Proposed residential indoor air quality guidelines for formaldehyde

4. Indoor Concentrations

4.1 Indoor Formaldehyde Concentrations in Canada

Some surveys of indoor formaldehyde were carried out in Canada in the 1980s following complaints and/or to assess the exposure arising from urea formaldehyde foam insulation (UFFI) and formaldehyde-containing wood products (Williams et al. 1981; Broder et al. 1988a; 1988b; 1988c). However, little Canadian data were collected after the UFFI ban in 1980 and the adoption of voluntary formaldehyde emission standards by particle board and MDF producers in the early 1990s (Table 3). These studies were all carried out for specific purposes (e.g. descriptive indoor air quality surveys) and covariate data were collected accordingly. None of these studies, therefore, presents a comprehensive picture of factors associated with indoor levels of formaldehyde.

Table 3. Indoor formaldehyde concentrations in Canada

Place and time	Houses	Sampling time and flow	Range (μg/m3)	Mean or median (μg/m3)	Reference
Quebec Feb.-Apr. 1995	73 apartments within 10 mid-rise residential buildings	5 to 7 days	25 to 86	37	Consortium Dessau-Siricon 1996
Vancouver, Ottawa and Toronto	24 apartments from 8 mid-rise buildings	7-day	12 to 74		Scanada Consultants Limited 1997
Windsor (Ontario) 1991-92,	22 homes where all inhabitants were non-smokers, 12 homes in which there was at least one smoker, and 18 offices/hotels where smoking was prohibited	24-hour samples	smoke-free: homes: 2.5-59.5 smoker homes: 6.6-107.2 Smoke-free offices: 5.9-87.0	Medians: smoke-free homes 22.8 smoker homes: 31.4 smoke-free offices: 14.1 Means smoke-free homes 27.1 smoker homes: 39.4 smoke-free offices: 17.6	Bell et al. 1994
Québec and surrounding towns	34 homes with a fireplace or a wood stove and 6 homes with no wood-burning appliance. No other combustion source (smoker, furnace, or attached garage) present.	24-hour active sampling at 0.4 L/min on the ground floor (in houses with a wood-burning appliance, samples collected while appliance in use)	houses with wood-burning appliance: 23.4 houses without wood-burning appliance: 19.5	Houses with wood-burning appliance: 8.2 (sd 4.6) Houses without wood-burning appliance: 9.9 (sd 5.5)	Lévesque et al. 2001
Prince Edward Island Jan.-Apr. 2002	55 homes where no smoker lived, and 4 homes inhabited by at least one smoker (total 59)	19.5 to 57.2 hours (median 23.8 hours) at 0.1 L/min	smoke-free homes: 5.5-87.5 smoker homes: 22.7-70.8	Medians no smokers 29.6 smokers: 38.2	Gilbert et al. 2005

In the 1990s, the Canadian Mortgage and Housing Corporation (CMHC) funded some surveys of air quality in residential buildings. In 73 apartments within 10 mid-rise residential buildings in the province of Quebec tested from February to April 1995, formaldehyde concentrations (sampling time 5 to 7 days) ranged from 25 to 86 μg/m³ , with a mean level of 37 μg/m³ (Consortium Dessau-Siricon 1996). In 24 apartments from 8 mid-rise buildings in Vancouver, Ottawa and Toronto, 7-day formaldehyde concentrations ranged from 12 to 74 μg/m³ (Scanada Consultants Limited 1997).

In Windsor in 1991-92, the Ontario Ministry of Environment and Energy measured indoor formaldehyde concentrations in 22 homes where all inhabitants were non-smokers and in 12 homes in which there was at least one smoker. Formaldehyde levels ranged from 2.5 to 59.5 μg/m³ (median 22.8 μg/m³ ) in smoke-free homes, and from 6.6 to 107.2 μg/m³ (median 31.4 μg/m³ ) in smoker homes. Formaldehyde levels were also measured in 18 offices and hotels where smoking was prohibited, and ranged from 5.9 to 87.0 μg/m³ (median 14.1 μg/m³ ) (Bell et al. 1994).

In Québec City, the Direction de la santé publique (DSP) measured 24-hour formaldehyde concentrations in 40 homes, of which 34 had a fireplace or a wood stove, and 6 had no wood- burning appliance. No other combustion source (smoker, furnace, attached garage) was present in any of these homes. In the houses with a wood stove, samples were collected while the appliance was in use. The highest formaldehyde concentration measured in that study was 23.4 μg/m³ . Average formaldehyde concentrations on the ground floor of houses with and without wood-burning appliances were 8.2 μg/m³ (SD 4.6 μg/m³ ) and 9.9 μg/m³ (SD 5.5 μg/m³ ), respectively (Lévesque et al. 2001).

