Abstract
Objective
To investigate the prevalence of elevated total cholesterol (TC) among workers in an automobile manufacturing plant, as well as to examine the impact of occupational exposure to hazardous factors on their TC levels.
Methods
This was a retrospective cross-sectional study. A latent class analysis was performed using occupational health examination data collected to classify the nine major occupational hazards into five distinct exposure profiles. Multivariate logistic regression models were employed to identify risk factors associated with elevated TC.
Results
A total of 191 workers (27.0%) were diagnosed with elevated TC in the study. Based on latent class analysis, occupational hazards were classified into five distinct groups: Low Exposure Group, Noise Exposure Group, Dust Exposure Group, Organic toxins Exposure Group, and Combined Noise and Harmful Gases Exposure Group. Multivariate logistic regression revealed that exposure to organic toxins and combined noise and harmful gases group was significantly associated with increased odds of elevated TC, with adjusted ORs (odds ratio) of 2.76 (95% confidence interval [CI]: 1.47-5.30, p=0.002) and 1.95 (95% CI: 1.05-3.70, p=0.038), respectively. Subgroup analyses indicated non-smokers and non-drinkers showed consistently elevated risk.
Conclusion
Occupational exposure to organic toxins, noise, and harmful gases may contribute to elevated TC levels in automobile manufacturing workers, highlighting the need for prevention strategies in occupational health settings.
Keywords
Introduction
Elevated total cholesterol (TC) is characterized by elevated levels of total cholesterol in the blood. Cell membranes, steroid hormones, and vitamin D synthesis depend on cholesterol. 1 Cholesterol is a major modifiable risk factor for coronary artery disease and ischemic heart disease. 2 Notably, in 2019, approximately 44% of global deaths attributable to ischemic heart disease were linked to high cholesterol levels. 3 The FOURIER trial demonstrated that intensive lipid-lowering therapy can reduce cardiovascular death, myocardial infarction, or stroke by 20% over two years (7.4% vs. 5.9%). 4
A large number of epidemiological studies indicate elevated TC is influenced by a variety of factors, including gender, 5 age, 6 Body Mass Index (BMI), 7 family history, 8 and lifestyle behaviors. 9 Occupational hazard exposures and lipid profile alterations remain poorly characterized. Low-dose occupational ionizing radiation may contribute to the development of lipid abnormalities. 10 Polycyclic aromatic hydrocarbons have been linked to elevated levels of TC, triglycerides (TG), and low-density lipoprotein cholesterol (LDL-C). 11 Occupational lead exposure correlates with serum cholesterol levels. 12
It is common for workers to be exposed to multiple hazards simultaneously. A worker in an automobile manufacturing plant may encounter several occupational hazards concurrently, and any health effects may be the result of the combined or interactive effects. Latent class analysis (LCA) is a probabilistic modeling technique used to classify study populations into distinct subgroups based on shared characteristics. The method maximizes inter-class heterogeneity while minimizing intra-class variability. 13 This study aims to assess multiple occupational hazard exposures among workers in a Nanjing automobile manufacturing plant based on LCA. Further, it investigates the effects of different exposure profiles on cholesterol levels, providing scientific evidence to understand elevated TC etiology and develop targeted interventions.
Subjects and methods
Subjects
The reporting of this study conforms to STROBE guidelines. 14 This was a retrospective cross-sectional study. We analyzed existing occupational health examination data and occupational hazard exposure records to assess the association between hazard exposures and elevated TC. This study included workers who underwent occupational health examinations at Nanjing Prevention and Treatment Center for Occupational Diseases in 2024. Participants were selected consecutively. The inclusion criteria were: workers with at least one year of continuous employment. The exclusion criteria were: (1) individuals who declined blood sampling on the day of examination; (2) workers with a history of hyperlipidemia prior to occupational exposure; and (3) female workers, who were excluded due to insufficient sample size (n < 10). In total, 708 participants were included. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2024). The study protocol was approved by the Ethics Committee of Nanjing Occupational Disease Prevention Institute (Approval No. 2025-005, Date: July 7, 2025). Because this was a retrospective study using existing data from occupational health examinations and occupational hazard records provided by the plant's administrative department, the requirement for written informed consent was waived by the Ethics Committee. All patient data were anonymized and de-identified prior to analysis. No personal identifiers, such as names, ID numbers, or contact details, were included in the dataset or the manuscript.
