Environmental Research 135 (2014) 289–295

Contents lists available at ScienceDirect

Environmental Research journal homepage: www.elsevier.com/locate/envres

Spatial and seasonal variability of urinary trihalomethanes concentrations in urban settings Xanthi D. Andrianou, Pantelis Charisiadis, Syam S. Andra, Konstantinos C. Makris n Water and Health Laboratory, Cyprus International Institute for Environmental and Public Health in association with the Harvard School of Public Health, Cyprus University of Technology, Irenes 95, Limassol 3041, Cyprus

art ic l e i nf o

a b s t r a c t

Article history: Received 14 July 2014 Received in revised form 19 September 2014 Accepted 22 September 2014

A complex network of sources and routes of exposure to disinfection by-products (DBP), such as trihalomethanes (THM) has been driving the wide variability of daily THM intake estimates in environmental epidemiological studies. We hypothesized that the spatiotemporal variability of THM exposures could be differentially expressed with their urinary levels among residents whose households are geographically clustered in district-metered areas (DMA) receiving the same tap water. Each DMA holds unique drinking-water pipe network characteristics, such as pipe length, number of pipe leaking incidences, number of water meters by district, average minimum night flow and average daily demand. The present study assessed the spatial and seasonal variability in urinary THM levels among residents (n ¼310) of geocoded households belonging to two urban DMA of Nicosia, Cyprus, with contrasting water network properties. First morning urine voids were collected once in summer and then in winter. Results showed that the mean sum of the four urinary THM analytes (TTHM) was significantly higher during summer for residents of both areas. Linear mixed effects models adjusted for age, season and gender, illustrated spatially-resolved differences in creatinine-adjusted urinary chloroform and TTHM levels between the two studied areas, corroborated by differences observed in their pipe network characteristics. Additional research is warranted to shed light on the contribution of spatially-resolved and geographically-clustered environmental exposures coupled with internal biomarker of exposure measurements towards better understanding of health disparities within urban centers. & 2014 Elsevier Inc. All rights reserved.

Keywords: Trihalomethanes Exposure Disinfection Urban Mixed effects

1. Introduction Potable water disinfection is a widely practiced public health intervention to reduce the risk of waterborne infections, but it may also lead to the formation of a suite of compounds, collectively called disinfection by-products (DBP) (Brown et al., 2011; Chowdhury et al., 2009). The four trihalomethanes (THM), namely, chloroform (TCM), bromodichloromethane (BDCM), dibromochloromethane (DBCM) and bromoform (TBM) comprise the only DBP class currently regulated in the European Union, because they have been associated with increased risk of developing malignancies (i.e. bladder cancer) and adverse pregnancy outcomes (EU Abbreviations: BDCM, Bromodichloromethane; BMI, Body mass index; BrTHM, Brominated trihalomethanes; DBCM, Dibromochloromethane; DBP, Disinfection by-products; DMA, District-metered area; GM, Geometric mean; ICC, Intraclass correlation coefficient; LOD, Limit of detection; LOQ, Limit of quantification; TBM, Bromoform; TCM, Chloroform; THM, Trihalomethanes; TTHM, Total trihalomethanes; UDWDS, Urban drinking water distribution system n Corresponding author. Fax: þ 357 25002676. E-mail address: [email protected] (K.C. Makris). http://dx.doi.org/10.1016/j.envres.2014.09.015 0013-9351/& 2014 Elsevier Inc. All rights reserved.

Council, 1998; Patelarou et al., 2011; Richardson et al., 2007; Villanueva et al., 2004). However, since the putative causal agent or combination of agents of adverse health effects in the DBP mixture is currently unknown, THM may only serve as a surrogate of this agent or combination of agents in disinfected water. The ubiquitous human exposure to a mixture of putative agents, exhibiting general mutagenicity and potential carcinogenicity by all routes has urged the IARC to consider disinfected water in its priority list for evaluation during 2015–2019 (IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, 2014). The choice of external THM exposure measurements (in tap water) has for quite a while dominated over the use of biomarkers of exposure in several environmental epidemiological studies (Makris and Andra, 2014). Whole blood and exhaled breath have been used as biospecimen for measurements of THM exposures with mixed success (Gordon et al., 2006; Rivera-Núñez et al., 2012). THM levels in urine have been measured and moderately correlated with the intake of the compounds through drinking water and occupation (Cammann and Hübner, 1995; Caro and Gallego, 2007; Polkowska et al., 2006, 2003). Tap water THM

290

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

measurements have been characterized by a wide variation between and within water distribution systems (Summerhayes et al., 2011; Włodyka-Bergier et al., 2014). The situation is further complicated when one considers multiple noningestion exposure THM sources, such as showering, swimming, bathing and various activities of household cleaning (Villanueva et al., 2007; Whitaker et al., 2003). The water boards of major EU cities divide their urban drinking water distribution system (UDWDS) into district metered areas (DMA), which are smaller autonomous sub-networks with distinct geographical boundaries that are easier to manage. Improved assessment of environmental exposures at various urban scales, particularly those at the small-area level could greatly facilitate the assessment of health disparities in urban settings (Voigtländer et al., 2014). Small-area health disparities are considered to be primarily driven by differences in individual exposure to small-area resources and stressors (Voigtländer et al., 2014). It is often the case that health disparities among neighborhoods within the same city are larger than reported health inequalities between cities (Jonker et al., 2013; Pickett and Pearl, 2001). Since several tap water uses or activities (i.e. showering, bathing, swimming, mopping, dishwashing) have been associated with elevated THM exposures and THM levels in tap water are influenced by water distribution network characteristics, we hypothesized that inclusion of DMA clusters into THM exposure assessment studies could shed additional light on the effects of smallarea characteristics on individual health. These geographicallydistinct urban DMA (one in the north and the other located south of the city) have been historically receiving the same tap water treated in a single water treatment plant, but they are differentiated with respect to the following UDWDS characteristics: pipe length, number of pipe leaking incidences, number of water meters by district, average minimum night flow and average daily demand (Pieri et al., 2014). It was speculated that household distance from the main chlorination tank and the number of pipe leaking incidences may also influence the formation of THM within UDWDS (Pieri et al., 2014; H. Wang et al., 2012; Z. Wang et al., 2012). The objective of this study was to assess the relative contribution of spatial and seasonal characteristics on the magnitude and variability of THM exposures using urinary THM measurements among residents of two DMA in Nicosia, Cyprus. 2. Methods

