DOI: https://doi.org/10.21203/rs.3.rs-2200006/v1
Lead is a heavy, toxic metal and its exposure to humans can lead to increased risk of cardiovascular disease development and mortality. Lead exposure has been shown to induce hyperhomocysteinemia (HHCy ) which further increases the risk of cardiovascular diseases. We aimed to investigate the mediation effect of blood lead induced HHCy on cardiovascular mortality in a national cohort. A total of 17,915 adults aged ≥ 20 who participated in the National Health and Nutrition Examination Survey (1999 to 2006). Information on mortality was ascertained via probabilistic matching to the death certificates from the National Death Index recorded up to December 31, 2015. Cox proportional hazards regression was performed to assess the association between blood lead levels and mortality. Mediation via HHCy was examined using a logit model. During a mean follow-up of 11.6 years, the incidences of CVD mortality were 0.73, 2.18, 3.03 and 4.94 per 1000 person-years across quarterlies of blood lead levels from low to high. Following multivariable adjustment, blood lead levels were strongly associated with CVD mortality in all mortality models (p trend < 0.001). This association remained statistically significant after further adjusting for quartiles of homocysteine (model 3; HR 1.38 (95% CI 1.01 - 1.89) p trend < 0.001). Furthermore, blood lead levels increased the odds of CVD mortality via homocysteine (indirect effect) (OR 1.42 (95% CI 1.30 - 1.55)), demonstrating the mediatory effect of homocysteine. This the first study that demonstrates that increased homocysteine mediates more than half of CVD mortality related to blood lead levels.
Although blood lead level (BLL) is continually decreasing in the United States [1], there are still large-scale sources of lead that contaminate air, soil and water, contributing to high BLL [2]. Lead is a heavy metal that is toxic to humans. There is a general consensus that BLL of 10 ug/dL for adults and 3.5 ug/dL for children are the reference values to identify high BLL. However, the Advisory Committee on Childhood Lead Poisoning Prevention (ACCLPP) of the Centers for Disease Control and Prevention (CDC) guidelines indicate that any blood lead level is toxic [3]. Lead can inhibit many enzymes and induce oxidative stress by forming complexes with proteins, amino acids, and thiol-containing compounds [4]. It increases reactive oxygen species, damaging various biomolecules like DNA, proteins and enzymes [5]. This causes physiological and biochemical changes affecting the nervous system, reproductive system, hematopoietic system and cardiovascular system.
Cardiovascular diseases (CVDs) are groups of diseases that affect heart and blood vessels. There are four main types of CVD; coronary heart disease, strokes and transient ischaemic attack, peripheral arterial disease and aortic disease [6]. During the last few decades, CVD mortality rates increased in both elderly and younger age groups in the USA population [7]. Among other environmental heavy metals, lead is a high risk factor to develop CVDs[8], [9] and increasing BLLs were linked to high CVD mortality rates [10].
Homocysteine is an amino acid produced during proteolysis and high homocysteine concentration in blood called hyperhomocysteinemia [11]. Lead exposure has been linked to HHcy in blood through its interaction with different proteins and disruption of homocysteine metabolism pathways [12]. Interestingly, products from homocysteine metabolism, such as methionine, reduces the damage caused by reactive oxygen species in cells exposed to lead [13]. Recently, the association between BLL and hyperhomocysteinemia was confirmed in an analysis that included over 9000 participants from the US National Health and Nutrition Examination Survey (NHANES) database [14]. In the literature, there are more than 100 diseases that are linked to HHcy with the most common diseases being CVD and neurological diseases [15]. However, the influence of HHcy on CVD remains unclear [11].
To our knowledge, there are no studies have investigated the mediation effect of BLL induced HHCy on cardiovascular mortality. Therefore, our aim was to explore the homocysteine mediation effect on the association between blood lead levels and cardiovascular mortality using NHANES (1999–2006).
