Investigating zinc’s role in mitigating blood lead levels’ toxicity on gut microbiota diversity: NHANES 2007–2010

Abstract

Introduction:

Zinc serves as a cofactor for numerous vital processes across species. Microorganisms that make up the gut microbiome rely on these zinc-dependent mechanisms to perform essential functions, contributing to a diverse and stable microbial environment. Environmental contaminants, such as lead, has been shown to disrupt diversity and stability. The purpose of this study was to determine whether zinc serves as an effect modifier against elevated blood lead levels (BLL) on gut microbiota diversity.

Methods:

The 2007–2008 and 2009–2010 NHANES datasets were utilized to conduct a cross-sectional complex survey analysis aimed at determining whether zinc intake acts as a protective factor against changes in microbiome diversity associated with BLL, using enterolactone (ENL) as a biomarker. A multiple linear regression was conducted to evaluate whether an interaction between BLL and zinc intake could predict ENL. The model included fiber intake and BMI as covariates.

Results:

BMI and fiber intake were identified as covariates. Fiber intake was a confounding variable in the relationship between zinc and ENL levels. Lead was found to decrease ENL levels (p = 0.002). The interaction between zinc and BLL was marginally significant (p = 0.089).

Conclusion:

This study suggests that lead’s impact on gut microbial diversity may depend on zinc status. These findings emphasize the importance of accounting for dietary confounders, such as fiber intake, to improve model accuracy and interpretation. While additional research is needed to confirm zinc’s potential protective role, public health strategies encouraging adequate zinc and fiber intake may in part help support microbial resilience and reduce lead’s effects on the gut microbiota.

1. Introduction

Approximately 10¹⁴ organisms inhabit the microbiota and exist in a symbiotic relationship with the host (Thursby and Juge, 2017). A diverse population of microbiota is associated positive health outcomes for the host (Durack and Lynch, 2019; Jin et al., 2024; Muller et al., 2021). Decreases in commensal bacteria diversity has been found in an array of medical conditions such as inflammatory bowel diseases cognitive decline, mental health issues, as well as metabolic and chronic diseases (Durack and Lynch, 2019; Jin et al., 2024; Muller et al., 2021; Lee et al., 2020). This suggests that diversity within the microbiota results in a stable microbial ecosystem, as varied species fulfill complementary roles, enhancing resilience to stressors (Houtz et al., 2022; Lozupone et al., 2012; Fassarella et al., 2021).

Lead is a pervasive environmental contaminant. It poses significant challenges to microbial communities by disrupting metabolic pathways, damaging cellular structures, and generating oxidative stress (Gao et al., 2017; Liu et al., 2021a). This disruption can result in a microbial imbalance, including the overgrowth of opportunistic species and a reduction in beneficial bacteria and compromise gut barrier integrity (Yu et al., 2021; Ghosh et al., 2023). Lead can then gain access to the bloodstream and impact the nervous, hematopoietic, renal, cardiovascular, reproductive, and skeletal systems (Liu et al., 2021a; Atsdr.; Flora et al., 2012).

Zinc is an essential cofactor for both the host and microbiota. To ensure sufficient zinc is available to support both host’s and the gut microbiota’s needs, which can utilize up to 20 % (Smith et al., 1972) of dietary zinc intake, the Recommended Dietary Allowance (RDA) for adults is 11 mg/day for males and 8 mg/day for females (Institute of Medicine US, 2001). Zinc absorption occurs primarily in the small intestine and is influenced by dietary factors such as the presence of phytates (which inhibit absorption), animal protein (which enhances it), and overall zinc status of the individual (Krebs, 2000). Deficiency in zinc can impair immune function, wound healing, and growth, and it has been linked to increased intestinal permeability and disruptions in gut microbial composition (Scarpellini et al., 2022). Conversely, chronic zinc excess, often due to over-supplementation, may interfere with the absorption of other essential minerals such as copper and iron, and can also alter gut microbial balance in ways that may not be beneficial (Lopez and Skaar, 2018).

