Eating disorders, nutrient intake, and cardiometabolic risks: a cross-sectional study in Turkish adults

Abstract

Background

Eating disorders can cause endocrine-metabolic complications, and comorbidities, potentially leading to obesity, type 2 diabetes, and cardiovascular disease (CVD). This study aimed to examine the associations between eating disorders, nutrient intake, and the risk of CVD and diabetes in adults aged 40–75 years.

Methods

This study included 3181 healthy Turkish adults aged 40–75 years. CVD risk was assessed using the Systematic Coronary Risk Evaluation (SCORE) and Atherosclerotic Cardiovascular Disease (ASCVD) Risk Estimator Plus. Type 2 diabetes risk was evaluated using the Finnish Diabetes Risk Score (FINDRISC), eating disorders were assessed with the Eating Disorder Examination Questionnaire (EDE-Q), and dietary intake was measured via 24-hour food recall.

Results

In the sample (50.4% male, mean age 50.33 ± 7.79 years), individuals at borderline 10-year ASCVD risk (estimated 5–7.4% 10-year risk) presented significantly greater body weight, body mass index, neck circumference, and waist-hip ratio than did those at low risk (p < 0.001). Additionally, the total EDE-Q scores of individuals with borderline 10-year ASCVD risk were significantly lower than those of individuals with low risk (p < 0.001). Total EDE-Q scores of individuals at very high risk of diabetes were higher than those of individuals at moderate risk (p < 0.05). SCORE and ASCVD risk were significantly weakly positively correlated with dietary cholesterol (r = 0.050, r = 0.064, respectively) and protein intake (r = 0.093, r = 0.142, respectively) (p < 0.05). Interestingly, and contrary to expectations based on previous literature, the total EDE-Q score was weakly negatively correlated with both SCORE (r = -0.101) and ASCVD (r = -0.092) risk scores (p < 0.001). This unexpected correlation could reflect selection bias, as healthier individuals exhibiting subclinical symptoms may have been more inclined to participate. In contrast, it was weakly positively correlated with the FINDRISC score (r = 0.163) (p < 0.001).

Conclusions

Increased anthropometric measurements were associated with elevated risks of CVD and diabetes. Higher dietary cholesterol and protein intake showed weak positive associations with CVD risk factors. Notably, eating disorders were positively associated with diabetes risk but unexpectedly showed a negative association with CVD risk. These findings suggest that disordered eating behaviors may influence cardiometabolic outcomes. Further research is warranted to explore the mechanisms underlying these associations. Given the cross-sectional nature of this study, causality cannot be inferred.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-025-24295-7.

Background

Recently, disordered eating behaviors have emerged as a significant public health concern, influenced by increasing societal pressures related to diet, health, and body image [1, 2]. Eating disorders (EDs) are psychiatric conditions characterized by abnormal eating and weight control-related behaviors, often accompanied by concerns about body shape and weight. The latest edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM)−5 identifies six major feeding and EDs: Anorexia Nervosa (AN), Bulimia Nervosa (BN), Binge Eating Disorder (BED), Avoidant/Restrictive Food Intake Disorder, pica, and rumination disorder [3]. The prevalence of EDs varies across populations. A study conducted by Qian et al. (2022) demonstrated that the lifetime prevalence of EDs was 1.69% and the prevalence of BED (1.53%) was higher than AN (0.16%) and BN (0.63%) [4]. In Türkiye, it is reported that 33.6% of women and 6.3% of men are on a diet; although dieting alone does not constitute an eating disorder, the pervasive focus on achieving a thin body image may increase vulnerability to disordered eating behaviors over time [5]. Additionally, most studies on EDs focus on young people, understanding these disorders in middle-aged and older adults is essential due to their heightened vulnerability to both EDs and cardiometabolic complications.

Although individuals with EDs may present with primarily psychiatric disorders, they are also associated with several physiological complications, including cardiomyopathy, hypertension, hormonal disturbances, and metabolic disorders. These complications may contribute to development of obesity, type 2 diabetes mellitus (T2DM), and cardiovascular diseases (CVDs) [6, 7]. Importantly, ED subtypes differ in how they influence nutrient intake and cardiometabolic outcomes. AN is typically associated with severe caloric restriction and low BMI, leading to complications such as bradycardia, hypotension, and reduced bone mineral density, despite often normal or favorable lipid profiles [8]. BN, involving recurrent binge eating and purging behaviors, can result in electrolyte imbalances, particularly hypokalemia, and increased cardiovascular strain due to compensatory mechanisms [9]. In contrast, BED is strongly associated with overweight and obesity, insulin resistance, dyslipidemia, and increased risk of T2DM and CVDs [10]. Thus, each subtype may influence cardiometabolic risk through distinct pathways ranging from nutrient deficiencies to excess energy intake and metabolic dysregulation warranting subtype-specific attention in population-based studies [11–13]. A growing body of evidence suggests a complex, bidirectional relationships between EDs and metabolic disorders, particularly T2DM [14–16]. Rates of EDs in individuals with T2DM have been reported to range between 1.4% and 25.0%, with BED being the most common subtype in this population [17–20].

EDs have also been linked to an increased risk of CVDs through behavioral mechanisms such as emotional eating, physical inactivity, and poor dietary quality, as well as through common risk factors including obesity, hypertension, and dyslipidemia [21–23]. Some evidence also suggests increased ED risk in patients with hypertension [24]. Weight loss interventions in obese adults with BED have been found to reduce CVDs risk by improving total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides, HbA1C, serum glucose, and heart rate [25]. Furthermore, numerous studies have shown that nutrition is effective in reducing the risk of hypertension and CVDs [26–28]. However, individuals with CVDs may also be more susceptible to developing EDs due to psychological stress, medication side effects, and strict dietary regimens. These factors may contribute to appetite dysregulation, increase the appeal of restricted foods, and foster disordered eating behaviors. These bidirectional relationship between EDs and cardiometabolic diseases are likely mediated by several mechanisms. Individuals with existing CVDs may become more vulnerable to EDs due to medication side effects and strict dietary regimens. This stress may precipitate or worsen disordered eating behaviors, such as emotional eating or loss of control eating. In addition, pharmacological treatments, including insulin, corticosteroids, and some antihypertensives, may affect appetite, weight regulation, or mood, contributing to eating disturbances. Furthermore, strict dietary regimens may trigger feelings of restriction, especially in individuals predisposed to EDs. These factors highlight the need for a holistic approach when addressing cardiometabolic risk in populations vulnerable to disordered eating [15, 29, 30].