This study used publicly available data from years (1999–2006) of National Health and Nutrition Examination Survey (NHANES). A total of 17,915 adults aged ≥ 20 years were included in the current analysis after excluding those without measures of blood lead and homocysteine. The survey is carried out annually using a complex multistage study design to assess health and nutritional status of the non-institutionalized US general population. Detailed NHANES protocols and procedures are available elsewhere [8]. In brief, a face-to-face interview was conducted, during which participants were asked to complete questionnaires, undergo medical examination and provide venous blood and other biological samples.
Blood lead was measured at the Environmental Health Sciences Laboratory of the CDC National Center for Environmental Health (NCEH). Lead levels in whole blood were measured on an inductively coupled plasma mass spectrometry (PerkinElmer, Norwalk, CT). Homocysteine was measured by a fully automated fluorescence polarization immunoassay (Abbott Diagnostics, Abbott Park, IL). Other data collected include sociodemographic characteristics, including age, sex, race/ethnicity, marital status, education, physical activity, current smoking, alcohol intake and poverty income ratio.
Information on mortality was ascertained via probabilistic matching to the death certificates from the National Death Index recorded up to December 31, 2015 [16]. The cause of death was coded according to the International Classification of Diseases, Tenth Revision (ICD-10). In the current analysis, cardiovascular disease (CVD) mortality was defined based on ICD codes I00–I99, while cancer-mortality was based on ICD codes C00–C97. This method has been validated and widely used [17], [18]
Blood lead level was categorised into quartiles with quartile 1 (Q1) and quartile 4 (Q4) as the lowest and highest, respectively. Sample characteristics was presented as mean (SD) or as percentages and the differences in subject characteristics were tested using one-way analysis of variance (ANOVA) or chi-square tests. Survey weight was used in all the multivariable analyses, except the mediation analysis. Follow-up duration was calculated as the difference between the survey date and the last known date alive or censored from the linked mortality file. Multivariate Cox proportional hazards regression was performed, to assess the association between blood lead levels and mortality. Time on survey was used as the time scale in the Cox regression. Three models were used in Cox regression analyses: model 1 was adjusted for age, gender and race; model 2 was further adjusted for income-poverty-ratio, leisure time physical activity, education, smoking, alcohol drinking and BMI; model 3 was further adjusted for homocysteine. Kaplan–Meier survival analysis was used to determine survival by quartiles of blood lead level. A logit model in a generalization method was used to determine multiple-adjusted odds ratios with 95% confidence intervals (OR [95% CI]) for CVD mortality according to total, direct and indirect (mediation via homocysteine) effects of blood lead levels [19]. All the analyses were performed using STATA (Version 17.0, Stata Corporation, College Station, TX, USA).
Based on the inclusion and exclusion criteria, a total of 17,915 participants who attended NHANES 1999–2006 were included in this study, with a mean age of 49.5 years. 47.6% of participants were men, and 52.4% were women. Non-Hispanic White participants comprised the majority of the sample size with 50.5% of study participants. The unweighted distributions of population characteristics and other covariates are shown in Table 1.