In both occasions first morning void samples were collected in 60 ml polypropylene vials. Participants were given written instructions for collecting the sample without leaving headspace and preserving the samples in the freezer, until their collection and transportation to our laboratory facilities by our personnel. Before the analysis, all samples were stored in deep freezers at  80°C. Because of the absence of chlorine to quench and pH levels being o6, no need was warranted to add preservatives to urine samples. The urine samples were analyzed for TCM, BDCM, DBCM, and TBM with a modified protocol of U.S. EPA method 551.1-1 (Charisiadis and Makris, 2014). Concentrations below the limit of detection (LOD) were assigned 1 1 to LOD and values below the limit of quantification (LOQ) were assigned to LOQ. 2 2 More specifically, for the summer sampling the LOD (LOQ) values were: 94 (282) ng L  1, 46 (136) ng L  1, 31 (94) ng L  1, and 40 (120) ng L  1 for TCM, BDCM, DBCM, TBM, respectively. Similarly, the values for the winter sampling were: 27 (80) ng L  1, 11 (32) ng L  1, 24 (71) ng L  1, and 13 (40) ng L  1. Pooled urine samples with negligible THM levels were fortified to a final concentration of 700 ng L  1 and used for quality control. For the summer sampling the recoveries were 103%, 100%, 98% and 92% for TCM, BDCM, DBCM, and TBM, respectively, while the mean surrogate recovery was 82%. The winter sampling recoveries were 103% for TCM, 105% for BDCM, 102% for DBCM and TBM, with 85% mean surrogate recovery. The intra- and inter-day variability of the analytical measurements were always o 3.5%, with the average total THM recoveries in pooled urine matrix were 95–101% with 4% average relative standard deviation, for the summer measurements. For the winter sampling the intra- and inter-day variability were always o4.5%, with 94–110% average THM recoveries (for the 700 and 1000 ng L  1) (standard deviation of 5%) in pooled urine matrix. Urinary creatinine was determined by the picric acid based spectrophotometric method (Jaffe method) (Angerer and Hartwig, 2010). 2.3. Statistical analyses Descriptive statistics, including the calculation of geometric and arithmetic means, standard deviations, median values and percentiles, were performed for the urinary concentrations of all THM analytes and for important covariates (gender, age, BMI). Total THM (TTHM) and brominated THM (BrTHM) were defined as the sum of TCM, BDCM, DBCM, TBM, and the sum of BDCM, DBCM and TBM, respectively. Urinary THM concentrations were creatinine-adjusted to account for the urinary dilution (creatinine levels reported in the supplementary information) and log-transformed (natural logarithm) to meet the normality criterion due to skewness. The t-test was used to evaluate differences between paired summer and winter log-transformed values of the study participants for each THM analyte and one-way intraclass correlation coefficient (ICC) was calculated to assess the between-season reproducibility of the measurements. The statistical analyses were performed for the pooled study population and sorted by area. Linear mixed effects models were constructed to account for the between- and within-subject random variability in the urinary THM measurements; thus, the association between the creatinine-adjusted log-transformed THM concentrations and fixed effects covariates, such as, age, BMI, gender, area and season was assessed (Peretz et al., 2002). Correlations between the creatinine-adjusted urinary concentrations of all compounds and the corresponding tap water levels measured in the households of each participant were assessed with the Spearman coefficient for both seasons in the pooled sample and for each area separately. The statistical analyses were performed with R with the packages ‘nlme′, ‘lme4′, ‘lmerTest′ and ‘irr′ (Bates et al., 2014; Gamer et al., 2012; Kuznetsova et al., 2014; Pinheiro et al., 2014; R. Core Team, 2013).

2.1. Study population and design

3. Results The present work was conducted as part of a larger cross-sectional study of human exposures to THM in the city of Nicosia, Cyprus; the recruitment and sampling activities took place during summer 2012 and winter 2013 (Charisiadis et al., 2014). Randomized recruitment of participants took place within the geographical borders of two pre-selected DMA areas (herein referred to as area 1 and area 2) with contrasting UDWDS characteristics. The selection of these two DMA out of the total of 24 DMA for the city of Nicosia was based on a riskhierarching algorithm that classified area 1 with a higher frequency of pipe leaking incidences and higher average night water flow (indicative of pipe leaking-induced water losses and pipe age) than those for area 2 (detailed description in Pieri et al., 2014). After obtaining individual written consents, the residential addresses in both areas were geocoded and face-to-face interviews were performed with each participant in their household. Recorded demographic characteristics, lifestyle factors and water consumption habits, were previously described by Charisiadis et al. (2014). The study protocol was approved by the Cyprus National Bioethics Committee. 2.2. Sample collection and analyses Urine sample collection was conducted twice, during July–early September 2012 (season 1, summer) and during January–early March 2013 (season 2, winter).

