Loss of zinc from zinc‐biofortified rice and conventional rice varieties due to various cooking techniques practiced in rural Bangladesh

Abstract

BACKGROUND

Zinc‐biofortified rice has the potential to provide a sustainable solution for zinc deficiency, particularly in Asian countries where rice is the primary staple food. However, cooking of raw rice is associated with some loss of micronutrients, including zinc. The literature suggests that the amount of water used during cooking can influence the extent of micronutrient loss in both fortified and conventional rice varieties. This study aimed to estimate the cooking loss of zinc, iron, and phytate, as an important inhibitor of zinc, in zinc‐biofortified and conventional rice varieties and to assess the effects of the amount of water used for cooking on the loss of these micronutrients. The secondary objective was to assess the impact on iron.

RESULTS

Significant differences in zinc content were observed between the biofortified and conventional raw rice varieties. The median zinc content (interquartile range) was 19.7 (19.3, 20.2) mg kg⁻¹ in the biofortified variety and 18.7 (18.3, 18.7) mg kg⁻¹ in the conventional variety (P = 0.012). Cooking resulted in a significant reduction in zinc, iron, and phytate content with higher losses observed when cooking with ‘excess’ water (39% for zinc and 29% for iron) than when cooking with ‘adequate’ quantities of water (19% for zinc and 22% for iron).

CONCLUSION

This is the first study to estimate the cooking loss of zinc in zinc‐biofortified rice in the Bangladeshi context and to estimate the effect of the amount of water used on this loss. Cooking resulted in a significant decrease in zinc and iron content, and there was a dose–response relationship between the amount of water used and the reduction in zinc and iron content. © 2026 The Author(s). Journal of the Science of Food and Agriculture published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

INTRODUCTION

Rice is an indispensable staple for nearly half of the global population and provides approximately 20% of the world's dietary energy supply. ¹ In some Asian countries, rice can contribute up to 70% of total calorie intake, ¹ with per capita consumption exceeding 100 kg per year. ² Although rice is a good source of energy, it contains only limited amounts of micronutrients such as zinc, iron, calcium, and phosphorus,^{3, 4} all of which are essential for growth, development and various physiological functions in humans. Rice can therefore meet only a small percentage of the recommended dietary allowance (RDA) for these micronutrients. Rice also contains anti‐nutrients such as phytate, which can inhibit the absorption of vital micronutrients such as zinc, iron, and calcium. ⁴

Zinc deficiency (ZD) is a global nutritional problem that negatively impacts physical growth, immune competence, reproductive function, and neurobehavioral development. ⁵ It has been estimated that approximately 17% of the global population is at risk of ZD, but the percentages are much higher in Asian and African countries. ⁶ Bangladesh, a low‐ and middle‐income country (LMIC) in South Asia, experiences a high level of zinc deficiency: 44% of preschool‐age children (PSAC) and 57% of non‐pregnant, non‐lactating women (NPNL) suffer from ZD. ⁷ Traditionally, the Bangladeshi diet is heavily plant‐based with a high reliance on rice as the staple food.^{8, 9} This contains 7–23.9 mg kg⁻¹ of zinc ¹⁰ and 840–2350 mg kg⁻¹ of phytate. ¹¹

However, the high reliance on rice also presents an opportunity to improve zinc intake through the consumption of zinc‐biofortified rice (ZBR). These rice varieties contain higher amounts of zinc than conventional rice (CR) varieties. The zinc accumulates during the growth period. Several ZBR varieties have been developed and released in Bangladesh by the Bangladesh Rice Research Institute (BRRI) with technical support from International HarvestPlus.^{12, 13} The zinc content in these varieties ranges from 19 to 27 mg kg⁻¹ of raw unpolished rice, which is much higher than that in CR. ¹⁴ However, it is also important to consider the interaction between iron and zinc, as they compete for the same transporter, and the effect of phytate, which inhibits zinc absorption, ⁵ before assessing zinc intake from ZBR.

