Eating Disorders and Their Association with Depression and Anxiety Among Medical Students: A Saudi Cross-Sectional Study

Abstract

Background: Eating disorders (EDs) are important mental illnesses that are often associated with depression and anxiety, leading to significant negative consequences. However, research on this topic in Saudi Arabia remains limited. This study aims to examine the risk of EDs among male and female medical students at King Saud University (KSU) and assess their risk factors and association with anxiety and depression. Methods: A cross-sectional study involving 425 participants was conducted, using a convenience sampling method. The study tools consisted of a questionnaire developed by the research team, the Eating Attitudes Test-26 (EAT-26), the Patient Health Questionnaire-9 (PHQ-9), and the Generalized Anxiety Disorder-7 scale (GAD-7). Results: Almost half (49.6%) were classified as high risk for EDs. Obesity was much higher among high-risk students than low-risk students (p < 0.001). Anxiety and depression were greater among high-risk students than low-risk ones. A higher body mass index (BMI) and depression greatly increased the risk of EDs (p < 0.001). Conclusions: The findings support the notion that medical students have a significant likelihood of developing EDs, especially if they have a high BMI and are depressed. The results show the importance of early identification and offering appropriate interventions to this vulnerable group.

1. Introduction

Eating disorders (EDs) are a group of psychiatric illnesses that include anorexia nervosa, bulimia nervosa, binge ED, avoidant/restrictive food intake disorder (ARFID), and other specified and unspecified EDs [1]. These illnesses affect 55.5 million individuals globally and have a lifetime prevalence of 6.1% in Saudi Arabia [2,3]. They are significant mental health conditions and can lead to serious mental and physical negative consequences [1,4]. Regarding mental consequences, EDs are linked to other psychiatric conditions such as post-traumatic stress disorder, self-harm behavior, depression, and anxiety [5,6,7]. In terms of physical consequences, EDs are associated with arrhythmia, hypotension, and electrolyte imbalances, among other physical manifestations [8]. Moreover, anorexia nervosa carries the highest mortality rate of any psychiatric disorder [9], with the leading causes for this mortality include suicide, cardiac-related causes, and pneumonia [10].

EDs are a substantial cause of morbidity, with the Global Burden of Disease (GBD) report estimating that EDs are the cause of 6.6 million cases with reduced disability-adjusted life years (DALYs) [2]. Furthermore, EDs result in detrimental consequences, such as a higher mortality risk and emotional and financial strains on families and societies [4,11]. Further, the quality of life, in terms of both social and physical functioning, was also found to be worse among those affected by EDs [12].

Globally, the management of EDs entails a specialized system of care. International standards include not only outpatient therapy but also Partial Hospitalization Programs (PHP), Intensive Outpatient Programs (IOP), and dedicated Residential Treatment Centers [13]. These specialized facilities allow for a multidisciplinary approach combining medical stabilization, nutritional rehabilitation, and evidence-based psychotherapies, such as Cognitive Behavioral Therapy (CBT) [13]. However, in the Middle East, the literature indicates that the recognition of EDs remains limited [14]. While psychiatric services are available in tertiary care hospitals, the region lacks a widespread network of specialized services [15]. Consequently, unlike the established networks found globally [13], the region faces unique challenges in early detection and management of EDs, often attributed to limited mental health awareness and stigma [14,16]. This structural gap highlights the urgent need for developing culturally appropriate specialized intervention programs to address the needs of those affected with EDs in the Middle East.

EDs are often accompanied by depression and anxiety [7], which share some risk factors such as low self-esteem, emotional dysregulation, and body dissatisfaction [17,18]. Depression may compromise one’s perception of self-worth and motivation, whereas anxiety may interfere with dietary patterns [7,14]. In addition, such co-occurrence complicates diagnosis and treatment, highlighting the need for integrated mental health screening [18]. Further, EDs with comorbid anxiety and depression tend to exhibit more severe symptoms, increased risk of suicidality, and a poorer prognosis, than those without such psychiatric comorbidities [19,20]. Nonetheless, although EDs may affect individuals of any demographic, they are more common among females [21], and tend to emerge most often in late adolescence and young adulthood [22]. Beyond gender and age, other factors also increase the vulnerability for EDs such as body dissatisfaction, low self-esteem, high perfectionism, and shape and weight concerns [23].