3.1. Study population A total of 310 adults (Z 18 years) participated in this study (154 participants from area 1 and 156 from area 2) (Table 1). A greater proportion of females (61%) was observed, while the male to female ratio was similar for both areas (0.66 in area 1 and 0.63 in area 2). The average BMI and age of the study participants was 26 kg m  2 and 50 years, respectively. The majority of participants from both areas were married (75% in area 1 and 82% in area 2) with similar educational background (Table 1). Area 1 is located in the old part of the city, and characterized by an older pipe network system, 22 km long, 2489 registered water meters, and 2.9 pipe leaking incidences per km (Charisiadis et al., 2014; Pieri et al., 2014). Area 2 has a newer and longer (136 km) pipe network compared to area 1 with 7694 water meters and a lower frequency of pipe leaking incidences (1.1 km  1). The greater pipe network length of area 2 was accompanied by a 3  greater daily average

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

Table 1 Selected characteristics of the study participants. Characteristic

Gender Female Male

Overall

Area 1

Area 2

n

%

n

%

N

%

189 121

61 39

93 61

60 40

96 60

62 38

Age (y); (mean, std. dev., min–max) 50 (16; 18-87) 50 (18; 18-83)

50 (15; 18-87)

BMI (kg m  2) o25 25-30 430

131 126 53

42 41 17

65 65 24

42 42 16

66 61 29

42 39 19

Education Primary Secondary University Other/missing

96 129 77 8

31 42 25 3

53 59 35 7

34 38 23 5

43 70 42 1

28 45 27 1

Marital status Single Married Divorcee Widower Missing

42 243 7 15 3

14 78 2 5 1

22 115 5 9 3

14 75 3 6 2

20 128 2 6 0

13 82 1 4 0

demand (150 m3 h  1) and 5  higher average minimum night flow per pipe kilometer than that in area 1. Urinary creatinine concentrations ranged from 0.03 g L  1 to 3.86 g L  1 (Table S6), with certain samples having relatively high concentrations of creatinine, while no samples were found to be over-diluted. 3.2. Spatial and seasonal variation in urinary trihalomethanes 3.2.1. Seasonal trends Significant (po 0.001) differences in creatinine-adjusted urinary THM levels between summer and winter were observed for both the pooled sample of participants and separately for each area (Table 2). The creatinine-unadjusted and adjusted mean urinary BrTHM concentrations were significantly (p o0.001) higher in winter than in summer (primarily due to TBM). The opposite was observed for TTHM (Table 2 and Table S1). The higher urinary TTHM levels observed during summer for both areas could be explained by the higher TCM values during that season (GM for the creatinine-adjusted TCM concentrations: 326 and 102 μg g  1, for summer and winter, respectively). On the contrary, the higher urinary BrTHM levels in winter were primarily driven by the trends of DBCM and TBM (Table 2). The differences in measured THM levels were corroborated by low ICC, indicating poor

291

reproducibility of the measurements between seasons (ICC ¼0.25, 95% CI [0.14–0.35] for the creatinine-adjusted TTHM). When data from both areas was pooled, season was a significant predictor (p o0.001) of urinary THM with positive coefficients for TCM, BDCM, and the TTHM in summer after adjusting for age, gender, BMI and area (Table 3); for DBCM, TBM and BrTHM the opposite trend was observed (i.e. negative coefficient for the summer). 3.2.2. Spatial trends The effect of area, as a proxy for the spatial differentiation of the two studied DMA, was found to be a significant predictor of TCM and TTHM levels, after adjusting for age, gender, BMI and season (Table 3). The spatial trends were further corroborated by area's 1 pipe network characteristics (older pipe network, higher frequency of water leaking events, and deteriorated condition of pipes) that prompted the formation of such compounds (Makris et al., 2014) (Table 3). Analysis of tap water THM levels showed that households of study participants in area 1 had higher TCM and TTHM concentrations (mean values of 21 μg L  1 and 70 μg L  1, respectively) compared to those levels in the households of participants from area 2 (mean value of 17 μg L  1 and 61 μg L  1 for TCM and TTHM, respectively) (data not shown). The Spearman coefficient that was calculated to assess possible correlations between the urinary and household tap water levels of each compounds in the pooled sample and both areas, ranged from  0.41 (p-value o0.001, creatinine-adjusted TCM levels; summer sampling of area 2) to 0.35 (p-value o0.001, creatinine-adjusted TCM levels; winter sampling of the pooled population). The winter levels of the total creatinine-adjusted urinary THM concentrations of the pooled sample were positively correlated with the tap water concentrations, however the value of the coefficient was low (ρ ¼0.14, pvalue ¼0.017). No significant correlation between the water and urinary THM levels was found for any of the compounds measured in area 1. In area 2, significant correlations were found for summer and winter TCM levels (ρ ¼  0.41, p-value o0.001 and ρ ¼0.31, pvalue o0.001, for summer and winter, respectively) and a negative correlation was also observed for the total THM levels measured during the summer sampling (ρ ¼  0.36, p-value o 0.001). When the mixed effects models were repeated with the inclusion of water THM levels, this covariate was not a significant predictor of the urinary THM concentrations, corroborating the limited correlation between the two variables. Linear mixed effects models performed by area with age, gender, BMI and season as fixed predictors were constructed again (Table S2). Similar trends as those observed in the pooled sample were observed, i.e. season, age and gender were significant predictors in all models except for the TCM model of area 2 where age was non-significant (p-value ¼0.07) and the model for BDCM for the same area where BMI was significant predictor (pvalue ¼0.03) and gender was non-significant (p ¼0.10). The t-tests

Table 2 Creatinine-adjusted urinary concentration differences between seasons in the pooled sample and the two areas. Both areas

Area 1

Summer Winter p-Value Creatinine-adjusted urinary concentration (ng g  1)a TCM BDCM DBCM TBM BrTHM TTHM a

608 131 77 32 239 847

243 61 119 147 326 569

o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 o 0.001

Area 2

Summer

Winter

p-Value

Summer

Winter

p-Value

685 179 80 33 293 977

312 37 115 140 292 604

o0.001 o0.001 o0.001 o0.001 o0.001 o0.001

532 83 73 31 186 718

176 85 122 153 360 535

o 0.001 o 0.001 o 0.001 o 0.001 o 0.001 0.009

The arithmetic mean of the creatinine-adjusted measurements is presented along with the p-value of the paired t-test for the log-transformed values.

