Residents with lower socioeconomic status (SES) living in social housing subsidized by the government or public funding agencies may be more exposed to pesticides used indoors because pesticides are applied due to structural defects, poor maintenance, etc.
In 2017, 28 particulate pesticides were measured in indoor air in 46 units of seven low-income social housing apartment buildings in Toronto, Canada, using portable air purifiers that were operated for one week. The pesticides analyzed were traditionally and currently used pesticides from the following classes: organochlorines, organophosphorus compounds, pyrethroids, and strobilurins.
At least one pesticide was detected in 89% of units, with detection rates (DRs) for individual pesticides reaching 50%, including traditional organochlorines and currently used pesticides. Currently used pyrethroids had the highest DFs and concentrations, with pyrethroid I having the highest particulate phase concentration at 32,000 pg/m3. Heptachlor, which was restricted in Canada in 1985, had the highest estimated maximum total air concentration (particulate matter plus gas phase) at 443,000 pg/m3. Concentrations of heptachlor, lindane, endosulfan I, chlorothalonil, allethrin, and permethrin (except in one study) were higher than those measured in low-income homes reported elsewhere. In addition to the intentional use of pesticides for pest control and their use in building materials and paints, smoking was significantly associated with the concentrations of five pesticides used on tobacco crops. The distribution of high-DF pesticides in individual buildings suggests that the main sources of the detected pesticides were pest control programs conducted by building managers and/or pesticide use by occupants.
Low-income social housing serves a critical need, but these homes are susceptible to pest infestations and rely on pesticides to maintain them. We found that 89% of all 46 units tested were exposed to at least one of 28 particulate-phase insecticides, with currently used pyrethroids and long-banned organochlorines (e.g., DDT, heptachlor) having the highest concentrations due to their high persistence indoors. Concentrations of several pesticides not registered for indoor use, such as strobilurins used on building materials and insecticides applied to tobacco crops, were also measured. These results, the first Canadian data on most indoor pesticides, show that people are widely exposed to many of them.
Pesticides are widely used in agricultural crop production to minimize damage caused by pests. In 2018, approximately 72% of pesticides sold in Canada were used in agriculture, with only 4.5% used in residential settings.[1] Therefore, most studies of pesticide concentrations and exposure have focused on agricultural settings.[2,3,4] This leaves many gaps in terms of pesticide profiles and levels in households, where pesticides are also widely used for pest control. In residential settings, a single indoor pesticide application can result in 15 mg of pesticide being released into the environment.[5] Pesticides are used indoors to control pests such as cockroaches and bed bugs. Other uses of pesticides include control of domestic animal pests and their use as fungicides on furniture and consumer products (e.g., wool carpets, textiles) and building materials (e.g., fungicide-containing wall paints, mold-resistant drywall) [6,7,8,9]. In addition, the actions of occupants (e.g., smoking indoors) can result in the release of pesticides used to grow tobacco into indoor spaces [10]. Another source of pesticide release into indoor spaces is their transportation from outside [11,12,13].
In addition to agricultural workers and their families, certain groups are also vulnerable to pesticide exposure. Children are more exposed to many indoor contaminants, including pesticides, than adults due to higher rates of inhalation, dust ingestion, and hand-to-mouth habits relative to body weight [ 14 , 15 ]. For example, Trunnel et al. found that pyrethroid/pyrethrin (PYR) concentrations in floor wipes were positively correlated with PYR metabolite concentrations in children’s urine [ 16 ]. The DF of PYR pesticide metabolites reported in the Canadian Health Measures Study (CHMS) was higher in children aged 3–5 years than in older age groups [ 17 ]. Pregnant women and their fetuses are also considered a vulnerable group due to the risk of early life pesticide exposure. Wyatt et al. reported that pesticides in maternal and neonatal blood samples were highly correlated, consistent with maternal-fetal transfer [18].
People living in substandard or low-income housing are at increased risk of exposure to indoor pollutants, including pesticides [ 19 , 20 , 21 ]. For example, in Canada, studies have shown that people with lower socioeconomic status (SES) are more likely to be exposed to phthalates, halogenated flame retardants, organophosphorus plasticizers and flame retardants, and polycyclic aromatic hydrocarbons (PAHs) than people with higher SES [22,23,24]. Some of these findings apply to people living in “social housing,” which we define as rental housing subsidized by the government (or government-funded agencies) that contains residents of lower socioeconomic status [ 25 ]. Social housing in multi-unit residential buildings (MURBs) are susceptible to pest infestations, mainly due to their structural defects (e.g. cracks and crevices in walls), lack of proper maintenance/repair, inadequate cleaning and waste disposal services, and frequent overcrowding [ 20 , 26 ]. Although integrated pest management programmes are available to minimise the need for pest control programmes in building management and thus reduce the risk of pesticide exposure, particularly in multi-unit buildings, pests can spread throughout the building [21, 27, 28]. Pest spread and associated pesticide use can negatively impact indoor air quality and expose occupants to the risk of pesticide exposure, leading to adverse health effects [29]. Several studies in the United States have shown that exposure levels to banned and currently used pesticides are higher in low-income housing than in high-income housing due to poor housing quality [11, 26, 30,31,32]. Because low-income residents often have few options for leaving their homes, they may be continually exposed to pesticides in their homes.
