Introduction

Ambient air pollution has long been recognized as an essential risk factor for mortality and progression of cardiovascular disease (CVD) (Dehghani et al. 2023; Dehghani et al. 2022). Ethylene oxide (EtO) is a chemical substance extensively utilized in numerous industrial processes, particularly in producing ethoxylated compounds, ethanolamines, and ethylene glycol ethers (Lynch et al. 2022; Kirman et al. 2021). However, in contrast to its widespread application, the health risks associated with EtO are of significant concern. Existing studies have confirmed the carcinogenicity of EtO in both in vitro and animal experiments (Brockmann et al. 2006; Lynch et al. 1984). Long-term exposure to EtO, especially within industrial production environments, has shown a positive correlation with risks of hematopoietic system tumors and breast cancer (Jones et al. 2023; Steenland et al. 2003). More alarmingly, EtO as an alkylating agent is believed to induce gene mutations and chromosomal aberrations by forming adducts with DNA (Kostoryz et al. 2007; Kostoryz et al. 2004). In recent years, apart from the health above risks, research has also suggested potential links between EtO and cardiovascular diseases (Zeng et al. 2021), chronic obstructive pulmonary disease (COPD) (Huang et al. 2023), and diabetes (Guo et al. 2021), but these relationships require further investigation.

Cardiovascular health (CVH) was first introduced by the American Heart Association (AHA) in 2010. It was measured through the “Life’s Simple 7” (LS7) score, encompassing seven critical health behaviors and factors (Lloyd-Jones et al. 2010). This initial scoring provided the medical community with a comprehensive and quantifiable method to assess and monitor the cardiovascular health of individuals and populations (Kaneko et al. 2022). However, with the deepening of research and accumulation of scientific evidence, the AHA revised this score in 2022, introducing the “Life’s Essential 8” (LE8) score. Building upon the foundation of LS7, LE8 incorporated sleep health as a significant new component and optimized the scoring algorithm (Lloyd-Jones et al. 2022). This update reflects a more profound understanding of cardiovascular health, emphasizing the importance of sleep to CVH. Furthermore, employing the LE8 score aids in more accurately evaluating and predicting the risk of cardiovascular diseases, offering both the public and health professionals a more comprehensive tool to promote the adoption of healthy lifestyles and behaviors (Wang et al. 2023).

Recent findings underscore EtO’s role in initiating inflammatory responses, which is pivotal in cardiovascular pathologies (Li et al. 2023; Zhu et al. 2022). Such inflammation, potentially exacerbated by EtO, has implications for atherosclerosis, arterial stiffness, and other cardiovascular anomalies (Ferrucci and Fabbri 2018). The LE8 score, representing a comprehensive measure of CVH, accentuates the need to explore EtO’s multifaceted implications in this realm. While NHANES-based studies have previously investigated EtO’s associations with conditions like COPD (Huang et al. 2023), its direct impact on CVH, especially within the LE8 framework, remains largely unexplored in the literature. Given the prevalent industrial exposure to EtO and its potential wide-ranging health implications, understanding its influence on CVH is crucial. Thus, this study leverages the National Health and Nutrition Examination Survey (NHANES) to probe the relationship between EtO exposure and CVH delineated by LE8 metrics.

Methods

Study population and design

The NHANES, a program orchestrated by the National Center for Health Statistics (NCHS), operates as a cross-sectional and nationally representative investigation designed to evaluate civilians’ health and nutritional status across the USA (Xie et al. 2023). This program, maintained since 1999, utilizes a multifaceted approach comprising physical examinations and interviews to glean demographic, socioeconomic, and health-related data (Xie and Zhang 2023). Detailed explications of the study design and methodology are accessible on the NHANES website. The scope of the current investigation extends to data derived from multiple NHANES cycles: 2013–2018. Additionally, we incorporated information from the NHANES Linked Mortality File, bridging the gap between NHANES participants and the National Death Index data until December 31, 2019. Participants, aged from 20 to 80 years. Exclusion criteria were stringently implemented, excluding 3711 individuals pregnant during the examination or those with missing values for specific variables such as CVH; 110 individuals without available data on ethylene oxide; 356 self-reported cardiovascular disease at baseline (Fig. 1). The recruitment process adhered to ethical guidelines with the survey protocol, securing approval from the NCHS Research Ethics Review Board. All participants were granted written informed consent, and our investigation aligns with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Fig. 1
figure 1

