lable at ScienceDirect

Environmental Pollution 205 (2015) 70e77

Contents lists avai

Environmental Pollution

journal homepage: www.elsevier .com/locate/envpol

Risk of human exposure to polycyclic aromatic hydrocarbons: A casestudy in Beijing, China

Yanxin Yu a, Qi Li a, Hui Wang a, Bin Wang b, *, Xilong Wang c, Aiguo Ren b, Shu Tao c

a College of Water Science, Beijing Normal University, Beijing 100875, PR Chinab Institute of Reproductive & Child Health/Ministry of Health Key Laboratory of Reproductive Health, School of Public Health, Peking University,Beijing 100191, PR Chinac Laboratory for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, Beijing 100871, PR China

a r t i c l e i n f o

Article history:Received 6 March 2015Received in revised form15 May 2015Accepted 17 May 2015Available online xxx

Keywords:Polycyclic aromatic hydrocarbonDietInhalationILCRCancer risk

* Corresponding author.E-mail address: binwangpku@foxmail.com (B. Wa

http://dx.doi.org/10.1016/j.envpol.2015.05.0220269-7491/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

Polycyclic aromatic hydrocarbons (PAHs) can cause adverse effects on human health. The relative con-tributions of their two major intake routes (diet and inhalation) to population PAH exposure are stillunclear. We modeled the contributions of diet and inhalation to the overall PAH exposure of the pop-ulation of Beijing in China, and assessed their human incremental lifetime cancer risks (ILCR) using aMont Carlo simulation approach. The results showed that diet accounted for about 85% of low-molecular-weight PAH (L-PAH) exposure, while inhalation accounted for approximately 57% of high-molecular-weight PAH (H-PAH) exposure of the Beijing population. Meat and cereals were the main contributorsto dietary PAH exposure. Both gaseous- and particulate-phase PAHs contributed to L-PAH exposurethrough inhalation, whereas exposure to H-PAHs was mostly from the particulate-phase. To reduce thecancer incidence of the Beijing population, more attention should be given to inhaled particulate-phasePAHs with considerable carcinogenic potential.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Polycyclic aromatic hydrocarbons (PAHs) emitted from theincomplete combustion of fossil fuels or biomass have attractedwidespread public concern because of their adverse effects onhuman health, including carcinogenicity, teratogenicity, andmutagenicity (Bostr€om et al., 2002). PAHs are ingested into thehuman body mainly through diet and inhalation (ACGIH, 2005).The relative contributions of the two routes to the total level of PAHexposure in the general population are crucial for PAH exposureassessment, especially in China with its high PAH emissions (Zhanget al., 2007, 2009). Unfortunately, the studies conducted to date arelimited and have yielded somewhat inconsistent conclusions.Studies conducted in some East Asian regions have revealed thatdietary exposure to PAHs contributes more to the overall exposurelevel of the local population; e.g., for the sum of the 16 EPA priority-controlled PAHs (SPAH16) in Tianjin, China (Li et al., 2005), benzo[a]pyrene equivalent PAH (BAPeq) in Taiyuan, China (Xia et al., 2010,2013), and pyrene, benzo[b]fluoranthene, and benzo[a]pyrene in

ng).

Tokyo, Japan (Suzuki and Yoshinaga, 2007). However, in the UnitedStates, the primary routes of exposure to low-molecular-weightPAHs (L-PAHs, generally PAHs with �4 benzene rings), includingnaphthalene, fluorene and pyrene, were inhalation, whereas BAPexposure was predominantly from food intake (Shin et al., 2013).This suggests that the major PAH exposure route varies amongpopulations in different areas and according to PAH molecularweight. Previous studies have focused on total PAHs, BAPeq, or alimited range of PAHs to investigate the relative contributions ofvarious exposure routes to the overall exposure level of the generalpopulation. A critical knowledge gap remains with respect to thecontribution of the 16 individual U.S. Environmental ProtectionAgency (EPA) priority-controlled PAHs.

The bio-accessibilities of different PAHs greatly depend on theirphysicochemical properties. For example, fine particulate matterwith a relatively high content of high-molecular-weight PAHs (H-PAHs, generally PAHs with >4 benzene rings) can penetrate deepinto the lungs when inhaled, resulting in greater bio-accessibilitythan L-PAHs (Ohura et al., 2005). Therefore, it was proposed thatPAHs in the particulate phase might pose a greater adverse healtheffect on the human body than PAHs in the gaseous phase (Li et al.,2005; Zhang et al., 2009). Regarding dietary exposure, the differ-ence in PAH bio-accessibility among food types is becoming a

Y. Yu et al. / Environmental Pollution 205 (2015) 70e77 71

concern. It has been reported that the PAH concentrations varyamong food types and according to PAH molecular weight (Xiaet al., 2010; Yu et al., 2011). For these reasons, a systematic anddetailed survey of the relative contributions of diet and inhalationis necessary.

Beijing is the capital of China, with the highest population(about 21million in 2013) and vehicle volumes (about 5.4million in2013) in northern China. Both the annual energy consumption andPAH emission density in this area account for a great proportion ofthe national total (Zhang et al., 2007). The annual-average con-centrations of the sum of the 15 EPA priority-controlled PAHs(SPAH15), excluding naphthalene, in Beijing urban air is reported tobe about 206 ng m�3, with a BAP concentration of about 6 ng m�3,which is considerably higher than the national standard (1 ng/m3)(Li et al. 2014). Yu et al. (2011) reported that the median concen-trations of SPAH15 in human milk, placenta, and umbilical cordblood for the Beijing population were 278, 819, and 1372 ng g�1 offat, respectively, which were higher by almost an order of magni-tude than corresponding levels in Japan and the United States,which may be caused by the higher PAH concentrations of variousfood types in the local area (Yu et al., 2011). An epidemiologicalinvestigation showed that there were 40,307 new cases of malig-nant tumors in Beijing in 2012 (Beijing Municipal Government,2014), which was twice the level of 10 years ago. Among them,lung cancer was ranked first, followed by colorectal, liver, stomach,and prostate cancer for males. The ranking order for females wasthyroid carcinoma, followed by breast, lung, colorectal, and uterinecancer. Animal experiments have shown that the position of a tu-mor is associatedwith the route of PAH exposure (Hecht,1999; Latifet al., 2010). In the Beijing population, we found that L-PAHs couldmore readily penetrate the barrier between placenta and umbilicalblood than H-PAHs (Yu et al., 2011). Our recent study showed thatPAH concentrations in maternal serum had a strong associationwith the increased risks of fetal neural tube defects, but no re-lationships between human serum PAH concentrations and indoorair pollutionwere found (Wang et al., 2015). To evaluate the overallPAH exposure, the route from food consumptionmust be taken intoconsideration. Therefore, the relative contributions of diet andinhalation routes to the overall PAH exposure level of the Beijingpopulation are important, and information is urgently required forenvironmental scientists, policy makers, and local residents ofBeijing.

