Registration Dossier

Data platform availability banner - registered substances factsheets

Please be aware that this old REACH registration data factsheet is no longer maintained; it remains frozen as of 19th May 2023.

The new ECHA CHEM database has been released by ECHA, and it now contains all REACH registration data. There are more details on the transition of ECHA's published data to ECHA CHEM here.

Diss Factsheets

Administrative data

Description of key information

There is sufficient evidence for skin sensitizing properties of formaldehyde in the guinea pig maximisation test (GPMT), in the Buehler test in guinea pigs and in the mouse local lymph node assay (LLNA). Formaldehyde is also a dermal allergen in humans. Anaphylaxis has been documented in case reports.
Animal studies do not indicate that formaldehyde may induce sensitization to the respiratory tract. A very limited number of case reports have been published on formaldehyde-related asthma but these data do not provide sufficient evidence that formaldehyde should be considered a risk factor for respiratory tract sensitization.


The data on skin and respiratory tract sensitization by formaldehyde have been comprehensively reviewed by the German MAK commission (MAK, 2010). With regard to skin sensitization it was concluded that allergic contact dermatitis is relatively frequently observed in patients by diagnostic patch testing. In addition, many experimental animal studies gave positive results for skin sensitisation. As asthma related to formaldehyde is concerned, it was concluded that the allergological findings do not provide a consistent pattern. In inhalation challenge tests generally immediate reactions were observed, dual or late reactions were rare. A differentiation to irritation often is difficult and specific IgE antibodies, if found, mostly did not correlate with the respiratory symptoms. Overall, a relationship of respiratory symptoms with formaldehyde is only in few cases sufficiently documented. The small number of reliable findings in comparison to the broad exposure potential would not warrant to classify formaldehyde as a respiratory sensitizer according to the criteria of the MAK commission.


Specialized studies with experimental animals did not provide evidence for a formaldehyde-induced respiratory allergy. In general, the results in human studies did not provide clear evidence for a formaldehyde-induced respiratory allergy. Reported symptoms in case reports might also be related to irritant effects.

Key value for chemical safety assessment

Skin sensitisation

Link to relevant study records
Reference
Endpoint:
skin sensitisation: in vivo (non-LLNA)
Type of information:
experimental study
Adequacy of study:
key study
Study period:
Jul 20, 1983 to Sep 08, 1983
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
guideline study with acceptable restrictions
Remarks:
no data on positive controls (but clearly positive results), no data on impurities of the test substance (but analytical grade), high concentration for intradermal induction
Qualifier:
according to guideline
Guideline:
OECD Guideline 406 (Skin Sensitisation)
Version / remarks:
May 12, 1981
Deviations:
yes
Remarks:
intradermal induction with high concentration
Qualifier:
according to guideline
Guideline:
other: EEC Directive 79-831, Annex V, Part B: Methods For The Determination Of Toxicity 4.1.7 Skin Sensitization
GLP compliance:
yes
Type of study:
guinea pig maximisation test
Justification for non-LLNA method:
The LLNA method was not available yet by the time the study was conducted.
Specific details on test material used for the study:
TEST MATERIAL
- Source: Riedel del Haen, Germany
- Purity: analytical grade, at least 35 % formaldehyde in aqueous solution
- Physical appearance: colourless, liquid

STABILITY AND STORAGE CONDITIONS OF TEST MATERIAL
- Storage condition of test material: in the dark at ca. 22°C
Species:
guinea pig
Strain:
other: Pirbright White, Hoe: DHPK(SPFLac)
Sex:
female
Details on test animals and environmental conditions:
TEST ANIMALS
- Source: Hoechst AG, Kastengrund, Germany
- Weight at study initiation: 247 g (219 - 279 g)
- Housing: in groups of 5 in Makrolon cages type IV
- Diet: ERKA 8300 (Robert Koch OHG, 47 Hamm), ad libitum
- Water: tap water, ad libitum
- Acclimation period: 5 days

ENVIRONMENTAL CONDITIONS
- Temperature: 20 - 24°C
- Humidity: 45 - 65 %
- Fully air-conditioned rooms
- Photoperiod: 12 hours / 12 hours
Route:
intradermal
Vehicle:
other: 50 % Freund`s adjuvant or 0.9 % NaCl
Concentration / amount:
5 % / 0.1 mL
Day(s)/duration:
study day 1 / once
Adequacy of induction:
highest concentration used causing mild-to-moderate skin irritation and well-tolerated systemically
Route:
epicutaneous, occlusive
Vehicle:
physiological saline
Concentration / amount:
5 % / 0.5 mL
Day(s)/duration:
study day 9 / 48 hours
Adequacy of induction:
highest concentration used causing mild-to-moderate skin irritation and well-tolerated systemically
No.:
#1
Route:
epicutaneous, occlusive
Vehicle:
physiological saline
Concentration / amount:
4 % (left flank) and 2 % (right flank) / 0.5 mL
Day(s)/duration:
study day 22 / 24 hours
Adequacy of challenge:
highest non-irritant concentration
No.:
#2
Route:
epicutaneous, occlusive
Vehicle:
physiological saline
Concentration / amount:
4 % (left flank) and 2 % (right flank) / 0.5 mL
Day(s)/duration:
study day 36 / 24 hours
Adequacy of challenge:
highest non-irritant concentration
No. of animals per dose:
20
Details on study design:
RANGE FINDING TESTS
The following test concentrations were applied to the right flank of two guinea pigs each dermal-occlusive for primary skin irritation: 10 % in 0.9 % NaCl, 6 % in 0.9 % NaCl and 1.2 % in 0.9 % NaCl. The right flank of the test animals was mechanically depilated, 0.5 mL of the test substance preparation was applied to a 2 x 2 cm cellulose swab, fixed to the flank and occluded (with foil and bandage) for 24 hours. The treated skin site was checked 24 hours after removal of the patch.
After treatment with the 1.2 % test substance solution, no irritation effects occurred. The dermal application of the 6 % solution resulted in hardly perceptible erythema, while the 10 % solution led to the formation of well-defined erythema. The results of this preliminary test led to the use of a 4 and 2 % test substance solution for dermal re-treatment in the main experiment.
To determine tolerability, the following 3 x 2 preparations were injected intradermally into two guinea pigs each:
1. site 1: 2 x 0.1 mL 50 % Freund's adjuvans, site 2: 2 x 0.1 mL 5 % test substance in 0.9 % NaCl, site 3: 2 x 0.1 mL 5 % test substance in 50 % Freund's adjuvans
2. site 1: 2 x 0.1 mL 50 % Freund's adjuvans, site 2: 2 x 0.1 mL 1 % test substance in 0.9 % NaCl, site 3: 2 x 0.1 mL 1 % test substance in 50 % Freund's ddjuvans
3. site 1: 2 x 0.1 ml 50 % Freund's adjuvans, site 2: 2 x 0.1 mL 0.2 % test substance in 0.9 % NaCl, site 3: 2 x 0.1 mL 0.2 % test substance in 50 % Freund's adjuvans
The injection sites were located within an area of 2 x 4 cm in the dorsal region near the shoulders of the animals.
Intradermal injections of 0.2, 1, and 5 % test substance in 0.9 % NaCl led to hardening, crusting and browning of the injection sites within 12 days. For the intradermal injections in the main experiment, the use of a 5 % preparation was used for the intradermal injections in the main experiment in order to compare the results of this with the results of studies conducted under these conditions published in the literature.

MAIN STUDY
A. INDUCTION EXPOSURE
INTRADERMAL
An area of 4 x 6 cm in the dorsal region near the shoulders of the animals was mechanically depilated. Intradermal injection of a total of 3 x 2 preparations was carried out per test animal. The injection sites were within an area of 2 x 4 cm. The injection sites were not covered. The application area was examined for local tolerance. Systemic toxic effects were recorded.
Control group:
site 1: 2 x 0.1 mL 50 % Freund's adjuvans
site 2: 2 x 0.1 mL 0.9 % NaCl
site 3: 2 x 0.1 mL 50 % Freund's adjuvans
Test group:
site 1: 2 x 0.1 mL 50 % Freund's adjuvans
site 2: 2 x 0.1 mL 5 % test substance in 0.9 % NaCl
site 3: 2 x 0.1 mL 5 % test substance in 50 % Freund's adjuvans

EPICUTANOUS
0.5 mL of the test substance preparation or the vehicle was applied to a a 2 x 4 cm cellulose swab. This swab covered the area of the intradermal iniection sites. An occlusive impermeable film and elastic bandage closed the application site for 48 hours.
Control group: 0.9% NaCl
Test group: 5 % test substance in 0.9 % NaC

B. CHALLENGE EXPOSURE
Two areas of approximately 5 x 5 cm on the left and right flanks of the test animals were mechanically depilated. 0.5 mL of the test substance preparation was applied to a 2 x 2 cm cellulose swab. An occlusive dressing with impermeable film and elastic bandage sealed the application site for for 24 hours.
Control and Test group : left flank : 4 % test substance, right flank: 2 % test substance
The second re-treatment was carried out after the irritation caused by the first re-treatment had completely subsided.

