Jump to main content.


Executive Summary Human Health Risk Assessment: Upper Hudson River

(August 1999)

This document presents the baseline Human Health Risk Assessment for the Upper Hudson River (HHRA), which is part of Phase 2 of the Reassessment Remedial Investigation/ Feasibility Study (Reassessment RI/FS) for the Hudson River PCBs site in New York.(1) This HHRA quantitatively evaluates both cancer risks and non-cancer health hazards from exposure to polychlorinated biphenyls (PCBs) in the Upper Hudson River, which extends from Hudson Falls, New York to the Federal Dam at Troy, New York. The HHRA evaluates both current and future risks to children, adolescents, and adults in the absence of any remedial action and institutional controls. The HHRA uses current U.S. Environmental Protection Agency (USEPA) policy and guidance as well as additional site data and analyses to update USEPA's 1991 risk assessment.

USEPA uses risk assessment as a tool to evaluate the likelihood and degree of chemical exposure and the possible adverse health effects associated with such exposure. The basic steps of the Superfund human health risk assessment process are the following: 1) Data Collection and Analysis to determine the nature and extent of chemical contamination in environmental media, such as sediment, water, and fish; 2) Exposure Assessment, which is an identification of possible exposed populations and an estimation of human chemical intake through exposure routes such as ingestion, inhalation, or skin contact; 3) Toxicity Assessment, which is an evaluation of chemical toxicity including cancer and non-cancer health effects from exposure to chemicals; and 4) Risk Characterization, which describes the likelihood and degree of chemical exposure at a site and the possible adverse health effects associated with such exposure.

The HHRA shows that cancer risks and non-cancer health hazards to the reasonably maximally exposed (RME) individual associated with ingestion of PCBs in fish from the Upper Hudson River are above levels of concern. Consistent with USEPA regulations, the risk managers in the Superfund program evaluate the risk and hazards to the RME individual in the decision-making process. The HHRA indicates that fish ingestion represents the primary pathway for PCB exposure and for potential adverse health effects, and that risks from other exposure pathways are generally below levels of concern. The results of the HHRA will help establish acceptable exposure levels for use in developing remedial alternatives for PCB-contaminated sediments in the Upper Hudson River, which is Phase 3 (Feasibility Study) of the Reassessment RI/FS.

Data Collection and Analysis

USEPA previously released reports on the nature and extent of contamination in the Upper Hudson River as part of the Reassessment RI/FS (e.g., February 1997 Data Evaluation and Interpretation Report, July 1998 Low Resolution Sediment Coring Report, August 1998 Database for the Hudson River PCBs Reassessment RI/FS [Release 4.1], and May 1999 Baseline Modeling Report). The Reassessment RI/FS documents provide current and forecasted concentrations of PCBs in fish, sediments, and river water and form the basis of the site data collection and analyses used in conducting the HHRA.

Exposure Assessment

Adults, adolescents, and children were identified as populations possibly exposed to PCBs in the Upper Hudson River due to fishing and recreational activities (swimming, wading), as well as from living adjacent to the Upper Hudson River and inhaling volatilized PCBs in the air. Cancer risks and non-cancer hazards were calculated for each of these populations. To protect human health and provide a full characterization of the PCB risks and hazards, both an average (central tendency) exposure estimate and an RME estimate were calculated. The RME is the maximum exposure that is reasonably expected to occur in the Upper Hudson River under baseline conditions.

The exposure pathways identified in the HHRA are ingestion of fish, incidental ingestion of sediments, dermal contact with sediments and river water, and inhalation of volatilized PCBs in air. For these exposure pathways, central tendency and RME estimates were calculated using point estimate analyses, whereby an individual point estimate was selected for each exposure factor used in the calculations of cancer risks and non-cancer health hazards. Incidental ingestion of river water while swimming was not evaluated because the river water meets drinking water standards for PCBs.

