Introduction

School cafeteria kitchens are indoor occupational settings where various harmful substances can be generated, particularly during cooking processes that release carbon monoxide (CO) and cooking oil fumes (COFs) [1]. COFs behave similarly to fine particulate matter (PM) and contain polycyclic aromatic hydrocarbons and volatile organic compounds, which can adversely affect human health with chronic exposure [2]. Evidence that exposure to COFs is associated with adverse health effects, such as lung cancer, is already widely accepted, primarily based on epidemiological studies from Chinese-speaking regions where cooking oil is heavily used [3-5].

A typical Korean meal includes steamed rice, soup, and two to three side dishes that are stir-fried, seasoned, or boiled. In other words, the tasks for each cook involve preparing a substantial volume of food, ranging from 10 to 100 servings, including dishes that are stir-fried and steamed. Daily menus are developed by a nutritionist to ensure a balanced diet, typically comprising one to two meat and one to two vegetable dishes. This culinary environment, including photographic documentation, has been previously described in the literature [6,7]. Previous studies have indicated an increased risk of lung cancer among cafeteria workers [8] and a higher prevalence of radiological abnormalities such as pulmonary nodules [6,9]. In addition to its respiratory effects, COF exposure has been associated with hepatic and reproductive system dysfunction [10]. CO is a well-established cardiovascular risk factor [11]. However, studies evaluating the cardiovascular impact of working as a cook, particularly studies investigating early structural changes using radiological markers, remain limited.

Radiological indicators that may serve as early markers of cardiovascular disease include vascular calcifications, the cardiothoracic ratio, and diameter of the ascending aorta. The cardiothoracic ratio has been reported to be negatively correlated with left ventricular function [12], whereas the aortic diameter is significantly associated with traditional cardiovascular risk factors, such as hypertension and dyslipidemia [13]. Because these parameters can be readily assessed using chest computed tomography (CT), they may be useful for evaluating potential or subclinical cardiovascular changes in specific occupational groups such as cafeteria workers.

Therefore, in this study, we aimed to assess subtle structural changes in the cardiovascular system of school cafeteria workers with long-term exposure to COFs by analyzing cardiovascular CT indicators and comparing them with those of an age- and sex-matched control group.

Methods

Ethics statement: This retrospective study was approved by the Institutional Review Board (IRB) of Keimyung University Hospital (IRB No: DSMC 2024-07-002). The requirement for informed consent was waived due to the retrospective nature of the study.

1. Study design and population

This study was based on the same cohort of school cafeteria workers and their age- and sex-matched controls used in our earlier investigation of chest CT findings [6].

School cafeteria workers who underwent low-dose chest CT screening at a tertiary referral hospital between January and December 2022 were included. The screening program for school cafeteria workers initially targeted workers aged ≥55 years or those with at least 10 years of employment. However, to examine the long-term effects of occupational exposure to COFs, we included only those aged 45 to 54 years with a confirmed employment duration of >10 years. Individuals aged 55 years or older were excluded due to uncertainty in their employment histories. A total of 88 cafeteria workers met the inclusion criteria and were enrolled. The control group used in a previous study was employed in the current analysis. These controls consisted of individuals without specific clinical indications who underwent low-dose chest CT scans at the hospital’s health promotion center as part of a health screening. Because the CT scans performed for disease evaluation are typically reimbursed under Korea’s national health insurance system for current and former smokers [14], these individuals are presumed to be free of known cardiopulmonary diseases. Control participants were selected from the hospital information system database of Keimyung University Dongsan Medical Center, which was built on an Oracle Database (Oracle Corporation, Austin, TX, USA) using structured query language (SQL) queries. All female individuals who underwent low-dose chest CT screening as part of a general health checkup between March 1, 2022, and July 8, 2024, were identified. Among these, 997 cases matched the sex and age (at the time of CT acquisition) of the school cafeteria workers who participated in the screening. Using SQL-based random sampling, each cafeteria worker was matched 1:1 with one of these control individuals. Finally, because the cafeteria worker group was restricted to those aged 45 to 54 years at the time of the CT scan, the corresponding age-matched control participants already paired with these workers were subsequently extracted and included in the final control group.

