Introduction

For seven decades, chromate production was an active industry in Jersey City, New Jersey and neighboring Kearny.1 By the mid-1970s, an estimated 2–3 million tons of chromium ore processing residue (COPR) had been generated. The waste, containing an estimated 2–7% chromium, was used as landfill in numerous sites throughout Hudson County. To date, over 100 sites contaminated with COPR have been identified in Jersey City. At some sites, concentrations over 50,000 p.p.m. total chromium were reported.1 It was not unusual for the waste to contain high levels (up to hundreds of p.p.m.) of hexavalent chromium, a known human inhalation carcinogen and dermal allergen.2, 3, 4, 5, 6

Exposure to Cr+6 has been an ongoing concern in and around Jersey City beginning in the late 1980s.1, 7, 8 A series of studies during the 1990s evaluated the exposure of Jersey City residents to chromium both before and during remediation.2, 3, 9, 10, 11, 12, 13 These 1990s studies have recently been reviewed as a model for exposure characterization.8 Since the public water supply was unaffected by the contamination, the chief routes of exposure were thought to be inhalation of particles, ingestion of particles (soil or airborne particles) and contaminated food, and dermal contact contaminated soil and water.1, 12 According to the Agency for Toxic Substances Disease Registry,14 chromium absorption is affected by the valence state, solubility, as well as route of exposure. After ingestion, Cr+6 is reduced in the stomach and less than 10% is absorbed. Small amounts of Cr+6 and Cr+3 may be absorbed dermally through unbroken skin. Cr+6 is more readily absorbed through the gastrointestinal tract and by the lungs than Cr+3. Once absorbed, Cr+6 appears to be essentially completely reduced to the trivalent form by the time it reaches the systemic circulation. Absorbed chromium is excreted in the urine as Cr+3. The half-life for urine elimination is approximately 10 h for Cr+3 and 39 h for Cr+6.14

Throughout the 1990s studies, chromium levels in house dust were used as an indicator of potential exposure. Exposure was verified through biomonitoring of urinary chromium concentrations. The studies found higher levels of total chromium in dust in homes near waste sites and higher levels in urine. Chromium in house dust in homes near waste sites was found to be strongly correlated with chromium concentration in the urine of young children living in these homes.13 The association between dust and urinary chromium concentrations was not significant in older children or adults. Behavioral characteristics such as spending more time indoors and hand-to-mouth behaviors were thought to increase exposure to contaminated house dust in this age group.3, 13 In homes not adjacent to COPR waste site, dust levels of chromium were not elevated and no association between dust and urine levels of chromium was found.13

In the 1990s, remediation of residential COPR sites was shown to decrease exposures.9 Subsequently, all COPR sites in residential locations received final or interim remediation either by excavation and clean fill or by capping. However, residents of Jersey City remained concerned about residual exposures to Cr+6. In earlier studies, the available analytical methodology did not allow us to measure Cr+6 in house dust wipe samples. However, the analytical methodology has advanced to permit quantitative measurement of Cr+6 in house dust samples. Thus, in response to residents’ concerns, a post-remediation investigation of Cr+6 exposure in Jersey City, the Chromium Exposure and Health Study (CEHS), was developed. The CEHS was designed in two phases. In phase I (CEHS-I) Jersey City homes (n=100) were compared with homes in reference communities (n=20) in central New Jersey with no known local sources of chromium.7 No significant differences in Cr+6 concentrations were found between formerly contaminated neighborhoods of Jersey City and the background locations. This paper reports on phase II of the study (CEHS-II), which examines the relationship between Cr+6 levels in dust and children’s chromium exposure in Jersey City in this post-remediation period. The goal of CEHS-II is to determine whether, after remediation, a significant association remains between chromium concentration in urine and chromium levels in house dust in young children. Although the results from CEHS-I indicated that Cr+6 levels in house dust did not differ between Jersey City and background locations, it is still possible that the nature of the Cr+6 particulates, or the conditions of the exposure differ between Jersey City and background locations such that internal exposure to Cr+6 remains elevated in Jersey City. The finding of a significant association between Cr+6 in house dust and children’s urinary chromium concentration would indicate that COPR contamination still affects children’s exposures, despite remediation. If a significant association is not found, then this provides further evidence that remediation reduced exposures.