Flowchart of participants selection. NHANES, National Health and Nutrition Examination Survey

Full size image

Cardiovascular health evaluation

CVH is gauged using the LE8 score, which integrates eight essential factors: diet, physical activity, tobacco/nicotine use, sleep patterns, BMI, non-HDL cholesterol levels, blood glucose, and blood pressure readings. The diet metric utilizes the Dietary Approaches to Stop Hypertension (DASH) diet score. Data on physical activity (minutes of moderate or intense weekly exercise), tobacco/nicotine habits (active tobacco consumption and exposure to secondhand smoke), sleep patterns (total sleep hours), and medication intake were sourced from standardized surveys. Weight, height, blood glucose, and blood pressure measurements were procured in specialized mobile centers adhering to established procedures. BMI is determined by dividing the individual’s weight in kilograms by their height in meters squared. Blood pressure averages were computed using all recorded readings at the initial assessment. Enzymatic methods were employed to measure serum cholesterol, and the difference between total and HDL cholesterol gave the non-HDL cholesterol. High-performance liquid chromatography was used to measure glycated hemoglobin. Comprehensive details and the scoring methodology for each CVH factor can be found in Table S1 and referenced studies (Ma et al. 2023; Sun et al. 2023; Zhang et al. 2023). Each CVH metric scores between 0 and 100. The composite CVH score is the average of all eight individual scores. According to American Heart Association guidelines, CVH scores of 80–100, 50–79, and 0–49 denote high, moderate, and low CVH levels (Lloyd-Jones et al. 2022).

Assessment of EtO

EtO exposure is ascertained via its Hemoglobin adduct (HbEtO). Owing to the protracted biological half-life of HbEtO, approximately 4 months, it serves as an efficacious surrogate marker for evaluating EtO exposure (Song et al. 2023). In the employed methodology, blood specimens are procured from participants after a fasting duration exceeding 9 h, subsequently preserved at − 30 °C, and dispatched to the National Center for Environmental Health. The quantification process harnesses the modified Edman reaction. After that, the isolation of EtO Edman derivatives is subjected to analytical scrutiny via high-performance liquid chromatography coupled with tandem mass spectrometry (HPLC-MS/MS), ensuring optimal precision in EtO detection. Concurrently, total hemoglobin quantification is undertaken, with commercial kits facilitating the standardization of EtO determinations. Finalized EtO concentrations are denoted in pmol/g Hb units. To guarantee the integrity and honesty of outcomes, rigorous internal and external quality control protocols are meticulously enforced throughout the analytical procedure.

Assessment of mortality

Mortality origins were defined based on the tenth version of the International Classification of Diseases (ICD-10). The terminus of the investigation included all-cause and cardiovascular disease-related mortalities. The latter encapsulated any death attributed to cardiovascular conditions, specifically assigned ICD-10 codes ranging from I00 to I99. Mortality data acquisition was facilitated by cross-referencing the NHANES datasets with the National Death Index, extending until December 31, 2019. All-cause mortality was classified as any recorded cause of death, while CVD mortality was identified by ICD-10 codes I00–I09, I11, I13, I20–I51, and I60–I69.

Covariables

Personal details, including age, gender, racial or ethnic identity, and educational level, were gathered through standard questionnaires. Racial or ethnic categories included White, Black, Mexican American, other Hispanic, and a category for other races (including Asian or multiracial). Educational attainment was segregated into three categories: those with less than a high school degree, those with a high school degree, and those with some college education or more. Information on family income was utilized to calculate the family income to poverty ratio, dividing the total family income by the official poverty threshold, which was further classified into low (< 1.3), intermediate (≥ 1.3 and < 3.0), and high (≥ 3.0) groups. Medical conditions such as the use of anti-hypertensive or lipid-lowering medications and the presence of diabetes were also included in the analysis. Physical examinations were conducted to gather parameters such as BMI.

