Optimization of double-cooking condition for low potassium potatoes using response surface methodology (RSM)

Abstract

This study aimed to establish optimal double-cooking condition using response surface methodology that maintained hardness while maximizing potassium reduction rate. The experimental design was based on the first cooking time (4.5–5.5 min) and rinsing time (20–60 s) through central composite design. This study suggested an optimal double-cooking condition of 5.5 min for first cooking and 57.57 s for rinsing. The model corroborated that the double-cooking condition significantly influenced dependent variables, including potassium reduction rate, hardness, and color (b-value). As the first cooking time increased, the potassium reduction rate increased and the hardness and b-value decreased. SEM revealed that double-cooked potato had more organized and netted structure. This structure could be helpful to maintain hardness, but relatively large amount of potassium could be leached out. The established optimal double-cooking condition for potatoes holds promise for broadening the dietary options for chronic kidney disease patients.

Introduction

Individuals suffering from chronic kidney disease (CKD) need to be cautious regarding potato consumption due to the potassium content, which can lead to hyperkalemia. Recent reports indicate that 10% of dialysis patients and 16% of non-dialysis CKD patients experience hyperkalemia (Vega et al., 2019). In case of CKD patients, declining kidney function gradually impairs the physiological mechanisms responsible for potassium balance, contributing to hyperkalemia (DuBose Jr, 2017). Hence, it is vital to eliminate potassium in potatoes before consumption for CKD patients’ diets.

Potatoes are one of the most widely consumed tubers in the world. Globally, potatoes boast a remarkable annual production of 350 million tons, showcasing a high yield per unit area (Choi et al., 2008). In South Korea, the primary cultivar is the Sumi potato, which introduced from the United States, and the potato production in Korea was 303,243 tons in 2022. Generally, comprising with 79.5% water, 21.5% carbohydrates, 2.5% protein, and 1.5% other components, potatoes stand as a significant source of diverse nutrients (Burlingame et al., 2009). They are rich in dietary fiber, minerals like potassium and phosphorus, as well as vitamin C. Furthermore, potatoes are recognized to contain many phytochemicals such as polyphenols and carotenoids (Burrowes and Ramer, 2008).

Potatoes usually undergo various cooking processes, including boiling, steaming, frying and roasting. The microstructure of the potato and the properties of the cell wall polymers can be affected during processing and contributes to the different texture of the potato after cooking (Bordoloi et al., 2012). Therefore, to comprehend this texture changes, it is essential to observe microstructural features such as cell shape and size and cell wall properties. To determine microstructure and texture, previous studies have used microscopes such as confocal scanning laser microscopy (CSLM) and scanning electron microscopy (SEM), as well as texture analyzers (Singh et al., 2008).

Previous studies have investigated the effects of different cooking methods on diverse minerals, vitamins, and phytochemicals in foods (Bethke and Jansky, 2008; Martínez-Pineda et al., 2020; Tian et al., 2016a; Vrdoljak et al., 2015). Retaining minerals was more successful in microwaving, grilling and baking, whereas the addition of water during cooking led to significant mineral reduction due to leaching (Tian et al., 2016a). A 50% decrease in potassium content when potatoes were boiled (Bethke and Jansky, 2008). A recent study explored various cooking methods-soaking, single-cooking, and frying-using fresh, canned, and frozen fried potatoes. The findings revealed a potential 70% reduction in potassium content after boiling followed by a brief soaking, while the soaking process alone didn’t adequately reduce potassium levels (Martínez-Pineda et al., 2020). Boiling in water or cooking in oil of potatoes, carrots, and green beans didn’t significant impact protein content but led to an average mineral reduction of about 30–40% (Vrdoljak et al., 2015).

In case of potatoes, double-cooking method is considered the most effective way to reduce potassium from potato tissues Bower (1989). This method entails discarding the boiling water, replacing it with room-temperature water, and then boiling the potatoes again. The loss of potassium content after following this double-cooking process revealed a significant reduction in potassium content. Specifically, when thinly sliced potatoes underwent double-cooking without prior soaking, there was an approximately 72% reduction in potassium content. However, detailed examination of the specific conditions of double-cooking and its impact on physicochemical characteristics remains unexplored. Therefore, this study investigated the effects of varying double-cooking conditions on potassium reduction rates, texture, and color in potatoes. The primary goal was to ascertain the most effective double-cooking condition that not only maximize potassium reduction but also ensure practicality in processing using response surface methodology (RSM). Additionally, to confirm the impact of each stage of the double-cooking process, the microstructure of the potatoes was determined.

