Re-thinking the link between exposure to mercury and blood pressure

Abstract

Hypertension or high blood pressure (BP) is a prevalent and manageable chronic condition which is a significant contributor to the total global disease burden. Environmental chemicals, including mercury (Hg), may contribute to hypertension onset and development. Hg is a global health concern, listed by the World Health Organization (WHO) as a top ten chemical of public health concern. Most people are exposed to some level of Hg, with vulnerable groups, including Indigenous peoples and small-scale gold miners, at a higher risk for exposure. We published a systematic review and meta-analysis in 2018 showing a dose–response relationship between Hg exposure and hypertension. This critical review summarizes the biological effects of Hg (both organic and inorganic form) on the underlying mechanisms that may facilitate the onset and development of hypertension and related health outcomes and updates the association between Hg exposure (total Hg concentrations in hair) and BP outcomes. We also evaluated the weight of evidence using the Bradford Hill criteria. There is a strong dose–response relationship between Hg (both organic and inorganic) exposure and BP in animal studies and convincing evidence that Hg contributes to hypertension by causing structural and functional changes, vascular reactivity, vasoconstriction, atherosclerosis, dyslipidemia, and thrombosis. The underlying mechanisms are vast and include impairments in antioxidant defense mechanisms, increased ROS production, endothelial dysfunction, and alteration of the renin–angiotensin system. We found additional 16 recent epidemiological studies that have reported the relationship between Hg exposure and hypertension in the last 5 years. Strong evidence from epidemiological studies shows a positive association between Hg exposure and the risk of hypertension and elevated BP. The association is mixed at lower exposure levels but suggests that Hg can affect BP even at low doses when co-exposed with other metals. Further research is needed to develop robust conversion factors among different biomarkers and standardized measures of Hg exposure. Regulatory agencies should consider adopting a 2 µg/g hair Hg level as a cut-off for public health regulation, especially for adults older than child-bearing age.

Introduction

Hypertension, or high blood pressure (BP), is one of the most common and manageable chronic conditions, and it contributes substantially to the total burden of disease worldwide. According to the World Health Organization (WHO), an estimated 1.28 billion adults aged 30–79 years have hypertension. Hypertension has long been recognized as a leading risk factor for cardiovascular diseases (CVDs), which affect millions globally and leads to many morbidities and fatalities (Forouzanfar et al. 2017). The WHO estimates that 17.9 million people died from CVDs in 2019, representing 32% of all global deaths (Abbafati et al. 2020). Among other major modifiable risk factors for CVD, including cigarette smoking, diabetes mellitus, and lipid abnormalities, high BP is associated with the strongest evidence for causation and has a high prevalence of exposure (Fuchs and Whelton 2020).

The prevalence of hypertension in adults aged 30–79 years was 32% in women and 34% in men (Zhou et al. 2021). The prevalence of hypertension varies across countries and regions, with higher rates observed in low- and middle-income countries. Hypertension prevalence is highest throughout central and eastern Europe, central Asia, Oceania, southern Africa, and some countries in Latin America and the Caribbean (Zhou et al. 2021). Many of these regions overlap with regions where fisheries and aquaculture have increased over the last two-to-three decades (Food and Agriculture Organization 2020). Many populations in regions with a high prevalence of hypertension also have fish and seafood as a staple food (Food and Agriculture Organization 2020).The etiology of hypertension is complex and multifactorial (Carretero and Oparil 2000). Recognized modifiable risk factors include unhealthy diets (high in salt, saturated fat, and trans fats, low in fruits and vegetables), physical inactivity, consumption of tobacco and alcohol, and being overweight or obese. Non-modifiable risk factors include a family history of hypertension, 65 years of age and older, and co-existing diseases such as diabetes or kidney disease (Carretero and Oparil 2000). Besides the traditional risk factors for hypertension, exposure to environmental chemicals may also play an important role (Abhyankar et al. 2012; Eum et al. 2008; Houston 2011; Navas-Acien et al. 2007). It is likely that their contribution has been underestimated (Fuller et al. 2022).

Mercury (Hg) is a chemical pollutant that is of concern to human health worldwide. The WHO has listed Hg as one of its top ten chemicals of public health concern (World Health Organization 2020). Human biomonitoring studies have established that most people worldwide are exposed to some amount of Hg and that there are notable vulnerable groups (e.g., Indigenous peoples and artisanal and small-scale gold miners (ASGM)) (Basu et al. 2018).

Hg released into the environment can contaminate water bodies and accumulate in aquatic ecosystems. This poses a risk to different fish species and other wildlife, affecting biodiversity and ecosystem health. This, in turn, poses a significant population health risk, particularly in coastal regions where fish and seafood are staple food. Workers in industries such as ASGM and chlor-alkali production are at a higher risk of occupational Hg exposure, including Hg poisoning. Chronic Hg exposure could potentially lead to an increased risk of hypertension, diabetes, CVDs, and certain types of cancer, posing a significant global burden on human health (Basu et al. 2023; Ha et al. 2017).

Although the health impacts of Hg have been primarily focused on its neurotoxicity, especially the effects of prenatal exposure on the neurodevelopment of children and youth, there is increasing evidence that exposure to Hg is a risk factor for hypertension and CVDs (Ha et al. 2017; Basu et al. 2023). The dose–response relationship between Hg exposure and CVDs and the possible underlying mechanisms were reported over 10 years ago (Houston 2011; Roman et al. 2011). We have published two systematic reviews and meta-analyses showing the significant relationship between hypertension, CVDs, and Hg exposure in the last five years (Hu et al. 2018, 2021), and more evidence has emerged since.

This study is a critical review carried out through database searches, synthesis of the evidence, and critical interpretation of the findings. This narrative review approach is useful for topics that require a meaningful synthesis of research evidence that may be complex or broad and that require detailed, nuanced description and interpretation (Sukhera 2022). The pathophysiology of hypertension is complex and involves processes that include inflammation, oxidative stress, endothelial dysfunction, altered activation of hormones that regulate salt, and water excretion by the kidney (such as those in the renin–angiotensin–aldosterone system), excessive activation of the sympathetic nervous system, or plaque buildup in the arteries (Harrison et al. 2021). Also, Hg is distributed in multiple organs and involves multiple pathways related to different health outcomes. Therefore, this study aims to review the dose–response of Hg at the tissue and organ level and the underlying mechanisms that can potentially contribute to hypertension onset reported in animal studies, update the latest epidemiological evidence between Hg exposure and hypertension, and critically evaluate the weight of evidence on Hg exposure as a risk factor for hypertension. Therefore, some key evidence of Hg’s impact on other health outcomes, such as CVD, with similar or related pathophysiology to hypertension from both animal and human studies was also summarized (less extensively) to add the weight of evidence that the association observed in epidemiology between Hg and hypertension is likely to be a cause–effect relationship.

Mercury speciation, toxicology and biomarkers

Hg is ubiquitously found in natural elements like soils, coal, and minerals. Its release into the atmosphere occurs naturally through geological actions, like volcanic eruptions, as well as through human activities, such as coal burning, waste incineration, and metal processing (ATSDR 2022; Driscoll et al. 2013). There are three forms of Hg: elemental (Hg⁰), inorganic (Hg¹⁺, mercurous Hg, and Hg²⁺ mercuric Hg), and organic. Organic Hg occurs when Hg is combined with carbon. Common types of organic Hg include methylmercury (MeHg/CH₃Hg⁺), ethylmercury (EHgC₂H₅Hg¹⁺), and dimethylmercury (C₂H₆Hg) (Clarkson and Magos 2006).

For the general populations, exposure to organic Hg primarily occurs via the consumption of larger fish species, where MeHg is biomagnified and predominates (Lin et al. 2021; Driscoll et al. 2013). MeHg is absorbed by marine species and biomagnified (Lin et al. 2021). MeHg exposure can commence during fetal development through cord blood and can persist via breast milk transmission to infants (Sundseth et al. 2017). Approximately 90–95% of MeHg is absorbed by the gastrointestinal tract. Major targets of MeHg-induced toxicity include the brain, reproductive system, and cardiovascular system (Government of Canada 2008; National Research Council (US) 2000). MeHg commonly forms bonds with sulfhydryl (thiol) groups and selenohydryl (selenol) groups (Ajsuvakova et al. 2020). Demethylating MeHg to Hg²⁺ can occur in the gastrointestinal tract (Nakamura et al. 1977; Rowland 1988) and the liver (Suzuki et al. 1984; Uchikawa et al. 2016). Hg²⁺ is then excreted in the feces and urine (Ballatori and Clarkson 1982; Yasutake et al. 1989). MeHg demethylation by gut flora had been indicated as the major elimination route. Typically, 90% or more of the Hg derived from a MeHg dose is excreted in feces as inorganic Hg (Ishihara 2000). Pope and Rand (2021) used data collected from 37 adult volunteers who consumed three fish meals with known MeHg concentrations and developed a PBPK model to estimate the metabolism and elimination rate of MeHg. They found that by day 50, 46 and 51% of the dose in the adult man and woman, respectively, was removed via biotransformation in the gut lumen, representing 73% of the total eliminated MeHg in each case. Overall, the whole-body half-life of MeHg is approximately 70–80 days (Government of Canada 2008; National Research Council (US) 2000). More recent human studies showed that the half-life of MeHg was 46.9, 38.9, and 31.5 days and steady-state blood MeHg of 2.6, 2.6, and 2.3 µg/l in men, women, and children, respectively, and gut microbiota can affect the elimination of MeHg and hence accounts for the individual variability (Pope and Rand 2021).

Additionally, the general population encounters inorganic Hg exposure to a lesser extent, often originating from sources like dental amalgams (ATSDR 2022; Park and Zheng 2012). Exposure to elemental Hg may occur within the general population through various means, including inhalation of vapors in ambient air, ingestion, and dental or medical procedures (Park and Zheng 2012). Workers in industries, such as ASGM, dental amalgam preparation, fluorescent light bulb manufacturing, chlor-alkali production, and thermometers and barometers manufacturing, may come into contact with Hg⁰ or other Hg-containing substances (Aubrac et al. 2022; Driscoll et al. 2013). Lipophilic Hg⁰ is rapidly distributed throughout the body to target the kidneys and brain. Excretion commonly occurs through exhalation, sweat, or saliva. Alternatively, Hg⁰ can be oxidized in erythrocytes to form Hg²⁺, which is then eliminated in the feces and urine (Ballatori and Clarkson 1982; Yasutake et al. 1989). Overall, the whole-body half-life of Hg⁰ is approximately 58 days (Government of Canada 2008; National Research Council (US) 2000).