ASSESSMENT OF THE SKIN REACTIONS ACCORDING TO DRAIZE
- Erythema and eschar formation:
0 = no erythema
1 = very slight erythema (barely perceptible)
2 = well-defined erythema
3 = moderate to severe erythema
4 = severe erythema (beet redness) to slight eschar formation (injuries in depth)
- Edema formation
0 = no edema
1 = very slight edema (barely perceptible)
2 = slight edema (edges of area well defined by definite raising)
3 = moderate edema (raised approximately 1 mm)
4 = severe edema (raised more than 1 mm and extending beyond the area of exposure)
Challenge controls:
yes
Positive control substance(s):
no
Group:
positive control
Remarks on result:
not measured/tested
Reading:
1st reading
Hours after challenge:
48
Group:
negative control
Dose level:
0
No. with + reactions:
0
Total no. in group:
20
Clinical observations:
no
Reading:
1st reading
Hours after challenge:
48
Group:
test chemical
Dose level:
4 %
No. with + reactions:
20
Total no. in group:
20
Clinical observations:
erythema up to grade 3, edema up to grade 2
Remarks on result:
positive indication of skin sensitisation
Reading:
1st reading
Hours after challenge:
48
Group:
test chemical
Dose level:
2 %
No. with + reactions:
16
Total no. in group:
20
Clinical observations:
erythema up to grade 2, edema up to grade 1
Remarks on result:
positive indication of skin sensitisation
Reading:
2nd reading
Hours after challenge:
72
Group:
negative control
Dose level:
0
No. with + reactions:
3
Total no. in group:
20
Clinical observations:
erythema up to grade 1
Reading:
2nd reading
Hours after challenge:
72
Group:
test chemical
Dose level:
4 %
No. with + reactions:
20
Total no. in group:
20
Clinical observations:
erythema up to grade 3, edema up to grade 2
Remarks on result:
positive indication of skin sensitisation
Reading:
2nd reading
Hours after challenge:
72
Group:
test chemical
Dose level:
2 %
No. with + reactions:
18
Total no. in group:
20
Clinical observations:
erythema up to grade 3, edema up to grade 2
Remarks on result:
positive indication of skin sensitisation
Reading:
rechallenge
Hours after challenge:
48
Group:
negative control
Dose level:
0
No. with + reactions:
6
Total no. in group:
20
Clinical observations:
erythema up to grade 2
Reading:
rechallenge
Hours after challenge:
48
Group:
test chemical
Dose level:
4 %
No. with + reactions:
20
Total no. in group:
20
Clinical observations:
erythema up to grade 4, edema up to grade 2
Remarks on result:
positive indication of skin sensitisation
Reading:
rechallenge
Hours after challenge:
48
Group:
test chemical
Dose level:
2 %
No. with + reactions:
11
Total no. in group:
20
Clinical observations:
erythema up to grade 4, edema up to grade 1
Remarks on result:
positive indication of skin sensitisation
Interpretation of results:
Category 1B (indication of skin sensitising potential) based on GHS criteria
Conclusions:
Sensitizing effects in the GPMT.
Executive summary:

In a reliable GLP conform study according to OECD TG 406, formaldehyde was tested for its potential to have a sensitizing effect with the help of the Guinea Pig Maximization Test based on the method of Magnusson and Kligman. Twenty females were induced by intradermal injection and topical application with 5 % formaldehyde solution, 10 female controls were shame treated. Challenge was performed with 2 or 4 % formaldehyde solution.


Controls gave valid results after both concentrations. In test animals 4 % formaldehyde solution resulted in positive reaction in all tested animals and 2 % formaldehyde solution in 80 % positive reactions in the first reading 48 hours after challenge application.


Under these test conditions and following the results described formaldehyde has a sensitizing effect on the skin of the guinea pig.

Endpoint conclusion
Endpoint conclusion:
adverse effect observed (sensitising)
Additional information:

SENSITISATION IN EXPERIMENTAL ANIMALS


 Guinea pig maximisation test (GPMT)


Hoechst (1983) used the GPMT for investigating skin sensitization. Twenty female guinea pigs were induced by intradermal injection and topical application with 5 % formaldehyde, 10 female controls were sham treated. Challenge was performed with 2 or 4 % solution in test animals and controls; controls showed no erythema at both concentrations. In test animals, 4 % formaldehyde resulted in a positive reaction in all tested animals and 2 % formaldehyde gave 80 % positive reactions at the 1st reading.


Bayer (1985 & 1987) using a similar experimental design gave also positive results, even after a 2nd challenge with 0.5 % formaldehyde.


Kimber et al. (1991) found a positive reaction after challenge with 2 % formaldehyde in physiological saline (non-irritant) in 9/9 animals induced via intradermal injection with 0.25 % formaldehyde and topical  application with 10 % formaldehyde in physiological saline (irritant).


Hilton et al. (1996) observed sensitising effects in all 10 treated animals using the same experimental design.


Marzulli et al. (1982) found a positive reaction after challenge with 5 % formaldehyde in physiological saline (non-irritant) in only 5/28 animals induced via intradermal injection with 5 and 10 % formaldehyde and topical application with 5 % formaldehyde in physiological saline, but as the scoring system is not specified, this results are doubtful.


BASF (1979) could not detect a positive reaction after challenge with 5 % formaldehyde induced via intradermal injection and topical application with 5 % formaldehyde, but as the documentation is insufficient for assessment, this study is not reliable.


 


CONCLUSION: Sensitizing in the GPMT.


 


Buehler test


Marzulli et al. (1982) used 10 guinea pigs per trial (3 replicates in total) to induce three times at day 1, 7, and 14 with 0.5 mL of a solution of 5 % formalin (37 % corresponding to 1.85 % formaldehyde under occlusive dressing (no data about local effects). For challenge at day 28 using the same exposure conditions, no skin reaction was detected in 30 animals. Here, formaldehyde (formalin) was not sensitising in this Buehler assay. The authors concluded that this procedure will only identify strong or moderate human contact sensitizers, but not weak ones.


Hilton et al. (1996) presented clearly positive results. Ten guinea pigs were induced by topical application of 5 % formaldehyde solution for 6 h under occlusive dressing. This procedure was repeated once weekly for 3 consecutive weeks. Five controls were sham treated. The challenge (6 h occlusive patch with 1 % test substance) was performed 12 - 14 days after the last induction. A positive reaction was found in 7/10 animals. No positive reaction was detected in 5 negative controls.


CONCLUSION: Sensitizing in the Buehler test if sufficient concentrations were used.


Local Lymph Node Assay (LLNA)


Kimber et al. (1991) conducted an interlaboratory comparison using the LLNA. In this LLNA 4 mice per dose received topical application of 25 μL of the test substance solutions (vehicle only, and 5, 10, 25 %) to the dorsum of both ears (daily applications on 3 consecutive days). At day 4 lymph nodes were excised 5 h after i.v. injection of 3H-labelled thymidine and lymph nodes pooled within each dose group. Incorporation of 3H-thymidine was measured via beta-scintillation counting. A significant increase in 3H-thymidine incorporation even at the lowest concentration of the test substance used for lymphocyte activation was observed. The results with formaldehyde were positive in 4 independent laboratories. The increase in stimulation index in all four laboratories was >3 even at the low concentration of 5 % formaldehyde.


Basketter et al. (2001) treated 4 mice per dose with 25 μL of the test substance solution at various concentrations (0, 0.04, 0.2, 0.4, 1.9, 3.7 % formaldehyde). The test solutions were prepared by dissolving formalin (37 % formaldehyde) in acetone/olive oil (4:1). The authors used a similar experimental design as in the study above. The stimulation index increased dose-dependently. No induction was detected at 0.04 % formaldehyde and the lowest sensitizing dose was 0.2 % with a stimulation of proliferation by a factor of 1.9. The EC3 (3-fold stimulation of proliferation as an index of the relative potency of a contact allergen) was 0.35 % formaldehyde corresponding to 87 μg/cm² (EC3 [%] x 250 [μg/cm²/% ] = EC3 [μg/cm²]). Correspondingly, the NOAEC of 0.04 % formaldehyde would be 10 µg/cm². This calculation was carried out according to ECHA (2008, Chapter R.8) assuming a dose volume of 25 µL (according to the standard LLNA protocol) and an estimated application area of 1 cm² for the mouse ear.


