SEARCH

SEARCH BY CITATION

Keywords:

  • acrylamide;
  • free asparagines;
  • reducing sugars;
  • sweet potato chips;
  • vegetable oils

Abstract

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

The objective of this study was to evaluate the precursors of acrylamide formation in sweet potato (SP) (Ipomoea batatas L. Lam) chips and to determine the effect of different types of vegetable oils (VOs), that is, palm olein, coconut oil, canola oil, and soya bean oil, on acrylamide formation. The reducing sugars and amino acids in the SP slices were analyzed, and the acrylamide concentrations of SP chips were measured. SP chips that were fried in a lower degree of unsaturation oils contained a lower acrylamide concentration (1443 μg/kg), whereas those fried with higher degree of unsaturated oils contained a higher acrylamide concentration (2019 μg/kg). SP roots were found to contain acrylamide precursors, that is, 4.17 mg/g glucose and 5.05 mg/g fructose, and 1.63 mg/g free asparagine. The type of VO and condition used for frying, significantly influenced acrylamide formation. This study clearly indicates that the contribution of lipids in the formation of acrylamide should not be neglected.

Practical Application

This study found that different properties of oil do contribute to significant difference of acrylamide formation in SP chips. Food industries that produce deep fried carbohydrate-rich food should consider using frying medium which produces lower acrylamide concentration, besides its nutrition values.


Introduction

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

Sweet potato (SP) roots are processed into French fries or chips by deep fat frying the SP slices at a temperature between 150 and 200 °C (Farinu and Baik 2007). The presence of acrylamide can be detected mostly in fried, deep-fat fried, roasted or oven-cooked carbohydrate-rich foods (Matthäus 2009). Sweet potatoes are another source of carbohydrate-rich food besides normal potatoes (Solanum tuberosum). Acrylamide formation, particularly during the heating of potato roots, has been examined in detail in many countries (Mulla and others 2011; Sanny and others 2012). Although Komthong and others (2012) measured acrylamide concentrations as high as 366 μg/kg in SP chips in a Thai market, acrylamide precursor, and temperature and period of frying were not reported because their report was based on a monitoring study.

The major mechanistic pathway established for acrylamide formation in foods is the Maillard reaction, with free asparagine as the main precursor (Mottram and others 2002; Stadler and others 2002). Acrylamide can also be formed by the deamination of 3-aminopropionamide (Granvogl and Schieberle 2006; Zamora and others 2009). In addition to these pathways, acrylamide can be formed via acrolein and acrylic acid (Yasuhara and others 2003). Various researchers (Yasuhara and others 2003; Ehling and others 2005; Matthäus 2009) concluded that lipid degradation products with 3 carbon units, that is, acrolein and acrylic acid, can generate acrylamide in the existence of asparagine without reducing sugar (RS). Although Weisshaar (2004) emphasised that alternative routes of formation via acrolein and acrylic are far less important than the asparagine pathway; recent studies (Zamora and Hidalgo 2008; Capuano and others 2010) have indicated that lipid oxidation positively influenced the formation of acrylamide. However, other researchers did not discover any significant negative effect of the oil type on acrylamide formation (Matthäus and others 2004; Mestdagh and others 2005; Matthäus 2009). To date, there is still some confusion and misunderstanding regarding the influence of the heating medium on acrylamide formation.

The objective of this study was to evaluate the precursors of acrylamide formation in SP chips, while determined the degree of unsaturation of vegetable oils (VOs) and its oxidation impact during consecutive frying on acrylamide formation. Four VOs were selected for the study, that is, palm olein (PO), coconut oil (COO), canola oil (CAO), and soya bean oil (SBO). The chemicals composition, RS and amino acids (AA) were analyzed before the SP slices were prepared for frying at temperature 180 ± 5 °C for 2 min. The oil samples before and after the 10 consecutive frying sessions were analyzed for their free fatty acid (FFA), fatty acid (FA) composition, peroxide value (POV), p-anisidine value (PAV), and iodine value (IV), while the deep-fried SP chips from each batch of frying were analyzed for the acrylamide concentration and color.

Materials and Methods

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

Materials

Acrylamide (>99%) and [13C3]-acrylamide (99% isotopic purity, 1 mg/mL) were obtained from Sigma–Aldrich (St. Louis, Mo., U.S.A.) and Cambridge Isotope Laboratories, Inc. (Andover, Mass., U.S.A.), respectively. VOs samples were purchased from local reliable retailers. All chemicals and solvents used were of analytical or high-performance liquid chromatography (HPLC) grade (Merck, Darmstadt, Germany). Fatty acid methyl esters (FAME) were obtained from Sigma–Aldrich (Bellefonte, Pa., U.S.A.).

