Due to the narrow time schedule or technical reasons, FeNO measurement and EBC sampling could not be applied in 14 and 5 workers, respectively. FeNO measurements revealed a median value of 14 (9; 21) ppb. EBC-pH could be determined in each sample with a median of 6.71 (6.25; 7.04). Based on the EBC volume collected, there were 110, 93, and 105 samples available for measurements of 8-iso-PGF_2α, PGE₂, and LTB₄, respectively. In all samples the respective biomarkers were detectable, but partly below the LOQ. In detail, values below the LOQ were 1/110, 55/93 and 89/105 for 8-iso-PGF_2α, PGE₂, and LTB₄, respectively. Concentration of 8-iso-PGF_2α and PGE₂ in EBC was 82.1 (61.9; 118.0) pg/mL and 26.1 (26.1; 54.8) pg/mL, respectively. LTB₄ was transformed into binary categories using the LOQ as cut off. FeNO concentrations were not correlated to pH or biomarker concentrations in EBC when referring to all subjects or subgroups (data not shown). The only significant correlation between pH and an EBC biomarker concentration was observed in case of PGE₂ in the subgroup of smokers (p = 0.007). Within EBC-biomarkers, 8-iso-PGF_2α was positively correlated to PGE₂ (p = 0.020). Concentrations of 8-iso-PGF_2α and PGE₂ were significantly higher in samples with LTB₄ > LOQ compared to <LOQ (108.9 (77.8; 262.6) pg/mL vs. 76.3 (59.5; 99.2) pg/mL, p = 0.012 and 78.8 (65.2; 125.0) pg/mL vs. 26.1 (26.1; 50.1) pg/mL, p < 0.0001).

A rather high current bioaerosol exposure could be suspected in 70/119 compost workers. No differences with respect to the exposure intensity (high vs. low) could be observed for FeNO (14 (9; 19) ppb vs. 14 (10; 22) ppb, p = 0.806), pH (6.74 (6.30; 7.00) vs. 6.66 (5.93; 7.16), p = 0.459), PGE₂ (26.1 (26.1; 54.8) pg/mL vs. 26.1 (26.1; 62.4) pg/mL, p = 0.780) or LTB₄ (> LOQ/< LOQ; 12/53 vs. 4/36, p = 0.278). Levels of 8-iso-PGF_2α were significantly increased in workers considered highly exposed to organic dust (86.6 (66.1; 128.8) pg/mL vs. 74.4 (56.3; 96.7) pg/mL, p = 0.047). Results of FeNO, pH measurements, and 8-iso-PGF_2α concentrations in EBC stratified according to smoking habits and atopy are depicted in Fig. 1. No further differences could be revealed in case of PGE₂ or LTB₄ after stratification of results according to smoking habits and atopy (data not shown). Smoking demonstrated a significant negative impact on FeNO levels both in low (p = 0.0001) and high (p < 0.0001) exposed workers. A trend for higher FeNO levels in atopic subjects was observed in low exposed workers (p = 0.152). EBC-pH was significantly lower in smoking compared to non-smoking subjects which was more apparent in high exposed subjects (6.46 (6.02; 6.95) vs. 6.83 (6.48; 7.05), p = 0.009) compared to low exposed subjects (6.58 (5.84; 7.01) vs. 6.89 (6.45; 7.27), p = 0.049). EBC-pH was not different when referring to atopic status and exposure intensity. No association between pH and cumulative smoking dose (pack-years) could be revealed (p = 0.848) in smokers. Concerning 8-iso-PGF_2α, differences between high and low exposed workers could be mainly attributed to differences in the subgroups of non-smokers (p = 0.087) and the subgroup of non-atopics (p = 0.072).

Correlations between current (CO vol%) or cumulative cigarette exposure (pack-years) and current (clean air supply) or cumulative (years of employment) occupational exposure and effect markers referring to all subjects under investigation (n = 119) are shown in Table 2. LTB₄ was not used for correlation analyses because of the high amount of values below the LOQ (84.8 %). Both current and cumulative cigarette exposure was negatively correlated to FeNO (p < 0.0001, each) and pH (0.005 and 0.010, respectively). No associations could be revealed between biomarkers in exhaled breath and EBC and occupational exposure.

After stratification according to atopic status, an association could be suspected between FeNO and time daily worked under clean air supply in atopic workers (r = −0.335, p = 0.071; Fig. 2a). Moreover, there was a statistically significant correlation between FeNO and duration of employment in this subgroup (r = 0.376, p = 0.041; Fig. 2b). No associations were observed in any subgroup between means of exposure and biomarkers assessed in EBC (data not shown).