**1. Introduction**

Humans are exposed on a daily basis to cadmium (Cd), a toxic heavy metal, mainly through the consumption of food containing Cd. Cd absorbed via food accumulates in the kidneys and may cause renal tubular dysfunction, called Cd nephropathy, in the inhabitants of Cd-polluted areas [1]. The most severe case of Cd toxicity is itai-itai disease, which is characterized by osteomalacia and renal anemia. It develops among patients with Cd nephropathy [2]. The heaviest Cd-polluted area in Japan was along the Jinzu River basin of Toyama prefecture, at which agricultural fields were contaminated by a large amount of Cd derived from an upstream mine. A large number of inhabitants in this area developed Cd nephropathy and 200 patients with itai-itai disease were officially recognized by 2020.

Although the Cd-polluted area in Toyama was completely restored in 2011, large but scattered Cd-polluted areas still remain in Akita prefecture, which is located in the northern part of Japan, due to the previous activities of mines and smelters [3]. Pollution levels are particularly high in the northern area of Akita prefecture, in which we performed local health examinations on female farmers as part of the Japanese multi-centered environmental toxicant study (JMETS) [4,5]. JMETS aimed for risk assessment of Cd by targeting females, who are generally more vulnerable to Cd toxicity than males. Health examinations were sequentially performed in two adjacent areas: the area along Yoneshiro River in Odate city in 2001–2002 (area A) [4] and that upstream of the river in Kazuno city and Kosaka town in 2003–2004 (area B) [5] (Figure 1). Cd pollution levels were higher in area B than in area A because area B was directly a ffected by two large mines and their a ffiliated smelters, while area A was secondarily contaminated through irrigation from the river that runs from area B. These studies revealed that many farmers were exposed to high levels of Cd through the consumption of self-harvested rice contaminated by Cd, some of whom had Cd nephropathy [4–6].

**Figure 1.** A map of area A (Odate city) and area B (Kazuno city and Kosaka town) in Akita, and their location in Japan (cited from CraftMAP). The thick solid line, dotted lines, and thin solid lines indicate the prefectural boundary, boundaries of municipalities, and courses of the river, respectively.

The provisional tolerable monthly intake (PTMI) of Cd reported by the Joint FAO (Food and Agriculture Organization)/WHO (World Health Organization) expert committee on food additives (JECFA) is now 25 μg/kg/month and was amended from the provisional tolerable weekly intake (PTWI) of 7 μg/kg/week in 2010 [7]. In Japan, the tolerable weekly intake (TWI) of Cd of 7 μg/kg/week remains in e ffect and was set by the Food Safety Commission of Japan (FSCJ) in 2008. In our previous study on area A in 2001, we performed dietary exposure assessments on Cd in subjects undergoing health examinations and obtained weekly Cd intake levels for comparisons with the PTWI of Cd at that time [4]. The findings revealed that 33–51% of subjects had Cd intake levels in excess of the PTWI. However, the exposure assessment was performed using a simplified method based on Cd concentrations in rice and miso, fermented soybean paste, and the consumption of these two foods to estimate individual Cd intake according to average Japanese Cd intake levels. Therefore, a more detailed exposure assessment of Cd is needed in this area to clarify the actual status of dietary Cd exposure in its inhabitants. A similar exposure assessment of Cd in area B, in which Cd pollution was heavier than in area A, needs to be conducted.

In the present study, we collected 100 types of local food items in 2 Cd-polluted areas in Akita, measured their Cd concentrations, and assessed Cd exposure levels in the subjects of previous local health examinations using an individual food analysis method based on Cd concentrations in individual food items and the intake amounts of these food items by subjects obtained from diet surveys conducted at health examinations. We also performed the Monte Carlo simulation [8]

to evaluate the probabilistic distribution of the Cd intake levels of these subjects in order to confirm the assessment. We demonstrated that Cd intake levels in these subjects were approximately PTWI or TWI, which showed that subjects exposed to Cd were at risk of adverse effects.

#### **2. Materials and Methods**

#### *2.1. Sampling and Handling of Food Items*

We selected approximately 80 local food items from the table of food groups of the Japan's National Health and Nutrition Survey provided by the Ministry of Health, Labor, and Welfare, that were eaten at a high frequency in diet surveys from health examinations conducted on local female farmers in area A in 2001–2002 [4]. We then deleted food items in which Cd concentrations were undetectable in the study on the absorption rate of dietary Cd among female farmers conducted in the winter of 2002–2003 [9]. We added food items that were assumed to have high Cd concentrations, such as seaweed, shellfish, mollusks, and livers, based on a previous study [10]. In November and December 2003, we purchased these food items at local markets in areas A and B based on the recommendations of a local female farmer who was familiar with traditional dietary patterns. We collected 100 food items for the measurement of Cd concentrations. We aimed to collect three of each food item in each area; however, this number increased or decreased depending on the inventory status. Therefore, the number of purchases ranged between 1 and 10, with an average of six. We collected 100 g of the edible portions of each food item and stored them at room temperature, 4 ◦C, or −20 ◦C depending on their perishability until the measurement of Cd concentrations.

