Health risks from multiroute exposure of potentially toxic elements in a coastal community: a probabilistic risk approach in Pangkep Regency, Indonesia

Abstract Populations in coastal areas are vulnerable to chronic exposure to potentially toxic elements (PTEs) from aquatic ecosystems. In this study, we estimated the health risks from Cd, Cr, Fe, and Pb exposure in river water, sediments, and edible aquatic biota (Anadara granosa and Chanos chanos). Human daily intake was determined by an interview process with 287 respondents that were selected by proportional cluster random sampling. The cumulative probability value for total risk exceeded the acceptable limit value, indicating that adverse health impacts from PTE exposure may occur in the Pangkep coastal population. The sensitivity analysis showed that exposure duration and Cr concentrations are the factors with the greatest influence in a recreational exposure scenario, whereas the frequency of exposure to shellfish and milkfish is the most influential factor for health risks in a residential scenario. Thus, reducing Cr concentrations in river water, restricting swimming duration, and limiting the frequency of shellfish and milkfish consumption are required to mitigate health risks in the future.


Introduction
The presence of potentially toxic elements (PTEs) in aquatic ecosystems is harmful not only to aquatic organisms but also to terrestrial organisms, including humans. This problem remains an issue of major concern worldwide because of the toxicity of such elements and their ability to accumulate in biological tissues. Thus, the presence of PTEs poses a threat to food chains. Several PTEs such as As, Hg, Pb, and Cd can Pangkep communities and are important sources of protein. Thus, the coastal community experiences greater potential exposure to PTEs through multiple environmental pathways, such as water and sediment while swimming, and/or the consumption of aquatic animals such as shellfish and milkfish taken from river water.
Human health risk assessments (HHRA) have been widely used by several researchers to assess potential adverse health effects that may occur in the future because of chronic exposure to trace elements from the aquatic environment (Yi et al. 2017;Ezugwu et al. 2019;Gao et al. 2019;Mgbenu and Egbueri 2019;Ukah et al. 2019;Egbueri 2020;Egbueri and Mgbenu 2020;Onyemesili et al. 2020;Astuti et al. 2021a;Enyigwe et al. 2021;Karmakar et al. 2021;Rauf et al. 2021b). Those previous studies revealed that many sources of PTE contamination exist in aquatic settings, such as mining; agricultural, domestic, and industrial activities; and geogenic sources. However, very little research has evaluated the health risks to coastal communities related to PTE contamination in the study area investigated here. Previous HHRA studies only assessed exposure to one particular metal (such as As or Cr) (Aripai et al. 2012;Bahar and Daud 2012). However, PTEs have the potential to cause health problems in the form of combined elements through exposure to multiple mediums and pathways. The complex health risks due to combinations of PTEs in coastal communities have not yet been assessed. Moreover, a prior study only estimated health risks using reference/standard values from the U.S. Environmental Protection Agency (USEPA) and other works of literature. They did not calculate risks on the basis of the real dietary intakes and exposure patterns of the local communities. Failure to address these factors can cause bias in health risk assessments. This is the first study that has calculated risks from a combination of PTEs and used the actual exposure patterns of the Pangkep community.
High uncertainties may be involved in conducting risk estimations. The uncertainty may originate from variation in environmental conditions, insufficient knowledge of the actual value of certain variables, and characteristics of the population (Sanaei et al. 2021). Previous studies of human health risks have not dealt with the uncertainty that may be present. Moreover, several studies use the standard value of human exposure patterns from USEPA to calculate risks, which may not be appropriate for the inhabitants of the study area. This can cause policy making designed to reduce health risks from environmental PTE exposure to be incompatible with the existing conditions in the region. Thus, a comprehensive estimation incorporating uncertainties associated with human health risk assessments is needed to more convincingly determine the possibility of further impacts from PTE exposure on human health.
Monte Carlo simulation (MCS) is a mathematical and logical model that has been widely used to assess uncertainty in health risk assessment models by collecting limited data to comprehensively predict various situations (Xiao et al. 2018;Sanaei et al. 2021). MCS can be applied to calculate uncertainties and conduct sensitivity analyses of random variables in human health risk assessments (Qu et al. 2012). This model was used by USEPA as the basis to consider health risk policy. Thus, to adequately consider the uncertainty that may be present, we used MCS to estimate health risks. The aims of this study are (1) to evaluate potential noncarcinogenic and carcinogenic risks due to PTE exposure through multiple exposure pathways and (2) to analyze uncertainties and the sensitivity of risk factors in the health risk assessment. Assessment of PTEs in coastal settings is of great importance. The results of this study may provide valuable information for aquatic ecosystem protection, assurance of food safety, and preservation of human health in the future.

