📄 medicine

Exposome-wide Association Patterns across Cardiovascular Disease Outcomes: Findings from the NHANES

This cross-sectional exposome-wide association study of 56,253 NHANES participants reveals that cardiovascular outcomes, particularly composite CVD and heart failure, exhibit distinct yet reproducible environmental exposure signatures, with blood cadmium and serum cotinine emerging as key risk factors across multiple disease endpoints.

Original authors: Hao Wang, Yulu Zheng, Jian Geng, Zhiyuan Wu, Yahong Zhu, Ziqi Zhou, Hang Qu, Lois Balmer, Aimin Yang, Xingang Li, Zheng Guo

Published 2026-07-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Hao Wang, Yulu Zheng, Jian Geng, Zhiyuan Wu, Yahong Zhu, Ziqi Zhou, Hang Qu, Lois Balmer, Aimin Yang, Xingang Li, Zheng Guo

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your body is like a house that has been standing for decades. Over time, it gets exposed to all sorts of things from the outside world: the air you breathe, the water you drink, the smoke from a neighbor's fire, and the chemicals in your cleaning products. Scientists call this total collection of environmental exposures your "exposome."

For a long time, researchers have tried to figure out which specific "invader" causes the house to develop cracks (heart disease). But usually, they only looked at one invader at a time, like checking just the roof or just the windows.

This paper is different. The researchers acted like a super-inspector who walked through the entire house at once, checking 138 different potential "invaders" (chemicals and toxins) to see which ones were most likely to be associated with different types of heart trouble. They used a massive database called NHANES, which is like a giant, national snapshot of the health of over 56,000 Americans collected over 20 years.

Here is what they found, broken down simply:

1. The "Big Picture" vs. Specific Rooms

The researchers looked at six different types of heart problems: a general "heart disease" category, heart attacks, angina (chest pain), heart failure, stroke, and coronary heart disease.

The Finding: Some "invaders" attacked the whole house (causing general heart disease), while others seemed to target specific rooms.
The Winners: The two conditions that showed the clearest, most consistent patterns were Composite Cardiovascular Disease (the general category) and Heart Failure. These were the "rooms" where the damage was most obvious and linked to the environment.

2. The Top Suspects (The "Bad Guys")

When the researchers looked at which chemicals were most often found alongside heart problems, a few names kept popping up, like repeat offenders in a police lineup:

Blood Cadmium: Think of this as a heavy metal toxin often found in tobacco smoke and industrial pollution. It showed up as a strong suspect for almost every type of heart problem they checked.
Serum Cotinine: This is a chemical marker that proves someone has been exposed to tobacco smoke (either by smoking or breathing secondhand smoke). It was consistently linked to heart trouble.
Nitromethane & Ethylene Oxide: These are other chemical markers linked to industrial pollution and smoke.

The Twist: Even when they looked only at people who had never smoked, some of these links (like Cadmium) were still there, suggesting these chemicals are harmful to the heart even without the smoke.

3. The "Good Guys" (Or Maybe Just Confusing?)

Some chemicals showed up as being less common in people with heart disease. The researchers found that higher levels of certain chemicals (like a specific type of PFAS called n-PFOA and urinary nitrate) were linked to lower odds of heart disease.

The Catch: The authors warn us not to think these chemicals are "superheroes" that protect the heart. Instead, they suggest this might be a trick of the data. For example, people with heart disease might change their diet or habits, which changes their chemical levels. It's like seeing that people who carry umbrellas get less wet, but it doesn't mean the umbrella caused the rain to stop; it just means they prepared for it. The paper says we need more study to understand these "protective" links.

4. Does It Matter Who You Are?

The researchers checked if these "invaders" affected men and women, or young and old people, differently.

The Result: Mostly, the bad guys attacked everyone similarly. However, there were some differences. For instance, the link between certain chemicals and heart failure seemed stronger in specific groups based on sex or age. It's like how a drafty window might make an elderly person feel much colder than a young person, even if the draft is the same.

