Date collection
This was a cross-sectional study based on follow-up of the manganese-exposed workers healthy cohort (MEWHC) in 2017, detailed description of the inclusion criteria and exclusion criteria for this cohort can be found in the previous study [22, 23]. Face-to-face interviews were conducted by trained staff, and we used standardized and structured questionnaires to collect the basic information, including demographic information, smoking, drinking status, high-fat diets frequency, medicine intake in the past two weeks, and self-report chronic diseases. We measured height and weight, waist circumference, and blood pressure using standard methods. Body mass index (BMI, kg/m2) was calculated from the height and weight measurements. The history of hypertension was defined according to the subjects' self-reported history and the results of blood pressure measured in the field, and the standard methods and definition of the Chinese guidelines for the prevention and treatment of hypertension to define hypertension (revised in 2018) was used[24]. We conducted a high animal fat diets frequency questionnaire to assess high fat diets intake. We divided the weekly high animal fat diets frequency into two groups, low high-fat diets frequency group was defined as the high-fat diets frequency less than 3 times per week, and high high-fat diets frequency group was defined as the high-fat diets frequency 3 times or more per week. Current smoking status was defined as smoking at least one cigarette per day for 6 consecutive months. Current drinker was defined as drinking more than 5 ml at a time and drinking at least 3 times a week for 6 months. Participants were classified as nonsmoker, former smoker or current smoker. A pack-year was defined as 20 cigarettes smoked every day for one year [25]. We further categorized participants' smoking status into three groups on the basis of median pack-years: nonsmoker, <18 pack-years, and ≥18 pack-years.
Our study subjects were recruited from the follow-up of manganese-exposed workers healthy cohort (ME WHC) in 2017. According to the exclusion criteria, subjects with cancer, or coronary heart disease, or stroke disease, or diabetes were excluded from the 828 workers, as well as those whose biological samples were undetected. Finally, a total of twenty-two subjects were excluded from our study. In the end, a total of 803 occupationally Mn-exposed workers were invited to participate in our study. All biological samples include blood and urine samples were collected after a night of fasting, and immediately were transported to the biobank and stored deep frozen at -80°C until be analyzed.
Measurement of Mn concentration in the respirable dust
We collected a total of 134 samples including 20 categories of relatively typical occupation by personal sampling. We selected three subjects in each occupation and measured their respiratory dust for 3 consecutive days. We used the air sampling pump (SENSIDYNE, GilAirPlus, US) and respirable dust sampling head (HXFC-WH, Wuhan, China), and glass fiber microporous filter membrane (diameter 0.22 mm, 37 mm diameter) to collect the respirable dust of the workers from the breathing zone of participants. Gas flow rates were set to 2 L/min and flow rates checked both before and after the sampling, the average time we collected 6.9h of sample per person per day. We tested only representative types of positions in the main workplace, individual exposure concentration of other positions was replaced by detected concentration with the same or similar workplace. We strictly abided by the national occupational health standard of the People's Republic of China. The standards are “Determination of airborne dust in the workplace part 2: Concentration of respirable dust” (GBZ/T 192.2-2007), and “Specifications of air sampling for hazardous substances monitoring in the workplace” (GBZ159-2004). The determine procedure followed the standard “Ambient air and stationary source emission-Determination of metals in ambient particulate matter-Inductively Coupled Plasma Mass Spectrometry (ICP-MS)” (HJ657-2013).
Digestion of filter membrane: the filter membrane was placed in a Teflon tube and added with 10 ml aqua regia. Microwave digestion instrument was used for digestion at 200 degrees Celsius for 30 minutes, diluted with ultrapure water to 50 ml and measured by ICP-MS (Perkin Elmer, NexION 2000, USA). We used the isotope 45Sc as the internal standard for the determination of Mn. Quality control (QC) measures included analysis of the initial calibration validation criteria, a repeat measurement for every 10 samples, and SRM1640A (natural trace element concentration in water). Mn metal had the limit of detection (LOD) of 0.076 μg/L. All samples concentration was higher than the LOD. Finally, we calculated the time-weighted average of Mn (as MnO2) (Mn-TWA) based on the three day average concentration. According to the National occupational health standards of the People's Republic of China, the occupational exposure limits permissible concentration-time weighted average(PC-TWA)for Mn and its compounds were set at 0.15mg/m3 (as MnO2) based on an 8-hour time weighting.(GBZ 2.1-2007). Therefore, we divided all subjects into two groups, low-exposure group was defined as Mn-TWA ≤0.15 mg/m3 and high-exposure was defined as Mn-TWA >0.15 mg/m3. The high-exposure group mainly included smelters, welders, human crushing workers, casting crane workers, material crane workers. The low-exposure group mainly included circulating cooling water system operator, chemical analysts, office workers, security guards and workers in other auxiliary positions.
