Association between heavy metal exposure and bacterial vaginosis: A cross-sectional study.

According to BV, Table 1 lists the descriptive features of the participants. Participants with higher BMI, serum lead, serum cadmium, total cholesterol, uric acid, and total calcium levels, lower PIR and HDL-cholesterol levels, belonging to non-Hispanic black and other racial backgrounds, having received education below a high school level, lacking vigorous and moderate physical activity in the past 30 days, and being unmarried, were found to have a higher likelihood of testing positive for BV compared to those who tested negative. Significant differences were not discovered in age, serum total mercury levels, and household size (p >0.05). BMI was calculated as the body weight in kilograms divided by the square of the height in meters. P value: if it is a continuous variable, it is obtained by the Kruskal Wallis rank sum test. If the theoretical number of counting variables is less than 10, it is received by the Fisher exact probability test. % for categorical variables: P value was calculated by chi-square test. Abbreviations: BMI, Body Mass Index; PIR, Poverty Income Ratio; HDL, High-Density Lipoprotein; GED, General educational development.; OR odds ratio; CI, confidence interval. The association between the collective and the three distinct categories of serum lead, total mercury, and cadmium concentrations and BV is presented in Table 2 . There was a notable positive association between the overall concentration of lead in the serum and the likelihood of developing BV, resulting in an 11% increase in the danger of BV [OR = 1.11 (1.02,1.19), p = 0.012]. The group with the highest lead levels in the serum exhibited the highest incidence of developing BV in comparison to the remaining two lower groups, exhibiting a statistically significant 35% rise in the possibility of disease development, as opposed to the group with low levels [OR = 1.35 (1.06,1.72), p = 0.016]. In all three models, a positive correlation between serum lead levels from 1.3 to 23.9 μg/dl and the occurrence of BV was significant and stable. The correlation between serum lead levels and the occurrence of BV was intuitively described using smoothing curves, as depicted in Fig 2 (p = 0.020). The association between serum lead and the risk of BV was exhibited by a smooth curve fitting model. The central red dots represent serum lead concentrations, with each point contributing to a continuous fitted curve. The region between the two blue dashed lines corresponds to the 95% confidence interval. The X-axis is serum lead levels (continuous variable), and the Y-axis is odds ratios (ORs). ORs were computed from the Model II in a multivariate logistic regression analysis. Unadjusted model: no covariates were adjusted. Model I: Part of covariates, age, race, education, and BMI were adjusted. Model II: All covariates, age, race, education, BMI, moderate activity over the past 30 days, PIR, marital status, HDL-cholesterol, total cholesterol, total number of people in the household, total calcium, uric acid, and vigorous activity over the past 30 days were adjusted. Abbreviations: BMI, Body Mass Index; PIR, Poverty Income Ratio; HDL, High-Density Lipoprotein; CI, confidence interval; OR, odds ratio. The study found that there was a passive association between serum total mercury levels ranging from 0.5 to 1.1 μg/l and a 2% decrease in the risk of BV, and the levels from 1.2 to 27.4 μg/l were negatively correlated with a 4% decrease in the risk of BV in model II. Nevertheless, the research findings did not demonstrate statistical significance(p>0.05). There was no statistically significant association found between overall serum total mercury levels and the risk of BV (p>0.05). There existed a clear correlation between the collective concentration of cadmium in the serum and the likelihood of developing BV, as indicated by a 48% rise in the risk of BV when adjusted for all covariates in the model [OR = 1.48 (1.25,1.76), p<0.001]. The serum cadmium levels ranging from 0.5 to 7.4 μg/l exhibited a 41% rise in the susceptibility to BV in model II, indicating a direct correlation with the risk of BV and statistical significance [OR = 1.41 (1.01,1.98), p = 0.047]. The positive relationship between serum cadmium and the risk of BV was steady in all three models. Fig 3 demonstrates the correlation between serum cadmium and the risk of BV intuitively through smoothing curves(p<0.001). The association between serum cadmium and the risk of BV was exhibited by a smooth curve-fitting model. The central red dots represent serum cadmium concentrations. The region between the two blue dashed lines corresponds to the 95% confidence interval. The X-axis is serum cadmium levels (continuous variable), and the Y-axis is odds ratios (ORs). ORs were computed from the Model II in a multivariate logistic regression analysis. Table 3 describes the stratification analysis for age, race, education, marital