In homes, residents may be exposed to high concentrations of pesticides over long periods of time because pesticide residues persist due to lack of sunlight, moisture, and microbial degradation pathways [33,34,35]. Pesticide exposure has been reported to be associated with adverse health effects such as neurodevelopmental disabilities (particularly lower verbal IQ in boys), as well as blood cancers, brain cancers (including childhood cancers), endocrine disruption-related effects, and Alzheimer’s disease .
As a party to the Stockholm Convention, Canada has restrictions on nine OCPs [42, 54]. A re-evaluation of regulatory requirements in Canada has resulted in the phase-out of nearly all residential interior uses of OPP and carbamate.[55] The Pest Management Regulatory Agency of Canada (PMRA) also restricts some indoor uses of PYR. For example, the use of cypermethrin for indoor perimeter treatments and broadcasts has been discontinued due to its potential impact on human health, particularly in children [56]. Figure 1 provides a summary of these restrictions [55, 57, 58].
The Y-axis represents the detected pesticides (above the detection limit of the method, Table S6), and the X-axis represents the concentration range of pesticides in the air in the particle phase above the detection limit. Details of the detection frequencies and maximum concentrations are provided in Table S6.
Our objectives were to measure indoor air concentrations and exposures (e.g., inhalation) of currently used and legacy pesticides in low socioeconomic status households living in social housing in Toronto, Canada, and to examine some of the factors associated with these exposures. The aim of this paper is to fill the gap in data on exposures to current and legacy pesticides in the homes of vulnerable populations, particularly given that indoor pesticide data in Canada are extremely limited [ 6 ].
The researchers monitored pesticide concentrations in seven MURB social housing complexes built in the 1970s at three sites in the City of Toronto. All buildings are at least 65 km from any agricultural zone (excluding backyard plots). These buildings are representative of Toronto social housing. Our study is an extension of a larger study that examined particulate matter (PM) levels in social housing units before and after energy upgrades [59,60,61]. Therefore, our sampling strategy was limited to collecting airborne PM.
For each block, modifications were developed that included water and energy savings (e.g. replacement of ventilation units, boilers and heating appliances) to reduce energy consumption, improve indoor air quality and increase thermal comfort [ 62 , 63 ]. The apartments are divided according to the type of occupancy: elderly, families and single people. The features and types of buildings are described in more detail elsewhere [24].
Forty-six air filter samples collected from 46 MURB social housing units in winter 2017 were analyzed. The study design, sample collection, and storage procedures were described in detail by Wang et al. [60]. Briefly, each participant’s unit was equipped with an Amaircare XR-100 air purifier fitted with 127 mm high-efficiency particulate air filter media (the material used in HEPA filters) for 1 week. All portable air purifiers were cleaned with isopropyl wipes before and after use to avoid cross-contamination. Portable air purifiers were placed on the living room wall 30 cm from the ceiling and/or as directed by residents to avoid inconvenience to residents and minimize the possibility of unauthorized access (see Supplementary Information SI1, Figure S1). During the weekly sampling period, the median flow was 39.2 m3/day (see SI1 for details of the methods used to determine flow). Prior to sampler deployment in January and February 2015, an initial door-to-door visit and visual inspection of household characteristics and occupant behaviour (e.g. smoking) was carried out. A follow-up survey was conducted after each visit from 2015 to 2017. Full details are provided in Touchie et al. [64] Briefly, the aim of the survey was to assess occupant behaviour and potential changes in household characteristics and occupant behaviour such as smoking, door and window operation, and use of extractor hoods or kitchen fans when cooking. [59, 64] After modification, filters for 28 target pesticides were analyzed (endosulfan I and II and α- and γ-chlordane were considered as different compounds, and p,p′-DDE was a metabolite of p,p′-DDT, not a pesticide), including both old and modern pesticides (Table S1).