The aims of this studywere to investigate: 1) the exposure levelsof the Beijing population to 15 individual U.S. EPA priority-controlled PAHs through the diet and inhalation routes; 2) themain contributors to dietary and inhaled exposure to PAHs in thepopulation; and 3) the potential cancer risk for the Beijing popu-lation caused by PAH exposure.

2. Materials and methods

2.1. Target population and PAHs of concern

The target population was residents of Beijing, with agesranging from 1 to 72 years. The population was divided into twogender groups (male and female). Each group was further dividedinto four subgroups by age: children (1e6 years old), adolescents(7e18 years old), adults (19e65 years old), and seniors (66e72years old). The following 15 U. S. EPA priority-controlled PAHs wereselected: acenaphthylene (ACY), acenaphthene (ACE), fluorene(FLO), phenanthrene (PHE), anthracene (ANT), fluoranthene (FLA),pyrene (PYR), benzo[a]anthracene (BAA), chrysene (CHR), benzo[b]fluoranthene (BBF), benzo[k]fluoranthene (BKF), benzo[a]pyrene(BAP), indeno[1,2,3-cd]pyrene (IcdP), dibenzo[a,h]anthrancene(DahA), and benzo[g,h,i]perylene (BghiP). Naphthalene was

excluded because of its higher volatility and poor quantification.Seven of the PAHs were classed as L-PAHs (ACY, ACE, FLO, PHE, ANT,FLA, and PYR) and eight as H-PAHs (BAA, CHR, BBF, BKF, BAP, IcdP,DahA, and BghiP).

2.2. Dietary exposure estimates

The residue levels in seven food categoriesdfruits, vegetables,cereals, fish, meat, eggs, and milkdof all selected PAHs were re-ported by us previously (Yu et al., 2011). To the best of ourknowledge, the reported PAH concentrations of the seven foodcategories in Beijing are the only comprehensively published data.The amounts of the various food categories consumed and bodyweight of all subgroups were obtained from the Chinese nationalhealth and nutrition survey report (Zhai and Yang, 2006), and thedetails are listed in Table S1 in the Supplementary Information. Thedata for the concentration of PAHs in foodstuffs were tested todetermine if they followed a logarithmic normal distribution, whilefood consumption and body weight were found to approximatelyfollow a normal distribution. The dietary exposure level (ED, ngperson�1 day�1) of PAHs is the sum of exposures from intake of theseven food categories as follows:

ED ¼X

Ci � FIi (1)

where Ci and FIi are the PAH concentration (ng g�1) and intake rate(g day �1) of food category (i) according to age group in Beijing. Thebody-weight (BW) adjusted dietary exposure to PAHs (ngkg�1 day�1) was calculated by dividing ED by body weight (ED/BW).It should be noted that the differences in the bioaccessibility ofPAHs in various food categories were neglected in our studybecause of the limited data available.

2.3. Inhaled exposure estimates

The inhaled exposure level (EI, ng person�1 day�1) was calcu-lated as:

EI ¼ �CgPAH þ CpPAH

�� BR (2)

where CgPAH (ng m�3) and CpPAH (ng m�3) are the concentrations ofgaseous-phase (gPAHs) and particulate-phase (pPAHs) PAHs inBeijing, respectively, as reported in our previous study (Liu et al.,2007a, b) and their detailed description was provided in the Sup-plementary Information. BR (m3 day�1) is the breathing rate ofvarious age groups in Beijing; i.e., 9.3, 15, 16.5, 13 (m3 day�1) forchildren, adolescents, adults, and senior males, respectively, and8.6, 12, 11, 9.9 (m3 day�1) for the corresponding subgroups in thefemale group, respectively (Wang et al., 2010, 2009). The BR vari-ation was assumed to be 10%. The body weight-adjusted inhaledexposure dose of PAHs (ng kg�1 day�1) was calculated by dividingEI by body weight (EI/BW). It was assumed that there was no dif-ference in the bioaccessibility of the gaseous- and particulate-phase PAHs and the same equation was used to assess their PAHexposure levels, therefore, these results should be interpolatedwith care.

2.4. Carcinogenic risk assessment

The incremental lifetime cancer risk (ILCR) was used to expressthe carcinogenic risk caused by PAH exposure, as follows:

ILCR ¼ ðE � SF � ED � CF � EFÞ=ðBW � ATÞ (3)

where E (ng person�1 day�1) is the exposure level of BAPeq; SF is the

Table 1The estimated median dietary exposure levels of polycyclic aromatic hydrocarbons (PAHs) for the eight subgroups of Beijing population.