Total | Q1 | Q2 | Q3 | Q4 | p-value | |
---|---|---|---|---|---|---|
N = 17,915 | N = 5,078 | N = 4,297 | N = 4,109 | N = 4,431 | ||
Age (years) | 49.5 (19.0) | 38.4 (16.2) | 48.7 (18.1) | 54.4 (17.7) | 58.4 (17.7) | < 0.001 |
Gender | < 0.001 | |||||
Men | 8,526 (47.6%) | 1,273 (25.1%) | 1,903 (44.3%) | 2,326 (56.6%) | 3,024 (68.2%) | |
Women | 9,389 (52.4%) | 3,805 (74.9%) | 2,394 (55.7%) | 1,783 (43.4%) | 1,407 (31.8%) | |
Race | < 0.001 | |||||
NH White | 9,039 (50.5%) | 2,661 (52.4%) | 2,201 (51.2%) | 2,119 (51.6%) | 2,058 (46.4%) | |
NH Black | 3,510 (19.6%) | 909 (17.9%) | 801 (18.6%) | 775 (18.9%) | 1,025 (23.1%) | |
Mex American | 3,962 (22.1%) | 1,066 (21.0%) | 933 (21.7%) | 898 (21.9%) | 1,065 (24.0%) | |
Other race/ethn | 1,404 ( 7.8%) | 442 ( 8.7%) | 362 ( 8.4%) | 317 ( 7.7%) | 283 ( 6.4%) | |
Education | < 0.001 | |||||
<11 grade | 5,600 (31.3%) | 1,067 (21.0%) | 1,241 (28.9%) | 1,337 (32.6%) | 1,955 (44.3%) | |
HS displ or GED | 4,244 (23.7%) | 1,156 (22.8%) | 999 (23.3%) | 1,059 (25.8%) | 1,030 (23.3%) | |
Some college | 4,687 (26.2%) | 1,660 (32.7%) | 1,163 (27.1%) | 972 (23.7%) | 892 (20.2%) | |
>college | 3,350 (18.7%) | 1,194 (23.5%) | 885 (20.6%) | 735 (17.9%) | 536 (12.1%) | |
Smoking | < 0.001 | |||||
Never | 9,252 (51.7%) | 3,463 (68.2%) | 2,326 (54.2%) | 1,858 (45.3%) | 1,605 (36.3%) | |
Former | 4,728 (26.4%) | 943 (18.6%) | 1,087 (25.3%) | 1,232 (30.0%) | 1,466 (33.2%) | |
Current smoker | 3,911 (21.9%) | 669 (13.2%) | 879 (20.5%) | 1,012 (24.7%) | 1,351 (30.6%) | |
Alcohol drinking | < 0.001 | |||||
No | 3,518 (19.6%) | 862 (17.0%) | 769 (17.9%) | 859 (20.9%) | 1,028 (23.2%) | |
Yes | 10,691 (59.7%) | 2,922 (57.5%) | 2,579 (60.0%) | 2,482 (60.4%) | 2,708 (61.1%) | |
Missing | 3,706 (20.7%) | 1,294 (25.5%) | 949 (22.1%) | 768 (18.7%) | 695 (15.7%) | |
BMI (kg/m2) | 28.4 (6.4) | 29.1 (7.2) | 28.7 (6.6) | 28.4 (5.8) | 27.5 (5.4) | < 0.001 |
Physical activity (METs minutes/week) | < 0.001 | |||||
<600 MET/Wk | 8,729 (48.7%) | 2,234 (44.0%) | 2,044 (47.6%) | 2,007 (48.8%) | 2,444 (55.2%) | |
600–1200 MET/Wk | 2,616 (14.6%) | 843 (16.6%) | 638 (14.8%) | 596 (14.5%) | 539 (12.2%) | |
>=1200 MET/Wk | 6,570 (36.7%) | 2,001 (39.4%) | 1,615 (37.6%) | 1,506 (36.7%) | 1,448 (32.7%) | |
Income to poverty ratio | < 0.001 | |||||
<1.30 | 4,687 (28.3%) | 1,194 (24.8%) | 1,051 (26.3%) | 1,041 (27.6%) | 1,401 (35.0%) | |
1.3–3.5 | 6,471 (39.0%) | 1,851 (38.4%) | 1,480 (37.1%) | 1,474 (39.1%) | 1,666 (41.6%) | |
>3.5 | 5,425 (32.7%) | 1,775 (36.8%) | 1,458 (36.6%) | 1,259 (33.4%) | 933 (23.3%) | |
Log transformed lead (ug/L) | 0.5 (0.7) | -0.3 (0.4) | 0.3 (0.1) | 0.8 (0.1) | 1.4 (0.4) | < 0.001 |
Log transformed homocysteine (umol/L) | 2.1 (0.4) | 1.9 (0.4) | 2.1 (0.3) | 2.2 (0.3) | 2.3 (0.4) | < 0.001 |
Data are presented as mean (SD) for continuous measures, and n (%) for categorical measures. |
Homocysteine levels showed a positive trend with blood lead levels. The mean concentrations of blood lead and homocysteine levels were 0.5 µg/L (SD 0.7) and 2.1 µmol/L (SD 0.4) (Log transformed), respectively. There were significant differences between blood lead levels and homocysteine quartiles. Indeed, participants in the highest quartile had blood lead levels of 1.4 µg/dL (SD 0.4) (Log transformed), and blood homocysteine level of 2.3 µmol/L (SD 0.4) (Log transformed). Furthermore, these participants were older, more likely to be male, have lower educational levels, lower physical activity and higher alcohol consumption.