There are also factors that can affect the zinc content of these varieties, and thus the efficacy of the intervention. These include the use of fertilizer, the zinc content of soil, the soil type, environmental conditions, and cultivation methods. ¹⁵ Post‐harvest processing factors, such as parboiling, temperature and duration of parboiling, milling, and polishing, can also impact the zinc content of raw rice.^{16, 17} Finally, during cooking, some zinc is lost from rice^{18, 19, 20, 21} and the extent of this loss depends on various cooking practices, including washing, soaking, and the amount of water used during cooking.^{22, 23, 24} In Bangladesh, two different techniques are used to cook rice. Initially, the rice is washed with water, usually to 5–6 times, until the water is clear. After washing, rice is cooked using a water‐to‐rice ratio of either 1:2 or 1:3, in which all the water is absorbed, and no excess water (gruel) remains once the rice is cooked. The other method involves cooking rice with ‘excess’ water (in a ratio of 1:4 to 1:6), where the ‘excess’ water or gruel is discarded after the rice is cooked. ²⁴ Cooking in ‘excess’ water can result in a greater loss of zinc from raw rice from both conventional and fortified rice varieties.^{22, 23, 24}

Currently, no study has determined the effect of cooking with different amounts of water on zinc loss from ZBR in Bangladesh. The present study was therefore conducted to estimate the cooking loss of zinc from ZBR and CR, and to assess the effect of cooking with ‘adequate’ water versus ‘excess’ water on this loss. Phytate was assessed as an important inhibitor of zinc absorption, and iron was assessed considering its potential interaction with zinc as secondary objectives. This study was carried out as a sub‐study of a community‐based randomized controlled trial (RCT) designed to evaluate the efficacy of ZBR in improving the zinc status of young NPNL women aged between 20 and 29 years in rural Bangladesh.

METHODS

Rice production and sampling

Two rice varieties were selected for this study: one ZBR variety, BRRI dhan100, and one CR variety, BRRI dhan29. The ZBR variety is not yet widely cultivated in the country, so it was produced under the direct supervision of the study team to ensure its purity. In contrast, BRRI dhan29 is one of the most commonly grown and consumed rice varieties in Bangladesh; a paddy of BRRI dhan29 was therefore procured directly from the local market in Birganj, Dinajpur.

BRRI dhan100 is a rice variety developed by the BRRI. It originated from a cross between two rice lines in 2006. The resulting plants were further crossed with BRRI dhan29 in 2007. Through several years of selection, plants that were healthy, resistant to diseases and insects, resistant to lodging, and produced long, slender grains were chosen. After repeated testing and selection, the variety was evaluated in field trials and released as BRRI dhan100 in 2020. ²⁵ In this study, BRRI dhan100 was selected as the ZBR variety because of its higher zinc content (25.7 mg kg⁻¹) and because it is not biofortified with other micronutrients, such as iron. Moreover, it is cultivated during the ‘boro’ season, where a greater proportion of rice cultivation takes place in the country. BRRI dhan100 also has a higher yield (7.7 t ha⁻¹), which could possibly facilitate uptake of the variety by farmers for cultivation and thus help wider adoption. ²⁵

Contract farming was implemented to grow the required amount of BRRI dhan100 paddy (12 000 kg) from certified seeds in the northern district of Nilphamari in Bangladesh. During the cultivation process, standard procedures were followed for the application of fertilizer, insecticides, and irrigation. Before harvesting, rogueing (the uprooting of paddy plants suggestive of other varieties based on their height and growth) was performed to further ensure the purity of the rice variety. Post‐harvest processing of BRRI dhan100 paddy was conducted in a local husking mill in eight batches. The process included pre‐steaming for 4–5 min, soaking in water overnight, parboiling for 6–7 min, and drying in sunlight before dehusking in a husking machine without any polishing. The raw rice was cleaned of bran and dust. A sample of raw rice (~500 g) was collected from each batch for laboratory testing of zinc, iron, and phytate.

The raw rice from the husking mill contained a high proportion of dead and black grains (~ 5%), which reduced the acceptability of the rice. Color sorting was therefore performed in a local auto rice mill (ARM) (Janata Auto Rice Mill, Birganj, Dinajpur). The process involved shaking the raw rice grains to further remove dust, followed by laser‐based color sorting (SORTEX R500, Bűhler, Uzwil, Switzerland). No milling or polishing was carried out throughout the process. The resulting clean, raw rice was packed in 20 kg sacks, which were letter coded, numbered, and weighed before storage. Eight samples (~500 g) were collected from randomly selected sacks and sent to the laboratory for testing of zinc, iron, and phytate in both raw and cooked rice.