One of the vulnerable populations for EDs is the medical students. Medical students have unique stressors that increase their susceptibility [24,25]. Such stressors among medical students include, but are not limited to, the lack of time and academic-related pressures [26]. Poor sleep quality is another critical factor; research indicates that poor sleep quality, which affects a vast majority of medical students, is significantly associated with higher levels of anxiety [27]. Another possible contributing factor that may increase susceptibility among medical students is the high rate of obsessive–compulsive behavior [28]. Further, several studies report that EDs symptoms among students have been increasing over time, with an estimated 10.4% risk in medical students [24,25]. A Lebanese study found that 17% of medical students are at high risk for EDs [14]. A Saudi study found rates up to 32.1%, with female students being more affected [16]. Another Saudi study conducted at Taif University, with only female participants, found 35.4% to be at risk of EDs [29].

In addition, many studies have indicated that medical students with EDs often experience mental health issues, such as anxiety and depression [30]. For instance, a study identified anxiety and depressive disorders as the most common psychiatric comorbidities in individuals with EDs [31], with academic stress, high expectations, and emotional exhaustion likely contributing [32,33]. Similarly, a broader survey among university students in Morocco identified a significant correlation between reported anxiety and depression and disordered-eating behaviors; around 32.2% of the students exhibited EDs symptoms, while 28.8% and 26.4% showed anxiety and depressive symptoms, respectively [34]. In a sample of university students in Spain, mild-to-severe depression or anxiety also significantly increased the odds of disordered-eating behaviors [35]. A study conducted in Lebanon revealed that medical students with EDs were commonly reported to be suffering from anxiety and depression as well [14]. Another study from South India revealed a relationship between stress, body image concerns, and EDs among medical students [32]. Other similar associations have been described; in a cross-sectional study of female medical students in Anhui, China, women at risk of subclinical EDs showed higher scores of depression and anxiety, as well as in slimming preoccupation and poorer relationships with parents [36]. A recent large multinational, cross-country study (including medical students from twelve countries, including Middle East and North Africa) indicated that almost a quarter of the students were at risk of EDs, with the risk being concentrated among those with other psychiatric or chronic conditions and receiving more exposure to thin-ideal body images [37]. Similarly, a study among university students in Bangladesh demonstrated that students with depressive symptoms, and anxiety symptoms had significantly higher risk of eating disorders compared to those without, highlighting the strong connection between emotional distress and EDs pathology [38].

Several factors could play a role in the underrecognition of EDs. The combination of cultural ideals plays an important role. For example, a study in Kuwait found that internalization of Western thin-appearance norms significantly predicts EDs pathology among adult women, suggesting that when thinness ideals are widespread, they may normalize extreme dieting behaviors and make early EDs symptoms less apparent to individuals and health professionals [39]. Dietary changes also is contributory factor; a narrative review of university students reported high consumption of energy-dense snacks, fast food, meal-skipping, and irregular eating patterns behaviors that overlap with disordered eating but are widely perceived as normal student dietary habits, potentially masking early ED symptoms [40]. Stigma further limits the recognition, as a systematic review demonstrated that individuals with EDs frequently experience social stigma and misconceptions that EDs are self-inflicted or are due to personal weakness, which reduces help-seeking and contributes to delays in recognition by health professionals [41]. Limited mental-health awareness also plays a role; for example, a qualitative study of university students found that participants exhibited poor awareness of EDs signs, causes, and appropriate help sources, and that this limited mental-health literacy may delay help-seeking and early recognition [42]. Together, earlier-mentioned factors may all play a role in hindering detection and management of EDs [43]. In the Saudi context, the widespread usage of social media—where daily usage rates are exceptionally high—has been linked to increased body dissatisfaction and the internalization of unrealistic beauty standards among Saudi women [44].

Although EDs symptoms are highly documented among medical students [14,16,25], very little attention has been given to whether affected students actually receive appropriate treatment. A study conducted in Saudi Arabia has revealed that there is a significant treatment gap; whereby a large proportion of medical students experiencing a high degree of psychological distress fail to seek help from a psychologist, which is usually because of stigma, lack of awareness about the services available, and the belief that they should manage their symptoms on their own [45]. As such, we assume that Saudi medical students rarely receive ED-needed help. The same pattern is observed internationally, as the majority of university students with EDs are not receiving treatment, usually because of a lack of time and not knowing where to get the help from [46]. Further, a qualitative study also pointed to the fact that when students seek help, they have to wait too long, and the general counseling services usually do not have the expertise in EDs [42]. Notably, digital campus-based interventions have demonstrated effectiveness in decreasing EDs symptoms and enhancing access to support [46].

