The World Health Organization (2010) indoor air guideline value on formaldehyde and recent scientific studies

Gunnar Damgrd Nielsen PhD, Dr Sc (Pharm) National Research Centre for the Working Environment, Denmark

Updated: 16.04.2012

WHO guidelines for indoor air quality. Selected pollutants.

WHO, Regional Office for Europe, Copenhagen, 2010.

A short-term (30 min) guideline of 0.1 mg/m3 was recommended for preventing sensory irritation in the general population. Also, the value was considered to prevent cancer.

Formaldehyde toxicokinetics (WHO 2010)

absorbed mainly in the upper airways (>90%) reacts with glutathione (HO-CH2-SG adduct) adduct is oxidized to formate (Km(rats)~ 2.6 ppm)) at levels above the Km value, levels increase disproportionate in nasal tissue inhalation causes no increase in blood FA T1/2~1.5 min in plasma formed endogenously blood level is about 2-3 mg/L and FA is found in exhaled air (mean~ 2 ppb; 75th percentile~6 ppb) reacts with proteins, RNA and DNA

N. GlossoN. trigeminus (V) pharyngeus (IX)

N. vagus (X)
Shusterman D. Review of the upper airway, including olfaction, as mediator of symptoms. Environ Health Perspect 2002; 110 (suppl.4): 649-653.

Portal-of-entry effects
I. Sensory irritation of eyes and upper airways Asthma: not supported by the WHO (2010) or recent studies

II.

III. Nasal genotoxicity and cancer

I. Sensory irritation of eyes and upper airways

Paustenbach et al. (1997) reviewed all data in animals and humans: No eye irritation at 0.5 ppm Proposed no irritation at < 0.1 ppm (24 h/day) in the general population. No especially sensitive group (including asthmatics) was identified. Epidemiological studies: Mixed exposures (less valuable)
Paustenbach D, Alarie Y, Kulle T, Schachter N, Smith R, Swenberg J, Witschi H, Horowitz SB. A recommended occupational exposure limit for formaldehyde based on irritation. J Toxicol Environ Health 1997, 50: 217-263.

Key study by WHO (2010): Lang et al. (2008)a

Controlled chamber study: 11M and 10 F, age:18-40 y, exposured for 4 h Eye and airway effects: NOAEL 0.5 ppm FA; LOAEL 0.5 ppm with peaks at 1 ppm AF: 5 and rounding to 0.1 mg/m3. The WHO guideline applies to each 30 min period of the day New chamber study : NOAEL 0.7 ppm; LOAEL 0.4 ppm with peaks at 0.8 ppm (hypo- and hypersensitive M)
a) Lang I, Bruckner T, Triebig G. Formaldehyde and chemosensory irritation in humans: a controlled human exposure study. Regul Toxicol Pharmacol 2008, 50: 23-36. b) Mueller JU, Bruckner T, Triebig G. Exposure study to examine chemosensory effects of formaldehyde on hyporsensitive and hypersensitive males. J Arch Occup Environ Health 2012, in press.

Sensory irritation: day after day exposure?

A 10-day study (1 h/day) in mice with a limonene-ozone mixture b) It contained volatile organic compounds and FA a) FA accounted for of the sensory irritation effect a) Low to very high sensory irritation effects were studied b) No increase in sensory irritation or airflow limitation over repeated exposures and no inflammation was observed b) Chamber studies adequately reflect long term effects.
a) Wolkoff et al. Acute airway effects of ozone-initiated d-limonene chemistry: importance of gaseous products. Toxicol Lett 2008, 181: 171-176. b) Wolkoff et al. Airway effects of repeated exposures to ozone-initiated limonene oxidation products as model of indoor air mixtures. Toxicol Lett 2012, 209: 166-172

II. Asthma
FA does not cause asthma in itself a) (except RADS) No lung function effects of FA (< 1 mg/m3) in asthmatics and nonasthmatics b) Pre-exposure to FA in asthmatics and post exposure to an allergen to which subjects were sensitized showed no consistent increase in allergen sensitivity b) No convincing association between FA exposures in homes and schools and asthma in children and adults b). New epidemiological studies: - One found a positive association in the group with the lowest (rural) FA exposure (Hulin et al. 2010):. - Two found no association (Hwang et al. 2011; Kim et al. 2011). Meta-analysis (McGwin et al., 2010): Significant association between FA exposure and asthma in children. Most influential studies were the Rumchev et al. (2002) and the Garret et al. (1999) studies.

a) Paustenbachet al. J Toxicol Environ Health 1997, 50: 217-263. b) Wolkoff and Nielsen. Environ Int 2010, 36: 788-799.

