United States house dust Pb concentrations are influenced by soil, paint, and house age: insights from a national survey
Results from the AHHS II survey indicate that homes with elevated house dust Pb concentration conditions remain present, although average house dust Pb concentrations in the United States have decreased (Table 1). Previous survey-level investigations of house dust Pb in the US, especially by dust mass, are limited. A previous HUD survey performed ~30 years ago (collection in 1993 and publication in 1998) details both dust Pb concentration and dust Pb loading data sets for 301 house dust samples [41]. The previous survey reported a mean Pb concentration of 364.2 µg Pb g−1 which is three times the mean concentration found in our investigation [28], suggesting a reduction in house dust Pb in the average US home between these studies. Overall decreases in house dust Pb concentrations are reasonable given efforts to reduce Pb sources over the past several decades [42]. However, homes where Pb dust concentrations remain elevated may be due to continued persistence of Pb hazards as shown in the AHHS II HUD Lead Findings report (paint, soil, etc.) [28, 42].
Another US investigation of 204 homes where children resided conducted in Rochester, New York in 1995 found substantially higher dust Pb concentrations and dust Pb loading (867.32 µg Pb g−1 and 130.0 µg Pb ft−2, respectively) than observed in our investigation [17, 26]. In addition to changes in dust Pb concentrations with time, elevated concentrations reported in the Rochester investigation may be a result of homes being in urban areas near anthropogenic sources of Pb contamination where soil and paint Pb are major contributors to house dust Pb. Outside of the US, a more recent national survey conducted by Health Canada provides a comprehensive evaluation of house dust Pb by mass and surface area (2013) [21]. Mean house dust Pb concentration for the Health Canada study was higher (210 µg Pb g−1); however, the Health Canada investigation specifically targeted urban areas, with approximately half of samples collected from homes within 2 km of industrial sites. Additionally, collection of the Health Canada samples was ~15 years after the aforementioned US studies, potentially representing temporal influences on house dust Pb concentration.
Across all of the aforementioned studies utilizing vacuum-collected house dust, it is important to consider that inter-study vacuum collection methodologies varied, which may limit comparisons. Such differences in collection were not assessed as part of this study. In summary, decreasing Pb concentration in house dust over time and proximity to Pb-associated anthropogenic activities (e.g., smelters, legacy contaminated urban areas, etc.) likely contributed to the lower mean dust Pb concentration observed in AHHS II.
Influence of house age and other residential characteristics on house dust Pb concentrations were further evaluated using AHHS II house age and region data. Consistent with other studies of Pb in house dust [20,21,22, 43], house age was found to be a significant predictor of dust Pb concentrations (Fig. 1). Notably, we observed an approximate 4-fold decrease in mean dust Pb concentration in homes built after the US CPSC issued (1978) stricter limits on house paint Pb concentrations (Fig. 1). This association is likely connected to the increased likelihood of lead-based hazards from interior and/or exterior residential sources in older homes, and/or increased time for paint to deteriorate and concentrate in house dust. The AHHS II and previous national surveys have demonstrated that paint-Pb levels increase with the age of housing, with the highest levels measured in pre-1940 housing [44, 45]. Even after ceasing use of Pb-based products, Pb is highly stable in soils and may be readily available for physical transport into homes [14, 18, 19, 46]. Additionally, home interiors once containing Pb-based paint are challenging to eliminate as contributors to house dust. Limited mass of Pb-based paint is required to incorporate into house dust and contribute to elevated dust Pb concentrations that adversely impact BLLs [17, 18, 24, 42]. This issue may be exacerbated by renovation efforts that generate fine particulate Pb-based paint deposition throughout homes, especially if sanding is performed or other proper precautions are not taken [12]. Older homes are associated with increased dust mass, which has been reported as a significant factor affecting increases in house dust Pb [21], providing another example of the connection between house age and house dust Pb.
Evaluation of census region and population density influence on house dust Pb concentrations indicate that house age is the primary factor predictive of dust Pb levels in US homes among the factors investigated as part of this study. Additional factors that may be associated with US census region or MSA beyond house age, including but not limited to population or industry density [12, 47, 48], likely have some influence on dust Pb levels in US homes (see SI), but such factors appear to be secondary to house age when evaluated at the scale of a nationwide US survey.
