3.1 Demographics, diagnosis and eating disorder psychopathology
The sample consists of 361 participants of which 250 (69.25%) have an eating disorder diagnosis and 111 (30.74%) are healthy controls. Of the 250 eating disorder patients, n= 84 were diagnosed with anorexia nervosa restrictive subtype (AN-R, 23.3%), n= 57 with anorexia nervosa binge-purge subtype (AN-BP, 15.8%), n= 38 with bulimia nervosa (BN, 10.5%), n= 20 with binge eating disorder (BED, 5.5%) and n= 51 with other specified feeding and eating disorders (OSFED, 14.1%). The eating disorder diagnoses differ in age (H(5) = 37.786, p < .001), BMI (H(5) = 251.860, p < .001) and illness duration (H(5) = 20.190, p < .001). Patients with binge eating disorder have highest age, BMI and illness duration and patients with both anorexia nervosa subtypes the lowest. Patients with bulimia nervosa and other specified feeding and eating disorders take a middle position. Means, standard deviations and outcomes of the pairwise comparisons are displayed in table 1.
Table 1.
Demographic data
|
AN-R (N=84)
|
AN-BP (N=57)
|
BN
(N=38)
|
BED
(N=20)
|
OSFED (N=51)
|
HC
(N=111)
|
|
M (SD) + range or %
|
M (SD) + range or %
|
M (SD) + range or %
|
M (SD) + range or %
|
M (SD) + range or %
|
M (SD) + range or %
|
Age
|
22.36 (5.59)
16-48
|
21.14 (4.54)
16-38
|
23.74 (7.98)
17-54
|
31.45 (11.16)
19-60
|
23.98 (6.48)
16-46
|
25.77 (8.40)
17-60
|
BMI (kg/m2)
|
16.69 (2.04)
11.0-21.6
|
16.53 (1.94)
11.0-19.8
|
21.69 (2.36)
18.6-28.3
|
36.49 (7.34)
25.2-49.8
|
21.46 (3.34)
13.8-33.3
|
23.12 (3.42)
18.5-35.6
|
Currently in treatment
|
72.6%
|
57.9%
|
60.5%
|
75%
|
64.7%
|
N.A.
|
In treatment since (years)
|
3.82 (3.67)
1-26
|
4.45 (4.89)
1-25
|
3.43 (3.96)
1-23
|
4.40 (5.42)
1-21
|
4.22 (4.40)
1-20
|
N.A.
|
Illness duration (years)
|
7.17 (5.54)
1-30
|
8.10 (5.63)
1-28
|
8.72 (6.83)
1-31
|
15.25 (9.65)
1-46
|
8.40 (5.97)
1-31
|
N.A.
|
Anorexia nervosa restrictive subtype (AN-R), anorexia nervosa binge-purge subtype (AN-BP), bulimia nervosa (BN), binge eating disorder (BED), other specified feeding and eating disorder (OSFED), healthy controls (HC). Age group comparisons: AN-BP vs. HC (p < .001); AN-BP vs. BED (p < .001); AN-R vs. HC (p = .019); AN-R vs. BED (p < .001); BN vs. BED (p = .012). BMI group comparisons: AN-BP vs. OSFED (p < .001); AN-BP vs. BN (p < .001); AN-BP vs. HC (p < .001); AN-BP vs. BED (p < .001); AN-R vs. OSFED (p < .001); AN-R vs. BN (p < .001); AN-R vs. HC (p < .001); AN-R vs. BED (p < .001); OSFED vs. BED (p < .001); BN vs. BED (p < .001); HC vs. BED (p = .003). Illness duration group comparisons: AN-R vs. BED (p < .001); AN-BP vs. BED (p = .007); OSFED vs. BED (p = .019); BN vs. BED (p = .037). All other comparisons were non-significant.
The diagnostic categories differ significantly on both the mean EDE-Q concerns score (H(5) = 148.847, p < .001), and all of its subscales [weight concerns (H(5) = 121.419, p < .001), shape concerns (H(5) = 129.271, p < .001) and eating concerns (H(5) = 170.061, p < .001)]. Patients with eating disorders are significantly more concerned about their eating, weight and shape than healthy controls, while the concerns of the different eating disorder diagnoses do not differ much from each other, except for patients with other specified feeding and eating disorders being less eating concerned than patients with bulimia nervosa. Means, standard deviations and outcomes of the EDE-Q subscales and the pairwise comparisons are depicted in table 2.