III. Portal-of-entry genotoxicity and carcinogenicity

Animal studies: Nasal genotoxicity is well established (non-linear exposure-response relationships) Nasal cancer: Key effect in cancer risk assessment by the WHO (2010) Human studies: Portal-of-entry genotoxicity is not that clear Nasopharyngeal cancer mainly based on the Hauptmann et al. 2004 study

Hauptmann M, Lubin JH, Stewart PA, Hayes RB, Blair A. Mortality from solid cancers among workers in formaldehyde industries. Am J Epidemiol 2004, 159: 1117-1130.

Nasal epithelial squamous cell carcinomas (SCCs) in rats exposed 6 h/day, 5 days/week to formaldehyde for 2 years a)
Formaldehyde (ppm) 0 0.3 0.7 2 6 10 14 15 Number with SCC/group size (%) 0/453 (0) 0/32 (0) 0/90 (0) 0/364 (0) Apparent NOAEL 3/325 (0.9) Apparent LOAEL 20/90 (22) 103/232 (44) 120/278 (43)

a) Studies: Kerns et al. (1983), Sellakumar et al. (1985), Monticello et al. (1996), Kamata et al.(1997); for full references c.f. Nielsen and Wolkoff. Arch Toxicol 2010, 84: 423-446.

New studies on toxicokinetics and gene expression by airborne FA:

Nasal FA-DNA adduct formation is highly non-linear a,b,c,d) Genotoxic and cytotoxic mode of action for nasal cancer a,b,c,d) No adduct at distant sites a,b,c,d) The toxicokinetic analysis indicated highly non-linear doseresponse relationship for increase of exogenous accumulated tissue FA e) Decreases nasal glutathione and increases gene expression e) Below 1 or 2 ppm FA: No increased risk of nasal cancer or in any other tissue, or effect on FA homeostasis within epithelial cells e)
a ) Lu et al. Toxicol Sci 2010, 116, 441-451, b) Lu et al. Chem Res Toxicol 2011, 24, 159-161, c) Moeller et al. Chem Res Toxicol 2011, 24, 162-164, d) Swenberg et al. Toxicol Sci 2011, 120(S1), S130-S145, e) Andersen et al. Toxicol Sci 2010, 118, 716-731.

Three largest, recently updated, occupational cohorts, simplified from WHO (2010)
Study
Exposure (ppm)

NCI cohort (N=25,619)

Average: 0.45, range: 0.01-4.25 23% had peaks 4 0.90 a,b)/1.07c) 1.19 b) 2.10 b,d) 0.95 b) 0.97 b) 1.42c) 0.85c) 0.94c) 1.02c) 1.15c) 0.90c)

UK cohort (N=14,014)
Range: 0.1 to >2

US garment workers (N=11,039)

Geometric mean: 0.15. GSD: 1.9. Past exposure: Substantial higher 0.89 0.00 (O/E: 0/0.16) 0.64 0.00 (O/E: 0/0.96) 0.88 0.98 0.55 0.85 1.09 1.44 ( 10 y exp + 20 y since first exp)

All cancers Nose and sinuses Pharynx Nasopharynx Larynx Lung Hodgkins disease Non-Hodgkins lymphoma Multiple myeloma Leukemia Lymphatic leukemia Myeloid leukemia

1.10 0.87 1.55 0.50 (O/E: 1/2) 1.07 1.22 0.70 0.98 0.86 0.91 0.89 (exp> 2 ppm)

a) Standardized mortality ratio. Where 95% CI does not include 1, it is indicated by bold. b) NCI updated to1994 (Hauptmann et al. 2003); not corrected data. c) NCI updated to 2004 (Freeman et al. 2009). d) Exact CI: 0.91-4.14.

Nasopharyngeal cancer death in 25,619 workers

The US NCI study a)
Peak exposure level ppm Non-exposed >0 - < 2.0 2.0 - < 4.0 4.0
P trend (0+FA conc.) P trend (FA conc.)

