**3. Results**

Food items and their Cd concentrations, divided into 10 subgroups, are listed in Tables 1–10. Cd concentrations in rice and rice products were high (Table 1). The average Cd concentrations in rice, 0.158 and 0.109 mg/kg, were lower than the safety standard of 0.4 mg/kg; however, 8.2 and 5.8% of rice had Cd concentrations that were above the safety standard in areas A and B, respectively. Among cereals, tubers, and roots, while the Cd concentrations of wheat flour and its products were not high, taro (satoimo in Japanese) had a high Cd concentration of 0.289 mg/kg (Table 2). The Cd concentrations of soybeans, including edamame, were high, whereas those of their processed foods were not, except for miso (Table 3). Cd was detected in all vegetables investigated, among which spinach, Japanese parsley, garland chrysanthemum, Japanese mustard spinach, and belvedere fruit showed high Cd concentrations (Table 4). Shiitake mushroom and seaweed (wakame, kombu, nori, and hijiki in Japanese) had markedly high Cd concentrations (Tables 5 and 6). Among fish and shellfish, salted squid guts, scallops with innards, oysters with innards, and freshwater clams, all of which had innards, had very high Cd concentrations, while fish meat itself did not (Table 7). Cd was not detected in many livestock food items, such as meat, eggs, and milk, except for the innards (Table 8). Cd was not detected in fruit (Table 9). High Cd concentrations were found in chocolate and tea leaves (Table 10). Cd was not detected in brewed tea.

We then calculated Cd intake levels by subjects who underwent health examinations, multiplied Cd concentrations in food items by individual food intake, and showed the results obtained in subgroups. The backgrounds of subjects in areas A and B, the food intake levels for whom were used to calculate Cd intake levels, are shown in Table 11. No significant di fferences were observed in age, energy intake, or rice intake between the 2 areas, whereas significant di fferences were noted in height (*p* = 0.047) and weight (*p* = 0.047), but were biologically negligible.

Cd intake from seaweed, fish, and shellfish was combined into one subgroup as seafood and that from livestock food was included in the subgroup as others, while that from fruit was excluded from calculations (Table 12). Among the subgroups, Cd intake from the subgroup of rice and rice products was the highest in both areas, and accounted for approximately 40–50% of the total Cd intake. Cd intake levels from the subgroups of vegetables and seafood were higher than those in the other subgroups. Cd intake from the subgroup of rice and rice products was significantly higher in area A than in area B, while those from other subgroups were similar between the two areas, except for vegetables, with Cd intake being significantly lower in area A than in area B. The median total Cd intake levels in areas A and B were 55.7 and 47.8 μg/day, respectively, with the former being significantly higher than the latter.


**Table 1.** Cadmium (Cd) concentrations in rice and rice products. Data are presented as arithmetic means, except for rice, which are presented as geometric means (GMs).

> \*: One sample is missing.

**Table 2.** Cadmium (Cd) concentrations in cereals, tubers, and roots. Data are presented as arithmetic means.


\*: Only from area A.

**Table 3.** Cadmium (Cd) concentrations in soybeans and soybean products. Data are presented as arithmetic means.


**Table 4.** Cadmium (Cd) concentrations in vegetables. Data are presented as arithmetic means.



**Table 4.** *Cont.*

\*: Only from area A.

**Table 5.** Cadmium (Cd) concentrations in mushrooms. Data are presented as arithmetic means.


**Table 6.** Cadmium (Cd) concentrations in seaweed and seaweed products. Data are presented as arithmetic means.


\*: Only from area A.

**Table 7.** Cadmium (Cd) concentrations in fish and shellfish. Data are presented as arithmetic means.



**Table 7.** *Cont.*

\*: Only from area B.

**Table 8.** Cadmium (Cd) concentrations in livestock food. Data are presented as arithmetic means.


\*: Only from area A.

**Table 9.** Cadmium (Cd) concentrations in fruit. Data are presented as arithmetic means.


**Table 10.** Cadmium (Cd) concentrations in others. Data are presented as arithmetic means.


\*: Only from area A.