Sample preparation

Purple flesh, standard SP roots originating from a plantation estate in Sepang, Selangor, were purchased from a local farmer. Approximately 70 kg of the fresh sweet potatoes were washed (but not peeled) and air dried (25 °C) before they were sliced to a thickness of 2 mm using a mechanical slicer (Berkel, Indianapolis, Ind., U.S.A.). Then 1 kg of SP slices were allocated for the chemical composition, RS and AA analyses. The samples for the RS and AA analyses were correctly pack and kept overnight at −20 °C and then lyophilized (Labconco, Kansas, Miss., U.S.A.) under a vacuum of 133 × 10−3 mbar at −50 °C for 60 h. The lyophilized samples were ground in a blender (Moulinex, Bagnolet, France), coded and stored in a zip-lock polyethylene bag at −20 °C before analysis. The remaining SP slices were used in the frying experiments.

Frying experiment

Slices of sweet potatoes were deep fried in PO, COO, CAO, and SBO. For each oil type, replicate experiments were conducted. In each frying experiment, 10 consecutive frying sessions of SP slices were performed. A calibrated K-type thermocouple (Fluke 51 II, Wash., U.S.A.) was used to monitor the changes of temperature during deep frying with immersion of the stainless steel thermocouple probe in the deep frying oil. The frying started 15 min after the oil temperature reached 180 °C and stabilized. A batch of 250 g of randomly chosen SP slices was deep fried at 180 ± 5 °C for 2 min at every 10 min intervals. Hence, the total frying time of 10 consecutive frying sessions was 115 min. An electric fryer (Roller Grill, Bonneval, France) with an oil capacity of 5 L was used. A fresh batch of oil was used in each frying experiment. After frying, the SP chips were dried on a paper towel, cooled to room temperature, crushed and homogenized in a blender (Moulinex, Bagnolet, France). All homogenized samples were coded and stored in a zip-lock polyethylene bag at −20 °C before being used for the acrylamide and color analyses.

In each frying experiment, 100 mL of the various unused VOs were collected before and after the 10 consecutive frying sessions. All oil samples were maintained in clean glass bottles in a dark room at 10 °C before conducting the oil analyses. Duplicate determinations of the FFA, FA composition, POV, PAV, and IV were performed on the collected VOs.

Chemical composition analysis

The chemical composition analysis, that is, estimation of the moisture, ash, protein, fat, carbohydrate, and fiber content, were conducted following the procedure described by the AOAC (2005).

Analysis of free AA

The free AA analysis was performed using the EZ:faast AA analysis kit (Phenomenex, Torrance, Calif., U.S.A.) (Farkas and Toulouee 2003). Lyophilized SP powder (500 mg) was extracted in a 10 mL of 25% solution of acetonitrile in water (v/v) for 1 h at 40 °C. Then the samples were prepared as instructed by the manufacturer and quantified using GC-FID (Agilent 6890N, Little Falls, Del., U.S.A.).

Analysis of glucose, fructose, and sucrose

The methods for the analysis of glucose, fructose, and sucrose were adapted from Sanny and others (2012). Freeze-dried samples (200 mg) were shaken in a vertical shaker (RS-1, Jeio Tech Co., Gyeonggi-do, Korea) at medium speed (ca. 285 pulses/min) in 10 mL acetonitrile/water (80:20, v/v) for 5 min. The suspension was then centrifuged (Kubota 2100, Tokyo, Japan) at 1700 RCF (g) for 10 min, and the supernatant was passed through a 0.45 μm nylon syringe filter (Sartorius Stedim Biotech, Goettingen, Germany). Aliquots (20 μL) of the filtrate were injected into a Waters HPLC instrument equipped with a refractive index detector (Milford, Mass., U.S.A.) and a Supelcosil LC-NH2 5 μm 250 mm × 4.6 mm ID column (Sigma–Aldrich, St Louis). A mobile phase of acetonitrile/water (80:20, v/v) was used as an isocratic mode of elution at a flow rate of 0.5 mL/min and maltose was used as an internal standard.

Analysis of oil

AOCS Official Methods (1993) were employed for the determination of the FFA content (Ca5a-40), POV (Cd 8b-90), and PAV (Cd 18–90) of the oil samples. The IV was measured according to PORIM Method p.3 2a (Siew and others 1995). The FA composition was determined by the conversion of the oil to FAME and quantified by GC-FID (Agilent 6890N). A medium polar cyanopropyl capillary column, that is, DB23 0.15 μm 0.25 mm × 60 m ID (Agilent Technologies, Santa Clara, Calif., U.S.A.), was used. Comparison with the retention times of a standard mixture was applied to identify the FAME peaks.