In addition to the point estimate analysis, a Monte Carlo analysis was performed to provide a range of estimates of the cancer risks and non-cancer health hazards associated with the fish ingestion pathway. The Monte Carlo analysis helps evaluate variability in exposure parameters (e.g., differences within a population's fish ingestion rates, number of years an angler is exposed, body weight) and uncertainty (i.e.,a lack of complete knowledge about specific variables).

Ingestion of Fish

For fish ingestion, both central tendency and RME estimates were developed for each of the parameters needed to calculate the cancer risks and non-cancer health hazards. Based on the 1991 New York Angler survey of fish consumption by licensed anglers (Connelly et al., 1992), the central tendency fish ingestion rate was determined to be approximately six half-pound meals per year and the RME fish ingestion rate was determined to be 51 half-pound meals per year.

For the point estimate analyses, cancer risks and non-cancer health hazards to an adult angler were calculated. Population mobility data from the U.S. Census Bureau for the five counties surrounding the Upper Hudson River and fishing duration data from the 1991 New York Angler survey were used to determine the length of time an angler fishes in the Upper Hudson River (i.e., exposure duration). The exposure duration for fish ingestion was 12 years for the central tendency exposure estimate and 40 years for cancer (7 years for non-cancer) for the RME estimate. Standard USEPA default factors were used for angler body weight. Future concentrations of PCBs in fish were derived from forecasts presented in the Baseline Modeling Report, which were then grouped by fish species and averaged over species for the entire Upper Hudson River. PCB losses during cooking were assumed to be 20% for the central tendency exposure estimate and 0% (no loss) for the RME estimate, based on studies reported in the scientific literature.

In the Monte Carlo analyses, each exposure parameter (e.g., ingestion rate, exposure duration, body weight) was represented by a range of values, each with an assigned probability, rather than as a single point estimate. Cancer risks and non-cancer hazards were calculated for anglers beginning at age 10. Differences in the length of time an angler fishes the Upper Hudson (exposure duration) were obtained from the 1991 New York Angler survey and the U.S. Census Bureau data. Differences in angler body weight through time were obtained from national health surveys summarized in the scientific literature. Future concentrations of PCBs in fish were derived from the Baseline Modeling Report. Fish species consumption variability was evaluated based on consumption patterns determined from the 1991 New York Angler survey and within-species PCB concentrations were averaged over location within the Upper Hudson River. The variability in fish ingestion rates was examined by considering surveys of fish ingestion rates in states other than New York. Variability in PCB cooking loss was determined from a review of the scientific literature.

Due to the lack of sufficient information available to define quantitative uncertainty distributions for several important exposure factors, such as exposure duration, an explicit two-dimensional Monte Carlo analysis which examines variability and uncertainty separately could not be performed. Instead, an expanded one-dimensional (1-D) analysis was completed using a sensitivity/uncertainty analysis. Each 1-D Monte Carlo simulation examined variability of PCB intake and was repeated for a range of possible input distributions for important exposure variables. A total of 72 separate combinations of the variable input parameters were examined in the 1-D analysis. Each 1-D simulation consisted of 10,000 simulated anglers, such that the entire 1-D Monte Carlo analysis consisted of 720,000 simulations.

Other Exposure Pathways

For the direct exposure scenarios for river water and sediment, the central tendency exposure estimates for adults and young children (aged 1-6) were assumed to be one day every other week for the 13 weeks of summer (7 days/year) and for the RME were assumed to be one day per week for the 13 weeks of summer (13 days/year). Adolescents (aged 7-18) were assumed to have about three times more frequent exposure, with a central tendency exposure estimate of 20 days/year and an RME estimate of 39 days/year. The risks due to possible inhalation of PCBs in air were evaluated for both recreational users of the river (swimmers and waders) as well as for residents living adjacent to the Upper Hudson River. The concentrations of PCBs in water and sediment were derived from the Baseline Modeling Report. The concentrations of PCBs in air were calculated from a combination of historical monitoring data and modeled emissions from the river using a USEPA-recommended air dispersion model. Standard USEPA default factors were used for certain exposure parameters (e.g., body weight) in the risk calculations for these pathways.