2. Computed tomography image analysis

All CT images were initially reviewed by a board-certified thoracic radiologist (JHH, with 12 years of experience) who was blinded to the group assignment. This radiologist performed all primary assessments, including the evaluation of calcifications, measurement of the ascending aortic diameter, and calculation of the cardiothoracic ratio. Subsequently, a second thoracic radiologist (JYK, with 14 years of experience), blinded to the case-control status, independently reevaluated each case to confirm the findings. For any case deemed to require correction or remeasurement, both radiologists jointly reassessed the images in a consensus review session, and the final values were determined through mutual agreement.

Coronary artery calcification (CAC) was visually assessed using a previously validated ordinal scale as follows: none (no visible CAC), mild (isolated speckles in any coronary artery), moderate (greater than mild but less than severe), or severe (contiguous, extensive, or diffuse calcification in at least one coronary segment). Given the small number of patients with CAC and the absence of severe findings, we analyzed the data using both an ordinal scale and binary classification (presence vs. absence of CAC) [15]. Calcification of the ascending aorta and cardiac valves was also assessed. Valve calcification was considered positive if noted in any of the cardiac valves, including the aortic, mitral, tricuspid, or pulmonary valve.

The ascending aortic diameter was measured on axial CT images at the level of the pulmonary artery bifurcation, perpendicular to the axis of the vessel [16]. The cardiothoracic ratio was calculated using a single axial slice where the heart appeared largest, defined as the maximum transverse cardiac diameter divided by the maximum internal thoracic diameter (Fig. 1) [12].

3. Chest computed tomography acquisition

Chest CT scans were acquired using 64- or 128-slice multidetector CT scanners (SOMATOM Force, SOMATOM Definition Edge, and SOMATOM Drive; Siemens Healthineers, Forchheim, Germany). All scans were performed during a single full-inspiratory breath-hold in the supine position, using a craniocaudal scanning direction. The imaging parameters included a tube voltage of 120 kVp, reference tube current of 20 mAs, and automatic exposure control. Reconstructions were performed at both 1.0 mm and 3.0 mm slice thickness using a b49f kernel. No intravenous contrast agent was administered.

4. Statistical analysis

Categorical variables were compared using chi-square tests, and continuous variables were compared using Student t-tests. A p-value<0.05 was considered statistically significant. The cardiothoracic ratio was analyzed as a continuous variable and subsequently dichotomized using thresholds of 0.50 and 0.56 to compare groups based on the presence of cardiomegaly [12]. For CAC, the participants were categorized based on an ordinal scale (grade 0, 1, or 2) and whether they had undergone coronary stent placement. For the binary analysis, calcification was considered present if any grade of calcification or a stent was observed. Calcifications of the ascending aorta and cardiac valves were assessed as binary variables (presence or absence of calcification). Furthermore, we classified participants as having “any cardiac calcification” if calcification was present in at least one of the following three regions: the coronary artery, ascending aorta, or cardiac valves. All statistical analyses were performed using R statistical software ver. 4.4.2 (https://r-project.org; R Foundation for Statistical Computing, Vienna, Austria), and a p-value<0.05 was considered statistically significant.

Results

The general and radiological characteristics of the study participants are presented in Table 1. All 176 participants (88 school cafeteria workers and 88 matched controls) were identical to those in the dataset used in a previous study investigating nodules and bronchial wall thickness on chest CT [6]. All participants were female, with a mean age of 51.41±2.42 years.

The mean diameter of the ascending aorta was significantly greater in the cafeteria worker group than in the control group (31.69±3.28 mm vs. 30.64±3.21 mm, p=0.032). Similarly, the transverse cardiac diameter was significantly larger in cafeteria workers (113.49±10.46 mm vs. 108.53±10.17 mm, p=0.002), whereas the chest transverse diameter did not differ between the groups (244.39±11.26 mm vs. 241.36±10.14 mm, p=0.063). The cardiothoracic ratio was also significantly higher in the cafeteria worker group (0.47±0.04 vs. 0.45±0.04, p=0.026). However, when using cardiothoracic ratio thresholds of 0.5 and 0.56 for binary comparisons, no statistically significant differences were observed between the groups. When applying a threshold of 0.50 for the cardiothoracic ratio, cardiomegaly was observed in 11 control participants (12.5%) and 15 cafeteria workers (17.0%) (p=0.524). Using a threshold of 0.56, cardiomegaly was identified in one control participant (1.1%) and two cafeteria workers (2.3%) (p>0.999).