5. How Strong is the Evidence?

To make sure they weren't just seeing ghosts, the researchers ran the data through a "stress test":

They checked if the results held up if they removed one year of data at a time (like removing a brick to see if the wall falls). The results stayed stable.
They checked if the results changed if they only looked at non-smokers. Many links remained.
They checked if the results changed if they only looked at men or only women. Most links were consistent.

The Bottom Line

This study didn't prove that these chemicals cause heart disease (because they looked at a snapshot in time, not a movie over time). However, it successfully identified a shortlist of environmental suspects that are consistently found in people with heart issues.

The main takeaway is that Cadmium and tobacco-related chemicals are the most consistent "bad actors" across the board. The study suggests that if we want to understand heart disease better, we need to stop looking at just one chemical at a time and start looking at the whole "environmental neighborhood" these chemicals come from.

What the paper does NOT say:

It does not tell doctors to prescribe specific treatments based on these chemicals yet.
It does not claim that avoiding these chemicals will definitely cure heart disease (since the study is a snapshot, not a cause-and-effect movie).
It does not say that the "protective" chemicals are good for you; it just notes the strange pattern and calls for more research.

Technical Summary: Exposome-wide Association Patterns across Cardiovascular Disease Outcomes

Problem Statement
Cardiovascular disease (CVD) remains the leading cause of global mortality, yet a significant proportion of its risk burden remains unexplained even after accounting for traditional factors like obesity, hypertension, and dyslipidemia. While environmental determinants (e.g., heavy metals, air pollutants, persistent organic pollutants) are increasingly implicated in CVD pathogenesis through mechanisms such as oxidative stress and endothelial dysfunction, current epidemiological literature suffers from two primary limitations:

Reductionist Scope: Most studies evaluate single chemicals or specific exposure classes in isolation, failing to capture the complex mixtures of environmental factors to which populations are simultaneously exposed.
Outcome Specificity: Research typically focuses on a single cardiovascular endpoint (e.g., only coronary heart disease or only heart failure), making it difficult to distinguish exposure signatures that are shared across cardiovascular conditions from those that are outcome-specific.

There is a lack of systematic, agnostic screening that evaluates a broad panel of environmental exposures across multiple distinct CVD phenotypes to identify both common and unique environmental determinants.

Methodology
The authors conducted a cross-sectional exposome-wide association study (ExWAS) using a harmonized, CVD-augmented analytic dataset from the National Health and Nutrition Examination Survey (NHANES).

Data Source: The study utilized data from 10 NHANES survey cycles (1999–2018), comprising a total of 56,253 participants.
Exposure Panel: A curated panel of 138 environmental biomarkers was analyzed, covering blood, serum, and urine matrices. These included heavy metals, per- and polyfluoroalkyl substances (PFAS), persistent organic pollutants (POPs), hemoglobin adducts, phenols, arsenic species, pesticide metabolites, tobacco-related biomarkers, and volatile organic compound metabolites.
Outcomes: Six cardiovascular outcomes were defined based on self-reported physician diagnoses: composite CVD (presence of any of the following), coronary heart disease (CHD), heart attack, angina, heart failure, and stroke.
Statistical Analysis:
- Primary Model: Survey-weighted logistic regression models were fitted for each exposure-outcome pair. Exposure concentrations were log2-transformed, allowing coefficients to be interpreted as odds ratios (ORs) per exposure doubling.
- Significance Thresholds: A nominal $P < 0.05$ was used as the primary screening criterion. False Discovery Rate (FDR) correction (Benjamini-Hochberg) was applied as a sensitivity analysis to identify robust associations.
- Robustness Checks: The study employed multiple sensitivity analyses, including:
  - Subgroup analyses by sex and age (<60 vs. ≥60 years).
  - Restriction to never-smokers to disentangle tobacco-related effects.
  - Leave-one-cycle-out analyses to assess stability across survey cycles.
  - Restricted cubic spline models to evaluate non-linear associations.
  - Chemical-category enrichment analysis.