Average exposure intensity ppm 0 >0 - < 0.5 RR and number of cases/person-y 1.00 (Reference) 2 /409,074 Not obtainable 0/279,992 0.38 NOAEL? 1/88,074 1.67 LOAEL? 6/88,568 P=0.126 P=0.066

RR and number of cases/person-y 1.00 (Reference) 2/409,074 Not obtainable 0/209,815

Not obtainable 0.5 -< 1.0 0/121,729 NOAEL? 1.83 LOAEL? 7/125,090 P=0.04 P<0.001 1.0

Non-linear exposure-response relationships

Hauptmann et al. Mortality from solid cancers among workers in formaldehyde industries. Am J Epidemiol 2004, 159, 1117-1130.

Limitations of the Hauptmann et al. (2004) study

Nasopharyngeal cancer (NPC): Validity? a)
Standardized mortality ratio (SMR): 2.10 (exact 95% CL: 0.91-4.14) Six of 10 NPC cases were from one (the Wallingford) of 10 plants. SMR in the Wallingford plant was 10 (4-22) and 0.65 (0.08-2.23) in the combined group of the other nine plants. (NOAEL) Many Wallingford cases had a short exposure Confounders in the Wallingford plant? The study missed 1006 death from the period 1980-94

Consistency with other studies? a)

UK cohort (N=14,014) half of expected NPCs (Coggon et al., 2003) The US garment worker cohort (N=11,039) no NPC (Pinkerton et al., 2004) US case-control study among embalmers who died from lymphohematopoietic cancers. Death from NPC OR(95%CL): 0.1 (0.01-1.2) (Hauptmann et al., 2009)
a) For references c.f. Nielsen and Wolkoff. Arch Toxicol 2010, 84: 423-446 and Golden. Crit Rev Toxicol 2011, 41: 672-721.

WHO 2010 and nasopharyngeal cancer (NPC) Based on the Hauptmann et al. (2004) study:
Effect supported from the rat studies No excess NPC at mean concentrations 1 ppm FA and at peak concentrations below 4 ppm For the guideline setting, the rats NOAEL at 1 ppm for histopathological effects and cell proliferation (below the 2 ppm NOAEL for SCC) was used as point of departure and considered to prevent NPC in humans

Lymphohematopoietic malignancies
Animal studies:
Not convincing, but if present only at 15 ppm and with a non-linear exposure-response relationship

Human studies:
The WHO (2010) analyzed the studies for: Exposure-response relationships? If non-linear, apparent NOAEL? Establishing levels, which the guideline level has to be below

Relative risk (RR) of lymphohematopoietic malignancies in humans average intensity, ppm

(Hauptmann et al. 2003; Freemen et al. 2009) a)
Disease Lymphohematopo ietic malignancies ICD-8: 200-209 2003 0.89 2009 0.99 Non-Hodgkins lymphoma ICD-8: 200 & 202 2003 1.03 2009 1.08 Hodgkins lymphoma ICD-8: 201 2003 0.33 2009 0.53 .11-2.6 1.00 NOAEL 3.62* 1.4-9.3 2.48 .84-7.3 0.03* 0.05* Myeloid leukemia ICD-8: 205 2003 0.40 2009 0.70 .23-2.16 1.00 Multiple myeloma ICD-8: 203 2003 2.04 2009 2.18* 1.01-4.7 1.00

Year 0

0.1-<0.5 Ref. gr. 0.5-<1.0

1.00

1.00

1.00

1.00

1.00 NOAEL 4.59* 1.6-13.2 3.06* 0.9-10.4 0.03* 0.03*

1.00

1.00

1.36

1.29

1.20

1.20

1.05 NOAEL 2.19* 0.9-5.3 0.11 0.11

1.21 .56-2.62 NOAEL 1.61 .76-3.39 0.40 0.43

1.37

1.40

1.0

1.25

1.07

0.80

0.71

1.39

1.49

P(0+FA) P(FA)

0.4 0.4

>0.50 >0.50

>0.5 >0.5

0.40 >0.5

>0.5 >0.5

>0.50 >0.50

a) Sometimes added 95% confidence limits. Figures marked with * do not include one. Hauptmann et al. J Natl Cancer Inst 2003, 95, 1615-1623. Corrected data are presented; had missed 1006 death. Freeman et al. J Natl Cancer Inst 2009, 101, 751-761.