Hilton et al. (1998) reported slightly higher EC3 values (but still in the same range) using formaldehyde solutions in acetone and dimethylformamide, i.g. 0.18 mol/L (corresponding to 0.54 % in acetone) and 0.11 mol/L (corresponding to 0.33 % in dimethylformamide). In a former study of the same group (Hilton et al. 1996) only SI values have been reported as well as the studies of Arts et al. (1997) and Dearman et al. (1999) do.


Boverhof et al. (2008) conducted an interlaboratory comparison using the LLNA. Using 1, 5 or 20 % formaldehyde in 1 % pluronic L92 surfactant in water revealed EC3 values between 3.8 and 12.3 in 5 different laboratories, being a wide range.


For risk assessment a decision has to be made which of the EC3 values obtained with different solvents should be used. For formaldehyde being very hydrophilic a highly polar solvent is most appropriate like acetone. There is no justification for the addition for a solvent containing 25 % olive oil corresponding to the acetone/olive oil system of Basketter et al. (2001) or using . Application in dimethylformamide, being a potent enhancer of skin penetration, would not correspond to exposure conditions in praxi (generally aqueous solutions). Therefore, the EC3 obtained with acetone as solvent is preferred for risk assessment. The EC3 of 0.54 % corresponds to an area dose of 135 µg/cm² as calculated according to ECHA (2008), which was also used by Griem et al. (2003). The authors compared quantitative data from humans and animals on skin sensitization for a large number of chemicals, including formaldehyde, and proposed a risk assessment methodology based on EC3 values from the LLNA. This value of 135 µg/cm² should be taken as the LOAEL for risk assessment based on the EC3. Correspondingly, the NOAEC for an acetone solution would be 0.06 % with an area dose of 15 µg/cm². A clearly higher EC3 of 0.96 % (corresponding to 240 µg/cm²) was reported by De Jong et al. (2007) using the acetone/olive oils system. In addition, in order to enhance possible low responses of weak sensitizers, animals were pre-treated on the dorsum of the ears with sodium dodecylsulfate 1 h before administration of the test solutions. Thus, taking the Hilton et al. (1998) data in acetone for risk assessment represents a conservative approach. 


De Jong et al. (2007) investigated the effect of prolonged repeated exposure to formaldehyde on the magnitude of response elicited in the LLNA with a different mouse strain. After a single exposure regime according to the standard (short-term) LLNA procedure (treatment on days 0, 1, and 2 and measuring the response on day 5) the EC3 value was 0.96 % and the EC2 value 0.6 % formaldehyde. After prolonged repeated exposure (treatment on day 0, 1, and 2 followed by a treatment once a week over 7 weeks and a 3-day treatment at days 56, 57, and 58 and measuring the response on day 61) the increase of cell proliferation at 0.6 % formaldehyde was compared to that obtained by the standard (short-term) procedure. The short-term procedure led to an increase of proliferation (compared to the vehicle control group) by a factor of 2.2 and the repeated prolonged exposure by a factor of 7 with 0.6 % formaldehyde. Thus, repetitive and prolonged exposure (in total 13 times over 8 weeks) increased the response by a factor of approximately 3 in comparison to short-term exposure. Obviously, the response to repeated exposures is leading to a plateau as 7 repetitions of the standard (“acute”) LLNA only led to an enhancement of effect by a factor of 3. According to the authors the finding for formaldehyde (and some other formaldehyde releasers) is in contrast to observations with some other skin sensitizers like benzocaine, dinitrochlorobenzene, or tetramethylthiuramdisulfide. These latter chemicals did not induce an increase of proliferative response by prolonged exposure under comparable experimental conditions. The authors postulate that the increased proliferated response of formaldehyde after repeated prolonged exposure may be explained by local crosslinking of proteins and other macromolecules in the skin leading to prolonged presence in the tissue. 


If these data are taken into account for risk assessment the threshold obtained by the standard LLNA procedure should be divided by 3 for extrapolation to repeated prolonged exposure. This factor should be sufficient as repeated exposure obviously led to a “plateauing” of the magnitude of response. The NOAEC of 0.04 % (Basketter et al., 2001) for the acetone/olive oil preparation would thereby be reduced to 0.013 % corresponding to 3.3 µg/cm². Transformation of this NOAEC to the acetone solution of Hilton et al. (1998) would give a NOAEC of 0.02 % (5 µg/cm²). The EC3 in acetone of Hilton et al. (1998) would be reduced to 0.18 % corresponding to an area dose of 45 µg/cm². These data suggest that formaldehyde is a strong sensitizer (cf ECHA, 2008; Table R. 8-23, p. 127).


 


CONCLUSION: Sensitizing in the LLNA (with EC3 values ranging between 0.33 % and 0.96 % depending on the solvent and the strains used, and a NOAEC in one experiment of 0.04 %).


 


Other tests investigating skin sensitization


Marzulli et al. (1982) exposed 30 guinea pigs to ten intradermal injections of each 0.05 mL µL of 0.1 % formaldehyde in saline within 3 weeks being followed by one last injection 2 weeks later. Ten animals showed a positive reaction.


In the same study they exposed 30 guinea pigs to four topical, occlusive applications of each 0.2 mL of 5 % formaldehyde in saline within one week, whereas the third application was combined with two injections of 0.075 mL Freund`s complete adjuvant, being followed by one last application 2 weeks later. Two animals showed a positive reaction.


In a third approach 30 guinea pigs were exposed to four topical, occlusive applications of each 0.2 mL of 5 % formaldehyde in saline being followed by a challenge with 0.5 mL of 5 % formaldehyde 3 weeks later. Four animals showed a positive reaction.


Lee et al. (1984) exposed 8 guinea pigs to two topical applications of each 100 µL of 37 % formaldehyde in saline being followed by a challenge with 20 µL of 2 % formaldehyde and revealed positive reactions in all animals.


 


SENSITISATION IN HUMANS


Formaldehyde is a dermal allergen in humans (WHO, 1989; IARC, 1995; ATSDR, 1999; OECD, 2002; summary of sensitisation in humans). Formaldehyde solution is a primary skin sensitizer inducing allergic contact dermatitis Type IV and may induce contact urticaria Type I (WHO, 1989). However, contact urticaria has been rarely associated with formaldehyde exposure (IARC, 1995; MAK, 2010). In the OECD SIDS Formaldehyde (2002) a threshold for the challenge concentration in patch tests on formaldehyde sensitized subjects was reported: 30 ppm (0.003%) in aqueous solution and 60 ppm (0.006%) for products containing formaldehyde (but no reference is given by OECD in this respect). However, other data on concentration-response relationships for skin allergic reaction in formaldehyde-sensitive patients induced by dermal exposures to formaldehyde suggested that a positive reaction to formaldehyde is rare below concentrations of 0.025-0.05% (ATSDR, 1999).


An indication for a threshold for induction of allergic skin reactions in sensitized persons can also be obtained from formaldehyde concentrations used for diagnostic patch testing. Trattner et al. (1998) compared the results obtained by simultaneous testing with 1% and 2% formaldehyde solutions in 3734 patch tests. Testing was done by occlusion in Finn Chambers and the patches were applied to the skin for 2 days. The number of positive reactions was the same for both concentrations, but 2% gave significantly more irritant reactions and therefore a 1% patch test concentration was proposed. But this concentration is a compromise between avoiding irritation (false positive) and sufficient sensitivity for detection of sensitization. As indicated in the review of De Groot et al. (2009) even an 1% test concentration may lead to false positive reactions in about 50% and as the study of Trattner et al. (1998) indicates about 40% of allergic patients may be missed in comparison to those identified by a test concentration of 2%.


Several studies have investigated the threshold for elicitation of contact allergic reactions in patients allergic to formaldehyde. Jordan et al. (1979) performed double-blind controlled tests on formaldehyde threshold responses in 9 allergic patient by repeated closed patch testing at the same site for 1 week (three patches applied during the week) with 0, 30, 60, and 100 ppm formaldehyde. Five patients were known to have a strong allergy to formaldehyde. The following incidences of positive reactions were found: 4/9 at 30 ppm; 5/9 at 60 ppm; 6/9 at 100 ppm. Two subjects reacting to 30 ppm were retested later and again had positive reactions. As the closed patch test represents exaggerated exposure conditions in comparison to many exposure scenarios, 11 subjects with delayed hypersensitivity to formaldehyde (including 7 from the closed patch testing) were tested in an open approach by spraying 30 ppm formaldehyde in an axilla (twice a day over 2 weeks). Only a very minimal response was produced in 2/11 subjects. The authors conclude that levels below 30 ppm should be tolerated by sensitive subjects if repeatedly applied to normal skin. It should be mentioned that an exaggerated closed test procedure with 3 applications over 1 week was used here in comparison to the standard closed patch testing once over 48 h. Finally, the description of the test method (no data for area of application and total amount administered) does not allow a calculation of the amount of formaldehyde applied per unit area.