Methods

Subject Recruitment

All residents of Jersey City with children under 7 years of age were eligible for the study. Repeated and multifaceted recruitment approaches were used over a 2-year period in an effort to include families living near remediated waste sites. This included using a televised public service announcement, community meetings, direct mailings to residents living near the two largest waste sites (one remediated and one undergoing final remediation), and the distribution of flyers to children in elementary schools, including pre-Kindergarten programs. Flyers were also made available at public health clinics and posted in local supermarkets. Additionally, participants were recruited in person from a large pediatric clinic in Jersey City. A single study visit was scheduled for all sample (dust and urine) and data collection. All phases of this study were reviewed and approved by the Institutional Review Board of Robert Wood Johnson Medical School and all study subjects signed an informed consent from.

Questionnaire Data

Two short questionnaires were administered by a field team member during the study visit. One questionnaire collected information about the home, including recent renovations, housing materials, cleaning methods, and ventilation. The second questionnaire was designed to gather information regarding the general time-activity patterns of participating children and the use of dietary supplements or vitamins. Parents were asked the number of hours the children played outside (outdoors) on a typical cold weather or winter day and on a warm weather day.

Dust Sample Collection and Analysis

A dust sample was collected from each of three areas of the home: the main entryway, the child’s main play area, and one other area (a bedroom, living room, or basement, if available). Samples were preferentially collected from elevated surfaces (furniture, shelves, etc.). If no suitable elevated surfaces were available in the location, then samples were collected from a hard floor. Dust samples were collected from three non-metal surfaces using two methods previously described.7 The preferred method was with the Lioy-Wainman-Weisel (LWW) sampler.15, 16 Sample collection consists of consecutive wipes within a defined template (30 × 5 cm, 150 cm2) with a series of three pre-weighed polyester filters. Filters are weighed after collection to determine the dust mass. In five cases, the sampling area could not fit the LWW template so a free hand wipe was collected. Analytical procedures are presented elsewhere.7, 17, 18 Briefly, dust samples were extracted with dilute (0.01%) nitric acid (pH=4.0 HNO3). Cr+6 was separated using an ion chromatograph (IC) with a CG5A guard column and analyzed by inductively coupled plasma mass spectrometry (ICP/MS). The analytical detection limit (ADL) for Cr+6 was calculated as three times the standard deviation of seven replicate injections of the lowest level standard, which was 0.038 ng. The method detection limit (MDL) for Cr+6 concentration was calculated to be 0.04 μg/g based on the ADL and the median of dust mass (4.81 mg) collected over the study period (0.192 ng/4.81 mg=0.04 μg/g). All the solvents used for sample preparation and analysis were checked for purity before use for field sample processing. Twenty field blanks (5.4% of the total dust samples) were collected throughout the study; this met the QA/QC goal of 5% field blank samples. The laboratory and field blank samples were analyzed using the same procedures as those for dust samples. No Cr+6 was detected in any of the field or lab blanks. Dust samples were analyzed for total dust loading (mg/m2) and Cr+6 concentration (μg/g) in dust. Cr+6 loading (μg/m2) was calculated based on Cr+6 concentration and total dust mass collected.

For additional quality control, 51 house dust samples (approximately 14% of the total dust samples) were collected side-by-side, and analyzed to estimate the method variability. We recognize that the spatial distribution of chromium in house dust is not homogeneous, and the side-by-side samples are not truly equivalent to duplicate samples. The mean±SD and median % difference between the side-by-side samples is 27±28% and 20%, respectively, with a range of 2–141%. This variability measured the method variation as well as the variability of chromium deposition on the surface.

Urine Sample Collection and Analysis

Spot urine samples were requested from all children under the age of 7 in the home. If possible, urine samples were collected during the dust sampling appointment. In the event the child would not be in the home during the dust sample collection, parents were mailed urine sample collection kits (instructions, a sealed specimen cup, and, if necessary, a pediatric sample collection bag) to collect the urine. Parents were asked to assist the child in collecting a sample in the specimen cup. Pediatric sample collection bags were provided for children not yet toilet-trained. If the sample could not be obtained during the appointment, parents were given the appropriate supplies and an appointment was made to pick up the sample at another time. Parents were asked to collect the sample no earlier than the day before the appointment and to refrigerate the sample until pick-up.