Statistical analysis

To guarantee the national representativeness of the findings, analytical considerations included sampling weights for the overrepresentation of certain demographic groups and complex sample structures. CVH was classified into low (< 50), intermediate (50–79), and high (≥ 80) based on the LE8 score. Categorical variables were recorded as percentages (weighted numbers), while continuous variables were presented as the mean ± standard deviation (SD). Missing covariate data were dealt with using multiple imputation techniques. The associations between ethylene oxide and the comprehensive CVH score and eight subscores were examined using linear regression models. Cox regression models were deployed to scrutinize the connection between ethylene oxide and mortality from all causes and cardiovascular mortality. Three regression models were utilized as follows: model 1 was unadjusted for covariates (but adjusted for the other seven subscores when evaluating CVH subscores); model 2 enhanced model 1 by including age, sex, and ethnicity; and model 3 further augmented model 2 by integrating education level, family income to poverty ratio, usage of hypertension or cholesterol-lowering medication, and diabetes status. Further sensitivity analyses were performed after excluding participants taking anti-hypertensive or lipid-lowering drugs. To investigate potential non-linear associations between EtO and CVH, a restricted cubic spline (RCS) was applied. Additionally, threshold effect analysis was deployed to explore possible inflection points within the non-linear relationship. All statistical procedures were conducted using R software version 4.3.0. All statistical tests were two-sided, with statistical significance denoted by P values < 0.05.

Results

Baseline characteristics

In the present analysis, 3748 adults, approximating 58.2 million U.S. adults, were included. The mean (SD) for the LE8 score was 65.78 (15.44), while the mean (SD) for HbEtO was 57.48 (103.17) pmol/g Hb. Figure 2 depicts the distribution of the CVH total score and eight subscores among all participants. The participant distribution, as determined by their CVH status (LE8 score), was as follows: 15.8% (equating to 9.2 million individuals) exhibited low CVH (LE8 < 50), 63.6% (approximating 37.0 million individuals) had moderate CVH (50 ≤ LE8 < 80), and 20.6% (corresponding to 12.0 million individuals) were classified with high CVH (LE8 ≥ 80). Initial evaluations revealed that those participants categorized with high CVH were generally younger and, more frequently, women and White. They were also more likely to possess higher educational achievement and more significant family income. However, these individuals exhibited a lower prevalence of diabetes, decreased utilization of anti-hypertensive or lipid-lowering medications, and diminished rates of all-cause and cardiovascular-related mortality, as detailed in Table 1.

Fig. 2
figure 2

Violin plot of distribution of Life’s Essential 8 total score and 8 subscores

Full size image
Table 1 Baseline characteristics of participants with different CVH levels estimated from LE8 score
Full size table

Association of HbEtO and the LE8 scores

Table 2 demonstrates the association of EtO with the total CVH score and the eight CVH subscores. In all models, ln-transformed HbEtO levels maintained a negative correlation association with total CVH score, and a 1-unit increase in ln-transformed HbEtO in the fully adjusted model was associated with a 3.69-point decrease in total CVH score [β = 3.69, 95% CI (− 4.15, − 3.23)]. In the association between HbEtO and the 8 CVH subscores, HbEtO was negatively correlated with 6 CVH subscores (DASH diet score, physical activity score, tobacco exposure score, sleep health score, blood lipid score, blood glucose score). Notably, the strongest negative correlation was observed between HbEtO and tobacco exposure score, where a 1-unit increase in ln-transformed HbEtO was correlated with a decrease of 23.92 in the tobacco exposure score [β = 23.92, 95% CI (− 24.84, − 22.99)]. Conversely, HbEtO demonstrated positive correlations with the body mass index score [β = 3.73, 95% CI (2.59, 4.87)] and blood pressure score [β = 1.35, 95% CI (0.44, 2.25)]. Sensitivity analysis was further performed, and the association between EtO and CVH was similar after excluding participants who took anti-hypertensive drugs or lipid-lowering drugs. The results showed that the negative effect of EtO on CVH may be stronger in healthy people [β = − 4.52, 95% CI (− 5.08, − 3.97)] (Table S2).

