When fitness becomes an obsession: a cross-sectional study investigating the risk of exercise addiction among athletes

Abstract

Objective

This study examines the relationship between the risk of exercise addiction and psychological factors like anxiety, depression, self-esteem and sleep quality in athletes.

Methods

This cross-sectional study was conducted in Lahore between August and November 2023, involving 282 athletes from five gyms engaged in bodybuilding, powerlifting or strength training. The study explored the relationship between exercise addiction risk and psychological factors, including anxiety, depression, body image distress, self-esteem, stress, obsessive-compulsive symptoms and sleep quality. Participants met specific inclusion criteria related to training frequency, duration and mental health status. Standardised questionnaires, including the Exercise Addiction Inventory, Hospital Anxiety and Depression Scale, Body Shape Questionnaire, Rosenberg Self-Esteem Scale, Obsessive-Compulsive Inventory-Revised, Pittsburgh Sleep Quality Index and Perceived Stress Scale, were used for data collection. Data were analysed using SPSS (V.24) with descriptive statistics, t-tests, analysis of variance (ANOVA), Pearson correlations and hierarchical multiple regression.

Results

267 completed the study. The majority were male (85.4%), with bodybuilding (50.2%) being the most common type of training. Participants reported high levels of psychological distress, including anxiety, body image concerns and stress. Exercise addiction risk was significantly associated with body image distress (r=0.45), anxiety (r=0.42) and stress (r=0.40), while self-esteem showed a negative correlation (r= –0.36). Hierarchical regression showed psychological factors accounted for 51% of the variance in addiction risk, with body image distress and anxiety being the strongest predictors. Bodybuilders and males reported significantly higher addiction risk scores compared with other groups.

Conclusion

Body image dissatisfaction, anxiety and stress are significant risk factors for exercise addiction, often reflecting emotional coping over performance goals.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Previous research has shown that psychological factors like anxiety, body image dissatisfaction and low self-esteem are linked to exercise addiction, especially in individualistic cultures. Exercise addiction risk is higher in appearance-focused sports such as bodybuilding.

WHAT THIS STUDY ADDS

• This study provides new evidence from a collectivist culture, showing that psychological predictors like body dissatisfaction and anxiety strongly relate to the risk of exercise addiction (REA) among strength athletes. It also highlights that elevated REA scores do not necessarily indicate dysfunctional exercise, emphasising cultural context in interpreting these risks.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The findings encourage sport psychologists and coaches to screen for psychological distress without prematurely labelling intense training as pathological. This nuanced understanding supports the development of culturally sensitive interventions that focus on mental health alongside athletic performance, helping to prevent potential negative outcomes without discouraging commitment.

Introduction

Exercise is widely recognised for its numerous physical and psychological benefits, including improved cardiovascular health, enhanced mental well-being and increased athletic performance. However, when exercise becomes excessive and compulsive, it can lead to a behavioural condition known as exercise addiction (EA).¹ Characterised by symptoms such as loss of control, withdrawal effects when unable to exercise and prioritising physical activity over health or relationships, EA is particularly prevalent among athletes and individuals engaged in high-intensity strength training disciplines like bodybuilding and weightlifting.²

The risk of exercise addiction is particularly prevalent among athletes, who often face immense pressure to perform at their peak. While dedication to training is essential for achieving competitive success, the line between commitment and addiction can become blurred. Elite athletes, in particular, are at greater risk due to the high demands of their sport and the cultural normalisation of extreme training regimens. Studies suggest that up to 42% of athletes may exhibit symptoms of exercise addiction depending on the sport and assessment tools used.³ This condition not only compromises physical health through overtraining and injuries but also negatively impacts mental well-being, often co-occurring with anxiety, depression or disordered eating.⁴

Exercise addiction in athletes stems from a mix of psychological, neurobiological and sociocultural influences. Many rely on exercise to manage stress, anxiety or low self-esteem, while perfectionism and body image concerns drive compulsive training. The brain’s reward system reinforces this behaviour through endorphin and dopamine release, creating dependency and withdrawal symptoms when training is restricted.⁵ Social pressures, media glorification of relentless training and competitive demands normalise excessive exercise.⁶ In weight-sensitive sports, athletes may overtrain to maintain performance, often prioritising exercise over recovery, blurring the line between dedication and addiction.⁷

