Participants
Two independent samples were used in this study. The ED sample was recruited in thirteen specialized centers for ED across the Netherlands between February 2021 and November 2021 and consisted of 156 participants with EDs. Inclusion criteria for participation were currently in treatment in one of the centers and age older than 18. Only a very small minority of participants were male (women n = 150, men n = 5, unknown n = 1). For this reason, we limited our ED sample to women in order to create a gender homogeneous sample. The 150 women were diagnosed according to DSM-5 criteria in the following categories: Anorexia Nervosa (AN) 62.7% (n = 94), Boulimia Nervosa (BN) 9.3% (n = 14), Binge Eating Disorder (BED) 13.2% (n = 20), Otherwise Specified Feeding and Eating Disorder (OSFED)14% (n = 21) and Avoidance Restrictive Food Intake Disorder (ARFID) 0.7% (n = 1). Mean age was 29.45 (SD 10.44, range 18–67) and self-reported mean body mass index (BMI) was 22.13 (SD 7.62, range 12.70-48.65). Ninety-five women followed an outpatient ED treatment and 55 an inpatient ED treatment, with an average treatment duration of 8.9 months (SD = 11.51, range 1–96).
The community sample consisted of 982 individuals from the general Dutch population, recruited between November 2019 and May 2020. The inclusion criteria were age older than 18 years and absence of ED symptoms, according to scores of the EDE-Q. Following these inclusion criteria, we excluded participants younger than 18 (n = 17) from the original sample and also removed participants with an EDE-Q score more than two standard deviations above the mean (n = 40). We did not exclude participants with a very high or very low BMI, when their EDE-Q scores gave no indication for ED symptoms. The sample of the remaining 925 participants consisted of 669 women and 256 men with an average age of 36 years (M = 36.53, SD = 14.71, range 18–76) and an average self-reported BMI of 23.97 (SD = 3.64, range 13.93–41.10). To be able to compare the community sample and the ED sample we excluded the men in the community sample in our psychometric analyses and used a sample of female participants with an average age of 36.03 (SD = 14.10, range 18–76) and an average self-reported BMI of 23.71 (SD = 3.83, range 13.93–41.10).
Measures
The FAS [21]; Dutch version: Alleva and Martijn[29] contains seven items rated on a 5-point Likert scale (1 = ‘strongly disagree’ to 5 = ‘strongly agree’). Scores from all items are averaged to produce a total score, with higher scores reflecting greater functionality appreciation. Internal consistency was good with Cronbach’s alpha .89 for the total sample (.88 women, .90 men) and a test-retest reliability (ICC) of .81 for women and .74 for men [21].
The BCS [16]; Dutch version: Dorhout, Basten, Bosscher and Scheffers[31] measures the degree of satisfaction with appearance and functionality of different parts of the body. The BCS consists of 40 items rated on a 5-point Likert scale (1 = ‘very dissatisfied’ to 5 = ‘very satisfied’). Higher scores indicate a higher body satisfaction. The construct and concurrent validity of the original scale are good[18, 32] Research on the Dutch version of the BCS in both an ED (n = 238) and a community (n = 1060) sample revealed three subscales: functional body satisfaction, weight related body satisfaction and non-weight related body satisfaction [1]). Internal consistency was good in both samples with Cronbach’s α = .90 for the total scale and Cronbach’s α = .83 − .85 for the subscales in the clinical sample [15].
The EDE-Q [30]; Dutch version: Aardoom, Dingemans, Slof Op ‘t Landt and Van Furth[33] measures ED symptoms. The EDE-Q consists of 36 items of which 22 determine the total score. These 22 items comprise four subscales, assessing restraint, shape concerns, weight concerns and eating concerns over the previous 28 days. Items are answered on a 7-point Likert scale ranging from 0, ‘not one day’, to 6, ‘every day’. Higher scores are indicative of higher ED psychopathology. Internal consistency of the Dutch version is good, with Cronbach’s α = .95 for the total scale and Cronbach’s α varying from .81 to .91 for the subscales [33].
Procedure
The study protocol was approved by the ethics committee of the Faculty of Social and Behavioural Sciences of Utrecht University (number 20–383). Furthermore, the Medical Ethics Review Committee of Utrecht University was consulted on this procedure; reference number WAG/mb/20/022653.
To collect clinical data the project was advertised in the thirteen centers for ED as a study on “Measuring exercise behaviour and body image in people with EDs”, by means of a poster and accompanying flyers including a QR-code. To participate in the study, ED patients could scan the QR-code on the flyer. After providing digital informed consent, information about the study objective and the voluntary and anonymous participation, participants were asked to complete an online survey via Qualtrics (www.qualtrics.com).
Data collection for the community sample was conducted using a snowball sampling method through e-mails sent to potential participants in the network of six bachelor students from Windesheim University of Applied Sciences, School of Human Movement and Education and the network of the first two researchers of this study. The e-mail included a link to the questionnaires preceded by an informed consent, information about the study objective and the voluntary and anonymous participation, and a request to readers to forward the e-mail to others in their network. Participants completed the survey through a secure online system (Formdesk) and could indicate if they were willing to fill out the survey for a second time. If this was the case, they got a personal code and received the survey again after two weeks. Hundred sixteen female participants completed the survey twice within a three-week interval. All survey materials were removed from the internet upon completion of the data collection phase.
Informed consent was obtained from all individual participants included in this study, both clinical and non-clinical. In both samples no participatory incentives were offered.
Statistical analysis
The factor structure of the Dutch FAS was examined using a two-step process that involved exploratory factor analysis (EFA) in the first step and confirmatory factor analysis (CFA) in the second [34]. To ensure adequate sample sizes for both EFA and CFA, the all-female ED sample was split, using a computer-generated random seed, and the same procedure was followed for the female participants of community sample. For the EFA this resulted in a split-half female ED sample (n = 75), and a split-half community sample of woman (n = 331). For the CFA the following split-half samples were used: a female ED sample (n = 75) and a female community sample (n = 338). For the EFA, maximum likelihood with oblique rotation was used as the factor extraction method[35] using SPSS 28.0. Numbers of factors retained were based on interpretation of the scree plot[36] and parallel analysis [37]. Cross-loadings were defined as an item that loads at > .32 on two or more factors [38]. For the CFA, we chose to report a broad range of indices and included root mean square error of approximation (RMSEA), standardized root mean square residual (SRMR), Comparative Fit Index (CFI) and Tucker Lewis Index (TLI). As a rule of thumb, RMSEA values < .08 suggest adequate and < .05 good model fit[39] An SRMR between .05 and .10 indicates an acceptable fit and values < .05 indicate good fit [40]. CFI and TLI values in the range between .90 and .95 may be regarded as indicating good model fit [41].
To examine construct validity, correlations between functional appreciation, body satisfaction and ED pathology were analyzed using Pearson product-moment correlation coefficient with correlations considered strong if r = .50 to 1.0, moderate if r = .29 − .49 and weak if r = .10 to .29 [42]. Test-retest reliability of the FAS was established by intraclass correlation (ICC; two-way mixed model, absolute agreement, single measurement) [43], ICC ≥ .75 was considered excellent, between .60 and .74 good, between .40 and .59 fair and below .40 poor [44].
Independent t-tests were used to analyze differences in FAS scores between men and women in the community sample and between the female community and female ED sample.