During a mean follow-up of 11.6 years (208,506 person-years), overall, 3700 deaths occurred. The incidences of all-cause mortality were 4.40, 9.10, 13.62 and 23.89 per 1000 person-years across quarterlies of blood lead levels from low to high (Table 2). Following multivariate-adjusted hazard ratios of mortality models due to all causes, there was a dose-response relationship between blood lead levels and all-cause mortality.
Quartiles of lead | ||||||||
---|---|---|---|---|---|---|---|---|
Q1 | Q2 | Q3 | Q4 | p trend | ||||
All-cause mortality | ||||||||
Number of cases | 387 | 731 | 970 | 1,612 | ||||
Incidence rate (per 1000 person years) | 4.40 | 9.10 | 13.62 | 23.89 | ||||
Model 1 | 1.00 | 1.21 | (1.02–1.44) | 1.32 | (1.13–1.56) | 1.98 | (1.66–2.37) | < 0.001 |
Model 2 | 1.00 | 1.12 | (0.94–1.34) | 1.17 | (0.98–1.39) | 1.60 | (1.33–1.94) | < 0.001 |
Model 3 | 1.00 | 1.07 | (0.89–1.28) | 1.09 | (0.91–1.29) | 1.40 | (1.17–1.68) | < 0.001 |
CVD mortality | ||||||||
Number of cases | 77 | 173 | 239 | 383 | ||||
Incidence rate (per 1000 person years) | 0.73 | 2.18 | 3.03 | 4.94 | ||||
Model 1 | 1.00 | 1.51 | (1.08–2.10) | 1.42 | (1.04–1.94) | 1.86 | (1.38–2.52) | < 0.001 |
Model 2 | 1.00 | 1.41 | (0.99–2.03) | 1.32 | (0.94–1.86) | 1.65 | (1.19–2.28) | 0.006 |
Model 3 | 1.00 | 1.31 | (0.92–1.87) | 1.20 | (0.87–1.66) | 1.38 | (1.01–1.89) | 0.103 |
Model 1 adjusted for age, gender and race | ||||||||
Model 2 further adjusted for income-poverty-ratio, leisure time physical activity, smoking, alcohol drinking and BMI | ||||||||
Model 3 further adjusted for quartiles of homocysteine |
Indeed, blood lead levels were strongly associated with higher all-cause mortality in the age, sex and race-adjusted model (model 1; HR 1.98 (95% CI 1.66–2.37), p trend < 0.001,). The positive relationship remained significant after further adjusting for other demographic characteristics (models 2; HR 1.60 (95% CI 1.33–1.94) p trend < 0.001), furthermore, this positive association remained statistically significant after further adjusting for homocysteine levels (model 3; HR 1.40 (95% CI 1.17–1.68) p trend < 0.001), Table 2. Kaplan-Meier survival analysis, Fig. 1, showed statistically significant lower survival rates for participants who had high blood lead levels (1.4 µg/dL, log transformed).
During a mean follow-up of 11.6 years (208,506 person-years), 872 deaths occurred due to CVD. The incidences of CVD mortality were 0.73, 2.18, 3.03 and 4.94 per 1000 person-years across quarterlies of blood lead levels from low to high, Table 2. After multivariable adjustment, blood lead levels were strongly associated with CVD mortality in all mortality models (p trend < 0.001) in a dose-dependent manner. When adjusted for age, sex and race-adjusted there was a statistically significant association between blood lead levels and CVD mortality (model 1; HR 1.86 (95% CI 1.38–2.52), p trend < 0.001,). This significant association continued after further adjustments for other demographic characteristics (models 2; HR 1.65 (95% CI 1.19–2.28) p trend < 0.001). Importantly, this association remained statistically significant after further adjusting for quartiles of homocysteine (model 3; HR 1.38 (95% CI 1.01–1.89) p trend < 0.001).