Although AHHS II vacuum-collected house dust is the primary focus of this study, paired dust wipe data collected by HUD was also evaluated [28]. Dust wipe data (Pb loading-basis) enabled vacuum dust Pb (concentration-basis) to be statistically related to current Pb hazard standards. While vacuum-collected house dust data represent an aggregate of all surfaces vacuumed, house dust wipe samples were collected from specific locations and surfaces in each home (Table 1) and are further described in the SI and a recent HUD report [28]. Statistically-significant correlations between vacuum-collected house dust and paired dust wipe data showcase relationships between the two dust Pb assessment methods. Correlations were stronger when wipe data was averaged by surface across rooms, which is reasonable given vacuum-collected samples in this study were a composite of dust collected throughout the home. Stronger correlations of vacuum dust Pb values with floor loading data than windowsill loading data are consistent with previous findings [17, 27, 41]. Collection of paired vacuum and dust wipe samples for dust Pb analysis from the same home is uncommon, with most investigations choosing one method to proceed with house dust Pb assessment. Lanphear et al. and Rasmussen et al. utilized both datasets to provide thorough insight into house dust Pb contamination and have used dust Pb concentration to predict dust Pb loading values, but this requires accurate determination of dust mass per residential surface area [17, 21, 27]. Given that compliance with current U.S. Pb dust hazard standards are assessed using dust wipe loading data (10 µg Pb ft−2 and 100 µg Pb ft−2 for floors and windowsills, respectively), statistically correlating vacuum-collected dust Pb concentrations to paired wipe data enabled concentration data to be related to hazard standards [28]. It is important to note, however, that while correlations were statistically significant, r values observed suggest the strength of this relationship is moderate in homes assessed.
Concentration data from vacuum collected dust was also related to paired loading data collected from the same home in the context of evaluating probabilities of dust Pb hazard exceedance (Fig. 3 and Tables 1 and S2). Collectively, probability of hazard exceedance and binning results suggest that homes with dust Pb wipe exceedances are likely to encounter higher Pb concentrations in house dust that exceed 400 µg Pb g−1. However, house dust Pb concentrations less than 400 µg Pb g−1 may still be of concern. Results indicate a probability of exceedance of at least ~25% at 150 µg Pb g−1 across both floor and windowsills combined (Fig. 3). Therefore, maximum acceptable probability of exceedance is important to consider when evaluating house dust Pb concentrations in the context of health standards.
Soil Pb has long been expected to be a mediator of house dust Pb, a relationship supported by correlations between mean soil Pb and Pb in vacuum dust (r = 0.64) observed in AHHS II [12, 13, 19, 20, 49, 50]. This finding provides needed data to support the long-standing assertion that soil Pb is an important mediator of house dust Pb [18, 22, 46, 50, 51]. In areas not located near industry, soil has been previously suggested as a less significant contributor to house dust, with Pb in dust expected to derive primarily from interior sources [13, 31]. Correlations between Pb in soil and dust observed in AHHS II, which includes homes across a range of population and urban densities, suggests soil is consistently important to house dust Pb. However, this does not signify that paint Pb sources are less important, rather, paint Pb and soil Pb concentration are related (Pearson correlation coefficient = 0.60; SI Table S3). The role of interconnected contributions of paint Pb and soil Pb to house dust are likely driven by house age, as decreasing house dust Pb concentration with decreasing house age was observed (Fig. 1). Soil Pb from the major entryway was found to have the strongest correlation with vacuum-collected house dust Pb (r = 0.66) compared to dripline, mid-yard, or children’s play area (r = 0.63, 0.55 and 0.56 respectively). Similar results were found for a house dust investigation in Philadelphia where elevated BLLs (µg dl−1) in children were found to be strongly associated soil Pb in the major entryway [22, 23]. Increased soil Pb concentration at the foundation drip line was also observed in AHHS II and was expected as this area typically has elevated Pb concentrations likely due to deterioration of exterior Pb-based paint [39, 52]. The relatively high mean Pb concentration of major entryway soil supports tracked-in sources as a potentially important pathway for soil Pb entry into the home and subsequent incorporation into dust [53]. Further investigation characterizing Pb sources and phases around the major entry area of households may provide the best prediction of house dust Pb concentration compared to other residential soil sampling locations. Additionally, targeted remediation of the entryway portions of a residence may have a substantial impact on house dust Pb, the primary contributor to elevated BLLs in children [17, 23, 24, 27].
Determination of Pb paint levels revealed that both interior and exterior paints sources are correlated with house dust Pb concentration; however, interior paint Pb had a stronger correlation (Tables 1 and 3). This observation is logical as interior Pb paint can directly contaminate indoor dust, whereas exterior Pb paint must first be tracked into the home. Stronger correlations with mean and 95th quantile values for Pb paint loading than median values suggests that associations between the Pb in paint and accumulation of Pb paint in dust may not be consistent across the range of Pb concentrations observed in each home. For example, not only may older paint in a home be more likely to have elevated Pb levels, but it may also be more likely to deteriorate, requiring only minute amounts to contribute substantially to Pb deposition in dust [17, 42].