BRRI dhan29 paddy was procured from the local market (Birganj, Dinajpur) through the same ARM owners (Janata Auto Rice Mill, Birganj, Dinajpur), and post‐harvest processing was performed following the same process. In the husking mill, the BRRI dhan29 paddy was processed in four batches to produce the required amount of raw rice, resulting in four samples of BRRI dhan29 raw rice from the husking mill. Color sorting was performed in the same ARM, following the same procedure, and eight samples of BRRI‐29 raw rice were collected from randomly selected sacks to test zinc, iron, and phytate in both raw and cooked rice. To summarize: a total of 16 samples of raw rice from the ARM were analyzed in the laboratory to assess the loss of zinc, iron, and phytate during the process of cooking, with eight being ZBR and eight being CR. An additional 12 samples of raw rice from the husking mill (ZBR = 8, CR = 4) were analyzed to explore if color sorting resulted in a reduction of zinc, iron, and phytate. Figure 1 shows the details of the samples tested in this study.

Figure 1.

Details of the workflow for the samples analyzed in the study.

Rice cooking

Raw rice samples were sent to the rice analysis laboratory (RAL) of the BRRI for nutrient analysis of the raw and cooked rice. To minimize potential measurement bias, the samples were assigned letter‐coded sample IDs, and the rice variety was not disclosed to the laboratory personnel. To cook rice with ‘adequate’ water, a rice‐to‐water ratio of 1:1.5 was used as per laboratory protocol, and rice samples were cooked in a rice cooker until all the water was absorbed and no gruel remained. This protocol was developed by adjusting for the differences between cooking rice in a rice cooker and in a conventional cooking pot, as supported by existing literature.^{26, 27} Cooking rice with ‘excess’ water involved a rice‐to‐water ratio of 1:5, in accordance with a previous study ²⁸ and laboratory practice; these samples were cooked in a traditional stainless‐steel rice‐cooking container. After cooking, all excess water (gruel) was drained.

Laboratory testing

The concentrations of Fe and Zn in the rice samples were determined using inductively coupled plasma optical emission spectroscopy (ICP‐OES) with a Shimadzu (Kyoto, Japan) ICP‐OES 9820 instrument. For this, 1.0 g of dried rice powder or 1.0 g of cooked rice was taken from a plastic zip‐lock package and weighed using an analytical balance. Each rice sample was placed in a microwave vessel along with a mixture of 6 mL of nitric acid (HNO₃) and 1 mL of hydrogen peroxide (H₂O₂). The microwave digestion program was conducted using the parameters outlined in Supporting Information, Table S1. When digestion was complete, the contents were transferred into polypropylene vials and diluted to the desired volume with ultrapure water to achieve total digestion. Subsequent analysis was performed using the Shimadzu ICP‐OES 9820 instrument, equipped with a mini plasma torch, concentric nebulizer, and cyclonic chamber. Supporting Information, Table S2 provides the detailed instrument configuration and operating parameters. The concentrated nitric acid (65%) and hydrogen peroxide (30%) were of trace metal grade, and ultrapure water was obtained from a Milli‐Q water‐purification system. External calibration standards were prepared by mixing single‐elemental stocks sourced from Merck and Sigma‐Aldrich (Dramstadt, Germany). These calibration standards were prepared in a 1% HNO₃ solution (Supporting Information, Table S3). Periodic Table Mix 1 for ICP (TraceCERT) certified reference material (CRM) was used as the standard solution. The details of the procedure have been described elsewhere. ¹⁴

For phytic acid analysis, 200 mg of rice powder was extracted with 5% trichloroacetic acid (TCA), incubated at 60 °C, centrifuged, and pooled into a 25 mL volumetric flask. A 20 mL aliquot was reacted with ferric chloride, heated at 95 °C for 45 min, filtered, and treated with potassium thiocyanate. Absorbance was measured at 485 nm using an ultraviolet‐visible (UV‐visible) spectrophotometer (Shimadzu 1900i), and phytic acid was quantified using a standard curve based on the assumption that four ferric ions bind to one phytic acid molecule (Fe₄P₆C₆H₆O₂₄), with results calculated on a dry weight basis. ²⁹ All analyses were performed in triplicate and the average value was reported.