PAHsa Male Female

Boys Adolescents Adults Seniors Girls Adolescents Adults Seniors

Dietary exposure without body-weight adjustment (ng person¡1 day¡1)SL-PAHs 1.38 � 104 1.82 � 104 1.75 � 104 1.62 � 104 1.29 � 104 1.59 � 104 1.50 � 104 1.41 � 104

SH-PAHs 7.32 � 102 1.04 � 103 1.03 � 103 9.03 � 102 6.80 � 102 8.83 � 102 8.66 � 102 7.90 � 102

SPAH15 1.44 � 104 1.92 � 104 1.83 � 104 1.70 � 104 1.37 � 104 1.67 � 104 1.58 � 104 1.47 � 104

BAPeq 1.29 � 102 1.81 � 102 1.81 � 102 1.66 � 102 1.20 � 102 1.53 � 102 1.53 � 102 1.42 � 102

Dietary exposure with body-weight adjustment (ng kg¡1 day¡1)SL-PAHs 5.47 � 102 3.44 � 102 2.48 � 102 2.31 � 102 5.39 � 102 3.27 � 102 2.47 � 102 2.31 � 102

SH-PAHs 2.92 � 101 1.96 � 101 1.47 � 101 1.30 � 101 2.83 � 101 1.82 � 101 1.45 � 101 1.30 � 101

SPAH15 5.74 � 102 3.64 � 102 2.62 � 102 2.45 � 102 5.68 � 102 3.45 � 102 2.60 � 102 2.41 � 102

BAPeq 5.21 � 100 3.43 � 100 2.58 � 100 2.39 � 100 5.00 � 100 3.16 � 100 2.53 � 100 2.34 � 100

a SL-PAHs: sum of low molecular-weight PAHs; SH-PAHs: sum of high-molecular weight PAHs, SPAH15: sum of all PAHs, and BAPeq: benzo[a]pyrene equivalent PAH.

Y. Yu et al. / Environmental Pollution 205 (2015) 70e7772

cancer slope factor of BAP, with a geometric mean of 7.27(mg kg�1 d�1)�1 and a geometric standard deviation of 1.53 fordietary exposure (Xia et al., 2010), and 3.14 (mg kg�1 d�1)�1 and 1.8for inhalation exposure (Chen and Liao, 2006), respectively; EF (dayyear�1) is the exposure frequency (365), ED (year) is the exposureduration (children, 6; adolescents, 12; adults, 47; seniors, 7), CF (mgng�1) is a conversion factor (i.e. 10�6), and AT (days) is the lifespanof carcinogens (i.e., 25,550 days). The standard deviations of E andBW and the geometric standard deviation of SF were used in aMonte Carlo simulation. Other parameters were set to be constant.

2.5. Data analysis

SPSS 13.0 was used for the statistical analysis, and a significancelevel of 0.05 was applied. A Monte Carlo simulation (10,000 runs)was conducted to estimate the exposure of the Beijing populationto the selected PAHs and to assess their incremental lifetime cancerrisk, with consideration of the variations of all the factors includedin the calculation.

3. Results and discussion

3.1. Dietary exposure to PAHs

Table 1 summarizes the median EDs of the sum of L-PAHs (SL-PAHs), sum of H-PAHs (SH-PAHs), sum of all PAHs (SPAH15), andbenzo[a]pyrene equivalent PAHs (BAPeq), with and without body-weight adjustment, among the eight subgroups [i.e., boys (BOY),male adolescents (MAO), male adults (MAU), male seniors (MSE),girls (GIR), female adolescents (FAO), female adults (FAU), and fe-male seniors (FSE)] in the Beijing population. The detailed statis-tical results for the calculated daily EDs of the 15 individual PAHs,SL-PAHs, SH-PAHs, SPAH15, and BAPeq for the eight subgroups arepresented in Table S2-S3. The overall order of EDs before body-weight adjustment (adolescents > adults > seniors > children)was different from that after body-weight adjustment(children > adolescents > adults > seniors) for any individual PAH(Table S3). For example, the adolescents [1.92 � 104 (MAO) and1.67� 104 (FAO) ng person�1 day�1] and children [1.44� 104 (MSE)and 1.37� 104 (FSE) ng person�1 day�1] had the highest and lowestmedian EDs of the SPAH15 among the four age groups in the maleand female groups without body-weight adjustment, respectively,while children [5.74 � 102 (BOY) and 5.68 � 102 (GIR) ngkg�1 day�1] were ranked first after body-weight adjustment. Thereason for the highest EDs in Beijing being found in adolescents isthat this group consumed a larger amount of foodstuffs with higherPAH levels than did the other age groups. However, when body-weight adjustment was considered, children had the highest EDsfor all selected PAHs among the four age groups because they hadthe lowest body weight. This difference suggests that body fat has

an important role in diluting PAHs in the body. This is consistentwith our previous study of dichlorodiphenyltrichloroethanes(DDTs), in which the levels of DDTs in the body appeared to bediluted by body fat to some degree (Tao et al., 2008). Therefore,children should attract more public concern because they experi-ence the highest risk of dietary PAH exposure.

With or without body-weight adjustment, males had higher EDsto SPAH15, BAPeq, SL-PAHs, and SH-PAHs than females (Table 1).This is mainly because males consume more foodstuffs than fe-males, which is consistent with previous reports that PAH intake bymales was higher than that of females (Falc�o et al., 2003; Martí-Cidet al., 2008; Xia et al., 2010).

Daily EDs of SL-PAHs [percentile range from 10 to 90% (PR): 3.00� 104e8.15 � 104 ng person�1 day�1 or 1.73 �102e6.91� 102 ng kg�1 day�1] were significantly higher than thoseof SH-PAHs (PR: 3.22 � 102e2.80 � 103 ng person�1 day�1 or6.85� 100e6.18� 101 ng kg�1 day�1) with or without body-weightadjustment for the whole Beijing population. The highest ED of theindividual PAHs was found for PHE, with the PR being2.31 � 103e5.90 � 104 ng person�1 day�1 or4.98� 101e1.32� 103 ng kg�1 day�1, which was about three ordersof magnitude higher than the lowest, DahA, which had a PR of2.18 � 100e2.12 � 101 ng person�1 day�1 or 4.69 � 10�2 -4.73� 10�1 ng kg�1 day�1. The order of the EDs for the 15 individualPAHs was PHE > FLO > FLA > PYR > ACE > ACY > ANT > CHR >BAA > BBF > BKF > BAP > IcdP > BghiP > DahA. Detailed infor-mation for the individual PAHs is presented in Table S3.

