Skip to main content

Table 1 Comparative features of First Australians (FA) and other Australians (OA) with and without an Eating Disorder (ED)

From: Prevalence of DSM-5 diagnostic threshold eating disorders and features amongst Aboriginal and Torres Strait islander peoples (First Australians)

  FA with ED (A) FA without ED (B) OA with ED (C) OA without ED (D) Statistic P Post-hocI
N 25 68 1071 4889    
Mean (SE) ANOVA F (df)
Age (years) 36.49 (3.59) 37.54 (3.26) 39.98 (2.03) 48.48 (2.42) 66.50 (36048) <.0001 A,B, C ≠ D
BMI (kg/m2) 29.10 (1.43) 27.45 (0.91) 28.52 (0.20) 26.66 (0.19) 32.28 (36048) <.0001 D ≠ C
Physical HRQoL 45.64 (3.15) 47.50 (2.13) 48.94 (0.57) 48.81 (0.60) 1.25 (36048) .29 n.a.
Mental HRQoL 48.47 (2.37) 49.44 (2.62) 50.06 (0.54) 52.87 (0.24) 34.29 (36048) <.0001 C ≠ D
Median (IQR) Moody’s Median (df)
Binge eatingIII 2.46 (1–3.15) 1.00 (1–1) 2.58 (2–3.22) 0.93 (0.89–0.98) 2119.71 (3) <.0001 A ≠ B,D B ≠ C, C ≠ D
Over-valuationIII 4.43 (2.44–5.39) 3.72 (1.61–4.67) 3.39 (2–4.56) 2.76 (0.64–3.87) 232.75 (3) <.0001 A,B,C ≠ D
n (%) χ2 F (df1,df2)
Purging 1 (4.00%) 0 (0%)) 51 (4.76%) 0 (0%) 3.05 (3,11) 0.074 n.a.
Diet/fasting 6 (24%) 0 (0%) 265 (24.74%) 20 (.40%) 2.17 (3,11) 0.150 n.a.
  1. IThe post-hoc test used for pairwise comparison of means was the Tukey-Cramer statistic and Bonferroni-corrected pairwise Median statistic was used post-hoc Moody’s median test. IIBinge eating 1 = none, 2 = <weekly, 3 = weekly, 4 = > weekly episodes; IIIOvervaluation 0 being not at all important and 6 being extremely or the most important issue; n.a. indicates not applicable as Rao-Scott chi-squared test cannot be performed because at least one cell has zero frequency. ED Eating Disorder, BMI Body Mass Index, HRQoL Health Related Quality of Life, df degrees of freedom, χ2 Chi Squared test, n number of participants