Definitions of variables
BMI was calculated as weight (kg) divided by height squared (m2). Normal weight was defined as BMI < 24.0 kg/m2, while overweight or obesity was defined as BMI ≥ 24.0 kg/m2. TC levels were measured using the cholesterol oxidase-peroxidase aminoantipyrine phenol enzymatic method on fasting venous blood samples collected in the morning. Elevated TC was defined according to international standards; a TC level ≥ 200 mg/dL was classified as elevated. This cut-off aligns with the lower limit for “borderline high” defined by guidelines from the Mayo Clinic (USA) 15 and the Ministry of Health (Singapore). 16 Smoking history was defined as smoking at least three cigarettes per day for a continuous or cumulative period of six months or more. Alcohol consumption was defined as drinking at least once a week during the past year.
Data collection
Demographic data, smoking history, alcohol consumption history, medical history, BMI, and blood pressure measurements were collected on the day of the medical examination. Occupational exposure records were obtained directly from the plant’s administrative department. The determination of exposure status was based on the plant’s historical occupational hazard evaluation reports and task assignments, providing qualitative binary indicators (exposed/not exposed) for each hazard rather than quantitative exposure measurements. Nine key occupational risk factors were selected after excluding those with exposure rates below 10%. The nine occupational hazardous factors to which workers were exposed included: noise, dust, benzene series (toluene and xylene), gasoline, formaldehyde, zinc oxide, nitrogen oxides, carbon monoxide, and isocyanates. The laboratory tests included complete blood counts, urinalysis, liver and renal function tests, and lipid profiles.
Statistical analysis
The statistical analyses were performed using the R software package (v4.3.3). Categorical variables were presented as frequencies and percentages, and differences between groups were analyzed using the Chi-square (χ2) test. Age was approximately normally distributed and presented as mean ± standard deviation (SD). Skewed distributed data are presented as median (interquartile range, Q1, Q3), and differences between groups were analyzed using the Mann-Whitney U test. Using the “poLCA” package in R, LCA was conducted using the exposure status of each worker as an indicator variable. Lo-Mendell-Rubin adjusted likelihood ratios (LMR) and Bootstrap likelihood ratio tests (BLRT) were used for model selection. P < 0.05 showed that a K-class model fit significantly better than a (K-1)-class model. The Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) were also evaluated, where lower values indicated a better fit. Entropy, which ranges from 0 to 1, was used to assess classification accuracy. The entropy value of 0.80 suggested a classification accuracy over 90%, with higher values suggesting greater precision.17–19
Analysis of influencing factors
Logistic regression models were constructed to evaluate the association between occupational risk factor classes and elevated TC levels during health examinations. TC status was defined as a binary outcome (normal TC = 0, elevated TC = 1). Four sequential models were developed: Model 1 included only LCA classes (Class 1 as reference); Model 2 further adjusted for age and BMI; Model 3 additionally incorporated smoking history (never smoked = 0, ever smoked = 1) and alcohol consumption frequency (never/occasional drinking = 0, frequent drinking = 1); and Model 4 further adjusted for prior hypertension diagnosis (no history = 0, yes = 1). Each model was independently analyzed to assess the incremental contribution of potential confounding variables on TC elevation.
Subgroup analysis
Subgroup analyses were conducted stratified by BMI (BMI ≥ 24 kg/m2 vs. < 24 kg/m2), smoking history, and alcohol consumption history. The effects of occupational risk factors on elevated TC were evaluated across these subgroups. Statistical tests were two-sided, with a p-value of 0.05 considered significant.
Results
LCA
Model fit indices and class sizes for LCA.