Previous studies indicate that the prevalence of exercise addiction varies across athletic disciplines and populations. For example, endurance sports, such as marathon running and cycling, show a higher risk for EA (14.2%) compared with other sports, like ball games (10.4%) and power disciplines (6.4%).⁸ However, not all high-volume exercisers develop EA; it is often the emotional relationship with exercise, such as feelings of guilt for missing a session or exercising despite injury, that distinguishes addiction from dedication.⁹

In Pakistan, the fitness culture has grown significantly over the past decade, particularly in urban centres like Lahore. With the increasing popularity of gyms and strength training among young adults, bodybuilders and weightlifters represent a population at heightened risk of developing EA. The cultural emphasis on achieving an ideal muscular physique, often influenced by global fitness trends, further exacerbates the risk of compulsive exercise behaviours.¹⁰ Despite this growing phenomenon, research on EA in Pakistan remains scarce, leaving a critical gap in understanding its prevalence, psychological correlation and impact on athletes’ well-being.

This study explores how psychological factors such as anxiety, depression, body image concerns, self-esteem, stress, obsessive-compulsive traits and sleep quality are linked to the risk of exercise addiction in bodybuilders and weightlifters. It looks at how gym culture, societal expectations and personal mindset influence the boundary between dedication and harmful exercise patterns. Grounded in the Biopsychosocial Model of addiction,¹¹ it highlights how biological, psychological and social influences together shape vulnerability to REA. The aim is to support strategies that promote healthier training behaviours and mental well-being among athletes.

Methodology

Study design and setting

This study employed a cross-sectional survey design conducted between August 2023 and November 2023, in Lahore. A total of 282 athletes were recruited from five different gyms, focusing on individuals engaged in bodybuilding, powerlifting and strength training disciplines. This study is grounded in Self-Determination Theory¹² to examine how unmet psychological needs for autonomy, competence and relatedness may contribute to exercise addiction risk by affecting motivation and related psychological factors.

Exercise type differences

These training types differ in both physical focus and psychological implications. Bodybuilding emphasises muscle size and aesthetics through high-volume workouts and is closely linked to body image concerns and a higher risk of exercise addiction.¹³ Powerlifting emphasises maximal strength in three lifts using heavy weights and low repetitions, with less concern for appearance but increased performance pressure.¹⁴ Strength training is more general, aimed at improving overall fitness, with lower psychological strain and a reduced risk of compulsive behaviour.¹⁵

Participants and inclusion criteria

The purpose was to examine the relationship between risk of exercise addiction and various psychological factors, including anxiety, depression, body image distress, self-esteem, stress levels, obsessive-compulsive tendencies and sleep quality. Participants were selected based on specific inclusion criteria to ensure the study focused on individuals at risk of developing exercise addiction. Eligible participants were 18 years or older and had been engaged in structured exercise for at least 12 months. They were required to be actively involved in bodybuilding, powerlifting or strength training and train for at least 5 hours per week. Additional inclusion criteria included a regular gym membership, psychological readiness (with no history of severe psychiatric disorders, such as schizophrenia or bipolar disorder) and proficiency in the Urdu language to complete the study questionnaires. Only those who provided informed consent and fully completed all assessments were included. Individuals with recent injuries or chronic medical conditions that prevented regular exercise were excluded from the study.

Sample size determination

The required sample size was determined using G*Power 3.1 software. Based on an effect size of 0.02, power of 0.90, an alpha level of 0.05 and four predictors in the regression model, the estimated sample size was 282 participants.

Measures

Data were collected through questionnaires to assess the risk of exercise addiction and its psychological correlates. The Exercise Addiction Inventory (EAI) was used to measure compulsive exercise behaviours on a six-item scale, with scores ranging from 6 to 30¹⁶. Higher scores indicated greater risk of exercise addiction, with a cut-off score of 24 or higher suggesting a high risk of exercise addiction. The Hospital Anxiety and Depression Scale (HADS) consisted of 14 items, with two subscales measuring anxiety and depression separately, each with scores ranging from 0 to 21.¹⁷ Scores between 8 and 10 indicated mild symptoms, scores between 11 and 14 indicated moderate symptoms, and scores between 15 and 21 indicated severe symptoms. The Body Shape Questionnaire (BSQ),¹⁸ a 34-item scale, assessed body image distress, with scores above 140 indicating significant body image concerns. Self-esteem levels were measured using the Rosenberg Self-Esteem Scale (RSES),¹⁹ a 10-item questionnaire on a four-point scale, where higher scores represented higher self-esteem. The Obsessive-Compulsive Inventory-Revised (OCI-R),²⁰ an 18-item measure, assessed obsessive-compulsive tendencies, with scores ranging from 0 to 72 and a cut-off of 21 or higher indicating clinically significant symptoms. The Pittsburgh Sleep Quality Index (PSQI) was used to evaluate sleep quality across seven components,²¹ with a global score above five indicating poor sleep quality. The Perceived Stress Scale (PSS), a 10-item measure, assessed overall stress levels, with scores of 27 or higher indicating high perceived stress.²²