Materials and methods

Materials and sample preparation

Fresh Sumi potatoes harvested in 2022 from Gangwon, Republic of Korea were used for the study. Initially, the fresh tubers underwent a series of following preparation steps using distilled water: washing, peeling, rinsing, and draining. For the actual cooking process, the peeled potatoes were cut into 1.5 cm cubes with a vegetable cutter. Following the cutting phase, the cubes were washed to eliminate surface starch and then soaked in room temperature water to prevent browning. The experimental procedure was executed using an induction cooker (Kitchen Flower, KEP-IH5000SKCP, Republic of Korea) by adding water at a 10:1 water-to-sample ratio to 50 g of the sample (Suzuki et al., 1999). Once the water reached boiling point, the potatoes underwent a sequence of cooking steps: initial cooking, followed by rinsing with room temperature water, and then a final cooking phase in boiling water (Burrowes and Ramer, 2008). At the conclusion of the cooking process, the potato cubes were promptly rinsed in room temperature water for 15 s (Oh, 1996) and subsequently drained on a sieve for 30 min. After draining, potassium content, color, and texture of samples were determined.

Experimental design

From preliminary experiments, a total cooking time of 6 min was established, which maintained the proper hardness level. Extended cooking beyond 6 min resulted in surface crumbliness and overcooking. Moreover, a longer initial cooking time highly correlated with a higher potassium reduction rate and second cooking time did not greatly affect the potassium reduction rate. Therefore, the experimental setup involved a factorial design integrating two independent variables (first cooking time and rinsing time) and five dependent variables (potassium reduction rate, hardness, L*, a*, and b* values). The experimental design employed in this study was based on variations in the first cooking time and rinsing time using central composite design (CCD) within response surface methodology (RSM) facilitated by Design Expert 13 (Stat-Ease, USA). To validate reproducibility, three replicates were performed at the central point, culminating in a total of 11 experiments. Each variable was assessed across three coded levels ranging from − 1 to 1, alongside − α and + α (α = 1.414), with corresponding coded and original levels were represented in Table 1. The range for the first cooking time spanned 4.5 to 5.5 min and the rinsing time varied from 20 to 60 s. The dependent variables were potassium reduction rate (Y₁), hardness (Y₂), and b* (Y₃). The obtained experimental data were analyzed through a following second-order model equation using RSM:

$$Y={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+{\beta }_{11}{X}_1^2+{\beta }_{22}22X_2^2+{\beta }_{12}{X}_1{X}_2+\epsilon$$

where, Y is the response, ${\beta }_0$ is a constant, ${\beta }_1$, and ${\beta }_2$ are linear regression, ${\beta }_{11}$, and ${\beta }_{22}$ are quadratic regression, and ${\beta }_{12}$ is interaction regression and $X_1$, and $X_2$ are the independent variables, the first cooking time and the rinsing time, respectively.

Table 1.

Central composite design used to evaluate the effects of first cooking time (t₁) and rinsing time (t₂) on the potassium reduction rate, hardness and color of potato

Runs	Independent variables
	Coded levels		Original levels
	t1	t2	t1 (min)	t2 (s)
1	0	0	5	40
2	0	1.414	5	68
3	0	0	5	40
4	0	− 1.414	5	12
5	− 1	1	4.5	60
6	0	0	5	40
7	− 1.414	0	4.3	40
8	1	1	5.5	60
9	1.414	0	5.7	40
10	1	− 1	5.5	20
11	− 1	− 1	4.5	20

Potassium content analysis

Potassium content was analyzed using Inductively Coupled Plasma spectrometer (Direct Reading Echelle ICP, Leeman, Mason, OH, USA). The specific wavelength used for potassium determination was 766.490 nm. Sample was prepared using the dry ashing digestion method prescribed by the Food Code (Ministry of Food and Drug Safety, South Korea). This method entails subjecting the samples to an electric furnace to incinerate the organic matrix, followed by dissolution with hydrochloric acid to eliminate it.

Texture analysis

Hardness was determined using a texture analyzer (CTX, AMETEK, Brookfield, Toronto, Canada). Each raw or cooked potato cube underwent compression using a 25.4 mm diameter cylinder probe, employing a 5 kg load cell. The trigger load was set at 0.5 g. Pre-test speed, test speed, and post-test speed were 0.5, 0.5, and 1 mm/s respectively. Based on preliminary tests, the maximum allowable deformation was established at 50% of the original height. Hardness was quantified by identifying the peak force recorded during the compression test. All tests were performed in triplicate to ensure accuracy and reliability of results.