One cafeteria worker had a coronary stent, which precluded an accurate CAC assessment. This participant was classified as having CAC for analysis. Excluding this individual, six cafeteria workers had CAC, five with mild (5.7%) and one with moderate (1.1%) calcification. In the control group, four individuals had CAC, one with mild (1.1%) and three with moderate (3.4%) calcification. The difference in CAC prevalence between groups was not statistically significant (p=0.193). We further categorized the participants into two groups based on the presence or absence of CAC and compared the prevalence between cafeteria workers and controls. Individuals with coronary stents were classified as having CAC. As a result, CAC was observed in seven cafeteria workers (8.0%) and four control participants (4.5%). The difference between groups was not statistically significant (p=0.533).

Ascending aortic calcification was observed in five cafeteria workers (5.7%) and two controls (2.3%), whereas valve calcification was detected in one cafeteria worker (1.11%) and two controls (2.3%). Differences in the prevalence of ascending aortic and valvular calcifications were not statistically significant (p=0.440 and p>0.999, respectively).

We combined coronary artery, valve, and ascending aortic calcifications into a single variable termed “any cardiac calcification” and compared its prevalence between the two groups. Calcification was observed in nine individuals (10.2%) in the cafeteria worker group and seven individuals (8.0%) in the control group. However, the difference between the groups was not statistically significant (p=0.793).

Discussion

This study investigated the subclinical cardiovascular changes in Korean school cafeteria workers using chest CT scans obtained during lung cancer screening. Although there was no statistically significant increase in the prevalence of calcifications or cardiomegaly, the cafeteria worker group demonstrated significantly higher values for both the cardiothoracic ratio and ascending aortic diameter when assessed as continuous variables.

The ascending aortic diameter is a vascular indicator that correlates with various cardiovascular risks, including cholesterol levels and CAC [13]. Similarly, the ascending aortic area has been significantly associated with carotid atherosclerosis [17]. In addition, the ascending aortic diameter has been linked to traditional cardiovascular risk factors, such as diastolic blood pressure, high-density lipoprotein cholesterol, and atherosclerotic disease [18]. The cardiothoracic ratio mainly reflects myocardial dimensions. Although this measure may be affected by the respiratory phase during image acquisition, it has been used as a prognostic factor for cardiovascular diseases in numerous epidemiological studies and is known to correlate with cardiac function [19]. As demonstrated in the present study, the cardiothoracic ratio measured on CT showed a significant negative association with ventricular ejection fraction [12].