Zinc’s role in gut microbiome diversity has been observed in several animal models, as well as in humans. In broilers, supplementation with zinc bacitracin at 50 ppm has been shown to improve alpha diversity by increasing beneficial bacteria like Lactobacillus and Faecalibacterium, which produce SCFAs such as butyrate, thus promoting gut health (Skalny et al., 2021). Similarly, human studies have demonstrated that zinc supplementation can modulate microbial diversity and improve gut function in conditions such as Crohn’s disease and irritable bowel syndrome (Scarpellini et al., 2022). In mice, control groups that received the standard zinc dose showed the greatest microbiome diversity, while zinc deficiency and high zinc doses led to decreased diversity. This suggests that there is an optimal range for zinc intake, with both extremes (deficiency and toxicity) disrupting the gut microbiome (Sauer and Grabrucker, 2019; Chen et al., 2021).

Zinc has been shown to protect against lead toxicity in various systems by reducing lead absorption and mitigating its toxic effects (Zhai et al., 2018). Research indicates that zinc supplementation lowers blood lead levels and restores disruptions caused by lead in oxidative stress, antioxidant enzyme activity, and organ function (Ugwuja et al., 2020; Wani et al., 2019; Cerklewski and Forbes, 1976). Specifically, zinc competes with lead for binding sites, enhances metallothionein production to sequester lead, and supports antioxidant defenses (Hudson et al., 2025). Studies in mice and rats have demonstrated zinc’s protective role by reversing lead-induced oxidative stress, improving biochemical parameters, and reducing organ damage (Ugwuja et al., 2020; Prasanthi et al., 2010). Additionally, zinc supplementation reduced calcium and lead resorption in the bones of fetal rats (Cerklewski and Forbes, 1976). In children, zinc status has been shown to influence the negative effects of lead on growth, with zinc deficiency exacerbating the detrimental impact of lead on stature (Cantoral et al., 2015).

Enterolignans, include ENL and enterodiol (END), are derivatives of plant lignin metabolism from gut microbes. Several studies have compared enterolignan concentrations in urine to 16S rRNA sequencing, which identifies and quantifies bacterial taxa by targeting a conserved region of the bacterial ribosomal RNA gene, allowing researchers to assess the composition and diversity of the gut microbiome. They found higher concentrations in the urine correlated with higher richness gut microbial community (Kase et al., 2022; Weiner et al., 2024; Hullar et al., 2015; Lampe, 2003; Shivappa et al., 2019). Their synthesis is only possible after the conversion of several other metabolites that require many species of bacteria in the gut microbiome to be present (Hullar et al., 2015; Li et al., 2022; Bess et al., 2020). It was also discovered ENL served as a better marker for diversity because its production relies on less abundant and more specialized bacterial taxa, whereas END is associated with common and dominant bacterial species (Clavel et al., 2005). Several of these specialized species, such as Bacteroides ovatus, Methanobrevibacter smithii, and Faecalibacterium prausnitzii, have been identified as key players in lignan metabolism but the exact pathways have not been mapped out as the process involves a consortium of bacteria, multiple pathways, and significant individual variability (Li et al., 2022). The ability of ENL to reflect general changes in microbiota diversity makes it a valuable tool for determining whether zinc mitigates lead’s negative impact on gut health in secondary data analysis using ENL as a biomarker.

This study is the first to explore whether zinc acts as an effect modifier, mitigating lead’s negative impact on microbial diversity at BLL found in the American population. The use of ENL as a biomarker for microbiome diversity introduces a novel approach to studying the relationship between zinc, lead, and gut health. Previous work using ENL from NHANES data has focused on associations with health outcomes such as asthma, heart disease, and mortality (Cardet et al., 2020; Frankenfeld, 2014). However, these studies have not investigated potential causes or predictors of ENL, such as environmental exposures. Zinc is a ubiquitous essential nutrient and is the preferred or sole co-factor for binding sites involved in many key pathways in both (Pajarillo et al., 2021). Therefore, ENL is being used to observe the relationship between lead, zinc, and gut microbiome diversity in this analysis. The findings have the potential to inform nutritional and public health strategies aimed at mitigating the effects of environmental toxin exposure.

