Food Chemistry 168 (2015) 294–301

Contents lists available at ScienceDirect

Food Chemistry journal homepage:

Effects of polishing, cooking, and storing on total arsenic and arsenic species concentrations in rice cultivated in Japan Shigehiro Naito a,⇑, Eri Matsumoto b, Kumiko Shindoh a, Tsutomu Nishimura b a b

National Food Research Institute, National Agriculture and Food Research Organization, 2-1-12 Kannondai, Tsukuba, Ibaraki 305-8642, Japan Japan Food Research Laboratories, 6-11-10 Nagayama, Tama-shi, Tokyo 206-0025, Japan

a r t i c l e

i n f o

Article history: Received 13 February 2014 Received in revised form 5 June 2014 Accepted 9 July 2014 Available online 18 July 2014 Keywords: Rice Total arsenic Inorganic arsenic Polishing Washing Cooking Storage

a b s t r a c t The effects of polishing, cooking, and storing on total arsenic (As) and As species concentrations in rice were studied adopting typical Japanese conditions. Total and inorganic As levels in three white rice samples polished by removing 10% of bran by weight were reduced to 61–66% and 51–70% of those in brown rice. The As levels in the white rice after three washings with deionized water were reduced to 81–84% and 71–83% of those in raw rice. Rinse-free rice, which requires no washing before cooking because bran remaining on the surface of the rice was removed previously, yielded an effect similar to that of reducing As in rice by washing. Low-volume cooking (water:rice 1.4–2.0:1) rice to dryness did not remove As. The As content of brown rice stored in grain form for one year was stable. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Rice is a major source of inorganic arsenic (iAs), particularly in Asia and other countries where it is a staple food (Benford et al., 2011). The International Agency for Research on Cancer (2012) has classified iAs (arsenite As(III) and arsenate As(V)) as human carcinogens. The predominant As species in rice are As(III), As(V), and dimethylarsinic acid (DMA) (Heitkemper, Vela, Stewart, & Westphal, 2001; Huang, Fecher, Ilgen, Hu, & Yang, 2012; Meharg et al., 2008; Narukawa, Hioki, & Chiba, 2011; Nishimura et al., 2010; Williams et al., 2005; Zavala, Gerads, Gürleyük, & Duxbury, 2008; Zhu et al., 2008). Some literature also includes monomethylarsonic acid (MMA) (D’Amato, Forte, & Caroli, 2004; Pizarro, Gómez, Palacios, & Cámara, 2003), arsenobetaine (AsB) Abbreviations: As, arsenic; As(III), arsenite; As(V), arsenate; AsB, arsenobetaine; AsC, arsenocholine; CCCF, Codex Committee on Contaminants in Food; CRM, certified reference material; DMA, dimethylarsinic acid; DP%, degree of polishing; FAPAS, food analysis performance assessment scheme; HPLC, high-performance liquid chromatography; IARC, International Agency for Research on Cancer; iAs, inorganic arsenic; ICP-MS, inductively coupled plasma-mass spectroscopy; JECFA, Joint FAO/WHO Expert Committee on Food Additives; LOD, limit of detection; LOQ, limit of quantitation; MAFF, Ministry of Agriculture, Forestry and Fisheries of Japan; MMA, monomethylarsonic acid; NMIJ, National Metrology Institute of Japan; RSDi, intermediate relative standard deviation; RSDr, repeatability relative standard deviation; SD, standard deviation; tAs, total arsenic. ⇑ Corresponding author. E-mail address: [email protected] (S. Naito). 0308-8146/Ó 2014 Elsevier Ltd. All rights reserved.