Although several studies have examined ED risk among medical students [17,24,25], few studies have explored how it intersects with symptoms of depression and anxiety within the Saudi context [16,29]. Identifying such overlap may help support early screening and guide targeted interventions for students. We opted to conduct this study about EDs in this particular group—namely, medical students—for various reasons. One of these reasons is the limited number of studies on this topic in general, and more so among medical students, in Saudi Arabia [16]. Another reason is that EDs tend to be more common in individuals with anxiety and depression [19,47,48] and that medical students have a higher level of anxiety and depression [27,30].

2. Materials and Methods

2.1. Study Design, Setting, and Participants

We conducted a cross-sectional study at King Saud University in Riyadh, Saudi Arabia, between November 2024 and June 2025. Our participants were first- to fifth-year male and female medical students. We excluded students from other colleges. We used convenience sampling. Using the formula n = (Z² × P × [1 − P])/d² n = ([1.96²] × [0.5] × [1 − 0.5])/(0.05²), 385 students were estimated to be needed with Z = 1.96, P = 0.5, (1 − P) = 0.5, and d = 0.05. The final sample in the study comprised 425 participants.

2.2. Study Instrument

The study tools consisted of a questionnaire developed by the research team, the Eating Attitudes Test-26 (EAT-26), the Patient Health Questionnaire-9 (PHQ-9), and the Generalized Anxiety Disorder-7 scale (GAD-7).

The questionnaire included questions on age, gender, academic year, place of residence (with family or in student housing), as well as height and weight which the research team used to calculate their body mass index (BMI).

The EAT-26 is a self-administered screening tool for EDs [49]. The internal consistency and validity of it has been demonstrated, with a Cronbach’s alpha of 0.85 [50,51]. The EAT-26 comprises two sections that measure different aspects of EDs behavior, such as the use of laxatives and vomiting [52]. It includes Part A (26 questions) and Part B (Behavioral Questions). Responses are evaluated on a 6-point scale (Always—3, Usually—2, Often—1, other responses—0). For question number 26 only, responses are interpreted as sometimes—1, rarely—2, never—3, and other answers—0. Participants are considered at risk of EDs if the total score is 20 or higher, or if they answered yes to any of the Part B questions [49].

The PHQ-9 is a depression screening tool that consists of nine questions, with the scoring ranging from 0 to 27 [53]. It exhibits high consistency and reliability, as well as a Cronbach’s α of 0.89 [53]. The total PHQ-9 scores from 1 to 4 indicate minimal depression, scores from 5 to 9 reflect mild depression, scores from 10 to 14 represent moderate depression, scores from 15 to 19 represent moderately severe depression, and scores from 20 to 27 denote severe depression [53].

The GAD-7 is another screening tool to screen for generalized anxiety, with the scoring ranging from 0 to 21. Higher scores indicate more severe anxiety [54]. The total GAD-7 scores from 0 to 4 indicate minimal anxiety, scores from 5 to 9 reflect mild anxiety, scores from 10 to 14 represent moderate anxiety, and scores from 15 to 21 denote severe anxiety. GAD-7 exhibits a good internal consistency, a Cronbach’s α of 0.92, a specificity of 82%, and a sensitivity of 89% [54].

The study tool was created using Google Forms. The link was sent to the participants via WhatsApp private messages rather than public or common WhatsApp groups. This method of distributing the link ensured that only the appropriate participants received it. We obtained the participants’ phone numbers from the College of Medicine’s class leaders of each year.

Ethical Consideration: The study was approved by the Institutional Review Board at the College of Medicine at King Saud University (IRB number E-24-9235; approval date: 24 November 2024; expiry date: 24 November 2025). This study was conducted in accordance with the ethical principles of the Declaration of Helsinki.

Consent to Participate: All participants provided electronic informed consent to participate in the study.

Data Analysis: The R Software version 4.4.0 (the R Foundation of Statistical Computing, Vienna, Austria) was used to analyze the data. The Cronbach’s alpha coefficient was used to measure the questionnaire scales’ internal consistency to assess their reliability; an alpha of 0.7 or higher was deemed satisfactory. The Shapiro test and visual examination of the Q–Q plots were used to assess the normality of continuous variables, which are displayed as means ± standard deviations (SDs) and medians. A bivariate analysis was conducted using the Mann–Whitney test, as the normality assumptions were not met. Frequency (%) is used to represent categorical variables. The difference between proportions was analyzed using Fisher’s exact test and the chi-square test. A logistic regression analysis was used to determine the relationship between various factors and ED risk, including all clinical variables in the multivariate logistic regression. We checked for multicollinearity using the Variance Inflation Factor (VIF); all VIF values were <5, indicating no significant multicollinearity. The reference category for BMI was normal weight; for gender, it was male; for academic year, it was 1st year; and for living situation was with parents. The two-tailed p-value was less than the traditional value of 0.05, implying that it is statistically significant.