2. Research methodology

2.1. Data collection

Data from the CDC’s NHANES 2007–2008 and 2009–2010 cycles were used to conduct a cross-sectional analysis. These cycles represent the years where zinc supplement intake and ENL levels overlapped in data collection. The NHANES datasets, which are based on a cross-sectional study design, employ a multistage, stratified, and clustered sampling approach with survey weights to ensure accurate representation of the American population. These methods were accounted for in the data analysis of this study. Zinc intake was assessed through 24-hour dietary recall interviews, where participants provided detailed reports of their food and beverage consumption. The USDA’s Food and Nutrient Database for Dietary Studies was used to calculate the nutrient content, including daily zinc intake, based on these records. BLL were measured using inductively coupled plasma mass spectrometry (ICP-MS), a highly accurate method for detecting lead concentrations in whole blood. Microbiome diversity was indirectly assessed through the measurement of ENL, a metabolite produced by gut bacteria, which was collected via urine samples collected at non-standardized times during participants’ visits to the Mobile Examination Center.

Zinc intake, fiber intake, BLL, BMI ENL, age, and creatinine were merged and uploaded into SPSS. The two 24-hour recalls for zinc food and supplement intake were both added and averaged to determine total zinc intake. Zinc intake was used as a measure in this study instead of blood zinc levels, as blood zinc is not a reliable indicator of zinc status due to its susceptibility to factors such as inflammation and recent food intake, (Berti et al., 2024) and because blood zinc levels were not collected during the relevant survey years. Missing food intake data were addressed using mean imputation, replacing absent values with the average intake for each cycle. For supplement intake data, Last Observation Carried Forward (LOCF) method was applied. The same was done for fiber intake. Individuals below the age of 18 and older than 50 were removed due to the changes in microbiome diversity at these stages of life (Wilmanski et al., 2021). ENL was adjusted for creatinine levels by dividing ENL, measured in ng/dL, by creatinine, measured in mg/dL, resulting in units of ng/mg/dL. This adjustment corrects for differences in urine dilution among individuals (Hullar et al., 2015; Sun et al., 2014). ENL was also common log transformed to account for its non-normal distribution. Lead levels were measured in the blood in μg/dL.

2.2. Complex survey analysis

A complex survey design was accounted for by applying the ’survey’ package in RStudio, which adjusts for the NHANES sampling weights, stratification, and clustering to ensure that the statistical estimates are representative of the U.S. population and that variance estimates are accurate. Descriptive statistics were run for all variables. Following diagnostic testing, a forward stepwise regression approach was employed to assess the relationships between the covariates and independent variables, focusing on their predictive capacity for ENL levels. An interaction term was incorporated to evaluate whether varying levels of zinc and BLL can predict ENL levels. The final model included BLL and zinc as independent variables, and accounted for fiber intake, BMI, and gender as covariates as well the zinc and BLL interaction term. AIC was used to assess model fit and compare the relative efficiency of models in approximating the data; although originally developed for simpler datasets, AIC is commonly recommended for use in complex surveys like NHANES to aid in model selection (Lumley and Scott, 2015). All statistical analyses and visualizations were performed using RStudio (Version 2023.09.1 +494) accessed through Anaconda.

3. Results

3.1. Descriptive analysis

The descriptive statistics revealed key trends in the population. The average BMI was 28.16 kg/m² (SD = 7.03), indicating that the population, on average, falls within the overweight category (BMI ≥ 25). The average daily fiber intake was 16.36 g/day (SD = 8.09), significantly below the recommended levels of 25 g/day for adult women and 38 g/day for adult men. In contrast, the average daily zinc intake was 16.08 mg/day (SD = 9.79), exceeding the recommended dietary allowance (RDA) of 8 mg/day for adult women and 11 mg/day for adult men. BLL averaged 1.24 μg/dL (SD = 0.99), with a range from 0.18 to 11.8 μg/dL. Lastly, the average age of participants was 33.47 years (SD = 8.99), with a range from 18 to 48 years (Table 1).

Table 1.

Descriptive statistics (n = 1804).

Variable	Weighted Mean	Weighted SD	Max	Min
Log10 ENL	2.28	0.81	4.47	−0.93
BMI (kg/m2)	28.16	7.03	68.63	15.49
AGE (years)	33.47	8.99	48	18
Total Fiber (g/day)	16.36	8.09	68.75	1.1
Total Zinc (mg/day)	16.08	9.79	113.02	1.05
BLL (μg/dL)	1.24	0.99	11.8	0.18

3.2. Multiple linear regression

Table 2 shows a total of 5 regression models in the forward stepwise process. The first model examined zinc intake alone as a predictor of ENL. Zinc demonstrated a statistically significant positive association with ENL (p = 0.025). In Model 2, both zinc and blood lead levels (BLL) were included. Lead exhibited a significant negative association with ENL (p = 0.014).