Before freezing, an aliquot of about 10 ml was sent to a commercial laboratory (Quest Diagnostics, Madison, NJ, USA) for creatinine concentration. The creatinine method used the kinetic spectrophotometric Jaffe method with limit of detection of 10 mg/dl. The remaining urine samples were frozen at −20oC until the time of analysis (up to 1 month). Urine samples were allowed to warm to about 16 oC before specific gravity was read using a Midget Urinometer (Fisher Scientific) with a range of 1.000–1.040 in increments of 0.001 with an accuracy of ±0.002. Total chromium was analyzed using a Perkin Elmer 5100 graphite furnace atomic absorption spectrometer with Zeeman correction. A matrix modifier (1000 p.p.m. Europium in 3% nitric acid) was used to reduce the background signal.19 A calibration curve was constructed that included a blank and four levels of chromium standards (0.25, 0.5, 1, and 2 μg/l). Correlation coefficients ranged from 0.9982 to 0.9998. Calibrations were verified using a Standard Reference Material (SRM) 2670a Low Level, “Toxic Elements in Urine (Freeze-Dried)” from the National Institute of Standards and Technology (NIST). Recoveries of the SRM (mean±SD, n=7) were found to average 96±8.6% ranging from 90% to 105%. For additional quality control, urine samples were analyzed in duplicate. The second aliquot was spiked with 0.5 μg/l chromium standard. Acceptable recovery measurements for the spiked sample are 60–130%. Samples with recoveries outside this range were rerun. Spike recoveries ranged from 62% to 126% with a mean of 93%. The limit of detection using this method was 0.02 μg/l.

Data Analysis

Statistical analyses were performed using SAS 9.2 (SAS Institute, Cary, NC, USA). Dust levels and urine concentrations were analyzed using non-parametric methods including the Wilcoxon–Mann–Whitney U-test (MWU), the Kruskal–Wallis (KW) one-way analysis of variance, and Spearman correlation. In all, 2 dust samples out of the total of 369 were below the MDL and were assigned one-half the MDL (0.02 μg/g and 1.1 ng/m2).20, 21 In all, 5 urine samples out of the total of 150 had chromium levels below the MDL and were assigned one-half the MDL (0.01 μg/l). Although creatinine correction is often used to address urine diluteness, the approach has been questioned for children, due to the variability in muscle development.22 As an alternative, specific gravity, which is based on solute concentration in the urine, was used. The corrected concentrations were examined for possible overcorrection. Samples were considered overcorrected if the specific gravity of the sample was in the 10th percentile or lower (1.005) and the corrected chromium concentration was in the 90th percentile or greater (0.48 μg/l). Six samples (one sample from a child less than 1 year old and five samples from 1 year old) were excluded based on possible overcorrection. All six had creatinine concentrations below 30 g/l. Univariate analyses were used for both uncorrected (Crur) and specific gravity-corrected (Crsg) urinary chromium concentrations. Urinary chromium concentrations were corrected to the mean specific gravity of all samples (1.016) by the formula: Crsg=Crur × (mean specific gravity−1)/(sample specific gravity−1).23

Urinary chromium concentrations (uncorrected and specific-gravity corrected) were compared with concentrations previously reported for children before remediation.13 Since only summary data were available for the earlier studies, a t-test, assuming unequal variances, was used to compare current levels with values reported by Stern et al.13 In the historical data set, 12 of the 67 children had values below the reported limit of detection (0.2 μg/l). In the publication, these samples had been assigned a value of 0.2 μg/l. Since this may have inflated the mean value, the mean for the uncorrected chromium concentrations was recalculated using 0.1 μg/l (one-half the limit of detection) for these 12 values as was done for samples with urinary chromium below the limit of detection in the current data set. The t-test was repeated using the corrected mean. No correction could be made for the specific-gravity corrected values.

The relationships between Cr+6 levels in dust and urinary chromium concentrations were examined using both Spearman correlation and multiple linear regression. Only urine samples collected within 7 days of the dust collection were used in the analyses. Multiple linear regression with urinary chromium concentration as the dependent variable that was conducted with backwards elimination was used. Only models that were significant (P<0.05) and retained a chromium dust level as an independent variable were evaluated. In some cases, two or three children from a single home participated in the study. Since the primary focus of this study was the exposure of the children rather than the condition of their homes and additionally, since exposure patterns vary by individual, when multiple children in the same home participated in the study, they were treated as independent observations in the statistical analysis. Multiple analyses were conducted in an exhaustive effort to disprove the study’s null hypothesis (i.e., there is no significant relationship between Cr+6 levels in house dust and total chromium concentrations in children’s urine). To reduce the likelihood of rejecting a possible association, no correction for multiple comparisons was made. In separate regression models, uncorrected and specific gravity-corrected urinary chromium concentrations, untransformed as well as natural log transformed, were used as the dependent variable. Each analysis was repeated using one of four exposure metrics to represent the Cr+6 levels in each home: the average concentration of all samples; the maximum concentration; the average loading; and the maximum loading. In each analysis, age and creatinine were independent variables in the regression model. When uncorrected urinary chromium was the dependent variable, specific gravity was also included as an independent variable. In the backwards elimination model, variables are eliminated if the P-value is <0.10. These analyses were performed multiple times by comparing the urine chromium metrics first to the entire dust sample data set, then to the following subsets: the child’s play area; the entryway; floor samples; and samples from all elevated surfaces.