Flyvholm et al. (1997) tested 20 patients allergic to formaldehyde with serial dilutions of 25, 50, 250, 500, 5000, and 10000 ppm formaldehyde under occluded conditions. 15 µl formaldehyde solutions were applied for 2 days by Finn Chambers (diameter 0.8 cm corresponding to an area of 0.5 cm²). Nineteen reacted to 10000 ppm, 9/20 to 5000 ppm, 2/20 to 500 ppm and 1/20 to 250 ppm. Retesting of the patient reacting to 250 ppm 1 year later showed negative results with 50, 100, and 250 ppm. In addition, non-occluded patch testing was performed by application of 15 µl of the formaldehyde solutions to a 1 cm² area of the forearm and allowing to dry at room temperature. Under these conditions no positive reactions were observed even at the highest concentration of 10000 ppm. The authors concluded that 250 ppm formaldehyde is the threshold concentration for occluded patch testing. The Finn Chamber device (area 0.5 cm²) used here allows the calculation of the amount of formaldehyde per unit area: 7.5 µg/cm² for 15 µl with a concentration of 250 ppm. Griem et al. (2003) assumed a NOAEC of 100 ppm corresponding to an area dose of 3µg/cm².


Finally, De Groot et al. (1988 cited in De Groot, 2009) found that 8 out of 35 formaldehyde allergic patients reacted with closed patch testing down to 1000 ppm, lower concentrations were not investigated.


Based on these findings De Groot et al. (2009) concluded that the safe formaldehyde concentration for sensitive patients remains largely unknown. Levels of 200-300 ppm free formaldehyde in cosmetic products have been shown to induce dermatitis from short-term use on normal skin.


In conclusion, by these data it is proposed to take 3 µg/cm² as NOAEL in formaldehyde sensitized persons for risk assessment.


A threshold concentration for induction of formaldehyde sensitization has been estimated to be less than 5% aqueous solution (OECD, 2002). This probably relates to formalin (and not formaldehyde) as used in the study of Kligman (1966, cited in OECD, 2002). The author described a human Maximization Test with formalin. For induction 5% formalin (formaldehyde concentration not given) was applied under occlusive dressing 5 times over 48 h preceded each by occlusive application of sodium lauryl sulfate over 24 h to induce moderate inflammation. The challenge test used 1% formalin after pretreatment with sodium lauryl sulfate for 1 h. Thereby 18/25 subjects became sensitized. As no further concentrations were tested, a threshold for induction of skin sensitization cannot be derived from this publication.


Marzulli and Maibach (1974) used the Draize test to investigate the induction concentration of formaldehyde in human volunteers by the human repeated-insult test (HRIPT). For induction the material was applied to the skin under occlusive conditions for 48 or 72 h. Ten epicutaneous applications were administered at the same site within 3.5 weeks. After a rest period of 2 weeks a challenge patch containing 0.37% formaldehyde was applied for 72 h. The following results were obtained for different induction concentrations of formaldehyde: 0.037%: 0/45; 0.37%: 4/89; 1.1%: 5/88; 1.9%: 4/52; 3.7%: 8/102 (subjects with positive reactions / subjects tested). Thus, induction of sensitization is not to be expected at a concentration of 0.037% even after repeated exposures under occlusive dressing over 2-3 days. The NOAEC of 0.037% was transformed to an area dose of 37 µg/cm² in reviews by Griem et al. (2003), Gerberick et al. (2001) and Basketter et al. (2005, 2008). In their review Basketter et al. (2008) note some shortcomings of the methodological description by Marzulli and Maibach (1974) but by taking account of the exposure being greater than in the standard HRIPT protocol they concluded that 37 µg/cm² is a reasonable conservative estimate of the NOAEC.


In a recent study (Thompson et al., 2002) patch testing in patients with suspected allergic contact dermatitis was performed between 1994 and 1996 (standard tray of 49 allergens including formaldehyde). Patients were exposed to 1% formaldehyde in water for 48 h and readings were performed 48 h after patch removal. In Kansas City 20 out of 138 patients with dermatitis showed a positive result in the patch test (corresponding to 14.5%); in other regions of the USA 268 out of 2973 patients patch tested with formaldehyde showed a positive result (corresponding to 9.0%). The difference is significant (p<0.05). Further details on the dermal sensitization rate for patients tested in dermatological clinics are given by MAK (2010). In the prevalence has nowadays declined to about 2%. Fora rate of 0.9% was reported for the general population and inof 0.3-0.6%.


Conclusion: In the USA 9% of patients with allergic contact dermatitis showed positive reaction in the standard patch test with 1% formaldehyde, in Kansas City 20 out of 138 patients (14.5%). Lower incidences have been described for.


With regard to skin sensitization MAK (2010) concluded that allergic contact dermatitis is relatively frequently observed in patients by diagnostic patch testing. In addition, many experimental animal studies gave positive results for skin sensitization.


 Zug et al. (2009, supporting) reported results from 4454 patients tested for 65 allergens by the North American Contact Dermatitis Group in 2005-2006. Formaldehyde ranged among the 15 most frequent allergens with an incidence rate of 9% (for comparison, the incidence rate for nickel sulphate was 19%, the highest incidence observed). Similar results were obtained by Arrandale et al. (2012, supporting): formaldehyde ranged among the 10 most frequent skin allergens. In 3676 patients 19 reacted positive to formaldehyde and the response was related to workplace exposure. When 6845 patients in Australia were patch tested between 1993 and 2006, formaldehyde was the most frequent contact allergen among the preservatives studied with 4.6% positive results (Chow et al., 2012, supporting).


Lundov et al. (2010, supporting) found by diagnostic patch testing that patients allergic to formaldehyde-releasers often had simultaneous allergy to formaldehyde itself. Nearly all patients who reacted positively to 2 formaldehyde-releasers were also allergic to formaldehyde. Thus concomitant allergy to formaldehyde-releasers and formaldehyde is common.


Isaksson et al. (2011, supporting) found that in 2504 patients that were tested by standard patch test series, 1.1% (27 subjects) reacted positively to a mixture of monomers and dimers from a resol resin based on phenol and formaldehyde. Of these 27 patients 2 reacted positively in addition to formaldehyde and 2 to a p-tert-butylphenol-formaldehyde resin. The authors propose that the mixture they used here should be added to the standard battery of substances for diagnostic patch testing.


Ponten et al. (2013, supporting) found that increasing the concentration of formaldehyde for diagnostic patch testsing from 1% (as commonly used in clinical practice) to 2% (wt/vol) significantly increased the positive reactions from 1.8% to 3.4%. Therefore, they propose that 2% test solutions should be used for clinical diagnostic patch testing.  


Kadivar and Belsito (2015, supporting) patch tested 2611 patients, among these 165 health care workers. Health care workers had significantly more workplace related allergic contact dermatitis, especially to formaldehyde, than non-health care workers.


Systemic sensitization (immediate type of allergy mediated by IgE)


An anaphylactic shock reaction after accidental i.v. application of formaldehyde during hemodialysis due to formaldehyde remaining in the system after disinfection with formaldehyde has been described in a case report (WHO, 1989). However, no data were given on the amount of formaldehyde applied. Furthermore, this type of sensitization seems to be very rare.


In the study of Kunisada et al. (2002) anaphylaxis due to formaldehyde released from root-canal disinfectant has been reported. The results of this study suggested anaphylaxis due to Type-I alIergy to formaldehyde.


 


SUMMARY


Classification and labelling of active substance according to Directive 1999/45/EC: R43 IARC (2006) presented a summary on toxic effects in animals as well as a summary on toxic effects in humans including data on sensitisation.


For risk assessment it is proposed to take an area dose of 3 µg/cm² in subjects allergic to formaldehyde and of 37 µg/cm² for induction of formaldehyde contact dermatitis based on human data. Based on animal data (LLNA test with acetone) an area dose of 135 µg/cm² for the EC3 and of 15 µg/cm² for the NOAEC for induction is proposed. Division by 3 to take into account repeated exposure would give an EC3 of 45 µg/cm² and a NOAEC of 5 µg/cm².