Odds ratios of CVD mortality according to total, direct and indirect (mediation of homocysteine) effects of blood lead levels are reported in Table 3. More than half of the increased all-cause mortality related to increased blood lead levels was due to increased levels of homocysteine after adjusting for age, gender and race, income-poverty-ratio, leisure time physical activity, smoking, alcohol drinking and BMI. Similarly, Blood lead levels increased the odds of CVD mortality with more than half of the increase was mediated by homocysteine.
Quartiles of blood lead | |||||||
---|---|---|---|---|---|---|---|
Q1 | Q2 | Q3 | Q4 | ||||
OR | OR | 95%CI | OR | 95%CI | OR | 95%CI | |
All-cause mortality | |||||||
Total | Ref | 1.49 | 1.23–1.79 | 1.73 | 1.46–2.04 | 2.73 | 2.38–3.13 |
Indirect | Ref | 1.28 | 1.24–1.32 | 1.44 | 1.38–1.51 | 1.67 | 1.57–1.77 |
Direct | Ref | 1.16 | 0.96–1.39 | 1.20 | 1.01–1.41 | 1.63 | 1.43–1.87 |
Percent contribution of indirection effect to total effect | 62.67** | 67.27** | 51.06** | ||||
CVD mortality | |||||||
Total | Ref | 1.38 | 1.02–1.87 | 1.60 | 1.15–2.23 | 1.96 | 1.39–2.75 |
Indirect | Ref | 1.18 | 1.13–1.24 | 1.29 | 1.21–1.36 | 1.42 | 1.30–1.55 |
Direct | Ref | 1.17 | 0.86–1.58 | 1.25 | 0.89–1.75 | 1.38 | 0.98–1.94 |
Percent contribution of indirection effect to total effect | 52.31 | 53.09 | 51.81 | ||||
Model 1 adjusted for age, gender and race, income-poverty-ratio, leisure time physical activity, smoking, alcohol drinking and BMI. |
This prospective cohort study used NHANES data (1999–2006) to investigate homocysteine mediation effect on the association between blood lead levels and cardiovascular mortality. Our findings demonstrated evidence that blood homocysteine levels mediate more than half of the association between blood lead levels and CVD mortality. To our knowledge this is the first study demonstrating homocysteine mediation of the association between blood lead levels and CVD mortality.
During a mean 11.6 years follow-up, blood lead levels was positively associated with CVD and all-cause mortality. Indeed, our analysis revealed a dose-dependent increase in mortality rates with blood lead levels with highest quartile of blood lead was associated with an incidence rate of 23.89 per 1000 years. Furthermore, blood lead levels were positively associated with increased homocysteine levels after adjusting for other covariates. Moreover, our date demonstrates a positive association of blood homocysteine with quartiles of blood lead levels, which may suggest that blood lead could be another contributing factor to increased blood homocysteine levels, furthermore, we found that older non-Hispanic White male participants were found to have the highest blood lead and homocysteine levels.
Our finding that blood lead was associated with increased homocysteine has been demonstrated by several studies [14], [20], [21]. In a study by Schafer et al. [12] reported a positive association between blood lead and homocysteine in a sample of 1,140 US adults. In another prospective cohort study involving 2280 American men demonstrated blood lead levels were positively associated with homocysteine levels [21]. Several other studies that investigated the association between blood lead level and homocysteine concluded a dose-response relationship exists between blood lead levels and homocysteine [22]. In a recent study by Li et al. suggesting that higher level of blood lead was associated with increased risk of hyperhomocysteinemia [14]. Therefore, as demonstrated by our findings and others it is clear that blood lead increases blood homocysteine.