Study area descriptions
The Pangkajene dan kepulauan (Pangkep) regency is a suburban area located in the north of Makassar city, South Sulawesi, Indonesia within latitudes of 04 40 0 -08 00 0 S and longitudes of 110 -119 48 0 67 00 E (BPS 2021). Figure 1 presents a map of sampling sites. The Pangkep regency is approximately 51 km from the provincial capital of South Sulawesi, with a total area of 12,362.73 km 2 (11,464.44 km 2 of sea and 898.29 km 2 of land/terrestrial area). This area lies within the Maros Pangkep karst national park, which is an area of magnificent karst features in Indonesia with spectacular karst tower outcrops. Geologically, the late Eocene-to middle Miocene-age Tonasa limestone formation is the main geologic formation in this area, which comprises solid limestone, coral, wackestones, calcarenite, calcirudite, and bioclastic (Wilson 1996;Wilson and Bosence 1997;Ahmad and Hamzah 2016;Nuhung 2016) deposits. The uppermost lithologies in the formation comprise packstones, wackestones, and grainstones. The carbonate outcrops in the Tonasa formation were formed by large benthic foraminifera and sometimes by coralline algae (Wilson 1996; Wilson and Bosence 1997). The carbonate platform in Sulawesi is part of the Australian margin that experienced rifting beginning in the Jurassic period and is one of the main hydrocarbon reservoirs in Southeast Asia (Wilson 1996;Polv e et al. 1997;Wilson and Bosence 1997). Carbonate platform outcrops in the Pangkajene area are characterized by vertical tower karst features in coastal areas (Wilson and Bosence 1997). A report by UNESCO (2019) stated that the Maros-Pangkep geopark is underlain by the Bantimala Melange tectonic complex that comprises low-to-ultrahighgrade metamorphic rocks, tectonic blocks of oceanic crust, and marine sedimentary rocks. Previous studies showed there are environmental degradations in the Maros Pangkep karst area which may harmful to the ecological system and human health due to cement production and limestone mining activities (WALHI 2018;Mallongi et al. 2019;Astuti et al. 2021a;Rauf et al. 2021aRauf et al. , 2021cMallongi et al. 2022;Rauf et al. 2022).
Hydrologically, this karst area has a high-capacity aquifer (100-500 L/s) (Nuhung 2016) that is dominated by subsurface drainage. Rainwater that falls on the limestone surface can pass through Dolina and Ponor and/or immediately infiltrate into rock cavities (Nuhung 2016). The aquifers and subsurface drainages in this area act as the catchment area for many rivers, including the Pangkajene, Pute, and Bantimurung rivers. The mainland is inhabited by a population of 345,775 (in 2020) dominated by people of Bugis-Makassar ethnicity who work in the livestock, agriculture, fishery, industry, mining, and tourism sectors (BPS 2021). The climate is characterized as a tropical monsoon climate : in 2020, the highest air temperatures occurred in October at 28.3 C and the lowest air temperature occurred in March at 26.9 C (BPS 2021). The amount of precipitation and rainy days in the Pangkep regency range from 8 to 966 mm/months and 3 to 27 days, respectively, in 2020 (BPS 2021). The soil in the Pangkep regency is highly fertile and is used to produce various crops and agricultural commodities such as rice, chili, shallots, garlic, spinach, watermelon, pomelo, mango, coconut, and medicinal plants (BPS 2021). In the fishery sector, milkfish is one of the main commodities in the Pangkep regency. Marble and limestone are the main mining resources in this area, although several minerals and other minable materials exist in the Pangkep regency, such as kaolin, basalt, coal, propylite, marble, clay/loam, limestone, feldspar, alluvial gold, gems, gravel, and radioactive minerals (Depdikbud 1976;DPU 2017;Astuti et al. 2021bAstuti et al. , 2021cMallongi et al. 2022).