status, BMI, HDL-cholesterol, total calcium, and uric acid between serum lead, total mercury, and cadmium and the risk of BV. In the case of serum lead and cadmium, the results of our stratified analysis reflected that the direction of the result was mostly consistent (OR>1). For Pb, the risk further increases in females who were 37 to 49 years old, accepted education less than high school, married and separated, BMI from 29.91 to 65.41 kg/m^2, HDL-cholesterol from 62 to 160 mg/dl, total calcium 9.2 to 9.4 mg/dl, and uric acid from 3.8 to 8.1 mg/dl. In addition, for Cd, the risk further increased in females who were 18 to 24 years old and 37 to 49 years old, non-Hispanic white and non-Hispanic black, accepted education less than high school, and high school diploma (including GED), widowed and divorced, never married, BMI from 18.51 to 65.41 kg/m^2, HDL-cholesterol from 48 to 160 mg/dl, total calcium from 8.1 to 9.1 mg/dl and 9.5 to 11.3 mg/ml, and Uric acid from 3.8 to 4.6 mg/dl. The logistic regression models constructed for subgroup analyses were adjusted for all covariates: age(excluded in analysis stratified by age), race(excluded in analysis stratified by race), education(excluded in analysis stratified by education status), BMI(excluded in analysis stratified by BMI), moderate activity over the past 30 days, PIR, marital status(excluded in analysis stratified by marital status), HDL-cholesterol (excluded in analysis stratified by HDL-cholesterol), total cholesterol, total number of people in the household, total calcium(excluded in analysis stratified by total calcium), uric acid(excluded in analysis stratified by uric acid) and vigorous activity over the past 30 days. Abbreviations: BMI, Body Mass Index; PIR, Poverty Income Ratio; HDL, High-Density Lipoprotein; GED, General educational development CI, confidence interval; OR, odds ratio. Females aged 25 to 36 exhibited statistically significant findings regarding the correlation between serum total mercury and the risk of BV (OR: 0.86, 95% CI: 0.76–0.98). However, the other result of the study was unstable and insignificant, and stratified analysis showed that OR fluctuations were more pronounced in different subgroups. The correlation between serum total mercury and the risk of BV among females aged between 25 and 36 is shown in Fig 4 (p = 0.104). The association between serum total mercury and the risk of BV among females aged between 25 and 36 was exhibited by a smooth curve fitting model. The central red dots represent serum total mercury concentrations, with each point contributing to a continuous fitted curve. The region between the two blue dashed lines corresponds to the 95% confidence interval. The X-axis is serum total mercury levels (continuous variable), and the Y-axis is odds ratios (ORs). ORs were computed from the Model II in a multivariate logistic regression analysis. Sensitivity analysis was conducted when the missing covariates data had been filled in the blank by multiple interpolations. The orientation of the sensitivity analysis results ( S1 Table ) is also generally coherent with the former results.

Bacterial vaginosis (BV), which is a frequent source of vaginal issues in women of reproductive age, can be either symptomatic or asymptomatic [ 1 ]. Roughly half of female individuals experience symptoms such as vaginal malodor, discharge, itching, and an increase in vaginal pH [ 2 ]. A systematic review conducted in 2013 revealed that the prevalence of BV exhibits considerable variation between and within countries worldwide: African-American (51%), Hispanic (32%), and white (23%) [ 3 ]. A study based on the NHANES database showed that the prevalence of bacterial vaginitis in American women between the ages of 14 and 49 years was approximately 29% from 2001 to 2004 [ 4 ]. BV can also lead to complications such as a heightened susceptibility to sexually transmitted diseases, encompassing HIV, pelvic inflammatory disease, preterm birth, and postoperative infections [ 5 – 12 ]. Moreover, after existing treatments, BV still has a high recurrence rate [ 13 , 14 ]. Globally, pollutants such as heavy metals exist in air, water, food, and soil, which explains their widespread exposure to humans [ 15 ]. It has been shown in the literature that heavy metal exposure causes great harm to human immunity, endocrine function, microbial communities, and other functions [ 16 – 22 ]. Empirical evidence suggests that exposure to heavy metals, specifically cadmium (Cd) and lead (Pb), may have detrimental effects on immune function, potentially leading to an increased susceptibility to infections [ 23 – 26 ]. A cross-sectional analysis conducted among non-smokers demonstrated that aside from causing harm to humans, heavy metals are similar to antibiotics and can also lead to an imbalance in the microbiota, which can negatively impact our bodies [ 27 ]. Evidence suggests that heavy metals like Pb and Cd can disrupt the equilibrium of human reproductive