Wang et al. [60] described the extraction and cleanup process in detail. Each filter sample was split in half and one half was used for the analysis of 28 pesticides (Table S1). Filter samples and laboratory blanks consisted of glass fiber filters, one for every five samples for a total of nine, spiked with six labeled pesticide surrogates (Table S2, Chromatographic Specialties Inc.) to control for recovery. Target pesticide concentrations were also measured in five field blanks. Each filter sample was sonicated three times for 20 min each with 10 mL of hexane:acetone:dichloromethane (2:1:1, v:v:v) (HPLC grade, Fisher Scientific). The supernatants from the three extractions were pooled and concentrated to 1 mL in a Zymark Turbovap evaporator under a constant flow of nitrogen. The extract was purified using Florisil® SPE columns (Florisil® Superclean ENVI-Florisil SPE tubes, Supelco) then concentrated to 0.5 mL using a Zymark Turbovap and transferred to an amber GC vial. Mirex (AccuStandard®) (100 ng, Table S2) was then added as an internal standard. Analyses were performed by gas chromatography-mass spectrometry (GC-MSD, Agilent 7890B GC and Agilent 5977A MSD) in electron impact and chemical ionization modes. Instrument parameters are given in SI4 and quantitative ion information is given in Tables S3 and S4.
Prior to extraction, labeled pesticide surrogates were spiked into samples and blanks (Table S2) to monitor recovery during analysis. Recoveries of marker compounds in samples ranged from 62% to 83%; all results for individual chemicals were corrected for recovery. Data were blank corrected using the mean laboratory and field blank values for each pesticide (values are listed in Table S5) according to the criteria explained by Saini et al. [65]: when the blank concentration was less than 5% of the sample concentration, no blank correction was performed for individual chemicals; when the blank concentration was 5–35%, data were blank corrected; if the blank concentration was greater than 35% of the value, data were discarded. The method detection limit (MDL, Table S6) was defined as the mean concentration of the laboratory blank (n = 9) plus three times the standard deviation. If a compound was not detected in the blank, the signal-to-noise ratio of the compound in the lowest standard solution (~10:1) was used to calculate the instrument detection limit. Concentrations in laboratory and field samples were
The chemical mass on the air filter is converted to the integrated airborne particle concentration using gravimetric analysis, and the filter flow rate and filter efficiency are converted to the integrated airborne particle concentration according to equation 1:
where M (g) is the total mass of PM captured by the filter, f (pg/g) is the pollutant concentration in the collected PM, η is the filter efficiency (assumed to be 100% due to the filter material and particle size [67]), Q (m3/h) is the volumetric air flow rate through the portable air purifier, and t (h) is the deployment time. The filter weight was recorded before and after deployment. Full details of the measurements and air flow rates are provided by Wang et al. [60].
The sampling method used in this paper measured only the concentration of the particulate phase. We estimated equivalent concentrations of pesticides in the gas phase using the Harner-Biedelman equation (Equation 2), assuming chemical equilibrium between the phases [68]. Equation 2 was derived for particulate matter outdoors, but has also been used to estimate particle distribution in air and indoor environments [69, 70].
where log Kp is the logarithmic transformation of the particle-gas partition coefficient in air, log Koa is the logarithmic transformation of the octanol/air partition coefficient, Koa (dimensionless), and \({fom}\) is the fraction of organic matter in particulate matter (dimensionless). The fom value is taken to be 0.4 [71, 72]. The Koa value was taken from OPERA 2.6 obtained using the CompTox chemical monitoring dashboard (US EPA, 2023) (Figure S2), since it has the least biased estimates compared to other estimation methods [73]. We also obtained experimental values of Koa and Kowwin/HENRYWIN estimates using EPISuite [74].
Since the DF for all detected pesticides was ≤50%, values
Figure S3 and Tables S6 and S8 show OPERA-based Koa values, the particulate phase (filter) concentration of each pesticide group, and the calculated gas phase and total concentrations. Gas phase concentrations and maximum sum of detected pesticides for each chemical group (i.e., Σ8OCP, Σ3OPP, Σ8PYR, and Σ3STR) obtained using the experimental and calculated Koa values from EPISuite are provided in Tables S7 and S8, respectively. We report measured particulate phase concentrations and compare the total air concentrations calculated here (using OPERA-based estimates) with air concentrations from a limited number of non-agricultural reports of airborne pesticide concentrations and from several studies of low-SES households [26, 31, 76,77,78] (Table S9). It is important to note that this comparison is approximate due to differences in sampling methods and study years. To our knowledge, the data presented here are the first to measure pesticides other than traditional organochlorines in indoor air in Canada.
