This study proposed an integrated modelling framework and a modified method for evaluating non-carcinogenic health risks from nonylphenol (NP)-contaminated food consumption. First, a fugacity-based multimedia model and a food web bioaccumulation model were adopted to predict the distribution of NP in the Can Giuoc river and the bioaccumulative concentrations in biota. Next, local people's exposure to NP was quantified using the accumulative concentrations and the data of fishery products intake from a questionnaire survey distributed among 203 local people. Then, human health risk was evaluated in terms of fishery products intake and intake frequency which were each derived from the same survey. The study revealed that human health risk would exist, although the obtained bioaccumulation factors for the consumed organisms were lower than the bioaccumulation criteria. Consuming 141 g or more per serving of riverine food products resulted in an average NP intake exceeding 0.005 mg/kg of body weight per day among 45-73% of the local adults, of whom pregnant women or young and potential mothers accounted for 10-21%. Seventy-nine percent was the highest rate of the population to be at risk under medium river flow rate when food-intake amount and intake frequency were taken into account. Ingesting 70 g per serving of more contaminated species, such as whiteleg shrimp and small fish, less frequently could lead to less risk exposure than ingesting 267 g per serving of less contaminated species, such as sand goby and climbing pearch, more frequently. By coupling food intake with intake frequency, the modified method enables the studying of human health risk from NP-contaminated food consumption to be conducted with more care, and so benefits risk communication at local level.
Keywords: Bioaccumulation factor (BAF); Biomagnification factor (BMF); Endocrine disruptor contamination; Food web bioaccumulation model; Fugacity-based multimedia model; Human health risk assessment; Tropical estuary ecosystem.