Table 2.
Eating Disorder Psychopathology
EDE-Q
|
AN-R
(N=82)
|
AN-BP (N=55)
|
BN
(N=35)
|
BED
(N=17)
|
OSFED (N=40)
|
HC
(N=98)
|
Eating concerns
|
2.97 (1.22)
|
3.60 (1.21)
|
3.94 (1.37)
|
3.27 (1.39)
|
2.60 (1.33)
|
0.68 (0.96)
|
Weight concerns
|
4.06 (1.54)
|
4.57 (1.42)
|
4.73 (1.14)
|
4.22 (1.62)
|
3.97 (1.51)
|
1.78 (1.47)
|
Shape concerns
|
4.58 (1.35)
|
4.99 (1.15)
|
4.83 (1.40)
|
4.57 (1.28)
|
4.30 (1.40)
|
2.10 (1.51)
|
Global score
|
3.87 (1.25)
|
4.39 (1.16)
|
4.50 (1.22)
|
4.02 (1.30)
|
3.62 (1.32)
|
1.52 (1.22)
|
Anorexia nervosa restrictive subtype (AN-R), anorexia nervosa binge-purge subtype (AN-BP), bulimia nervosa (BN), binge eating disorder (BED), other specified feeding and eating disorder (OSFED), healthy controls (HC). EDE-Q group comparisons on the three concerns subscales and global score. AN-R vs. HC (all p < .001; r between -.58 and -.67), AN-BP vs. HC (all p < .001; r between -.70 and -.83), BN vs. HC (all p < .001; r between -.65 and -.83), BED vs. HC (all p < .001; r between -.45 and -.54), OSFED vs. HC (all p < .001; r between -.49 and -.52), OSFED vs. BN (eating concerns p = .015; r = .38). All other comparisons were non-significant.
3.2. Anxiety: eating disorder specific fears
Specific eating disorder anxieties are assessed with the EFQ subscales fear of weight gain, fear of social consequences, fear of personal consequences, fear of physical sensations, fear of social eating, and the FOFM subscales trait anxiety about eating, and feared concerns.
All eating disorder diagnoses score significantly higher than healthy controls on the EFQ subscales fear of weight gain (H(5) = 121.551, p < .001), fear of social consequences (H(5) = 130.138, p < .001), fear of personal consequences (H(5) = 158.707, p < .001), fear of physical sensations (H(5) = 128.624, p < .001) and fear of social eating (H(5) = 123.938, p < .001). There are no significant differences between the eating disorder diagnoses. Likewise, all eating disorder diagnoses score significantly higher than healthy controls on the FOFM subscales trait anxiety about eating (H(5) = 194.643, p < .001) and feared concerns (H(5) = 181.657, p < .001). Patients with binge eating disorder score lower on eating anxiety than patients with the anorexia nervosa binge-purge subtype, while all other comparisons between the eating disorder diagnoses are non-significant. Means, standard deviations of the EFQ and FOFM subscales and the outcomes of the pairwise comparisons are shown in table 3.
Table 3.