Occupational related to lymphohematopoietic malignancies

Disease
Non-Hodgkins lymphoma Hodgkins lymphoma Leukemia

Exposure
Benzene, trichloroethylene, agriculture/farmers, abattoir (butcher) workers (viruses?), herbicides and pesticides Agriculture/farmers, abattoir workers (viruses?), EBV Benzene (AMLa)), gasoline (3-5% benzene), ionizing radiation (AML), dioxins (TCDD)?, butchers (viruses?), agriculture/farmers (CLL), 1,3butadiene, embalmers and formaldehyde exposed workers, treatment with alkylating agents, rubber industry (CLL, benzene?), arsenic, cigarette smoking (9% of AML?), pesticides Engine exhaust, teachers, pesticides,

Multiple myeloma

a) Acute myeloid leukemia (AML), chronic lymphoid leukemia (CLL).

1) Descatha et al. Occupational exposures and haematological malignancies: overview on human recent data. Cancer Causes Control 2005, 16, 939-953. 2) Blair et al. Chemical exposures and risk of chronic lymphocytic leukaemia. Br J Haematol 2007, 139, 753-761. 3) Natelson Benzene-induced acute myeloid leukemia: a clinicians perspective. Am J Hematol 2007, 82, 826-830. 4) Steinmaus et al. Meta-analysis of benzene exposures and non-Hodgkin lymphoma: biases could mask an important association. Occup Environ Med 2008, 65, 371-378. 5) Van Maele-Fabry et al. Risk of leukaemia among pesticide manufacturing workers: a review and meta-analysis of cohort studies. Environ Res 2008, 106, 121-137. 6) Merhi et al. Occupational exposure to pesticides and risk of hematopoietic cancers: meta-analysis of case-control studies. Cancer Causes Control 2007, 18, 1209-1226. 7) Lichtman MA. Cigarette smoking, cytogenetic abnormalities, and acute myelogenous leukemia. Leukemia 2007, 21, 1137-1140.

The WHO (2010) indoor air guideline value

Sensory irritation is the critical effect, which also prevents cancer NOAEL: 0.5 ppm, using an AF: 5 and rounding: Guideline value: 0.1 mg/m3 (0.08 ppm) (30-min average concentration) Prevention of nasal cancer will prevent other types of cancer (WHO 2010) Point of departure: the 1 ppm NOAEL for histopathological effects and nasal cell proliferation, using an AF of 3 for interspecies variation (local effect) and an AF of 2 for variation in the human population and thus 0.17 ppm (0.2 mg/m3) would prevent cancer Biologically motivated model: 0.25 mg/m3 predicts an additional risk 10-6 for continuous exposure of non-smokers a) Supporting evaluation of cancer effects The 1 ppm point of departure was used, but no interspecies AF was used as the rat is the sensitive species. Being precautionary, an AF of 10 was used for intra human variations, suggesting that 0.1 ppm (0.12 mg/m3) would prevent cancer b) and supported by the results from the biologically motivated model a)
a) Conolly et al. Toxicol Sci 2004, 82: 279-296. b) Nielsen and Wolkoff. Arch Toxicol 2010, 84: 423-446.

Backup slides

Exposure concentrations in homes and dwellings

Area Type of estimate Concentration Reference (g/m3 FA) 20-40 65 ~170 ~30 ~240 10-80 100 5-50 Sarigiannis et al. (2011) Salthammer et al. (2010) (WHO 2010)

EU, Canada Geometric mean, and USA median or average 95th percentile Maximum Japan China Mean Mean

EU, Canada Mean and USA 95th percentile EU Mean

Sarigiannis et al. Exposure to major volatile organic compounds and carbonyls in European indoor environments and associated health risk. Environ Int 2011, 37: 743-765. Salthammer et al. Formaldehyde in indoor environment. Chem Rev 2010, 110: 2536-2572.

