This study involved secondary analysis of data collected for the development and validation of the V-EDS described in detail in McLean et al. [10]. Ethics approval was obtained from the Monash University Human Research Ethics Committee (Project ID: 30651) and participants were informed of the purpose of the study and provided informed consent prior to participating.

Participants were advertised through previous participant databases, social media advertisements (e.g., Vegans in Australia Facebook group), and eating disorder charity networks to complete an online survey on vegetarian and vegan eating behaviours and attitudes. Participants were required to be 18 years or over, residing in Australia, and adhering to a vegetarian or vegan diet to be eligible. Participants who identified as “meat-reducers” (i.e., flexitarians, semi-vegetarians, pescatarians) were excluded from the study owing to our specific focus on vegetarian and vegan eating behaviours.

Participant demographic characteristics Participants responded to several demographic characteristic questions (e.g., age, gender, ethnicity, religion, education level) and specific questions about their dietary adherence including duration of adherence and motivations.

Vegetarian vegan eating disorder screener (V-EDS) [10]. The V-EDS is an 18-item self-report measure designed to assess symptoms of eating disorder pathology in vegetarians and vegans over the past seven days. The V-EDS comprises six dietary characteristic items and 12 behavioural and attitudinal items. The dietary characteristic items can be used to provide clinical information around the respondent’s dietary attitudes (e.g., “Your vegetarian/vegan diet is a part of your identity”) and are along a 5-point Likert scale from strongly disagree to strongly agree. The following behavioural and attitudinal items (e.g., “Has the way you thought about food become intrusive?”) measure the presence of eating disorder pathology, rated along a 5-point Likert scale from no days to every day. The V-EDS has been found to support a unidimensional factor structure with strong initial internal consistency (α = 0.95–0.96) and convergent validity (0.87–0.88), and moderate discriminate validity (0.45–0.55). Internal consistency in the present study was excellent (α = 0.95, ω = 0.95).

Briefly, participants were advertised with a link to the online survey and responded to demographic characteristic questions, including age, gender, ethnicity, religion, and highest completed education, followed by the V-EDS. Participants self-reported their eating disorder diagnosis if applicable.

Statistical analyses were performed using SPSS Version 27 [15]. Descriptive statistics were calculated, including means and percentages for participant characteristics (age, gender, dietary status) and medians for V-EDS scores, and presented by study group (clinical, non-clinical). Differences in participant characteristics by group were examined using t-tests for continuous variables (age) and chi-square tests for categorical variables (gender, dietary status). Receiver operating characteristic (ROC) curve analysis was used to compute possible cut-off values for the V-EDS. The ROC is a plot that displays the trade-off between the sensitivity and its 1-specificity across a range of threshold values. The area under the curve (AUC) was calculated to provide an overall test performance statistic based on the trapezoidal rule. The discrimination power of the V-EDS is interpreted according to the AUC as non-informative (AUC = 0.50), poor (0.50 < AUC < 0.70), good (0.70 < AUC < 0.90), excellent (0.90 < AUC < 1.00), and perfect (AUC = 1.00; [16]). For this study, the optimal cut-off was identified using the Youden index which puts equal weight on sensitivity and specificity (sensitivity + sensitivity − 1), resulting in an index score between 0 and 1 [17, 18]. The cut-off point for having an acceptance Youden index is 50%, with higher scores indicating better performance. Statistical significance level was set at p < 0.05.