Specific eating anxieties measured with the eating disorder fear questionnaire (EFQ) and fear of food measure (FOFM)
EFQ
|
AN-R
(N=84)
|
AN-BP (N=57)
|
BN
(N=38)
|
BED
(N=20)
|
OSFED (N=49)
|
HC
(N=110)
|
Fear of weight gain
|
6.08 (1.29)
|
6.24 (1.29)
|
6.46 (0.87)
|
6.20 (1.09)
|
6.11 (1.37)
|
3.91 (1.81)
|
Fear of social consequences
|
5.06 (1.57)
|
5.41 (1.51)
|
5.47 (1.44)
|
5.92 (1.22)
|
5.36 (1.46)
|
2.87 (1.47)
|
Fear of personal consequences
|
5.53 (1.34)
|
5.64 (1.47)
|
5.55 (1.29)
|
5.14 (1.56)
|
5.22 (1.43)
|
2.42 (1.52)
|
Fear of physical sensations
|
6.26 (1.25)
|
6.50 (1.02)
|
6.35 (0.96)
|
6.25 (1.03)
|
5.97 (1.35)
|
3.84 (1.93)
|
Fear of social eating
|
4.51 (1.98)
|
5.10 (2.19)
|
4.43 (2.10)
|
5.05 (1.55)
|
4.57 (1.92)
|
1.89 (1.51)
|
FOFM
|
AN-R
(N=84)
|
AN-BP (N=57)
|
BN
(N=37)
|
BED
(N=19)
|
OSFED (N=46)
|
HC
(N=104)
|
Trait eating anxiety
|
38.71 (11.02)
|
41.05 (13.10)
|
40.97 (11.71)
|
28.10 (13.69)
|
33.67 (11.86)
|
12.08 (7.30)
|
Feared concerns
|
42.26 (11.67)
|
46.21 (12.38)
|
42.08 (11.99)
|
35.84 (10.44)
|
39.26 (12.14)
|
16.50 (9.12)
|
Anorexia nervosa restrictive subtype (AN-R), anorexia nervosa binge-purge subtype (AN-BP), bulimia nervosa (BN), binge eating disorder (BED), other specified feeding and eating disorder (OSFED), healthy controls (HC). EFQ group comparisons. AN-R vs. HC (all p < .001; r between -.56 and -.72), AN-BP vs. HC (all p < .001; r between -.63 and -.74), BN vs. HC (all p < .001; r between -.51 and -.64), BED vs. HC (p < .001 and r between -.42 and -.59), OSFED vs. HC (all p < .001; r between -.47 and -.59). All other comparisons were non-significant. FOFM group comparisons (eating anxiety & feared concerns subscale). AN-R vs. HC (all p < .001; r between -.77 and -.81), AN-BP vs. HC (all p < .001; r between -.85 and -.86), BN vs. HC (all p < .001; r between -.68 and -.78), BED vs. HC (eating anxiety p = .001 and r = -.36; feared concerns p < .001 and r = -.41), BED vs. AN-BP (eating anxiety p = .045 and r = .34), OSFED vs. HC (all p < .001; r between -.61 and -.63). All other comparisons were non-significant.
To sum up, all eating disorder patients show more eating disorder specific anxieties than healthy controls. The diagnostic categories did not differ in specific eating anxieties on six out of seven subscales, but patients with binge eating disorder showed less trait anxieties about eating than patients with the anorexia nervosa binge-purge subtype.
3.3 Anxiety: general fears
General anxieties are assessed with the DASS anxiety subscale and the Intolerance of Uncertainty scale.
The diagnostic categories significantly differ on the DASS anxiety subscale (H(5) = 107.595, p < .001). The eating disorder diagnoses anorexia nervosa (both subtypes), bulimia nervosa and other specified feeding and eating disorders do not differ in general anxiety and they all show significantly more anxiety than the healthy controls. Patients with binge eating disorder show significantly less anxiety than patients with the anorexia nervosa binge purge subtype, but their scores do not significantly differ from the other eating disorder diagnoses and from the healthy controls. So, binge eating disorders are in the middle between most other eating disorders and the healthy controls without significantly differing from both (except for the anorexia nervosa binge purge subtype). In addition, all diagnostic categories do not differ from each other in intolerance of uncertainty, while they all score significantly higher on intolerance of uncertainty than healthy controls (H(5) = 110.806, p < .001). Means, standard deviations of the DASS and IUS and outcomes of the pairwise comparisons are displayed in table 4.