Finally, inorganic Hg in both Hg¹⁺ and Hg²⁺ forms are formed through the oxidation of Hg⁰ or demethylation of MeHg in the tissues of animals or plants. The absorption rate of inorganic Hg is low at around 5%, and the major target organ of Hg²⁺ is the kidneys; at high doses, Hg²⁺ can result in kidney failure and gastrointestinal damage (Government of Canada 2008; National Research Council (US) 2000). Hg²⁺ is eliminated in the feces and urine (Ballatori and Clarkson 1982; Yasutake et al. 1989). The whole-body half-life of Hg²⁺ is approximately 1–2 months (Government of Canada 2008; National Research Council (US) 2000).

Hg exposure can be assessed in two major ways. The first approach involves estimating external dose by measuring Hg concentrations in various mediums, such as food, air, or water, and then multiplying the concentrations by the corresponding frequency of exposure. The second method focuses on estimating the body burden of Hg by analyzing biological samples such as hair, urine, blood, nails, cord tissues, or placenta (Basu et al. 2023; Ha et al. 2017). Pharmacokinetics models have been developed to predict the body burden of Hg in various tissues from external doses of Hg exposure for both inhaled Hg vapor (Jonsson et al. 1999; Leggett et al. 2001) and MeHg (Carrier et al. 2001; Wang and Wang 2015). These models provide quantifiable parameters for exposure, absorption, distribution, metabolism, and excretion, thereby allowing for a more robust comparison between studies reporting different Hg exposure sources and doses.

Measurements of total Hg in blood and urine can be used as biomarkers of total exposure to all forms of Hg without additional information on the magnitudes of exposure to any specific forms of Hg. MeHg concentration in whole blood or total Hg concentration in red blood cells or hair are preferred biomarkers for MeHg exposure. Inorganic Hg in blood (or plasma/serum) and inorganic Hg or total Hg in urine are considered biomarkers of exposure to inorganic forms of Hg (ATSDR 2022).

Conversion ratios or equations are published to quantify the numeric relationships among various Hg biomarkers. The WHO recommends a hair-to-blood ratio of 250 for the conversion of hair Hg levels (in µg/g) to those in whole blood (in µg/L) (Joint FAO/WHO Expert Committee on Food Additives 2004). Although this ratio is widely adopted, there is literature suggesting that there is vast inter-individual variation (Liberda et al. 2014; Singh et al. 2023; Yaginuma-Sakurai et al. 2012). Hg concentrations in toenails (in µg/g) and in urine (µg/L) could be converted to hair Hg (in µg/g) using a regression model (Ohno et al. 2007).

Animal studies on mercury and hypertension

To establish the cause–effect relationship between Hg exposure and hypertension, we searched the literature for animal dosing studies. We found 17 major studies published between 1983 and 2023 investigating the relationship between Hg exposure and hypertension using rodent models (Table 1). Different strains of rats were used, and the Hg dose was administered by drinking water, oral gavage, or intramuscular injection. The exposure duration ranged from 13 to 320 days. More studies (13) used inorganic Hg than MeHg (7). Most of them (14/17) found a significant relationship between Hg treatment and increased BP.

Table 1.

Characteristics of published rodent studies which assess blood pressure included in the review

Reference	Species	Exposure route	Co-exposure or interaction with other metals	Dose	Form	Outcome	Total exposure = (dose × length × weight (assume 300 g)) For drinking water assume 30 ml/day
Carmignani et al. (1983)	Sprague–Dawley rats	Drinking water		50 μg/ml for 320 days	Inorganic	No significant change in BP	Total exposure = 50 µg/ml × 30 ml/day × 320 days = 420,000 µg
Carmignani et al. (1992)	Wistar rats	Drinking water		200 μg/ml for 180 days	Inorganic	Increase in arterial BP	Total exposure = 200 µg/ml × 30 ml/day × 180 days = 1,080,000 µg
Machado et al. (2007)	Wistar rats	Intravenous injection		680 ng/kg/BW/day	Inorganic	SBP/DBP increase	Total exposure = 680 ng × 1 day × 0.3 kg = 204 ng = 0.204 µg
Simões et al. (2016)	Wistar rats	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	Diastolic arterial BP increase	Initial dose = 4.6 µg/kg × 0.3 kg = 1.38 µg Subsequent doses = 0.07 µg/kg/day × 0.3 kg × 30 days = 0.63 µg Total exposure 30 days = 1.38 µg + 0.63 µg = 2.01 µg
Rizzetti et al. (2017b)	Wistar rats	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 or 60 days	Inorganic	SBP increase after 60 days only	Total exposure 30 days = 2.01 µg Total exposure 60 days = 2.64 µg
Bello et al. (2023)	Wistar rats	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	SBP/DBP increase	Total exposure 30 days = 2.01 µg
Schereider et al. (2021)	Wistar rats (female)	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	No significant change in SBP	Total exposure 30 days = 2.01 µg
Vassallo et al. (2019)	Spontaneously hypertensive rats [Wistar rats (control)]	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	SBP increase in spontaneously hypertensive rats only	Total exposure 30 days = 2.01 µg
Fardin et al. (2020)	Spontaneously hypertensive rats	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	SBP increase	Total exposure 30 days = 2.01 µg
Simões et al. (2020)	Spontaneously hypertensive rats [Wistar rats (control)]	Intramuscular injections		Initial dose of 4.6 μg/kg and subsequent doses of 0.07 μg/kg/BW/day for 30 days	Inorganic	SBP increase in spontaneously hypertensive rats only	Total exposure 30 days = 2.01 µg
Tamashiro et al. (1986)	Spontaneously hypertensive rats [Wistar rats (control)]	Injection		5 mg/kg/BW/day for 10 days	Methyl	NA	Total exposure = 5 mg/kg × 10 days × 0.3 kg = 15 mg = 15,000 µg
Wakita (1987)	Wistar rats	Subcutaneous injection Oral gavage		5 mg/kg 11 times for 13 days (acute) 0.5 mg/kg/BW/day (chronic) for 23–28 days	Methyl	SBP increase following cessation	Total exposure acute = 5 mg/kg × 11 times × 0.3 kg = 16.5 mg = 16,500 µg Total exposure chronic = 0.5 mg/kg × 23–28 days × 0.3 kg = 3.45–4.2 mg = 3450–4200 µg
Grotto et al. (2009)	Wistar rats	Oral gavage		100 μg/kg/BW/day for 100 days	Methyl	SBP increase	Total exposure = 100 µg/kg × 100 days × 0.3 kg = 3000 µg
Grotto et al. (2011)	Wistar rats	Oral		Contaminated fish for 84 days [uncontaminated fish (control)]	Methyl	SBP increase	N/A
Wildemann et al. (2015a)	Wistar rats	Drinking water		7, 14, 29, 57, 357, 2000, 4000, 8000 μg/kg/BW/day for 28 days 4, 7, 14, 29, 57, 357, 1607 μg/kg/BW/day for 28 days	Inorganic Methyl	MeHg caused SBP/DBP increase	4 µg/kg × 28 days × 0.3 kg = 33.6 µg 7 µg/kg × 28 days × 0.3 kg = 588 µg 14 µg/kg × 28 days × 0.3 kg = 1176 µg 29 µg/kg × 28 days × 0.3 kg = 2436 µg 57 µg/kg × 28 days × 0.3 kg = 4788 µg 357 µg/kg × 28 days × 0.3 kg = 2998.8 µg 1607 µg/kg × 28 days × 0.3 kg = 13,498.8 µg 2000 µg/kg × 28 days × 0.3 kg = 16,800 µg 4000 µg/kg × 28 days × 0.3 kg = 33,600 µg 8000 µg/kg × 28 days × 0.3 kg = 67,200 µg
Wildemann et al. (2015b)	Wistar rats	Drinking water	Yes	57 or 357 μg/kg/BW/day for 28 days 29 or 357 μg/kg/BW/day for 28 days	Inorganic Methyl	MeHg alone caused SBP increase	29 µg/kg × 28 days × 0.3 kg = 2436 µg 57 µg/kg × 28 days × 0.3 kg = 4788 µg 357 µg/kg × 28 days × 0.3 kg = 2998.8 µg
Wildemann et al. (2016)	Wistar rats	Drinking water	Yes	357 or 4000 μg/kg/BW/day for 28 days 7 or 357 μg/kg/BW/day for 28 days	Inorganic Methyl	Inorganic Hg increased DBP MeHg alone caused SBP/DBP increase	7 µg/kg × 28 days × 0.3 kg = 588 µg 29 µg/kg × 28 days × 0.3 kg = 2436 µg 357 µg/kg × 28 days × 0.3 kg = 2998.8 µg 4000 µg/kg × 28 days × 0.3 kg = 33,600 µg

To support the observed cause–effect relationships, we reviewed other studies reporting Hg distribution to the cardiovascular system, its effects and the underlying mechanisms. We group the findings under the following sections: Hg distribution in organs related to hypertension development, structural changes, functional changes, and factors that affect hypertension.

Hg distribution in organs related to hypertension development

Many organs are involved in the development of hypertension. Three, which we will cover here, are the 1) heart, 2) kidneys, and 3) liver. Evaluating Hg distribution in each is important to understanding Hg-induced hypertension. There are scarce data on the distribution of Hg in human tissues. Work conducted on a victim of Minamata disease following the Minamata disaster in Japan has revealed high levels of inorganic Hg in the liver and kidneys. Lower levels of Hg were observed in the brain. However, approximately 50% of it was MeHg (Marumoto et al. 2020). Studies from non-polluted regions with larger sample sizes also found inorganic Hg in the brain and kidney and additional Hg deposits in the thyroid and pancreas (Pamphlett 2021).