Measurement of Serum lipids
Serum samples were stored in a deep freeze -80 °C until be tested. Serum lipids were measured by using chemistry automatic analyzer (Hitachi 7600-020, Kyoto, Japan) at the testing center of the Department of Clinical Laboratory at the First Affiliated Hospital of Guangxi Medical University in Nanning. Serum lipids include Total Cholesterol (T-CHO), Triglyceride (TG), Low-density lipoprotein cholesterol (LDL-C), and High-density lipoprotein cholesterol (HDL-C) [26]. Serum lipids were classified according to The 2016 Chinese Guideline for the Management of dyslipidemia in Adults (Chinese guideline). High LDL-C was defined as Low-density lipoprotein cholesterol ≥4.14 mmol/L. high TG was defined as triglyceride ≥2.26 mmol/L, high T-CHO was defined as total cholesterol ≥6.22 mmol/L, and low HDL-C was defined as High-density lipoprotein cholesterol <1.04 mmol/L. Dyslipidemia was well-established with strong risk factors of cardiovascular diseases, and cardiovascular diseases are sequelae of dyslipidemia[27]. The dyslipidemia guideline suggested that a target of LDL-C should be set according to individual ASCVD risk, and the individual ASCVD risk was evaluated adjusted age, gender, BMI, history of hypertension, and smoking status[26].
Statistical analysis
Mann-Whitney U test was used to compare the lipids levels of low Mn exposure and high Mn exposure group. Logistic regression models was used to estimate Mn exposure levels and the adjusted odds ratio (ORs), 95% confidence interval (CIs) of no achieving LDL-lowering targets risk, high LDL-C risk, high TG risk, high T-CHO risk, and low HDL-C risk, respectively. Specifically, because of the high inter variable correlation between age and seniority, the only seniority was adjusted in models. Finally, the potential confounders for models adjustment included gender, seniority, BMI, hypertension, medicine intake in the past two weeks, high-fat diets frequency, smoking status, and drinking status.
We conducted a stratified analysis according to strata of gender, seniority, BMI, hypertension, medicine intake in the past two weeks, high-fat diets frequency, smoking status, and drinking status. In our research, we considered the biologic interaction in potential risk factors. Rothman has argued that we should focus on epidemiological interaction or interaction on an additive scale by testing whether the joint effect from exposure to both factors was greater than the sum of their independent effects, and suggested the use of relative excess risk for interaction (RERI) in assessing additive interaction. The previous reports gave RERI a detailed explanation and calculation method [28-30]. Display quotations of over 40 words, or as needed. In short, RERI was calculated as follows:
RERI= e (β1+ β2+β3) – e (β1) – e (β2) +1
Where β1 is the coefficient of the effect of Mn-TWA levels measure, β2 is the coefficient of potential risk factor and β3 is the coefficient of the cross-product of the Mn-TWA levels and potential risk factor. And we estimated 95% confidence intervals by using the standard delta method. When the relative excess risk for interaction is an estimate of more than zero, there is an additive scale interaction between the two risk factors, and the 95% confidence interval is positive and does not contain zero. RERI of 0 indicates exact additivity and there is no additive scale interaction. Finally, we assessed the interaction between Mn-TWA levels and each of the potential interacting factors of smoking effects (both smoking status and pack-years), drinking status, high-fat food frequency, and BMI on high TG risk in the logistic regression models by using R software. Analyses were conducted using R (version 3.4) and SPSS (19.0), two-sided, p <0.05 was considered statistically significant.