hormones, which will contribute to the reduction of fertility or increase the risk of endocrine-secreting cancers or several side effects [ 28 – 30 ]. Currently, it has been determined that BV is a condition where dysbiosis is present in the typical bacterial population found in the vagina [ 31 ]. The abnormal operation of vaginal mucosal immune [ 32 ] and hormonal milieu [ 33 ] in women take part in the vaginal microbiota dysbiosis that participates in the occurrence of BV. Serum exposure to heavy metals can indirectly affect the immune and endocrine function of the human body by affecting the microbiome [ 34 , 35 ] or directly affecting the immune and endocrine function of women, thereby increasing the risk of BV in theory [ 36 , 37 ]. Nonetheless, more empirical research is needed to examine the direct correlation between serum heavy metal exposure and BV. Consequently, the primary intention of this paper is to explore the correlation of serum heavy metal exposure with the danger of BV by utilizing data from the National Health and Nutrition Examination Survey (NHANES) from 2001 to 2004 among U.S. females. In the meantime, to reduce bias and make the results more scientifically robust, the potential confounders based on their associations with the outcomes of interest or a change in effect estimate of more than 10% were selected in the analysis. Covariates were included as potential confounders in the final analysis models if they changed the estimates of heavy metals on BV by more than 10% or were significantly associated with BV [ 38 , 39 ].

The cross-sectional study unveiled a substantial correlation between serum Pb and Cd with elevated susceptibility to BV. At the same time, Hg was identified as not significantly related to the risk of BV among females. According to the research, while there are more than a couple of heavy metals, we suggest decreasing environmental and occupational exposure to heavy metals as much as possible.

The authors bear responsibility for all facets of the work, making sure that any doubts regarding the precision or comprehensiveness of any research element are properly examined and answered. The Institutional Research Ethics Review Board of the National Center for Health Statistics of the Centers for Disease Control and Prevention authorized each protocol for the NHANES. The study has been approved by the Ethics Review Board of the NCHS. Written informed permission was acquired by each subject. The website https://www.cdc.gov/nchs/nhanes/about_nhanes.htm has all further information. The source of our data was selected from the NHANES database ( https://www.cdc.gov/nchs/nhanes/about_nhanes.htm ), a study that focused on the health topics of the general populations of all ages in the United States. The NHANES is an investigation programme designed to assess the health and nutritional status of adults and children in the United States of America. The survey analyses a nationally representative sample of approximately 5,000 individuals on an annual basis. The sample for the survey is selected in such a way as to be representative of the population of the United States, encompassing all age groups ( https://www.cdc.gov/nchs/nhanes/about_nhanes.htm ). The study was approved by the NCHS Ethics Review Board. The flowchart outlining the exclusion criteria is shown in Fig 1 . The data of this study were obtained from two cycles of NHANES, 2001–2002 and 2003–2004. Our research examined a total of 21,161 individuals who had the potential to be suitable participants and had completed both the interview and the Mobile Examination Centers (MECs) examination. Furthermore, we excluded individuals with incomplete information regarding BV (n = 18,355) and serum heavy metal levels (n = 99). Following the exclusion of populations lacking crucial covariate data which included age, body mass index (BMI), Poverty Income Ratio (PIR), HDL cholesterol (High-Density Lipoprotein), and so on, a total of 2,493 18 years to 49 years female participants were ultimately enrolled in our research. The population analyzed corresponds to 92% of the eligible population. These data can be accessed for free on the Internet through the NHANES data files. It was easy for us to find the measurement procedures and diagnosis of BV in the NHANES documentation [ 40 – 42 ]. The MECs of NHANES collected samples of vaginal swabs from adult female participants aged 18–49. After being obtained, these swabs were placed onto pH paper, ensuring that any collected samples were effectively preserved. Once the samples were secured onto the pH paper, they were attached to slides using delicate yet precise techniques. Subsequently, these slides were subjected to Gram staining. Bacterial vaginitis was defined strictly by determining the Nugent score, which was most often used in assessing the situation of Gram-stained vaginal secretions. Nugent’s approach investigated the decrease in lactobacilli and gram-positive rods, which are considered part of