 


 

Respiratory sensitisation

Link to relevant study records

Referenceopen allclose all

Endpoint:
respiratory sensitisation: in vivo
Type of information:
experimental study
Adequacy of study:
key study
Study period:
not specified
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
study well documented, meets generally accepted scientific principles, acceptable for assessment
Qualifier:
no guideline available
Principles of method if other than guideline:
Male Balb/C mice were exposed to the test substance and the role of air humidity and allergic sensitization on the acute airway response were investigated.
GLP compliance:
not specified
Specific details on test material used for the study:
TEST MATERIAL
- Formaldehyde was supplied by Kin-Tek (TX, USA) as a certified Trace Source™ permeation tube with paraformaldehyde
Species:
mouse
Strain:
Balb/c
Sex:
male
Details on test animals or test system and environmental conditions:
TEST ANIMALS
- Source: Taconic, Denmark
- Housing: Polypropylene cages (380 × 220 × 150 mm) with pinewood sawdust bedding
- Diet: Altromin no. 1324, Altromin, ad libitum
- Water: ad libitum

ENVIRONMENTAL CONDITIONS
- Temperature: 20 - 24°C
- Humidity: 30 - 50 %
- Photoperiod: 12 hours / 12 hours
Route of induction exposure:
other: intraperitoneal and inhalation
Route of challenge exposure:
inhalation
Vehicle:
other: air
Concentration:
Target concentrations: 0.4, 1.8 and 7 ppm, Measured concentrations: similar under dry and humid environments both at the low (0.42 ± 0.01 ppm) and medium (1.8 ± 0.09 ppm) test substance levels, at the highest test substance level, the concentrations were 4.0 ppm and 5.7 ppm in the dry and humid environments, respectively
No. of animals per dose:
30
Details on study design:
SENSITISATION
- Mice (n = 30) were immunized to OVA by intraperitoneal (i.p.) injections of 1 μg OVA in combination with 270 μg Al(OH)3 in 100 μL 0.9 % saline on day 0
- Mice were boosted i.p. on days 14 and 21 with 0.1 μg OVA in 100 μL 0.9 % saline
- Finally, the animals were exposed 20 min to an aerosol of 0.2 % OVA on days 28 and 29 using a Pari Star nebulizer
- For the non-sensitized control mice (n = 30), saline (0.9 %) was used for the three i.p. injections and the aerosol exposures
- On days 29 and 30, the sensitized and non-sensitized mice were housed at either low (b10) or high (85 - 89) % relative humidity

EXPOSURE
- On day 31, the mice were exposed to the test substance
- Test substance was generated from a Kin-Tek (TX, USA) gas standard generator (Model 491MB) by use of a permeation tube and dry air, and led to a 24 L exposure chamber
- The airflow rates in the chamber were set at 18.8 - 23.2 L/min
- The chamber exposure concentrations of the test substance were monitored pre- and post-test substance exposures and every 10 min during exposures by 10 min air sampling of 4.4 L on dinitrophenylhydrazine (DNPH) sampling cartridges
- Samples were analyzed by HPLC using a diode array detector
- Mice were inserted into body plethysmographs that were connected to the exposure chamber
- The respiratory parameters were obtained for each mouse from a pneumotachograph connected to each head-out plethysmograph that allows continuously monitoring of the parameters
- The exposures were preceded by a period that allowed the mice to adapt to the plethysmographs
- A 15 min period was used to establish baseline (control) values of the respiratory parameters
This period was followed by a 60 min exposure period and a 30 min recovery period

OTHER
- Bioassay: the respiratory effects were studied in a mouse bioassay, which allows the detection of respiratory effects on the upper airway (sensory) irritation as well as effects in the conducting airways and at the alveolar level
- Respiratory parameters: continuous computerized monitoring of the breathing pattern of unanesthetized mice was used to study the following effects
- Sensory irritation pattern: sensory irritants decrease the respiratory rate (f, breaths per min) in mice due to a reflex causing a break at the end of the inspiratory phase
- Airflow limitation: narrowing of the bronchi causes increased resistance that reduces the expiratory flow (VD, mL/s)
- Pulmonary irritation: may be quantified by “time of a pause” (TP, s), which is characterized by a short break at the end of expiration
- Assessment of airway inflammation: bronchoalveolar lavage (BAL) was performed to assess whether the exposures caused inflammation in the lungs
Challenge controls:
yes
Positive control substance(s):
not specified
Negative control substance(s):
not specified
Results:
Lung inflammation:
No effect was seen on test substance exposure and humidity on the degree of lung inflammation. Only the sensitization procedure increased the number of BAL cells.

Airway effects of FA in sensitized mice:
A) Humid environment:
- In the humid environment, little effect of allergic sensitization was seen in the upper respiratory tract, as suggested by the superimposed Time of Break (TB) responses. As typical for nose irritation, the maximum response was reached within the first 20 min of exposure period and thereafter gradually reduced during the remaining part of the period. After cessation of exposure (at time = 60 min), the TB response rapidly approached the baseline level. No apparent difference was revealed by comparison of the AUCs from the OVA and saline sensitized mice, respectively. Similar results were seen at the two lowest FA concentrations.
- Sensitization to OVA amplified the effect of FA in the conducting airways at the highest FA concentration but not at lower concentrations.
- The stronger response from the lungs at the higher FA concentrations was also to some extent reflected by a marginally elongated Time of Pause.
B) Dry environment:
- Saline control mice had a high, short-lasting peak in TB, suggesting a vigorous, transient nose irritation
- OVA-sensitization reduced the airway responsiveness to FA in the conducting airways compared to the saline control mice. Thus, both at the medium and high FA exposure concentrations, the saline sensitized control mice showed a more pronounced reduction in VD compared to the OVA-sensitized mice.

Effect of relative air humidity on airway response to FA
A) Saline control mice: No effect of humidity was apparent on nasal irritation in the non-sensitized mice even at high concentrations of FA
B) The humidity had an impact on the conducting airways at FA >1.8 ppm. Hence, mice housed in a dry environment had a more pronounced decrease in expiratory flow at the highest FA concentration compared to mice housed in the humid environment. The amplifying effect of a dry environment on lower airway effects was also apparent from the TP; it showed that the highest FA concentration induced a stronger pulmonary irritation in the dry environment compared to the humid environment.
Positive control results:
not specified
Negative control results:
not specified
Interpretation of results:
GHS criteria not met
Conclusions:
Formaldehyde induced airway irritation was both influenced by air humidity and prior allergic airway inflammation, but only at high concentrations of ≥ 1.8 ppm. There was no interaction at 0.4 ppm, a concentration relevant for occupational and indoor exposures.
Executive summary:

In a reliable study, the role of air humidity and allergic sensitization on the acute airway response to inhaled formaldehyde vapor was investigated in mice. Mice were sensitized to the immunogen ovalbumin (OVA) by three intraperitoneal injections followed by two aerosol challenges. Once sensitized, the mice were housed at high (85–89%) or low (<10%) relative humidity, respectively for 48 h prior to a 60-min exposure to either 0.4, 1.8 or about 5 ppm formaldehyde. Before, during and after exposure, breathing parameters were monitored.


Formaldehyde induced airway irritation was both influenced by air humidity and prior allergic airway inflammation, but only at high concentrations of ≥ 1.8 ppm. There was no interaction at 0.4 ppm, a concentration relevant for occupational and indoor exposures.

Endpoint:
respiratory sensitisation: in vivo
Type of information:
experimental study
Adequacy of study:
key study
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
study well documented, meets generally accepted scientific principles, acceptable for assessment
Qualifier:
no guideline available
Principles of method if other than guideline:
Male Balb/C mice were exposed to the test substance, airway hyper-responsiveness, pulmonary tissue damage, eosinophil infiltration and cytokine expression were investigated
GLP compliance:
not specified
Specific details on test material used for the study:
TEST MATERIAL
- Source: Sigma, ST. Louis, MO, USA
- 10 % formaldyhyde
Species:
mouse
Strain:
Balb/c
Sex:
male
Details on test animals or test system and environmental conditions:
TEST ANIMALS
- Source: Hubei Experimental Animal Center, Wuhan, China
- Age at study initiation: 7 - 8 weeks
- Weight at study initiation: 18 - 20 g
- Housing: pathogen-free room
- Diet: standard rodent chow, ad libitum
- Water: distilled water, ad libitum

ENVIRONMENTAL CONDITIONS
- Temperature: 20 - 25°C
- Humidity: 50 - 70 %
- Photoperiod: 12 hours / 12 hours
Route of induction exposure:
inhalation
Route of challenge exposure:
inhalation
Vehicle:
other: air
Concentration:
0.5 and 3.0 mg/m³
No. of animals per dose:
6
Details on study design:
PREPARATION OF GASEOUS TEST SUBSTANCE
- To perform whole-body exposures, gaseous test substance at pre-set concentrations of 0.5 and 3.0 mg/m³ (based on Chinese Occupational Threshold Limit Values) was generated with 10 % test substance that was diluted with filtered house air prior to delivery into a small exposure chamber
- Animals were placed in a glass vessel (exposure chamber) connected to the generator and exposed to the gaseous test substance
- During the experiment, exposure parameters were controlled automatically based on chamber readings. Air temperature, humidity, and ventilation rate were maintained at 22.5 - 23.5°C, 40 - 45 %, and 1.0 L/min