(Mihucz et al., 2007; Pizarro et al., 2003), or arsenocholine (AsC) (Mihucz et al., 2007). The level of iAs in consumed rice depends on environmental conditions (As in soil and water), agricultural conditions (materials, cultivars, and control of irrigation water), and rice-processing and preparation methods (Benford et al., 2011). Reports of iAs concentrations in raw rice for risk assessment have been increasing, thus increasing the need for determining iAs in rice in order to assess the risk to humans (Carbonell-Barrachina et al., 2012; Juhasz et al., 2006; Laparra, Vélez, Barberá, Farré, & Montoro, 2005; Rintala, Ekholm, Koivisto, Peltonen, & Venäläinen, 2013; Schoof et al., 1998, 1999; Signes, Mitra, Burló, & CarbonellBarrachina, 2008; Smith et al., 2006; Torres-Escribano, Leal, Vélez, & Montoro, 2008; Williams et al., 2005). However, few reports have focused on the effects of rice-processing and preparation methods on iAs content (Ackerman et al., 2005; Juhasz et al., 2006; Laparra et al., 2005; Mihucz et al., 2007; Narukawa et al., 2011; Raab, Baskaran, Feldmann, & Meharg, 2009; Signes et al., 2008; Smith et al., 2006; Sun et al., 2008; Torres-Escribano et al., 2008). The Codex Committee on Contaminants in Food (CCCF), whose agenda included discussion of arsenic (As) in rice in the 1990s but discontinued it in 1999, has resumed this discussion since the Joint FAO/WHO Expert Committee on Food Additives (JECFA) re-evaluated the health risk of As exposure in 2010. CCCF has asked member countries, especially rice-producing countries, to submit


S. Naito et al. / Food Chemistry 168 (2015) 294–301

relevant data and information regarding As contamination in rice. Specifically, there is an urgent need to accumulate data on iAs in rice. Although Japan is a major rice-producing country, only a few studies have focused on total As and iAs concentrations in rice cultivated in Japan (Arao, Kawasaki, Baba, Mori, & Matsumoto, 2009; Kuramata, Abe, Matsumoto, & Ishikawa, 2011; Nagaoka, Nishimura, Matsuda, & Maitani, 2008; Narukawa et al., 2011; Nishimura et al., 2010). In the present study, we investigate the effects of the following rice-processing and preparation treatments on total As and As species concentrations in rice by adopting a main rice cultivar (Koshihikari; (Oriza sativa L.) and a typical condition in Japan: polishing, processing rinse-free white rice, washing and low-volume cooking (water:rice 1.4–2.0:1), and storing rice in grain form for one year.

was adopted. Each sample was mixed just after procurement and repacked in rice bags, which were additionally placed in polyethylene bags for storage until treatment. Two rice bags (60 kg) of each sample (samples 1, 2, and 4) were selected randomly and then sorted using a grain colour sorter (ES-01AM, Satake, Hiroshima, Japan) that sorts sound rice grain from unsound rice grain (e.g., broken, immature, or coloured grains). Japanese rice-milling factories are usually equipped with grain sorters. The percentages of sound rice grain were 98.8% for sample 1, 99.9% for sample 2, and 95.0% for sample 4. The sorted brown rice samples were used for polishing, washing, and cooking; unsorted brown rice was used for rinse-free rice (sample 2) and storage (sample 3). Rice samples in polyethylene bags were kept at 10 °C until treatment and at 2– 8 °C from treatment to analysis.

2. Materials and methods

2.2. Polishing

The study design for treatment is presented in Table 1. We studied polishing and cooking as main treatments that cover most rice processing from harvest to table. In addition, we studied rinse-free rice, washing, and storage as supplement treatments. Throughout the study, two analytical samples from each experiment were analysed. The experiments of individual treatments were performed in triplicate or in five replicates. Deionized water used throughout this study was purified in a Milli-Q system (Merck Millipore Corporation, Tokyo, Japan).

White rice with a degree of polishing (DP%) by weight of 90% or 95%, was prepared from brown rice samples 1, 2, and 4 using a household rice mill (RSKM5B, Satake) at Satake Corp. (Hiroshima, Japan). White rice with 95DP% (90DP%) was obtained from milling 5% (10%) of outer layers from brown rice. Brown rice (750 g) was polished in the mill per run, and four runs were performed in one experiment. Five-replicate experiments were conducted on different days. White rice of each experiment was sifted using a sieve with a 0.5 mm wire diameter and a 2.0 mm aperture. The remaining white rice grains were used for washing and cooking. White rice grains are generally sorted after polishing in Japanese rice-milling factories.