Here, we will focus on the findings of Hg distribution in rodent models. In rats following Hg⁰ vapor exposure, Hg has been measured in the heart, kidneys, and liver (Iranmanesh et al. 2013). Following MeHg exposure in rats, Hg was found in the heart, liver, and kidneys as either inorganic Hg or MeHg. Following thimerosal exposure in rats, Hg was found in the heart, liver, and kidneys as MeHg, inorganic Hg, or ethylmercury (Rodrigues et al. 2010). Thimerosal has also been found in similar organs following exposure in mice (Carneiro et al. 2014). Interestingly, spontaneously hypertensive rats exposed to 5 mg/kg/BW/day over 10 days had higher Hg levels in their kidneys and liver compared to control Wistar rats (Tamashiro et al. 1986). Overall, elemental, inorganic, and organic forms of Hg can be found in the heart, liver, and kidneys directly or via metabolism to alternate Hg forms.

Two key factors determine Hg distribution in these critical areas: 1) age and 2) sex. First, age plays an important role in Hg accumulation. In a study, two groups of rats, young (10 weeks old) and old (36 months old), were sacrificed, and the number of trace elements in various organs was analyzed. Researchers found that old rats had significantly more Hg in their livers and blood than young rats (Rakic et al. 2020). Similar results were found in a Harbour Seal study where Hg in the liver and heart significantly increased with age, probably due to bioaccumulation (Marino et al. 2011). In humans, a similar trend has been observed, in which inorganic Hg increases with age. However, it was reported that the peak in inorganic Hg began to decrease in multiple organs in individuals older than 80 years old (Pamphlett 2021). Second, sex is a factor that can alter Hg toxicokinetics. Whole-body clearance of Hg is potentially faster in females than in males (Thomas et al. 1987). In C57BL/6N mice 5 min after treatment with MeHgCl, the level of Hg in the kidneys of females was significantly lower. This led to significantly less Hg in the urine of females at 24 h and significantly more Hg in the liver and blood of females (Hirayama and Yasutake 1986). Moreover, male Wistar rats (Bello et al. 2023; Simões et al. 2016, 2020), but not females (Schereider et al. 2021), show changes in BP following inorganic Hg exposure. This sexual dimorphism is potentially caused by lower renal organic anion transport 1 (Oat1) and Oat3 levels in females, as Oat1 and Oat3 are transporters involved in the renal uptake of Hg (Hazelhoff et al. 2012).

Structural changes

Changes in heart and kidney weight have been extensively studied following Hg exposure. Two early studies examined the structural effect of MeHg and HgCl₂ in early postnatal rats. Here, MeHg and HgCl₂ were administered in varying doses (1 and 2.5 mg/kg/BW/day) starting from P1 until wean (Bartolome et al. 1984; Slotkin et al. 1985). In HgCl₂-treated rats, researchers observed a slight increase in heart and kidney weight at early postnatal, which gradually decreased; this was followed by a slight decrease in heart weight compared to the controls (Bartolome et al. 1984). In MeHg-treated rats, similar effects on heart and kidney weights were observed (Slotkin et al. 1985). In a long-term generational study conducted in rats, the parents’ generation was given 0, 0.1, 0.5, and 2.5 ppm MeHgCl and bred to create F1 progeny. F1 was later bred to generate F2. Significant increases in the weight of the hearts and kidneys were observed in the Fl and F2 offspring following 2.5 ppm treatment (Verschuuren et al. 1976a, b).

Changes in collagen and elastin have also been observed in HgCl₂-exposed rodents across multiple studies. Rats given 1 mg/kg/BW/day HgCl₂ for 12 weeks had an increase in total collagen and elastin in the heart muscle (Olczyk et al. 1994). This effect was also observed after a shorter exposure period. Here, rats exposed to 1.23 mg/kg/BW/day HgCl₂ for 28 days also had increased collagen deposits in the aorta (Arbi et al. 2021).

More severe heart malformations have been observed following in utero exposure to Hg. MeHg, Hg⁰, and Hg²⁺ all readily distribute into the heart and kidneys of prenatally exposed offspring resulting in congenial deformities (Feng et al. 2004; Rutkiewicz and Basu 2013; Yoshida et al. 1986). Embryonic mice treated with 7.5 mg/kg/BW/day MeHg from embryonic day 5.5 (E5.5) to E10.5 had significantly thinner ventricular walls. Moreover, total heart volume was significantly lower at E14.5 following treatment with MeHg (Qiu et al. 2022). Hamsters injected with a single dose (15 mg/kg/BW) of mercuric acetate were sacrificed at E12. The embryos showed ventral wall defects and pericardial cavity distention (Gale 1980). Outside of rodent models, a reduction in the thickness of the ventricular walls has been observed in multiple species, including chickens (Ronconi-Krüger et al. 2022) and fish (Weis and Weis 1977).

Functional changes

Apart from structural changes, alterations in force generated by contracting papillary muscles have also been observed. Papillary muscles are located in the ventriculus of the heart and contract during systole. Female rat papillary muscles treated with varying concentrations of HgCl₂ showed that lower concentrations (1 μM) could increase contracting papillary force, while higher concentrations of HgCl₂ (5–10 μM) would decrease contracting papillary force. Potentially by modulating Ca²⁺ release from the sarcoplasmic reticulum (Oliveira et al. 1994; Oliveira and Vassallo 1992; Vassallo et al. 1999). The sarcoplasmic reticulum is an organelle within smooth muscle cells that releases Ca²⁺ to trigger the contraction of the muscle fibers. Excessive Ca²⁺ release from vascular smooth muscle cells has been associated with vasoconstriction, which could potentially influence hypertension development.

Similar results were observed following exposure of isolated atria to MeHgOH⁻. At lower concentrations (0.5 and 2 ppm), MeHg increased the frequency of contractions, while at high concentrations (>2 ppm), MeHg decreased the frequency of contractions (Su et al. 1979). Overall, MeHg seems to induce a biphasic effect on muscle contraction (Ottolini et al. 2019).

Factors that influence the onset of hypertension

Multiple studies in rodents have linked Hg exposure to increased hypertension risk. For example, chronic MeHgCl exposure caused a significant increase in SBP 42 days after cessation (Wakita 1987). Grotto et al. found in a series of experiments that exposure to MeHgCl (100 μg/kg/BW/day) (Grotto et al. 2009) or MeHg-contaminated fish (Grotto et al. 2011) significantly increased SBP over time. Similar results were seen at even lower doses observed by Wildemann et al. (2015a, b). Here, rats were exposed to MeHg or HgCl₂ for 4 weeks. The researchers observed that 7 μg/kg/BW/day MeHg and above significantly increased SBP relative to control, while 14 μg/kg/BW/day MeHg and above significantly increased diastolic blood pressure (DBP) relative to control. This effect was not seen following treatment with HgCl₂ (Wildemann et al. 2015a).

While many studies have shown no effect of HgCl₂ on BP (Carmignani et al. 1983; Schereider et al. 2021; Wildemann et al. 2015a) other studies have found a relationship between HgCl₂ and hypertension (Bello et al. 2023; Machado et al. 2007). Male Wistar rats were given HgCl₂ for 180 days; this caused an increase in systemic arterial BP (Carmignani et al. 1992). Similar results were observed in an in vivo model of hypertension called the spontaneously hypertensive rat model. Here, rats were exposed to an initial dose of 4.6 μg/kg/BW HgCl₂ and subsequent doses of 0.07 μg/kg/BW/day HgCl₂ for 4 weeks. The researchers found that HgCl₂ accelerated the development of hypertension marked by increased SBP while increasing reactive oxygen species (ROS) production in mesenteric resistance arteries (Fardin et al. 2020). However, a follow-up study showed that non-spontaneously hypertensive (normotensive) rats do not show the same phenomenon (Simões et al. 2020). However, they do show increased mean arterial pressure and heart rate after 30 days of HgCl₂ exposure (Simões et al. 2016), and increased SBP after prolonged exposure (60 days) (Rizzetti et al. 2017b). Sex differences also contribute to MeHg-induced BP impairments. MeHgCl 5 mg/kg/BW/day was administered over 10 days to spontaneously hypertensive rats. Male spontaneously hypertensive rats experienced more toxicity, indicated by higher rates of mortality, larger weight loss, and earlier neurologic signs (Tamashiro et al. 1986). Overall, some studies found that even low doses of inorganic Hg can cause changes in SBP/DBP in rodent models (Bello et al. 2023; Machado et al. 2007; Rizzetti et al. 2017b; Simões et al. 2020). In the studies summarized in the current review, MeHg can also affect BP, but this has been studied at high dosage experiments (Grotto et al. 2009; Wakita 1987). In the three studies that evaluated the effect of both MeHg and inorganic Hg on BP at various dosages, two major conclusions were established: 1) MeHg-exposed rats had increased BP at multiple doses and 2) inorganic Hg showed fewer cardiovascular effects than MeHg (Wildemann et al. 2015a, b, 2016).

Hypertension is influenced by a wide range of factors and CVDs, many of which are interconnected and contribute to the disease’s onset and progression. Other factors that influence the onset of hypertension are various, including 1) vasoconstriction, 2) atherosclerosis, 3) dyslipidemia, and 4) thrombosis. Vascular reactivity refers to the ability of blood vessels to change their diameter and ultimately alter blood flow. Vasoconstriction is a type of vascular reactivity where blood vessels are narrowed due to the contraction of smooth muscle cells. This can contribute to or be caused by hypertension onset (Giles et al. 2012). Multiple factors contribute to vascular reactivity. Angiotensin, 5-HT (serotonin), and ACh (acetylcholine) are all critical molecules, and their role will be highlighted below.

First, angiotensin is a hormone that can cause vasoconstriction by binding to receptors on smooth muscle cells of blood vessels. Second, 5-HT is a neurotransmitter and vasoconstrictor that also binds to receptors located on smooth muscle cells. Finally, ACh is a neurotransmitter that acts on endothelial cells to promote the release of nitric oxide (NO). In smooth muscle cells, NO increases the concentration of cyclic guanosine monophosphate (cGMP), which ultimately causes vasodilation. Multiple studies have found that HgCl₂ alters vascular reactivity in both the peripheral nervous system (PNS) and the central nervous system (CNS).

In the rat aorta, HgCl₂ was shown to cause oxidative stress via NAPDH oxidation, resulting in decreased NO production, leading to endothelial dysfunction, and enhanced vascular reactivity following phenylephrine exposure (a drug used for vasoconstriction) (Cordeiro et al. 2019). Similar HgCl₂-induced impairments have been observed in rat aortic rings (Lemos et al. 2012), and isolated rat tail vascular bed (Da Cunha et al. 2000).