Color analysis

The color assessment of both raw and cooked potato samples was conducted using a colorimeter (CR-20; Konica Minolta, Tokyo, Japan). The CIE Lab system was employed with an illuminant of D65 standard and 10° observer angle. The L value represents brightness, a value denotes the presence of redness or greenness, and b value signifies yellowness or blueness. Color measurements were obtained from three different potato cubes and performed in triplicate to ensure consistency and reliability of data.

Microstructure

The internal changes in raw and cooked potatoes, including alterations during the double-cooking process, were examined using a Scanning electron microscope (TM3000, Hitachi, Tokyo, Japan). For analysis, five potato slices (15 × 15 × 1 mm³) were collected at different stages of optimal double-cooking condition timeline; (1) raw, (2) after the first cooking, (3) after rinsing, (4) after the second cooking, (5) single cooking for 6 min. These slices were lyophilized using freeze dryer (PVTFD10R, Ilshin Bio Base, Yangju, Korea) at − 70 °C for 48 h. Microstructural images of each samples were captured at 15 kV with magnifications set at 200× and 500× (Zhang et al., 2021).

Optimization and verification

The regression analysis and analysis of variance (ANOVA) were carried out to fit the model and ascertain the statistical significance of the model terms. The RSM software generated 3D response surface plots and numerical optimization data. The effects of independent variables on dependent variables, along with the statistical significance of linear, quadratic, and interactions of each factor on the responses, were determined at 99% confidence level. Any statistically insignificant dependent variable was eliminated from the model. Each trial was conducted in triplicate to ensure reliability. Verification of the optimum double-cooking condition was executed. Potato samples subjected to the identified optimal first cooking and rinsing times were experimentally analyzed, and the expected values were statistically compared to the experimentally verified values (Gan et al., 2007).

Results and discussion

Experimental data

Potassium reduction rate, hardness, and color of double-cooked potatoes with varied first cooking and rinsing times are shown in Table 2. Notably, the potassium reduction rates of double-cooked potatoes exceeded those of the control (single-cooking for 6 min) across all conditions. This confirms the efficacy of the double-cooking method in potassium leaching in potato. This method appears simpler and more effective in leaching potassium compared to single-cooking, bypassing the need for overnight presoaking or intricate treatments (Burrowes and Ramer, 2008).

Table 2.

Experimental results of response variables for double-cooked potatoes

Runs	Cooking conditions	Reduction rate (%)	Hardness (g)	L*	a*	b*
	Control (6 min)	38.84 ± 0.66g*	1397.10 ± 224.00ab	63.64 ± 2.93a	− 4.58 ± 0.06abc	1.22 ± 0.76abc
1	5.0 min/40 s	44.00 ± 0.22d	1766.87 ± 200.82a	63.85 ± 1.99a	− 4.76 ± 0.11cde	2.00 ± 0.81abc
2	5.0 min/68 s	48.25 ± 0.44c	1187.33 ± 87.19b	64.22 ± 0.99a	− 4.89 ± 0.18e	2.60 ± 1.49a
3	5.0 min/40 s	47.93 ± 0.18c	1731.47 ± 109.19ab	62.74 ± 3.17a	− 4.81 ± 0.10de	1.64 ± 0.72abc
4	5.0 min/12 s	42.85 ± 0.46e	1549.53 ± 72.23ab	65.22 ± 1.49a	− 4.83 ± 0.13de	1.26 ± 0.27abc
5	4.5 min/60 s	40.95 ± 0.07f	1453.30 ± 197.01ab	63.99 ± 2.36a	− 4.67 ± 0.05bcd	2.19 ± 0.16abc
6	5.0 min/40 s	43.88 ± 0.36d	1872.83 ± 218.37a	65.11 ± 3.15a	− 4.52 ± 0.08ab	1.85 ± 0.91abc
7	4.3 min/40 s	41.27 ± 0.20f	1830.67 ± 248.95a	63.07 ± 3.48a	− 4.48 ± 0.20a	2.43 ± 1.35ab
8	5.5 min/60 s	59.31 ± 0.06a	1457.87 ± 232.21ab	62.44 ± 3.41a	− 4.72 ± 0.15cde	1.75 ± 0.80abc
9	5.7 min/40 s	52.98 ± 0.32b	1398.63 ± 149.55ab	62.82 ± 3.59a	− 4.62 ± 0.12abc	1.04 ± 0.75bc
10	5.5 min/20 s	40.46 ± 0.23f	1663.03 ± 154.10ab	63.99 ± 2.70a	− 4.68 ± 0.08bcd	0.82 ± 1.21c
11	4.5 min/20 s	48.03 ± 0.47c	1729.17 ± 255.88ab	64.48 ± 1.65a	− 4.81 ± 0.10de	1.43 ± 0.94abc