Sampling methods
This study was conducted in the Pangkajene dan Kepulauan (Pangkep) regency in South Sulawesi Province, Indonesia. The sampling was done in November 2020 and June 2021. A total of 18 samples of river water and 18 samples of river sediment, three samples of shellfish (Anadara granosa), and four samples of milkfish (Chanos chanos) were collected. The shellfish were taken from three sampling sites on the coast and they represent the biota most frequently consumed by residents. Additionally, we collected edible organisms (shellfish and milkfish) from coastal areas. Shellfish and milkfish make up the main foodstuffs in the Pangkep regency because they are cheap and abundantly available in that area. More broadly, milkfish is a major commodity in the Pangkep regency. Milkfish samples were taken from the two largest aquaculture fields in a downstream area of the Pangkajene river. River water and river sediment were collected twice, once in November 2020 (wet season) and once in June 2021 (dry season). Shellfish and milkfish were collected only once, during the month of harvest in June 2021.
The water samples were placed in high-density polyethylene (HDPE) bottles (1L). To sterilize equipment and prevent contamination, pre-cleaned bottle was prepared. The riverine water was collected using a water sampler from 10 to 30 cm below the water surface, whereas sediments were taken by Ekman grab from the sediment surface (10-30 cm depth). Approximately 500 g of sediment, 350 g of milkfish, and 500 g of shellfish samples were stored in clean polyethylene plastic, after which samples were stored in a cooler box with ice and directly transported to a laboratory.

Sediment analysis
All sediment samples were separated from rocks, grass, and other litter, and then oven-dried at 40 C. All dried samples were crushed into powder and sieved through a 0.5-mm-mesh sieve. The samples were then ground by a blender into a fine powder and sieved using a 0.5-mm-mesh sieve. Afterward, 0.5 g of sediment was added to a solution of 5 mL 65% HNO 3 and 1 mL 60% HClO 4 . This method complies with the established acid digestion method outlined by APHA (1999). Adding the HNO 3 and HClO 4 solution is necessary to digest silicate minerals or difficult-to-oxidize organic matter (OM) (APHA 1999). This method is also used as the established heavy-metal analysis protocol in Indonesia (Departemen Pertanian RI 2009). The digested samples were heated on a hot plate at 200 C until they formed a clear ±3 mL solution. One milliliter of the clear solution was added to 25 mL distilled water, filtered, and homogenized.

Water sample analysis
A 5 mL water sample was added to 5 mL HNO 3 and 1 mL HClO 4 . The acid digestion method was performed according to the APHA (1999) protocol and Indonesia's standard for heavy metal analysis by Departemen Pertanian RI (2009). The solution was then heated on a hot plate at 200 C until 1-3 mL of clear solution was formed. Then, 25 mL of distilled water was added to 1 mL of the clear solution.

Anadara granosa and Chanos chanos analysis
The flesh of Anadara granosa and Chanos chanos was ground using a blender, and 5 g of the ground flesh was added to a solution of 10 mL of 65% HNO 3 and 2 mL 60% HClO 4 and left overnight. Acid digestion was conducted according to the APHA (1999) protocol and Indonesia's standard for heavy metal analysis by Departemen Pertanian RI (2009). The digested solutions were stirred and heated at 200 C until a clear solution formed. After digestion, 1 mL of the clear solution was added to 25 mL of distilled water, filtered, and homogenized. Cd, Cr, Fe, and Pb concentrations were conducted using an AAS Thermo scientific iCE 3000 series instrument with wavelength absorption at 248.3 (Fe), 283.3 (Pb), 228.8 (Cd), and 357.9 (Cr).