In Prince Edward Island in winter 2002, Health Canada sampled 59 homes for 19.5 to 57.2 hours (median 23.8 hours). Samples were collected in the main living room of the homes. Formaldehyde concentrations ranged from 5.5 to 87.5 μg/m³ with a median of 29.6 μg/m³ in homes where no smoker lived (n=55), and from 22.7 to 70.8 μg/m³ with a median of 38.2 μg/m³ in the 4 homes inhabited by at least one smoker (Gilbert et al. 2005).

With the exception of the study of Lévesque et al. (2001) in Québec, results from the 1990s and early 2000s consistently indicate that formaldehyde concentrations in Canadian homes range between 2.5 and 88 μg/m³ with an average between 30 and 40 μg/m³ . The lower levels found in the Québec study may be explained in part by the absence of a combustion source other than wood-burning appliances.

4.2 Determinants of Indoor Formaldehyde Levels

4.2.1 Building characteristics

A Swedish research group studied determinants of indoor formaldehyde levels in Uppsala, Sweden. Two-hour formaldehyde levels were measured in 62 dwellings in 1991-92, and 88 people inhabiting these dwellings completed questionnaires on factors likely to affect exposure, such as building materials, indoor painting in the last 12 months, mechanical ventilation, presence of carpets, and presence of smokers. Formaldehyde concentrations ranged between <5 and 110 μg/m³ , and were higher in houses with wall-to-wall carpets (Norbäck et al. 1995). A logistic regression was performed to assess the association between individual building characteristics and indoor formaldehyde, adjusting for all other significant factors. Wooden house, wall-to-wall carpets and painted wood were independently associated with formaldehyde concentration increments of 7 μg/m³ (95% CI
1-13), 13 μg/m³ (95% CI 4-22) and 16 μg/m³ (95% CI 7-25), respectively. No significant influence of building age, mechanical ventilation or environmental tobacco smoke was found (Wieslander et al. 1997). Also, the association between formaldehyde concentration and classroom furnishing was investigated in 181 classrooms randomly selected in 48 schools. Formaldehyde concentrations were measured over 4 hours, and ranged from <3 to 72 μg/m³ (geometric mean 3 μg/m³ ). Also, in each classroom, a "shelf factor" was calculated as the length of open shelves in relation to room volume and a "fleece factor" as m² of fabrics in relation to room volume. After adjustment for season and air exchange rate, formaldehyde concentrations were positively correlated with the "fleece factor" (p=0.013) and the "shelf factor" (p<0.001). The authors hypothesized that formaldehyde might be adsorbed onto indoor surfaces and re-emitted in the indoor environment (Smedje and Norbäck 2001).

In France, 72-h formaldehyde samples were collected in 61 dwellings located in the Paris region (Clarisse et al. 2003). Geometric mean formaldehyde levels were 21.7 μg/m³ (SD 1.9 μ g/m3) in kitchens, 24.3 μg/m³ (SD 1.9 μg/m³ ) in living rooms and 24.5 μg/m³ (SD 2.0 μg/m³ ) in bedrooms. A multiple linear regression analysis showed that temperature (p=0.01) and the age of floor coverings (p=0.02) were significantly associated with formaldehyde levels, while CO₂ (p=0.36), the type of floor covering (p=0.32), the presence of pressed wood products (p=0.90), the age of wall coverings (p=0.55) and smoking (p=0.30) were not.

4.2.2 Season

In the United Kingdom, indoor formaldehyde levels were measured every 4 to 6 weeks during 3 years in five homes. Mean formaldehyde concentrations in years 1, 2 and 3 were 17, 19 and 17 μg/m³ , respectively; no long-term trend was observed. However, there was a clear seasonal pattern: concentrations measured from April to September were significantly higher (p<0.05) than those measured from October to March (Brown et al. 1995).

4.2.3 Outdoor air supply

In Montréal, the impact of outdoor air supply on levels of indoor air contaminants was determined by manipulating experimentally the mechanical ventilation system of two major office buildings. Among the contaminants monitored, formaldehyde was the most tightly associated with air exchange: 96% of the variance of formaldehyde concentrations was explained by outdoor air supply, compared to 87% of CO₂ variance and less than 30% for all other chemicals monitored (Menzies et al. 1996). A German study of 252 houses sampled from 1986 to 1993 found a significant negative correlation between formaldehyde levels (ranging from 12-649 μg/m³ ) and air exchange rate expressed in hours -1 (r=-0.2105, p<0.01). No significant association was found between formaldehyde and temperature, or between formaldehyde and relative humidity (Salthammer et al. 1995).