It is crucial to assess the risk of developing T2DM and CVDs in individuals with EDs to enable the implementation of appropriate precautions, early diagnosis and effective treatment strategies. Despite the increasing recognition of EDs as a significant public health concern, the majority of existing research has concentrated on adolescents and young adults [26, 31], resulting in a critical gap in the literature regarding how EDs impact health outcomes in middle-aged and older adults [32–35]. Middle-aged and older adults (aged 40–75) are also at elevated risk for cardiometabolic conditions such as T2DM and CVDs, making this population especially relevant for investigating these complex associations. In older adults, psychological stress related to disease burden and lifestyle changes may exacerbate disordered eating behaviors, such as emotional eating. Additionally, medications such as insulin, corticosteroids, and certain antihypertensives can influence appetite, body weight, or mood, potentially contributing to disordered eating patterns [36, 37]. This study aims to examine the associations between disordered eating behaviors, nutrient intakes, and estimated risks of CVD and T2DM in Turkish adults aged 40–75 years.

Materials and methods

Study design and sample selection

The sample size of the study was calculated as 3780 individuals in total, based on 99% power, an effect size of 0.085 and α = 0.05, using the biostatistical power analysis program G-Power. This cross-sectional study was conducted in Türkiye between June and October 2023 using a voluntary sampling method. To enhance representativeness and reduce sampling bias, participants were recruited from a range of family health centers located in urban and rural areas across different regions of Türkiye (e.g., Marmara, Central Anatolia, Southeast Anatolia). Efforts were made to include individuals from varied socioeconomic backgrounds by selecting centers in both high and low-income districts. The exclusion criteria included individuals under the age of 40 years or over the age of 75 years; those with a history of chronic CVDs such as stroke, myocardial infarction, coronary artery bypass, percutaneous coronary intervention, coronary angiography, heart failure, or peripheral artery disease; those with cancer, liver, kidney diseases or diabetes; those with diagnosed EDs; those in a special diet program; pregnant or breastfeeding women; and those with a history of alcohol or drug addiction. In addition, individuals with previously diagnosed EDs were excluded to focus on disordered eating behaviors in the general population and to avoid potential confounding effects associated with ongoing treatment or severe psychopathology. A total of 3533 healthy adults aged 40 to 75 years, who did not meet any of the exclusion criteria, had their blood lipid parameters measured within the past three months, and agreed to participate, were included in the study. The recruitment of participants is demonstrated in the flow diagram (Fig. 1).

Fig. 1.

Participants’ recruitment flow chart

Data collection and assessment

Data were obtained through a survey using a face-to-face interview method conducted with individuals residing in Türkiye who had consented to participate in the study. Nutritionists involved in data collection received structured training sessions before the study. This training covered the use of standardized questionnaires, interview techniques to minimize interviewer bias, proper use of measurement tools, and adherence to uniform procedures. The questionnaire collected data on individuals’ general descriptive characteristics, physician-diagnosed disease status, vitamin and mineral supplement intake, anthropometric measurements, physical activity levels, nutritional habits, smoking and alcohol consumption, CVDs and T2DM risk, EDs, and 24-hour retrospective food consumption records.

Assessment of anthropometric measurements

Body weight (kg) and height (cm) were measured with participants in light clothing and without shoes, using a calibrated digital scale (Sinbo SBS-4452) and a portable stadiometer, respectively. Circumference measurements, including neck (NC), waist (WC), and hip circumference (HC), were taken using a non-elastic tape measure (SECA 206), with participants standing upright and arms relaxed at their sides. BMI (kg/m²) and waist-to-hip ratio (WHR) were calculated. BMI was calculated as body weight (kg) divided by height squared (m²), and WHR was calculated as WC divided by HC. The WHO classification for BMI was used [38]. WC was evaluated according to the WHO classification. While WC ≥ 80 cm for women and ≥ 94 cm for men is considered risky in terms of metabolic complications, WC ≥ 88 cm for women and ≥ 102 cm for men is considered a high risk [39]. NC > 37 cm in men and > 34 cm in women is considered a risk factor for overweight and obesity [40].

Assessment of physical activity level

A brief two-question physical activity assessment tool was used during face-to-face interviews to assess participants’ physical activity levels. The total score, which was obtained by summing up the points obtained according to the answers provided for each question, was listed as “inadequately active” or “adequately active” for points of 0–3 or ≥ 4, respectively [41]. Although the two-question physical activity tool used in this study does not have a formal validation study in Turkish populations, it has been utilized in several thesis-based and community-level studies due to its practicality and ease of application in field settings [42]. It was considered suitable for this study given the large sample size and the need for brief, face-to-face assessment tools.

Assessment of cardiovascular risk status

The Systematic Coronary Risk Evaluation (SCORE) and the Atherosclerotic Cardiovascular Disease (ASCVD) Risk Estimator Plus are tools that estimate an individuals’ optimal risk, 10-year risk, and lifetime risk. The SCORE program developed by the European Society of Cardiology [43, 44] and ASCVD program developed by the American College of Cardiology were used to assess risk [44–46].

Risk calculations were based on responses to questions about age; sex; systolic and diastolic blood pressure; total cholesterol, HDL-C, and LDL-C levels; diabetes history; smoking status; hypertension treatment; statin use; and aspirin use. Blood pressure data required for the SCORE and ASCVD risk calculations were obtained via a standardized protocol. Systolic blood pressure was measured in both arms while the participants were in a seated position with a calibrated sphygmomanometer after a minimum of 15 min of rest. The mean of the two measurements was taken and recorded in mm/Hg. The biochemical parameters required for both scales (total cholesterol, HDL-C, LDL-C) were obtained from participants’ files on the basis of their biochemical parameters recorded within the past three months.

According to the SCORE calculation, < 1% indicates of low risk, 1–5% represents moderate risk, ≥ 5% and < 10% signifies high risk, and ≥ 10% indicates of very high risk [44]. According to the ASCVD calculation, a total score of less than 5% is considered low risk, a score between 5% and 7.4% is considered borderline risk, a score between 7.5% and 19.9% is considered moderate risk, and a score greater than 20% is considered high risk [47]. These risk estimation tools were selected due to their wide clinical use, validation in large population studies, and relevance to the age group included in our study population.

Assessment of type 2 diabetes risk status

To assess the risk of T2DM, the Finnish Diabetes Risk Score (FINDRISC), which has been validated in Turkish and is recommended for use by the Turkish Endocrinology and Metabolism Society, was employed [48, 49]. The scale contains a total of eight questions, including age, BMI, physical activity status, fruit and vegetable consumption, medication use, and family history of diabetes. The total score ranges between 0 and 26. The total score obtained indicates the 10-year risk of developing T2DM and is classified as follows: less than 7 points indicates very low risk, 7 to 11 points indicates low risk, 12 to 14 points indicates moderate risk, 15 to 20 points indicates high risk, and greater than 20 points indicates very high risk [48]. The FINDRISC tool was chosen due to its simplicity, validation in multiple European populations, including Türkiye, and its suitability for estimating 10-year type 2 diabetes risk in adults based on easily obtainable clinical and lifestyle factors.

