[Paper Review] Leveraging Machine Learning to Identify Gendered Stereotypes and Body Image Concerns on Diet and Fitness Online Forums

Minh Duc Chu, Cinthia Sánchez+4 more|arXiv (Cornell University)|Jul 4, 2024

Obesity and Health Practices4 citations

TL;DR

This study uses transformer-based NLP and community membership analysis to map 46 Reddit diet and fitness forums along gender and body ideal axes, revealing that thinness-oriented communities—especially feminine-leaning ones—express significantly higher levels of negative emotions and toxicity, while muscularity-focused forums show less negativity regardless of gender. The findings highlight gendered emotional dynamics in online body image discourse and inform more nuanced content moderation strategies.

ABSTRACT

The pervasive expectations about ideal body types in Western society can lead to body image concerns, dissatisfaction, and in extreme cases, eating disorders and other psychopathologies related to body image. While previous research has focused on online pro-anorexia communities glorifying the "thin ideal," less attention has been given to the broader spectrum of body image concerns or how emerging disorders like muscle dysmorphia ("bigorexia") present on online platforms. To address this gap, we analyze 46 Reddit forums related to diet, fitness, and mental health. We map these communities along gender and body ideal dimensions, revealing distinct patterns of emotional expression and community support. Feminine-oriented communities, especially those endorsing the thin ideal, express higher levels of negative emotions and receive caring comments in response. In contrast, muscular ideal communities display less negativity, regardless of gender orientation, but receive aggressive compliments in response, marked by admiration and toxicity. Mental health discussions align more with thin ideal, feminine-leaning spaces. By uncovering these gendered emotional dynamics, our findings can inform the development of moderation strategies that foster supportive interactions while reducing exposure to harmful content.

Motivation & Objective

• To investigate how gendered body ideals shape emotional expression and toxicity in online diet and fitness communities.
• To examine whether communities centered on the thin ideal or muscular ideal differ systematically in emotional tone and mental health discourse.
• To assess how moderation practices might be improved by accounting for gendered emotional patterns and body ideal alignment.
• To explore the intersection of body image concerns, mental health challenges, and online emotional expression across diverse subreddits.
• To develop a framework for identifying and supporting vulnerable users in online communities without reinforcing harmful gender stereotypes.

Proposed method

• The study analyzes 46 Reddit subreddits related to diet, fitness, and mental health using public data collected under Reddit’s terms of service.
• Community membership structures are analyzed to project subreddits along gender and body ideal axes, using a method adapted from Waller and Anderson (2021).
• Transformer-based language models (e.g., BERT-based classifiers) are applied to detect emotions and toxicity in user-generated content across subreddits.
• Emotion and toxicity scores are quantified and correlated with community characteristics, including gender orientation and body ideal focus.
• The analysis includes both quantitative projection of communities on axes and qualitative interpretation of emotional patterns in relation to mental health discourse.
• Ethical safeguards include anonymization of data, IRB approval, and regular team check-ins to mitigate researcher distress.

Experimental results

Research questions

• RQ1 RQ1: Are there systematic differences in the structure, content, and membership of diet and fitness communities, and where do they fall on the body ideal spectrum?
• RQ2 RQ2: Are there systematic differences in the emotions and toxicity expressed in different communities, and what do these reveal about body ideals and gender stereotypes?
• RQ3 RQ3: How do responses to members’ posts differ in terms of emotions and toxicity across communities?
• RQ4 RQ4: Is toxicity associated with harmful content, particularly in communities discussing serious mental health issues?

Key findings

• Thinness-oriented subreddits, especially those with feminine-leaning membership, exhibit significantly higher levels of negative emotions and toxicity compared to muscularity-focused forums.
• Muscularity-oriented communities show lower overall negativity, regardless of the gender orientation of their members, suggesting a protective effect of the muscular ideal on emotional tone.
• Communities discussing serious mental health issues—such as anorexia and bigorexia—align more closely with thinness-oriented, feminine-leaning forums in emotional expression, indicating shared emotional dynamics.
• The emotional profiles of communities are more strongly linked to body ideals than to gender identity, with muscular ideal forums showing consistent emotional neutrality across gender lines.
• Toxicity is systematically associated with harmful content, particularly in thinness-promoting communities, where it often co-occurs with disordered eating discourse.
• Emotion and toxicity classifiers show robustness to potential gendered linguistic bias, as evidenced by consistent performance across male- and female-dominated subreddits in validation tests.