Quality assurance
For accurate analysis, blanks and sample duplicates were included in the PTE analytical process. To sterilize equipment and prevent contamination, the glassware was properly precleaned by soaking in 6% HCl for at least 24 h and then rinsing three times with double-distilled water and oven-drying at a temperature of 105 C. Analytical blanks were analyzed similarly as for sample analysis but without samples. The percent recovery achieved for each PTE was >90% with linear calibration curves (R 2 of 0.999).

Estimation of human health risk
Two exposure scenarios were considered in this study: (1) a recreational scenario, which results from incidental ingestion and dermal contact when swimming in the Pangkajene river, and (2) a residential scenario that arises from the consumption of shellfish and milkfish (Supporting Information Figure S1). When people are in recreational activities such as swimming or fishing, there is a potential for pollutant exposure through dermal contact with water and sediment (USEPA 2021a). They are also potentially exposed by PTE through incidental ingestion of water or sediment (USEPA 2015(USEPA , 2021aCanpolat et al. 2020). Due to the Pangkep population does not use the river water as the main drinking water resource, thus, PTEs exposures from river water and sediment were only included in the recreational exposure scenarios. Whereas, the health risks from shellfish and milkfish consumption were included in the residential exposure scenario.
The hazard quotient (HQ) and carcinogenic risk (CR) values were used in the present study to estimate the noncarcinogenic (Cd, Pb, Cr, and Fe) and carcinogenic (Pb, Cd, and Cr) risks, respectively, due to PTE exposure. Noncarcinogenic and carcinogenic health risks in the recreational exposure scenario were calculated using Eqs.
Carcinogenic risks in the recreational exposure scenario: In dermal exposure, body surface area was calculated using follows equation Eq. (15) (Reading and Freeman 2005): where SA is the body surface area (m 2 ), BW is the body weight (Kg), H represents body height (m). Then, SA is multiplied by 10,000 to convert in cm 2 . In residential exposure scenario, noncarcinogenic and carcinogenic health risk were estimated using Eqs. (16)-(21) (USEPA 1989(USEPA , 2015(USEPA , 2021b(USEPA , 2021d: Noncarcinogenic risks in the residential exposure scenario: Carcinogenic risks in the residential exposure scenario: TCR ¼ CR mðmilkfishÞ þ CR mðShellfishÞ (21) All abbreviations, units, and values related to the risk assessment (THQ, HI, CR, CR m , TCR) equations were expressed in Table 1.

Human data collection
Human data were obtained by interviewing 287 individuals, including adults and children, to obtain daily exposure pattern data used in the risk assessment metrics (ingestion rate (IR), body weight (BW), exposure frequency (EF), exposure duration (ED), surface area (SA), averaging time (AT), exposure time (ET), and exposure event (EV)). Data for this study were collected using proportional cluster random sampling from two village populations (Mappasaile and Tekolabbua villages).

Uncertainty and sensitivity analysis
The uncertainty and sensitivity analysis was conducted using Crystal Ball 11.1.2 (Oracle Inc Software, USA) in a Microsoft Excel (2019 version) environment. After estimating the HQ and CR, the MCS technique was run with 10,000 iterations to calculate THI and TCR, whereas considering the distribution of independent health risk variables in the study area (IR, BW, EF, ED, AT, ET, and EV) that were directly obtained from the interview process. A final human health risk result for each metal   (RAIS 2020) was produced by comparing the lower bound uncertainty (the 5th percentile) and the upper bound uncertainty (the 95th percentile) with the permissible value (<1 for HQ and <1 Â 10 À06 for CR). In this study, the estimated values for HQ, CR, THI, and TCR are presented for the 5% and 95% percentiles. A sensitivity analysis was used to determine the most influential factor in the health risk estimations (Qu et al. 2018;Tajdar-Oranj et al. 2018). A high correlation coefficient indicates a greater influence on the final health risk calculation. Using USEPA human health risk assessment methods with MCS provides a probability distribution of HQ and CR values to estimate the uncertainty related to human health risks due to PTE exposure from various pathways (Qu et al. 2018).