292

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

Table 3 Linear mixed model effects parameters for the pooled (both areas) sample. (ln) creatinine-adjusted urinary concentration TCM BDCM

Area (Area 1) Age (y) Gender (female) BMI (kg m  2) Season (summer)

DBCM

TBM

BrTHM

TTHM

β

p-Value

β

p-Value

β

p-Value

β

p-Value

β

p-Value

β

p-Value

0.956 0.013 0.451  0.005 1.162

o0.001 o0.001 o0.001 0.64 o0.001

–0.306 0.013 0.347  0.012 1.615

0.003 o 0.001 o 0.001 0.24 o 0.001

0.020 0.017 0.494  0.004  0.534

0.73 o0.001 o0.001 0.46 o0.001

0.011 0.018 0.462  0.005  1.606

0.84 o 0.001 o 0.001 0.39 o 0.001

0.042 0.016 0.438  0.008  0.425

0.48 o 0.001 o 0.001 0.19 o 0.001

0.292 0.015 0.455  0.009 0.268

o0.001 o0.001 o0.001 0.18 o0.001

confirmed the spatial differentiation of urinary TCM, BDCM and TTHM levels between the two areas in both (summer and winter) seasonal sampling schemes (Table S3). In the aforementioned mixed effects models, subjects were included as a random variable and the magnitude of the betweenand within-subject variability was assessed (Table 4). In none of the models did the between-subject variability exceed 36% of the total, which was the case for TBM among participants of area 2. Moreover, TBM was, consistently, the analyte with the highest between-subject variability in the pooled sample (33%). TCM showed low between-subject variability ranging from 3% of the total for the pooled sample model, and up to 11% and 8% for the models of areas 1 and 2, respectively. When the models were repeated with the inclusion of season as a random variable it accounted for a higher percentage of the total variability than the between-subject variability, which decreased in all models (Tables S4 and S5).

4. Discussion This is the first study that differentially assigned data on urinary biomarkers of THM exposures to two geographicallydefined clusters of urban dwellers, based on their respective DMA's drinking-water pipe network characteristics. Significant differences between the two urban areas in creatinine-adjusted urinary TTHM and TCM levels were observed (area 14 area 2, Table S3), even after adjusting for age, sex and season, corroborating the older pipe network characteristics of area 1. However, this was not the case for urinary BrTHM; we speculated that the lipophilic nature of BrTHM and the similar quality characteristics (soluble bromine content and natural organic matter) of the finished tap water delivered to both areas, could partially explain the insensitivity of BrTHM to spatial changes in urinary BrTHM levels of the two studied areas; the total soluble bromide concentrations in tap water did not differ between the two areas, being 0.16 and 0.17 mg L  1 in area 1 and 2, respectively, while total organic carbon concentrations of tap water were 3.6 and 2.9 mg L  1 in area 1 and 2, respectively. Area-resolved differences in everyday activities involving dermal uptake that could also

explain the aforementioned observation about BrTHM were also not observed (data not shown). Habitual patterns of showering, swimming in indoor pools and household cleaning during summer did not differ between residents of the two areas (data not shown), hinting towards differences in tap water TTHM and TCM levels and fluctuations in diurnal within-subject TTHM levels to explain their spatially-resolved urinary levels. Similar questionnaire-based information for winter sampling (i.e. patterns of water contact related activities and possible THM exposures) was not available for inclusion in the analysis. Additionally, differences in urine/air partition coefficients could partially explain the between-subject variability associated with each THM (Batterman et al., 2002); The TBM with the highest urine/air partition coefficient (20.74) demonstrated the highest percent of between-subject variability (33%), while TCM with the lowest urine/air partition coefficient (3.14) of the four THM exhibited the lowest between-subject variability (3%). The older pipe age, the higher frequency of occurrence for pipe leaking incidences and the higher average minimum night flow (indicator of water leaking incidences) were greatly manifested in the pipe network of area 1 and much less in area 2, supporting the higher urinary TTHM levels of area 1. The older pipe network and the frequency of water pipe leaking events in area 1 could be indicative of enhanced biofilm growth on pipe surfaces and increased residual chlorine consumption rates, leading to greater THM formation potential (H. Wang et al., 2012; Z. Wang et al., 2012). Residual chlorine contact in finished water with bacterial exopolymeric biomass of biofilm colonies (Pseudomonas strains) enhanced the formation of biofilm-generated THM (Z. Wang et al., 2012). Residual disinfectant consumption as a function of water age was observed in water distribution networks, allowing for: (i) increased densities of opportunistic microbes (H. Wang et al., 2012) and (ii) concomitantly enhanced THM formation in distant neighborhoods away from the main chlorination tank (Pieri et al., 2014). Water residence time was shown to influence the magnitude of water TTHM concentrations between sites served by the same water distribution network, indicating the effect of intra-system variation on water THM levels (Loyola-Sepulveda et al., 2013).