The relative contributions of the seven food categories to theEDs of SPAH15, BAPeq, SL-PAHs, and SH-PAHs for the whole Beijingpopulation are shown in Fig. 1. A similar pattern in the contributionof the seven food categories to EDs for SPAH15 and SL-PAHs wasobserved, with the order being meat (34.9e43.4%) > cereals(21.2e26.5%) > fish (11.7e13.9%) > milk (8.3e16.9%) > eggs(6.2e7.8%) > vegetables (2.6e3.8%) > fruits (1.5e1.7%). This wasslightly different from the order for BAPeq [cereals(49.0e54.4%) > meat (16.1e21.1%) > fish (10.1e10.4%), vegetables(8.2e10.3%) >milk (3.3e7.3%), eggs (3.8e5.1%) > fruits (0.6e0.7%)],and SH-PAHs [meat (32.7e40.8%) > cereals(20.5e34.7%) > vegetables (8.7e11.4%), fish (8.2e9.1%) > eggs(6.1e8.1%) >milk (2.6e5.4%) > fruits (0.7e0.9%)]. This suggests thatcereals and vegetables contribute more to the EDs of H-PAHs thanL-PAHs. An important reason for the observed difference is that theresidue levels of L-PAHs and H-PAHs vary according to food type.Themedian concentration of SL-PAHs in the seven food types, fromhigh to low, was meat (34.4 ng g�1), fish (34.1 ng g�1), eggs (13.8 ngg�1), cereals (10.4 ng g�1), milk (7.80 ng g�1), fruits (6.77 ng g�1),and vegetables (3.31 ng g�1), while that of SH-PAHs was fish(2.76 ng g�1), meat (1.95 ng g�1), cereals (0.942 ng g�1), eggs (0.648ng g�1), vegetables (0.526 ng g�1), milk (0.117 ng g�1), and fruits(0.0735 ng g�1) (Yu et al., 2011).

Fig. 1. Relative contributions of the seven food categories to the dietary exposure levels of the sum of low-molecular weight PAHs (SL-PAHs) (A), sum of high-molecular weightPAHs (SH-PAHs) (B), sum of 15 PAHs (SPAH15) (C), and benzo[a]pyrene equivalent PAHs (BAPeq) (D) for Beijing population.

Y. Yu et al. / Environmental Pollution 205 (2015) 70e77 73

The intake dose of BAPeq of the Beijing population was three-tofive-fold lower than that of the Taiyuan population and two ordersof magnitude lower than that of the Tianjin population, because theconcentrations of BAPeq in various food categories from Beijingwere significantly lower than those from Taiyuan and Tianjin; thiswas particularly true for fruits, fish, meat, and milk (Li et al., 2005;Li, 2007; Xia et al., 2010) (Table S4). However, the dietary PAHexposure level of the Beijing population is higher than that re-ported from developed countries. For example, the dietary intakedose of the 16 EPA priority-controlled PAHs through fish con-sumption for the general Korean population was reported to be15.3 ng kg�1 d�1 (Moon et al., 2010), which is lower than the ED ofSPAH15 via fish consumption by residents of Beijing (ranging from22.2 to 52.8 ng kg�1 d�1) (Yu et al., 2011). In addition, the Beijingpopulation ingested more BAPeq than did the residents of six citiesin Korea (Yoon et al., 2007) and a study group in Catalonia, Spain(Martí-Cid et al., 2008). This is not only because the Beijing popu-lation consumed food types with a higher BAPeq concentration butalso because they ingest larger amounts of various foodstuffs thanthe population of Korea (Table S4).

3.2. Inhalation exposure to PAHs

Table 2 summarizes the median EIs of SL-PAHs, SH-PAHs,SPAH15, and BAPeq among the eight subgroups of the Beijing pop-ulation. The detailed statistical results of the daily EIs of PAHs (15individual PAHs, SL-PAHs, SH-PAHs, SPAH15, and BAPeq, with andwithout body-weight adjustment, for the eight subgroups are

provided in Table S5eS6. The overall orders of the age groups for EIswithout body-weight adjustment for males (adults >adolescents > seniors > boys) were different from females(adolescents > adults > seniors > girls). For the EIs with body-weight adjustment, the orders for the male and female groupswere identical; i.e., (children > adolescents > adults > seniors). Thisdecreasing trend of EIs with an increase in age was consistent withthat of dietary exposure (Table 1). Males had higher EIs of SPAH15,BAPeq, SL-PAHs, and SH-PAHs than females with or without body-weight adjustment, likely because males have a higher respirationrate than females (Wang et al., 2009, 2010).

Daily EIs of SL-PAHs (PR: 2.03 � 103e3.91 � 103 ngperson�1 day�1 or 3.87 � 101e8.84 � 101 ng kg�1 day�1) weresignificantly higher than those of SH-PAHs (PR:8.53 � 102e1.64 � 103 ng person�1 day�1 or 1.60 �101e3.68 � 101 ng kg�1 day�1) for the whole population. Thehighest EI of the PAHs was found for PHE, with the PR being5.32 � 102e1.02 � 103 ng person�1 day�1 or 1.01 �101e2.31 � 101 ng kg�1 day�1, which was about one order ofmagnitude higher than the lowest value, for ANT, which had a PR of4.27 � 101e8.19 � 101 ng person�1 day�1 or 8.16 � 10�1 -1.88� 100 ng kg�1 day�1. The order of EIs for the 15 individual PAHswas PHE > FLA > PYR > FLO > ACY > BBF >CHR > BAP > BKF > BAA > ACE > BghiP > IcdP > DahA > ANT, whichdiffered from that of the EDs. The EIs of the SL-PAHswere about 2.4-fold higher than those of the SH-PAHs, although the differenceswere smaller than those for the EDs. The detailed information forthe individual PAHs is presented in Table S6.

Table 2The estimated median inhaled exposure levels of polycyclic aromatic hydrocarbons (PAHs) for the eight subgroups of Beijing population.