Urinary chromium concentrations were also compared by exposure subgroups.3 Two classifications of exposure subgroups were based on the four dust metrics. Homes with Cr+6 dust levels exceeding the 75th percentile were defined as having high dust levels. Those exceeding the 90th percentile were defined as the highest dust levels. Using non-parametric tests (MWU), uncorrected and specific-gravity corrected urinary chromium concentrations (for samples collected within 7 days of the dust sample) were compared between the high (upper quartile) and all other homes and between the highest (upper decile) dust level homes and the remaining homes.

Results

Hexavalent Chromium Concentration and Loading in Dust Samples

In this phase II study, a total of 369 dust samples were collected from 123 homes between 18 August 2008 and 23 September 2010. Cr+6 was detected in all homes. Cr+6 concentrations (μg/g) were not significantly different between floors and elevated surfaces (Table 1). However, floor samples had significantly lower total dust loadings (mg/m2) (Mann–Whitney U MWU P=0.0011) and lower Cr+6 loadings (μg/m2) (MWU; P=0.0001). Relatively few floor samples were collected from either the background (23%) or phase I (10%) studies compared with the phase II study (47%). Therefore, only elevated surfaces were used to compare the studies. Overall, Cr+6 concentrations were not significantly different among the studies. However, both total dust loading and Cr+6 loadings were significantly lower in phase II (KW test; P<0.0001) for both dust and Cr+6 loadings. Both floor and elevated surface samples were used to compare areas of the home. No significant difference was found between entryway, child’s play area, and other living area samples for any outcome variable (Cr+6 concentration, Cr+6 loading, and dust loading). The investigation of housing and landscaping characteristics as possible predictors of Cr+6 in house dust is presented in Supplementary Table 1.

Table 1 Descriptive statistics for dust samples for current phase II study, phase I (elevated surfaces) and the background location study (elevated surfaces).1
Full size table

Cr Concentration in Urine Samples

Urine samples were collected from 150 children (77 boys; 73 girls) in 119 homes. Urine samples were collected from 1 child in 90 homes; from 2 children in 27 homes; and from 3 children in 2 homes. In four homes, dust samples were collected, but no urine sample was obtained. Most urine samples (82%) were collected within a day of the dust sampling. Seven urine samples were collected more than a week after the dust sample (8–106 days after the study visit). The children’s ages ranged from 1 month to 6 years with a median age of 4 years. The urine metrics were specific gravity, creatinine concentration, and total chromium concentration (uncorrected and specific gravity corrected) (Table 2). There was no significant sex difference, season difference, or morning versus evening sample collection time difference for any of the urinary metrics. Chromium concentration was not significantly correlated between children in the same home.

Table 2 Urine concentrations uncorrected and corrected for specific gravity.a
Full size table

Chromium concentration (uncorrected and SG corrected) was not correlated with age. However, creatinine and specific gravity were positively correlated with age (Spearman’s rho=0.432; P<0.0001 and rho=0.336; P<0.0001, respectively) and with each other (rho=0.777; P<0.001). Creatinine-corrected chromium concentration was also negatively correlated with age (Spearman’s rho=−0.177; P=0.0349). Both uncorrected and specific gravity corrected-urinary chromium concentrations were analyzed for associations with children’s reported activities (Table 3). Children reported to play outdoors in warm weather had marginally higher uncorrected chromium concentrations (P=0.089). In contrast to the warm weather results, children playing reported to play outdoors in cold weather had significantly lower corrected chromium concentrations than those who did not play outdoors (P=0.036).