Table 4 Variance component estimates for the linear mixed effect models of the pooled sample and by area. Both areas

TCM BDCM DBCM TBM BrTHM TTHM

Area 1

Area 2

Betweensubject

Withinsubject

% Between-subject Betweenvariability subject

Withinsubject

% Between-subject Betweenvariability subject

Withinsubject

% Between-subject variability

0.05 0.00 0.12 0.12 0.08 0.10

1.55 1.66 0.27 0.25 0.38 0.47

3 0 31 33 17 17

0.75 1.55 0.28 0.26 0.50 0.39

11 0 29 31 10 20

2.04 1.56 0.25 0.23 0.25 0.54

8 4 34 36 32 15

0.09 0.00 0.12 0.12 0.06 0.10

0.17 0.06 0.13 0.13 0.12 0.10

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

Seasonal effects on urinary TTHM and BrTHM levels were noted for the pooled sample and for each of the two areas, separately. Consistently, linear mixed effects models showed higher urinary TCM and TTHM in summer, while, lower urinary BrTHM in summer were observed for both areas, indicating a seasonal pattern in THM exposures. No data on seasonal patterns of urinary THM exist in the literature. However, similar seasonal patterns in blood TTHM have been previously reported where summer blood TTHM levels were higher than those in winter for a population of postpartum women (Rivera-Núñez et al., 2012). Being the most volatile THM analyte, TCM is expected to partition to the gaseous phase during periods of elevated temperatures (Mediterranean summer), highlighting inhalation as a major noningestion route of TCM exposure. The lower BrTHM urinary levels in summer could perhaps be explained by the fact that these compounds are less volatile than TCM and subject to preferential dermal absorption due to their lipophilic properties (Silva et al., 2013). Additionally, soluble bromine levels during summer for both areas were below detection limit (o2 mg L−1) and the ratio of desalinated to conventionally treated tap water was higher during that season. Complementarily, linear mixed effects models showed that within-subject variability for urinary TTHM was consistently greater than the between-subject variability, when adjusting for covariates such as area, age and BMI. Area was not a significant predictor of urinary BrTHM levels, as mentioned earlier, therefore the observed spatial differences, as indicated by the significance of area in the linear mixed effects model of the TTHM, may be primarily driven by the contribution of TCM. Socioeconomic status may also impact the magnitude of THM exposures, but questionnaire responses did not reveal significant differences in education and marital status between residents of the two urban areas (data not shown). The two studied areas are about 15 km apart and no major socioeconomic disparities are expected among their residents. The possible link between water THM levels and societal factors was earlier examined in different areas of Massachusetts, BA, USA, but no such association was found (Evans et al., 2013). In Spain, subjects with higher socioeconomic status that resided in areas with higher levels of DBP had higher THM exposure through non-ingestion routes (inhalation and dermal absorption), stemming from their increased consumption of THM-free bottled water and enhanced showering frequency and indoor swimming pool use (Castaño-Vinyals et al., 2011). Additionally, physiological differences in the metabolic pathways of all four THM compounds may hinder our progress in improving exposure assessment for the benefit of community health studies. The quantitative characterization of THM disposition in humans has received much attention and it is often shown to vary according to personal habits and lifestyle factors, genetic polymorphisms, CYP enzyme phenotypic activity, medications used and disease status (Blount et al., 2011; Riederer et al., 2014). In physiologically-based toxicokinetic models that were constructed to assess uptake and metabolism via indoor exposures, parent THM compounds were fully metabolized after oral absorption (Haddad et al., 2006). Moreover, the fraction of the metabolized dose through dermal absorption and inhalation was found to be associated with the lipophilicity of the compounds. TCM (the least lipophilic) was the least absorbed and TBM (the most lipophilic) the most metabolized; the fraction of absorbed dose metabolized via inhalation was 0.73 for TCM and 0.86 for TBM (Haddad et al., 2006). The genetic background of individuals through the phenotypic activity of select enzymes, in particular, that of CYP2E1 may be impacting THM metabolism (Backer et al., 2008; Gemma et al., 2003; Leavens et al., 2007). The metabolic activation of chloroform by CYP2E1 in mice (Constan et al., 1999) and human liver microsomes (Gemma et al., 2003) was the rate-

293

limiting step in determining the magnitude and extent of toxicity of chloroform's metabolites (phosgene) to liver and kidney function. Backer et al. (2008) found that polymorphisms in the genes CYP2D6 and GSTT1 were predictors of blood THM concentrations after controlling for exposures to THM through activities such as swimming activity or hot water intake (Backer et al., 2008). The activity of the CYP2E1 enzyme was also associated with the metabolism of TCM and BDCM; impaired CYP2E1 activity affected TCM metabolism, suggesting that its activity increased with body weight (Allis and Zhao, 2002; Brill et al., 2012; Gemma et al., 2003; Leavens et al., 2007). Urinary THM measurements as biomarkers of THM exposures have not been extensively used in health studies, therefore, their use as a biomarker integrating across exposure sources and pathways remains limited. Considering the short half-lives (o1 h) of THM and the fact that we collected first morning urine void spot samples, the observed urinary THM levels may represent the background concentration of THM, which is the result of volatilized mass of THM during various tap water uses, including the frequency and extent of use of cleaning products and activities. Further, the use of urine as a matrix for THM measurements presents us with practical advantages, such as its non-invasive sampling nature. The magnitude of THM in urine has been reported to be comparable to that measured in the systemic circulation (Cammann and Hübner, 1995). Cammann and Hübner (1995) measured both blood and urine THM concentrations in workers and swimmers in indoor swimming pools; measurements were moderately correlated (r ¼0.6, calculated with data read from figures in the paper). Use of urinary THM measurements was also practiced by Caro and Gallego (2007) who studied workers and swimmers at an indoor swimming pool with controlled sampling during and after work shift. The urinary THM concentrations in Caro and Gallego (2007) reflected occupational and perhaps higher exposures than those experienced by the general population (Caro and Gallego, 2007). Our group has shown that urinary THM measurements in a (sub)population group in Nicosia, Cyprus quantitatively reflected specific questionnaire responses capturing volunteer's lifestyle habits that related to THM formation (Charisiadis et al., 2014). Our lab is currently performing experiments tackling with the within-subject diurnal variability of urinary THM levels when individuals participated into routine daily household cleaning and personal care (showering) activities. An important limitation of our study is the lack of information on lifestyle and personal habits, such as the duration and type of each daily activity contributing to daily THM intake estimates for both sampling periods. Another limitation is the number of samples that were available per study participant. Additionally, people commonly use more than one sources of drinking water in a day, for example at work or at school the water source could perhaps differ from the one used at home. THM levels from all the drinking water sources used per subject could not be measured for in the present study. First morning voids may not be indicative of the intra-subject diurnal variability in THM intake across the entire day. Thus, the use of two or more repeated measurements from the same subjects and the inclusion of multiple measurements of the exposure through all possible routes (i.e. THM levels in household tap water, in the “away-from-home” water source, or in inhaled breath during showering) could help better address the issue of exposure misclassification. The implications of this study are important, because historically-established geographic boundaries within a city as defined by those of the drinking-water pipe infrastructure, i.e., the DMA, could find use for the improved exposure assessment of tap water contaminants, including DBP. The need for epidemiologic studies that address the challenges of exposure assessment for tap and bottled water contaminants is well documented (Villanueva et al.,