Statistical analysis

The results of the laboratory analysis were entered into an MS Excel file, which was converted to a Stata file for subsequent analysis using Stata 19 (StataCorp). The phytate‐to‐zinc molar ratio (PZMR) was calculated by converting the phytic acid and zinc content to moles, dividing by 660.1 (the molar weight of phytic acid) and 65.4 (the molar weight of zinc), respectively, and then dividing the phytic acid molar content by the zinc molar content. Given the small number of samples for each variety (n = 8), non‐parametric tests were conducted throughout the analysis. Descriptive statistics were reported as medians and interquartile ranges (IQR). Wilcoxon rank‐sum tests were performed to determine the differences in the content of phytate, zinc, iron, and PZMR in raw rice between different groups: ZBR versus CR or husking mill versus ARM. The Skillings–Mack test was performed separately for each nutrient to assess differences between raw and cooked rice across the two cooking methods, accounting for repeated measurements on the same samples. ³⁰ The Skillings–Mack test was preferred over the Friedman test, because it uses weighted ranking and yields higher test statistics and lower P‐values. ³¹ The pairwise Wilcoxon signed‐rank test was conducted as a post hoc test to explore the differences in nutrient content among raw rice, cooked rice with adequate water, and cooked rice with excess water.

RESULTS

Difference between ZBR and CR (in raw and cooked rice)

In the raw rice samples from the ARM, the ZBR contained significantly higher amounts of zinc than the CR variety (P = 0.012). Specifically, the ZBR variety had a median zinc content of 19.7 (IQR 19.3–20.2) mg kg⁻¹, whereas the CR variety had a median zinc content of 18.7 (IQR 18.3–18.7) mg kg⁻¹. Although the phytate levels did not differ, higher zinc levels led to increased bioavailability in the ZBR variety, as indicated by a significantly lower PZMR (P = 0.034). Specifically, the ZBR variety exhibited a median PZMR of 2.9 (IQR 2.7–3.1), whereas the CR variety exhibited a median PZMR of 4.5 (IQR 3.0–6.5). No difference in iron content was observed between the two varieties (Table 1).

Table 1.

Differences in phytate, zinc, iron content, and phytate‐to‐zinc molar ratio (PZMR) in raw rice between zinc‐biofortified rice (ZBR) and conventional rice (CR)

Variable	Raw rice from ARM (n = 16)			Cooked rice with adequate water † (n = 16)			Cooked rice with excess water ‡ (n = 16)
	ZBR (n = 8) Median (IQR)	CR (n = 8) Median (IQR)	P	ZBR (n = 8) Median (IQR)	CR (n = 8) Median (IQR)	P	ZBR (n = 8) Median (IQR)	CR (n = 8) Median (IQR)	P
Phytate (mg kg⁻¹)	589.0 (567.1, 602.8)	847.7 (574.1, 1199.3)	0.401	584.0 (554.7, 592.8)	797.9 (577.1, 1188.3)	0.142	581.2 (568.7, 589.5)	846.9 (570.0, 1213.8)	0.294
Zinc (mg kg⁻¹)	19.7 (19.3, 20.2)	18.7 (18.3, 18.7)	0.012	16.0 (15.3, 16.7)	14.8 (14.1, 15.4)	0.021	11.7 (11.4, 12.2)	10.9 (10.6, 11.6)	0.074
Iron (mg kg⁻¹)	5.0 (4.3, 5.2)	4.6 (4.4, 5.0)	0.674	3.7 (3.6, 3.9)	3.9 (3.7, 4.4)	0.294	3.3 (3.2, 3.7)	3.7 (3.3, 3.9)	0.601
PZMR	2.9 (2.7, 3.1)	4.5 (3.0, 6.5)	0.034	3.5 (3.4, 3.6)	5.3 (3.8, 8.4)	0.059	5.0 (4.8, 5.1)	7.8 (4.9, 10.9)	0.142

ARM, auto rice mill; PZMR, phytate–zinc molar ratio; ZBR, Zinc‐biofortified rice; CR, Conventional rice.

^† Rice‐to‐water ratio 1:1.5.

^‡ Rice‐to‐water ratio 1:5.