2.1. Samples For selecting representative rice samples in Japan, we first chose the rice cultivar Koshihikari (O. sativa L.) because its share of total production is 40% and the shares of cultivars other than Koshihikari do not exceed 10% (Ministry of Agriculture, Forestry and Fisheries of Japan (MAFF), 2010). Furthermore, there were few genotypic differences in the level of either total As or iAs in the grains of 10 major Japanese rice cultivars including Koshihikari (Kuramata et al., 2011). Next, we determined the target total As levels (0.1 mg/kg, 0.3 mg/kg, and 0.5 mg/kg) in brown rice by reference to reports that (1) total As concentrations in 17 rice samples (brown rice, white rice, glutinous rice, and nonglutinous rice) produced in Japan ranged from 0.10 to 0.54 mg/kg (Nishimura et al., 2010) and (2) total As concentrations in brown rice from the major rice cultivars produced in Japan ranged from 0.04 to 0.33 mg/kg, with a mean of 0.16 mg/kg (n = 199) (MAFF, 2006). This MAFF survey reflected regional variations rather than genotypic variations (Kuramata et al., 2011). We thus collected four brown rice samples (ID = 1–4) with different total As levels from different regions in Japan (Table 1), and used sample 2 or sample 3 as a sample with a representative total As level in domestic rice. Sample 1 (180 kg = 30 kg/bag  6 bags) was purchased from one farmer, and samples 2 (1050 kg = 30 kg/bag  35 bags) and 3 (180 kg = 30 kg/bag  10 bags) were purchased from another farmer. Sample 4 (90 kg = 30 kg/bag  3 bags) was procured from an agricultural research institute where water-saving cultivation

2.3. Rinse-free rice treatment Rinse-free white rice requires no washing for cooking because bran remaining on the surface was removed during production. Methods of producing rinse-free white rice in Japan include the wash-and-dry method, the adsorbent method using rice bran or heated tapioca, and the brushing method. In this study, bran was removed by the adsorbent method using bran other than that of the sample. We studied rinse-free treatment as a supplemental experiment because of the minimum sample amount (180 kg/run) required for this processing, and used brown rice sample 2 as a representative sample for this treatment. Brown rice sample 2 was mixed, sorted, and polished using a rice mill (DCM-75, Toyo Rice Cleaning Machine) and then processed for rinse-free white rice (‘‘BG’’ rinse-free rice equipment with a production capacity of 3 ton/h, Toyo Rice Cleaning Machine) at a rice-milling factory of Toyo Rice Cleaning Machine Co., Ltd. (Wakayama, Japan). Brown rice (180 kg) was polished, sorted, and processed in one experiment; and five-replicate experiments were conducted on different days. DP% of white rice at each run could not be obtained because the rice mill at the factory is not equipped with a balance. Bran removed from white rice during production of rinse-free white rice could not be collected because

Table 1 Samples used for each treatment. Sample ID

1 2 3 4 a

Total As in brown rice (mg/kg)a

Moisture in brown rice (%)

Production area

0.503 0.235 0.250 0.048

13.4 14.1 13.6 12.5


Expressed as mg/kg dry matter.