Table 2.

Forward stepwise regression model selection.

Model	Variable	Coefficient	SE	t-value	p-value	AIC
1	Intercept	2.201	0.039	56.37	< 0.001	4340.13
	Zinc	0.005	0.002	2.361	0.025
2	Intercept	2.267	0.041	55.075	< 0.001	4334.26
	Zinc	0.005	0.002	2.403	0.023
	Lead	−0.055	0.021	−2.625	0.014
3	Intercept	2.854	0.099	28.828	< 0.001	4279.94
	Zinc	0.005	0.002	1.978	0.056
	Lead	−0.061	0.019	−3.203	0.003
	BMI	−0.020	0.002	−7.061	< 0.001
4	Intercept	2.740	0.099	27.822	< 0.001	4270.35
	Zinc	0.002	0.003	0.615	0.543
	Lead	−0.063	0.019	−3.399	0.002
	BMI	−0.020	0.003	−7.042	< 0.001
	Fiber	0.009	0.003	2.706	0.011
5	Intercept	2.675	0.095	28.014	< 0.001	4269.45
	Zinc	0.005	0.003	1.52	0.140
	Lead	−0.014	0.035	−0.393	0.697
	BMI	−0.019	0.003	−7.072	< 0.001
	Fiber	0.009	0.003	2.709	0.012
	Lead: Zinc	−0.003	0.002	−1.762	0.089

Model 3 included BMI as an additional predictor, revealing a strong negative association with ENL (p < 0.001). In Model 4, the introduction of fiber intake led to zinc losing its statistical significance (p = 0.543), while fiber remained positively associated with ENL (p = 0.011). The inclusion of fiber also reduced the zinc coefficient from 0.004 to 0.002, a change of approximately 50 %.

Model 5 included all covariates plus the interaction between zinc intake and BLL. The interaction term showed marginal statistical significance (p = 0.089), and BLL was no longer significant (p = 0.697). This outcome reflects the moderate collinearity between BLL and the interaction term. Importantly, due to model hierarchy rules, retaining the interaction term requires including the main effects for both zinc and BLL, even if their coefficients are set to zero in the statistical output. As a result, the zinc coefficient in Model 5 does not represent a true estimate of zinc’s independent effect and should not be interpreted in isolation. The model could not be simplified further without violating these statistical principles.

The Akaike Information Criterion (AIC) was used to compare relative model performance while accounting for the study’s complex survey design. AIC values decreased from 4340.13 in Model 1–4269.45 in Model 5, indicating that Model 5 provided a better balance between model fit and complexity. It is important to note that AIC does not measure explained variance or absolute model fit; rather, it evaluates which model is more efficient at approximating the data given the number of parameters used (Xue and Zhou, 2024).

4. Discussion

4.1. Enterolactone as a biomarker

ENL levels vary significantly among individuals due to several factors. Important contributors include the use of antibiotics, which can disrupt gut bacteria, and lifestyle variables such as diet (especially intake of lignan-rich foods), BMI, smoking, sex, and age. These factors influence how effectively lignans are converted and absorbed, leading to the wide range of ENL levels observed across different people (Hålldin et al., 2019). The regression model found a positive correlation between fiber and ENL. Fiber intake has shown to increase diversity in the gut microbiome (Fu et al., 2022; Cronin et al., 2021). Inversely, increased BMI, related to obesity, has been associated with decreased diversity in the gut microbiota (Liu et al., 2021b; DiBaise et al., 2008). This consistency suggests that ENL is a reliable biomarker of gut microbiome function and diversity.

4.2. zincs’ role in protecting diversity

The interaction term in Model 5 was marginally significant (p = 0.089), indicating that the effect of lead on ENL levels may vary depending on zinc intake. Specifically, higher zinc intake appears to attenuate lead’s negative impact on ENL, while lower zinc intake may intensify it. Although the p-value was borderline, the reduction in AIC for Model 5 suggests that including the interaction improved the model’s explanatory power. The negative coefficient for the interaction supports the idea that zinc offers a protective effect, moderating the adverse influence of lead on microbiome diversity.