Statistical analysis
IBM SPSS version 24 (SPSS Inc, USA) software was performed for statistical analysis. This study presented a parametric test result (at a significant level >.05). The statistical analysis was used to assess descriptive analysis (mean, standard deviation, minimum and maximum value). Then, the uncertainty and sensitivity tests were analyzed by Crystal Ball software (version 11.1.2).

Results and discussion
PTEs concentrations in water, sediment, and aquatic biotas Tables 2 and 3 provide the concentration of selected PTEs (Cd, Cr, Fe, and Pb) in water, sediments, shellfish, and milkfish. The highest mean concentration in river water was found for Cr (1.64 mg/L), followed by Fe (1.31 mg/L), Pb (0.79 mg/L), and Cd (0.42 mg/L). According to Table 2, the mean concentration of PTEs in river water exceeded the water quality standards set by the WHO, USEPA, and the Indonesian standard, indicating that the water was highly polluted by PTEs and unacceptable as drinking water. Compared with world average concentrations, the PTE concentrations in the Pangkajene river were very high. Compared with prior studies by Aripai et al. (2012) and Salamahu (2012), the Pb and Cd concentrations in river water were higher than in previous studies, indicating that actions taken to control PTEs have not been fully successful in recent years. Cr and Pb concentrations in the Pangkajene river were lower than a river in India studied by Pandion et al. (2022); however, the Cd concentration in the Pangkajene river was higher than in the Indian river. High concentrations of PTEs may be influenced by a surplus amount of rainwater to the river system during wet season (Pandion et al. 2022). Aside from the factor of excess rainwater, anthropogenic factors such as untreated sewage disposal and direct industrial discharge to the river, as well as natural activities like underwater currents and sediment movement, may elevate PTE concentrations in the water column (Najmeddin et al. 2018;Pandion et al. 2022). Many anthropogenic activities take place in the area surrounding the Pangkajene river, such as limestone and marble mining; cement production; and domestic, agricultural, and fishery activities ). Additionally, wastewater treatment remains rarely practiced in the Pangkep regency. These conditions may influence PTE accumulation in the Pangkajene river. The mean concentration of Cd, Cr, Fe, and Pb in sediments of the Pangkajene river followed the order: Fe > Cr > Pb > Cd. The Cd concentration was lower than the detection limit in the sediment column, indicating that Cd mainly occurs in the aqueous phase in the river ecosystem and is readily available to aquatic organisms. This finding is inconsistent with a study by Maddusa and Asrifuddin (2008), which reported higher Cd concentrations in sediment than in the water. This condition may be influenced by the chemical properties of Cd, which is known as a labile PTE that migrates rapidly within the environment (Jiang et al. 2013). Several studies have demonstrated that Cd is easily transported into water is highly bioavailable to aquatic organisms (Jiang et al. 2013;Liu et al. 2019). Cd, Cr, and Pb concentrations in sediment were lower than the probable effect level (PEL) value (Table 2). When PTE concentrations in sediments are lower than the PEL, adverse effects on benthic organisms will be infrequent (Kuang et al. 2021).