Cultivation year

2012 2011 2012 2012

Main treatment

Supplement treatment



Rinse free


s s

s s


s s






S. Naito et al. / Food Chemistry 168 (2015) 294–301

bran of rice other than the sample had been placed previously in the equipment and was partially mixed with bran of the sample. 2.4. Washing We studied washing treatment as a supplemental experiment because washing is a part of cooking. Brown rice, white rice with 95DP% and 90DP% (sample 2), and white rice with 90DP% (sample 1) were used as representative samples for this treatment. Raw rice was washed by placing 450 g portions of rice in an electric compact rice washer (KOM-3, Izumi Products Co., Nagano, Japan) and then adding 600 mL of deionized water. The first washing lasted 10 s, and the second and third washings lasted 30 s each. These washing times were determined according to the recommendation of the rice-washer manual. The water was decanted, 600 mL more water was added, then 40 g washed rice was sampled at each washing time. Triplicate experiments were conducted on different days. Two analytical samples (washed rice and washing water) for each washing in each experiment were analysed. 2.5. Cooking Brown rice, white rice with 95DP% and 90DP% (samples 1, 2, and 4), and rinse-free white rice (sample 2) were cooked. Raw rice except for rinse-free white rice was washed under the same conditions as described for washing. However, washed-rice sampling was not performed. The washed rice or rinse-free white rice was soaked in deionized water and cooked automatically using an electric rice cooker (NP-NC10, Zoujirushi, Osaka, Japan). The cooker was selected based on brand share and price share information for electric rice cookers in Japan. The following rice:water ratios by weight were used: 1:2 for brown rice, 1:1.6 for white rice with 95DP%, and 1:1.4 for white rice with 90DP% and rinse-free white rice. Excess water was evaporated to dryness. The rice:water ratio and the automatic mode of the rice cooker for each rice sample were determined according to the recommendation of the ricecooker manual. Two cooked rice samples (120 g) in each experiment were freeze-dried for analysis. Triplicate cooking experiments for each rice sample were conducted on different days. 2.6. Storage We studied storage treatment as a supplemental experiment because our internal quality control data in analysis of As content in rice had suggested storage stability of As content in rice. Brown rice sample 3 obtained shortly after harvest was used as a representative sample for this treatment. Brown rice sample 3, harvested in September 2012, was stored at 15 °C or 25 °C from October 2012 to October 2013. In Japan, rice is usually stored as brown rice until polishing for distribution, and brown rice in grain form is usually stored for less than one year. Storage conditions of 15 °C and 70 RH% are adopted for government-controlled rice in Japan. The annual mean temperature in Tokyo is 16.6 °C, and the average temperature in summer is 26.2 °C, based on weather statistics in the last decade. Five rice bags (30-kg brown rice per bag) were stored at each temperature. Humidity was not controlled, although each rice bag stored at 15 °C was additionally placed in a polyethylene bag in order to reduce variation of moisture in rice. At the beginning of storage, two analytical samples from each of the five rice bags were taken for analysis, and homogeneity amongst the five bags was confirmed. For each storage period (one month, eight months, and twelve months), three bags were randomly selected from the five bags; then two analytical samples from each bag were taken for analysis.