Note. AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion; BLRT: Bootstrap Likelihood Ratio Test; LMR: Lo-Mendell-Rubin test. Class sizes are derived from the predicted class memberships.
The conditional probability distributions of the five categories for each item.
Demographic characteristics of the normal group and the elevated TC group
Demographic characteristics of the normal group and the elevated TC group.
Note. BMI: Body Mass Index.
Risk factors associated with elevated TC
Logistic regression analysis of risk factors for elevated TC in workers.
Note. Model 1: unadjusted; Model 2: Model 1 + age + BMI; Model 3: Model 2 + Smoking + Alcohol.
Significant codes: p<0.01: “**”; p<0.05: “*”.
Subgroup analysis of elevated TC risk factors
In the overall population, after adjusting for age, BMI, smoking history, alcohol consumption history, and prior hypertension diagnosis, the risk of elevated TC was significantly higher in Class 4 (Organic toxins Exposure Group) and Class 5 (Combined Noise and Harmful Gases Exposure Group) compared to Class 1 (Low Exposure Group), as detailed in Model 4 of Table 3.
Subgroup analyses stratified by BMI, smoking history, and alcohol consumption history showed that all p values for interaction were greater than 0.05. Notably, among workers with normal BMI, the risk was particularly pronounced in Class 4 (Organic toxins Exposure) (OR=3.03, 95% CI: 1.06-9.31, p=0.043). In overweight/obese workers, Class 4 remained significantly associated with elevated TC (OR=2.58, 95% CI: 1.18-5.84, p=0.019). Among non-smokers, the association between exposure and elevated TC was particularly strong. Class 3 (Dust exposure) showed an OR of 2.77 (95% CI: 1.19-6.90, p=0.022), while Class 4 showed an OR of 4.13 (95% CI: 1.65-11.01, p=0.003). Among non-drinkers, both Class 4 and Class 5 remained significantly associated with elevated TC. Class 4 showed an OR of 2.82 (95% CI: 1.47-5.54, p=0.002), while Class 5 showed an OR of 1.95 (95% CI: 1.03-3.76, p=0.043). The forest plot of subgroup analyses is presented in Figure 2. Subgroup analyses of factors influencing elevated TC in workers.
Sensitivity analyses
Baseline characteristics of the study population stratified by high TC status (sensitivity analysis).
Note. TC, Total Cholesterol; IQR, Interquartile Range; SD, Standard Deviation.
Sensitivity analyses of the association between latent classes and TC using different outcome definitions.
Discussion
Prior studies investigating the impact of occupational hazards on workers’ health have primarily focused on isolated exposures, such as benzene, noise, and dust. Often, workers are exposed to more than one hazardous factor. In this study, LCA was applied to nine major occupational hazards encountered in an automobile manufacturing plant, identifying five classes of exposure. Elevated TC has been found to be prevalent in the study population, with elevated TC levels significantly associated with exposure to organic toxins, as well as noise and harmful gases combined. Mixed occupational exposures elevate TC levels, thus providing a new indicator for identifying risk factors for dyslipidemia and cardiovascular metabolic abnormalities. Further research on mechanisms and dose-response relationships will contribute to scientific evidence for cardiovascular prevention strategies and practical guidance for enterprises seeking to improve workplace environments and monitor employee health.
The study found that 191 out of 708 workers (27.0%) with occupational medical examinations had elevated TC. In China’s Patient-centered Evaluative Assessment of Cardiac Events Million Persons Project, 20 a nationwide survey conducted between 2014 and 2019 reported a prevalence of hypercholesterolemia of 7.1% among individuals aged 35 to 75. As a result of this significant discrepancy, occupational populations may be exposed to greater health risks. In the United States, coronary heart disease mortality rates declined after 1968, with a reduction of over 40% between 1980 and 2000. Modifiable risk factors accounted for approximately 44% of this decline, with cholesterol reduction contributing 24%. 21 A healthy cholesterol level reduces cardiovascular disease mortality.