*Same letters in the same column are not significantly different at p < 0.05

L*, a*, b* are the color expresion

The potassium reduction rate increased, while hardness and yellowness decreased with prolonged first cooking time. The highest potassium reduction rate was observed at 5.5 min of first cooking time coupled with 60 s of rinsing time. During double-cooking, the first cooking in boiling water triggers destruction of the potato tissue, facilitating the elution of potassium present within the tissue, ultimately resulting in reduced potassium content. Hardness ranged from 1187.33 to 1872.83 g, with the lowest hardness was observed at 5.0 min of first cooking time and 68 s of rinsing time. After cooking, the decline in potato hardness can be attributed to various factors such as starch gelatinization, alterations in turgor pressure, changes in cell size and cell wall structure, and the breakdown of the middle lamella. Interestingly, L value (brightness) and a value (redness) exhibited no significant changes with cooking. However, b value (yellowness) decreased with increasing the first cooking time and decreasing the rinsing time. This indicates a progressive fading of yellowness as the duration of the first cooking time increased.

Model fitting of dependent variables

The second-order polynomial response surface model was applied to represent an equation for each variable. To validate the model’s fit, regression analysis and ANOVA were conducted. The p-value, lack of fit, and R² results for each response variable are presented in Table 3. The 2FI model for potassium reduction rate, quadratic model for hardness, and linear model for b value were selected. However, the p-values and R² values of L value and a value were 0.2754–0.2862 and 0.0993–0.1997, respectively, indicating that their models are not significant. The lack of fit test checks whether the model is unable to represent data from points not included in the experimental conditions (Varnalis et al., 2004). Similarly, the lack of fit tests for L value and a value revealed the lower F-values less than 0.05, confirming that these models did not adequately represent the predicted results. This suggests the existence of other factors influencing these responses (Le and Jittanit, 2015), leading to their exclusion from the optimization model. Conversely, the p-values for potassium reduction rate, hardness, and b value were below 0.05 (ranging from 0.0003 to 0.0133) and their R² values ranged from 0.8642 to 0.9246, indicating the suitability of these models. These results showed a high proportion of variability in each response can be explained by the data, affirming the accuracy of these models (Nath and Chattopadhyay, 2008).

Table 3.

Model fitting of double-cooking from the results of 3 response variables

Response	Model	P-value	Lack of fit	R2	Equation in terms of actual factors (t1: first cooking time, t2: rinsing time)
Reduction rate	2FI	0.0003	0.7356	0.9246	$Y_1=136.51-19.01t_1-3.12t_2+0.65t_1{t}_2$
Hardness	Quadratic	0.0372	0.2383	0.8546	$Y_2=-4132.57+2400.92t_1+23.25t_2+1.77t_1{t}_2-263.99t_1^2-0.48t_2^2$
b*	Linear	0.0003	0.3901	0.8642	$Y_3=4.61-0.76t_1+0.02t_2$

L*, a*, b* are the color expresion

Potassium reduction rate

Effect of the double-cooking conditions on the potassium reduction rate are shown in Table 2. The potassium reduction rate of single-cooked control sample was 38.84% and those of double-cooked samples were ranged from 40.46 to 59.31%. The response surface model equation for potassium reduction rate is outlined in Table 3. ANOVA results showed that the impact of first cooking time (p ≤ 0.01) was greater than that of rinsing time (p ≤ 0.01) at linear level. Additionally, the interaction term of the variables was also found to have a significant effect (p ≤ 0.001). The model equation for the potassium reduction rate was found to fit well, explaining 93% of the variability.

Earlier studies have noted that potatoes exhibit increased potassium leaching when their skin is removed, accompanied by higher cooking water volume and longer cooking durations (Jin et al., 2016). Moreover, a more pronounced decrease in potassium content was observed with the double-cooking method (Burrowes and Ramer, 2008). It’s reasonable to assume that changing the water during the double-cooking process alters potassium concentration, facilitating greater leaching from the inside of potato. Therefore, more potassium would have been leached from the inside of the potato. Also, cooked and cooled potatoes became more permeable to moisture, allowing the replaced water to penetrate and dissolve potassium more effectively (Kaur et al., 2002).