In rice cooked with an ‘adequate’ quantity of water, the ZBR contained a significantly higher amount of zinc than the CR variety (P = 0.021). Specifically, the ZBR variety had a median zinc content of 16.0 (IQR: 15.3–16.7) mg kg⁻¹, whereas the CR variety had a median zinc content of 14.8 (IQR: 14.1–15.4) mg kg⁻¹. The other parameters – phytate, iron and bioavailability – did not differ significantly. However, the differences in zinc content between the varieties were absent when rice was cooked with ‘excess’ water (Table 1).

Effect of color sorting

To test whether the color sorting process affected the nutrient content in raw rice, the phytate, zinc, and iron content and the PZMR of raw rice sampled from the husking mill and the ARM were compared. The results showed that all the parameters were unaffected in the ZBR variety; however, the iron content was reduced by the process of color sorting in the CR variety (Supporting Information, Fig. S1).

Effect of cooking

The Skillings–Mack tests showed significant differences in zinc and iron content, and in zinc bioavailability (as indicated by the PZMR), among raw rice, rice cooked with ‘adequate’ amounts of water, and rice cooked with ‘excess’ water (Table 2). The positive value for the zinc content of raw rice (+4.94) indicates a higher zinc level, whereas the strongly negative value for rice cooked with ‘excess’ water reflects substantial zinc loss. The value of 0.0 for rice cooked with ‘adequate’ quantities of water suggests only a partial loss of zinc from the raw rice. In contrast, phytate levels were not significantly different among raw rice, rice cooked with ‘adequate’ quantities of water and rice cooked with ‘excess’ water.

Table 2.

Skillings–Mack test results for phytate, zinc, iron and phytate‐to‐zinc molar ratio (PZMR) in raw, cooked rice with ‘adequate’ water, and cooked rice with ‘excess’ water

Nutrient	WSum/SE			Skillings–Mack value	P
	Raw rice from ARM	‘Adequate’ water †	‘Excess’ water ‡
Phytate (mg kg⁻¹)	+0.93	0.00	−0.92	1.139	0.558
Zn (mg kg⁻¹)	+4.94	0.00	−4.90	30.459	<0.001
Fe (mg kg⁻¹)	+4.94	−0.61	−4.29	28.913	<0.001
PZMR	−4.64	−0.31	+4.90	30.459	<0.001

ARM, auto rice mill; PZMR, phytate‐to‐zinc molar ratio.

^† Rice‐to‐water ratio 1:1.5.

^‡ Rice‐to‐water ratio 1:5.

Post‐hoc comparison using the paired Wilcoxon signed‐rank test showed that there was a significant difference among raw rice, rice cooked with ‘adequate’ quantities of water, and rice cooked with ‘excess’ water in terms of zinc, iron, and zinc bioavailability expressed as PZMR, although no change in phytate was observed (Table 3). For example, compared to raw rice, there was a 19% reduction in zinc content of rice cooked with adequate water (P < 0.001) as the median reduced from 18.77 to 15.24 mg kg⁻¹. When the rice was cooked with excess water, the zinc content reduced even more to 11.43 mg kg⁻¹, resulting in a zinc loss of approximately 40% (P < 0.001). A similar gradient was observed for iron. In contrast, as the zinc content was reduced and the phytate levels were similar, bioavailability of zinc decreased, indicated by an increase in PZMR, following the gradient from raw rice to rice cooked with ‘adequate’ quantities of water, to rice cooked with ‘excess’ water.

Table 3.

Predicted nutrient concentrations and percentage loss in different cooking methods

Nutrient	Cooking method	Median (IQR)	% Loss vs raw
Phytate (mg kg⁻¹)	Raw rice from ARM	596.92 (567.04, 757.52)	Refa
	Adequate water †	586.99 (557.29, 715.43)	1.67%a
	Excess water ‡	583.66 (568.67, 743.18)	2.22%a
Zn (mg kg⁻¹)	Raw rice from ARM	18.77 (18.63, 19.61)	Refa
	Adequate water †	15.24 (14.91, 16.57)	18.81%b
	Excess water ‡	11.43 (10.88, 12.11)	39.08%c
Fe (mg kg⁻¹)	Raw rice from ARM	4.74 (4.38, 5.19)	Refa
	Adequate water †	3.69 (3.63, 4.15)	22.2%b
	Excess water ‡	3.37 (3.24, 3.95)	28.8%c
PZMR	Raw rice from ARM	3.07 (2.83, 4.01)	Refa
	Adequate water †	3.68 (3.38, 4.64)	−19.84%b
	Excess water ‡	5.02 (4.75, 6.60)	−63.59%c

Values with different superscript letters within the same nutrient are significantly different. The value of Raw rice from ARM is considered as Reference for calculating percent reduction in different cooking techniques.