In the CNS, MeHg exposure for 3 weeks in a rodent model promoted angiotensin-induced vasoconstriction in isolated basilar arterial. Similarly, ACh-induced relaxation in the basilar arterial rings was decreased in MeHg-treated mice (Islam et al. 2016). In another study, L-NAME (NO synthase inhibitor) was added to basilar arteries isolated from both control and HgCl₂-treated rats. L-NAME caused enhanced 5-HT-induced vasoconstriction. This indicates that the presence of NO normally counteracts the vasoconstrictive effects of 5-HT in these arteries. However, inhibition of NO production causes enhanced vasoconstriction in Hg-treated rats (Wiggers et al. 2016).

Atherosclerosis is the thickening or hardening of arteries where plaques, which consist of cholesterol and other substances form. This results in reduced blood flow which could potentially contribute to or be caused by hypertension development (Poznyak et al. 2022). At the beginning of atherosclerotic plaque formation, a critical step is the adhesion of monocytes to endothelial cells and the subsequent migration and transformation of monocytes into macrophages.

An in vitro model found that MeHg treatment significantly induced the adhesion of monocytes to human microvascular endothelial cells and increased pro-inflammatory cytokines and NF-κB activation (Fowler et al. 2021). In vivo studies have shown a similar relationship between MeHg and atherosclerosis. Here, atherosclerosis-prone apolipoprotein E (ApoE) knockout mice were compared to atherosclerosis-resistant C57BL/6 mice. Following surgery to induce atherosclerosis the mice were split into 1) control or 2) MeHg drinking water (20 ppm). They found that atherosclerosis lesions were more extensive in the aorta and carotid sites of MeHg-treated ApoE knockout mice and MeHg-treated C57BL/6 control mouse line (Silva et al. 2020).

Dyslipidemia is characterized by abnormal levels of lipids including cholesterol and triglycerides in the blood. Dyslipidemia impairs levels of low-density lipoprotein (LDL), commonly referred to as “bad” cholesterol, and high-density lipoprotein (HDL), commonly referred to as “good” cholesterol. High levels of cholesterols can contribute to atherosclerotic plaque formation and potentially hypertension development (Otsuka et al. 2016).

MeHg exposure (20 ppm) in ApoE knockout mice elevated total cholesterol (TC), LDL, and HDL levels (Roque et al. 2021). This is consistent with a study using both Swiss and C57BL/6 mice treated with high-dose MeHg (40 ppm) for 21 days, which showed that MeHg-treated mice had high total HDL and non-HDL (including LDL) plasma levels (Moreira et al. 2012). Overall, MeHg (at high doses) seems to lead to dyslipidemia in rodent models.

Lipid-lowering drugs may present as a potential rescue for MeHg-induced hypercholesterolemia. A study using Swiss mice treated with 40 ppm MeHg and Probucol (a lipid-lowering drug) prevented the development of hypercholesterolemia (Moreira et al. 2012).

Thrombosis occurs when a blood clot forms and obstructs blood flow. Procoagulant factors that trigger platelet activation may contribute to hypertension by increasing the risk of blood clots, narrowing blood vessels, and promoting inflammation. One procoagulant factor that has been linked to hypertension is thrombin. In human erythrocytes, exposure to 0.25–5 μM HgCl₂ for 1–48 h resulted in increased thrombin generation (Lim et al. 2010). Furthermore, in rats, 1.148 mg/kg/BW/day HgCl₂ for 28 days caused enhanced platelet activation (Arbi et al. 2017). In combination, these results suggest that Hg can promote procoagulant factors leading to platelet activation and potentially blood clot formation.

Mechanism of Hg-induced hypertension

It is well established that Hg binds to thiol and seleno groups of proteins with high affinity and disrupts the structure and activity of enzymes, transporters, and other proteins dependent on functional thiol and seleno groups. Therefore, adverse outcomes are often caused by multiple mechanisms, including intracellular calcium homeostasis, mitochondrial function, oxidative stress, and neurotransmitter release (Kang et al. 2021).

The most studied mechanism is the effects of Hg on oxidative stress and mitochondrial dysfunction. HgCl₂ is known to increase the formation of reactive oxygen species (ROS), such as hydrogen peroxide (H₂O₂) and superoxide anion (O₂⁻), through the NAPDH oxidase enzyme (Cordeiro et al. 2019; Rizzetti et al. 2017a). Another potential contributor to ROS production is via the electron transport chain, which has been well studied in neurons, but evidence for this outside of the CNS is lacking (Mori et al. 2007). In this review, we will focus on the studies related directly to CVDs only. MeHg can also trigger mitochondrial membrane potential disruption (Nishimura et al. 2019), decrease mitochondrial function in cardiomyocytes (Truong et al. 2015), and cause mitochondrial damage in the myocardium (Arbi et al. 2021). Disruptions in mitochondrial potential and function were seen in conjunction with increased ROS production (Nishimura et al. 2019; Truong et al. 2015).

Thiol depletion can also contribute to ROS formation through the inactivation of antioxidant defense molecules, including superoxide dismutase (SOD), glutathione (GSH), and catalase (CAT) [see review, (Ajsuvakova et al. 2020)]. Wiggers et al. (2008) showed that only EC-SOD was significantly changed in the rat aorta following HgCl₂ exposure (Wiggers et al. 2008). Other studies which have investigated cardiovascular dysfunction and HgCl₂/MeHg exposure have seen reductions in general SOD (Jindal et al. 2011; Zhang et al. 2013). Hg has an affinity for GSH, to form the Hg-GSH complex (Ballatori and Clarkson 1982). As a result, this complex reduces the total amount of GSH, which has been observed in multiple models of cardiovascular dysfunction across both inorganic and organic forms of Hg (Ghizoni et al. 2017; Jindal et al. 2011; Karaboduk et al. 2015; Zhang et al. 2013). Finally, multiple cardiac models have also shown decreased CAT levels following HgCl₂ exposure (Karaboduk et al. 2015; Zhang et al. 2013).

Multiple protective agents have been employed to rescue impaired antioxidant homeostasis following Hg exposure. First, in rat myocardial tissues, both MeHg and HgCl₂ caused decreased levels of GSH and SOD, which was rescued by co-treatment with melatonin, which is known to detoxify ROS (Jindal et al. 2011). Second, vitamin E has been proposed in multiple papers to alleviate Hg-induced cardiotoxicity (Jin et al. 2012; Karaboduk et al. 2015); via increased SOD, CAT, and GPx levels (Karaboduk et al. 2015). Finally, multiple forms of selenium have been shown to have beneficial effects on both MeHg and HgCl₂-induced toxicity dependent on dosage (Jin et al. 2012; Karaboduk et al. 2015). Selenite (a form of selenium) could similarly raise SOD, CAT, and GPx levels in rodents following HgCl₂ exposure (Karaboduk et al. 2015).

We propose that Hg toxicity is mediated through three major mechanisms: 1) endothelial cell dysfunction, 2) alteration of the renin–angiotensin system, and 3) inflammation. We summarize the potential effects and the underlying mechanisms of Hg on the cardiovascular system in Fig. 1 and discuss the details of each mechanism below.

Fig. 1.

The proposed mechanism by which mercury (Hg) effects the cardiovascular system. Hg is able to enhance an ACE and renin. These are released from the kidney and lungs respectively and ultimately result in increased angiotensin II which has been shown to trigger ROS formation in vascular smooth muscle cells. This inhibits NO and cGMP contributing to vasoconstriction. ROS production can trigger NF-kB leading to transcription of pro-inflammatory cytokines and atherosclerosis. In vascular endothelial cells, Hg forms complexes with GSH and selenium inhibiting ROS reduction. Furthermore, NAPDH oxidase generates ROS which have been shown to trigger increased intracellular Ca²⁺; this has been shown to lead to cytotoxicity and eicosanoid production. Abbreviations: AA, arachidonic acid; ACE, angiotensin I-converting enzyme; Ca²⁺, calcium; cGMP, cyclic guanosine monophosphate; COX, cyclooxygenase; GPx, glutathione peroxidase; GSH, glutathione; GSSG, oxidized glutathione; H₂O, water; H₂O₂, hydrogen peroxide; Hg, mercury; HgSe, mercury selenide; IL-6, interleukin 6; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; NO, nitric oxide; PLA2, phospholipase A2; PLD, phospholipase D; ROS, reactive oxygen species; TNF-a, tumor necrosis factor alpha. Created with BioRender.com

Endothelial dysfunction

Lipid peroxidation is a process that occurs when free radicals and ROS oxidize polyunsaturated fatty acid (PUFA) chains of cell membrane lipids. This process generates reactive lipid peroxides, including malondialdehyde (MDA), contributing to cellular damage. Increased levels of MDA have been readily observed following exposure to both MeHg and HgCl₂ (Jindal et al. 2011; Vassallo et al. 2019). Rats exposed to MeHg via contaminated fish had a significantly increased SBP over time combined with a significant increase in plasma MDA (Grotto et al. 2011).

As introduced previously, NO is a molecule produced by endothelial cells that acts as a vasodilator. In smooth muscle cells, NO increases the concentration of cGMP, which ultimately causes vasodilation. Multiple studies have found that NO is reduced following both HgCl₂ and MeHg. An in vitro experiment using aortic rings exposed to HgCl₂ showed increased ROS production and reduced downstream NO (Lemos et al. 2012). Furthermore, a dose-dependent decrease in NO production has also been observed in a culture of human endothelial cell, EA.hy926, following MeHg (0.1–10 µM) exposure (Van Dao et al. 2016). Other models of Hg-induced toxicity have observed similar effects (Cordeiro et al. 2019; Da Cunha et al. 2000; Lemos et al. 2012). However, it is important to note that contradictory studies have observed increased NO production following acute MeHg exposure to induce vasodilation (Omanwar et al. 2013).