294

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

2014). Increased urbanization may be associated with deleterious health outcomes, but standardized methodology to allow comparisons across various urban scales, geographies and cultures is currently lacking (Cyril et al., 2013). Large-scale urban regeneration and neighborhood renewal programs attempt to lead to improvements in living conditions, economic opportunities, and population health in cities. If exposures to water or other contaminants are found to be clustered within geographically-distinct urban areas (i.e. water DMA) the gathered information could provide a unique opportunity for the development of targeted and cost-effective population health interventions. In the present study, the urinary THM measurements of residents in two DMA with contrasting pipe network characteristics were found to be spatially and seasonally differentiated within the boundaries of the same city. If current urban geography and zip code classification schemes allow for clustering of neighborhoods with similar exposure characteristics, health disparities and socioeconomic inequalities could perhaps more effectively be addressed allowing for optimal resource allocation. The way forward is the development of protocols and methodologies of “all-inclusive” studies that take into consideration all parameters implicated in exposure, social and health disparities.

Competing financial interests None.

Acknowledgments Konstantinos C. Makris would like to express his gratitude for the financial support from the Cyprus Research Promotion Foundation, the EU Structural Funds and Cohesion Fund in Cyprus according to the Cypriot Operational Program “Sustainable Development and Competitiveness” (AEIFORIA/ASTI/0311(BIE)/20).

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.envres.2014.09.015.

References Allis, J.W., Zhao, G., 2002. Quantitative evaluation of bromodichloromethane metabolism by recombinant rat and human cytochrome P450s. Chem. Biol. Interact. 140, 137–153. http://dx.doi.org/10.1016/S0009-2797(02)00022-4. Angerer, J., Hartwig, A. (Eds.), 2010. The MAK-Collection for Occupational Health and Safety, Part IV: Biomonitoring Methods, Vol. 12. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, pp. 169–184. Backer, L.C., Lan, Blount, Q., Nuckols, B.C., Branch, J.R., Lyu, R., Kieszak, C.W., Brinkman, S.M., Gordon, M.C., Flanders, S.M., Romkes, W.D., Cantor, K.P., M., 2008. Exogenous and endogenous determinants of blood trihalomethane levels after showering. Environ. Health Perspect. 116, 57–63. http://dx.doi.org/ 10.1289/ehp.10049. Bates, D., Maechler, M., Bolker, B., Walker, S., 2014. lme4: Linear mixed-effects models using Eigen and S4. Batterman, S., Zhang, L., Wang, S., Franzblau, A., 2002. Partition coefficients for the trihalomethanes among blood, urine, water, milk and air. Sci. Total Environ. 284, 237–247. http://dx.doi.org/10.1016/S0048-9697(01)00890-7. Blount, B.C., Backer, L.C., Aylward, L.L., Hays, S.M., LaKind, J.S., 2011. Human exposure assessment for DBPs: factors influencing blood trihalomethane levels. In: Nriagu, J.O. (Ed.), Encyclopedia of Environmental Health. Elsevier, Burlington, pp. 100–107. Brill, M.J.E., Diepstraten, J., Rongen, A., van, Kralingen, S., van, Anker, J.N., van den, Knibbe, P.C.A.J., 2012. Impact of obesity on drug metabolism and elimination in adults and children. Clin. Pharmacokinet. 51, 277–304. http://dx.doi.org/ 10.2165/11599410-000000000-00000.