CHALLENGE PROTOCOL
- Animals were randomly allocated into six groups: (1) saline control; (2) ovalbumin (OVA)-immunized (OVAimm) only; (3) 0.5 mg test substance/m³ exposure; (4) OVAimm + 0.5 mg test substance/m³; (5) 3.0 mg test substance/m³ test substance exposure; and, (6) OVAimm + 3.0 mg test substance/m³
- In the OVA-immunized only (OVAimm) group, each animal was injected intravenously on days 10 and 18 with 200 μL of a normal saline solution containing 20 mg OVA and 20 mg aluminum hydroxide each time, followed by an aerosol challenge with 1 % OVA for 30 min/d on Days 22 - 28 using an ultrasonic nebulizer
- The saline control group was treated in the same manner, but exposed to saline instead of OVA on Days 10, 18, and 22 - 28
- To determine the effect of allergen-combined test substance exposure, the OVAimm + 0.5 mg test substance/m³ (FA0.5+OVA) and OVAimm + 3.0 mg test substance/m³ (FA3.0+OVA) groups were treated the same as the OVAimm only group, except that the test substance exposure (6 h/d) period varied from Day 1 to 21
- In the test substance-0.5 and test substance-3.0 only groups, saline was substituted for OVA
- To prevent adversely affecting measures of lung cytokines/pulmonary architecture due to any possible influence from the methacholine used in airway hyperresponsiveness (AHR) assessments, dedicated groups of animals (for each of the above regimens) were employed
- One set of 36 animals was treated for 28 days according to the above-regimen(s) and then used for AHR tests and BALF sample collection; a second set of 36 animals was treated as above and then used directly (i.e., non-lavaged) for cytokine (left lung) and histopathologic (right lung) analyses

AIRWAY HYPER-RESPONSIVENESS MEASUREMENTS
- Airway hyper-responsiveness as an indicator of lung function was tested 24 h after the final aerosol exposure using an AniRes 2005 lung function system
- Animals were anesthetized and then connected to a computer-controlled ventilator via a cannula that had been surgically implanted in the trachea
- The respiratory rate was pre-set at 95 times/min and the expiration/inspiration time ratio was set at 1.5:1
- When the baseline parameters were stable, a 27-gauge needle was inserted into the jugular vein and methacholine was administered through a catheter; increasing dosages of 0, 0.025, 0.05, 0.1, and 0.2 mg MCh/kg were injected every 5 min
- Animals airway responsiveness in each dosing cycle was examined by measuring expiratory resistance (Re), inspiratory resistance (Ri), and dynamic lung compliance (CLdyn)
- The relative area (R-area) was defined as the area under the peak curve of Re or Ri and beyond the baseline level during the 250-s timeframe (the missing 30 s to fill out the ‘5 min’ dosing cycle was used to freshly prepare the next MCh dose before its injection)
- The distance from the wave trough to the baseline level of the CLdyn curve was adopted to quantitatively assess lung compliance

BRONCHOALVEOLAR LAVAGE FLUID (BALF) COLLECTION AND CELL COUNTS
- BALF was collected immediately after the conclusion of the airway hyper-responsiveness measurements
- Specifically, after the final 0.2 mg MCh/kg challenges, the animals were removed from the operation box and sacrificed
- Intratracheal cannulation was then performed using a 1 mL syringe; once the cannula was place, 0.5 mL saline was injected into the trachea and lungs
- The chest of each animal was gently massaged and rinsed for ca. 1 min, and then the liquid withdrawn with the same syringe
- The process was repeated three times; the total quantity of BALF was 1.2 - 1.3 mL for each sample
- Cell samples in the BALF were collected by centrifugation (1000 rpm, 10 min, 4°C) and re-suspended in 1 mL saline
- Total cell counts were obtained from 500 µL of the cell suspension using a blood cell analysis system
- Another 500 μL was prepared for eosinophil counts
- Eosinophils were stained using Eosin Y sodium solution, and then assessed using a hemacytometer and a DM 4000B microscope
- Neutrophils or monocytes were not counted due to the limited BALF

PULMONARY HISTOPATHOLOGICAL ANALYSES
- Right lungs from non-lavaged mice were isolated for histopathology slice preparation
- All samples were incubated in fixative (saturated 2,4,6-trinitrophenol:formalin:glacial acetic acid [15:5:1 v/v/v]) for 24 h at room temperature, then cut into pieces, and stained with hematoxylin and eosin
- The stained pieces were then embedded in paraffin, sectioned into 10-μm slices and observed using the DM 4000B microscope
- Results were then reported in terms of slight, moderate, or severe changes
- Specifically, the designation: ‘Slight’ refers to a slight thickening of the airway wall without damage/folding on the cuticular layer of the airway, or cell infiltration around airway and vessels; ‘Moderate’ refers to a bit of thickening of the airway wall and there some folding on the cuticular layer, and a little cell infiltration around airway and vessels; and, ‘Severe’ refers to obvious thickening of the airway wall and severe damage/ folding on the cuticular layer, with mucus inside the airway and great amounts of cell infiltration around airways and vessels

DETECTION OF THE CYTOKINE LIVELS
- IL-4, IL-6 and Interferon (IFN)-gamma by ELISA

STATISTICAL ANALYSIS
- Both a oneway analysis of variance (ANOVA) and a Student’s t-test were applied to evaluate significant differences
- A p value < 0.05 was used to evaluate statistical significance
Challenge controls:
yes
Positive control substance(s):
none
Negative control substance(s):
not specified
Results:
ASSESSMENT OF AIRWAY HYPER_RESPONSIVENESS
- The results show that Re and Ri rose in a dose-dependent manner, while CLdyn decreased because of an increase in the distance from the wave through to the baseline level
- All of the OVA-treated animals, including those in the OVA only, test substance 0.5 + OVA, and test substance 3.0 + OVA groups, exhibited greater airway responses to MCh when compared with those of the saline controls
- The test substance 0.5 + OVA and test substance 3.0 + OVA groups had a significantly enhanced airway response compared with the OVA-immunized only group and the saline control group
- The test substance 3.0 + OVA animals showed a very marked change in their airway responses compared with mice in the OVAimmunized-only group, resulting in an extreme decline in lung function
- The two groups exposed to test substance only (test substance 0.5 and test substance 3.0) did not display any significant changes (relative to controls), even though their responsiveness was higher than that of the saline animals but lower than that of the OVA-immunized-only animals

PULMONARY HISTOPATHOLOGY
- Histopathological changes associated with changes in lung function (re: airway hyper-responsiveness) were assessed
- There were clear pulmonary histopathological changes, including inflammatory cell infiltration, airway wall thickening, and remodeling (each of which are associated with asthma in humans) in all OVA-treated groups (i.e., OVA-immunized-only, test substance 0.5 + OVA and test substance 3.0 + OVA)
- The most severe changes, including cell infiltration and airway remodeling, were found in the test substance 3.0 + OVA animals
- There were slight changes observed in the test substance 3.0 group compared with the saline control group

EOSINOPHIL INFILTRATION IN BALF
- The ratios of eosinophils to total cells contained in BALF of the test substance exposure groups (i.e., test substance 0.5, test substance 0.5 + OVA, test substance 3.0, and test substance 3.0 + OVA) were all significantly increased compared with ratios among the saline controls
- OVA immunization significantly intensified the increased eosinophil infiltration induced by test substance (p < 0.01)

CYTOKINE DETECTION
- Both test substance exposure and OVA-combined test substance exposure enhanced IL-4 and IL-6 levels in pulmonary tissue homogenates
- Test substance 3.0 exposure, but not test substance 0.5 exposure, induced a significant increase in IL-4 and IL-6 levels, even compared with the OVA immunized only group
- Among all the groups, test substance 3.0 + OVA exposure induced the highest IL-4 and IL-6 levels
- IFNγ levels were not significantly different among the groups
Positive control results:
not specified
Negative control results:
not specified
Interpretation of results:
GHS criteria not met
Conclusions:
Exposure can induce and aggravate asthma in Balb/c mice when it is combined with ovalbumin immunization.
Executive summary:

In a reliable study, the underlying role of formaldehyde exposure in occupational asthma, especially when it is combined with allergen exposure, was explored. Balb/c mice were randomly divided into six groups (n = 6/group): (1) saline control; (2) ovalbumin (OVA)-immunized (OVAimm) only; (3) 0.5 mg formaldehyde/m³ exposure; (4) OVAimm + 0.5 mg formaldehyde/m³; (5) 3.0 mg formaldehyde/m³ formaldehyde exposure; and, (6) OVAimm + 3.0 mg formaldehyde/m³. These low and high exposure formaldehyde levels were adopted from current (0.5 mg/m³) and original (3.0 mg/m³) Chinese Occupational Threshold Limit Values. Experiments were conducted after 3 week of combined exposure and a 1-week challenge with aerosolized OVA.