Assessment of eating disorder status

The Eating Disorder Examination Questionnaire (EDE-Q) was used to assess participants’ ED status [50]. In 2022, Esin et al. conducted a formal validity and reliability study of the Turkish version of the scale [51], confirming its psychometric adequacy and for use in Turkish populations. The short form, comprising 13 questions, includes five subdimensions: eating restraint, shape and weight over-evaluation, body dissatisfaction, bingeing, and purging. EDE-Q scoring was based on the total score for all items on the scale, as well as for each subdimension. For each subdimension of the EDE-Q, a mean score was calculated by dividing the total score of that subscale by the number of items it contained. The total score was then computed by averaging the mean scores of the five subdimension. Higher scores indicate greater levels of eating-related psychopathology [51].

Assessment of nutrient intakes

Nutrient intake was assessed using the 24-hour food recall method, conducted over three non-consecutive days—two weekdays and one weekend day—to enhance accuracy and better reflect usual intake. Before completing the food consumption records, participants were guided by trained researchers who had undergone comprehensive training to ensure consistency and minimize interviewer bias. This training included familiarization with the food portion photo catalog, standardized kitchen measurement tools and interviewing techniques to effectively prompt participants without leading their responses. Participants were guided by researchers and provided with information on portion sizes using a food portion photo catalog and standardized kitchen measurements (water glass, tablespoon, etc.) [52] before completing the food consumption record. The Computer-Assisted Nutrition Program (BeBIS) was used to calculate energy and nutrient intakes obtained from participants’ food consumption records. The program has been widely used in Turkey and is based on validated food composition databases, ensuring reliable nutrient estimates. No extreme energy intake values were identified; therefore, no exclusions based on energy intake were applied. The daily recommended nutrient intakes for Turkish adults were compared with the Türkiye Nutrition Guide-2022, and the percentages were calculated [53]. Energy intake values were examined but none met exclusion criteria based on predefined cutoffs commonly used in dietary assessment studies (e.g., < 800 kcal or > 4000 kcal/day).

Statistical analyses

To evaluate the findings obtained in the study, statistical analysis was conducted using the SPSS (Statistical Package for the Social Sciences) for Windows version 21.1. Descriptive statistics were presented as the mean (X̅) and the standard error (SEM). The frequency data were presented as number (n) and percentage (%). The normality of distribution in the data was tested with the Kolmogorov-Smirnov test. For comparisons involving more than two groups that did not follow a normal distribution, the Kruskal-Wallis test was used. The Mann-Whitney U test with Bonferroni correction was used for pairwise comparisons between groups. Spearman correlation analysis was used to assess the relationships between variables due to the non-normal distribution of the data, and a correlation graph was generated using the RStudio program. Effect sizes offer important insight into practical significance for interpreting statistical results; according to Cohen’s classification, a correlation coefficient (r) of approximately 0.10 is considered weak, 0.30 is considered moderate, and 0.50 or above is considered strong [54]. Although statistically significant, the observed correlations were weak in magnitude, suggesting limited practical or clinical relevance. Relationships between EDs and independent variables were analyzed by multiple linear regression analysis. Although multicollinearity was not detected, nutrient intake variables were excluded from the regression model due to their consistently weak correlations with ED scores, which would contribute minimally to model explanatory power. The results were examined in the 95% confidence interval at the p < 0.05 significance level.

Results

General characteristics of the study population

A total of 3181 individuals (1579 females and 1602 males) were completed in the study. Table 1 shows the general characteristics of the participants. The study population, with a mean age of 50.3 ± 7.8 years, comprised 50.4% males. The study revealed that 77.8% of participants were insufficiently active, and the mean BMI was 26.4 ± 4.3 kg/m².

Table 1.

General characteristics in overall population

Variables		Overall (n = 3181)
Age, years		50.3 ± 7.8
Gender	Female	1579 (49.6)
	Male	1602 (50.4)
Marital status	Married	2606 (81.9)
	Non-married	575 (18.1)
Education	Primary school	944 (29.7)
	High school	1005 (31.6)
	University	1060 (33.3)
	Postgraduate	172 (5.4)
Employment status	Labouer	1200 (37.7)
	Employee	525 (16.5)
	Trades	347 (10.9)
	Retired	407 (12.8)
	Student	7 (0.2)
	Not working	695 (21.8)
Smoking status	Yes	1108 (34.8)
	No	2073 (65.2)
Disease status	Yes	513 (16.0)
	No	2702 (84.9)
Diseases	Obesity	58 (1.8)
	Gastrointestinal system diseases	115 (3.6)
	Thyroid diseases	113 (3.6)
	Neurological-psychological diseases	112 (2.9)
	Other diseases	101 (3.2)
Supplement use	Yes	1188 (37.3)
	No	1993 (62.7)
Supplements	Multivitamin-Mineral	732 (61.6)
	Omega-3	127 (10.6)
	Vitamin D	299 (25.1)
	Probiotic	59 (4.9)
Anthropometric measurements	Body weight (kg)	76.1 ± 14.2
	BMI (kg/m2)	26.4 ± 4.3
	Neck circumference (cm)	35.7 ± 4.7
	Waist-hip ratio	0.8 ± 0.1
Physical activity	Insufficient active	2474 (77.8)
	Enough active	707 (22.2)
Total SCORE points		4.3 ± 0.1
10-year ASCVD risk score		4.3 ± 0.1
Lifetime ASCVD risk score		37.7 ± 0.3
Optimal risk score		2.4 ± 0.1
Total FINDRISC score		9.3 ± 0.1
Total EDE-Q score		2.6 ± 0.04
Eating Restraint		3.7 ± 0.1
Shape and Weight Over-evaluation		3.0 ± 0.1
Body Dissatisfaction		3.8 ± 0.1
Bingeing		2.1 ± 0.1
Purging		0.4 ± 0.02

Data are expressed as number (%) or mean ± standard error (SEM)

ASCVD Atherosclerotic Cardiovascular Disease, BMI Body Mass Index, EDE-Q Eating Disorder Examination Questionnaire, FINDRISC Finnish Diabetes Risk Score, SCORE Systematic Coronary Risk Evaluation

Anthropometric measurements and ED status according to cardiovascular and diabetes risk

Table 2 presents the anthropometric characteristics and ED status of individuals, categorized by their cardiovascular risk levels. The body weight, BMI, NC, and WHR of individuals in the high-SCORE risk category were significantly greater than those of the other groups (p < 0.001). Individuals with moderate SCORE risk had significantly higher total EDE-Q (p < 0.001), eating restraint (p < 0.001), shape and weight over-evaluation (p < 0.001), body dissatisfaction (p = 0.009), purging (p = 0.001) scores compared to those in the high-risk category.