It has been previously suggested that exposure to various pollutants generated in kitchen environments can adversely affect multiple organ systems, including the cardiovascular system [20]. Previous studies have indicated respiratory effects [6,9] and carcinogenic risks [8] in Korean cooking staff. However, limited research has been conducted on cardiovascular system-related markers, symptoms, and disease outcomes among individuals working in school kitchens, particularly in Korea. Cafeteria workers may experience short-term high-level exposure to PM and CO [1]. High ambient concentrations of PM ≤2.5 μm (PM_2.5) are known to trigger atherogenic conditions such as increased endothelial apoptosis, an antiangiogenic plasma profile, and elevated levels of circulating monocytes and T cells [21]. A study conducted in China comparing workers with COF exposure before and after shifts in kitchens found a correlation between increased PM and polycyclic aromatic hydrocarbon levels and reduced heart rate variability, a cardiac autonomic marker associated with arrhythmia and sudden cardiac death [22]. Another study indicated that short-term exposure to aerosols from indoor gas stove cooking caused an increase in systolic blood pressure within 60 minutes of exposure [23], which is consistent with the findings of other studies on ultrafine particles from electric stove use [24]. Data collected from school kitchens in Korea demonstrated that CO concentrations in some areas could exceed both the U.S. National Institute for Occupational Safety and Health (NIOSH) time-weighted average (TWA) threshold of 35 parts per million (ppm) and the ceiling limit of 200 ppm [1]. CO binds to hemoglobin to form carboxyhemoglobin, thereby increasing myocardial oxygen demand and the risk of ischemia, exerting negative inotropic effects, and promoting thrombosis by increasing platelet stickiness. The NIOSH-recommended TWA of 35 ppm was intended to prevent chronic heart disease [25]. Thus, cooking staff may experience sufficient CO exposure levels to exert clinical effects. Animal studies have shown that exposure to low levels of CO (50 or 100 ppm for 4 hours over 5 consecutive days) in female Wistar rats results in hypertrophy of the left and right ventricles and interventricular septum. In these experiments, the mean duration of exercise before the onset of pain was reduced and pain duration was significantly prolonged after 100 ppm CO exposure compared to air exposure [26]. CO exposure has also been shown to induce direct vasodilation in situ [27]. Long-term exposure to such hemodynamically active substances may lead to subclinical changes in the cardiovascular system, and ultimately, an increased risk of cardiovascular disease. Given their prolonged exposure to multiple cardiovascular risk factors in cooking environments, cafeteria workers may develop subclinical structural changes in the heart due to cumulative and chronic physiological stressors.

This study had several limitations, the most significant of which is the retrospective design. This design prevented the collection of personal, occupational, and medical histories, thereby limiting the analysis to imaging data only. Consequently, it was not possible to perform multivariable regression adjustments or calculate propensity scores to account for major confounders such as smoking history. Although the cafeteria worker and control groups were matched 1:1 by sex and age, the absence of smoking data precluded a more detailed analysis, such as stratification by exposure duration or accounting for smoking status. In a previous retrospective analysis of chest CT scans obtained for lung cancer screening, there was no standardized control for the respiratory phase or gating, which may have introduced measurement errors. However, these errors are expected to be nondifferential. In addition, the study lacked data on key cardiovascular risk factors, such as body habitus, cholesterol levels, smoking history, family history, and occupational history of the control group, which are potential confounders. However, cafeteria work is considered a physically demanding job for women [7], and a healthy worker effect or survival bias may be present [28]. Moreover, physical activity at work is known to protect against cardiovascular disease [29,30]. Another study indicated that school cooks had a lower body mass index, smoking rate, and comorbidity burden than clerical workers [8], and the levels were not substantially different from those of the general population. Thus, based on a synthesis of prior studies, such bias is more likely to have led to an underestimation of cardiovascular risk in school cafeteria workers. Additionally, the present study used a single-institution approach. Owing to the absence of similar prior research, this can be considered a pilot study, for which an appropriate effect size and sample size could not be prospectively calculated. Therefore, for outcomes that did not demonstrate a statistically significant difference, such as calcification of the coronary artery or aortic arch, our findings will be valuable in estimating the necessary effect and sample sizes for planning future large-scale studies. Another limitation is that interobserver reliability metrics, such as the intraclass correlation coefficient, were not calculated because one radiologist performed all measurements, whereas a second radiologist reviewed and confirmed them. However, this limitation is unlikely to have significantly affected the results, as the measurements were straightforward, with minimal potential for interobserver disagreement.

Despite these limitations, the identification of significant differences in both the ascending aortic diameter and cardiothoracic ratio supports the presence of subclinical cardiovascular alterations in the cooking staff. This study provided pilot data suggesting that occupational exposure to substances generated during cooking poses cardiovascular risks beyond the respiratory system. Future studies using larger samples and longitudinal designs that follow workers into older age, when cardiovascular disease is more prevalent, are needed to clarify the clinical relevance of these findings and overcome the healthy worker effect. Public health efforts should consider whether targeted cardiovascular prevention strategies are warranted for this occupational group.

Article information

Conflicts of interest

No potential conflict of interest relevant to this article was reported.

Funding

None.

Author contributions

Conceptualization, Data curation, Methodology, Supervision: KB; Investigation, Software, Visualization: JHH; Validation: JYK; Writing-original draft: JHH; Writing-review & editing: JYK, KB

Fig. 1.