**Table 11.** Backgrounds of female farmers who underwent health examinations in areas A and B.

\*: *p* < 0.05 versus area A (unpaired Student's *t*-test).

**Table 12.** Daily cadmium intake per person (μg/day) in female farmers in areas A and B.


\*: *p* < 0.05 versus area A (median test). \*\*: including seaweed, fish, and shellfish. \*\*\*: including manju, livestock food, chocolate, and flavor seasonings.

We then compared the results obtained with Cd intake by the general population in Japan (Figure 2) [15,16]. Cd intake has been gradually decreasing in Japan: 31.1 μg/day (16.2 μg/day from rice and 14.9 μg/day from other food items) in 1981, 21.1 μg/day (7.8 μg/day from rice and 13.3 μg/day from other food items) in 2007, and 17.8 μg/day (5.7 μg/day from rice and 12.1 μg/day from other food items) in 2015. This was mainly attributed to a reduction in Cd intake from rice. Cd intake levels from rice and rice products in areas A (28.3 μg/day) and B (19.4 μg/day) were 3.6- and 2.5-fold higher, respectively, than that (7.8 μg/day) by the general population in 2007, while Cd intake levels from other food items in areas A and B (approximately 24.0 μg/day) were 1.8-fold higher than that (13.3 μg/day) by the general population. Total Cd intake levels in areas A and B were approximately 2.5-fold higher than that by the general population.

**Figure 2.** Cadmium intake levels per person from rice and rice products and from others in female farmers in cadmium-polluted areas A and B in Akita, Japan, shown by medians, and comparisons with average cadmium intake in the Japanese population.

We also calculated weekly Cd intake per body weight, using individual values of body weight, to compare the PTWI of JECFA at that time or the current TWI in Japan of 7 μg/kg BW/week. Median weekly Cd intake levels were 7.2 and 6.0 μg/kg BW/week in areas A and B, respectively, with both being approximately PTWI or TWI (Table 13). The distributions of weekly Cd intake levels in areas A and B were shown in histograms (Figure 3). The exclusion of data obtained from two subjects in area A that were extremely large and considered to be outliers resulted in similar distributions in both areas that skewed to the higher side. The percentages of subjects with weekly Cd intake levels above PTWI or TWI were 51.7 and 38.0% in areas A and B, respectively (*p* < 0.05, χ*2* test) (Table 13).

**Table 13.** Weekly cadmium intake per body weight (μg/kg BW/week) in female farmers in areas A andB and their distribution.


 *p* <

**Figure 3.** Distribution of weekly cadmium intake per body weight in female farmers in cadmium-polluted areas A and B in Akita, Japan.

We further divided weekly Cd intake per body weight into age-classified groups and examined differences between them (Table 14). In area A, weekly Cd intake per body weight was higher in subjects aged 40 or older than in younger subjects, while no significant differences were observed between age-classified groups in area B. Weekly Cd intake per body weight in subjects aged 50 or older was above PTWI or TWI in area A.


**Table 14.** Age-classified weekly cadmium intake per body weight (μg/kg BW/week) in female farmers in areas A and B.

\*: *p* < 0.05 versus 20–29 years (Steel-Dwass test).

We then performed the Monte Carlo simulation using the same data and the results obtained are shown in Figure 4. We excluded 2 outliers in area A, which markedly skewed their probability density distributions. As a result, we obtained estimated median weekly Cd intake levels of 7.0 and 6.0 μg/kg BW/week in areas A and B, respectively (Table 15), which were similar to the results described above.

**Figure 4.** Probability density distributions of weekly cadmium intake per body weight in female farmers in cadmium-polluted areas A and B in Akita, Japan, estimated using the Monte Carlo simulation.

**Table 15.** Results of the Monte Carlo simulation for weekly cadmium intake per body weight (μg/kg BW/week) in female farmers in areas A and B.


\*: Two outliers were excluded.