Brown, D., Bridgeman, J., West, J.R., 2011. Predicting chlorine decay and THM formation in water supply systems. Rev. Environ. Sci. Biotechnol. 10, 79–99. http://dx.doi.org/10.1007/s11157-011-9229-8. Cammann, K., Hübner, K., 1995. Trihalomethane concentrations in swimmers' and bath attendants' blood and urine after swimming or working in indoor swimming pools. Arch. Environ. Health Int. J. 50, 61–65. http://dx.doi.org/ 10.1080/00039896.1995.9955013. Caro, J., Gallego, M., 2007. Assessment of exposure of workers and swimmers to trihalomethanes in an indoor swimming pool. Environ. Sci. Technol. 41, 4793–4798. http://dx.doi.org/10.1021/es070084c. Castaño-Vinyals, G., Cantor, K.P., Villanueva, C.M., Tardon, A., Garcia-Closas, R., Serra, C., Carrato, A., Malats, N., Rothman, N., Silverman, D., Kogevinas, M., 2011. Socioeconomic status and exposure to disinfection by-products in drinking water in Spain. Environ. Health 10, 18. http://dx.doi.org/10.1186/1476-069X-10-18. Charisiadis, P., Andra, S.S., Makris, K.C., Christodoulou, M., Christophi, C., Kargaki, S., Stephanou, E., 2014. Household cleaning activities as non-ingestion exposure determinants of urinary trihalomethanes. Environ. Sci. Technol. 48, 770–780. http://dx.doi.org/10.1021/es404220z. Charisiadis, P., Makris, K.C., 2014. A sensitive and fast method for trihalomethanes in urine using gas chromatography–triple quadrupole mass spectrometry. J. Chromatogr. B 947–948, 17–22. http://dx.doi.org/10.1016/j.jchromb.2013.11. 060. Chowdhury, S., Champagne, P., McLellan, P.J., 2009. Models for predicting disinfection byproduct (DBP) formation in drinking waters: a chronological review. Sci. Total Environ. 407, 4189–4206. http://dx.doi.org/10.1016/j.scitotenv.2009. 04.006. Constan, A.A., Sprankle, C.S., Peters, J.M., Kedderis, G.L., Everitt, J.I., Wong, B.A., Gonzalez, F.L., Butterworth, B.E., 1999. Metabolism of chloroform by cytochrome P450 2E1 is required for induction of toxicity in the liver, kidney, and nose of male mice. Toxicol. Appl. Pharmacol. 160, 120–126. http://dx.doi.org/ 10.1006/taap.1999.8756. Cyril, S., Oldroyd, J.C., Renzaho, A., 2013. Urbanisation, urbanicity, and health: a systematic review of the reliability and validity of urbanicity scales. BMC Public Health 13, 513. http://dx.doi.org/10.1186/1471-2458-13-513. EU Council, 1998. Council Directive 98/83/EC of 3 November 1998 on the quality of water intended for human consumption. Evans, A.M., Wright, J.M., Meyer, A., Rivera-Núñez, Z., 2013. Spatial variation of disinfection by-product concentrations: exposure assessment implications. Water Res. 47, 6130–6140. http://dx.doi.org/10.1016/j.watres.2013.07.032. Gamer, M., Lemon, J., Fellows, I., Singh, P., 2012. irr: Various Coefficients of Interrater Reliability and Agreement. Gemma, S., Vittozzi, L., Testai, E., 2003. Metabolism of chloroform in the human liver and identification of the competent P450s. Drug Metab. Dispos. 31, 266–274. http://dx.doi.org/10.1124/dmd.31.3.266. Gordon, S.M., Brinkman, M.C., Ashley, D.L., Blount, B.C., Lyu, C., Masters, J., Singer, P. C., 2006. Changes in breath trihalomethane levels resulting from household water-use activities. Environ. Health Perspect. 114, 514–521. Haddad, S., Tardif, G.-C., Tardif, R., 2006. Development of physiologically based toxicokinetic models for improving the human indoor exposure assessment to water contaminants: trichloroethylene and trihalomethanes. J. Toxicol. Environ. Health A 69, 2095–2136. http://dx.doi.org/10.1080/15287390600631789. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, 2014. IARC, Lyon, France (Internal Report No. 14/002). Jonker, M.F., Congdon, P.D., van Lenthe, F.J., Donkers, B., Burdorf, A., Mackenbach, J. P., 2013. Small-area health comparisons using health-adjusted life expectancies: a Bayesian random-effects approach. Health Place 23, 70–78. http://dx.doi. org/10.1016/j.healthplace.2013.04.003. Kuznetsova, A., Brockhoff, P.B., Christensen, R.H.B., 2014. lmerTest: tests for random and fixed effects for linear mixed effect models (lmer objects of lme4 package). Leavens, T.L., Blount, B.C., DeMarini, D.M., Madden, M.C., Valentine, J.L., Case, M.W., Silva, L.K., Warren, S.H., Hanley, N.M., Pegram, R.A., 2007. Disposition of bromodichloromethane in humans following oral and dermal exposure. Toxicol. Sci. 99, 432–445. http://dx.doi.org/10.1093/toxsci/kfm190. Loyola-Sepulveda, R., Mudge, S.M., Bravo-Linares, C., 2013. Seasonal variation of disinfection by-products in drinking water in Central Chile. Water Qual. Expo. Health 5, 1–9. http://dx.doi.org/10.1007/s12403-012-0082-y. Makris, K.C., Andra, S.S., 2014. Limited representation of drinking-water contaminants in pregnancy–birth cohorts. Sci. Total Environ. 468–469, 165–175 http://dx.doi.org/10.1016/j.scitotenv.2013.08.012. Makris, K.C., Andra, S.S., Botsaris, G., 2014. Pipe scales and biofilms in drinkingwater distribution systems: undermining finished water quality. Crit. Rev. Environ. Sci. Technol., 0. http://dx.doi.org/10.1080/10643389.2013.790746. Patelarou, E., Kargaki, S., Stephanou, E.G., Nieuwenhuijsen, M., Sourtzi, P., Gracia, E., Chatzi, L., Koutis, A., Kogevinas, M., 2011. Exposure to brominated trihalomethanes in drinking water and reproductive outcomes. Occup. Environ. Med. 68, 438–445. http://dx.doi.org/10.1136/oem.2010.056150. Peretz, C., Goren, A., Smid, T., Kromhout, H., 2002. Application of mixed-effects models for exposure assessment. Ann. Occup. Hyg. 46, 69–77. http://dx.doi.org/ 10.1093/annhyg/mef009. Pickett, K.E., Pearl, M., 2001. Multilevel analyses of neighbourhood socioeconomic context and health outcomes: a critical review. J. Epidemiol. Commun. Health 55, 111–122. http://dx.doi.org/10.1136/jech.55.2.111. Pieri, P., Andra, S.S., Charisiadis, P., Demetriou, G., Zambakides, N., Makris, K.C., 2014. Variability of tap water residual chlorine and microbial counts at spatially resolved points of use. Environ. Eng. Sci. , http://dx.doi.org/10.1089/ees.2013.0480 (140324140110003).