Toxicity Assessment

The toxicity assessment is an evaluation of the chronic (7 years or more) adverse health effects from exposure to PCBs (USEPA, 1989b). In Superfund, two types of adverse health effects are evaluated: 1) the incremental risk of developing cancer due to exposure to chemicals and 2) the hazards associated with non-cancer health effects, such as reproductive impairment, developmental disorders, disruption of specific organ functions, and learning problems. The cancer risk is expressed as a probability and is based on the cancer potency of the chemical, known as a cancer slope factor, or CSF. The non-cancer hazard is expressed as the ratio of the chemical intake (dose) to a Reference Dose, or RfD. The chronic RfD represents an estimate (with uncertainty spanning perhaps an order of magnitude or greater) of a daily exposure level for the human population, including sensitive populations (e.g., children), that is likely to be without an appreciable risk of deleterious effects during a lifetime. Chemical exposures exceeding the RfD do not predict specific diseases. USEPA's Integrated Risk Information System, known as IRIS, provides the primary database of chemical-specific toxicity information used in Superfund risk assessments. The most current CSFs and RfDs for PCBs were used in calculating cancer risks and non-cancer hazards in the HHRA.

PCBs are a group of synthetic organic chemicals consisting of 209 individual chlorinated biphenyls called congeners. Some PCB congeners are considered to be structurally similar to dioxin and are called dioxin-like PCBs. USEPA has classified PCBs as a probable human carcinogen, based on a number of studies in laboratory animals showing liver tumors. Human carcinogenicity data for PCB mixtures are limited. USEPA (1996) described three published studies that analyzed deaths from cancer in PCB capacitor manufacturing plants (Bertazzi et al., 1987; Brown, 1987; Sinks et al., 1992). Recently, Kimbrough et al. (1999) published the results of an epidemiological study of mortality in workers from two General Electric Company capacitor manufacturing plants in New York State. Due to the limitations of the Kimbrough et al. (1999) study identified by USEPA in its review (e.g., more than 75% of the workers never worked with PCBs, the median exposure for those who worked with PCBs was only a few years, and the level of PCB exposure could not be confirmed), USEPA expects that the study will not lead to any change in its CSFs for PCBs, which were last reassessed in 1996.

Risk Characterization

Point Estimate Calculations

Ingestion of fish contaminated with PCBs resulted in the highest lifetime cancer risks. The RME estimate of the increased risk of an individual developing cancer averaged over a lifetime based on the exposure assumptions is 1 10-3, or one additional case of cancer in 1,000 exposed people. The RME risks associated with the dioxin-like PCBs are comparable. The central tendency (average) estimate of risk is 3 10-5, or 3 additional cases of cancer in 100,000 exposed people. For known or suspected carcinogens, acceptable exposure levels for Superfund are generally concentration levels that represent an incremental upper bound lifetime cancer risk to an RME individual of between 10-4 and 10-6. The central tendency cancer risks and non-cancer hazards are provided to more fully describe the health effects associated with average exposure. Estimated cancer risks relating to PCB exposure in sediment and water while swimming or wading, or from inhalation of volatilized PCBs in air by residents living near the river, are much lower than those for fish ingestion, falling generally at the low end, or below, the range of 10-4 to 10-6. A summary of the point estimate cancer risk calculations is presented below.
 

Point Estimate Cancer Risk Summary
Pathway Central Tendency Risk  RME Risk
Ingestion of Fish 3 10-5 (3 in 100,000) 1 10-3 (1 in 1,000)
Exposure to Sediment* 4 10-7 (4 in 10,000,000) 1 10-5 (1 in 100,000)
Exposure to Water* 1 10-8 (1 in 100,000,000) 2 10-7 (2 in 10,000,000)
Inhalation of Air* 2 10-8 (2 in 100,000,000) 1 10-6 (1 in 1,000,000)
*Total risk for child (aged 1-6), adolescent (aged 7-18), and adult (over 18).