Table 4
General anxieties measured with the depression, anxiety, stress scale (DASS) and intolerance of uncertainty scale (IUS)
DASS
|
AN-R
(N=84)
|
AN-BP (N=56)
|
BN
(N=36)
|
BED
(N=18)
|
OSFED (N=46)
|
HC
(N=102)
|
Anxiety
|
19.31 (11.64)
|
24.82 (11.62)
|
21.50 (10.27)
|
12.89 (11.73)
|
18.30 (10.95)
|
6.84 (7.21)
|
IUS
|
AN-R
(N=84)
|
AN-BP (N=57)
|
BN
(N=38)
|
BED
(N=20)
|
OSFED (N=51)
|
HC
(N=111)
|
|
93.24 (18.63)
|
96.21 (21.39)
|
88.68 (19.85)
|
89.00 (23.88)
|
87.57 (18.91)
|
64.21 (18.44)
|
Anorexia nervosa restrictive subtype (AN-R), anorexia nervosa binge-purge subtype (AN-BP), bulimia nervosa (BN), binge eating disorder (BED), other specified feeding and eating disorder (OSFED), healthy controls (HC). DASS anxiety group comparisons. AN-R vs. HC (p < .001; r = -.53), AN-BP vs. HC (p < .001; r = -.70), BN vs. HC (p < .001; r = -.55), BED vs. HC (p = .615; r = -.19), BED vs. AN-BP (p = .007, r = -.41) OSFED vs. HC (p < .001; r = -.47). All other comparisons were non-significant. IUS group comparisons. AN-R vs. HC (p < .001; r = -.61), AN-BP vs. HC (p < .001; r = -.64), BN vs. HC (p < .001; r = -.45), BED vs. HC (p < .001; r = -.40), OSFED vs. HC (p < .001; r = -.46). All other comparisons were non-significant.
3.4: Avoidance behaviors
Avoidance behaviors are assessed with the FOFM food avoidance behaviors subscale and the EDE-Q eating restraint subscale.
The diagnostic categories differ in food avoidance behaviors (H(5) = 174.846, p < .001). The eating disorder diagnoses anorexia nervosa (both subtypes), bulimia nervosa and other specified feeding and eating disorders do not differ in food avoidance behaviors and they all show significantly more avoidance behaviors than healthy controls. The binge eating disorder diagnosis is in the middle again: binge eating disorders do not differ from healthy controls and they do not differ from the bulimia nervosa and other specified feeding and eating disorders. But, they show significantly less food avoidance behaviors than both anorexia nervosa subtypes. Further, the eating disorder diagnoses differ in eating restraint (H(5) = 129.777, p < .001). Patients with anorexia nervosa (both subtypes), bulimia nervosa and other specified feeding and eating disorders do not differ from each other in eating restraint and all score significantly higher than the control group, whereas patients with binge eating disorder do not significantly differ from healthy controls (p=0.06). Means, standard deviations and outcomes of the pairwise comparisons are provided in table 5.
Table 5
Avoidance behaviors measured with the food avoidance behaviors subscale from the fear of food measure (FAB) and the eating restraint subscale from the EDE-Q
FOFM
|
AN-R
(N=84)
|
AN-BP (N=57)
|
BN
(N=37)
|
BED
(N=19)
|
OSFED (N=46)
|
HC
(N=104)
|
FAB
|
31.14
(7.56)
|
30.65
(9.15)
|
27.35
(7.87)
|
18.53
(7.78)
|
27.22
(8.07)
|
12.10 (7.30)
|
EDE-Q
|
AN-R
(N=82)
|
AN-BP (N=55)
|
BN
(N=35)
|
BED
(N=17)
|
OSFED (N=40)
|
HC
(N=98)
|
Eating restraint
|
3.42 (1.45)
|
3.98 (1.63)
|
3.96 (1.22)
|
2.58 (1.35)
|
3.39 (1.55)
|
1.24 (1.23)
|
Anorexia nervosa restrictive subtype (AN-R), anorexia nervosa binge-purge subtype (AN-BP), bulimia nervosa (BN), binge eating disorder (BED), other specified feeding and eating disorder (OSFED), healthy controls (HC). FOFM group comparisons (food avoidance behaviors subscale). AN-R vs. HC (p < .001; r = -.83), AN-BP vs. HC (p < .001; r = -.77), BN vs. HC (p < .001; r = -.57), BED vs. HC (p = .626), BED vs. AN-R (p < .001; r = .45), BED vs. AN-BP (p < .001; r = .48), OSFED vs. HC (p < .001; r = -.60). All other comparisons were non-significant. EDE-Q group comparisons (eating restraint subscale). AN-R vs. HC (p < .001; r = -.61), AN-BP vs. HC (p < .001; r = -.74), BN vs. HC (p < .001; r = -.67), BED vs. HC (p = .061; r = -.27), OSFED vs. HC (p < .001; r = -.55). All other comparisons were non-significant.