Data analysis

Data were analysed using SPSS V.24.0. Descriptive statistics summarised participant demographics and psychological variables. One-way analysis of variance (ANOVA) examined differences in risk of exercise addiction by exercise type, and independent t-tests compared REA by gender. Pearson correlation coefficients were used to assess the relationships between REA and psychological factors. A hierarchical multiple regression analysis was conducted to examine the contribution of psychological and demographic variables to the risk of exercise addiction. Statistical significance was set at p value<0.05.

Results

A total of 282 participants were initially recruited for the study. However, 15 participants were excluded due to incomplete survey responses or withdrawal from the study, resulting in a final sample size of 267 athletes (response rate: 94.7%). Reasons for dropout included lack of time (n=6), loss of interest (n=4) and incomplete questionnaire responses (n=5).

Table 1 presents the baseline characteristics of the participants. The average age was 26.4 years, with a predominance of males (85.4%). On average, participants had 3.2 years of training experience, engaged in 5.6 training sessions per week, totalling 8.5 hours weekly. Notably, 25.5% of the participants reported involvement in competitive sports, ranging from local and regional events to national-level championships, primarily in bodybuilding and powerlifting. This subset of athletes actively participates in organised competitions, which may influence their training intensity, psychological stress and motivation. Regarding exercise type, bodybuilding was the most common (50.2%), followed by powerlifting (28.8%) and strength training (21.0%). These demographics provide a comprehensive overview of the study population, which is important for interpreting subsequent analyses.

Table 1. Baseline characteristics of study participants.

Variable	Mean±SD/n (%)
Total participants	267
Age (years)	26.4±4.7
Gender
Male	228 (85.4%)
Female	39 (14.6%)
Training experience (years)	3.2±1.1
Training frequency (sessions/week)	5.6±1.3
Training duration (hours/week)	8.5±2.3
Competitive participation	68 (25.5%)
Exercise type
Bodybuilding	134 (50.2%)
Powerlifting	77 (28.8%)
Strength training	56 (21.0%)

The results presented in table 2 align with the study’s aim by demonstrating a clear connection between the risk of exercise addiction and its psychological correlates. The elevated mean EAI score suggests a notable potential risk of addiction to exercise among athletes. High anxiety, depression and body image distress scores indicate that psychological distress is prevalent in this group. The variability in self-esteem, significant obsessive-compulsive tendencies, poor sleep quality and high stress levels further reinforces the study’s objective of examining the mental health challenges associated with compulsive exercise behaviour.

Table 2. Descriptive statistics of psychological assessments.

Variable	Mean±SD	Range
Exercise Addiction Inventory (EAI)	22.3±5.6	12–30
Hospital Anxiety and Depression Scale (HADS)	14.8±3.9	8–21
Body Shape Questionnaire (BSQ)	87.6±15.2	56–125
Rosenberg Self-Esteem Scale (RSES)	18.3±4.7	10–30
Obsessive-Compulsive Inventory-Revised (OCI-R)	20.5±6.1	12–36
Pittsburgh Sleep Quality Index (PSQI)	7.2±2.3	3–12
Perceived Stress Scale (PSS)	19.8±4.5	10–30

Table 3 shows significant correlations between the risk of exercise addiction and psychological factors. Body image distress (r=0.45, p value=0.005) had the strongest positive association, followed by anxiety (r=0.42, p value<0.001), stress (r=0.40, p value<0.001) and depression (r=0.38, p value=0.001), indicating that individuals with higher distress levels are more prone to compulsive exercise. Conversely, self-esteem (r=−0.36, p value=0.004) showed a negative correlation, suggesting that lower self-esteem increases addiction risk. These findings highlight the psychological vulnerabilities linked to excessive exercise and the need for targeted mental health interventions.