The evaluation of non-cancer health effects involved comparing the average daily exposure levels (dose) to determine whether the estimated exposures exceed the Reference Dose (RfD). The ratio of the site-specific calculated dose to the RfD for each exposure pathway is summed to calculate the Hazard Index (HI) for the exposed individual. An HI of one (1) is the reference level established by USEPA above which concerns about non-cancer health effects must be evaluated. 

Ingestion of fish resulted in the highest Hazard Indices, with an HI of 10 for the central tendency point estimate and an HI of 116 for the RME point estimate. The total HIs for exposure to sediment, water, and air are all below one. Non-cancer hazards due to inhalation of PCBs were not calculated because IRIS does not contain a toxicity value for inhalation of PCBs. A summary of the point estimate non-cancer hazards is presented below.
 

Point Estimate Non-Cancer Hazard Summary
Pathway Central Tendency Non-Cancer Hazard Index RME Non-Cancer  

Hazard Index

Ingestion of Fish 10 116
Exposure to Sediment* 0.05 0.2
Exposure to Water* 0.007 0.02
Inhalation of Air  Not Calculated Not Calculated
*Values for child and adolescent, which are higher than adult for these pathways.

 

Monte Carlo Estimation

In the Monte Carlo analysis, a distribution of cancer risks and non-cancer health hazards was calculated for the fish ingestion pathway. The tables below summarize the low-end (5th percentile), midpoint (50th percentile), and high-end (> 90th percentile) cancer risks and non-cancer hazards. At a given percentile, the risks or hazards are higher than that presented in the table for 100 minus the given percentile. For example, as shown for the base case in the table below, the calculated incremental cancer risk at the 95th percentile is 9 x 10-4, which means that the cancer risks for only the top 5th percentile are greater than that value.
 

Monte Carlo Cancer Risk Summary - Fish Ingestion
Risk Percentile Low Estimate Base Case High Estimate
5th Percentile 7 10-7  5 10-6  5 10-5 
50th Percentile 1 10-5  6 10-5  4 10-4 
90th Percentile 7 10-5  5 10-4  2 10-3 
95th Percentile 1 10-4  9 10-4  3 10-3 
99th Percentile 3 10-4  4 10-3  1 10-2 

 
 

Monte Carlo Non-Cancer Hazard Summary - Fish Ingestion
Risk Percentile Low Estimate Base Case High Estimate
5th Percentile 0.1 1 7
50th Percentile 2 11 51
90th Percentile 5 31 117
95th Percentile 11 82 233
99th Percentile 19 136 366

Comparison of Point Estimate and Monte Carlo Analyses

The Monte Carlo base case scenario is the one from which point estimate exposure factors for fish ingestion were drawn, thus the point estimate RMEs and the Monte Carlo base case estimates are comparable. Similarly, the point estimate central tendency (average) and the Monte Carlo base case midpoint (50th percentile) are comparable. For cancer risk, the point estimate RME for fish ingestion (1 10-3) falls approximately at the 95th percentile from the Monte Carlo base case analysis. The point estimate central tendency value (3 10-5) and the Monte Carlo base case 50th percentile value (6 10-5) are similar. For non-cancer hazards, the point estimate RME for fish ingestion (116) falls between the 95th and 99th percentiles of the Monte Carlo base case. The point estimate central tendency HI (10) is approximately equal to the 50th percentile of the Monte Carlo base case HI of 11.

Major Findings of the HHRA

The HHRA evaluated both cancer risks and non-cancer health hazards to children, adolescents and adults posed by PCBs in the Upper Hudson River. USEPA has classified PCBs as probable human carcinogens and known animal carcinogens. Other long-term adverse health effects of PCBs observed in laboratory animals include a reduced ability to fight infections, low birth weights, and learning problems. The major findings of the report are:

 
 

1. A separate human health risk assessment is being conducted for the Mid-Hudson River (Federal Dam at Troy, New York to Poughkeepsie, New York).

For information about this page, contact: kluesner.dave@epa.gov

 

LocalNavigation


Jump to main content.