2.7. Total arsenic determination Rice samples (raw rice, washed rice, and cooked rice after freeze-drying) were milled to particles of less than 0.5 mm using a food mill (Force Mill FM-1, Osaka Chemical Co., Ltd., Osaka, Japan). The pretreatment and measurement conditions reported in a previous study (Nagaoka et al., 2008) were partly modified as follows: 1 g of rice powder, 0.2 g of bran, or 10 g of washing water for extraction; perchloric acid (1 mL, 60%) instead of hydrogen peroxide (2 mL, 30%); potassium iodide solution of 20% instead of 40%; and 10% ascorbic acid solution of 2.5 mL instead of 5 mL. Total As in rice, bran, and washing water was determined using an atomic absorption spectrometer (AA-220, Varian Technologies) equipped with a vapour generation accessory (VGA-77, Varian Technologies). Factors for converting between dry mass and wet mass were obtained by measuring the mass loss after drying portions of the samples at 135 °C for 1 h. The correction factors were also used in the As speciation analysis. The performance characteristics were as follows: limit of detection (LOD, 3r) (mg/kg) = 0.004 (rice), 0.02 (bran), and 0.0004 (washing water); recovery (%) = 98.8 (rice), 95.8 (bran), and 90.8 (washing water); repeatability relative standard deviation (RSDr) (%) = 2.7 (rice), 3.0 (bran), and 3.6 (washing water); and intermediate relative standard deviation (RSDi) (%) = 4.7 (rice), 4.0 (bran), and 4.5 (washing water). The LOD value was derived from eightreplicate blank analyses. The mean recovery was obtained from triplicate analyses at two added concentrations (2- and 20-fold limit of quantitation (LOQ)). The RSDr and RSDi means were calculated from the data analysed in eight replicates daily for three days each at two added concentrations (2- and 20-fold LOQ). 2.8. Inorganic arsenic speciation Arsenic speciation analyses of As(III), As(V), DMA, and MMA were carried out using an HPLC (Agilent 1200 Series, Agilent Technologies Japan, Tokyo, Japan) coupled with an ICP-MS (Agilent 7500ce, Agilent Technologies Japan, Tokyo, Japan). The analytical conditions reported in the previous studies (Nagaoka et al., 2008; Nishimura et al., 2010) were modified as follows: a rice sample (0.5 g) was mixed with 0.15 mol/L nitric acid; a bran sample (0.1 g) was mixed with 0.3 mol/L nitric acid; a washing water sample (0.5 g) condensed at 100 °C from 2.5 g of the sample was mixed with 0.15 mol/L nitric acid; heating temperature for extraction from samples (rice, bran, or washing water) was 100 °C for 2 h; extract was treated by adding 2 mL deionized water and centrifuging at 3500 rpm for 10 min, followed by collecting the supernatant. The residue was treated in the same way two more times, and then all the supernatants were mixed. The amount of iAs was calculated as the sum of As(III) and As(V). The protocol for this analysis in rice was evaluated in an international collaborative study (Ukena et al., 2014). The performance characteristics were as follows: LOD (mg/ kg) = 0.001 (iAs in rice), 0.002 (DMA or MMA in rice), 0.03 (iAs in bran), 0.02 (DMA or MMA in bran), 0.0004 (iAs in washing water), 0.0002 (DMA in washing water), and 0.0003 (MMA in washing water); recovery (%) = 97.0 (iAs in rice), 96.5 (DMA in rice), 94.0 (MMA in rice), 98.2 (iAs in bran), 104.3 (DMA in bran), 100.8 (MMA in bran), 95.3 (iAs in washing water), 98.3 (DMA in washing water), and 96.5 (MMA in washing water); RSDr (%) = 1.7 (iAs in rice), 4.7 (DMA in rice), 19 (MMA in rice), 1.7 (iAs in bran), 2.1 (DMA in bran), 3.4 (MMA in bran), 3.2 (iAs in washing water), 3.1 (DMA in washing water), and 2.8 (MMA in washing water); RSDi (%) = 3.3 (iAs in rice), 7.4 (DMA in rice), 25 (MMA in rice), 2.6 (iAs in bran), 2.5 (DMA in bran), 3.8 (MMA in bran), 3.5 (iAs in washing water), 3.2 (DMA in washing water), and 2.9 (MMA in


S. Naito et al. / Food Chemistry 168 (2015) 294–301

washing water). Each performance characteristic was calculated in the same way as in total As analysis. 2.9. Analysis quality assurance and quality control All analyses were performed at the Tama Laboratory of the Japan Food Research Laboratories, which is an ISO 17025 accredited laboratory for analysing As in food using atomic absorption spectrometry. This laboratory participated in the FAPAS proficiency testing programs for determining total As and iAs in powdered rice in 2011 through 2013 (all z scores 62). Each analytical batch contained a procedural blank, spiked samples or CRM (NMIJ 7503-a for Total As, iAs, and DMA in white rice; NMIJ 7531-a for total As in brown rice), and one pair of analytical portions for checking repeatability. Data of the batch were accepted if the spiked recovery was between 80% and 110% or the analytical value of the CRM was within the 95% confidence interval of the CRM certified value, and the difference between one pair of analytical portions was below 20% of the mean value of the two portions. 2.10. Statistical analysis Microsoft Excel 2010 and 2013 (Microsoft, Tokyo, Japan) were used for data analysis. 3. Results and discussion MMA was not detected in any samples. As concentration is expressed as mg/kg dry matter for rice and mg/kg for rinse water. 3.1. Polishing Total As and As species concentrations amongst brown rice and white rice with 95DP% and 90DP% indicated that total As and iAs concentrations decreased, depending on DP%; however, DMA content varied little (Table 2). In Japan, white rice with 90DP% is a staple food, although some people prefer brown rice or white rice