Table 3. Correlations between REA and psychological variables (with 95% CI).

Variable	REA (r)	95% CI	P value
Anxiety (HADS)	0.42	(0.35, 0.53)	<0.001
Depression (HADS)	0.38	(0.32, 0.51)	0.001
Body Image Distress (BSQ)	0.45	(0.29, 0.50)	0.005
Self-Esteem (RSES)	−0.36	(−0.49, −0.20)	0.004
Stress (PSS)	0.40	(0.26, 0.52)	<0.001

Note. Bootstrapped 95% CI based on 1000 resamples.

BSQ, Body Shape Questionnaire; HADS, Hospital Anxiety and Depression Scale; PSS, Perceived Stress Scale; REA, risk of exercise addiction; RSES, Rosenberg Self-Esteem Scale.

Before interpreting the hierarchical regression analysis, the key assumptions of linear regression were evaluated. Visual inspection of scatterplots confirmed linearity between predictors and the outcome. The Durbin-Watson statistic (value=1.94) indicated independence of residuals. Residual plots showed no evidence of heteroscedasticity, supporting the assumption of homoscedasticity. Multicollinearity was not a concern, as all Variance Inflation Factor (VIF) values were below 2. Normality of residuals was confirmed through Q-Q plots and a non-significant Shapiro-Wilk test (p value>0.05). Additionally, no significant outliers were detected based on standardised residuals, Cook’s distance or leverage values. These diagnostics support the validity of the regression results presented below.

Table 4 showed that psychological factors are stronger predictors of the risk of exercise addiction than demographic or training variables. Gender and exercise type explained 10% of the variance, while the addition of psychological factors increased this to 51%. Body image distress (β=0.33) and anxiety (β=0.31) were the strongest predictors, followed by stress and depression. Group comparisons revealed that bodybuilders had the highest risk scores (M=23.1), followed by powerlifters (M=22.5) and strength trainers (M=21.0), with significant differences. Males scored slightly higher than females, though the effect size was small. These results suggest that emotional and cognitive factors, like body dissatisfaction and anxiety, play a larger role in exercise addiction risk than exercise type or frequency alone.

Table 4. Hierarchical regression predicting risk of exercise addiction.

Predictor	β	SE	95% CI
Step 1: Control variables
Gender (male vs female)	0.12	0.07	(−0.01, 0.25)
Exercise type (reference: strength training)
Bodybuilding	0.15	0.08	(0.01, 0.29)
Powerlifting	0.13	0.07	(−0.01, 0.26)
Step 2: Psychological predictors
Body Image Distress (BSQ)	0.33	0.06	(0.21, 0.45)
Anxiety (HADS-A)	0.31	0.08	(0.17, 0.44)
Stress (PSS)	0.29	0.09	(0.12, 0.45)
Depression (HADS-D)	0.27	0.07	(0.13, 0.41)

Note. Bootstrapped 95% CI based on 1000 resamples.

BSQ, Body Shape Questionnaire; HADS-A, Hospital Anxiety and Depression Scale – Anxiety subscale; HADS-D, Hospital Anxiety and Depression Scale – Depression subscale; PSS, Perceived Stress Scale.

Model statistics

• Model R² (Step 1) = 0.10, p=0.026

• Model R² (Step 2) = 0.51, p=0.002

• ΔR² = 0.41, p<0.001

Table 5 presents group differences in the risk of exercise addiction (REA) based on exercise type and gender. Participants involved in bodybuilding reported the highest REA scores (M=23.1, SD=5.7), followed by powerlifters (M=22.5, SD=5.3) and strength trainers (M=21.0, SD=5.2). The difference across exercise types was statistically significant, F(2, 264) = 4.21, p value=0.016, with a small effect size (η² = 0.031). Additionally, male participants showed significantly higher REA scores (M=23.0, SD=5.5) than females (M=20.2, SD=4.9), t265 = 2.03, p value=0.044, with a small effect size (Cohen’s d=0.31). These findings suggest that both exercise type and gender play a role in REA differences.

Table 5. Group comparisons of therisk of exercise addiction by exercise type and gender.

