Three large construction companies in Norway, engaged in outdoor rock drilling, were invited to participate in this cross-sectional study, and they all agreed. At the time of the study, the companies employed in total 140 male rock drillers, defined as working in the immediate vicinity of a drilling rig. These subjects were invited to participate. Four subjects declined participation, and 13 were no longer exposed at the time of the examinations and were therefore excluded. Thus, the study comprises 123 currently CS exposed subjects. Referents (N = 48), that according to our knowledge, had never been occupationally exposed to particulate matter (PM) or CS, were recruited among administrative personnel, typically foremen and other non-exposed heavy construction workers. They were working at the same construction sites as the rock drillers. None of the potential referents declined to participate.

Details on job descriptions and exposure have been published (Ulvestad et al. 2020). In short, rock drillers operate different types of moveable drilling equipment at work sites scattered throughout the country. Blasting leaders and guardrail installers work in the vicinity of the drilling operation. Six different job categories could be defined, depending on how the jobs were performed. The airborne exposure of everyone was closely connected to the job categories. This information was used for the calculation of cumulative exposure estimates.

Medical examinations, including blood sampling, and recording of background data were carried out at the regional offices of the companies immediately after the end of a four days’ work week. Height and weight were measured for the calculation of body mass index (BMI). Previous and current jobs, occupational dust exposure, and the type of drilling equipment the rock drillers had used during their work career, were recorded. Smoking habits and medical history were also recorded. Spirometry, without the use of bronchodilators, was carried out as previously described (Ulvestad et al. 2020). Finally, venous blood samples were collected for the assessment of biomarkers.

Venous blood samples (10 mL) were collected in vacuum tubes without additives (Becton Dickinson and Company, Franklin Lakes, NJ, USA) and coagulated for 45 min before centrifugation (2000 g for 15 min). Serum was aliquoted into 4.0 mL NUNC® polypropylene cryotubes (Thermo Fisher Scientific, Waltham, MA, USA), and thereafter frozen before long-term storage at -80 °C at the National Institute of Occupational Health (Oslo, Norway).

Protein levels of Chemokine ligand 17 (CCL17), CCL18, CCL22, matrix metalloproteinase (MMP) 9, Pentraxin 3 (ptx3) and TGF-β were measured using DuoSet ELISA kit obtained from R&D systems (Stillwater, MN, USA). MMP1 was measured using ProcartaPlex™ Multiplex Immunoassay (Affymetrix, eBiosciences, Japan) and MMP-12 using Luminex magnetic beads analysed on a BIO-PLEX (Biorad, Ca, USA). V-PLEX Proinflammatory Panel 1 Human Kit was used for quantitative determination of interferon gamma (IFN-γ), IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and TNF (Meso Scale Diagnostics, USA). The expression of IL-1 β, IL-2, IL-4, 12p70 and IL-13 were low, and some of the values were below lower limit of detection (LOD). All analysis were performed according to manufactures instruction. A large proportion of IL-1β (N = 131/171) and IL-12p70 (N = 83/171) measurements were below LOD and were therefore not considered in the statistical analysis.

High sensitivity CRP in serum was determined at a clinical commercial laboratory (Unilabs, Oslo, Norway) using CardioPhase TM, Advia 2400 (Siemens Healthcare Diagnostics Inc., Tarrytown, NY, USA) with a method’s LOD of 0.1 mg/L.

Assessment of exposure to CS has been described previously (Ulvestad et al. 2020). Shortly, 23 selected construction sites of the three participating companies were surveyed between 2015 and 2018. Each job task was assessed by personal air sampling. Altogether, 98 and 93 air samples from the respirable and thoracic aerosol fraction, respectively, were collected, and the PM mass determined. Respirable α-quartz was measured by X-ray diffraction spectrometry in 67 air samples according to the silver filter NIOSH Method 7500 (NIOSH, 2003) with a Malvern Panalytical X’Pert3Powder diffractometer, equipped with a PIXcel1D detector and an Empyrean X-ray tube (Malvern Panalytical B.V., Eindhoven, Netherlands). Procedures for the collection of air samples and determination of PM and α-quarz have been described (Ulvestad et al. 2020).

The concentrations measured in the air samples were used to calculate the mean exposure level of each of the six defined job categories. The information collected from the occupational history interview was used together with the arithmetic mean (AM) exposure for each year in each drilling job to calculate individual cumulative exposure to respirable crystalline silica by adding up each year of exposure. Further details on the calculations of cumulative exposure estimates have been published (Ulvestad et al. 2020).

The distributions of the variables were examined visually, Q-Q-plots were assessed, and skewness of the distributions were calculated. Data were log-transformed when the skewness exceeded 2.0. Geometric means (GM) are presented for those variables, while arithmetic means (AM) are presented otherwise. Minimum and maximum values (Min-Max) as well as 95% confidence intervals (95% CI) are also presented. Differences between two groups were assessed with the Studentʹs t-test, while analysis of variance (ANOVA) was used when more than two groups were compared. Chi-square test was used to assess differences between groups with respect to categorical variables, but Fisherʹs exact probability test was applied when expected figures were small (< 5). Associations between concentrations of biomarkers and several independent variables simultaneously were calculated with multiple linear regression analysis (backwards procedure). Independent variables included in the models were ever being exposed to crystalline silica (1/0), being a current smoker (1/0), age, BMI (kg/m2), current self-reported infection (1/0), self-reported use of asthma medication (1/0), and the time of blood sampling (in minutes after midnight). If ever being exposed to CS was associated with any of the biomarker concentrations, the variables cumulative exposure and current exposure were added separately to the model instead of ever being exposed. The multiple linear regression analysis was then performed among exposed rock drillers only. The rock drillers were finally stratified into three equally large groups according to cumulative exposure (low, medium, and high) for the visualization of dose-response associations adjusted for relevant confounders, using general linear models for adjustments. Two-tailed p-values < 0.05 were chosen as the level of statistical significance. The statistical data package SPSS, version 25.0 (IBM Corporation, Somers, NY, USA) was used for the statistical calculations.