Table 2.

The anthropometric measurements and EDE-Q scores of participants according to CVD risk status

	SCORE					ASCVD
	Low (n = 74)	Moderate (n = 2150)	High (n = 694)	Very high (n = 263)	p-value**	Low (n = 2151)	Borderline (n = 367)	Moderate (n = 493)	High (n = 43)	No risk (n = 127)	p-value**
Body weight (kg)	63.9 ± 1.3	74.7 ± 0.3*	80.4 ± 0.5*	80.1 ± 0.9	< 0.001	74.0 ± 0.3a	81.1 ± 0.7a, i	82.3 ± 0.6	79.4 ± 2.2	71.9 ± 1.1i	< 0.001
BMI (kg/m2)	23.5 ± 0.5	26.0 ± 0.1§	37.3 ± 0.2§	27.9 ± 0.3	< 0.001	26.0 ± 0.1b	27.2 ± 0.2b, j	27.6 ± 0.2	27.3 ± 0.7	25.2 ± 0.4j	< 0.001
Neck circumference (cm)	32.3 ± 0.4	35.1 ± 0.1¥	36.9 ± 0.2¥	37.5 ± 0.3	< 0.001	34.9 ± 0.1c	37.2 ± 0.2c, k	37.9 ± 0.2	38.5 ± 0.67	34.5 ± 0.4k	< 0.001
Waist-hip ratio	0.8 ± 0.01	0.9 ± 0.00γ	0.9 ± 0.00γ	0.9 ± 0.01	< 0.001	0.9 ± 0.002d	0.9 ± 0.004d, l	0.9 ± 0.004	0.9 ± 0.02	0.9 ± 0.008l	< 0.001
EDE-Q, total	2.7 ± 0.4	2.8 ± 0.1‡	2.3 ± 0.1‡	1.9 ± 0.1	< 0.001	2.7 ± 0.1e	2.3 ± 0.1e, m	2.2 ± 0.1	1.9 ± 0.4	3.18 ± 0.26m	< 0.001
Eating Restraint	4.4 ± 0.6	4.03 ± 0.1†	3.1 ± 0.2†	2.6 ± 0.2	< 0.001	3.9 ± 0.1f	3.1 ± 0.2f, n	2.9 ± 0.2	2.6 ± 0.5	4.5 ± 0.5n	< 0.001
Shape and Weight Over-evaluation	3.2 ± 0.5	3.3 ± 0.1β	2.6 ± 0.1β	1.9 ± 0.2	< 0.001	3.2 ± 0.1g	2.5 ± 0.2g, o	2.4 ± 0.2	1.7 ± 0.5	3.8 ± 0.4o	< 0.001
Body Dissatisfaction	3.5 ± 0.5	3.9 ± 0.1φ	3.6 ± 0.2φ	3.1 ± 0.3	< 0.001	3.9 ± 0.1h	3.3 ± 0.2h	3.4 ± 0.2	2.8 ± 0.6	4.2 ± 0.4	< 0.001
Bingeing	1.9 ± 0.5	2.1 ± 0.1	1.9 ± 0.1	1.7 ± 0.2	< 0.001	2.04 ± 0.1	2.0 ± 0.2	1.9 ± 0.2	2.1 ± 0.6	2.7 ± 0.3	0.795
Purging	0.3 ± 0.2	0.5 ± 0.03δ	0.3 ± 0.04δ	0.3 ± 0.1	< 0.001	0.5 ± 0.03	0.3 ± 0.1p	0.3 ± 0.1	0.4 ± 0.2	0.7 ± 0.2p	0.008

Data are expressed as mean ± standard error (SEM)

ASCVD Atherosclerotic Cardiovascular Disease, BMI Body Mass Index, EDE-Q Eating Disorder Examination Questionnaire, SCORE Systematic Coronary Risk Evaluation

^**Kruskal Wallis Test

^{*, §, ¥,‡, †, β, φ, δ, γ} = Different symbols in the same row correspond to significant differences (p < 0.001)

^a−p = Different letters in the same row correspond to significant differences (p < 0.05)

Individuals in the low 10-year ASCVD risk subgroup presented significantly lower body weights, BMIs, NCs, and WHR than those in the borderline subgroup did (p < 0.001). Compared with those with borderline risk, participants with low 10-year ASCVD risk presented significantly higher scores on the total EDE-Q, eating restraint, shape and weight overevaluation, and body dissatisfaction subgroups (p < 0.001, p < 0.001, p < 0.001, p < 0.001, respectively). A comparison of individuals at borderline risk of ASCVD without risk groups revealed that individuals without ASCVD risk had higher total EDE-Q scores (p < 0.001), as well as higher scores on the eating restraint (p = 0.002), shape and weight over-evaluation (p < 0.001), and purging (p = 0.002) subdimensions (Table 2).

The anthropometric characteristics and ED status of individuals according to diabetes risk categories are shown in Table 3. Individuals with moderate diabetes risk had significantly higher anthropometric measurements [body weight, BMI, NC (p < 0.001)], total EDE-Q scores (p = 0.004), and body dissatisfaction (p < 0.001) than those at low risk. Similarly, individuals with very high diabetes risk had significantly higher body weight (p < 0.001), BMI (p < 0.001), total EDE-Q (p = 0.006), body dissatisfaction (p = 0.007) and bingeing (p < 0.001) scores than individuals with moderate diabetes risk did.

Table 3.