PAHsa Male Female

Boys Adolescents Adults Seniors Girls Adolescents Adults Seniors

Inhaled exposure without body-weight adjustment (ng person¡1 day¡1)SL-PAHs 2.20 � 103 3.55 � 103 3.91 � 103 3.08 � 103 2.03 � 103 2.84 � 103 2.61 � 103 2.34 � 103

SH-PAHs 9.23 � 102 1.49 � 103 1.64 � 103 1.29 � 103 8.53 � 102 1.19 � 103 1.09 � 103 9.81 � 102

SPAH15 3.33 � 103 5.36 � 103 5.90 � 103 4.65 � 103 3.07 � 103 4.29 � 103 3.93 � 103 3.54 � 103

BAPeq 4.52 � 102 7.30 � 102 8.01 � 102 6.32 � 102 4.18 � 102 5.83 � 102 5.36 � 102 4.81 � 102

Inhaled exposure with body-weight adjustment (ng kg¡1 day¡1)SL-PAHs 8.84 � 101 6.74 � 101 5.59 � 101 4.38 � 101 8.44 � 101 5.88 � 101 4.30 � 101 3.87 � 101

SH-PAHs 3.68 � 101 2.80 � 101 2.32 � 101 1.83 � 101 3.53 � 101 2.46 � 101 1.78 � 101 1.60 � 101

SPAH15 1.34 � 102 1.02 � 102 8.34 � 101 6.62 � 101 1.27 � 102 8.87 � 101 6.40 � 101 5.75 � 101

BAPeq 1.81 � 101 1.39 � 101 1.14 � 101 9.09 � 100 1.73 � 101 1.21 � 101 8.86 � 100 7.93 � 100

a SL-PAHs: sum of low molecular-weight PAHs; SH-PAHs: sum of high-molecular weight PAHs, SPAH15: sum of all PAHs, and BAPeq: benzo[a]pyrene equivalent PAH.

Y. Yu et al. / Environmental Pollution 205 (2015) 70e7774

The relative contributions of gPAHs and pPAHs to the inhaledexposure levels of the male and female groups were identical. Forthe whole population, the inhaled dose of gPAHs accounted for75.5% (SL-PAHs), 0.9% (SH-PAHs), 49.9% (SPAH15), and 0.4% (BAPeq)of the total inhaled exposure, while pPAHs accounted for 24.6% (SL-PAHs), 99.1% (SH-PAHs), 50.1% (SPAH15), and 99.6% (BAPeq). Thissuggests that gPAHs were the major contributor to the inhaled L-PAHs but made a negligible contribution to inhaled H-PAHscompared to pPAHs. This is because L-PAHs were the predominantcompounds in the vapor phase, while H-PAHs were dominant inthe particulate phase (Liu et al., 2007a, b).

Out results are higher than the inhaled BAPeq exposure reportedin Tianjin, China [i.e., 322 (for children) and 519 (for adults) ngperson�1 day�1 (sampled in 2005) (Bai et al., 2009). Because theatmospheric concentrations of BAP and DahA, which have highertoxicity equivalence factors in Beijing (Liu et al., 2007a, b) weretwo-to three-fold higher than those in Taiyuan (Xia et al., 2013), theinhaled doses of BAPeq by the Beijing population ranged from two-to four-fold higher compared to the Taiyuan population. The dailyinhalation exposure levels of the population in Taiwan-Taichung toBAPeq sampled in 2002e2003 were 252 (children), 1590 (adoles-cents) and 1628 (adults) ng person�1 day�1, respectively (Chen andLiao, 2006), which were higher than those in Beijing [i.e., 452(children), 730 (adolescents), and 801 (adults) ng person�1 day�1].This is mainly because the atmospheric BAPeq level(59.4 ± 37.6 ng m�3) in Beijing (Liu et al., 2007a, b) was substan-tially lower than that in Taiwan-Taichung (60.3 mg m�3) (Fang et al.,2004a; Fang et al., 2004b; Tsai et al., 2004). Compared to the lowlevels of inhaled BAPeq (5.44 ng person�1 day�1) in Japan, which isdue to its lower atmospheric BAPeq concentration (0.360 ng m�3 inJapan in 2003) (Suzuki and Yoshinaga, 2007), the Beijing popula-tion experiences greater inhalation exposure to PAHs.

Table 3The estimated median total exposure levels of polycyclic aromatic hydrocarbons (PAHs)

PAHsa Male

Boys Adolescents Adults Seniors

Inhaled exposure without body-weight adjustment (ng person¡1

SL-PAHs 1.67 � 104 2.31 � 104 2.27 � 104 2.02 � 1SH-PAHs 1.91 � 103 2.93 � 103 3.06 � 103 2.55 � 1SPAH15 1.86 � 104 2.58 � 104 2.58 � 104 2.28 � 1BAPeq 6.31 � 102 9.90 � 102 1.06 � 103 8.68 � 1

Inhaled exposure with body-weight adjustment (ng kg¡1 day¡1)SL-PAHs 6.67 � 102 4.35 � 102 3.24 � 102 2.89 � 1SH-PAHs 7.63 � 101 5.51 � 101 4.35 � 101 3.63 � 1SPAH15 7.45 � 102 4.88 � 102 3.67 � 102 3.24 � 1BAPeq 2.53 � 101 1.87 � 101 1.51 � 101 1.24 � 1

a SL-PAHs: sum of low molecular-weight PAHs; SH-PAHs: sum of high-molecular we

3.3. Overall exposure to PAHs

Table 3 summarizes the median total exposure levels (ETs) ofSL-PAHs, SH-PAHs, SPAH15, and BAPeq among the eight subgroupsof the Beijing population. The detailed statistical results of the dailyETs (sum of EI and ED) of PAHs (15 individual PAHs, SL-PAHs, SH-PAHs, SPAH15, and BAPeq for the eight subgroups are listed inTables S7-S8. In males without body-weight adjustment, adoles-cents had a slightly higher ET of SL-PAHs and a lower ET of SH-PAHs than adults. However, their ETs of SPAH15 were almostidentical, and both were higher than the values of the other twogroups. In females without body-weight adjustment, the orders ofthe four age groups for SL-PAHs, SH-PAHs, SPAH15, and BAPeq wereconsistent (i.e., adolescents > adults > seniors > girls). After body-weight adjustment, the ET order for both the male and femalegroups was consistent with that for the EI and ED; i.e.,(children > adolescents > adults > seniors). Overall, male groupshad higher ETs than female groups for all individual PAHs(Table S8).