ARM, auto rice mill; PZMR, phytate‐to‐zinc molar ratio.

^† Rice‐to‐water ratio 1:1.5.

^‡ Rice‐to‐water ratio 1:5.

DISCUSSION

This is the first study to report changes in the zinc, phytate, and iron content of ZBR varieties developed in Bangladesh in relation to different cooking methods. It shows that cooking rice with a water ratio of 1:5, and discarding the gruel, resulted in a higher loss of zinc and iron from raw rice than cooking it with a water ratio of 1:1.5. It also changed the zinc bioavailability, although no significant change was observed in the phytate content.

This study was part of a community‐based RCT to assess the efficacy of ZBR in improving zinc status in young NPNL women in rural Bangladesh. The ZBR, BRRI dhan100, has an estimated zinc concentration of 25.7 mg kg⁻¹.^{25, 32} Laboratory analysis of color‐sorted raw rice indicated a significantly higher zinc concentration in the ZBR than in the CR variety. However, the mean zinc concentration of 19.84 mg kg⁻¹ (SD 1.07 mg kg⁻¹) in the ZBR was much lower than the predicted value of 25.7 mg kg⁻¹. Zinc accumulation in ZBR grains depends on several factors, such as soil properties (zinc, pH), fertilizer use and type, and water management.^{33, 34, 35} In this study, zinc sulfate fertilizer was used at a dosage that followed the standard guidelines for the rice variety. ³² Water management practices adhered to the prescribed protocol, although a brief period of inadequate water availability occurred during the reproductive phase of the paddy plants as a result of a heatwave and drought, which was promptly alleviated through irrigation. This may have affected zinc accumulation in the BRRI dhan100 paddy. ³⁶ However, soil zinc and pH data were not available for the cultivation area.

In addition to the agronomic and environmental determinants of zinc accumulation in grains, the analytical methodology can also influence the estimation of zinc content. X‐ray fluorescence (XRF) was used to assess the zinc content of BRRI dhan100 as 25.7 mg kg⁻¹ ³² whereas the present study measured the zinc concentration using ICP‐OES, considered as the gold standard and offering greater precision than other methods. ³⁷ A more recent study that adopted the same technique also reported a similar zinc content (20.1 mg kg⁻¹) in brown rice of the same variety. ¹⁴

Several studies have shown that the cooking process significantly reduces micronutrients, such as zinc and iron, in conventional and fortified raw rice cooked with either adequate ¹⁹ or excess water. ²⁰ This is consistent with the present study, although the reported proportion of loss ranged widely (13% to 64%). Studies that compared different cooking techniques also reported findings similar to the present study, in that increasing the amount of water during cooking and discarding the gruel was associated with increased zinc loss. Shahriar et al.²⁴ reported the effects of different cooking techniques, such as the number of times the rice was washed before cooking, and the different amounts of water used for cooking, on loss of toxic materials and essential micronutrients in conventional rice. They found the highest reduction in zinc (17%) when the rice was washed five times and cooked with a 1:4–1:6 ratio with the gruel being discarded after cooking. ²⁴ Another study on conventional rice conducted by Mwale et al. ²³ found a gradual increase in zinc loss as the amount of water used in cooking increased, with 7.7% in 1:6 and 14.2% in 1:10 water use. Cooking with adequate water also resulted in zinc loss, but it was much smaller. Azam et al.²² reported a similar gradient in their study involving fortified and non‐fortified rice. In fortified rice, the zinc loss ranged from 13.5% to 21.0% when raw rice was cooked with adequate water (in a rice cooker) after washing. In non‐fortified rice, the proportion was similar (13.5% to 26.0%). However, when cooked with excess water following washing, the zinc loss was much higher (58.1% to 65.9% in fortified rice and 27.1% to 42.6% in non‐fortified rice), which is similar to the findings of the present study. ²²

