
Added sugar (% E) 110 (78, 140)b 13, 922b 0.68 (0.49, 0.80)c
Alcohol (% E) 47 (24, 92)b 0.4, 5266b 0.65 (0.46, 0.79)c
Dietary fibre density (g/MJ) 100 (93, 107)b 60, 167b 0.66 (0.46, 0.79)c
  1. aMean and LOA agreement based on raw data (absolute differences, original units): 4-d diary average (app) - average of two 24-h recalls
  2. bMean agreement and LOA based on log transformed data (relative differences, %): 4-d diary average (app): average of two 24-h recalls
  3. cSpearman’s rank correlation coefficient