These findings align with the understanding that, although zinc is not directly responsible for maintaining microbial diversity, stability, or gut barrier integrity, it acts as an assistant, serving as a cofactor for vital processes such as microbial enzyme activity, DNA repair, oxidative stress defense, and gut barrier strength (Pajarillo et al., 2021; Velasco et al., 2018; Porcheron et al., 2013; Marreiro et al., 2017). It supports thousands of proteins, including enzymes and transcription factors, and plays key roles in immune function, gut barrier integrity, and microbial balance (Wan and Zhang, 2022). Zinc-binding proteins comprise approximately 5–6 % of total proteins in prokaryotes (Andreini et al., 2008) and plays a vital role in supporting the maintenance of bacterial populations that enhance gut resilience (Wan and Zhang, 2022; Reed et al., 2015; Gomes et al., 2021; Koren and Tako, 2020). These bacteria contribute to gut health by strengthening the intestinal barrier, (Chai et al., 2024; Cheng et al., 2023; Matar et al., 2024a) promoting antioxidant activity, (Skalny et al., 2021) and biosorption of toxicants (Xing et al., 2017; Matar et al., 2024b). While these functions do not directly increase bacterial diversity, they create conditions that support the growth of beneficial microbes, enabling them to thrive and outcompete pathogenic species for space and resources within the gut (Molnar-Nagy et al., 2022). These vital processes are important in assisting in resilience against stressors such as toxins and pollutants which can disrupt microbial communities (Rocca et al., 2019).

The preserved gut barrier likely prevented lead from entering the bloodstream, as evidenced by the lower BLLs observed in this study. Preventing lead absorption is vital, as the gut serves as a major entry point for lead into the body (Toxicological Profile for Lead, 2002). Once lead enters the bloodstream, it can wreak havoc on vital organs such as the kidneys, and brain, triggering a cascade of medical conditions, including cognitive impairment, cardiovascular disease, and chronic kidney disease (Harari et al., 2018; Navas-Acien et al., 2007; Vlasak et al., 2019). By maintaining gut resilience and preventing lead from crossing the intestinal barrier, zinc may play a minor role in preventing an avalanche of health consequences.

4.3. Confounding factor

Fiber was found to be a confounding factor due to the 50 % change in zinc’s coefficients when it was added to the model. This means that zinc may not have a meaningful impact on ENL levels when fiber is accounted for. The substantial change in the estimate suggests that zinc’s apparent influence on ENL was likely an overestimation prior to including fiber in the model.

Controlling for fiber ensures that the observed interaction accurately reflects zinc and lead’s combined effect, independent of confounding influences. When fiber is included in the model, it may account for much of the variability in diversity that could otherwise be attributed to zinc. As a result, zinc’s independent contribution to diversity becomes “masked” or overshadowed by fiber’s more dominate effect. However, zinc’s moderating role on lead’s impact is marginally significant because its effect is conditional as it emerges specifically in the presence of lead. This suggests that zinc’s protective influence is not about directly increasing diversity but rather about mitigating the negative effects of lead. Fiber, which explains direct contributions to diversity, does not interfere with zinc’s ability to modify lead’s toxicity.

4.4. Limitations

Measuring microbial diversity is inherently challenging, and this study is limited by its reliance on ENL as a proxy measure, rather than using direct microbiome profiling methods such as 16S rRNA sequencing. This restriction narrows the scope of the findings, as ENL does not fully capture the complexity of microbial diversity. Also, despite controlling for known confounders, other unmeasured variables such as gender, socioeconomic status, ethnicity, overall dietary patterns, physical activity, medication use (e.g., antibiotics), and stress levels may still influence ENL levels and microbial diversity (Parizadeh and Arrieta, 2023; Nobre and Alpuim Costa, 2022). The study design introduces several limitations as well, particularly due to its cross-sectional nature and reliance on self-reported dietary intake data. The cross-sectional design limits causal inferences, while the reliance on dietary data introduces potential recall bias. Additionally, the lack of longitudinal data restricts the ability to explore the dynamic, long-term effects of zinc and lead on gut microbial diversity and stability. These limitations suggest the need for future research incorporating direct microbiome analysis, longitudinal data, and the inclusion of other variables that may influence microbiota diversity.