The anthropometric measurements and EDE-Q scores of participants according to diabetes risk status

	FINDRISC
	Very low risk (n = 951)	Low risk (n = 1289)	Moderate risk (n = 486)	High risk (n = 387)	Very high risk (n = 68)	p-value*
Body weight (kg)	68.8 ± 0.4	75.9 ± 0.4a	81.2 ± 0.6a, f	85.9 ± 0.8	89.2 ± 1.6f	< 0.001
BMI (kg/m2)	23.6 ± 0.1	26.1 ± 0.1b	28.5 ± 0.2b, g	30.4 ± 0.2	33.0 ± 0.5g	< 0.001
Neck circumference (cm)	33.9 ± 0.1	35.7 ± 0.1c	37.1 ± 0.2c	37.8 ± 0.2	37.7 ± 0.5	< 0.001
Waist-hip ratio	0.8 ± 0.003	0.9 ± 0.002	0.9 ± 0.004	0.9 ± 0.01	0.9 ± 0.01	< 0.001
EDE-Q, total	2.2 ± 0.1	2.5 ± 0.1d	2.9 ± 0.1d, h	3.4 ± 0.1	4.2 ± 0.4h	< 0.001
Eating Restraint	3.5 ± 0.2	3.6 ± 0.1	3.9 ± 0.2	4.1 ± 0.2	5.2 ± 0.7	< 0.001
Shape and Weight Over-evaluation	2.5 ± 0.1	2.9 ± 0.1	3.1 ± 0.2	3.9 ± 0.2	4.3 ± 0.6	< 0.001
Body Dissatisfaction	2.7 ± 0.1	3.6 ± 0.1e	4.7 ± 0.2e, i	5.5 ± 0.2	6.3 ± 0.7i	< 0.001
Bingeing	1.6 ± 0.9	2.0 ± 0.9	2.2 ± 0.2j	2.9 ± 0.2	4.8 ± 0.7j	< 0.001
Purging	0.4 ± 0.1	0.4 ± 0.04	0.4 ± 0.1	0.5 ± 0.1	0.5 ± 0.2	0.156

Data are expressed as mean ± standard error (SEM)

BMI Body Mass Index, EDE-Q Eating Disorder Examination Questionnaire, FINDRISC Finnish Diabetes Risk Score

^*Kruskal Wallis Test

^a−j = Different letters in the same row correspond to significant differences (p < 0.001)

Nutrient intakes by CVD and diabetes risk levels

Compared with those at high risk, those at very high SCORE risk had greater fiber intake (p = 0.004). Individuals with no 10-year ASCVD risk had lower carbohydrate, fiber (p < 0.001), potassium and iron (p = 0.003) contents, while their lipid (p = 0.009) intake was greater than that of those with moderate risk. Energy intake was greater in participants with high diabetes risk than in those with moderate risk (p = 0.007) (Table 4). Although some nutrient intake differences were statistically significant, the magnitude of these differences was small and may have limited practical relevance.

Table 4.

Percentage of daily energy and nutrient requirements meet according to cardiovascular and diabetes risk categories

	Energy and Nutrients
	%E	%Carb	%Pro	%Lipid	%Fiber	%B12 vit	%Folate	%C vit	%Na	%K	%Ca	%Mg	%Fe
SCORE
Low (n = 74)	80.7	42.2	17.3	41.1	72.5	103.3	82.8	79.8	141.1	62.9	60.2	84.9	85.4
Moderate (n= 2150)	84.9	43.1	17.3	40.2	75.1	120.2	85.7	66.3	164.7a	65.1	72.6	78.9	89.2b
High (n = 694)	86.3	43.9	17.0	39.7	77.7c	122.5	89.5	62.3	174.2a	68.3	73.5	81.8	93.6b
Very high (n = 263)	86.7	45.7	16.7	38.1	84.0c	124.7	88.5	53.3	166.8	68.8	74.0	81.5	94.2
Total (n = 3181)	85.3	43.5	17.2	39.9	76.3	120.7	86.7	64.9	166.4	66.1	72.6	79.9	90.5
p-value*	0.262	0.002	0.091	0.006	< 0.001	0.283	0.058	0.082	0.001	0.004	0.014	0.653	0.002
ASCVD
Low (n = 2151)	84.9	43.2	17.2	40.2	74.5	116.9	85.1	92.0	162.0	64.6	71.8	79.1	87.9
Borderline (n = 367)	86.0	43.8	17.1	39.7	79.1	127.3	92.4	86.3	178.9	68.0	75.6	81.7	96.2
Moderate (n = 493)	87.3	44.8d	16.9	38.7e	83.1f	129.2	89.2	88.7	180.2	71.1g	74.4	82.2	97.9h
High (n = 43)	85.2	44.8	16.7	38.9	83.9	103.9	91.4	99.4	165.8	69.5	75.2	78.1	89.4
No 10-y risk (n = 127)	82.2	41.2d	18.3	40.8e	70.3f	137.9	85.5	79.4	149.8	63.6g	70.5	80.1	89.0h
Total (n = 3181)	85.3	43.5	17.2	39.9	76.3	120.7	86.7	90.5	166.4	66.1	72.6	79.9	90.5
p-value*	0.423	0.001	0.174	0.015	< 0.001	0.003	0.005	0.225	< 0.001	< 0.001	0.059	0.535	< 0.001
FINDRISC
Very low (n = 951)	82.3	42.7	17.6	40.2	72.6	113.7	85.6	88.9	158.8	63.8	69.1	77.8	88.0
Low (n = 1289)	82.7	43.2	17.3	40.1	75.1	122.3	86.1	88.1	168.2	65.1	72.3	78.3	90.3
Moderate (n = 486)	88.5i	43.9	16.9	39.6	80.3	125.5	87.1	91.0	165.8	69.9	76.5	82.0	91.4
High (n = 387)	94.0i	44.9	16.2	39.4	82.4	120.4	89.4	97.5	176.7	68.2	75.3	84.1	93.4
Very high (n = 68)	104.9	46.0	16.5	37.9	88.1	154.5	92.4	112.0	183.9	74.4	86.7	100.9	103.6
Total (n = 3181)	85.3	43.5	17.2	39.9	76.3	120.7	86.7	90.5	166.4	66.1	72.6	79.9	90.5
p-value*	< 0.001	0.001	< 0.001	0.157	< 0.001	0.134	0.717	0.006	< 0.001	< 0.001	< 0.001	< 0.001	0.005

Data are expressed as percentage (%)

ASCVD Atherosclerotic Cardiovascular Disease, FINDRISC Finnish Diabetes Risk Score, SCORE Systematic Coronary Risk Evaluation, %E Energy (%), %Carb Carbohydrate (%), %Pro Protein (%), %Na Sodium (%), %Ca Calcium (%), %K Potassium (%), %Mg Magnesium (%), %Fe Iron(%)

^*Kruskal Wallis Test

^a−i = Different letters in the same column correspond to significant differences (p < 0.001)

Correlations among anthropometric measures, nutrient intakes, eating disorder scores, and disease risks

Figure 2 presents a visual summary of pairwise correlations among eating disorder scores, anthropometric measurements, nutrient intakes, and cardiometabolic risk scores, highlighting both expected and unexpected associations across variables. A moderate positive correlation was found between SCORE and ASCVD risk and WHR (r = 0.302; p < 0.001, r = 0.371; p < 0.001, respectively), NC (r = 0.309; p < 0.001, r = 0.370; p < 0.001, respectively), whereas weak correlations were observed for protein intake (r = 0.093; p < 0.001, r = 0.142; p < 0.001 respectively), and cholesterol intake (r = 0.050; p = 0.005, r = 0.064; p < 0.001 respectively). However, the clinical relevance of this very weak correlation may be limited. The FINDRISC score was found to be positively and significantly correlated with BMI (r = 0.604; p < 0.001, strong), NC (r = 0.370; p < 0.001, moderate) and energy intake (r = 0.119; p < 0.001, weak) (p < 0.001).