Variable	Group	Mean (M)	SD	Test statistic	P value	Effect size
Exercise type	Bodybuilding	23.1	5.7
	Powerlifting	22.5	5.3	F(2, 264) = 4.21	0.016	η² = 0.031
	Strength training	21.0	5.2
Gender	Male	23.0	5.5	t(265) = 2.03	0.044	d=0.31
	Female	20.2	4.9

Discussion

The present study investigated psychological and demographic predictors of the risk of exercise addiction (REA) among strength-based athletes in Pakistan, including those engaged in bodybuilding, powerlifting and general strength training. These findings contribute to the ongoing discourse surrounding compulsive exercise behaviour, particularly in high-training-volume populations and support previous research emphasising the interplay between psychological vulnerability and rigorous physical activity.²³

Our results show that psychological factors, especially body image dissatisfaction and anxiety, are more strongly associated with REA than training-related or demographic variables. This aligns with prior research indicating that body dysmorphic symptoms and anxiety are significant predictors of exercise addiction in gym-goers.²⁴ Consistent with a previous study by von Ranson et al,²⁵ we also observed positive correlations between REA and depression, perceived stress and obsessive-compulsive traits, suggesting that maladaptive affect regulation mechanisms may contribute to compulsive training behaviour.

We also found a significant negative correlation between self-esteem and REA, reinforcing theoretical models of behavioural addiction, which propose that low self-worth can drive excessive engagement in rewarding behaviours as a coping mechanism.²⁶ This tendency to seek temporary emotional relief through exercise may lead to the development of dependency patterns over time.

Although psychological variables had the strongest predictive power, demographic and training-related variables played a secondary role. Specifically, participants engaged in bodybuilding reported significantly higher REA scores than those involved in powerlifting or general strength training. This supports the notion that appearance-oriented sports may heighten vulnerability to compulsive exercise due to societal and subcultural pressures to attain an ideal physique.²⁷

Gender differences were observed, with males reporting higher REA scores than females. This partially supports earlier findings,²⁸ although some studies have found no gender differences or even higher REA prevalence among females in different athletic contexts.²⁹ Such inconsistencies highlight the influence of sport-specific demands, individual motivation and sociocultural gender norms on REA vulnerability.

While previous studies³⁰ have recently critiqued the focus on exercise addiction in athletes, arguing that high training levels are often goal-oriented and non-pathological, our study deliberately adopted a risk-based approach. The REA framework allows for early identification of psychological distress without necessarily labelling athletes as clinically addicted. As Lukács et al ²³ noted, REA exists along a continuum, with both adaptive and maladaptive manifestations depending on the individual and situational context.

Importantly, our findings correspond with those from individualistic societies where REA has been linked to anxiety, depression and body dissatisfaction.³¹ However, cultural differences warrant caution in generalising these findings. In collectivist contexts like Pakistan, communal values, family expectations and social conformity may influence exercise behaviour in unique ways.³² For instance, high commitment to training may reflect adherence to group-based ideals rather than psychological dysfunction. This perspective supports the argument that high involvement in exercise should not automatically be equated with pathology.³⁰

Our aim is not to pathologise structured athletic training, but to recognise when it may stem from psychological distress. Elevated anxiety, stress and body dissatisfaction among participants suggest that, for some, exercise serves as an emotional escape rather than performance enhancement. This is consistent with the Affect Regulation Model, which suggests excessive exercise may be used to manage negative emotions, potentially leading to compulsive patterns over time.³³

Limitation

The sample was taken from five gyms in Lahore, specifically those engaged in bodybuilding, powerlifting and strength training. This limits the generalisability of the findings to other populations, such as recreational exercisers or individuals from different sports. Additionally, the results are specific to Pakistani gyms, where cultural differences in body image perception, stress management and exercise attitudes may affect the applicability of these findings to other regions. These factors highlight the need for future research involving more diverse populations to enhance the study’s broader relevance.

Conclusion

This study found that body image dissatisfaction, anxiety and stress are stronger predictors of exercise addiction risk than training habits or demographics. It suggests that for some, compulsive exercise serves as a form of emotional coping, particularly in appearance-focused sports. Conducted in a collectivist culture, the study also shows how cultural context influences the meaning and impact of excessive exercise, underscoring the need for culturally sensitive prevention strategies.

Data availability statement

Data are available upon reasonable request.