Physicochemical properties of FA

Properties Reaction with Melting point Boiling point Water solubility Vapour pressure

Values OH, -NH2 or = NH, -SH a) -92C a,b) -19C a,b) 400g/L at 20C a,b) 3886 mmHg at 25C a,b)

a) WHO 2010 b) http://chem.sis.nlm.nih.gov/chemidplus/

Nasal formaldehyde (FA) metabolism

Inhaled FA (CH2O)

Mucus layer FA (CH2(OH)2) (formaldehyde acetal)

Adducts with mucus layer and cell surface constituents from gene expression1,2) DNA adducts3) Gene expression1,2)

Epithelial cell membrane CH2(OH)2

GSH

Reacts with DNA and proteins

HO-CH2-SG
FDH

HCOO- + GSH 1. Andersen et al. Toxicol Sci 2010, 118, 716-731. 2. Andersen et al. Toxicol Sci 2008, 105, 368-383. 3. Swenberg et al. Toxicol Sci 2011,120(S1): S130-S145.

Lymphohematopoietic malignancies in animal studies Drinking water studies a,b):

Overall, the drinking water studies showed no increase in lymphohematopoietic malignancies in two well-conducted long-term studies (Til et al. 1989; Tobe et al. 1989). Where significant, the effects were at the high FA levels (Soffritti et al. 2002) and apparently, exposure-response relationships were non-linear. However, these results were from a study with severe limitations (IARC 2006).

Inhalation studies a,b):

The occurrence of lymphohematopoietic malignancies in inhalation studies in rats and mice is not convincing (c.f. Kerns et al. 1983; Sellakumar et al. 1985; Kamata et al. 1997). Exposures up to 15 ppm FA.

Similar conclusion in a recent review c) In rats SCC and potentially lymphohematopoietic malignancies
a) WHO (2010) b) Nielsen GD, Wolkoff P. Arch Toxicol 2010, 84, 423-446. c) Golden R. Crit Rev Toxicol 2011, 41: 672-721.

Decreased survival may mask development of lymphohematopoietic cancer

If corrected for survival, only lymphohematopoietic malignancies in the highest exposure group across all inhalation studies in rats and mice, i.e. non-linear exposureresponse relationship and NOAEL

Monticello et al. Correlation of regional and nonlinear formaldehyde-induced nasal cancer with proliferating populations of cells. Cancer Research 1996, 56, 1012-1022.

Buccal and/or nasal genotoxicity in humans

Increased micronucleus formation (MN):
Limited consistency within and across epidemiological studies a) Studies have potential confounders a) Controlled exposure study at 4 h/day for 10 working days at 0.15-0.5 ppm FA with peaks at 0.6 and 1 ppm. Buccal cells: No significant increase b) Where present (WHO, 2010), associated with high mean and/or high peak exposures c) Recent studies: High peak exposures in recent epidemiological studies d,e) Controlled exposure study 4h/day for 5 days. FA 0.7 ppm, peaks op to 0.8 ppm. Nasal epithelial cells: No increase in MN Nasal biopsies: No remarkable change in gene expression f)
a) Speit and Schmid. Mutat Res 2006, 613: 1-9; b) Speit et al. Mutat Res 2007, 627: 129-135; c) Nielsen and Wolkoff. Arch Toxicol 2010, 84: 423-446;d) Viegas et al. J Occup Med Toxicol 2010, 5: 25; e) Ladeira et al. Mutat Res 2011, 721: 15-20; f) Zeller et al. Mutagenesis 2011, 26: 555-561.

Lymphohematopoietic malignancies in humans?

Meta-analysisa) and updated analysisb) the highest exposure group from each study:
Increased RR for leukemiaa,b), especially myeloid leukemiaa,b), and multiple myelomaa), but not for Hodgkinsa), non-Hodkins lymphomaa) and lymphatic leukemiab). WHO (2010): If FA driven, a non-linear exposure-response relationship and NOAEL.

Highly exposed Chines workers c) (FA median: 1.3 ppm; 90th percentile: 2.5 ppm)
had decreased lymphocyte, granulocyte, platelet and red blood cell counts. Cultivated blood stem cells had increased monosomy of chromosome 7 and trisomy of chromosome 8 in exposed to a median of 2.1 ppm and a 90th percentile of 4.1 ppm. WHO (2010): No exposure-response relationship, high exposures (extreme peaks?), and limited transparency. Later evaluations: Study with severe limitations. d,e)