Fig. 2.

Pairwise correlations among CVDs-diabetes risks, eating behavior, anthropometric measurements, and nutrients

A positive and significant correlation was observed between the total EDE-Q score and BMI (r = 0.277; p < 0.001, weak to moderate), WC (r = 0.101; p < 0.001, weak), and body weight (r = 0.142; p < 0.001, weak) limited clinical relevance when the relationship between EDs and anthropometric characteristics were examined. A significant negative correlation was found between the eating restraint score and WHR (r = −0.103; p < 0.001, weak). BMI was positively correlated with the shape and weight over-evaluation score (r = 0.195; p < 0.001, weak), the body dissatisfaction score (r = 0.319; p < 0.001, moderate) and the bingeing score (r = 0.253; p < 0.001, weak to moderate). The total EDE-Q score was negatively correlated with SCORE (r = −0.101; p < 0.001, weak) and ASCVD (r = −0.092; p < 0.001, weak) risk, whereas it was positively related with FINDRISC (r = 0.163; p < 0.001, weak) (Fig. 2). This unexpected negative correlation may reflect characteristics of the study population, such as health-conscious individuals with subclinical eating behaviors or lower cardiometabolic burden. Additionally, selection bias related to voluntary participation cannot be ruled out.

Considering the correlations between EDs status and CVDs risk, weak negative correlations were observed between SCORE risk and the eating restraint (r = −0.127; p < 0.001), shape and weight over-evaluation (r=−0.106; p < 0.001), body dissatisfaction (r = −0.064; p < 0.001), and purging (r = −0.056, p = 0.001) scores. Similarly, ASCVD risk was weakly negatively correlated with the eating restraint (r = −0.123; p < 0.001), shape and weight over-evaluation (r = −0.099; p < 0.001), body dissatisfaction (r = −0.065; p < 0.001), and purging (r = −0.045, p = 0.013) scores. For diabetes risk, a significant weak to moderate positive correlation was found between the FINDRISC score and the eating restraint (r = 0.073; p < 0.001), shape and weight over-evaluation (r = 0.107; p < 0.001), body dissatisfaction (r = 0.218; p < 0.001), and bingeing (r = 0.146; p < 0.001) scores (Fig. 2). A weak positive correlation was found between bingeing and carbohydrate intake (r = 0.122, p < 0.001). Although some correlations were statistically significant, their small effect sizes suggest limited clinical relevance and should be interpreted with caution.

Predictors of ED scores

Multiple linear regression analysis was conducted to predict ED scores based on anthropometric measures and CVD risk, diabetes risk with all models adjusted for age and sex. BMI, WC, body weight, and SCORE score were significant predictors of EDs (F = 62.171, p < 0.001). Each unit increase in participants’ BMI was associated with a 0.28-point increase in the EDE-Q score (B = 0.28, 95% CI: 0.24 to 0.32, p < 0.001). In contrast, each unit increase in WC (B = −0.02, 95% CI: −0.03 to −0.01, p < 0.001), body weight (B = −0.02, 95% CI: −0.03 to − 0.01, p = 0.001), and SCORE risk (B = −0.08, 95% CI: −0.12 to − 0.05, p < 0.01) were significant negative predictors of EDE-Q scores. A table summarizing the regression results is provided as a supplement (Supplementary Table 1).

Discussion

CVDs and T2DM are among the most prevalent chronic public health problems in both developed and developing countries worldwide. However, the possible effects of EDs and nutritional behaviors on the risk of CVD and diabetes have not been sufficiently investigated. In present study, the total EDE-Q scores of individuals with borderline 10-year ASCVD risk were significantly lower than those of individuals with low risk. SCORE and ASCVD risk were significantly weakly positively correlated with dietary cholesterol and protein intake. The total EDE-Q score was weakly negatively correlated with SCORE and ASCVD risk scores and positively correlated with the FINDRISC score.

Anthropometric measurements are frequently employed to assess obesity, which is a risk factor for CVDs [55]. In a cohort study designed to evaluate ten-year ASCVD risk, BMI emerged as the most significant risk parameter among CVD risk factors [56]. Another study of 22.476 participants aged 30–64 years with no history of CVD reported that overweight or obese individuals faced a greater risk of developing CVD than did those with a normal body weight [57]. Furthermore, research on NC, another effective parameter for determining cardiometabolic risk, revealed that individuals with low SCORE risk levels had significantly lower NCs [58]. Several studies have shown associations between various anthropometric measures and cardiometabolic risk, including BMI, WC, WHR, and NC [59–61]. In another study (n = 1446), higher values of body weight, BMI, WC, and NC were associated with increased CVDs risk [61]. Our findings, consistent with prior literature, confirmed the association between anthropometric indicators and cardiometabolic risk. These results underscore the importance of regularly evaluating individuals’ anthropometric measurements.

As with cardiovascular risk, anthropometric measures serve as essential tools for evaluating diabetes risk [62]. In a study evaluating the relationship between anthropometric measurements and diabetes risk, a moderate positive correlation was found between WC, the WHR and BMI and the FINDRISC score [63]. Similarly, Jurca-Simina et al. [64] reported that the T2DM risk score increased with increasing WHR and body fat percentage (p < 0.001). In another study, the effects of body weight and WC on diabetes risk in non-diabetic and prediabetic individuals were evaluated, and a relationship was found between the FINDRISC score and BMI of prediabetic individuals [65]. Consistent with the literature, our study revealed that, as the degree of CVD and diabetes risk increased, body weight, BMI, NC, and the WHR increased. Among the anthropometric measurements, NC had the strongest relationship with CVD risk, whereas BMI had a similar relationship with diabetes risk.