Workers in the automotive manufacturing industry are commonly exposed to organic toxins during painting, maintenance, and interior assembly. These toxins include toluene, xylene, and gasoline. As compared to those exposed to Class 1, individuals in Class 4 exposed to gasoline, formaldehyde, isocyanates, toluene, and xylene had a significantly higher risk of developing elevated TC. Cumulative exposure to these chemicals was associated with elevated cholesterol levels with an OR of 2.76 (95% CI: 1.47–5.30).
Previous studies have demonstrated significant positive correlations between volatile organic compounds (VOCs) and both TC and high-density lipoprotein cholesterol (HDL-C). 22 Benzene, toluene, and xylene are principal aromatic hydrocarbons found in gasoline. 23 Exposure to BTEX (benzene, toluene, ethylbenzene, and xylenes) has been associated with elevated TG, TC, and LDL-C levels. 24 Evidence suggests that benzene enhances the expression of fatty acid transport proteins and β-oxidation enzymes. 25 This increased fatty acid transport into mitochondria results in two key metabolic consequences: first, the oxidation of fatty acids within mitochondria generates acetic acid, a precursor for hepatic cholesterol biosynthesis; 26 second, excessive mitochondrial fatty acid oxidation disrupts the mitochondrial antioxidant system, 27 leading to mitochondrial dysfunction, elevated reactive oxygen species (ROS), and lipid peroxidation. In addition, VOCs have been shown to increase white blood cell counts and promote the release of inflammatory cytokines, 28 which may further perturb lipid metabolism. These inflammatory responses are potentially involved in the regulation of cholesterol synthesis, transport, and clearance.
Formaldehyde and isocyanates do not directly elevate cholesterol levels, but new research suggests that oxidative stress may trigger Wnt/β-catenin signaling, which in turn leads to cholesterol accumulation. 29 Exposure to formaldehyde causes systemic redox imbalance and oxidative damage, 30 while exposure to isocyanates reduces glutathione, superoxide dismutase, and catalase levels. 31 Through indirect mechanisms, these effects may contribute to elevated TC.
Occupational noise exposure alters lipid metabolism. Each one-unit increase in cumulative noise exposure corresponds to a 0.013 increase in TC/HDL-C. 32 Also, noise exposure induces significant oxidative stress in the human body, 33 which may contribute to elevated cholesterol levels. However, the risk of elevated TC in the noise exposure group did not differ significantly from that in Class 1 (p = 0.097). A possible influence of noise exposure on cholesterol homeostasis cannot be conclusively ruled out due to the small sample size. Notably, in the sensitivity analysis defining the outcome as TC ≥ 240 mg/dL, Class 2 (Noise Exposure) also exhibited a significant association. However, this result should be interpreted with extreme caution. The prevalence of high TC (≥240 mg/dL) in the reference group (Class 1) was exceptionally low (only 2 cases, 1.8%), resulting in a statistically unstable estimate with a very wide confidence interval (OR = 4.57, 95% CI: 1.03–20.36). In contrast, the main analysis using the standard cutoff (TC ≥ 200 mg/dL) did not show a significant association for Class 2. Therefore, the current evidence is insufficient to conclude a definitive link between noise exposure and severe hypercholesterolemia, and this finding may be attributed to random variation due to the limited number of cases in the reference group. Further studies with larger sample sizes are needed to verify this association.
Elevated TC risk was significantly higher in the Class 5 combined noise and harmful gases exposure group than in the Class 1 group, with an OR of 1.95 (95% CI: 1.05–3.70). Nitrogen dioxide exposure is associated with elevated levels of serum cholesterol and LDL-C. 34 Nitrogen oxides, sulfur dioxide, and carbon monoxide also contribute to metabolic disorders, such as hypertension and dyslipidemia, primarily through systemic inflammation. A reduction in cholesterol efflux capacity, antioxidant activity, and anti-inflammatory potential is particularly associated with these effects. 35 Elevated TC may be exacerbated by the combined effects of noise and harmful gases on oxidative stress, resulting in oxidative damage and disrupted lipid metabolism.