The mechanism of the positive association between blood lead level and homocysteine concentration has not yet been fully elucidated. However, several possible mechanisms have been suggested that may explain this association [12]. One such suggestion, is the interaction of lead with sulfhydryl group containing proteins. It has been shown that lead inhibits δ-aminolevulinic acid dehydratase (ALAD), an enzyme required for heme synthesis, thereby affecting cystathionine β-synthase function, which is an enzyme required to catalyze the first step of the transsulfuration pathway, which catalyzes the condensation of homocysteine to form cystathionine, this leads to accumulation of homocysteine [12]. Another mechanism that has been suggested is the direct inhibition of homocysteine metabolism, homocysteine contains a sulfhydryl group, lead has an affinity for sulfhydryl group, leading to inhibition of homocysteine and its accumulation in the blood [22].
CVD is one of the leading causes of mortality worldwide. In the U.S., deaths due to CVD and other related vascular diseases were the leading cause of death, and it is estimated that 2030, 43.9% of the US population will have some form of CVD (Benjamin et al., 2017). Environmental factors such as exposure to lead has been associated with the development of CVD[24] Our data shows after adjusting for age, sex and other demographic characteristics, blood level levels of 4.1 µg/dL were strongly associated with CVD mortality (HR 1.65 (95% CI 1.19–2.28), p trend < 0.001,). Our data supports previous study by Lanphear et al. reporting blood lead levels lower than 5 µg/dL were associated with increased risk of mortality [25].
Several other studies have reported a positive association between blood lead and all-cause, CVD and cancer mortality at lower blood lead levels [26], [27]. Previous studies using NHANES data demonstrated that blood lead levels were associated with higher all-cause and CVD mortality [27]. A recent study by Aoki et al. analyzing 1999 to 2010 NHANES data reported a linear association between blood lead levels and increased CVD mortality [28]. Therefore, as demonstrated by our study and previous reports of increased mortality rates from cardiac events in individuals with blood lead levels lower than 5 mg/dl, suggests no blood lead levels could be considered as safe.
Homocysteine has been reported as an independent risk factor for atherosclerosis development[29], leading to an increased risk of CVDs including myocardial infarction, stroke, and peripheral vascular disease. Elevated homocysteine levels may induce vascular damage via several mechanisms including endothelial dysfunction and damage to vascular smooth muscle [30]. Some proposed mechanisms for homocysteine-induced damage included increased oxidative stress, inhibition of nitric oxide synthesis and proliferation of vascular smooth muscle cells [31]. Previous studies have demonstrated a positive association between serum homocysteine and increased arterial stiffness [32], [33]. Our study presents the first report demonstrating homocysteine mediation effect on the association between blood lead levels and cardiovascular mortality in a large representative sample of U.S population. After adjusting for age, gender and race, income-poverty-ratio, leisure time physical activity, smoking, alcohol drinking and BMI, blood lead levels increased the odds of CVD mortality via homocysteine (indirect effect) (OR 1.42 (95% CI 1.30–1.55)).
The strengths of the study include the large sample size and relatively long follow-up duration as well as the representative US sample. There were few limitations associated with our study. Although it is established that blood lead levels increase homocysteine, in this study we didn’t seek to establish this. Therefore, bias due to unmeasured confounding factors may still be present. Another potential limitation is the presence of genetic abnormalities that are known to increase homocysteine levels, including CBS and MTHFR mutations.
More than half of the increased CVD mortality related to blood lead was mediated by increased homocysteine levels in the US population.
Author contribution
All authors contributed to the study. Sapha Shibeeb: Conceptualization of the study, data analysis, writing—review and editing. Atiyeh Abdallah: Conceptualization of the study, writing (review and editing). Zumin Shi: Conceptualization of the study, methodology, writing—review and editing, data analysis. The authors have read and approved the final draft of the manuscript.
Funding The authors received no funding support to publish this data
Data availability Data will be made available on request
Consent for publication All the authors approved this submission.
Conflict of interest The authors declare no competing interests