The daily ETs of SL-PAHs (PR: 1.57 � 104e2.31 � 104 ngperson�1 day�1 or 2.82 � 102e6.67 � 102 ng kg�1 day�1) weresignificantly higher than those of SH-PAHs (PR: 1.77 �103e3.06 � 103 ng person�1 day�1 or 3.37 �101e7.63 � 101 ng kg�1 day�1), with or without body-weightadjustment, for the whole population. PHE had the highest ETsamong all of the individual PAHs, with the PR being3.15 � 103e6.00 � 104 ng person�1 day�1 or6.69 � 101e1.34 � 103 ng/kg day�1, which was two orders ofmagnitude higher than the lowest ETs for DahA (PR:3.13 � 101e1.77 � 102 ng person�1 day�1 or 6.48 � 10�1 -3.88� 100 ng kg�1 day�1). The order of the ETs for the 15 individualPAHs was PHE > FLO > FLA > PYR > ACE > ACY > ANT > CHR > BBF

for the eight subgroups of Beijing population.

Female

Girls Adolescents Adults Seniors

day¡1)04 1.57 � 104 1.98 � 104 1.86 � 104 1.72 � 104

03 1.77 � 103 2.39 � 103 2.26 � 103 2.06 � 103

04 1.74 � 104 2.23 � 104 2.08 � 104 1.92 � 104

02 5.88 � 102 7.98 � 102 7.53 � 102 6.81 � 102

02 6.54 � 102 4.08 � 102 3.06 � 102 2.82 � 102

01 7.32 � 101 4.92 � 101 3.74 � 101 3.37 � 101

02 7.27 � 102 4.57 � 102 3.40 � 102 3.14 � 102

01 2.45 � 101 1.65 � 101 1.24 � 101 1.12 � 101

ight PAHs, SPAH15: sum of all PAHs, and BAPeq: benzo[a]pyrene equivalent PAH.

Y. Yu et al. / Environmental Pollution 205 (2015) 70e77 75

> BAA > BKF > BAP > BghiP > IcdP > DahA. This was almostidentical to the ranking order for the dietary intake of individualPAHs.

The relative contributions of EI and ED to the ET of the 15 indi-vidual PAHs among the four age groups were similar (Table S9). Inthe male and female subgroups, the contributions of EI and ED tothe ETof the 15 individual PAHs exhibitedminor differences (Fig. 2).This suggested that the intake of L-PAHs occurredmostly by dietaryexposure, whereas inhalation exposure was more important for H-PAHs. For the whole Beijing population, diet accounted for about84.7% of the SL-PAH intake and 42.6% of the SH-PAH intake, whileinhalation accounted for about 15.3 and 57.4%, respectively. Theseresults were similar to the 75.0% contribution of dietary exposure toSPAH16 for the Tianjin population (Li et al., 2005). Different patternsof contribution have been reported in developed countries. Morethan 90.0% of PYR, BBF, and BAP (as H-PAHs) were ingested in thediet of non-smoking university students in Japan (Suzuki andYoshinaga, 2007). In the U.S. population, L-PAHs (e.g., NAP, FLO,PHE, and PYR) were ingestedmainly by inhalation (more than 97%),whereas BAP (as an important H-PAH) was ingested in the diet(more than 95%) (Shin et al., 2013). It was reported that about 98%of PAH exposure was in the diet route in Spain (Linares et al., 2010).These differences cannot be easily explained by our data, andtherefore further in-depth studies are necessary.

3.4. Exposure risk assessment

The probability distribution patterns of the ILCR induced by theETs of BAPeq for the male and female groups of the Beijing

Fig. 2. Relative contributions of diet and inhalation to the total exposure levels of the 15 indin Beijing population.

Fig. 3. Cumulative probability of incremental lifetime cancer risk (ICLR) of total exposure dopanel) subgroups in Beijing population.

population were similar (Fig. 3). The ILCR of the ET for the wholeBeijing population ranged from 1.0 � 10�7 to 1.0 � 10�3, and mostof the risk was below the serious or priority risk level (10�4), asregulated by the U.S. EPA. The percentages of the population withan ICLR due to an ET above the serious acceptable risk level were0.5% (BOY), 1.3% (MAO), 14.0% (MAU), 0.0% (MSE), 0.3% (GIR), 0.8%(FAO), 10.5% (FAU), and 0.0% (FSE), respectively. Adults, especiallymales, had the highest cancer risk, whereas other age groups had alow risk (<2%). However, the percentages of all subgroups with anICLR due to an ET above the acceptable risk level (10�6) were >90%.This suggests that this population had a potential cancer risk higherthan the acceptable level. The order of the median ILCR for the fourage subgroups was adults (3.37 � 10�5) > adolescents(1.10 � 10�5) > children (7.97 � 10�6) > seniors (4.38 � 10�6). Thisorder is different from that of the ET (children >adolescents > adults > seniors). This is mainly because adults havethe highest exposure duration (47 years), followed by adolescents(12 years). The detailed statistical results for the ILCRs induced bythe ED and EI of BAPeq are presented in Table S10. The percentagesof the population with an ICLR from both ED and EI above theacceptable risk level ranged from ~80% in seniors to ~100% in adults.However, EI contributed more to the total ILCR than ED because themore highly toxic H-PAHs were ingested by inhalation (Fig. 3).

The cancer risk resulting from the ED of PAHs for the Beijingpopulation in this study was lower than that of the Taiyuan pop-ulation (20.5e39.4% above the serious cancer risk level) in China(Xia et al., 2010), while the two populations had a comparable riskfrom EI (Xia et al., 2013). Compared to the dietary ILCR (2.30� 10�5)as reported in Korea (Moon et al., 2010) and Egypt (1.31 �

ividual polycyclic aromatic hydrocarbons (PAHs) for the male (A) and female (B) groups

ses to benzo[a]pyrene equivalent PAHs (BAPequ) for male (left panel) and female (right

Y. Yu et al. / Environmental Pollution 205 (2015) 70e7776

10�5e8.92 � 10�5) (Khairy and Lohmann, 2013), the Beijing pop-ulation has a slightly lower cancer risk in terms of the ILCR (PR:1.81 � 10�6e1.78 � 10�5). The cancer risk of inhalation exposure toPAHs for the Beijing population was lower than that reported inTaiwan (Chen and Liao, 2006) and Egypt (Khairy and Lohmann,2013). However, the ILCR caused by PAHs for the Beijing popula-tion was considerably higher than that reported in the Spanishpopulation (10�9e10�7) (Linares et al., 2010).