Measurement of ascending aortic diameter and cardiothoracic ratio on chest computed tomography. (A) The ascending aortic diameter is measured at the level just below the pulmonary artery bifurcation, using the outer wall-to-outer wall distance on axial images, perpendicular to the aortic axis. (B) The cardiothoracic ratio is calculated by dividing the maximum transverse cardiac diameter by the maximum internal thoracic diameter on the axial slice where the heart appears the largest.

References

• 1. Lee IJ, Lee SG, Choi BH, Seo HK, Choi JH. Hazard levels of cooking fumes in Republic of Korea schools. Saf Health Work 2022;13:227–34.ArticlePubMedPMC
• 2. Chiang TA, Wu PF, Ko YC. Identification of carcinogens in cooking oil fumes. Environ Res 1999;81:18–22.ArticlePubMed
• 3. Jia PL, Zhang C, Yu JJ, Xu C, Tang L, Sun X, et al. The risk of lung cancer among cooking adults: a meta-analysis of 23 observational studies. J Cancer Res Clin Oncol 2018;144:229–40.ArticlePubMedPMCPDF
• 4. Lee T, Gany F. Cooking oil fumes and lung cancer: a review of the literature in the context of the U.S. population. J Immigr Minor Health 2013;15:646–52.ArticlePubMedPDF
• 5. Yu IT, Chiu YL, Au JS, Wong TW, Tang JL. Dose-response relationship between cooking fumes exposures and lung cancer among Chinese nonsmoking women. Cancer Res 2006;66:4961–7.ArticlePubMedPDF
• 6. Hong JH, Kim JY, Baek K. Radiological changes in chest computed tomography findings of school cooks: a comparative study with age-matched controls. Saf Health Work 2025;16:228–35.ArticlePubMedPMC
• 7. Moon YH, Yang YJ, Do SY, Kim JY, Lee CG, Chae HJ, et al. Evaluation of the prevalence of musculoskeletal symptoms, presumptive diagnosis, medical care use, and sick leave among female school meal service workers. Ann Occup Environ Med 2019;31:1.ArticlePubMedPMCPDF
• 8. Jang J, Kim EM, Kim J, Min J, Kim I. Lung cancer risk in female school cooks: a nationwide retrospective cohort study in the Republic of Korea. Saf Health Work 2025;16:105–12.ArticlePubMedPMC
• 9. Kim M, Kim Y, Kim AR, Kwon WJ, Lim S, Kim W, et al. Cooking oil fume exposure and Lung-RADS distribution among school cafeteria workers of South Korea. Ann Occup Environ Med 2024;36:e2. ArticlePubMedPMCPDF
• 10. Rim K. Toxicity of cooking oil fumes and prevention of occupational diseases: a literature review. Toxicol Env Heal Sci 2023;15:101–8.ArticlePDF
• 11. Sørhaug S, Steinshamn S, Nilsen OG, Waldum HL. Chronic inhalation of carbon monoxide: effects on the respiratory and cardiovascular system at doses corresponding to tobacco smoking. Toxicology 2006;228:280–90.ArticlePubMed
• 12. Chae M, Na JU, Lee JH, Shin DH. A study on computed tomography cardiothoracic ratio in predicting left ventricular systolic dysfunction. Clin Exp Emerg Med 2023;10:52–9.ArticlePubMedPMCPDF
• 13. Turkbey EB, Jain A, Johnson C, Redheuil A, Arai AE, Gomes AS, et al. Determinants and normal values of ascending aortic diameter by age, gender, and race/ethnicity in the Multi-Ethnic Study of Atherosclerosis (MESA). J Magn Reson Imaging 2014;39:360–8.ArticlePubMedPMC
• 14. Lee J, Lim J, Kim Y, Kim HY, Goo JM, Lee CT, et al. Development of protocol for Korean Lung Cancer Screening Project (K-LUCAS) to evaluate effectiveness and feasibility to implement national cancer screening program. Cancer Res Treat 2019;51:1285–94.ArticlePubMedPMCPDF