the normal flora, as well as the decline in gram-variable morphotypes and the presence of gram-negative bacteria associated with BV [ 43 ]. The scores assigned as 0–6 were regarded as within the normal range, whereas a score of 7–10 indicated a positive outcome that suffered from BV. According to the NHANES website, we collected serum heavy metals data, including total mercury, cadmium, and lead. The website comprehensively explains the laboratory methods, elaborating on the specific techniques and procedures used for conducting experiments. Serum total mercury, lead, and cadmium were measured by the Centers for Disease Control and Prevention National Center for Environmental Health. The quantification measure of Pd and Cd were detected with a PerkinElmer SIMAA 6000 simultaneous multi-element atomic absorption spectrometer with Zeeman background correction and Hg was determined by flow injection cold vapor atomic absorption analysis with online microwave digestion in 2001–2002. Inductively coupled plasma‒mass spectrometry (ICP‒MS) was utilized to ascertain the concentrations of Hg, Pb, and Cd in the entire blood in 2003–2004. Results below the lower limit of detection are substituted with a value equal to the limit of detection divided by the square root of two. To reduce the bias of the outcome and the occurrence of false-positive and ensure the authenticity and reliability of the findings in the study [ 39 , 44 , 45 ], we selected the confounders based on their associations with the outcomes of interest or a change in effect estimate of more than 10%. Covariates were included as potential confounders in the final models if they changed the estimates of heavy metals on BV by more than 10% or were significantly associated with BV [ 38 , 39 ]. Considering the two criteria, participants’ age, BMI, PIR, HDL cholesterol [ 46 ], total cholesterol [ 47 , 48 ], uric acid [ 49 ], total calcium [ 48 ], race, education, marital status [ 50 , 51 ], total number of people in the household, vigorous activity over the past 30 days and moderate exercise over the past 30 days [ 52 ] were selected as our study’s covariates. In order to determine an individual’s BMI, it is imperative to divide their weight, which is quantified in kilograms, by the square of their height, measured in meters. The study classified participants into three racial groups: non-Hispanic Black, non-Hispanic White, and individuals of other ethnic. The marital status of the respondents was categorized as married, divorced, widowed, separated, never married, or living with a partner. Education level was categorized into three groups: individuals with less than high school, high school diploma, and education beyond high school. Results for HDL cholesterol, total cholesterol, uric acid, and complete calcium were extracted from the NHANES laboratory database. EmpowerStats software ( https://www.empowerstats.net/cn/ ) and the statistical package R ( www.r-project.org ) were utilized to establish all analyses. With the aim of ascertaining the statistical significance of the investigation, a threshold of p<0.05 was considered. Descriptive analysis was conducted on continuous variables through mean, standard deviation, and categorical variables through weighted percentages (%). The relationship between clinical outcome and exposure was reflected by the odds ratio (OR) value [ 53 ]. In order to obtain a more precise assessment of the outcome, a 95% confidence interval was computed. The correlation between serum heavy metals and BV was examined using a multivariable logistic regression analysis. We constructed three models, and the details are as follows. The unadjusted model did not incorporate any modifications to the variables. Demographic variables were considered to adjust in model I, which included age, race, BMI, and education level. Model II underwent a comprehensive adjustment process, encompassing all covariates that were included. Meanwhile, tertiles of serum heavy metals levels were established in accordance with the distribution of the investigated population. The potential nonlinearity was resolved by employing three smooth curve fittings. Aiming to alleviate potential bias in the study, stratified analyses were generated with the use of covariates. Stratified analyses were carried out to account for the confounding factors of age, race, education, marital status [ 54 ], BMI [ 55 ], HDL -cholesterol, total calcium, and uric acid. In accordance with the WHO classification of obesity status, the study population was divided into 6 groups on the basis of BMI [ 56 , 57 ]: <18.5(underweight), 18.5–24.9(normal), 25.0–29.9(overweight), ≥30.0(obesity). Additionally, the research categorized marital status as married, widowed and divorced, separated, never married, and living with partner. Sensitivity analysis was performed after the missing data of covariates were populated by multiple interpolations when the missing data of heavy metals and BV were eliminated.

(DOCX)