Key Contributions and Results
The study systematically mapped the associations between 138 environmental biomarkers and six CVD outcomes, revealing heterogeneous but interpretable exposure signatures.

Prevalence: Weighted prevalence estimates were 8.7% for composite CVD, 3.5% for CHD, 3.4% for heart attack, 2.4% for angina, 2.4% for heart failure, and 2.8% for stroke.
Association Counts:
- Composite CVD: Yielded 27 nominal associations, with 11 surviving FDR adjustment.
- Heart Failure: Yielded 26 nominal associations, with 16 surviving FDR adjustment (the highest number of robust findings).
- Other Outcomes: CHD, angina, stroke, and heart attack showed fewer FDR-significant associations (0, 4, 4, and 4 respectively).
Leading Exposures:
- Positive Associations: The strongest and most reproducible positive associations across multiple outcomes (particularly composite CVD, heart failure, heart attack, and stroke) involved blood cadmium and serum cotinine. Other notable positive signals included blood nitromethane and ethylene oxide hemoglobin adduct (EO-Hb).
- Negative Associations: Linear perfluorooctanoate (n-PFOA) and triclosan showed consistent negative associations with composite CVD and other outcomes. Urinary nitrate and certain PFAS (e.g., PFHxS) were negatively associated with heart failure.
Robustness and Heterogeneity:
- Of the 92 nominal associations carried forward for robustness evaluation, 87 remained directionally stable in leave-one-cycle-out analyses, and 85 remained significant in core covariate-adjusted models.
- Smoking: 30 associations remained significant among never-smokers. However, several top associations (e.g., blood cadmium, triclosan) were attenuated or lost significance in the never-smoker subset, suggesting strong confounding by smoking history.
- Subgroups: Significant exposure-by-sex interactions were observed for several biomarkers (e.g., serum cotinine and total mercury in composite CVD; urinary barium and n-PFOS in heart failure). Age-dependent interactions were less frequent but notable for blood cadmium in composite CVD.
Chemical Enrichment: Enrichment analysis highlighted blood heavy metals, urine metals, polychlorinated biphenyls (PCBs), and organochlorine pesticides as classes significantly enriched for associations across multiple cardiovascular outcomes.

Significance and Claims
The authors claim that this study provides the first comprehensive ExWAS evaluation of environmental exposure patterns across a broad spectrum of CVD phenotypes using a nationally representative sample.

Prioritization of Endpoints: The study suggests that composite CVD and heart failure are the most sensitive and reproducible endpoints for detecting environmental signals in cross-sectional data, likely due to their ability to capture cumulative cardiovascular injury.
Shared vs. Specific Signatures: The findings demonstrate that while some exposures (notably blood cadmium and serum cotinine) drive risk across multiple cardiovascular conditions, others exhibit outcome-specific patterns.
Tobacco and Oxidative Stress: The consistent positive association of tobacco-related biomarkers (cadmium, cotinine) and reactive hemoglobin adducts supports the hypothesis that oxidative stress and tobacco toxicants are central mechanisms in environmental CVD etiology.
Caveats on Inverse Associations: The authors explicitly caution that the observed negative associations for PFAS and urinary nitrate likely do not indicate a cardioprotective effect. Instead, they may reflect residual confounding, reverse causation, or complex pharmacokinetics, highlighting the limitations of cross-sectional designs in establishing causality.
Future Direction: The study positions its findings as a foundation for future longitudinal and mechanistic research. It aims to prioritize specific biomarkers and endpoints for further investigation into modifiable environmental determinants of cardiovascular risk, rather than claiming immediate clinical application or causal proof.

The paper concludes that while the cross-sectional design and reliance on self-reported diagnoses limit causal inference, the robustness of the signals across subgroups and cycles provides a compelling, data-driven map of the environmental landscape of CVD.