3. Results

3.1. Sociodemographic and Anthropometric Characteristics of the Study Participants

A total of 1375 students were contacted and 425 participants completed the survey, yielding a completion rate of 30.9%. The mean age of participants was 21.23 ± 1.50 years (minimum: 18; maximum: 26); 47.8% were males and 52.2% were females. Approximately 19.3% of the students were overweight, and 12.7% were obese. The mean height, weight, and BMI were 166.78 ± 9.56, 67.83 ± 20.60, and 24.14 ± 6.24, respectively. It was also found that 31.3% of the students were in the third year, and 20.5% were in the second year. Most students, 91.1%, lived with their parents, while 7.1% lived in dormitories. The mean EAT-26 score was 15.53 ± 11.95. See

Table 1. Sociodemographic and anthropometric characteristics of the study participants.

	Total N = 425
Age	21.23 ± 1.50, 21
Gender
Male	203 (47.8%)
Female	222 (52.2%)
Weight (kg)	67.83 ± 20.60, 63
Height (cm)	166.78 ± 9.56, 167
BMI (kg/m2)	24.14 ± 6.24, 23.1
Underweight	46 (10.8%)
Normal	243 (57.2%)
Overweight	82 (19.3%)
Obese	54 (12.7%)
Academic year
1st year	75 (17.6%)
2nd year	87 (20.5%)
3rd year	133 (31.3%)
4th year	68 (16.0%)
5th year	62 (14.6%)
Living situation
With parents	387 (91.1%)
Dorms/student housing	30 (7.1%)
Other	8 (1.9%)
EAT-26 mean score	15.53 ± 11.95, 13

BMI, body mass index; EAT-26, 26-item Eating Attitudes Test.

for the sociodemographic and anthropometric characteristics of the participants.

3.2. GAD-7 and PHQ-9 Scores

Table 2. Statistics of GAD-7 (anxiety) and PHQ-9 (depression) scores.

	Cronbach’s Alpha	Mean ± SD, Median	Min–Max
GAD-7	0.92	7.90 ± 5.60, 7	0–21
PHQ-9	0.87	8.82 ± 6.10, 8	0–27

GAD-7, Generalized Anxiety Disorder 7-item scale; PHQ-9, Patient Health Questionnaire 9-item scale.

displays the GAD-7 and PHQ-9 results. The PHQ-9 scale had good internal consistency with a Cronbach’s alpha of 0.87, whereas the GAD-7 scale had excellent internal consistency with a Cronbach’s alpha of 0.92. The median GAD-7 and PHQ-9 scores were 7.90 ± 5.60 and 8.82 ± 6.10, respectively. Sorting the results according to the severity of anxiety symptoms, 134 (31.53%) of the 425 participants who completed the GAD-7 reported no anxiety symptoms, 150 (35.29%) had mild anxiety symptoms, 79 (18.59%) had moderate anxiety, and 62 (14.59%) had severe anxiety. The PHQ-9 revealed that 116 (27.29%) participants had no depression, 135 (31.76%) participants had mild symptoms of depression, 102 (24.00%) had moderate symptoms, 46 (10.82%) had moderately severe symptoms, and 26 (6.12%) had severe symptoms.

3.3. Prevalence and Association of Eating Disorder Risk with Sociodemographic Characteristics

Table 3. Prevalence and relationship between eating disorder risk and socio-demographic characteristics.

	Low Risk N = 214	High Risk N = 211	p-Value
Age	21.21 ± 1.45, 21	21.25 ± 1.55, 21	0.788 a
Gender			0.589 b
Male	105 (49.1%)	98 (46.4%)
Female	109 (50.9%)	113 (53.6%)
BMI (kg/m2)			<0.001 * b
Underweight	33 (15.4%)	13 (6.2%)
Normal	130 (60.7%)	113 (53.6%)
Overweight	35 (16.4%)	47 (22.3%)
Obese	16 (7.5%)	38 (18.0%)
Academic year			0.886 b
1st year	36 (16.8%)	39 (18.5%)
2nd year	42 (19.6%)	45 (21.3%)
3rd year	67 (31.3%)	66 (31.3%)
4th year	38 (17.8%)	30 (14.2%)
5th year	31 (14.5%)	31 (14.7%)
Living situation			0.233 c
With parents	198 (92.5%)	189 (89.6%)
Dorms/student housing	11 (5.1%)	19 (9.0%)
Others	5 (2.3%)	3 (1.4%)
Prevalence of eating disorders risk based on EAT-26
Part A	297 (69.9%)	128 (30.1%)
Part B	258 (60.7%)	167 (39.3%)
Entire	214 (50.4%)	211 (49.6%)