In vascular endothelial cells, HgCl₂, MeHg, and thimerosal were shown to cause an intracellular Ca²⁺ influx caused by ROS formation and diminished thiols. This Ca²⁺ influx activates PLD in a dose-dependent manner. Overall, this leads to endothelial cell dysfunction and eicosanoid production (Hagele et al. 2007; Peltz et al. 2009). Calcium chelating agents, when co-administered with either MeHg, HgCl2, or thimerosal, were able to attenuate PLD activation (Peltz et al. 2009). Ca²⁺ and thiol depletion are also involved in the Hg-induced activation of the phospholipase A2 (PLA2) pathway. This pathway begins with PLA2 cleaving arachidonic acid (AA) from cell membrane phospholipids. AA can then be metabolized through the COX pathway, which converts AA into various prostaglandins and thromboxanes. These molecules play important roles in inflammation and platelet aggregation (Hagele et al. 2007; Mazerik et al. 2007, 2008). MeHgCl stimulates TXA₂ and prostacyclin (PGI2) in the heart under in vitro conditions, and this effect was inhibited by both thromboxane synthetase and phospholipase A2 inhibitors (Ally et al. 1993). HgCl₂ was shown to increase thromboxane A2 (TXA₂) following decreased NO. TXA₂ is known to contribute to platelet aggregation and vasoconstriction (Cordeiro et al. 2019).

Alteration of the renin–angiotensin system

Hg can alter the renin–angiotensin system by enhancing angiotensin I-converting enzyme (ACE 1) and renin 2. These are released from the kidney and lungs, respectively, and ultimately result in increased angiotensin II, which has been shown to trigger ROS formation in vascular smooth muscle cells, which inhibits NO and cGMP, contributing to vasoconstriction. Chronically HgCl₂-treated rats showed increased plasma angiotensin I-converting enzyme (ACE), which is a potential contributing factor in angiotensin II formation (Carmignani et al. 1992). Other studies have shown increased plasma renin and a simultaneous BP increase following MeHgCl exposure; this could further promote angiotensin II synthesis (Wildemann et al. 2016). Furthermore, downstream angiotensin II is known to decrease NO synthesis via ROS formation. In the aortic ring, HgCl₂ exposure caused increased vasoconstriction and angiotensin II, ultimately reducing NO synthesis (Lemos et al. 2012). Similar results have been observed in rats where an initial dose of 4.6 μg/kg/BW HgCl₂ and subsequent doses of 0.07 μg/kg/BW/day HgCl₂ were given for 30 days. Left coronary arteries had increased NO production triggered by increased O₂⁻ (Furieri et al. 2011). Enhanced Rho-kinase activity and inhibition of NO have also been observed in mice following 21 days of 40 ppm MeHg (Islam et al. 2016).

Angiotensin II receptor (AT-1R) may play a critical role in this pathway. Angiotensin II decreases NO by binding to the AT-IR on the surface of vascular smooth muscle cells. This activates ROS, which then scavenges NO or indirectly limits NO synthase (Fig. 1). Co-treatment with losartan, an AT-1 receptor blocker, prevented the deleterious effects of HgCl₂ on vascular reactivity and oxidative stress (Rizzetti et al. 2018).

Inflammation

ROS production following Hg exposure can trigger NF-kB, leading to the transcription of pro-inflammatory cytokines (TNF-a and IL-6) and atherosclerosis (Baiyun et al. 2018). ROS are also critical to triggering the cytoprotective pathway Keap1/Nrf2. Kelch-like ECH-associated protein 1 (Keap1) contains a cysteine residue that Hg can bind to. Under stress conditions, electrophiles and ROS bind to Keap1, which causes nuclear factor erythroid 2-related factor 2 (Nrf2) to migrate from the cytoplasm into the nucleus, leading to transcription of glutamate–cysteine ligase (GCL) and antioxidant defense enzymes (Ni et al. 2010; Toyama et al. 2007; Wang et al. 2009). However, decreased Nrf2 and decreased downstream antioxidant targets (HO-1 and NQO1) were observed following HgCl₂ exposure in liver cells, potentially due to the crosstalk between Nrf2 and NF-κB (Baiyun et al. 2018; Zhang et al. 2017).

Luteolin is a natural antioxidant, and co-administration of luteolin and HgCl₂ in Wistar rats has been shown to rescue histopathological alterations and oxidative stress infected by HgCl₂ alone. Luteolin can elevate Nrf2 levels and inhibit NF-kB. Overall, it prevents inflammation and oxidative stress caused by HgCl₂ alone (Baiyun et al. 2018).

Brief summary

In the sections above, we highlighted the potential relationship between various forms of Hg on hypertension development. At the organ and tissue level, Hg has been shown to cause structural and functional changes. Moreover, we emphasize the multi-faceted impact of Hg on cardiovascular health, including its effects on vascular reactivity, vasoconstriction, atherosclerosis, dyslipidemia, and thrombosis. At the cellular and molecular levels, we highlight the similarity in the ability of both inorganic and organic Hg to induce impairments in antioxidant defense mechanisms, leading to increased ROS production, which causes endothelial dysfunction and alteration of the renin–angiotensin system, leading to inflammation, and resulting in vasoconstriction. While both inorganic Hg and MeHg have been reported to have similar effects at the cellular and molecular levels, as highlighted above, key differences exist. Mainly, limited cellular uptake and intracellular distribution limit the role of inorganic Hg. This is primarily due to its lack of lipophilicity. This may provide some insight into why MeHg has more extensive cardiovascular effects, which was highlighted in multiple rodent studies (Wildemann et al. 2015a, b, 2016).

The significance of these findings lies in the comprehensive understanding of Hg’s impact on organ, tissue, cellular, and molecular functioning leading to cardiovascular dysfunction. Overall, clarifying the molecular initiating event is critical to understanding and effectively treating Hg-induced cardiovascular dysfunction. A summary of these effects is presented in the form of an adverse outcome pathway in Fig. 2.

Fig. 2.

Adverse outcome pathway (AOP) of MeHg exposure and hypertension onset. Abbreviations: AO, adverse outcome; AOP, adverse outcome pathway; KE, key event; MIE, molecular initiating event

Population studies on mercury and hypertension

Our 2018 systematic review included 29 studies, a significant positive association between Hg and BP was identified, and a non-linear dose–response relationship is plausible for both SBP as a continuous measurement or hypertension (defined as SBP ≥ 140 mmHg or DBP ≥ 90 mmHg) (Hu et al. 2018). We recommended that hair Hg at 2 μg/g be considered as a potential candidate regulatory value for international agencies [e.g., the WHO and the Environmental Protection Agency (EPA)] to update their guidelines on Hg exposure (Hu et al. 2018).

In this review, we aim to update the systematic review and elaborate on the following key aspects: 1) investigate how the source and type of Hg exposure and the use of different biomarkers influence the association between Hg and hypertension; 2) identify populations at higher risk due to increased exposure duration, elevated doses, or exposure during vulnerable life stages; 3) explore how simultaneous exposure to Hg and other heavy metals might interact, potentially amplifying or mitigating Hg’s impact on BP; 4) examine the interaction between Hg exposure and dietary factors, e.g., omega-3 fatty acids and selenium, and how these might modulate the association between Hg and hypertension; 5) understand the relationship between Hg exposure and health conditions related to hypertension, as well as broader cardiovascular health outcomes. Finally, we weigh the evidence of causality and discuss the potential regulatory value of Hg and public health recommendations to reduce the health burden associated with Hg exposure.

Update on literature published since 2018

Using the same search criteria, we found 16 new studies reporting findings on the relationship between Hg exposure and hypertension in different populations between 2018 and 2024 (Table 2). Most of the studies (9) were from the United States, 3 were from China, and the rest were from South Korea, Canada, Kazakhstan, and Brazil. The sample size ranged from 84 to 8371 participants. Five studies reported urinary Hg to reflect inorganic Hg exposure (Bulka et al. 2019; Farzan et al. 2021; Shinetova et al. 2020; Wang et al. 2021; Nunes et al. 2022), the rest of the studies reported MeHg or total Hg exposures measured in blood. At least nine of these studies could potentially provide additional information on the dose–response relationship between Hg exposure, BP, and hypertension. The rest of the studies are not eligible for dose–response meta-analysis due to several reasons: 1) the study participants are pregnant women only (Liu et al. 2019; Louopou et al. 2020; Wang et al. 2020), 2) the biomarker reported in the study cannot be converted to hair Hg (Ma et al. 2022, 2023; Nunes et al. 2022), and 3) lack of quantitative Hg exposure data (Guo et al. 2022b).

Table 2.