Fe in sediment was higher than the PEL value, indicating that harmful effects may be expected to occur frequently in sediment-dwelling benthic organisms. High concentrations of PTEs in the sediment may result from the deposition of suspended materials containing PTEs and downstream water movement (Yi et al. 2011). The concentration of Pb in this study was higher than that in prior studies (Salamahu 2012;Fitriani et al. 2013Fitriani et al. , 2014Ali 2017), and the Fe concentration in this study was also higher than that reported in an earlier study by Amina (2012). A study by Paena et al. (2015) reported that sediment in the Pangkajene river contained a high concentration of OM and phosphate, which might influence the accumulation of PTEs in the river ecosystem. Compared with results reported by Yenny et al. (2020), Pb and Cr concentrations in the Pangkajene river sediment were higher than those in the Citarum river. As the Citarum river is among the most polluted rivers in Indonesia, these findings demonstrate an alarming state of contamination in the Pangkajene river. Table 3 lists the mean PTE concentrations in shellfish (Anadara granosa) and milkfish (Chanos chanos) from the Pangkajene river. Based on the mean concentrations, the order of PTEs in shellfish samples was Fe > Cr > Cd > Pb, and the order in milkfish samples was Fe > Cr > Pb > Cd. The Cd concentration in shellfish in this study was higher than that reported in prior studies by Maddusa and Asrifuddin (2008) and Haspullah et al. (2018). Pb and Cr concentrations in shellfish in the present study are also higher than those in shellfish studied by Haspullah et al. (2018). This result reveals that Cd, Cr, and Pb concentrations in Anadara granosa in the years of this study are higher than in the prior studies, indicating that PTE pollution in the Pangkajene river is becoming increasingly severe in recent years.
In comparison with other studies, Fe and Pb concentrations in shellfish from the Pangkajene river were higher than in shellfish from the Java sea and Madura straits (Yona et al. 2020(Yona et al. , 2021. Cd and Cr concentrations in shellfish studied here were higher than those in a study by Soegianto et al. (2020), whereas Cd and Pb concentrations in milkfish were lower than those in previous studies (Syarif 2015;Zulfiah et al. 2017;Haeriah 2019). A study by Awaluddin et al. (2020) showed that pond water and sediments in the Pangkep regency had high Pb concentrations of 0.0182 mg/L and 25.210 mg/kg, respectively. A high concentration of Pb in pond water and sediment may influence the PTE concentration in milkfish (Awaluddin et al. 2020).
The concentrations of PTEs in shellfish and milkfish shown in Table 3 indicate that all PTE concentrations in shellfish and milkfish were lower than the food quality criteria values set by EFSA, except for Fe. Aquatic organisms such as fish regulate their intake of trace elements from water via intake of suspended foods and ion transfer through lipophilic membranes (Yi et al. 2011), whereas bottomfeeder organisms like Anadara granosa take up nutrients and contaminants from sediments (Mrozi nska and Bąkowska 2020; Soegianto et al. 2020). Hence, the high concentration of PTEs in sediments and water contribute to bioaccumulation in aquatic organisms. Table 4 reports the exposure pattern of Pangkajene communities, with noncarcinogenic indicator values shown in Table 5. All HQ values were below 1 in both exposure scenarios, except for the HQ ingestion of Cr in the residential exposure scenario.