Irregular and unbalanced eating behaviors have long been recognized as contributing to negative health outcomes. Many of deaths among individuals with EDs have been linked to cardiovascular complications [29, 66]. Mechanisms include electrolyte disturbances due to vomiting, dietary restriction, medication use (e.g., use of laxatives, weight loss pills), and refeeding syndrome [67]. Structural and functional changes resulting from EDs may be responsible for the risk of CVD. These changes include heart rate and rhythm, hemodynamic, and peripheral vascular abnormalities, electrolyte imbalances, increased carotid artery stiffness, reduced aortic stiffness, vagal hyperactivity and endothelial dysfunction. In addition, ED-related weight gain and obesity as comorbid conditions increase the risk of CVD [9]. Large-scale investigations of the associations of EDs with thirty-one different chronic somatic conditions in adults (n = 36,309) have been conducted. The study revealed that individuals with BED had a significantly greater incidence of hypertension and minor CVDs [68]. Consistent with prior research, a large-scale study across 14 countries (n = 24,124) reported elevated hypertension prevalence among individuals with BED [69].

Similar to previous findings indicating increased lipid concentrations in individuals with AN [70], our study highlights the importance of assessing cardiovascular risks in populations with EDs. However in the study by Udo et al., AN and BN were not associated with elevated serum cholesterol and triglyceride levels, contrary the meta-analysis findings. Only BED was found to be correlated with higher serum triglyceride and cholesterol levels [68]. In a cohort study, it was observed that the risk of CVD increased at different periods after AN diagnosis, especially the risk of ischemic heart disease increased after 60 months [71]. Another study found a significant genetic correlation between BED with AN and myocardial infarction, but no significant genetic correlation was found between AN without BED and any CVD event [8]. Unlike the literature, our study revealed that individuals with high total and ED subdimension scores (eating restraint, shape, weight over-evaluation, body dissatisfaction, purging) had lower SCORE and 10-year ASCVD risk values.

It must be remembered that the EDE-Q measures a range of disordered eating attitudes and behaviors, among them which might not be sufficient to diagnose an ED. Consequently, subclinical behaviors rather than full syndrome EDs may be the cause of higher EDE-Q scores in our study cohort. Interestingly, an inverse association was observed between cardiovascular risk and disordered eating behaviors. Furthermore, certain subclinical disordered eating behaviors, particularly dietary restraint or restrictive intake, may contribute to lower BMI and improved lipid profiles, which in turn can reduce calculated risk scores. These behaviors, while pathological in nature, might inadvertently align with cardiometabolic risk reduction. Thus, the inverse correlation observed may reflect an overlap between maladaptive eating attitudes and weight-control efforts that lower traditional cardiometabolic risk markers. This unexpected association may partly be explained by selection bias, as individuals with higher health consciousness or milder, subclinical disordered eating symptoms may have been more likely to participate. Moreover, restrictive eating behaviors, which often reduce BMI and other cardiometabolic risk factors, could also contribute to the inverse relationship observed. The potential influence of psychological stress and medication use on both eating behaviors and cardiometabolic health should also be considered. These differing results may be attributed to the focus of existing studies on specific EDs, whereas our study evaluated eating problems collectively. Differences in CVD assessment methods and sample characteristics between studies may also contribute to the variability in results. Evaluations based on specific conditions, such as BED or AN, may yield different clinical outcomes. Some disordered eating behaviors, particularly restrictive eating patterns, can result in lower body weight and BMI, which are traditionally associated with reduced cardiovascular risk, the inverse relationship observed between EDE-Q scores and CVD risk [72]. While several correlations reached statistical significance, the strength of some associations was weak, and their clinical relevance should be interpreted with caution. The clinical reflections of ED subdimensions and types differ from each other. While an inverse relationship is seen in some types of ED, a positive relationship is seen in others. The ED type and clinical presentation of the sample in this study may have affected the risk of CVD. Therefore, a detailed examination of the subdimensions of ED is important to shed light on this issue. Our findings diverge from much of the prior literature by showing an inverse relationship between disordered eating behaviors and CVD risk. Many previous studies have focused on clinical ED diagnoses such as BED and AN (For instance, BED has been associated with obesity, hypertension, and increased cardiometabolic risk [73], whereas AN often coexists with lower BMI and, in some cases, favorable lipid profiles but elevated risk for cardiac complications [74]. In our study, higher scores on eating restraint and shape/weight over-evaluation may reflect subclinical or health-conscious behaviors that contribute to lower body weight and, consequently, lower CVD risk. This highlights the importance of assessing ED subtypes and symptom dimensions separately when exploring their cardiometabolic effects.

EDs directly affect individuals’ eating behavior, body weight and indirectly affect individuals’ risk of T2DM [75]. In a study involving adults, 6.9% of individuals with low diabetes risk and 13.5% of individuals with high diabetes risk had a risk of eating behavior disorders. A statistically significant relationship was identified between diabetes risk and eating behavior disorders [76]. Another study aimed at identifying individuals at high risk for diabetes through HbA1c measurement and determining the prevalence of EDs revealed a weak positive correlation between ED scores and HbA1c values [77]. Consistent with the literature, our study revealed that ED scores increased with increasing diabetes risk. A significant positive correlation was found between diabetes risk and scores for eating restraint, shape and weight over-evaluation, body dissatisfaction, and bingeing. Impaired carbohydrate metabolism may contribute to the onset and persistence of BN and other EDs. These findings suggest that individuals at risk for diabetes may engage in unhealthy eating practices. In managing a condition such as diabetes, where maintaining blood glucose balance is crucial, it is essential for individuals to adopt healthy eating behaviors and a balanced diet to mitigate disease risk. Therefore, measures to detect and address eating behavior disorders are important for reducing diabetes risk.

The risk of diabetes varies depending on the type of EDs. In a meta-analysis by Nieto-Martínez and colleagues, BED (OR: 3.69) and BN (OR: 3.45) increased the risk of T2DM, whereas AN was associated with a decreased risk (OR: 0.71) [78]. In our study, the score for the binge eating subdimension increased with higher diabetes risk, which consistent with the literature. In contrast to the effects observed in BED and BN, individuals with AN typically engage in excessive energy restriction. Speakman et al. reported that energy restriction leads to reductions in adipokines and improvements in insulin resistance related to hyperglycemia [79], which is also a mechanism suggesting that AN may reduce the incidence of T2DM. This mechanism suggests that AN may reduce the incidence of T2DM. A similar mechanism is thought to apply to CVD risk as well. However, given that preventing one disease by inducing another is neither feasible nor safe, these results should be examined more thoroughly.