Two studies have examined the cooking loss of zinc in ZBR in Indonesia. Fitri Mahiyudi et al. ³⁸ compared the loss of zinc in conventional, fortified, and biofortified rice cooked using an ‘adequate’ amount of water (1:1.25) and reported that all rice varieties lost zinc in relation to raw rice, and the highest proportion of zinc loss occurred in the biofortified rice (30%), followed by conventional rice (25.7%) and fortified rice (13.8%). However, Stephanie et al. ³⁹ compared the effect of different cooking methods (rice pot and rice cooker adopting a 1:1.25 ratio, and steamer adopting a 1:2 ratio) on zinc loss in conventional and biofortified rice. They reported a higher loss in the steamer (52.6% in biofortified rice and 75.41% in conventional rice) than in the rice pot (38.39% in biofortified rice and 66.10% in conventional rice) or rice cooker (33.42% in biofortified rice and 65.87% in conventional rice). However, the differences in zinc content among the different cooking methods were not statistically significant within or across rice varieties.

Reviewing the existing literature indicates that there is significant variability in the estimates of the proportion of zinc loss due to cooking. However, a general trend was observed whereby an increase in the amount of water used for cooking resulted in increased zinc loss. This was also demonstrated in the present study. During cooking, the zinc and other micronutrients present in the rice grain leach out into the water; therefore, cooking with excess water and discarding the gruel results in increased loss. ²² It was also observed that, compared with fortified or conventional rice, the zinc loss proportion was much higher in biofortified rice. Fortification of paddy in these studies was done using zinc‐enriched solution during soaking or parboiling, which results in the accumulation of zinc in the endosperm layer, bound to the gelatin layer, ⁴⁰ which resists leaking into water due to heating during cooking. On the other hand, biofortification involves the accumulation of zinc during the growing phase, where most of the zinc is accumulated in the outer layer, which is also rich in phytate. ⁴¹ During cooking, phytate breaks down and releases zinc, and a higher proportion of zinc is lost if rice is cooked with excess water that is to be discarded as gruel.

Phytate is an important anti‐nutrient that has been discussed in relation to zinc nutrition. Plant‐based foods are typically rich in phytate, with large amounts present in whole grain cereals. ⁵ Cooking usually results in the loss of phytate, as it breaks down at high temperatures. ⁴² A similar trend was observed in the present study, as a gradual decrease in phytate content was observed when a greater amount of water was used for cooking, although the difference was not statistically significant. Although no change in phytate was found, a decrease in zinc content as a result of cooking resulted in changes in zinc bioavailability from ZBR. The results indicated that PZMR changed from 3.68 in rice cooked with adequate water to 5.02 in rice cooked with ‘excess’ water. This translates to a decrease in bioavailability from 50% to 30%, ⁴³ indicating that cooking ZBR with ‘excess’ water not only decreased its zinc content but also decreased the amount of zinc absorbed from it.

In addition to zinc and phytate, the iron content of the rice varieties in the study was analyzed as a secondary objective. There was no difference in the iron content of raw rice between varieties. However, the iron content was reduced significantly from raw rice during cooking, and the loss was significantly higher when rice was cooked with ‘excess’ water. This was attributed to the mechanism described above for zinc and this is supported by the existing literature.^{19, 21, 22, 23} A recently published study reported an increase in iron content after cooking rice; this was attributed to groundwater iron. ⁴⁴ In contrast with iron, levels of zinc in groundwater in Bangladesh are negligible. ⁴⁵

Zinc deficiency is one of the major nutritional problems in LMICs, and ZBR is a potentially sustainable solution. As ZBR contains 16.10 mg of zinc per kg of cooked rice (in an ‘adequate’ amount of water), and average raw rice consumption in Bangladesh is 329 g, ⁴⁶ this would translate to consumption of 15.9 mg of zinc from rice cooked with ‘adequate’ water. Considering the PZMR of rice cooked with ‘adequate’ quantities of water, the bioavailability will be around 50%, which translates to approximately 8 mg of absorbed zinc. ⁴³ This would serve as 72% of the RDA for an adult male and 99% for an adult female. ⁴⁷ However, if rice is cooked with ‘excess’ water, the bioavailability decreases to 30% ⁴³ and the proportion of RDA fulfilled would drop to 30% for an adult male and 41% for an adult female. ⁴⁷ Given the importance of the amount of cooking water in retaining important micronutrients, policymakers may consider promoting specific cooking practices as a strategy to reduce micronutrient losses at the household level by integrating the recommendations into dietary guidelines and nutrition education.