The subgroup analysis revealed that compared with Class 1, Classes 4 and 5 did not show a statistically significant increase in the risk of elevated TC among workers who frequently smoked and consumed alcohol. In contrast, Classes 4 and 5 in the non-smoking and non-drinking group had a higher risk of elevated TC than Class 1, or demonstrated a trend toward increased risk. Smoking increases serum levels of TC and LDL-C, as well as low levels of HDL-C. Oxidative stress induced by smoking may exacerbate lipid metabolic disturbances. 22 Alcohol consumption has also been linked to dyslipidemia. 36 Regular smokers and drinkers may have already experienced long-term cumulative damage, making their responses to additional risk factors less pronounced. Moreover, existing research indicates that BTEX exerts a greater influence on cardiovascular disease risk among smokers and non-drinkers. In particular, ethylbenzene and o/p-xylene show more pronounced effects on individuals without alcohol consumption habits who also suffer from diabetes and hypertension. 24
These preventive and control measures are recommended based on the aforementioned research. Replace toxic paints with water-based or non-toxic alternatives, as well as strengthen the management of hazardous substances in general. Optimize job assignments to minimize noise and toxic gas exposure simultaneously. Likewise, this strategy may benefit traffic police officers and industrial workers. Improve lifestyle management by promoting weight control, smoking cessation, and moderate alcohol consumption. Protect against oxidative damage with targeted antioxidant protection.
Strengths of the study
Despite numerous established risk factors for dyslipidemia, there is limited evidence for occupational exposures to abnormal lipids. By investigating the relationship between occupational hazardous agents and TC levels, this study provides actionable insight for the prevention and early management of dyslipidemia.10,11 Moreover, examining multiple co-occurring occupational hazards in a cumulative manner could also provide a better reflection of real-world exposure conditions, thus addressing a critical gap in occupational epidemiology. It lays a foundation for mechanistic studies on how occupational exposure disrupts lipid metabolism and informs precision-oriented occupational health interventions.
Limitation of the study
The study has several limitations. First, this study is limited by its retrospective cross-sectional design. Exposure and outcome were measured simultaneously, so a cause-effect relationship cannot be established. The design is also susceptible to recall and selection biases. Therefore, pre-existing unknown elevated TC among workers is difficult to exclude. In addition, this study relied on occupational health examination data and did not gather information on diet, physical activity, shift work, and length of service. Furthermore, we were unable to adjust for socioeconomic differences by job title due to data limitations. Lack of these data limited the study’s accuracy. In the case of occupational hazards that cannot be quantified, it is not possible to verify a dose-effect relationship with cholesterol, and only an association can be demonstrated. Further studies can use a multivariate regression model to eliminate confounding variables like diet, exercise, and clinical treatment history, quantify harmful exposure factors, and accurately reveal the causal relationship between the study subjects and lipid levels. Finally, LDL-C is the most clinically relevant component of TC. 37 The plant measured only TC and did not test LDL-C due to economic constraints. To confirm these findings, longitudinal studies with comprehensive exposure assessments and lipid profiling are needed.
Conclusion
This study found a prevalence of elevated TC of 27.0% among the automobile manufacturing workers. Occupational exposure to organic toxins and noise combined with harmful gases has been linked to elevated cholesterol levels. Noise and dust may also influence cholesterol metabolism. Therefore, it is essential to develop and implement targeted occupational health interventions aimed at reducing lipid levels and safeguarding cardiovascular health.
Footnotes
Ethical considerations
The Ethics Committee of the Nanjing Occupational Disease Prevention Institute approved this study (Approval No. 2025-005, Date: July 7, 2025).
Author contributions
HJ conceived and designed the study, collected and analyzed the data, drafted the manuscript. DL collected and analyzed the data, YC drafted the manuscript. Critical revision of manuscript: all authors. All authors read and approved the final manuscript.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author, Y. Chen. The data are not publicly available due to privacy or ethical restrictions.