BMI, body mass index; * p < 0.05. ^a, Mann-Whiteny U test; ^b, Chi-square test; ^c, Fisher’s exact test.

shows prevalence and relationship between eating disorder risk and socio-demographic characteristics. Based on the assigned score from the EAT-26, Part A, it was estimated that 30.1% of medical students are at risk for ED and should seek evaluation by a specialist for further diagnosis. According to the accepted scores on behavioral questions from the EAT-26 test, Part B, it was estimated that 39.3% of medical students had a risk of developing an ED. Based on the overall scores and interpretation of the EAT-26 questionnaire, it was found that almost half (49.6%) of the medical students met at least one criterion that could indicate the probable existence or susceptibility to an ED. These individuals should see a specialist for further diagnosis. The Cronbach’s alpha coefficient of the EAT-26 was 0.86, demonstrating good internal consistency.

Univariate analysis was conducted to determine the association of various factors with ED risk. A statistically significant difference was observed in the prevalence of ED risk between students of different BMI groups. The high-risk group had more obese students (18%) compared to the low-risk group (7.5%). On the contrary, underweight students were less represented in the high-risk group at 6.2% compared to 15.4% in the low-risk group. No statistically significant differences were found in the prevalence of ED risk between males and females (p > 0.05) or regarding age, academic year, or living situation.

3.4. Association of Eating Disorder Risk with Anxiety and Depression Disorder Symptoms

Table 4. Relationship between GAD-7 (anxiety) and PHQ-9 (depression) scores and eating disorder risk among study participants.

	Eating Disorder Risk		p-Value
	Low Risk N = 214	High Risk N = 211
GAD-7
1. Feeling nervous, anxious, or on edge	1.12 ± 0.88, 1	1.34 ± 0.98, 1	0.020 * a
2. Not being able to stop or control worrying	0.95 ± 0.92, 1	1.35 ± 0.97, 1	<0.001 * a
3. Worrying too much about different things	1.23 ± 0.94, 1	1.61 ± 1.04, 1	<0.001 * a
4. Trouble relaxing	0.99 ± 0.94, 1	1.35 ± 1.02, 1	<0.001 * a
5. Being so restless that it is hard to sit still	0.65 ± 0.84, 0	0.99 ± 1.03, 1	<0.001 * a
6. Becoming easily annoyed or irritable	0.98 ± 0.92, 1	1.26 ± 1.02, 1	0.004 * a
7. Feeling afraid as if something awful might happen	0.79 ± 0.91, 1	1.24 ± 1.06, 1	<0.001 * a
Total score	6.70 ± 5.23, 6	9.12 ± 5.71, 9	<0.001 * a
Severity			<0.001 * b
No anxiety (0–4)	83 (38.79%)	51 (24.17%)
Mild anxiety (5–9)	83 (38.79%)	67 (31.75%)
Moderate anxiety (10–14)	27 (12.62%)	52 (24.64%)
Severe anxiety (15–21)	21 (9.81%)	41 (19.43%)
PHQ-9
1. Little interest or pleasure in doing things	0.85 ± 0.84, 1	1.19 ± 0.96, 1	<0.001 * a
2. Feeling down, depressed, or hopeless	0.81 ± 0.80, 1	1.20 ± 0.97, 1	<0.001 * a
3. Trouble falling or staying asleep, or sleeping too much	1.19 ± 1.03, 1	1.51 ± 1.09, 2	0.002 * a
4. Feeling tired or having little energy	1.24 ± 0.95, 1	1.61 ± 1.01, 2	<0.001 * a
5. Poor appetite or overeating	0.61 ± 0.80, 0	1.25 ± 1.02, 1	<0.001 * a
6. Feeling bad about yourself or that you are a failure or have let yourself or your family down	0.86 ± 0.98, 1	1.25 ± 1.12, 1	<0.001 * a
7. Trouble concentrating on things, such as reading the newspaper or watching television	0.88 ± 0.93, 1	1.30 ± 1.05, 1	<0.001 * a
8. Moving or speaking so slowly that other people could have noticed. Or the opposite, being so fidgety or restless that you have been moving around a lot more than usual	0.34 ± 0.67, 0	0.69 ± 0.95, 0	<0.001 * a
9. Thoughts that you would be better off dead, or of hurting yourself	0.30 ± 0.68, 0	0.60 ± 0.95, 0	<0.001 * a
Total score	7.07 ± 5.27, 6	10.60 ± 6.37, 10	<0.001 * a
Severity			<0.001 * b
No depression (0–4)	82 (38.32%)	34 (16.11%)
Mild depression (5–9)	74 (34.58%)	61 (28.91%)
Moderate depression (10–14)	34 (15.89%)	68 (32.23%)
Moderately severe depression (15–19)	19 (8.88%)	27 (12.80%)
Severe depression (20–27)	5 (2.34%)	21 (9.95%)