Airway hyper-responsiveness, pulmonary tissue damage, eosinophil infiltration, and increased interleukin (IL)-4 and IL-6 levels in lung tissues were found in the OVA + 3.0 mg formaldehyde/m³ hosts as compared to values seen in the OVA-immunized only mice.


The results here suggest that formaldehyde exposure can induce and aggravate asthma in Balb/c mice when it is combined with OVA immunization.

Endpoint conclusion
Endpoint conclusion:
no adverse effect observed (not sensitising)
Additional information:

RESPIRATORY SENSITISATION IN EXPERIMENTAL ANIMALS


Test methods to reliably differentiate between skin and respiratory sensitizers and to identify the latter are still under development. MAK (2010) has summarized the results obtained for formaldehyde. Notwithstanding the still exploratory status of these protocols, for formaldehyde no evidence for specific induction of airway sensitization can be derived from the animal studies.


Larsen et al. (2013) tested acute airways effects of formaldehyde in mice sensitized by 3 intraperitoneal injections of ovalbumin on day 0, 14, and 21. On day 28 and 29 the animals were exposed to an aerosol of ovalbumin (20 min) and on day 31 to formaldehyde for 1 h (0.4, 1.8, and 5 ppm). On day 29 and 30 the mice were hold under conditions of high (85-89%9 and low (<10%) relative humidity. Before, during, and after exposure breathing parameters of the upper respiratory tract and specific markers of lung and nose irritation were measured. Non-immunised animals served as controls. Formaldehyde induced airway irritation was both influenced by air humidity and prior allergic airway inflammation, but only at high concentrations of ≥1.8 ppm. There was no interaction at 0.4 ppm, a concentration relevant for occupational and indoor exposures.


Liu et al. (2011) exposed mice at 0, 0.5, and 3.0 mg/m³ over 21 days. On day 10 and 18 ovalbumin was injected intravenously for induction of sensitisation and on days 22-28 the animals were challenged with aerosolised ovalbumin. Control groups with and without formaldehyde inhalation and ovalbumin induction were included. The method followed that of Qiao et al. (2009) for rats (see 7.9.2). After the last challenge exposure lung function was measured in trachea-cannulated mice after intravenous application of O-acetyl-ß-methacholine and bronchi-alveolar lavage fluid was collected. Histopathology of the lung was performed and various cytokines in the lung were measured. At the high formaldehyde concentration airway hyper-responsiveness, pulmonary tissue damage, eosinophil infiltration and increased interleukin IL-4 and IL-6 in lung tissue were found in the ovalbumin + formaldehyde treated animals as compared to mice only immunised to ovalbumin. The authors conclude that formaldehyde can induce and aggravate asthma in mice.


Wu et al. (2013) carried out a mechanistic study to identify the mechanism and receptors involved in the asthma model of Liu et al. (2011). A similar experimental protocol was used with a formaldehyde exposure scheme of 5 d/week (6 h/d) over 4 weeks, ovalbumin induction on days 10, 18, and 25, and challenge on days 29-35. During formaldehyde exposure HC-030031 (ankyrin 1 receptor antagonist) and capsazepine (vanilloid 1 receptor antagonist) were injected daily by the intraperitoneal route. Control groups without formaldehyde, ovalbumin or the receptor antagonists were included. An extended set of parameters was measured as compared to that of Liu et al. (2011) including immunohistochemistry, cytokines and neuropeptides. Regarding the combined action of formaldehyde and ovalbumin the results of Liu et al. (2011) were confirmed and the pulmonary reaction was reduced by the receptor antagonists.


In summary, these studies cannot be taken as an experimental proof that formaldehyde may lead to respiratory sensitisation. Some of these investigations are barely acceptable because they used an exposure procedure without analytical verification of formaldehyde in the exposure system. But above all, none of these studies used appropriate control substances that may lead to dermal sensitisation with or without being respiratory sensitizers.


Lee et al. (1984) exposed guinea pigs by the inhalation, dermal, and injection (into footpads with Freund’s complete adjuvant) routes. Pulmonary hypersensitivity was assessed by measuring a potential increase in respiration rate. No pulmonary hypersensitivity was detected by challenge with 2 or 4 ppm formaldehyde for any of the induction routes and formaldehyde-serum albumin adducts were not observed after inhalation treatment. On the other hand, dermal sensitization was observed by all of the different induction routes. Formaldehyde led to skin sensitization in guinea pigs without causing respiratory hypersensitivity.


Hilton et al. (1996) exposed mice (n = 6 per group) twice at day 1 and 7 via the skin to irritant concentration of formaldehyde solutions for induction. At day 14 the animals were killed and serum IgE concentration (indicator for respiratory allergy) was measured. In contrast to the positive control TMA (trimellitic anhydride), a known human respiratory allergen, the IgE concentrations in mice exposed to formaldehyde solutions or to the contact allergen DNCB (2,4-dinitrochlorobenzene) were not increased. It is concluded that formaldehyde lacks a significant potential to cause sensitization of the respiratory tract.


Arts et al. (1997) revealed similar results. In rats, topical application of formaldehyde did not lead to an increase in serum IgE concentration in contrast to the effects observed after trimellitic anhydride treatment. But according to the authors further studies may be needed with chemicals that have both irritant and sensitizing properties at about similar concentrations or may act through non-IgE-mediated immune mechanisms.


Arts et al. (2008) developed a respiratory local lymph node assay. Mice were exposed to the test chemical by repeated inhalation and 3 days after the last exposure cell proliferation was determined in the draining mandibular lymph nodes. While several respiratory sensitizers led to an increased proliferation rate, formaldehyde did not.


Several authors investigated cytokine profiles as an indicator for respiratory sensitization.


Hilton et al. (1996) applied 50 μL of the test solution (10, 25, 50 % formalin in DMF vehicle or 25 % trimellitic anhydride [positive control] in vehicle) on both shaved flanks of mice (n = 10 per group) and 5 days later the treatment was repeated. Further 5 days later, 25 μL of the same test solution was applied to the dorsum of both ears, once daily for 3 consecutive days. Thirteen days after initiation of exposure, mice were sacrificed and draining auricular lymph nodes were excised and pooled for each group. Single cell suspensions were prepared of lymph node cells (LNC). Concentrations of 2 different cytokines, interferon gamma and interleukin 10, in the cell culture supernatants of draining lymph node cells were determined by sandwich ELISA (enzyme-linked immunosorbent assay) method. Formalin provoked the production of interferon gamma but did not increase the interleukin 10 concentration. The reverse pattern was detected with the positive control TMA: slight increase in interferon gamma but a clear increase in interleukin 10. These data suggested that formaldehyde did not induce IgE response in mice and that no sensitization of the respiratory tract in mice is induced.


Dearman et al. (1999) made a similar experiment. Mice were treated dermally with formaldehyde, 2,4-dinitrochlorobenzene and trimellitic anhydride. Production of different cytokines (interferon-gamma, interleukin-4 and-10) was determined in the local lymph nodes. A similar cytokine profile was obtained for formaldehyde and 2,4-dinitrochlorobenzene, clearly different to that of trimellitic anhydride. The authors concluded that formaldehyde does not elicit significant allergic sensitization to the respiratory tract.


Dearman et al. (2005) compared frequencies of intracellular cytokine (IL-4 and IFN-gamma)-positive CD4+and CD8+lymphocytes in the draining lymph nodes of mice after topical treatment. The reactions to two contact sensitizers (dinitrochlorobenzene and formaldehyde) and two respiratory sensitizers were compared and both types showed different patterns. The respiratory sensitizers led to a much higher frequency of IL-4-expressing CD4+cells as compared to the contact sensitizers. Both types led to a similar increase in IFN-gamma positive CD4+and CD8+lymphocytes. These different patterns might be used to discriminate skin and respiratory sensitizers.


Xu et al. (2002) investigated the mRNA expression of different cytokines in spleen and draining lymph nodes of mice after cutaneous exposure to formaldehyde. They suggested that the induction of IFN-gamma and IL-4 are common immunological features of contact allergens.


Persoz et al. (2009) exposed human alveolar epithelial cells in vitro for 30 min to 200 µg/m³, which leads to a mitochondrial activity reduction of 30 %.


In summary, even under consideration that generally accepted animal testing procedures for defining respiratory sensitizers are not yet available, the data from animal experiments indicate that formaldehyde does not act as a respiratory sensitizer.