Table 2 The associations between ln-transformed HbEtO and the Life’s Essential 8 Cardiovascular Health (CVH) score
Full size table

Figure 3 depicts the non-linear associations between EtO, the overall CVH score, and the 8 CVH subscores. It is worth noting that an inverted J-shaped relationship was evident between ln-transformed HbEtO and the CVH overall score, with an inflection point at 3.15 pmol/g Hb (log-likelihood ratio < 0.01). When ln HbEtO levels were below 3.15 pmol/g Hb, no significant association was found between ln-transformed HbEtO and the CVH overall score [β = − 0.07, 95% CI (− 1.61, 1.47)]. However, when ln HbEtO levels exceeded 3.15 pmol/g Hb, a negative correlation was established between ln-transformed HbEtO and the CVH overall score [β = 4.49, 95% CI (− 5.09, − 3.90)].

Fig. 3
figure 3

Dose–response relationships between ln-transformed ethylene oxide levels and Life’s Essential 8 total score and 8 subscores. Beta (solid lines) and 95% confidence levels (shaded areas) were adjusted for age, gender, race, education level, ratio of family income to poverty, use of anti-hypertensive or lipid-lowering medication, and diabetes

Full size image

Association of HbEtO with all-cause mortality and cardiovascular mortality

Among all 3748 participants in the cohort, 87 all-cause mortality cases and 14 cardiovascular-related mortality cases were recorded over a follow-up duration ranging from 0 to 85 months, with a median follow-up time of 49 months. No significant association was observed between HbEtO and all-cause mortality in the fully adjusted model. However, a positive correlation was identified between HbEtO and cardiovascular-related mortality, wherein a one-unit increase in ln-transformed HbEtO was associated with a 71% increase in the risk of cardiovascular mortality. Participants in the highest quartile of HbEtO levels experienced an 11.66-fold higher risk of cardiovascular death than those in the reference group [HR = 12.66, 95% CI (1.17, 137.38)] (Tables 3 and 4).

Table 3 Threshold effect analysis of ln-transformed HbEtO on total CVH score using the two-piecewise linear regression model
Full size table
Table 4 The associations between ln-transformed HbEtO with all-cause mortality and cardiovascular mortality
Full size table

Discussion

In this nationally representative study, we discerned an inverse association between EtO exposure, as quantified by HbEtO levels, and the overall CVH score delineated by the LE8 metrics. Specifically, elevated EtO exposure was correlated with deteriorated CVH, significantly beyond an identified inflection point of 3.15 pmol/g Hb. This relationship was particularly pronounced in the context of tobacco exposure, while intriguingly, positive correlations emerged with metrics like body mass index and blood pressure. Additionally, while EtO exposure did not significantly influence all-cause mortality, it was robustly associated with increased cardiovascular mortality. These findings emphasize the potential cardiovascular implications of EtO, a chemical of widespread industrial relevance, highlighting the importance of monitoring and mitigating its health impacts.