* p < 0.05. GAD-7, Generalized Anxiety Disorder 7-item scale; PHQ-9, Patient Health Questionnaire 9-item scale, ^a, Mann-Whiteny U test; ^b, Chi-square test.

shows a comparison of anxiety disorder symptoms, evaluated through the GAD-7, between students with a high risk of an eating disorder and those with a low risk. Students with a high risk of an eating disorder had higher mean scores on all items and a higher prevalence of moderate and severe anxiety compared to those with a low risk. Specifically, 19.43% of students with a high risk of an eating disorder had severe anxiety while only 9.81% of students with a low risk fell into the severe anxiety category. These differences were statistically significant (p < 0.05).

The high- and low-risk groups were compared in terms of depression symptoms, as assessed by the PHQ-9. High-risk students had higher mean scores for all PHQ-9 items compared to low-risk students. Furthermore, more students with moderate, moderately severe, and severe depression were among those who were at high risk of an eating disorder than those who were at low risk. In particular, only 2.34 percent of the low-risk group experienced severe depression compared to 9.95 percent of participants with a high risk of an eating disorder. These differences were statistically significant (p < 0.05) (Table 4).

3.5. Logistic Regression Analysis of Characteristics Associated with Eating Disorder Risk

The univariable logistic regression results revealed that BMI, anxiety, and depression were significantly associated with the risk of eating disorders. Underweight students were less likely to have eating disorders than students with a normal weight (OR = 0.45; p = 0.024). However, obese students were 2.73 times more likely to have eating disorders than students with a normal weight (OR = 2.73; p = 0.002). Students who had high anxiety (OR = 1.08; p < 0.001) and depression (OR = 1.11; p < 0.001) scores were more likely to have eating disorders.

According to the multivariate logistic regression model, depression and BMI were significantly associated with the likelihood of having an eating disorder. Overweight students were 1.86 times more likely to have eating disorders than students with a normal weight (OR = 1.86; p = 0.022). Similarly, obese students were 3.08 times more likely to have eating disorders than students with a normal weight (OR = 3.08; p = 0.001). On the contrary, underweight students were less likely to have eating disorders than students with a normal weight (OR = 0.42; p = 0.011). Students with a high depression score were more likely to have eating disorders (OR = 1.11; p < 0.001) Although anxiety was associated with eating disorder risk in univariate analysis, it became non-significant after adjustment for depression, suggesting overlapping psychological constructs (

Table 5. Predictors of eating disorders among the participants (logistic regression analysis).

	Univariate			Multivariate
	OR	95% CI	p-Value	OR	95% CI	p-Value
Age	0.98	0.86–1.12	0.788
Gender
Male	Reference	-	-
Female	1.11	0.76–1.63	0.589
BMI (kg/m2)
Underweight	0.45	0.22–0.88	0.024 *	0.42	0.19–0.85	0.020 *
Normal weight	Reference	-	-	-	-	-
Overweight	1.54	0.93–2.57	0.091	1.86	1.10–3.17	0.022 *
Obese	2.73	1.47–5.28	0.002 *	3.08	1.61–6.15	0.001 *
Academic year
1st year	Reference	-	-
2nd year	0.99	0.53–1.84	0.972
3rd year	0.91	0.52–1.60	0.742
4th year	0.73	0.38–1.41	0.347
5th year	0.92	0.47–1.81	0.816
Living situation
With parents	Reference	-	-
Dorms/student housing	1.81	0.85–4.03	0.131
Others	0.63	0.13–2.60	0.529
Total anxiety score	1.08	1.05–1.13	<0.001 *	1.01	0.96–1.07	0.617
Total depression score	1.11	1.07–1.15	<0.001 *	1.11	1.06–1.17	<0.001 *

OR, odds ratio; CI, confidence interval; BMI, body mass index; * p < 0.05.