Beyond potassium reduction via double-cooking, potatoes retain potential for inclusion in the diets of individuals with kidney disease owing to the bioavailability of potassium. Heat treatment disrupts cell walls and can lead to complex formations with other components, affecting the bioavailability of potassium (Naismith and Braschi, 2008). Reported percentages of bioavailable potassium range between 65 and 70% in most fruits and vegetables (Ceccanti et al., 2022).

Hardness

Effects of the double-cooking conditions on hardness of potatoes are presented in Table 2. Single cooked control sample exhibited a hardness of 1397.10 g, while, those of double-cooked samples were ranged from 1187.33 to 1872.83 g. The response surface model equation for hardness is shown in Table 3. ANOVA results showed the significance of the effect of rinsing time (p ≤ 0.05) and quadratic effect of rinsing time (p ≤ 0.05) on hardness. The model equation for hardness was found to be a good fit, explaining 85% of the variability. Raw potatoes commonly exhibit cell rupture and separation during cooking due to the destabilization of pectin material under heat treatment (Singh et al., 2008). Moreover, textural changes in cooked potatoes are attributed to starch gelatinization and retrogradation (Kaur et al., 2002). Multiple factors, including starch content, degree of starch swelling and gelatinization, pectin degradation, breakdown of the cell wall and cell separation can influence potato texture (Singh et al., 2008). Consequently, after cooking, hardness significantly decreases due to these changes.

The differences in hardness between single-cooked and double-cooked potatoes are possibly due to the presence of the cooling process. Cooling by an intermediate rinsing procedure promotes retrogradation of starch and removes the leached free starch, thereby lowering starch content. Starch retrogradation diminishes turgor pressure, allowing cells to round-off during subsequent cooking (Lamberti et al., 2004). Hardness elevation occurs through the gradual formation of a gel structure during cooling due to short-term starch retrogradation, i.e., retrogradation of amylose. This structure might stem from leached amylose during the first cooking, contributing to the increased hardness observed in most double-cooked potatoes (Jankowski, 1992).

Color

Effects of the double-cooking conditions on the color of potatoes are shown in Table 2. Since the models were deemed nonsignificant in case of the L value and a value, only the b value revealed the significant effect and was ranged from 0.82 to 2.60. The response surface model equation for b value is presented in Table 3. ANOVA results emphasized that the effect of rinsing time (p ≤ 0.001) held greater significance than the effect of first cooking time (p ≤ 0.01) at the linear level. The model equation for the b value was found to be significant, explaining 86% of the variability. The faded yellowness is attributed to the inactivation of browning enzymes, such as polyphenol oxidase. During the blanching process, color could be influenced by enzyme inactivation and adjustment in the reducing sugar content of the surface (Andersson et al., 1994). Although b value had a significant difference depending on cooking conditions, the color values of the potato samples appeared generally similar with slight variations. Visual inspection did not reveal significant color differences based on cooking conditions. Therefore, the impact of the color variables was not considered in the subsequent optimization process.

Consequently, both first cooking time and rinsing time significantly affect the potassium reduction rate, hardness and b value of potatoes. Increase in the first cooking time resulted in decreased potassium reduction rate and b value but decreased hardness. Conversely, increase in the rinsing time resulted in decreased potassium reduction rate but increased hardness and b value. Double-cooking method reduced potassium content to a greater extent and increased hardness of potatoes compared to single-cooking method (Table 4).

Table 4.

Analysis of variance for 5 response variables

Variables/factors	F-values
	Reduction rate	Hardness	L*	a*	b*
Model	27.90***	5.88*	1.76	1.70	25.45***
X1	25.31**	4.30	1.12	0.14	21.06**
X2	12.80**	9.45*	2.31	0.00	29.84***
X1X2	45.6***	0.10	0.43	0.65
$X_1^2$		1.84	2.26	2.21
$X_2^2$		15.48*	1.30	3.35