In the particle phase, the maximum detected concentration of Σ8OCP was 4400 pg/m3 (Table S8). The OCP with the highest concentration was heptachlor (restricted in 1985) with a maximum concentration of 2600 pg/m3, followed by p,p′-DDT (restricted in 1985) with a maximum concentration of 1400 pg/m3 [57]. Chlorothalonil with a maximum concentration of 1200 pg/m3 is an antibacterial and antifungal pesticide used in paints. Although its registration for indoor use was suspended in 2011, its DF remains at 50% [55]. The relatively high DF values and concentrations of traditional OCPs indicate that OCPs have been widely used in the past and that they are persistent in indoor environments [6].
Previous studies have shown that building age is positively correlated with concentrations of older OCPs [6, 79]. Traditionally, OCPs have been used for indoor pest control, particularly lindane for the treatment of head lice, a disease that is more common in households with lower socioeconomic status than in households with higher socioeconomic status [80, 81]. The highest concentration of lindane was 990 pg/m3.
For total particulate matter and gas phase, heptachlor had the highest concentration, with a maximum concentration of 443,000 pg/m3. Maximum total Σ8OCP air concentrations estimated from Koa values in other ranges are listed in Table S8. Concentrations of heptachlor, lindane, chlorothalonil, and endosulfan I were 2 (chlorothalonil) to 11 (endosulfan I) times higher than those found in other studies of high- and low-income residential environments in the United States and France that were measured 30 years ago [77, 82,83,84].
The highest total particulate phase concentration of the three OPs (Σ3OPPs)—malathion, trichlorfon, and diazinon—was 3,600 pg/m3. Of these, only malathion is currently registered for residential use in Canada.[55] Trichlorfon had the highest particulate phase concentration in the OPP category, with a maximum of 3,600 pg/m3. In Canada, trichlorfon has been used as a technical pesticide in other pest control products, such as for the control of non-resistant flies and cockroaches.[55] Malathion is registered as a rodenticide for residential use, with a maximum concentration of 2,800 pg/m3.
The maximum total concentration of Σ3OPPs (gas + particles) in air is 77,000 pg/m3 (60,000–200,000 pg/m3 based on Koa EPISuite value). Airborne OPP concentrations are lower (DF 11–24%) than OCP concentrations (DF 0–50%), which is most likely due to the greater persistence of OCP [85].
The diazinon and malathion concentrations reported here are higher than those measured approximately 20 years ago in low socioeconomic status households in South Texas and Boston (where only diazinon was reported) [ 26 , 78 ]. The diazinon concentrations we measured were lower than those reported in studies of low- and middle-socioeconomic status households in New York and Northern California (we were unable to locate more recent reports in the literature) [ 76 , 77 ].
PYRs are the most commonly used pesticides for bed bug control in many countries, but few studies have measured their concentrations in indoor air [86, 87]. This is the first time that indoor PYR concentration data have been reported in Canada.
In the particle phase, the maximum \(\,{\sum }_{8}{PYRs}\) value is 36,000 pg/m3. Pyrethrin I was the most frequently detected (DF% = 48), with the highest value of 32,000 pg/m3 among all pesticides. Pyrethroid I is registered in Canada for the control of bed bugs, cockroaches, flying insects, and pet pests [55, 88]. Additionally, pyrethrin I is considered a first-line treatment for pediculosis in Canada [89]. Given that people living in social housing are more susceptible to bed bug and lice infestations [80, 81], we expected the concentration of pyrethrin I to be high. To our knowledge, only one study has reported concentrations of pyrethrin I in indoor air of residential properties, and none have reported pyrethrin I in social housing. The concentrations we observed were higher than those reported in the literature [90].
Allethrin concentrations were also relatively high, with the second highest concentration being in the particulate phase at 16,000 pg/m3, followed by permethrin (maximum concentration 14,000 pg/m3). Allethrin and permethrin are widely used in residential construction. Like pyrethrin I, permethrin is used in Canada to treat head lice.[89] The highest concentration of L-cyhalothrin detected was 6,000 pg/m3. Although L-cyhalothrin is not registered for home use in Canada, it is approved for commercial use to protect wood from carpenter ants.[55, 91]
The maximum total \({\sum }_{8}{PYRs}\) concentration in air was 740,000 pg/m3 (110,000–270,000 based on the Koa EPISuite value). Allethrin and permethrin concentrations here (maximum 406,000 pg/m3 and 14,500 pg/m3, respectively) were higher than those reported in lower-SES indoor air studies [26, 77, 78]. However, Wyatt et al. reported higher permethrin levels in the indoor air of low-SES homes in New York City than our results (12 times higher) [76]. The permethrin concentrations we measured ranged from the low end to a maximum of 5300 pg/m3.