X.D. Andrianou et al. / Environmental Research 135 (2014) 289–295

Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2014. nlme: Linear and nonlinear mixed effects models. Polkowska, Ż., Kozłowska, K., Mazerska, Z., Górecki, T., Namieśnik, J., 2003. Relationship between volatile organohalogen compounds in drinking water and human urine in Poland. Chemosphere 53, 899–909. http://dx.doi.org/ 10.1016/S0045-6535(03)00586-1. Polkowska, Ż., Kozłowska, K., Mazerska, Z., Górecki, T., Namieśnik, J., 2006. Volatile organohalogen compounds in human urine: the effect of environmental exposure. Chemosphere 62, 626–640. http://dx.doi.org/10.1016/j.chemosphere. 2005.05.036. R. Core Team, 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Richardson, S.D., Plewa, M.J., Wagner, E.D., Schoeny, R., Demarini, D.M., 2007. Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap for research. Mutat. Res. 636, 178–242. http://dx.doi.org/10.1016/j.mrrev.2007.09.001. Riederer, A.M., Dhingra, R., Blount, B.C., Steenland, K., 2014. Predictors of blood trihalomethane concentrations in NHANES 1999–2006. Environ. Health Perspect. , http://dx.doi.org/10.1289/ehp.1306499. Rivera-Núñez, Z., Wright, J.M., Blount, B.C., Silva, L.K., Jones, E., Chan, R.L., Pegram, R. A., Singer, P.C., Savitz, D.A., 2012. Comparison of trihalomethanes in tap water and blood: a case study in the United States. Environ. Health Perspect. 120, 661–667. http://dx.doi.org/10.1289/ehp.1104347. Silva, L.K., Backer, L.C., Ashley, D.L., Gordon, S.M., Brinkman, M.C., Nuckols, J.R., Wilkes, C.R., Blount, B.C., 2013. The influence of physicochemical properties on the internal dose of trihalomethanes in humans following a controlled showering exposure. J. Expo. Sci. Environ. Epidemiol. 23, 39–45. http://dx.doi. org/10.1038/jes.2012.80. Summerhayes, R.J., Morgan, G.G., Lincoln, D., Edwards, H.P., Earnest, A., Rahman, M. B., Byleveld, P., Cowie, C.T., Beard, J.R., 2011. Spatio-temporal variation in trihalomethanes in New South Wales. Water Res. 45, 5715–5726. http://dx. doi.org/10.1016/j.watres.2011.08.045.

295

Villanueva, C.M., Cantor, K.P., Cordier, S., Jaakkola, J.J.K., King, W.D., Lynch, C.F., Porru, S., Kogevinas, M., 2004. Disinfection byproducts and bladder cancer: a pooled analysis. Epidemiology 15, 357–367. Villanueva, C.M., Gagniere, B., Monfort, C., Nieuwenhuijsen, M.J., Cordier, S., 2007. Sources of variability in levels and exposure to trihalomethanes. Environ. Res. 103, 211–220. http://dx.doi.org/10.1016/j.envres.2006.11.001. Villanueva, C.M., Kogevinas, M., Cordier, S., Templeton, M.R., Vermeulen, R., Nuckols, J.R., Nieuwenhuijsen, M.J., Levallois, P., 2014. Assessing exposure and health consequences of chemicals in drinking water: current state of knowledge and research needs. Environ. Health Perspect. 122, 213–221. http://dx.doi. org/10.1289/ehp.1206229. Voigtländer, S., Vogt, V., Mielck, A., Razum, O., 2014. Explanatory models concerning the effects of small-area characteristics on individual health. Int. J. Public Health 59, 427–438. http://dx.doi.org/10.1007/s00038-014-0556-8. Wang, H., Masters, S., Hong, Y., Stallings, J., Falkinham, J.O., Edwards, M.A., Pruden, A., 2012a. Effect of disinfectant, water age, and pipe material on occurrence and persistence of Legionella, mycobacteria, Pseudomonas aeruginosa, and two amoebas. Environ. Sci. Technol. 46, 11566–11574. http://dx.doi.org/10.1021/ es303212a. Wang, Z., Kim, J., Seo, Y., 2012b. Influence of bacterial extracellular polymeric substances on the formation of carbonaceous and nitrogenous disinfection byproducts. Environ. Sci. Technol. 46, 11361–11369. http://dx.doi.org/10.1021/ es301905n. Whitaker, H.J., Nieuwenhuijsen, M.J., Best, N.G., 2003. The relationship between water concentrations and individual uptake of chloroform: a simulation study. Environ. Health Perspect. 111, 688. Włodyka-Bergier, A., Bergier, T., Kot, M., 2014. Occurrence of volatile organic chlorination by-products in water distribution system in Krakow (Poland). Desalin. Water Treat. 52, 3898–3907. http://dx.doi.org/10.1080/19443994. 2014.887490.

Spatial and seasonal variability of urinary trihalomethanes concentrations in urban settings.

A complex network of sources and routes of exposure to disinfection by-products (DBP), such as trihalomethanes (THM) has been driving the wide variabi...
341KB Sizes 4 Downloads 8 Views