*Significant at p < 0.05 level

**Significant at p < 0.01 level

***Significant at p < 0.001 level

Optimization and validation

Response surface plots for potassium reduction, hardness, and b value are shown in Figs. 1, 2 and 3, respectively. In these plots, red indicates higher values and a blue indicates lower values. The results of numerical optimization process according to the RSM program are presented in Table 5. The determination of optimal conditions aimed to achieve the highest overall desirability. Additionally, optimization was guided by maximum potassium reduction and an acceptable lower limit of hardness, in which hardness value was comparable to that of 6 min cooked sample. Consequently, the optimal double-cooking condition was 5.5 min for the first cooking time, 57.57 s for the rinsing time and 0.5 min for the second cooking time, respectively. These optimal conditions highlight the desirability of longer first cooking and rinsing times. Predicted values and experimental values under these optimal conditions were 57.62 and 56.26% of potassium reduction, 1377.10 and 1453.77 g of hardness, and 1.75 and 1.98 of b value, respectively. The overall desirability was 0.91 (Table 6).

Fig. 1.

Response surface plot of the effects of first cooking and rinsing times on potassium reduction rate

Fig. 2.

Response surface plot of the effects of first cooking and rinsing times on hardness

Fig. 3.

Response surface plot of the effects of first cooking and rinsing time on b value

Table 5.

Optimal double-cooking condition of potato and its predicted values

First cooking time (min)	Rinsing time (s)	Reduction rate (%)	Hardness (g)	b*	Desirability
5.50	57.57	57.64	1397.10	1.75	0.91

L*, a*, b* is the color expresion

Table 6.

Comparison between predicted values and experimental values

Responses	Values
	Predicted	Experimental
Reduction rate (%)	57.64	56.26 ± 0.28
Hardness (g)	1397.10	1453.77 ± 158.03
b*	1.75	1.98 ± 0.93

L*, a*, b* is the color expresion

Microstructure

To examine the effect of the first cooking, rinsing and the second cooking during the double-cooking on microstructure of potato tissue, scanning electron microscopy (SEM) was employed (Fig. 4). The microstructure of potato tissue and the properties of cell wall polymers as well as cell arrangement greatly influenced by the double-cooking. In the raw potato, the cell wall outline was well preserved, accompanied by numerous starch granules. Each cell exhibited a mix of small mature and many immature starch granules (Singh et al., 2005). After the first cooking, some starch granules remained, alongside observable destruction of certain cell walls and an emptied internal structure. After rinsing, the surface shrunk slightly and some wrinkles were formed due to the alignment of amylose and amylopectin during cooling through the retrogradation process (Wang and Copeland, 2013). After the second cooking, the aligned structure was disrupted, and some netted and filamentous structure emerged. In contrast, the single-cooked control sample without rinsing showed significant destruction and fusion of internal structures, leading to a substantial decrease in hardness compared to the double-cooked potato. It has been reported that single-cooking of potatoes induced starch swelling and gelatinization, as well as disruption of the cell walls and middle lamella (Tian et al., 2016b). Consequently, double-cooked potato showed the development of more netted structures while maintaining cell integrity, resulting in sustained hardness and increased leaching of intracellularly bound potassium. This phenomenon suggests that double-cooked potatoes achieve a higher potassium reduction rate owing to the formation of a netted structure facilitating improved water penetration into the cellular interior.

Fig. 4.

SEM micrographs of potato slice at various steps; A raw; B after the first cooking (5.5 min); C after the rinsing (58 s); D after the second cooking (0.5 min); E control (6 min single-cooking)

Response Surface Methodology (RSM) was employed to optimize double-cooking conditions for low-potassium potatoes while ensuring suitable quality, such as hardness and color. It was observed that the potassium reduction rate, hardness, and b value were significantly reliant on the independent variables. Through the RSM, an optimal double-cooking condition comprising 5.5 min of the first cooking, 57.57 s of the rinsing, and 0.5 min of the second cooking was proposed, and hardness was maintained while maximizing potassium reduction rate in this condition. The RSM model validated the influence of the double-cooking process on dependent variables such as potassium reduction rate, hardness, and b value. An increase in the first cooking and rinsing times correlated with increase in potassium reduction rate and decrease in hardness. Additionally, an increase in the first cooking time coupled with a decrease in rinsing time resulted in reduced b value. SEM results confirmed that double-cooked potatoes had a more aligned and netted structure compared to single-cooked potato, enabling significant potassium leaching while maintaining its hardness. Providing the low-potassium potatoes with proper texture could be beneficial for patients with Chronic Kidney Disease (CKD), who often encounter symptoms like anorexia and nutritional deficiencies, diversifying their diet to potentially alleviate these issues.

Declarations

Conflict of interest

The authors have declared no conflict of interest.