Characteristics of studies published after 2018 and included in the review

Reference	Population	Exposure route	Co-exposure or interaction with other metals	Mean age or age range (years)	Male (%)	N	Biomarker (unit)	Form	Mercury concentration	Outcome	Definition of HPT	Blood pressure measurement	Variables adjusted for
Bulka et al. (2019)	USA	General population	Yes	54.8	52.7	1088	Urine (µg/h)	Inorganic	GM 17.8	HPT	130/85 mmHg, BP medication		Age, gender, race/ethnicity, family income: poverty ratio, total caloric intake, educational attainment, smoking status, alcohol consumption, physical activity status, survey cycle, and BMI
Kim et al. (2019)	South Korea	Population living in mercury polluted regions		≈65	31.6	7822	Blood (µg/L)	Total	Mean 5.2 (SD: 4.5)	HPT	140/90 mmHg, BP medication	Measured with standard mercury sphygmomanometers	Age, sex, smoking status, alcohol drink, and income
Liu et al. (2019)	USA	Pregnant women	Yes	28.3	0	1274	RBC (µg/L)	Total	(range: 0.3, 27.8)	HPT	140/90 mmHg		Age at delivery (continuous), self-reported race (black, nonblack), education (below high school, high school, college or above), parity (nulliparous, multiparous), pregnancy body mass index (continuous), and smoking status during pregnancy (never, former, current)
Louopou et al. (2020)	Canada	Pregnant women		31.9	0	1817	Blood (µg/L)	Methyl		HPT, BP	140/90 mmHg, BP medication	Average of 2 readings using a mercury sphygmomanometer	Age, BMI, fish consumption, weight gain, coffee intake, education, household income, ethnicity, parity, multiple child pregnancy, maternal smoking
Shinetova et al. (2020)	Kazakhstan	Population living in mercury polluted regions		54	56	84	Urine (µg/L)	Inorganic	GM 2.5 (SD: 0.9)	HPT	130/85 mmHg, BP medication	Medical records
Wang et al. (2020)	China	Pregnant women	Yes	20–30	0	854	Blood (µg/L)	Total	Case Median 1.52 (IQR: 0.97, 2.36) Control Median 1.49 (IQR: 0.96, 2.08)	HPT	140/90 mmHg	Medical records	Pre-pregnancy BMI, education, household monthly income per capita, and gestational age
Yao et al. (2020)	USA	General population	Yes	8-17	50.9	7076	Blood (µg/L)	Total/Methyl	Mean 0.44	BP	NA	Average of three measurements with standard mercury sphygmomanometer
Desai et al. (2021)	USA	General population	Yes	12.5	49.8	1642	Blood (µg/L)	Total	Median 0.37	BP	NA	Average of three measurements with standard mercury sphygmomanometer	Age, sex, race, BMI, total energy intake, cycle, education of household head, and income to poverty ratio
Farzan et al. (2021)	USA	General population		5.5	48.6	395	Urine (µg/L)	Inorganic	Mean 0.07	BP	NA	Average of five measurements with	Age, height, weight; child sex, birth weight, gestational age; maternal education and smoking during pregnancy; and urine specific gravity
Wang et al. (2021)	USA	General population	Yes	49.4	0	1317	Urine (µg/L)	Inorganic	Mean 1.2	BP	NA	Average of 2 readings using a mercury sphygmomanometer	Age at baseline, race/ethnicity, study site, specific gravity (log-transformed), education, smoking status, alcohol drinking, physical
Xu et al. (2021)	USA	Occupational population	Yes	>30	74	957	Blood (µg/L)	Total		HPT, BP	140/90 mmHg, BP medication	Average of two measurements	Age, sex, race, education, income, BMI
Guo et al. (2022a, b)	USA	General population	Yes	≥20	49.6	8371	Blood (µg/L)	Total	GM 3.90	BP	NA	Average of second- and third-time measurements in 5 min interval using standard mercury sphygmomanometer	Age, gender, smoking status
Ma et al. (2022)	China	Pregnant women	Yes	30	0	438	Plasma (µg/L)	Total	Case Median 0.33	HPT	140/90 mmHg	Medical records	Maternal age at enrollment, gestational age at blood sample collection, household income, pre-pregnancy BMI, parity, passive smoking, and gestational diabetes mellitus
Nunes et al. (2022)	Brazil	General population		59	42	112	Urine (µg/g creatinine)	Inorganic	Median	BP	NA		Age, sex, BMI, smoking and alcohol consumption
Tang et al. (2022)	USA	Asian population	Yes	>20	47.3	1422	Blood (µg/L)	Total/Methyl	Mean 1.95	HPT, BP	130/80 mmHg, BP medication	Average of three measurements with standard mercury sphygmomanometer	Age, sex, education, annual household income, cotinine level, alcohol use, BMI, omega-3 fatty acids and selenium
Ma et al. (2023)	China	Prenatal exposure		5–6	51.6	2535	Maternal serum (µg/L)	Total	GM 1.03	HPT, BP	NA	Average of 3 measurements using electronic sphygmomanometer	Maternal age, maternal education, monthly family income, parity, smoking during pregnancy and family history of hypertension. child age, BMI, child sex and feeding patterns

BMI, body mass index; BP, blood pressure; HPT, hypertension; GM, geometric mean; IQR, inter-quartile range; SD, stan study

In five out of the nine studies, the highest exposed group had average Hg levels equal to or above 2 μg/g hair Hg (Bulka et al. 2019; Farzan et al. 2021; Kim et al. 2019; Shinetova et al. 2020; Wang et al. 2021). These five studies all reported a positive association between Hg and either SPB/DBP or hypertension. Two studies investigated residents of industrially polluted areas (Kim et al. 2019; Shinetova et al. 2020), and the remaining three investigated general populations. Four of the studies used urinary Hg as a biomarker of Hg exposure, while one used blood Hg (Kim et al. 2019). In the other four studies, the average Hg level in the highest exposed group was below the cut-off (Desai et al. 2021; Tang et al. 2022; Xu et al. 2021; Yao et al. 2020). Three studies reported no or inverse association between Hg, BP, or hypertension (Desai et al. 2021; Xu et al. 2021; Yao et al. 2020), while one study reported a positive association in a subgroup of participants (Non-Hispanic Asian) and no association in the rest of the study populations (Tang et al. 2022). All four studies were from the US, and three were based on the NHANES (Desai et al. 2021; Tang et al. 2022; Yao et al. 2020).

The findings of the nine additional studies reconfirmed our conclusion on the overall association between Hg exposure, BP, and hypertension. Hair Hg 2 μg/g was a reasonable cut-off for the association between Hg exposure, BP, and hypertension. The shape of the dose–response relationship curve may be worth revisiting as well. However, several uncertainties and research gaps remain. First, evidence suggesting no or an inverse association at a lower dose (<1 μg/g hair Hg) is the most abundant, followed by a clear positive association from higher end exposure (>5 μg/g hair Hg), and the least evidence with inconsistent findings from medium-exposed populations. Second, the population exposed to low, medium, and high Hg was likely to be heterogeneous. Many of the studies with low Hg exposure participants were likely from the few national biomonitoring studies, e.g., NHANES, or other studies conducted in populations consuming a typical “western” diet that consists of only a small amount of fish and seafood. Participants with higher end exposure were either residents living close to industrial contaminated sites or workers being exposed occupationally. The medium-exposed participants were likely coastal or Indigenous populations who were exposed to MeHg through fish and seafood consumption. Besides the heterogeneity in the number of studies available and study population by exposure doses, several other key factors may influence the relationship between Hg exposure, BP, and hypertension.

In the following sections, we will investigate how the source and type of Hg exposure and the use of different biomarkers influence the association between Hg and hypertension and discuss the effects of different confounding factors.

Sources, forms, and biomarkers of mercury exposure

One main challenge to establishing a robust dose–response relationship is that Hg exists in various chemical forms, and each of these forms is characterized by distinct exposure sources, target organs, toxicological properties, and metabolic pathways. With currently available epidemiology studies, it is not feasible to elaborate on whether Hg exposure from different sources or different forms of Hg, affects BP following the same or similar pathways. However, for public health purposes, scientists would like to explore if Hg’s impact on BP or hypertension is similar, given that all populations are exposed to a certain level/dose of Hg, regardless of its source and form. Furthermore, many of the epidemiology studies assess Hg exposure with biomarkers, which adds another layer of uncertainty to estimating the overall relationship between Hg, BP, and hypertension.

It is important to delineate the impact of biomarkers in two aspects. First, whether the association between Hg and BP is consistent within the same study measuring different biomarkers, and second, whether the association remains similar across different studies measuring different biomarkers with similar exposure levels. As for the impact of exposure sources and forms of Hg, most scientists are mainly interested in comparing the associations between MeHg exposure from fish and seafood and inorganic Hg exposure from occupational sources at moderate to higher end exposure. Four underlining assumptions are needed to support the investigation of these associations. First, urinary Hg primarily reflects exposure to inorganic Hg, e.g., exposure from mining. Second, hair Hg or blood MeHg primarily reflects exposure to MeHg, e.g., from fish and seafood consumption. Third, blood total Hg reflects exposure to all forms of Hg. Finally, different biomarkers are convertible.

Evidence from different biomarkers in the same study population

Several studies reported more than one biomarker for the same participants, including hair and blood Hg (Bautista et al. 2009; Choi et al. 2009; Valera et al. 2011a), blood and urinary Hg (Kobal et al. 2004; Park et al. 2013), and hair and urinary Hg (Goodrich et al. 2013; Rajaee et al. 2015; Siblerud 1990). These studies covered a wide range of Hg exposure levels and were from both general population and occupational exposures. Generally, when comparing the BP between high-exposure and low-exposure groups classified by different biomarkers within the same study, the estimates were always in the same direction and with similar magnitudes. For example, it was estimated that SBP was 3.4 mmHg (95% CI: −0.5 to 7.3) higher in the high-exposure group measured by hair Hg and 2.7 mmHg (95% CI: −1.2 to 6.6) higher measured by blood Hg (Bautista et al. 2009). The differences in SBP were −2.0 mmHg (95% CI: −3.1 to −0.9) and -3.2 mmHg (95% CI: −5.2 to −1.2) between exposure groups defined by blood and urinary Hg, respectively (Park et al. 2013). The corresponding estimates were 1.2 mmHg (95% CI: −5.7 to 8.1) and 1.0 mmHg (95% CI: −3.9 to 5.9) between exposure groups defined by hair and urinary Hg (Rajaee et al. 2015). The results were similar in the other studies identified for DBP, with two exceptions in the occupational setting. One was that the result from blood Hg was opposite to urinary Hg, which is the preferred biomarker for miners exposed to inorganic Hg (Kobal et al. 2004). The other was between hair and urinary Hg among dentists (Goodrich et al. 2013). In general, the association between Hg exposure and BP was consistent across three major biomarkers (hair, blood, and urinary Hg) in the non-occupational populations, given the same or similar exposure level. Therefore, it is reasonable to synthesize evidence from different studies using different biomarkers with proper conversion. Hair Hg level seems to be an appropriate biomarker for the comparison across studies.

Evidence from different biomarkers in populations with similar exposure levels

The association between Hg exposure and BP was examined thoroughly in both the United States (US) population and the Indigenous population with high fish consumption, including Inuit living in the Arctic. Hg exposure in the US general population is low regardless of which biomarkers were measured (Bautista et al. 2009; Desai et al. 2021; Mordukhovich et al. 2012; Mozaffarian et al. 2012; Park et al. 2013; Vupputuri et al. 2005; Xu et al. 2021; Yao et al. 2020). Also, regardless of the biomarkers used and the study design, no or inverse association between Hg, BP, or hypertension has been reported in the US general population consistently. These studies contribute substantially to the body of literature investigating Hg exposure and BP at lower doses (less than 2 μg/g hair Hg or equivalent), and the findings were relatively homogenous as many of them were using the National Nutrition and Health Examination Survey. On the other hand, Indigenous populations are generally exposed to Hg higher than 2 μg/g hair Hg or equivalent through their traditional diet (Houde et al. 2022; Pirkle et al. 2016). Blood Hg (Hu et al. 2017; Valera et al. 2009, 2011b, 2013) and hair Hg (Choi et al. 2009; Fillion et al. 2006; Valera et al. 2011a) were commonly assessed in these populations to reflect their dietary exposure to MeHg. Positive associations between Hg exposure and BP outcomes were consistently reported in these studies, except for one that reported no significant association among 313 Inuit adults in northern Canada (Valera et al. 2013). Studies from the Nordic population (Daneshmand et al. 2016; Pedersen et al. 2005; Virtanen et al. 2012a, b) and South Korea (Eom et al. 2014; Hong et al. 2013; Kim et al. 2019; Nguyen and Kim 2022; Park and Choi 2016) in general fell into moderate Hg exposure, and they generally showed positive associations.