This study had several limitations, which should be consideredwhen interpreting the findings. First, both the air and food sampleswere collected during 2005e2006. The concentrations of PAHswere likely to be different from previous studies because greateffort weremade by the Beijing government to improve the local airquality for the Olympic Games in 2008. Second, variances in thebioavailability of PAHs among food categories and inhaled air werenot considered in the modeling; therefore, the circulating PAHlevels were not considered. However, our study also had severalstrengths. First, the sampling periods of the food and air samplesfrom Beijing were matched, which enabled investigation of therelative contributions of the PAH exposures from the diet andinhalation. Second, selection bias was minimal because seven foodcategories (i.e., fruits, vegetables, cereals, fish, meat, eggs, andmilk)and inhaled gaseous and particulate phases were considered sys-tematically, unlike other studies. Third, the relative contributions ofthe 15 individual EPA priority-controlled PAHs from the diet andinhalation routes were evaluated. Therefore, our study may presentthe reference to evaluate the improvement of air pollution controlin the following years.

4. Conclusions

The order of the overall exposure of the Beijing population to 15individual PAHs was PHE > FLO > FLA > PYR > ACE > ACY >ANT > CHR > BBF > BAA > BKF > BAP > BghiP > IcdP > DahA. Withbody-weight adjustment, the orders of the PAH exposure levels bythe inhalation and dietary routes among the four age groups wereidentical (i.e., children > adolescents > adults > seniors). Males hadhigher exposure levels to all individual PAHs than females. Dietmainly accounted for the exposure to L-PAHs (about 84.7% for SL-PAHs), while exposure to H-PAHs was mainly from inhalation(about 57.4% for SH-PAHs). Meat and cereals were themain sourcesof dietary exposure to PAHs. Both gaseous- and particulate-phasePAHs contributed to the inhalation of L-PAHs, whereas inhaled H-PAH or BAPeq exposure was almost exclusively due to particulate-phase PAHs. The majority of the ILCRs from overall daily exposureto BAPeq for the four subgroups of the Beijing population werebelow the serious risk level. However, almost 90% were higher thanthe acceptable level, which suggested that the potential cancer riskof the Beijing population should attract wide public concern.

Conflicts of interest

The authors declare they have no actual or potential competingfinancial interests.

Role of the funding sources

The funding agencies have no role in study design, imple-mentation, data analysis, and interpretation.

Acknowledgment

This researchwas supported by grants from the National NaturalScience Foundation of China (No. 41371466, 41401583, 41390240).

Appendix A. Supplementary data

Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.envpol.2015.05.022.

References

ACGIH, 2005. Polycyclic Aromatic Hydrocarbons (PAHs) Biologic Exposure Indices(BEI). American Conference of Governmental Industrial Hygienists, Cincinnati,OH.

Bai, Z.P., Hu, Y.D., Yu, H., Wu, N., You, Y., 2009. Quantitative health risk assessment ofinhalation exposure to polycyclic aromatic hydrocarbons on citizens in Tianjin,China. Bull. Environ. Contam. Toxicol. 83, 151e154.

Beijing Municipal Government, 2014. Report on Hygiene and Health of BeijingPopulation in 2013. People's Health Press, Beijing.

Bostr€om, C.E., Gerde, P., Hanberg, A., Jernstr€om, B., Johansson, C., Kyrklund, T.,Rannug, A., T€ornqvist, M., Victorin, K., Westerholm, R., 2002. Cancer riskassessment, indicators, and guidelines for polycyclic aromatic hydrocarbons inthe ambient air. Environ. Health Perspect. 110 (Suppl. 3), 451e488.

Chen, S.C., Liao, C.M., 2006. Health risk assessment on human exposed to envi-ronmental polycyclic aromatic hydrocarbons pollution sources. Sci. Total En-viron. 366, 112e123.

Falc�o, G., Domingo, J.L., Llobet, J.M., Teixid�o, A., Casas, C., Müller, L., 2003. Polycyclicaromatic hydrocarbons in foods: human exposure through the diet in Catalonia,Spain. J. Food Prot. 66, 2325e2331.

Fang, G.C., Wu, Y.S., Chen, M.H., Ho, T.T., Huang, S.H., Rau, J.Y., 2004a. Polycyclicaromatic hydrocarbons study in Taichung, Taiwan, during 2002-2003. Atmos.Environ. 38, 3385e3391.

Fang, G.C., Chang, K.F., Lu, C., Bai, H., 2004b. Estimation of PAHs dry deposition and B[a]P toxic equivalency factors (TEFs) study at urban, industry park and ruralsampling sites in central Taiwan, Taichung. Chemosphere 55, 787e796.

Hecht, S.S., 1999. Tobacco smoke carcinogens and lung cancer. J. Natl. Cancer Inst.91, 1194e1210.

Khairy, M.A., Lohmann, R., 2013. Source apportionment and risk assessment ofpolycyclic aromatic hydrocarbons in the atmospheric environment of Alexan-dria, Egypt. Chemosphere 91, 895e903.

Latif, I.K., Karim, A.J., Zuki, A.B., Zamri-Saad, M., Niu, J.P., Noordin, M.M., 2010.Pulmonary modulation of benzo[a]pyrene-induced hemato- and hepatotoxicityin broilers. Poult. Sci. 89, 1379e1388.

Li, X.R., Li, B.G., Tao, S., Guo, M., Cao, J., Wang, X.J., Liu, W.X., Xu, F.L., Wu, Y.N., 2005.Population exposure to PAHs in Tianjin area. Acta Sci. Circumst. 989e993 (inChinese).

Li, X.R., 2007. Spatial Distribution Pattern of Emission, Dispersion and Exposure ofPolycyclic Aromatic Hydrocarbons in Tianjin, China Beijing. Peking University,China (in Chinese).

Li, W., Wang, C., Wang, H.Q.J., Chen, J.W., Shen, H.Z., Shen, G.F., Huang, Y., Wang, R.,Wang, B., Zhang, Y.Y., Chen, H., Chen, Y.C., Su, S., Lin, N., Tang, J.H., Li, Q.B.,Wang, X.L., Liu, J.F., Tao, S., 2014. Atmospheric polycyclic aromatic hydrocarbonsin rural and urban areas of northern China. Environ. Pollut. 192, 83e90.