Point estimate of human health risk
This indicates that chronic exposure to Cr through milkfish and shellfish consumption can potentially induce adverse effects and noncarcinogenic risks to humans in the future (HQ > 1). This result is inconsistent with other studies (Bandara and Pathiratne 2019;Mwakalapa et al. 2019;Pandion et al. 2022) that found HQ values lower than 1. Based on the HI value in the recreational scenario, PTE (Cd, Cr, Fe, and Pb) exposure from both ingestion and dermal routes was safe from noncarcinogenic health effects. This result is consistent with the findings of several other studies (Alidadi et al. 2019;Canpolat et al. 2020;Shakeri et al. 2020). Based on the study by Paustenbach et al. (2003), dermal exposure poses an insignificant health risk due to the minor amount of Cr that can penetrate deep skin layers. Another study stated that speciation of PTEs plays a role in health risk estimations because it influences the bioavailability, water-soluble properties, and the mobile fraction of PTEs in the human body (Sah et al. 2017). For example, Cr(VI) is more toxic than Cr(III), because it can be absorbed into deep skin layers (Paustenbach et al. 2003;Alidadi et al. 2019). Thus, dermal contact may make a small contribution to total health risks. Additionally, a study by Tong et al. (2021) demonstrated that ingestion is the primary exposure route of PTEs in an aquatic system because the values of HQ ingestion of PTEs were higher than those of HQ dermal for both children and adults.
The carcinogenic risk indicators are presented in Table 6. In the recreational exposure scenario, the concentration of Cd in river water significantly contributed to carcinogenic risk (CR > 1 Â 10 À04 ). Like the noncarcinogenic risk, the contribution of river sediment exposure to carcinogenic risk is small. This result is consistent with the findings of several other studies (Wojciechowska et al. 2019;Yang et al. 2019;Ustao glu et al. 2020). Based on the CR m value, exposure to river water through incidental ingestion and dermal contact is not safe (CR m > 1 Â 10 À04 ). This finding is contrary to that of a previous study by Canpolat et al. (2020), which stated that there is no carcinogenic risk of PTE exposure through recreation activities. Conversely, exposure to river sediment is still deemed safe (CR m < 1 Â 10 À06 ). The CR, CR m , and TCR values tend to be higher in adults than in children. As demonstrated in a prior study (Tong et al. 2021), this trend illustrates that adults are more susceptible to carcinogenic effects of PTE exposure from water systems.
In the residential exposure scenario, the order of CR shellfish decreased as Cd > Cr > Pb, whereas the order of CR milkfish decreased as Cr > Pb > Cd. Based on the CR m value, the consumption of shellfish and milkfish is not safe because it is potentially associated with carcinogenic effects in humans (CR m > 1 Â 10 À04 ). Thus, action to reduce PTE concentrations in the river ecosystem is required immediately. The TCR values in the recreational and residential scenarios indicate that adults have a higher carcinogenic risk than children. These findings further support the idea that cancer vulnerability increases in midlife and at older ages (White et al. 2014). Cancer development is a complex process; nevertheless, the life course approach can be used to investigate the influence of prenatal and early life exposure on cancer in adults (White et al. 2014;Bhatt 2019). Age can reflect the ED and accumulation of CR factors; additionally, a multifactor process of CR may influence changes in normal cell processes that lead to cancer, including DNA damage accumulation and mutation over time.

Probabilistic risk assessment
Using a probabilistic risk approach based on MCS, the mean, 5th, and 95th percentile values of the THI and TCR distributions for different population groups are shown in Figure 2. In the recreational exposure scenario, the mean THI values estimated by the probabilistic approach for adults and children were 1.81 and 1.75, respectively (Figure 2(A)), whereas in the residential scenario, the mean THI values calculated using MCS were 3.05 (adult) and 3.97 (children). Based on those values, both children and adults were at risk of noncarcinogenic health problems related to PTE exposure (THI > 1).
The THI values for adults and children in either exposure scenario were greater than 1 at the 95th percentile, indicating that 95% of the different age group populations experienced noncarcinogenic risk from multiple exposure pathways and that noncarcinogenic risk may not have occurred in 5% of the population (Sanaei et al. 2021). For noncarcinogenic risk, the sensitivity test results (Figure 3) revealed that the exposure frequency to milkfish (EF-MF) is the main factor influencing noncarcinogenic risk for adults in the residential scenario, with a contribution of 31.00%. Exposure frequency to shellfish (EF-SF) was the main factor in total noncarcinogenic risk in children, with a contribution of 30.90%. In the recreational exposure scenario, ED as the main factor related to noncarcinogenic risk for the adult population, with a contribution of 35.10%, and the Cr concentration in river water (Cr-RW) contributed most to noncarcinogenic risk in child populations, with a contribution amounting to 22.4%. Figure 2(B) presents the cumulative probability value for estimated TCR in recreational and residential scenarios. In the recreational exposure scenario, the mean values for adults and children using a probability risk approach are 1.12 Â 10 À03 and 3.98 Â 10 À04 , respectively. In the residential exposure scenario, the mean values for adults and children are 5.41 Â 10 À03 and 3.08 Â 10 À03 , respectively. Those values showed that both in the recreational and residential scenarios TCR values were higher than 1 Â 10 À04 , indicating that a potential CR exists from multiple PTE exposure pathways in the Pangkep coastal communities. Figure 4 depicts the TCR value sensitivity analysis results. Exposure frequency to shellfish is the most influential factor for both adult and child populations in the residential scenario, with contributions of 67.7% and 45.6%, respectively. In the recreational scenario, ED is the most important factor for adult and child populations, with contributions of 74.2% and 33.5%, respectively. This result is in line with a study by Gao et al. (2019).