High sodium intake is a significant dietary factor that increases the risk of cardiometabolic diseases [80]. A meta-analysis found that for every 1-gram increase in dietary sodium intake, the risk of CVD increased by 6.0% [81]. The mechanisms through which high sodium intake contributes to the development of CVD are diverse, including volume expansion, changes in renal function and sodium balance, impaired renin‒angiotensin‒aldosterone system response, central stimulation of sympathetic nervous system activity, and potential inflammatory processes [82]. According to Türkiye Nutrition Guide-2022, 64.1% of adults aged 18–65 consume more than 4000 mg of sodium daily [53]. In our study, the sodium intake of individuals with high SCORE risk was found to be greater than that of those with medium risk. Our findings, which are consistent with the literature, demonstrate a significant linear relationship between dietary sodium intake and CVD risk. Bakırhan and İncedal Irgat (n = 1339) reported that a one-unit increase in saturated fat intake was associated with a 1.02-fold increase in 10-year CVD risk, whereas an increase in protein intake was associated with a 0.96-fold decrease (p < 0.05) [83]. Another study reported that dietary modification of mono- and polyunsaturated fats was inversely related to CVD risk [84]. In our study, dietary cholesterol and protein intake were significantly positively associated with cardiovascular risk (Fig. 2). Although the effects of nutrients on CVD risk vary, it can be concluded that both the amount and type of lipid intake are crucial dietary factors in determining CVD risk [85–88].

In obese individuals, higher levels of non-esterified fatty acids, glycerol, hormones and pro-inflammatory cytokines are released from adipose tissue, which may contribute to the development of insulin resistance and subsequently diabetes [88]. A study involving an average of 11.9% energy restriction observed a decrease in CVD risk markers, such as insulin resistance and T2DM [89].

Similarly, in our study, the energy intake of individuals at high risk of diabetes was greater than that of individuals at moderate risk, and diabetes risk was found to be positively and significantly correlated with energy intake. Energy restriction may have a positive effect on reducing proinflammatory markers and the risk of developing diabetes and cardiometabolic diseases. High dietary energy intake may trigger chronic inflammation via the hypoxia response in adipose tissue, and fatty acids and glucose metabolites may contribute to inflammation by activating serine kinases in cells. In contrast to high energy, energy restriction reduces circulating levels of inflammatory cytokines and reduces proinflammatory responses in various tissues [77, 83, 89, 90]. Although individual nutrient intake levels (e.g., sodium, cholesterol, and protein) were assessed in our study, these findings may also reflect broader dietary patterns among the study population. For example, the high sodium and low fiber intake observed may be inconsistent with dietary patterns such as the Mediterranean or DASH diets, which emphasize fruit, vegetable, whole grain, and healthy fat consumption. Future studies should further explore how adherence to established dietary patterns may influence cardiometabolic risk in relation to disordered eating behaviors.

Our study has several strengths, including a large sample size and being the first to comprehensively investigate the relationship between EDs and the risk of CVD and diabetes. However, this study also has several limitations. A major limitation is the cross-sectional design, which precludes the ability to determine causality. Longitudinal studies are essential to clarify the temporal and potentially causal relationships between disordered eating, nutrient intake, and cardiometabolic risk. The generalizability of our findings may be limited due to the voluntary sampling method and focus on a specific population in Türkiye. This approach may introduce selection bias, as individuals with higher health awareness may have been more likely to participate. One limitation of this study is the exclusion of individuals with previously diagnosed EDs, which may limit the generalizability of the findings. As a result, the study focuses primarily on subclinical or undiagnosed disordered eating behaviors. Additionally, while we adjusted for key confounders, residual confounding is still possible. Factors such as socioeconomic status, psychological stress, physical activity, sleep, and medication use were not fully captured. Furthermore, clinical diagnoses of CVD and T2DM were not made; instead, validated risk scores were used, which may not entirely reflect true disease status. While these scores are widely applied, they may not fully capture actual disease incidence. Another potential limitation is the use of a 24-hour food recall method for evaluating eating patterns. This method relies on participants’ memory and self-reports. Self-reported food consumption records may not always reflect daily nutrition, which may result in reporting over- or under-intake. Observational dietary intake could be useful for more accurate results. Differences in individual responses, which may be influenced by genetic, epigenetic, and environmental factors, could account for inconsistencies with the results of other studies. While not formally validated in Turkish, this method has been previously used among Turkish adults, supporting its practical applicability [91–94]. Despite these limitations, this study raises an important point: linking eating behavior disorders and food intake with the risk of CVD and diabetes may be an important strategy for mitigating disease risk. In order to reduce the risk of diseases such as CVDs and diabetes, practices and preventive policies that detect inadequate and excessive nutritional intake and eating behavior disorders should be developed in the clinic and in the field, and activities to increase public awareness should be carried out. Eating behavior disorders and inadequate/excessive food intake detected in the clinic should be treated first. Encouraging health workers in this regard would be an important first step.

Conclusion

This study revealed complex associations between anthropometric indicators, ED symptomatology, and cardiometabolic risk levels among middle-aged adults in Türkiye. Disordered eating symptoms particularly subclinical presentations, appeared to be less prevalent among individuals with higher cardiometabolic risk, while subclinical eating disturbances were more common in those with lower or moderate risk levels, suggesting a complex and possibly non-linear relationship between eating behaviors and cardiometabolic health. Further studies should investigate the relationship between subclinical disordered eating behaviors and cardiovascular risk, utilizing longitudinal designs and clinical assessments.

Clinically, establishing specialized units that assess EDs alongside CVD and diabetes is crucial. We recommend integrating routine ED screening into primary care cardiometabolic assessments using brief validated tools such as the SCOFF or EAT-26 questionnaires. Multidisciplinary teams including nutritionists, psychologists, and cardiologists should be structured to provide holistic care. In addition, public health initiatives should focus on community-based education and screening programs to raise awareness about the comorbidity of EDs and cardiometabolic diseases.

Abbreviations

AN

Anorexia Nervosa

ASCVD

Atherosclerotic Cardiovascular Disease

BED

Binge Eating Disorder

BMI

Body mass index

BN

Bulimia Nervosa

CVD

Cardiovascular disease

CVDs

Cardiovascular diseases

EDs

Eating disorders

EDE-Q

The Eating Disorder Examination Questionnaire

FINDRISC

The Finnish Diabetes Risk Score

NC

Neck circumference

SCORE

Systematic Coronary Risk Evaluation

T2DM

Type 2 diabetes mellitus

WC

Waist circumference

WHR

Waist-to-hip ratio

Authors’ contributions

ÖA, HB contributed to the conceptualization of the study. BNK and SÖ curated the data. FÖA and EK performed the formal analysis and prepared all tables and figures. FÖA, EK, FESK, and HB wrote the original draft of the manuscript. FESK and HB supervised the review & editing at the manuscript. All authors contributed to the article and approved the submitted version.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The research was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. All individuals who agreed to participate in the study were read the informed consent form, and written informed consent was obtained from each participant prior to their inclusion in the study. The ethical permission for the study was approved by the decision of the Istanbul Medipol University Non-Interventional Research Ethics Committee (decision number: 55, document number: E-10840098-202.3.02-627).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.