).

4. Discussion

Our study aimed to examine the ED risk among medical students using the EAT-26. We found a notably high ED risk; namely, 49.6% of our participants. This high figure observed in our study is consistent with the elevated rates documented internationally among medical students. For instance, in Ukraine, the ED risk was reported to be 36.9% [55], while in Romania, there was a 37.1% overall ED risk among medical students [56]. Our high figure is also consistent with the regional literature; a Saudi study on medical students reported a 32.1% ED risk [16]. Although the authors used a different tool than ours to screen for EDs—namely, the “Sick, Control, One stone, Fat, and Food” (SCOFF) tool—a study of female university students in Jazan, Saudi Arabia, found 47.9% to be at risk [57]. However, our figure (49.6%) is higher than those of some other studies; the global pooled average for medical students’ ED risk is estimated at 10.5% [25,58]. Our figure is also higher than European estimates, such as 23.8%, which was reported in a European multi-center study [59]. We hypothesize that the reason for our figure to be higher than some other studies is multifactorial. First, in our study, we used convenience sampling, which might introduce a certain degree of selection bias, meaning that students who are already symptomatic (in terms of anxiety, depression, or EDs) might be more interested in participating in the study, potentially raising the observed rate compared to a truly random sample. Second, our high figure could be justified by the global trends of EDs, where the incidence of EDs appears to be increasing over time [2,25]. That said, we recommend future Saudi studies to consider utilizing more rigorous sampling methods, such as random sampling techniques, to better assess the risk of EDs. In addition, moving beyond using screening tools and utilizing structured or semi-structured interviews could also yield more precise results.

In our study, high BMI strongly predicted ED risk. Obese and overweight students were 3.52 and 2.16 times more likely to be at risk compared to those with a normal or underweight BMI. Reports of similar links were observed in the United States [43], Bangladeshi [18], and Middle Eastern studies [25]. Thus, overweight and obese students should be the groups targeted for future screening and intervention.

Depressive symptoms were also a strong predictor; for each PHQ-9 point, the risk of an ED increased by 11%. A meta-analysis involving medical students across multiple countries confirmed that depressive symptoms were significantly comorbid with ED symptomatology [25]. The findings of this study [25] corroborate the view that depression is a key factor in ED development. Similarly, a cross-sectional study at the Lebanese American University, utilized the PHQ-9 and found an association between emotional distress and disordered eating [14]. However, they did not isolate depression as an independent predictor in this study [14]. Such findings highlight the importance of screening for both depression and EDs in such populations.

High-risk students had higher anxiety scores in the univariate analysis, but not the multivariate analysis. In contrast to our findings, a study in Anhui, China, found that anxiety, together with depression and BMI, remained a significant predictor of ED risk among female medical students, after adjusting for covariates [36]. Likewise, a study among medical students found that both anxiety and depression, measured via the Hospital Anxiety and Depression Scale (HADS) were significantly associated with increased ED risk among medical students [7]. We hypothesize that multiple reasons led to the discrepancy between our findings and those from other studies, namely, that anxiety was not a significant predictor in our multivariate model. First, in our study, we specifically screened for GAD. Compared to our study, other studies screened for anxiety in the general sense, rather than GAD [36]. Nonetheless, some studies screened for other anxiety disorders—for example, social anxiety disorder—which has a strong link to EDs [25,60]. Another reason could be related to the screening tool used; in our study, we used the GAD-7, while other studies used either different scales [7] or multiple scales [25]. Another explanation might be related to the students underreporting their anxiety symptoms as some of them might consider it normal or stigmatizing to report [41].

We did not find a statistically significant association between gender and ED risk in the univariate analysis. However, females showed a tendency toward a higher risk, though it did not reach statistical significance. These results show that gender could act as a confounding variable, with its effect influenced by other factors such as BMI or psychological distress. The findings from international and regional research support a gender-based disparity in ED risk, with females exhibiting higher prevalence. A study among nutrition students in the United States, found that female students were more prone to body image dissatisfaction, a key contributor to disordered eating [43]. A study at the Lebanese American University, reported a high prevalence of EDs among female medical students [14]. In the Gulf region, a similar study found that female students at King Abdulaziz University in Jeddah, Saudi Arabia, had a higher risk of developing EDs compared to males [16]. Even though it did not reach statistical significance in our study, the higher ED risk among females may suggest a meaningful pattern. From a clinical standpoint, gender-specific vulnerabilities should be considered in the future when designing screening tools and preventive interventions for EDs. Further, future Saudi studies should target an equal gender representation among the sample and investigate the related factors. Such factors could include internalized body ideals, concerns about physical appearance, and gender-related stigmas.