Risk management
The 95th percentile value represents the highest exposure that is reasonably predicted to occur at the site and is known as the reasonable maximum exposure (RME) (USEPA 2001). An RME value that exceeds the target level risk or the allowable risk means that implementation of risk reduction measures is needed. Based on the equation by Ditjen (2012), a recommended value for restricting shellfish consumption (EF-SF: 3 days/year for adults and EF-SF: 7 days/year for children) should be put in place to avoid carcinogenic risk. Aside from limiting shellfish consumption, several  studies recommend implementing a well-balanced diet to avoid chronic diseases such as cardiometabolic disease and death (Zhai et al. 2015;Micha et al. 2017). A recent review by Zhai et al. (2015) stated that the consumption of vegetables, fruits, and other edible plants containing vitamins and essential metals can reduce the toxicity of metals in the human body. Vitamins C and E are antioxidants that have the benefit of scavenging free radicals and decreasing lipid peroxidation. For the recreational scenario, limiting ED may protect the population from carcinogenic risk with recommended ED values of 13 years (adults) and 9 years (children).
To protect populations from noncarcinogenic risks, reducing milkfish consumption frequency (EF-MF) to 170 and 140 days/year for adults and children, respectively, must be implemented. Additionally, reducing the Cr concentration in river water needs to be implemented immediately. A study by Yoshinaga et al. (2018) recommends using a magnesium (Mg)-and iron (Fe)-based hydrotalcite-like compound as a low-cost and highly effective depurative method to reduce Cr pollution in the river. Additionally, strong regulatory actions, PTE concentration monitoring in river ecosystems, as well as the use of wastewater treatment plants, should be performed by the local authority.

Conclusions
The mean value of PTE concentrations in river water was found to be above the water quality standard set by the WHO, indicating that the river water is highly polluted and unacceptable for daily consumption purposes. Except for Fe, the mean PTE concentrations in sediments were lower than the PEL values, indicating that negative effects on sediment-dwelling benthic organisms are not expected to occur frequently. The mean PTE values in the aquatic organisms (Anadara granosa and Chanos chanos) were below the limit for food quality standards, except for Fe concentrations in shellfish. Based on THI and TCR values, the coastal area communities of the Pangkep regency are at high risk of noncarcinogenic and carcinogenic effects from multiple PTE exposure pathways. This conclusion is also supported by the cumulative risk value determined using the Monte Carlo simulation, which demonstrates mean values of THI and TCR value above the allowable values (THI > 1 and TCR > 1 Â 10 À04 ). The sensitivity test results revealed that ED and the concentration of Cr in river water (Cr-RW) are the most influential factors in noncarcinogenic risk in the recreational scenario, whereas the consumption frequency of milkfish and shellfish were the greatest noncarcinogenic and carcinogenic risk factors in the residential scenario. Thus, we recommend reducing Cr concentrations in the river ecosystem, restricting swimming durations, and limiting the consumption frequency of shellfish and milkfish to overcome future health risks.