The consistent associations observed within populations with similar exposure levels suggest that, although the use of different biomarkers may introduce variations to the estimated magnitudes of association between Hg exposure and BP outcomes to a certain extent, it is unlikely to change the direction and significance of the association within a given exposure range. The differences in association reported between the US population and other populations with higher dietary Hg exposure suggest that Hg exposure level is a more important factor influencing the association between Hg exposure and BP outcomes.

Evidence of population exposed to different sources and forms of mercury

Since the studied populations are exposed to varying sources of Hg that are in different forms, the studies reported in the literature often assessed Hg exposure using different biomarkers. For instance, individuals exposed to inorganic Hg via mining activities typically exhibit considerably higher Hg exposure levels compared to those exposed to MeHg through fish consumption. Conversely, individuals who consume minimal or no fish typically show notably lower Hg exposure levels. In this context, observed disparities in the associations between Hg exposure and BP outcomes could be attributed to variations in either Hg exposure levels, in sources or forms, or both.

Therefore, in this review, we focused on evaluating the associations reported within populations exposed to comparable levels of either inorganic Hg or MeHg. This approach aimed to mitigate potential confounding factors arising from dissimilar exposure levels, allowing for a more accurate and balanced assessment of the associations between these distinct forms of Hg and their respective health outcomes.

At the higher end of Hg exposure, Hg miners from Slovenia in the 1970s had an average annual urinary Hg level of over 26 μg/L, certain residents of Minamata, Japan, and Amazon, Brazil could have a hair Hg level over 30 μg/g (Fillion et al. 2006; Yorifuji et al. 2010). All three studies reported a positive association with similar magnitudes, with odds ratio comparing the highest to lowest exposure groups 2.1, 2.5, and 3.8, respectively. In a moderate exposure scenario, individuals with dental amalgam had hair Hg 1.4 μg/g and an average of 5.7 mmHg higher in SPB, compared to those without (Siblerud 1990). In populations exposed to similar MeHg from fish intake (Daneshmand et al. 2016; Hong et al. 2013; Hu et al. 2017; Park and Choi 2016). The estimated differences in SBP comparing high–to-low-exposure groups in these studies were also similar, ranging from 2.0 to 4.4 mmHg.

At the low end of Hg exposure, it is difficult to distinguish the sources and contributions of different Hg forms to the overall exposure. Also, given that null or inverse associations were consistently reported in populations with hair Hg below 2 μg/g (Hu et al. 2018), it can be assumed that populations with low Hg exposure should have no concern about effects on BP, regardless of the form of Hg.

Brief summary

In summary, the associations observed remained largely consistent across various biomarkers, both within the same study population/participants and among populations with similar exposure levels. Additionally, when comparing populations exposed to broadly similar levels of Hg but differing in the forms and sources, the associations exhibited remarkable similarity. This consistency in associations across diverse biomarkers and exposure scenarios suggests the plausibility of establishing a dose–response relationship between Hg exposure and BP outcomes, using either measured or converted hair Hg concentrations.

Populations at higher risk

Since exposure dose is the driving factor of the association between Hg and hypertension, it is worth identifying the population groups that are more at risk due to their dietary habits, occupation, or geographical location. Monitoring and addressing Hg exposure in such populations are crucial for mitigating health risks associated with long-term Hg exposure.

Coastal populations and fishing communities are at risk of elevated Hg exposure due to fish and marine mammal consumption as Hg biomagnifies through the food chain (Driscoll et al. 2013; Lavoie et al. 2013). This includes Indigenous populations and communities reliant on subsistence fishing (ATSDR 2022). Workers in industries such as mining, smelting, or manufacturing products involving Hg are at risk of higher exposure due to direct Hg contact (ATSDR 2022; Driscoll et al. 2013). Individuals involved in small-scale gold mining often use Hg to extract gold from ore, leading to substantial exposure through inhalation or skin contact (ATSDR 2022; Esdaile and Chalker 2018). Certain regions might have higher levels of Hg exposure due to natural deposits, industrial pollution, or mining activities. Residents in such affected areas are at higher risk of Hg exposure as well (ATSDR 2022; Aubrac et al. 2022). Certain traditional medicines and cosmetic products might contain Hg as an ingredient, and as a result, consumers of these products may be exposed to Hg (Bastiansz et al. 2022).

In addition to the populations with a higher risk of elevated Hg exposure, certain populations may be more sensitive to the adverse effects of Hg due to various factors. Certain groups of individuals are sensitive to any stressor, e.g., the elderly and individuals with the existing health conditions. Hg exposure may exacerbate their health conditions through various pathways. Some individuals may have genetic variations affecting their ability to metabolize or excrete Hg efficiently, leading to increased susceptibility to Hg toxic effects (Andreoli and Sprovieri 2017). Several recent studies suggest that Hg exposure is related to hypertensive disorders of pregnancy or postpartum BP. However, results on this topic are inconsistent, and more research is needed (Dantas et al. 2022; Liu et al. 2019; Louopou et al. 2020; Ma et al. 2022; Wang et al. 2020). Exposure to Hg during crucial early developing stages, e.g., fetus and infants, may also have additional long-term adverse effects.

Hg crosses the placental barrier and accumulates in fetal tissues, especially the developing brain. Evidence suggests that prenatal exposure may lead to long-term unfavorable neurological and cardiovascular outcomes (Counter and Buchanan 2004; Davidson et al. 2004; Gallego-Viñas et al. 2019). Several studies reported prenatal Hg exposure was associated with alterations in BP in children or adolescents (Farzan et al. 2021; Ma et al. 2023; Sørensen et al. 1999; Thurston et al. 2007). Among these two studies were from highly exposed populations—Seychelles and the Faroe Islands (Sørensen et al. 1999; Thurston et al. 2007). Negative findings (no association) or partial associations (significant results in certain subgroups) were reported as well (Grandjean et al. 2004; Gregory et al. 2016; Kalish et al. 2014; Valera et al. 2011b, 2012). The mixed findings observed suggest the possibility of a dose-dependent relationship between prenatal Hg exposure and BP during childhood.

Mixed findings were reported for postnatal exposure as well. Two studies reported a positive significant association (Farzan et al. 2021; Poursafa et al. 2014), and interestingly, the association observed in prenatal exposure in Seychelles and the Faroe Islands became null in children and adolescent with postnatal exposure assessment (Grandjean et al. 2004; Valera et al. 2011b, 2012). This variability in results suggests that high fish intake or other public health interventions during childhood may play a crucial role in mitigating the adverse effect of prenatal Hg exposure to BP outcomes. It is also worth clarifying whether prenatal and postnatal exposure to Hg influences BP similarly.

Chemical mixtures and interaction with other heavy metals

The most commonly studied Hg chemical mixtures in animals are Hg and lead (Pb) and Hg and cadmium (Cd). Co-exposure to Hg and Pb has been studied in rats. Here, rats were exposed to a mixture of Hg²⁺, Pb, and MeHg for 4 weeks via drinking water. Exposure to MeHg alone raised SBP and pulse pressure, while exposure to the mixture had no effect relative to control. Potentially because Hg²⁺ and Pb work as an antagonist of MeHg (Wildemann et al. 2015a, b). Similar results were observed in a follow-up study whereby MeHg-induced hypertension was not seen following exposure to Hg²⁺, Pb, and MeHg chemical mixtures (Wildemann et al. 2016).

Co-exposure of Hg and Cd has also been extensively studied. As highlighted previously, exposure to HgCl₂ alone has been shown to increase collagen and elastin in the heart muscle (Olczyk et al. 1994). Here, co-exposure to both Cd and HgCl₂ for 28 days leads to further collagen deposition (Arbi et al. 2021). A study in rabbits showed that following both HgCl₂, Cd, and HgCl₂ + Cd exposure TC was elevated, while HDL and LDL were lowered; suggesting that both Cd and Hd can also contribute to dyslipidemia (Ali et al. 2020).

Silver nanoparticles (SiNPs) are a more recent area of study. A recent study exposed human cardiomyocytes to MeHg and SiNPs. This caused enhanced ROS production and lowered SOD and GPx (Yang et al. 2018). Overall, studies that assess co-exposure to multiple chemicals may more accurately replicate real-world toxicant exposure and may clarify epidemiological results.

Many of the human studies investigated the association between mixtures of heavy metals and cardiovascular outcomes using biomonitoring data with different matrices of biomarkers (e.g., blood, urine, and serum) and various statistical methods. The available evidence suggests a positive association between hypertension (or increased BP) and the concurrent exposure of Hg with other metals, particularly Pb, arsenic, and Cd (Bulka et al. 2019; Park and Choi 2016; Poursafa et al. 2014; Shiue 2014; Yao et al. 2021). This co-exposure to heavy metals was reported to be related to other metabolic conditions (Bulka et al. 2019; Yao et al. 2021; Nunes et al. 2022; Qu et al. 2022; Shiue 2014; Poursafa et al. 2014; Rotter et al. 2015), heart diseases, and mortality, as well (Yao et al. 2021). A good proportion of the positive findings were from low Hg populations, e.g., the US population (Bulka et al. 2019; Guo et al. 2022a; Park and Choi 2016; Shiue 2014; Yao et al. 2021). This underscored the importance of the chemical mixture approach in assessing Hg’s toxic effect at exposure levels relevant to general populations, which are likely to be well below regulatory values. Nonetheless, these findings need to be confirmed with further high-quality prospective studies and in other populations with similar biomonitoring data. There are studies showing either no association from heavy metal mixtures or the positive association was attributed primarily to metals other than Hg (Desai et al. 2021; Guo et al. 2022b; Liu et al. 2022; Mordukhovich et al. 2012; Nunes et al. 2022; Qu et al. 2022; Rotter et al. 2015).