US case-control study in embalmers k): Among lymphohematopoietic

malignancies, myeloid leukemia was associated with embalming. Trend was significant with peak FA exposures (driven by +/- FA). WHO (2010): Low number of cases in the control groups. No exposure-dependent trend within FA groups. Similar OR in each FA metric group. Limitations in exposure assessment (peaks were estimated). Later publication: additional limitations. d,j)
a) Zhang et al. Mutat Res 2009, 681: 150-168. b) Schwilk et al. J Occup Environ Med 2010, 52: 878-886. c) Zhang et al. Cancer Epidemiol Biomarkers Prev 2010, 19: 80-88. d) Golden R. Crit Rev Toxicol 2011, 2011, 41: 672-721. e) Speit et al. Cancer Epidemiol Biomarkers Prev 2010, 19: 1882-1884. k) Hauptmann et Al. J Natl. Cancer Inst 2009, 101: 1696-1708. l) Colle et al. Regul Toxicol Pharmacol 2010, 58: 161-166.

Acute leukemia is a bone marrow diseases with immature blood cells

Acute myeloid/myelogenous/myeloblastic leukemia (AML) (red, some white cells and
platelets)

Acute lymphocytic/lymphatic/lymphoblastic leukemia (ALL)

(lymphocytic line)

Genetic damage
Chromosomal aberrations often translocations, i.e. rearrangement of pieces between
two non-homologous chromosomes; AML, e.g. t(8;21), inv(16) and t(15;17), and ALL, e.g. t(9;22), t(12;12), t(1;19), t(17;19) and t(8;14). Normal karyotype in AML (30-50%) and in ALL (20-45%). Gene mutations genes controlling crucial signalling pathways and key transcription factors.

Epigenetic changes in transcriptional regulation of tumor suppressor genes and

oncogenes

DNA methylation (DNA methyl transferases), e.g. of cytosine to 5-methylcytosine in the

CpG (G: guanosine) islands in promotors. Histone modification methylation, acetylation/deacetylation, ubiquitylation, phosphorylation, sumoylation and ADP-ribosylation. microRNAs are regulators, e.g. in cell apoptosis, differentiation and proliferation. They are involved in leukemogenesis and tumour suppression.
Chen J, Odenike O, Rowley JD. Leukaemogenesis: more than mutant genes. Nature Reviews 2010, 10, 23-34

Relative risk (RR) of lymphohematopoietic malignancies in humans peak exposure, ppm

(Hauptmann et al. 2003; Freemen et al. 2009) a)
Disease Lymphohematopoi etic malignancies ICD-8: 200-209 2003 0.99 2009 1.07 Non-Hodgkins lymphoma ICD-8: 200 & 202 2003 1.01 2009 1.06 Hodgkins lymphoma ICD-8: 201 2003 0.36 0.04-3.2 1.00 NOAEL 3.28 0.96-11 3.30 0.98-11 0.009* 0.04* 2009 0.67 .12-3.6 1.00 NOAEL 3.30* .04-10.5 3.96* 1.3-12.0 0.004* 0.009* Myeloid leukemia ICD-8: 205 2003 0.60 0.1-3.1 1.00 NOAEL 1.98 0.7-5.6 2.79* 1.1-7.2 0.009* 0.02* 2009 0.82 .25-2.7 1.00 Multiple myeloma ICD-8: 203 2003 2.58 0.95-7.0 1.00 2009 2.74* 1.2-6.3 1.00

year 0

0.1-<2.0 Ref. gr. 2.0-<4.0

1.00

1.00

1.00

1.00

1.34 NOAEL 1.48* 1.04-2.2 0.03* 0.03*

1.17 NOAEL 1.37* 1.031.8 0.04* 0.02*

1.02

1.08

1.30 .58-2.9 NOAEL 1.78 0.87-3.6 0.07 0.13

1.55

1.65

4.0

0.95

0.91

2.03 0.89-4.6 >0.5 0.14

2.04* 1.01-4 >0.50 0.08

P(0+FA) P(FA)

>0.5 >0.5

>0.50 >0.50

a) Sometimes added 95% confidence limits. Figures in red and marked with * do not include one. Hauptmann et al. J Natl Cancer Inst 2003, 95, 1615-1623. Corrected data are presented; had missed 1006 death. Freeman et al. J Natl Cancer Inst 2009, 101, 751-761.

Other risk assessment approaches

Index report France US EPA