Although STR biocides are not registered for use in the home in Canada, they may be used in some building materials such as mold-resistant siding [75, 93]. We measured relatively low particulate phase concentrations with a maximum \({\sum }_{3}{STRs}\) of 1200 pg/m3 and total air \({\sum }_{3}{STRs}\) concentrations up to 1300 pg/m3. STR concentrations in indoor air have not been previously measured.
Imidacloprid is a neonicotinoid insecticide registered in Canada for the control of insect pests of domestic animals.[55] The maximum concentration of imidacloprid in the particulate phase was 930 pg/m3, and the maximum concentration in general air was 34,000 pg/m3.
The fungicide propiconazole is registered in Canada for use as a wood preservative in building materials.[55] The maximum concentration we measured in the particulate phase was 1100 pg/m3, and the maximum concentration in general air was estimated to be 2200 pg/m3.
Pendimethalin is a dinitroaniline pesticide with a maximum particulate phase concentration of 4400 pg/m3 and a maximum total air concentration of 9100 pg/m3. Pendimethalin is not registered for residential use in Canada, but one source of exposure may be tobacco use, as discussed below.
Many pesticides were correlated with each other (Table S10). As expected, p,p′-DDT and p,p′-DDE had significant correlations because p,p′-DDE is a metabolite of p,p′-DDT. Similarly, endosulfan I and endosulfan II also had a significant correlation because they are two diastereoisomers that occur together in technical endosulfan. The ratio of the two diastereoisomers (endosulfan I:endosulfan II) varies from 2:1 to 7:3 depending on the technical mixture [94]. In our study, the ratio ranged from 1:1 to 2:1.
We next looked for co-occurrences that might indicate the co-use of pesticides and the use of multiple pesticides in a single pesticide product (see the breakpoint plot in Figure S4). For example, co-occurrence could occur because the active ingredients could be combined with other pesticides with different modes of action, such as a mixture of pyriproxyfen and tetramethrin. Here, we observed a correlation (p < 0.01) and co-occurrence (6 units) of these pesticides (Figure S4 and Table S10), consistent with their combined formulation [75]. Significant correlations (p < 0.01) and co-occurrences were observed between OCPs such as p,p′-DDT with lindane (5 units) and heptachlor (6 units), suggesting that they were used over a period of time or applied together before the restrictions were introduced. No co-presence of OFPs was observed, with the exception of diazinon and malathion, which were detected in 2 units.
The high co-occurrence rate (8 units) observed between pyriproxyfen, imidacloprid and permethrin may be explained by the use of these three active pesticides in insecticidal products for the control of ticks, lice and fleas on dogs [95]. In addition, co-occurrence rates of imidacloprid and L-cypermethrin (4 units), propargyltrine (4 units) and pyrethrin I (9 units) were also observed. To our knowledge, there are no published reports of co-occurrence of imidacloprid with L-cypermethrin, propargyltrine and pyrethrin I in Canada. However, registered pesticides in other countries contain mixtures of imidacloprid with L-cypermethrin and propargyltrine [96, 97]. Furthermore, we are not aware of any products containing a mixture of pyrethrin I and imidacloprid. The use of both insecticides may explain the observed co-occurrence, as both are used to control bed bugs, which are common in social housing [86, 98]. We found that permethrin and pyrethrin I (16 units) were significantly correlated (p < 0.01) and had the highest number of co-occurrences, suggesting that they were used together; this was also true for pyrethrin I and allethrin (7 units, p < 0.05), while permethrin and allethrin had a lower correlation (5 units, p < 0.05) [75]. Pendimethalin, permethrin and thiophanate-methyl, which are used on tobacco crops, also showed correlation and co-occurrence at nine units. Additional correlations and co-occurrences were observed between pesticides for which co-formulations have not been reported, such as permethrin with STRs (i.e., azoxystrobin, fluoxastrobin, and trifloxystrobin).
Tobacco cultivation and processing rely heavily on pesticides. Pesticide levels in tobacco are reduced during harvesting, curing, and final product manufacturing. However, pesticide residues still remain in the tobacco leaves.[99] Additionally, tobacco leaves may be treated with pesticides after harvest.[100] As a result, pesticides have been detected in both tobacco leaves and smoke.
In Ontario, more than half of the 12 largest social housing buildings do not have a smoke-free policy, putting residents at risk of exposure to second-hand smoke.[101] The MURB social housing buildings in our study did not have a smoke-free policy. We surveyed residents to obtain information about their smoking habits and conducted unit checks during home visits to detect signs of smoking.[59, 64] In winter 2017, 30% of residents (14 out of 46) smoked.
Post time: Feb-06-2025