#### *2.2. Measurement of Cd Concentrations*

The measurement of Cd concentrations in food items, except for rice, was conducted by Japan Food Research Laboratories (Tokyo, Japan). Next, 10 to 20 g of the edible portion of food, which was precisely measured in a Kjeldahl flask, was added to 200 mL of nitric acid (HNO3) and heated. After the vigorous reaction was completed, 5 mL of sulfuric acid (H2SO4) was added to the flask and heated again until the color changed to light yellow. After cooling, the inside of the flask was washed well with deionized water (less than 1 μs/cm of conductance) and heated again until H2SO4 was released as white smoke. The residual was then dissolved in an appropriate quantity of deionized water to make a sample solution. The solution was moved to a separatory funnel and 10 mL of 50% diammonium hydrogen citrate and Thymol blue indicator (0.1 g of Thymol blue in 100 mL of ethyl alcohol) were then added. After neutralization with ammonium solution, the volume of the solution was increased to 100 mL by the addition of deionized water, and this was followed by 5 mL of 3% ammonium pyrrolidine-N-dithiocarbamate (APDC) solution/ammonium sulfate and 10 mL of butyl acetate with shaking for 5 min. After being left to stand, the butyl acetate layer was collected for the measurement of Cd concentrations using a flame atomic absorption spectrometer (AA-890, Nippon Jarrell-Ash Co., Ltd., Tokyo, Japan). The original standard Cd solution (Kanto Chemical Co., Inc., Tokyo, Japan) was diluted with 1% hydrochloric acid (HCl) to make 0.4 and 0.8 μg/mL standard Cd solutions. Quality control was achieved in the analysis using sugar (commercial products) added with Cd as an alternative to certified reference materials. Its additional recovery was maintained within 90–110%. The detection limits for cereals, other food items, and drinking water were 0.01 mg/kg, 0.005 mg/kg, and 0.001 mg/L, respectively.

#### *2.3. Health Examinations and Diet Surveys*

Health examinations performed on female farmers in areas A and B in 2001–2004 were described previously [4.5]. In the present study, we used data obtained on age, height, weight, and the results of a diet history questionnaire (DHQ) from 725 and 438 subjects in areas A and B, respectively. DHQ is designed to assess food and nutrient intake levels in the previous month based on the quantity and semiquantitative frequency of the consumption of 110 food items commonly eaten in Japan [11]. Estimates of intake for food, energy, and selected nutrients were calculated using an ad hoc computer algorithm for DHQ based on Standard Tables of Food Composition in Japan, which has already been validated [12,13]. We used data obtained on individual food intake levels to calculate Cd intake levels. Among subjects, 12 and 6 with extremely low or high energy intake levels (≤1000 or ≥3500 kcal/day) were excluded in areas A and B, respectively, in addition to one whose consumption of rice was zero in area A, resulting in 712 and 432 subjects for analyses.

#### *2.4. Calculation of Cd Intake Levels*

We calculated the Cd intake levels of individual subjects by multiplying the intake of each food item by its Cd concentration and then summed Cd intake levels from all foods consumed. However, the food items for which intake levels were assessed by DHQ and those for which Cd concentrations were measured were not always in one-to-one correspondence. Therefore, we adjusted mismatches for reconciliation, as described below. Regarding boiled barley-rice, which is assumed to consist of 70% rice and 30% wheat, the Cd concentrations of rice and wheat flour, respectively, were multiplied. Since the intake levels of udon (wheat noodles) and soba (buckwheat noodles) were collectively assessed by DHQ, they were divided into two halves, each of which was multiplied by the Cd concentration of udon or soba. The intake levels of various noodles, including Chinese noodles or spaghetti, which were separately assessed by DHQ, were multiplied by the Cd concentration of udon, assuming that these foods were similarly made of wheat flour. The intake levels of butter rolls, croissants, pizza, pancakes, and okonomiyaki (Japanese-style pancakes) were also multiplied by the Cd concentration of white bread, while those of snack foods, Japanese sweets, cakes, cookies, and doughnuts were multiplied by the Cd concentration of manju (Japanese sweet bun). Regarding sweet potatoes, taro, yams, and Chinese yams, which were assessed collectively by DHQ, the average Cd concentration of sweet potatoes, taro, and yams was used. The average Cd concentration of silken and cotton tofu was used for tofu (soybean curd). The average Cd concentration of spinach, garland chrysanthemum, Japanese mustard spinach, Bok choy (Chinese cabbage), and Japanese leek was used for leafy green vegetables, which were collectively assessed by DHQ. Concerning the intake of wakame seaweed, the Cd concentration of raw wakame seaweed was used. The intake of mushrooms was evenly divided into shiitake mushroom and other mushrooms, and the Cd concentrations of shiitake mushroom and maitake mushroom were multiplied. Regarding the intake of shrimp and fish eggs, the average Cd concentrations of prawns and shrimp and of cod and salmon roe, respectively, were used. Since the intake of shellfish included oysters and other shellfish in DHQ, the Cd concentrations of oysters with innards and the average Cd concentrations of scallops without innards, Japanese littleneck clams, and freshwater clams were respectively multiplied. The average Cd concentration of beef liver, pork liver, chicken liver, Hinai chicken liver, and the innards of Hinai chicken was used for the intake of liver. Food items for which Cd concentrations were not detected, such as animal meat, eggs, milk, and green tea (after brewing), were not included in the calculation of total Cd intake. Cd intake from garlic, okra, belvedere fruit, kelp, hijiki seaweed, agar-agar, mozuku seaweed, and scallops with innards was also excluded because they were not evaluated by DHQ; however, their Cd concentrations were measured. The arithmetic means (AMs) of the Cd concentrations of these food items from both areas were multiplied by their corresponding food intake, while rice Cd concentrations in individual subjects, which were obtained in previous health examinations, were used to calculate individual Cd intake from rice. When a subject consumed brown rice, the Cd concentration of which is generally reduced by 10% due to polishing [14], the Cd concentration adjusted to be equivalent to polished rice was multiplied for calculations. One rice sample was missing in area A, which was substituted with the geometric mean (GM) of the Cd concentration in rice. The Cd concentrations of some foods with masses that are changed by cooking were corrected using the table of mass changes in individual food items from the Standard Tables of Food Compositions in Japan (7th edition, Japanese Ministry of Education, Culture, Sports, Science and Technology).