Table 3 Children’s activities and urinary chromium concentration.
Full size table

The CEHS-II uncorrected concentrations (mean±SD, 0.22±0.16, n=150) were significantly lower (t-test; P<0.001) than those reported by Stern et al.13 in the 1990s (0.46±0.37, n=67) (Figure 1). The mean for the historic concentration was corrected by using one-half the limit of detection for the 12 samples below the LOD. Using the corrected mean (0.44 μg/l), the levels in the current data set were still significantly lower (t-test, P<0.001). Similar results were obtained for the specific gravity-corrected samples (0.23±0.23, n=150 versus 0.61±0.72, n=67).

Figure 1
figure 1

Comparison of current urinary chromium concentration in children 0–6 years with pre-remediation levels.

Full size image

Comparing Chromium in Urine and Dust

Regression Model

No significant correlations between chromium levels in dust and urinary chromium concentrations were found (Supplementary Table 2; Figure 2). Only regression models using the samples collected in the child’s play area and uncorrected urinary chromium concentration retained the chromium dust metrics as a significant factor (Table 4). The models describe an inverse relationship between each dust metric (average concentration, maximum concentration, average load, and maximum load) and uncorrected urinary chromium concentrations. This result is counter-intuitive and appears likely to be a chance association, particularly since the negative association disappeared when three influential points were removed from each model (the three highest concentration/loading values, respectively). When the same regression model was used without a dust variable (only age and creatinine as the dependent variables), little change in the adjusted R2 was noted. Similar results were observed using the natural log-transformed uncorrected urinary chromium concentrations. Thus, the dust values had little impact on the variance of the urinary chromium concentration.

Figure 2
figure 2

Child play area samples: uncorrected urinary chromium concentration versus average chromium concentration in house dust.

Full size image
Table 4 Significant (P<0.05) regression models for uncorrected urinary chromium concentration (μg/l) based on backward elimination using dust samples collected in the child’s play area.
Full size table

Subgroup Analysis

No significant difference in uncorrected or specific gravity-corrected urinary chromium concentrations was found between the high dust level homes (Cr+6 concentration or loading >75th percentile) and the remaining homes for any of the four dust metrics (see Supplementary Table 3). Children in homes with the highest average and highest maximum Cr+6 concentrations (Cr+6 concentrations >90th percentile) had significantly lower concentrations of uncorrected and specific gravity-corrected urinary chromium concentrations.

Discussion

Dust Results

These results confirm many of the findings from CEHS-I regarding Cr+6 levels in house dust in Jersey City and in house dust in homes background locations.7 Low but detectable levels of Cr+6 were found in all homes. Since all remediation had been completed before CEHS-I, no change in chromium dust levels was expected. Using only elevated surface samples, the Cr+6 concentrations in CEHS-II were not significantly different from those found in CEHS-I either in Jersey City or in background locations. However, the dust and Cr+6 loadings were significantly lower in the CEHS-II. The lower loadings may be attributed to the different areas of the home sampled in the phase II study. In CEHS-I, elevated surfaces were selected based on visible dust loading regardless of the room within the home. In CEHS-II, elevated surfaces in entryways and child play areas were targeted, regardless of the dust loading. Although the hexavalent chromium levels in Jersey City were not significantly different from those found in background areas, reducing exposure to a carcinogen to as low a level as reasonably achievable is desirable. Therefore, study participants in homes with elevated Cr+6 levels (a single sample greater than the NJ soil remediation guideline of 20 μg/g) were counseled on reducing exposure.

Urine Chromium

Before remediation, Stern et al.3, 13 found a positive association between chromium in dust and urine of young children. Median urinary chromium concentrations of 0.3 μg/l and greater were reported for children in Jersey City. In contrast, the CEHS-II found not significant association between concentrations of Cr+6 in house dust and urine of young children. Further, the median urinary concentration in CEHS-II, 0.19 μg/l, is similar to those reported for children from control (non-exposed) populations from the 1990s (Figure 1).13 While a t-test comparing the concentrations found a significant difference in urine levels in the 1990s and CEHS-II, the results need to be interpreted with caution since the data were not normally distributed. However, these results are consistent with findings in the 1990s that following remediation of chromium waste sites in Jersey City, chromium concentrations and loadings in house dust decreased to background levels.9, 11 The findings of CEHS-II suggest that there is no positive relationship between chromium in dust and chromium in urine.