with 93–97DP%. The total As and iAs concentrations in white rice with 90DP% (95DP%) were reduced to 61–66% and 51–70% (80– 84% and 75–89%) of those in brown rice (Table 2). To our knowledge, studies using brown and white rice prepared directly from brown rice are limited (Narukawa et al., 2011; Sun et al., 2008). The percentages of reductions presented above correspond well with the results of Narukawa et al. (2011), indicating that in a white rice sample with 90DP% total As concentration was reduced to 62% and iAs concentration was reduced to 60%, by polishing the brown rice with 0.173 mg/kg of total As, cultivated in Japan. Sun et al. (2008) reported total As and As species concentrations in brown rice, and white rice with 93DP% and its bran prepared from two Chinese and three Bangladeshi brown-rice samples. Their results indicated that total As and iAs concentrations were the highest in bran, followed by brown and white rice where concentrations were reduced to 60–82% and 43–83% of those in brown rice. They also reported that DMA content was fairly uniform throughout the grain, and MMA content was less than LOD in rice except for one brown rice, and near LOQ in bran. These results exhibited the same trend as our samples 1 and 2 in Table 2; however, the percentage reduction of iAs content reached more than 50% for their two samples, despite 93DP%. Total As and iAs concentrations in bran were 3–10 times higher than those in brown rice (Table 2). Moreover, concentrations in bran of white rice (samples 1 and 2) with 95DP% were higher than those of white rice with 90DP%. This result suggests that more iAs is located in the outer layer of brown rice, which supports the findings of Meharg et al. (2008) that As was preferentially localised at the surface, in the region corresponding to the pericarp and aleurone layer, and most of the total As might be in the form of iAs in these layers. Meharg et al. (2008) reported that the percentage of iAs was higher in brown rice than in polished rice, and decreased as total As levels in rice grains increased, but where the polished rice was not prepared directly from brown rice. The results of samples 1 and 2 in Table 2 support these findings, although the percentage differences are not as large.

Table 2 Concentrations of total As and As species in rice or in bran with different degrees of polishing. Sample ID

Sample type


Total Asb Mean



% tAs SD


Inorganic Asb e



% iAs

iAs/tAs (%)


DMAb Mean




Extraction ratio (%)c

Mass balance of tAs (%)d Meanf



Brown White White

100 95 90

0.487 0.411 0.296

0.018 0.017 0.013

100 84 61

0.431 0.325 0.221

0.020 0.014 0.008

100 75 51

89 79 75

0.073 0.074 0.079

0.0053 0.0050 0.0034

100 101 108

103 97 101

– 105 104

– 4 4


Brown White White

100 95 90

0.223 0.179 0.147

0.009 0.005 0.005

100 80 66

0.208 0.156 0.132

0.007 0.008 0.006

100 75 63

93 87 90

0.016 0.016 0.017

0.0007 0.0009 0.0011

100 100 106

100 96 101

– 100 100

– 4 4


Brown White White

100 95 90

0.040 0.033 0.025

0.002 0.002 0.003

100 83 63

0.044 0.039 0.031

0.002 0.002 0.002

100 89 70

110 118 124


Effects of polishing, cooking, and storing on total arsenic and arsenic species concentrations in rice cultivated in Japan.

The effects of polishing, cooking, and storing on total arsenic (As) and As species concentrations in rice were studied adopting typical Japanese cond...
306KB Sizes 4 Downloads 3 Views