Worth mentioning, our findings should be interpreted with cautions. In terms of internal validity, our study is cross-sectional, which might hinder establishing causal relationships between EDs, depression, and anxiety. While we controlled for key confounders such as age, gender, BMI, academic year, and living situation, residual confounding by unmeasured factors remains possible. Additionally, the use of self-administered measures for EDs, depression, and anxiety introduces the potential for recall bias, which may affect the accuracy of the observed associations. In terms of external validity, the ability to generalize our findings might be limited by the fact that participants were recruited from a single academic institution in Saudi Arabia, and had specific demographics. Therefore, when applying these results to broader populations, caution should be exercised.

Strengths and Limitations

This is one of the few Saudi studies to assess EDs along with depression and anxiety in medical students, using validated tools and a reasonable sample size. In essence, the study provides physicians, policymakers, and staff members at academic institutions with valuable insights.

Even so, there are limitations to this study. One limitation is that the type of study was cross-sectional, which hinders the ability to make causal inferences. Future studies should employ longitudinal study designs, which can follow the changes over time and help identify potential causal pathways between risk factors and the development of EDs. Using convenience sampling is another limitation in our study, which might have caused selection bias and limited the generalizability of our findings as the selected students may not fully represent the entire student population in Saudi Arabia. Future Saudi studies should consider using random or stratified sampling techniques, thus making the sample more representative and reducing the risk of selection bias. In addition, our sample consisted of students invited through WhatsApp, which could introduce a selection bias, as students with an interest in mental health may be more willing to participate, and hence may be overrepresented. Future studies should consider this limitation and find a better method to obtain a more balanced representative sample. Additionally, because the research took place at one university, the conclusions might lack generalizability. Future research should consider conducting such a study in multicenter settings across different regions in Saudi Arabia to ensure broader representation and allow for comparisons between different institutions and academic environments. Another limitation is that the study was conducted exclusively among medical students, which further limits the generalizability of the findings to the wider university population. Future studies should include representative students from various colleges and various institutions to allow broader comparisons. Additional limitation in our study is that, due to the methodology used—namely, the structure of the developed survey—we were able to calculate the completion rate but not the response rate; in particular, our study tool was composed of mandatory questions, and the participants could not submit the survey until they completed all the questions. Future studies could find a better way to create the study tool to allow the response rate to also be calculated. Furthermore, we relied on the EAT-26, which is a screening tool but is not diagnostic; i.e., the EAT-26 measures risk but cannot confirm ED diagnoses. As such, we recommend future Saudi studies to consider using a more rigorous methodology, such as structured or semi-structured diagnostic interviews to assess the topic in a more precise way. Another limitation is that we only screened for GAD and did not examine other anxiety disorders, such as social anxiety disorder, which is commonly comorbid with EDs [60]. As such, we recommend that future Saudi studies evaluate other forms of anxiety, especially social anxiety disorder. Another limitation in our study is that we used self-reported data, which may be susceptible to recall bias. Clinical interviews and assessments should be included in future studies along with objective data such as medical and behavioral records, reducing bias and improving the accuracy of research results. We also calculated the BMI based on the height and weight reported by the participants, which could lead to inaccurate data compared to if the research team measured these variables themselves. Future studies should consider taking such measurements directly, rather than depending on the participants’ reported values. A last limitation to acknowledge is that certain potential confounders were not assessed; for example, media exposure (including social media) [61] and the level of academic stress. As such, we recommend that future Saudi studies consider assessing these factors.

5. Conclusions

Our study aimed to assess the ED risk of medical students and analyze the associated factors with a focus on depression, anxiety, and demographic characteristics. The findings indicated that medical students have a higher ED risk than that reported in previous international studies, as shown in a systematic review and meta-analysis [58]. We also found that students with a high BMI and depressive symptoms are more likely to be at risk of EDs.

Considering our results, we suggest employing early screening, introducing specific psychoeducational programs, and creating culturally appropriate interventions to address the body and mental health-related concerns of those with EDs. Such strategies can help promote healthier eating behaviors and improve the mental well-being of those affected by, or at risk of, EDs in Saudi Arabia. Although our study sample consisted exclusively of medical students, we believe the recommendations mentioned earlier could benefit university students in general, not only those in medical colleges.