Many factors, in addition to age, BMI, and fiber intake can impact ENL levels that were not included in this study. Hormonal influences are one example; a positive correlation has been found between serum ENL and sex hormone-binding globulin (SHBG) in postmenopausal women (Zeleniuch-Jacquotte et al., 2004). Medications, like oral antimicrobials, have also been shown to decrease ENL levels (OUP Academic) (Kilkkinen et al., 2002). Lifestyle factors such as smoking and alcohol intake are associated with lower ENL concentrations (Horner et al., 2002). Additionally, inflammation may lower ENL levels, with studies indicating a connection between increased cytokines and reduced ENL (Eichholzer et al., 2014).

Several factors have also been shown to impact blood lead levels as well. Dietary factors, including other essential micronutrients such as calcium and iron intake, have been shown to impact lead absorption, with higher calcium (Bruening et al., 1999) and iron (Słota et al., 2021). Lifestyle factors, such as smoking and alcohol use, have been linked to increased lead absorption (Schuhmacher et al., 1993). Environmental exposures, including residential proximity to lead sources, can significantly elevate BLL (Wani et al., 2015). Exploring these factors could help clarify BLL variability.

The analysis employed linear regression models, which assume a linear relationship between exposure and outcome. However, essential nutrients typically have a hormetic curve, providing benefit at adequate levels and potential harm at deficient and excesses levels (Hayes, 2010). Dose-response relationships for zinc would require a broader distribution of exposure values to adequately assess whether the system experiences stress or protective effects at higher or lower thresholds. Zinc deficiency (Briefel et al., 2000) and elevated lead exposure (Lanphear et al., 2018; Rossi, 2008) are relatively uncommon in the U.S. population, resulting in a narrow range of values. This limited variability reduces the statistical power to detect non-linear or threshold effects and precludes a robust evaluation of potential dose-response patterns across extreme exposures (Sullivan and Feinn, 2012). Further research with a wider exposure range data would be necessary to explore this potential dose-response dynamic.

5. Conclusion

This study provides suggestive evidence that zinc intake may mitigate the loss of gut microbiome diversity associated with lead exposure, as indicated by the marginally significant interaction between zinc and BLL in predicting ENL levels (p = 0.089). Based on this p-value, there is approximately a 91 % probability that this interaction is not due to chance. This interaction highlights the complex interplay between dietary nutrients and environmental toxins, underscoring the need to consider such relationships in public health interventions.

In this study, Americans exhibited relatively low blood lead levels (mean BLL = 1.24 μg/dL), yet a significant relationship between lead exposure and microbial diversity was still observed. In populations with higher lead exposure, such as those in industrial areas or regions with limited environmental protections, lead’s potential to disrupt gut health may leave individuals even more vulnerable to its systemic health consequences (Ghosh et al., 2023; Kar-Purkayastha et al., 2012; Aliyu Haruna and Musa, 2021). Zinc supplementation and increased consumption of zinc-rich foods could serve as one facet of a comprehensive public health strategy aimed at enhancing gut microbial resilience and protecting against lead’s systemic toxic effects (Ugwuja et al., 2020). Zinc works in concert with other essential nutrients, functioning as a supportive element within a complex biological system that maintains gut integrity and host health (Scarpellini et al., 2022).

Future research could begin by examining additional subpopulations within the existing dataset to evaluate whether ENL is a reliable biomarker across diverse groups. Factors such as socioeconomic status, area of residence (urban or rural), ethnicity, gender, and dietary patterns could provide valuable insights into how ENL levels different populations (Parizadeh and Arrieta, 2023; Nobre and Alpuim Costa, 2022). Identifying these variables would help clarify the broader applicability of ENL as a biomarker. Building on this, future research should focus on populations with higher lead exposure and lower zinc intake, as such conditions may better capture dose-response relationships.

Given the observational nature of NHANES data, causal relationships between zinc and lead cannot be established with certainty. Longitudinal studies or experimental research are necessary to determine whether zinc has a definitive protective effect against lead toxicity. Establishing this causal relationship would provide critical insight into the potential of zinc as a modifiable factor for reducing lead burden in exposed populations, thereby highlighting the need for further investigation beyond cross-sectional analysis.

Data availability

Data will be made available on request.