The study adopted non‐commercial processing of paddy to minimize micronutrient loss during the process, as suggested by a previous study. ¹⁷ However, the resultant raw rice contained a considerable portion of black and dead rice, which would render the product unacceptable. It was therefore decided to carry out color sorting, without any polishing, in an auto rice mill, to improve the quality of the product. The study explored whether there was any effect of color sorting on the concentration of zinc, iron, and phytate in rice of both varieties. The zinc and phytate concentrations did not differ between husking mill rice and auto rice mill rice in either variety. However, the iron content of the CR decreased, although the statistical difference was marginal (P = 0.048) (Supporting Information, Fig. S1). The reason for this could be that the process of color sorting involved a step in which the raw rice was shaken to remove dust, which could have also removed some of the bran on the surface of the grains. It is well known that iron concentration is much higher in bran than in other parts of the grain, and the iron content in the bran is seven times higher than that of the hull and endosperm, compared with zinc, which is only three times higher. ⁴⁸ This also has policy implications, as our research shows that the combination of non‐commercial post‐harvest processing and color sorting can lead to micronutrient retention as well as consumer satisfaction.

One of the limitations of this study is the small sample size (n = 4) for the CR raw rice variety collected from husking mill, which might have affected iron content comparisons (see above). Another limitation is that, when cooking rice, standard laboratory procedures were followed to ensure methodological rigor. These might not reflect the traditional cooking practices in rural households. Despite these limitations, this study provides valuable early evidence of the effects of different cooking methods on zinc loss in ZBR in Bangladesh. Moreover, rice samples were obtained through real‐time, non‐commercial post‐harvest processing, unlike many existing studies, which processed rice under standard laboratory conditions. This approach provides a more realistic household scenario and a more accurate representation of zinc loss and bioavailability from biofortified rice at the point of consumption. The results of this study will therefore be valuable in estimating zinc intake from interventions with greater precision.

CONCLUSION

This study provided valuable insights into the loss of zinc in ZBR due to cooking, and the effect of the amount of water used for cooking on this loss. Consistent with studies conducted on fortified and conventional rice, the use of ‘excess’ amounts of water was found to be associated with an increased loss of zinc and iron, and with lower zinc bioavailability. These results can be incorporated into policies leading to changes in cooking practices for better micronutrient retention. They will also be useful for estimating the actual zinc intake from interventions such as ZBR and interpreting treatment effects in future research.

FUNDING INFORMATION

The study was funded by the Fondation Nestlé as part of a community‐based randomized controlled trial to test the efficacy of zinc‐biofortified rice in improving zinc status in young non‐pregnant, non‐lactating women in rural Bangladesh. The donor had no role in the design, implementation, analysis, or interpretation of this study.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

AUTHOR CONTRIBUTIONS

Prasenjit Mondal, Sabuktagin Rahman, and Faruk Ahmed conceptualized the study. Prasenjit Mondal, Sabuktagin Rahman, Faruk Ahmed, Malay Kanti Mridha, and Hai N Phung supervised the rice production process and sample collection. Abdullah Al Mahim was involved in the post‐harvest processing of paddy varieties. Habibul Bari Shozib and Md. Mariful Islam were involved in the laboratory analysis of the samples. Prasenjit Mondal and Melaku Engidaw conducted the data analysis and drafted the manuscript. All authors reviewed the draft and provided input while preparing the manuscript.

ETHICS STATEMENT

Ethical approval for this study was obtained from Griffith University (approval number 2023/611) and from the Institutional Review Board of BRAC James P Grant School of Public Health, BRAC University, Bangladesh (approval number IRB‐2024‐IS‐08).

Supporting information

Figure S1. Effect of color sorting on the concentration of phytate, zinc, iron, and Phytate zinc molar ratio in Zinc‐biofortified rice and Conventional rice varieties.

Table S1. Microwave digestion parameters for processing rice sample

Table S2. Instrument configuration and ICP‐OES parameters

Table S3. Calibration standard concentration (mg kg⁻¹).

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.