Although vacuum- collected dust Pb was generally found to be more strongly correlated with soil Pb than paint Pb, both Pb from soil and paint are important mediators of dust Pb concentrations. Furthermore, Pb in soil and paint were strongly correlated. Specifically, mean and 95th quantile soil and paint Pb Pearson correlation coefficients were 0.60 and 0.69, respectively (SI Table S3), suggesting paint Pb contributes to soil Pb as exterior paint deteriorates and deposits into soil systems [52] before infiltrating homes and accumulating in house dust. Other major contributors of soil Pb, in addition to paint Pb, include leaded gasoline, Pb smelters, and other industrial activities [52]. The strong correlation of soil and paint Pb provides valuable insight showcasing interconnection between Pb sources that may concentrate in house dust. Pb phases indicative of Pb-based paint are still found in prior Pb speciation investigations of both soils and house dust and likely contribute to sorbed Pb concentrations in soils as paint containing Pb phases degrade [14, 30, 31]. Additionally, identification of paint Pb species in house dust may not necessarily be an indicator of contamination originating from within the home, as exterior paint Pb surfaces are, on average, more concentrated than interior sources (Table 1) and are correlated with soil Pb (SI Table S3). This finding is critical for future speciation analyses probing Pb sources within and around residences, where limited investigations have revealed the presence of Pb phases commonly attributed to white paint pigment [14, 30].
House dust Pb remains a persistent issue contributing to increased BLLs in the United States and globally [5]. The AHHS II study provides a window into the current state of Pb in and around residences. Here we evaluated the relationship between house dust Pb concentrations and two common residential Pb sources: soil and Pb-based paint. Despite the banning of Pb-based paint ~45 years ago, results of this survey suggest that dust Pb concentrations in homes continue to be influenced by Pb-based paint and soil contaminated with Pb (SI Table S3), in particular through soil transport from the major entryway. House dust Pb concentrations and loading have decreased with time, but Pb sequestered in soils, in conjunction with Pb-based paint, continue to be mediators of house dust Pb contamination. This is especially true for older homes built prior to the banning of Pb-based paint (pre-1978) in residential and public properties, where legacy Pb-based paint in residence interior and exteriors is more likely. Pb readily sorbs to many environmental constituents that facilitate stability in soil systems [14, 54, 55] and may be available for physical transport into home interior; however, this stability is not expected to persist upon ingestions and/or inhalation [14, 56, 57]. Sorbed Pb and paint Pb phases (e.g., cerussite, hydrocerussite) have been found to be highly bioavailable [14]. Therefore, results presented here suggest that soil Pb remediation, potentially via in situ chemical remediation methods [54, 58], may be an important path to mitigating further Pb contamination of house dust via the soil to dust pathway.
Current USEPA hazard standards of house dust Pb are based on Pb loading data (µg Pb ft−2) using house dust wipes; however, our results demonstrate that probability of dust Pb loading exceeding hazard standards can be statistically estimated using house dust Pb concentration data from the same home. We observed an ~25% probability of exceedance for dust Pb concentrations of 150 µg Pb g−1. Therefore, the predictive model (Fig. 3) may be useful in cases where only house dust Pb concentration data is available and improved with future paired dust Pb concentration and loading datasets. Future investigations probing specific Pb phases in house dust and relationships to residential sources will be essential to assessing exposure and developing effective remediation strategies.
Dust samples from household vacuum bags can provide a cost-effective measure of contaminant concentrations [26, 59]. While important, there are weaknesses to using vacuum-collected house dust. An important disadvantage of household vacuum samples is the lack of information on dust age and provenance (i.e., area sampled) required for calculating dust and elemental loading rates. Vacuum collection methods are generally more variable than wipe methods, and don’t allow for dust Pb to be evaluated per surface area. This limits the direct usage of vacuum-collected house dust to current hazard standards based on Pb loading values.
AHHS II included collection of additional data shown in previous studies to relate to dust Pb levels in homes, including socio-economic data. Such data was collected by means of a questionnaire completed by the homeowner or representative, with responses to questions being completely voluntary. Upon review of the available data, we determined there was insufficient socio-economic data available to include these as factors assessed in this study. This also limited the number of variables that could be considered in multivariate models relating factors to vacuum-collected dust Pb concentrations.
Collection of dust from the homeowner’s vacuum provided mass quantities necessary for inorganic analyses and advanced characterization techniques that are either more challenging or unachievable using wipe methods. Using dust from homeowner’s vacuum also provides a quick, easy, low-cost method and low burden effort on the homeowner for collecting dust from homes. Therefore, results of Pb in vacuum-collected house dust presented here may facilitate future Pb dust investigations, particularly in larger surveys where such benefits may be of particular value.