 


RESPIRATORY SENSITISATION IN HUMANS


There are only a few case reports of bronchial asthma (2 renal dialysis nurses, a plastic moulder, a printer, a worker in a phenol formaldehyde manufacturing plant, and a carpenter) indicating respiratory tract sensitization. In acute exposure challenges at exposure levels of 3 ppm formaldehyde, all above mentioned subjects showed marked changes in FEV1 or airflow rates. In a study on 230 patients occupationally exposed to formaldehyde and who had reported respiratory symptoms consistent with asthma only 12 subjects showed a decrease inPeak Expiratory Flow Rate (PEFR) > 15% after acute challenge with 2 ppm formaldehyde. However, the mechanism of sensitization in this study is uncertain (ATSDR, 1999; WHO, 1989). In contrast to these positive results no challenge-induced deficits in FEV1 or airflow rates were demonstrated in 3 other studies on formaldehyde exposed subjects with respiratory problems (ATSDR, 1999).


The formation of formaldehyde-specific IgE antibodies in groups of formaldehyde-exposed subjects has been investigated in several studies. In general, the results did not provide evidence for a formaldehyde-induced respiratory allergy. However, there is some evidence for formation of specific IgE in children exposed to low level of formaldehyde in indoor air (Wantke et al., 1996; in ATSDR, 1999).


In contrast, Doi et al. (2003) concluded that it appears unlikely, that formaldehyde is a relevant allergen in childhood asthma.In this study specific IgE against formaldehyde was measured by CAP RAST in 122 asthmatic children and 33 nonallergic children in. A questionnaire on clinical features of their asthma, their living conditions, and symptoms of mucosal irritation was performed. The personal factors were similar in the two groups. The median level of total IgE of asthmatics (755 IU/mL) was significantly higher than that of controls (108 IU/mL). However, formaldehyde-specific IgE was detected in only two asthmatic children: one exhibited 0.42 UA/mL of formaldehyde-specific IgE and the other 0.46 UA/mL. Both values are only slightly increased compared with the limit given for positive results of 0.35 UA/mL.


One of these 2 children had severe asthma and frequent complaints of mucosal irritation in places where the formaldehyde level was likely to be high. On the other hand, the other child had mild asthma and rare complaints of mucosal irritation. In contrast, it is also stated that both patients did not have severe asthma or frequent symptoms of mucosal irritation. There was no significant difference between the control group and the group of asthmatic children in prevalence of IgE to formaldehyde. These results suggest that the prevalence of IgE to formaldehyde may be rare in Japanese children and FA may not play an important role as an allergen causing asthma or as an irritant.


MAK (2010) reviewed the literature on formaldehyde and asthma and concluded that the allergological findings do not provide a consistent pattern. In inhalation challenge tests generally immediate reactions were observed, dual or late reactions were rare. A differentiation against irritation often is difficult and specific IgE antibodies, if found, mostly did not correlate with the respiratory symptoms. Overall, a relationship of respiratory symptoms with formaldehyde is only in few cases sufficiently documented. The small number of reliable findings in comparison to the broad exposure potential would not warrant to classify formaldehyde as a respiratory sensitizer according to the criteria of the MAK commission.


Recently a theory was developed how formaldehyde might possibly exacerbate asthma and induce bronchoconstriction through influencing thiol biology (Thompson and Grafström, 2008; Thompson et al., 2008). This theory is based on the fact that that formaldehyde dehydrogenase (FDH) has a dual role in formaldehyde metabolism and also partially regulates nitrosothiol homeostasis by catalyzing the reduction of the endogenous nitrosylating agent S-nitrosoglutathine (GSNO). Thereby formaldehyde may influence FDH mediated catabolism of GSNO which acts as an endogenous bronchodilator. But taking into consideration that neither animal experiments nor observations in humans lend sufficient support to an asthmagenic action of formaldehyde, this hypothesis remains speculative.


Wolkoff and Nielsen (2010, key) reviewed the studies related to effects of formaldehyde on asthmatics especially at low exposures of <100 ppb. After a detailed analysis of the conflicting results in the volunteer studies reported by Ezratty et al. (2007) and Casset et al. (2006) they concluded that exposure of asthmatics to formaldehyde would not lead to an exacerbation of the lung function. They also reviewed case control and cross sectional studies that had suggested a possible association of low formaldehyde concentrations and asthma. By taking into consideration the complex exposure situations and potential confounding factors, they concluded that at exposures <100 ppb such associations between children and adults in homes and schools have generally not been convincing mainly due confounding factors and susceptibility of the findings to chance effects. Furthermore, they did not identify major differences between children, adults, and asthmatic.


This conclusion was upheld in a follow-up review of the same group (Nielsen et al., 2013, key) again concluding that associations between formaldehyde exposure and exacerbation of asthma have not been convincingly demonstrated in children or adults in schools or at home, corresponding to an assessment of Golden (2011, key).


This was also a conclusion of the review of Heinrich (2011, key) who noted that instead consistent environmental risk factors were tobacco smoke, living close to busy roads and in damp homes with visible mould.


Schram-Bijerk et al. (2013, key) came to similar conclusions. They developed a method to estimate the burden of disease (BoD) by the most important indoor air pollutants, namely dampness, carbon monoxide, radon/thoron, formaldehyde and environmental tobacco smoke in The Netherlands. Following a method developed by WHO (Öberg M. Jaakkola MS. Prüss-Üstün A. et al. Second-hand smoke: assessing the environmental burden of disease at national and local levels. Geneva: World Health Organization. 2010) they took into account health outcomes and estimates for exposure levels, % of population exposed, exposure-effect relationships, severity of effect, and duration of disease. The health effect under consideration for formaldehyde was asthma in children. The largest BoD was attributed to environmental tobacco smoke followed by radon and thoron from soils and building materials, dampness and carbon monoxide. According to their estimate, formaldehyde did not contribute to the BoD. They further compared their analysis with BoD estimates from a previous study (EnVIE; de Oliveira Fernandes E, Jantunen M, Carrer P. et al. ENVIE-co-ordination action on indoor air quality and health effects. Publishable final activity report. Brussels: European Commission, 2009). The EnVIE estimates were 3 times higher than those of the authors. The most important factors according to EnVIE were combustion products from outdoor sources, bio-aerosols due to dampness and by outdoor sources, volatile organic compounds, radon from soils, pathogens, and carbon monoxide.


Many of the newly identified studies have already been discussed in detail by Nielsen et al. (2013) and the assessments here basically rely on those of Nielsen et al. (2013).


A meta-analysis of McGwin et al. (2010, supporting) with two different models (fixed-effects and random effects model) gave an OR of 1.03 (95 % CI 1.02–1.04) and of 1.17 (95 % CI 1.01–1.36), respectively.


In a cross-sectional study among 2,453 Korean school children respiratory symptoms were obtained by means of a questionnaire. The mean formaldehyde concentration was 28 µg/m³ in class rooms and the outdoor level was 4.3 µg/m³. There was no association between class room FA exposures and wheeze, asthma, but wheeze was significantly associated with reported indoor dampness and mold growth in the home environments (Kim et al. 2011, supporting).


In a cross-sectional study comprising 6,590 children, past-year rhinoconjunctivitis and past year asthma incidences were obtained by questionnaire. Exercise-induced asthma and skin prick test reactions were investigated. A significantly increased odds ratio (OR: 1.19) was observed for past-year rhinoconjunctivitis at formaldehyde exposures exceeding 28.4 µg/m³, but not for past-year asthma (OR: 0.90). Exercise-induced asthma was not correlated with FA exposure. The authors concluded that school air quality may affect rhinitis (formaldehyde) and asthma-related (particulates, acrolein and NO2) morbidity (Annesi-Maesano et al. 2012, supporting).


Recently, German UBA (2016) performed a comprehensive review of epidemiological studies investigating the association between formaldehyde exposure and the induction or exacerbation of asthma in children including several of the studies mentioned above. UBA concluded that there is no clear association between formaldehyde exposure in the indoor environment and asthma in children. It was stated that the above mentioned epidemiological studies (e.g. Krzyzanowski et al., 1990; Annesi-Maesano et al. 2012 etc.) suffer from small sample size, implausible formaldehyde concentrations, and the fact that other substances or factors initiating asthma and asthma-like complaints were not adequately considered. Results derived from controlled human exposure studies as well as animal experiments support their opinion.

Justification for classification or non-classification

According to current knowledge and application of the criteria given in Annex I of Regulation (EC) No. 1272/2008, the classification Skin Sens 1A, exceeding the classification given in Regulation (EC) No. 1272/2008, Annex VI, Table 3.1 is required.