• 15. Fresno CU, Tijmes FS, Thavendiranathan P, Akhtari S, Karur GR, Torres FS, et al. Visual ordinal scoring of coronary artery calcium on contrast-enhanced and noncontrast chest CT: a retrospective study of diagnostic performance and prognostic utility. AJR Am J Roentgenol 2022;219:569–78.ArticlePubMed
• 16. Wolak A, Gransar H, Thomson LE, Friedman JD, Hachamovitch R, Gutstein A, et al. Aortic size assessment by noncontrast cardiac computed tomography: normal limits by age, gender, and body surface area. JACC Cardiovasc Imaging 2008;1:200–9.ArticlePubMed
• 17. Kuipers AL, Carr JJ, Terry JG, Nair S, Barinas-Mitchell E, Wheeler V, et al. Aortic area as an indicator of subclinical cardiovascular disease. Open Cardiovasc Med J 2022;16:e187419242203100. ArticlePDF
• 18. Chen T, Yang X, Fang X, Tang L, Zhang Y, Weng Y, et al. Potential influencing factors of aortic diameter at specific segments in population with cardiovascular risk. BMC Cardiovasc Disord 2022;22:32.ArticlePubMedPMCPDF
• 19. Truszkiewicz K, Poręba R, Gać P. Radiological cardiothoracic ratio in evidence-based medicine. J Clin Med 2021;10:ArticlePubMed
• 20. Li A, Kosonen R. Pollutions of cooking oil fume and health risks. In: Li A, Kosonen R, editors. Kitchen pollutants control and ventilation: a ventilation guide to Asian & European kitchen environment. Springer; 2019. p. 61–150.
• 21. Pope CA, Bhatnagar A, McCracken JP, Abplanalp W, Conklin DJ, O'Toole T, et al. Exposure to fine particulate air pollution is associated with endothelial injury and systemic inflammation. Circ Res 2016;119:1204–14.ArticlePubMedPMC
• 22. Chan CC, Lin LY, Lai CH, Chuang KJ, Wu MT, Pan CH, et al. Association of particulate matter from cooking oil fumes with heart rate variability and oxidative stress. Antioxidants (Basel) 2021;10:1323.ArticlePubMed
• 23. Naseri M, Sultanbekovna AA, Malekipirbazari M, Kenzhegaliyeva E, Buonanno G, Stabile L, et al. Human exposure to aerosol from indoor gas stove cooking and the resulting cardiovascular system responses. Toxicol Rep 2024;13:101716.ArticlePubMedPMC
• 24. Gabdrashova R, Nurzhan S, Naseri M, Bekezhankyzy Z, Gimnkhan A, Malekipirbazari M, et al. The impact on heart rate and blood pressure following exposure to ultrafine particles from cooking using an electric stove. Sci Total Environ 2021;750:141334.ArticlePubMed
• 25. National Institute for Occupational Safety and Health. 1988 OSHA PEL Project documentation [Internet]. Cincinnati, OH: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 1988 [cited 2024 Oct 22]. https://www.cdc.gov/niosh/pel88/630-08.html.
• 26. Anderson EW, Andelman RJ, Strauch JM, Fortuin NJ, Knelson JH. Effect of low-level carbon monoxide exposure on onset and duration of angina pectoris. A study in ten patients with ischemic heart disease. Ann Intern Med 1973;79:46–50.ArticlePubMedPDF
• 27. Lin H, McGrath JJ. Vasodilating effects of carbon monoxide. Drug Chem Toxicol 1988;11:371–85.ArticlePubMed
• 28. Li CY, Sung FC. A review of the healthy worker effect in occupational epidemiology. Occup Med (Lond) 1999;49:225–9.ArticlePubMed
• 29. Li J, Siegrist J. Physical activity and risk of cardiovascular disease--a meta-analysis of prospective cohort studies. Int J Environ Res Public Health 2012;9:391–407.ArticlePubMedPMC
• 30. Fransson EI, Alfredsson LS, de Faire UH, Knutsson A, Westerholm PJ. Leisure time, occupational and household physical activity, and risk factors for cardiovascular disease in working men and women: the WOLF study. Scand J Public Health 2003;31:324–33.ArticlePubMedPDF

Figure & Data

References

Citations

Citations to this article as recorded by