The study found no significant seasonal, gender, or age differences in urinary chromium concentrations. Similar results were noted in an earlier study during chromium remediation.11 The effect of outdoor play on chromium concentrations is less clear. In the 1990s, before remediation, a positive correlation between time outdoors and children’s urinary chromium concentrations was observed.3 Further investigation conducted during remediation of surrounding sites confirmed a positive correlation for young children in the summer but no significant correlation in the fall season.11 The 1990s studies suggested the presence of an outdoor source of chromium exposure. The lack of association reported during the fall season was thought to be attributed to either the reduced number of hours outdoors during the season or the removal of the chromium source.11 In CEHS-II, children reported to play outdoors in cold weather were reported to have significantly lower, not higher, levels of uncorrected chromium (P=0.036). No significant associations were found for outdoor play in warm weather or hours of outdoor play in the preceding week. The lack of a positive association between urinary chromium levels and outdoor play in CEHS-II suggests that the children sampled did not encounter significant outdoor sources of chromium post-remediation.

The results suggest that the current levels of Cr+6 in house dust are not significant predictors of chromium exposure and probably not a significant source of total chromium exposure. Thus, the current situation is different from the pre-remediation era when house dust levels were higher and when studies demonstrated a statistically significant positive relationship between house dust concentrations of total chromium and urinary chromium concentrations in children.3, 13 The success of these studies in establishing a relationship between levels of total chromium in house dust and internal exposure has been attributed to the presence of a population with elevated exposures.3 Earlier work found declining levels of total chromium in house dust during and after remediation.9, 11 At low levels of chromium contamination, low variation in concentrations of Cr+6 in house dust would be difficult to detect by biomonitoring children’s urinary chromium, due to the variation in urinary chromium concentrations within the normal population from dietary sources and physiologic variability. One methodological difference between the current study and pre-remediation studies is the form of chromium measured in the dust. Although Cr+6 was the contaminant of concern, the 1990s studies were able to measure only total Cr, not Cr+6, in house dust and used total Cr as an indirect indicator of exposure to Cr+6. In contrast, the present study measured Cr+6 in house dust, increasing the accuracy of the exposure assessment.

Given the lack of association of urine Cr with house dust Cr+6, it is interesting to speculate as to the possible sources of variability in urine Cr of the children in this study. Cr is a trace nutrient. However, earlier studies failed to show an association between any particular food type and urine Cr.3, 24 Furthermore, dietary Cr is likely to be entirely Cr+3 and therefore, poorly absorbed from all dietary sources. Thus, while variability in diet is likely to contribute to variability in urine Cr, the nature of this contribution is unclear. Sources of Cr+6 (and to a lesser extent, Cr+3) in the environment other than house dust are likely to contribute to urine Cr. These include ambient air and Cr+6 in industrial products. Finally, it is not clear that any of the approaches employed in this study were completely successful in adjusting for urine diluteness. Thus, some portion of the variability in urine Cr concentration is likely to reflect residual unadjusted variability in urine diluteness.

As in the phase I and background studies, Cr+6 was found in all the homes tested including in homes in areas of New Jersey with no known Cr+6 source of contamination.7 Despite its widespread presence in homes, the sources of Cr+6 in households have not been identified, although the reported use of Cr+6 in wood stains may be a contributor.7 Other more general sources of chromium, including hexavalent chromium, are residential combustion, power plants, and metal industries.14 Oxidation of both anthropogenic and naturally occurring trivalent chromium may also contribute to the levels found in house dust. Identification of the sources of Cr+6 in background house dust is challenging. These background sources of Cr+6 in house dust may be significant in the persistence of Cr+6 allergic contact dermatitis.6

Conclusions

In contrast to our findings in pre-remediation studies of Cr exposure in Jersey City, there no longer appears to be an association between Cr+6 in house dust and Cr in the urine of children in the same homes. The urine chromium levels are lower than in the pre-remediation period. Consistent with our more recent findings, this appears to be associated with the remediation of chromium waste sites in Jersey City and to reflect the reduction of Cr+6 exposure in Jersey City to New Jersey urban background levels. In the larger context, this study serves to cap the series of studies of exposure to chromium waste in Jersey City (summarized before this study in Stern et al.8). As such, this study further emphasizes the utility of house dust sampling as an instrument for identifying exposure and tracking the efficacy of remediation. Further, and importantly, this study strengthens the case we have set forth previously8 for the combining house dust sampling with urine biomonitoring as a means for overcoming the inherent deficiencies in the use of each metric separately and for focusing the information that each can provide.