To our knowledge, this is the first study investigating the association between EtO exposure and cardiovascular health scores. Previous research on the health implications of EtO has predominantly focused on its carcinogenicity, particularly in breast and lymphatic and hematopoietic cancers (Marsh et al. 2019; Jinot et al. 2018; Weiderpass et al. 2011). While there have been studies exploring the association of EtO occupational exposure with CVD and cardiovascular mortality (Gardner et al. 1989; Steenland et al. 2004), the general population also faces risks from EtO exposure through smoking or exposure to volatile organic compounds during renovation (Shintani 2017). Two recent population-based studies from the US have filled this research gap in the general population, suggesting a significant positive correlation between EtO exposure and angina pectoris, myocardial infarction, and hypertension (Zeng et al. 2021; Wu et al. 2022). However, no association was found with stroke and coronary heart disease. Given the cross-sectional design of these studies, the impact of EtO on the cardiovascular system remains ambiguous. In this study, we employed the LE8 score to quantitatively assess participants’ cardiovascular health levels, eliminating potential reverse causation from baseline CVD. Results indicated a negative correlation between EtO exposure and CVH when exposure reached a certain threshold. CVH scores have been demonstrated in multiple prospective studies from various regions, including the USA (Xia et al. 2023), UK (Wang et al. 2023), Finland (Isiozor et al. 2023), and China (Xing et al. 2023), to be associated with all-cause mortality and cardiovascular mortality, and related to a reduction in life expectancy. Our study found a significant correlation between increased EtO exposure and cardiovascular mortality but not with all-cause mortality. This outcome is somewhat consistent with the notion that ethylene oxide leads to a decline in CVH, resulting in increased cardiovascular mortality among participants. However, given the limited sample size of deceased participants, we advise a cautious interpretation of the Cox regression model results. We could not employ mediation analysis to elucidate potential causal associations due to statistical limitations.

The mechanisms underlying the observed associations between EtO exposure and cardiovascular health, as delineated by the LE8 metrics, warrant further exploration. A consistent negative relationship was evident between EtO exposure and the overall CVH score, suggesting a potentially detrimental impact on cardiovascular health with increased exposure. Notably, the most pronounced negative correlation was observed between EtO exposure and the tobacco exposure score. Since EtO is a by-product of tobacco smoke, this strong association underscores the compounded cardiovascular risk from direct tobacco exposure and its resultant EtO emissions (Jain 2020). Several plausible mechanisms might underpin these associations. Firstly, oxidative stress induced by EtO can compromise endothelial function, a cornerstone of cardiovascular health (Lynch et al. 1984; Cheang et al. 2022; Adedara and Farombi 2010). Moreover, inflammation, frequently triggered by EtO exposure, can adversely affect the vascular tone and renal function, both pivotal for cardiovascular well-being (Kuiper et al. 2008; Rasool et al. 2021). Interestingly, our study revealed positive associations of EtO exposure with BMI and blood pressure scores, implying that increased EtO exposure might be linked to improved CVH in these domains. For the BMI score, potential pathways include metabolic disruptions caused by EtO or its metabolites, which might influence lipid metabolism, and the inflammatory reactions linked to EtO exposure that could affect adipose tissue function (Tangvarasittichai 2015; Chen et al. 2003). In addition to the role of inflammation and oxidative stress, the blood pressure score and interactions with the renin-angiotensin system, a primary regulator of blood pressure, might be a contributory factor (Zhu et al. 2022). While our findings suggest that EtO exposure adversely affects several CVH dimensions, the precise underpinning mechanisms demand further elucidation.

Our study carries several strengths. Foremost, we utilized nationally representative data, providing a comprehensive understanding of EtO exposure and its association with cardiovascular health. This study is pioneering in exploring the relationship between EtO exposure and the LE8-defined CVH scores, offering novel insights into the potential cardiovascular implications of this ubiquitous industrial compound. The LE8 score, which encompasses multiple cardiovascular health metrics, provides a holistic perspective on cardiovascular health, enhancing the robustness of our findings. However, some limitations merit consideration. As with any cross-sectional study design, causality cannot be unequivocally established. This design, coupled with the reliance on self-reported measures for specific metrics, might introduce potential biases. In addition, due to database limitations, we were unable to include all factors that have a potential impact on outcomes, such as family history of CVD. Another potential limitation is the possible inflation of hazard ratios might need to be considered (Tzeng 2021). Lastly, as EtO exposure can vary over time, relying on a single measure might not fully encapsulate the cumulative exposure and its effects on CVH.

Conclusion

In summary, our study reveals a negative association between elevated EtO exposure and cardiovascular health, as delineated by the LE8 metrics, significantly beyond a certain threshold. The pronounced implications of tobacco exposure, coupled with potential impacts on body mass index and blood pressure scores, emphasize the multifaceted cardiovascular effects of EtO. Furthermore, while the study identifies a robust link between EtO and heightened cardiovascular mortality, the exact mechanistic pathways await deeper exploration.