Interaction with omega-3 fatty acids and selenium

Certain foods and nutrients, on the other hand, interact with Hg, either directly or indirectly, to mitigate Hg’s adverse effect on BP. Omega-3 fatty acids and selenium are the two most investigated nutrients, which could potentially counteract Hg’s adverse effect due to their anti-inflammatory and anti-oxidative properties. Furthermore, selenium can bind with Hg and form less toxic complexes. It is reasonable to assume that most of the findings in this section were more relevant to MeHg exposure, although studies might report blood total Hg or other Hg exposure biomarkers.

Omega-3 fatty acids

The interaction between omega-3 fatty acids and Hg on BP can be investigated through two approaches. First, by synthesizing studies that directly address their interaction, and second, by comparing studies that reported Hg exposure from fish consumption against those with similar Hg exposure from other sources. One common approach to investigating the direct interaction was to stratify the study participants by Hg and omega-3 fatty acids (fish consumption) levels and then compare the outcome of interest among different categories. No clear interaction between Hg and omega-3 fatty acids on hypertension was observed (Virtanen et al. 2012b). The protective effect of omega-3 fatty acids on hypertension was attenuated by Hg (Xun et al. 2011).

More studies were available reporting the interaction on other cardiovascular or metabolic outcomes, e.g., stroke, coronary heart disease (Downer et al. 2017; Guallar et al. 2002; Virtanen et al. 2005; Wennberg et al. 2012), and blood cholesterol (Smith et al. 2009). Except for Downer 2017, the rest studies reported consistently that Hg was associated with an increased risk of cardiovascular outcomes, while omega-3 fatty acids were inversely associated with these outcomes, with the two effects offsetting each other.

In summary, the epidemiological evidence on the interaction between Hg and omega-3 fatty acids in relation to hypertension is not as clear as it is for other cardiovascular outcomes, such as myocardial infarction. More studies specifically focused on BP outcomes are needed. Additionally, most studies examining the interaction between Hg and omega-3 have stratified their exposures into high or low categories based on the distribution within the study population. This leaves uncertainty about the precise levels at which the interaction begins. More studies with Hg exposure at or around 2 µg/g hair mercury will help to better quantify such interaction. Furthermore, it remains unclear whether Hg and omega-3 interact by regulating the same or similar pathways with offsetting effects on hypertension or if they work individually through separate pathways, ultimately resulting in an offsetting effect. More experimental studies on that perspective will shed light on the underlying mechanism.

Selenium

The approach to examine the interaction between selenium and Hg is similar to that for omega-3 fatty acids. Significant interactions between selenium and Hg were reported in two studies at moderate-to-high Hg exposure levels, with mean hair Hg of 12 µg/g and mean blood Hg of 7.8 µg/L, respectively (Afridi et al. 2014; Hu et al. 2017). The findings are inconclusive in populations with low Hg exposures. No interaction was observed in two studies in the US population (Downer et al. 2017; Mozaffarian et al. 2012). However, the potential offsetting effect between Hg and selenium cannot be completely ruled out for two reasons. First, Hg may not have any adverse effect at such an exposure level. Second, most of the US population has adequate selenium intake; Hg may exhibit certain adverse effects on BP in selenium-depleted populations. Another study reported a three-way interaction between omega-3 fatty acids, selenium, and Hg on hypertension in a 20-year follow-up study (Xun et al. 2011). In that study, the protective effect of omega-3 fatty acids on hypertension was most pronounced in the high selenium, low Hg group and was most diminished in the low selenium, high Hg group. This finding provides indirect evidence of an interaction between Hg and selenium in relation to hypertension (Xun et al. 2011).

Similar to the interaction between Hg and omega-3 fatty acids, more epidemiological studies with varying levels of Hg and selenium exposures are needed. These studies should aim to first confirm the presence of such an interaction and second to quantify the exact dose at which this interaction starts. More mechanism studies are needed to understand how Hg and selenium interact on various biological pathways associated with BP outcomes. It is also important to investigate whether omega-3 fatty acids and selenium interact with inorganic and organic forms of Hg similarly. Understanding this distinction could provide valuable insights into the mechanisms of their interaction and their potential impact on health outcomes.

Besides that, several unique features of selenium are worth considering for future studies. First, there are different forms of selenium compounds, and animal studies suggest that the effectiveness of each compound in reducing Hg toxicity is different (Magos et al. 1981; Yamashita et al. 2013). More human studies are needed to understand the impact of selenium intake from different sources on their health implications. Second, in populations in which fish and seafood are their primary source of selenium intake, certain indexes such as the selenium/Hg ratio were proposed to guide fish choice (Barone et al. 2021; Gochfeld and Burger 2021; Ralston et al. 2008). Third, selenium may exhibit a toxic effect on certain health outcomes at a high dose (EFSA Panel on Nutrition et al. 2023).

Mercury exposure and conditions related to hypertension

There has been a recognized pattern of hypertension co-existing with other risk factors, such as obesity, elevated blood glucose, and unfavorable blood cholesterol levels (Huang 2009). These combined conditions substantially elevate the likelihood of developing diabetes, heart disease, stroke, or all three (Mottillo et al. 2010). If the association between Hg exposure and hypertension is not coincidental, it is expected that Hg exposure would demonstrate correlations with other metabolic conditions and CVDs. The present review provides some key references from this perspective without a comprehensive literature search and quality assessment.

There is evidence suggesting that Hg levels are significantly higher among individuals with type-2 diabetes compared with those without; however, the association between Hg exposure and type-2 diabetes is not conclusive (Ghorbani Nejad et al. 2022; Guo et al. 2023; Roy et al. 2017). There is increasing evidence outlining the positive association between Hg exposure and adverse blood cholesterol profiles (Cho 2017; Farkhondeh et al. 2020; Park et al. 2016; Sohn et al. 2020; Xu et al. 2023). A systematic review and meta-analysis identified a positive association between Hg exposure and obesity (Jeon and Park 2023). Another systematic review and dose–response meta-analysis from our group revealed that chronic exposure to Hg was associated with an increased risk of all-cause mortality and fatal/nonfatal IHD (Hu et al. 2021). The risk of multiple cardiovascular endpoints starts to increase consistently at a hair Hg concentration of 2 μg/g (Hu et al. 2021).

These findings strongly suggest that the link between Hg exposure and hypertension is not coincidental. There is a growing indication that Hg exposure might serve as a notable risk factor not just for hypertension but also for other metabolic conditions, CVDs, and even mortality. However, the emerging evidence was largely from biomonitoring data from the US and Korean populations. More high-quality prospective studies from populations with different Hg exposure levels are needed to solidify these associations and draw a more robust conclusion.

Weight of evidence for causality

The plausibility of the association between Hg exposure and BP outcomes was evaluated with the Bradford Hill criteria (Fedak et al. 2015; Hill 1965). Upon comprehensive evaluation of all evidence identified in the current review, alongside previous reviews and peer-reviewed articles, hypertension, SBP, and DBP were discussed as one outcome. Criteria 1, Strength of association—the available evidence suggested a moderate-to-strong magnitude (strength) of association between Hg exposure and BP, especially at hair Hg over 2 µg/g or equivalent. Criteria 2, Consistency—the available evidence suggested the association was consistent in several ways: 1) the association observed in the same population or population with similar exposure levels were consistent, 2) the association observed from different biomarkers were consistent, and 3) the association observed for hypertension, SBP, and DBP were consistent. Criteria 3, Specificity—Hg appeared to be involved in multiple mechanism pathways and play a role in multiple metabolic risk factors. The association between Hg exposure and BP was not specific. Criteria 4, Temporality—most epidemiological studies currently available are cross-sectional. However, there are also plenty of prospective studies suggesting a positive association between Hg exposure and BP outcomes. Criteria 5, Biological gradient and plausibility—significant dose–response relationship between Hg exposure and BP outcomes were reported. In vitro, in vivo, and animal studies provided supportive evidence on various pathways that are linked to BP outcomes. Criteria 6, Coherence—mechanistic studies provide supporting evidence to epidemiological studies. Criteria 7, Experiment —experimental evidence from animal studies is conclusive. Experimental evidence in human exposure was rare, although studies investigating Minimata disease or comparing populations living within and outside of Hg-polluted areas may be considered quasi-experiment evidence. Criteria 8, Analogy—the evidence suggesting interactions between omega-3 fatty acids, selenium, and Hg provides certain insight into the mechanism of association between Hg exposure and BP.

Conclusion

This review has presented conclusive evidence on the cause–effect of Hg exposure and hypertension in animal studies, including the mechanistic studies that provide supportive evidence on this association. Strong evidence from epidemiological studies shows that Hg exposure is associated with an increased risk of hypertension and elevated SBP/DBP with a dose–response relationship. The association between Hg exposure and BP is mixed at lower exposure levels; however, increasing evidence suggests that Hg can play a role in BP outcomes co-exposed with other metals, even at low doses. The availability of evidence is also uneven across three forms of Hg (elementary, inorganic, and organic) and different biomarkers. Based on biomarker conversion, a dose–response relationship was derived, assuming that different forms of Hg exhibit similar toxic effects on BP outcomes.

Further research to develop more robust conversion factors among different biomarkers and standardized/harmonized measures of Hg exposure will help to better understand the association between Hg and BP. For risk management purposes, a practical framework is lacking to integrate mechanism and epidemiological studies to derive a robust regulatory value on Hg’s toxic effect on cardiovascular endpoints. The current evidence continues to support the proposed 2 µg/g hair Hg or equivalent as the cut-off level for public health regulation. Regulatory agencies should consider adopting this level, specifically for adult men or women older than child-bearing age.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Conflict of interest

None of the authors has a conflict of interest to declare.

Authorship

X.F.H., A.L., and H.M.C. conceived the study. X.F.H. and A.L. performed the literature searches. X.F.H. and A.L. screened articles for inclusion. X.F.H., A.L., and H.M.C. wrote the first draft of the manuscript. All authors revised the manuscript and approved the final version.

Ethics of human subject participation

This study did not require ethical approval, as the data used have been published previously and are, hence, already in the public domain. Consent is not required when conducting a review.