Linares, V., Perell�o, G., Nadal, M., G�omez-Catal�an, J., Llobet, J.M., Domingo, J.L., 2010.Environmental versus dietary exposure to pops and metals: a probabilisticassessment of human health risks. J. Environ. Monit. 12, 681e688.

Liu, S.Z., Tao, S., Liu, W.X., Liu, Y.N., Dou, H., Zhao, J.Y., Wang, L.G., Wang, J.F.,Tian, Z.F., Gao, Y., 2007a. Atmospheric polycyclic aromatic hydrocarbons inNorth China: a winter-time study. Environ. Sci. Technol. 41, 8256e8261.

Liu, Y.N., Tao, S., Yang, Y.F., Dou, H., Yang, Y., Coveney, R.M., 2007b. Inhalationexposure of traffic police officers to polycyclic aromatic hydrocarbons (PAHs)during the winter in Beijing, China. Sci. Total Environ. 383, 98e105.

Martí-Cid, R., Llobet, J.M., Castell, V., Domingo, J.L., 2008. Evolution of the dietaryexposure to polycyclic aromatic hydrocarbons in Catalonia, Spain. Food Chem.Toxicol. 46, 3163e3171.

Moon, H.B., Kim, H.S., Choi, M., Choi, H.G., 2010. Intake and potential health risk ofpolycyclic aromatic hydrocarbons associated with seafood consumption inKorea from 2005 to 2007. Archives Environ. Contam. Toxicol. 58, 214e221.

Ohura, T., Noda, T., Amagai, T., Fusaya, M., 2005. Prediction of personal exposure toPM2.5 and carcinogenic polycyclic aromatic hydrocarbons by their concentra-tions in residential microenvironments. Environ. Sci. Technol. 39, 5592e5599.

Shin, H., McKone, T.E., Bennett, D.H., 2013. Evaluating environmental modeling andsampling data with biomarker data to identify sources and routes of exposure.Atmos. Environ. 69, 148e155.

Suzuki, K., Yoshinaga, J., 2007. Inhalation and dietary exposure to polycyclic aro-matic hydrocarbons and urinary 1-hydroxypyrene in non-smoking universitystudents. Int. archives Occup. Environ. Health 81, 115e121.

Tao, S., Yu, Y.X., Liu, W.X., Wang, X.L., Cao, J., Li, B.G., Lu, X.X., Wong, M.H., 2008.Validation of dietary intake of dichlorodiphenyltrichloroethane and metabolitesin two populations from Beijing and Shenyang, China based on the residuals inhuman milk. Environ. Sci. Technol. 42, 7709e7714.

Tsai, P.J., Shih, T.S., Chen, H.L., Lee, W.J., Lai, C.H., Liou, S.H., 2004. Assessing andpredicting the exposures polycyclic aromatic hydrocarbons (PAHs) and theircarcinogenic potencies from vehicle engine exhausts to highway toll stationworkers. Atmos. Environ. 38, 333e343.

Wang, B., Jin, L., Ren, A.G., Yuan, Y., Liu, J.F., Li, Z.W., Zhang, L., Yi, D.Q., Wang, L.L.,Zhang, Y.L., Wang, X.L., Tao, S., Finnell, R.H., 2015. Levels of polycyclic aromatic

Y. Yu et al. / Environmental Pollution 205 (2015) 70e77 77

hydrocarbons in maternal serum and risk of neural tube defects in offspring.Environ. Sci. Technol. 49, 588e596.

Wang, B.B., Duan, X.L., Jiang, Q.J., Huang, N., Qian, Y., Wang, Z.S., Zhang, J.L., 2010.Inhalation exposure factors of residents in a typical region in northern China.Res. Environ. Sci. 23, 1421e1427 (in Chinese).

Wang, Z.S., Wu, T., Duan, X.L., Wang, S., Zhang, W.J., Wu, X.F., Yu, Y.J., 2009. Researchon inhalation rate exposure factors of Chinese residents in environmentalhealth risk assessment. Res. Environ. Sci. 22, 1171e1175 (in Chinese).

Xia, Z.H., Duan, X.L., Qiu, W.X., Liu, D., Wang, B., Tao, S., Jiang, Q., Lu, B., Song, Y.,Hu, X., 2010. Health risk assessment on dietary exposure to polycyclic aromatichydrocarbons (PAHs) in Taiyuan, China. Sci. Total Environ. 408, 5331e5337.

Xia, Z., Duan, X., Tao, S., Qiu, W., Liu, D., Wang, Y., Wei, S.Y., Wang, B., Jiang, Q., Lu, B.,Song, Y., Hu, X., 2013. Pollution level, inhalation exposure and lung cancer riskof ambient atmospheric polycyclic aromatic hydrocarbons (PAHs) in Taiyuan,China. Environ. Pollut. 173, 150e156.

Yoon, E., Park, K., Lee, H., Yang, J.H., Lee, C., 2007. Estimation of excess cancer risk ontime-weighted lifetime average daily intake of PAHs from food ingestion. Hum.Ecol. Risk Assess. 13, 669e680.

Yu, Y.X., Wang, X.L., Wang, B., Tao, S., Liu, W.X., Wang, X., Cao, J., Li, B.G., Lu, X.X.,Wong, M.H., 2011. Polycyclic aromatic hydrocarbon residues in human milk,placenta, and umbilical cord blood in Beijing, China. Environ. Sci. Technol. 45,10235e10242.

Zhai, F.Y., Yang, X.G., 2006. Report on Nutrient and Health of Chinese Population in2002 (Part II: Diet Pattern and Nutrient Intake). People's Health Press, Beijing(in Chinese).

Zhang, Y.X., Tao, S., Cao, J., Coveney Jr., R.M., 2007. Emission of polycyclic aromatichydrocarbons in China by county. Environ. Sci. Technol. 41, 683e687.

Zhang, Y.X., Tao, S., Shen, H.Z., Ma, J.M., 2009. Inhalation exposure to ambientpolycyclic aromatic hydrocarbons and lung cancer risk of Chinese population.Proc. Natl. Acad. Sci. U. S. A. 106, 21063e21067.