Environmental Research

Hyunok Choi*, Norbert Schmidbauer , Carl-Gustaf Bornehag


Background: The question regarding the true sources of the purported microbial volatile organic compounds (MVOCs) remains unanswered.

Objective:To identify microbial, as well as non-microbial sources of 28 compounds, which are commonly accepted as microbial VOCs (i.e. primary outcome of interest is Σ 28 VOCs).

Methods: In a cross-sectional investigation of 390 homes, six building inspectors assessed water/mold damage, took air and dust samples, and measured environmental conditions (i.e., absolute humidity (AH, g/m3), temperature (°C), ventilation rate (ACH)). The air sample was analyzed for volatile organic compounds (μg/m3) and; dust samples were analyzed for total viable fungal concentration (CFU/g) and six phthalates (mg/g dust). Four benchmark variables of the underlying sources were defined as highest quartile categories of: 1) the total concentration of 17 propylene glycol and propylene glycol ethers (Σ17 PGEs) in the air sample; 2) 2,2,4-trimethyl-1,3-pentanediol monoisobutyrate (TMPD-MIB) in the air sample; 3) semi-quantitative mold index; and 4) total fungal load (CFU/g).

Results: Within severely damp homes, co-occurrence of the highest quartile concentration of either Σ17 PGEs or TMPD-MIB were respectively associated with a significantly higher median concentration of Σ 28 VOCs (8.05 and 13.38 μg/m3, respectively) compared to the reference homes (4.30 and 4.86 μg/m3, respectively, both Ps ≤0.002). Furthermore, the homes within the highest quartile range for Σ fungal load as well as AH were associated with a significantly increased median Σ 28 VOCs compared to the reference group (8.74 vs. 4.32 μg/m3, P=0.001). Within the final model of multiple indoor sources on Σ 28 VOCs, one natural log-unit increase in summed concentration of Σ17 PGEs, plus TMPD-MIB (Σ 17 PGEs + TMPD-MIB) was associated with 1.8-times (95% CI, 1.3–2.5), greater likelihood of having a highest quartile of Σ 28 VOCs, after adjusting for absolute humidity, history of repainting at least one room, ventilation rate, and mold index (P-value =0.001). Homes deemed severely mold damaged (i.e., mold index =1) were associated with 1.7-times (95% CI, 0.8—3.6), greater likelihood of having a highest quartile of Σ 28 VOCs, even though such likelihood was not significant (P-value =0.164). In addition, absolute humidity appeared to positively interact with mold index to significantly elevate the prevalence of the highest quartile category of Σ 28 VOCs.

Conclusion: The indoor concentration of Σ 28 VOCs, which are widely accepted as MVOCs, are significantly associated with the markers of synthetic (i.e. Σ17 PGEs and TMPD-MIB), and to less extent, microbial (i.e., mold index) sources.

Link to the full article

* Denotes CSDA Associates and Staff