Analysis of acrylamide

The sample treatment procedure was conducted according to the method described by Sanny and others (2012). A subsample (0.5 g) of crushed SP chips were placed in a 50 mL centrifuge tube, and 10 mL of water containing 500 ng 13C3-labeled acrylamide as the internal standard (final concentration = 50 ng/mL) was added. The samples were mixed in a vertical shaker (RS-1, Jeio Tech Co., Gyeonggi-do, Korea) for 10 min at medium speed (ca. 285 pulses/min). The homogenate was centrifuged in a refrigerated centrifuge (3 to 18K, Sigma, Gillingham Dorset, UK) at 11200 RCF (g) for 30 min at 4 °C. Approximately 2 mL of an aliquot beneath the oil layer was taken using a syringe, which was filtered through a 0.45 μm nylon syringe filter (Sartorius Stedim Biotech, Goettingen, Germany); the filtrate was then collected. Both Oasis HLB and Oasis MCX cartridges were conditioned with 2 mL of methanol and equilibrated with 2 mL of water. The filtrate (2 mL) was passed through an Oasis HLB cartridge (Waters) gravitationally and discarded. The acrylamide-containing fraction from the Oasis HLB cartridge was eluted with 2 mL of water; the eluate (eluate 1) was collected and then loaded onto Oasis MCX cartridge (Waters). The eluate (eluate 2) was collected and transferred to an amber vial for the LC-MS/MS analysis. The quantification and identification of acrylamide was performed by a TSQ Quantum Ultra (Thermo Scientific, San Jose, Calif., U.S.A.) triple-quadrupole mass spectrometer with an Accela High Speed LC quaternary high pressure pump and an Accela autosampler (Thermo Finnigan, San Jose, Calif., U.S.A.). An atmospheric pressure chemical ionization (APCI) source was used, and the data analysis was performed with the Xcalibur software (Thermo Scientific). Hypercarb column (2.1 mm × 50 mm ID, 5 μm) (Thermo Electron, Bellafonte, Pa., U.S.A.) was used with 100% water as mobile phase. The acrylamide present was analyzed using APCI in the positive ion mode. The selective reaction monitored mode (SRM) was acquired with the characteristic fragmentation transitions of m/z 72 > 55 ([M+H−NH3]+) for acrylamide and m/z 75 > 58 for [13C3]-acrylamide. Six-point linear calibration plots were constructed (r2 > 0.999) for the calibration standards ranging from 5 to 500 ng/mL. The limits of detection and limit of quantification were calculated as 2 and 5 μg/kg acrylamide, respectively.

Analysis of color

The procedure was adapted from Mulla and others (2011). The color of the ground SP chips was measured using a Hunter Lab UltraScan Pro colorimeter with EasyMatch QC software (Hunter Associate Laboratory Inc., Reston, Va., U.S.A.). The Hunter Lab notation was applied, where L* indicates the level of lightness or darkness (0 for black, 100 for white), a* represents the level of redness (positive values) or greenness (negative values), and b* represents the level of yellowness (positive values) or blueness (negative values). In 1976, the Commission Internationale d'Eclairage (CIE) has adopted L*a*b* is an international standard for color measurements. The colorimeter was calibrated using a white and black plate, and the same background was used for all analyses.

Statistical analysis

General linear model (GLM) analysis of variance (ANOVA) was conducted, and differences among the mean concentrations of acrylamide for the 4 types of VOs were assessed using Tukey's multiple comparisons test. One-way ANOVA was used to evaluate the significant differences of the oil properties before and after 10 consecutive frying sessions. Pearson correlations were used to study the relationship between the increase of consecutive frying sessions and the color parameter (L*, a*, and b*) to acrylamide concentration. A value P < 0.05 was considered to be significant. The statistical analyses were performed using Minitab Statistical Software v.14 (Minitab Inc., State College, Pa., U.S.A.).

Results and Discussion

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

RS and free asparagine in SP roots

Table 1 shows that the moisture content in sweet potatoes was the highest, followed by carbohydrates, crude protein, crude fiber, ash, and crude fat. The results imply that sweet potatoes are another carbohydrate-rich food source (Lebot 2010). A preliminary sugars analysis of this study variety has been done and we observed no existence of maltose in raw SP roots. Hence, maltose was used as the internal standard for SP sugars analysis. Furthermore, literature reported that maltose was formed in cooked SP roots and not found in raw SP roots (Picha 1985). Sucrose, glucose, and fructose are the carbohydrates that occur naturally in sweet potatoes (Picha 1986). The sucrose concentration in raw sweet potatoes was 52 mg/g, whereas those of glucose and fructose were 4.17 and 5.05 mg/g (on a dry weight basis), respectively. The high concentrations of RS (glucose and fructose) are generally found in SP roots. A report from 21 SP cultivars showed that RS was varied from 16% to 45% (dry weight basis) in SP roots (Aina and others 2009). The sucrose concentration in SP roots is approximately 10 times higher than the sucrose concentration in potato roots (4.80 mg/g), approximately 5 times higher than in rye (11.5 mg/g) and wheat flour (9.76 mg/g) (Elmore and others 2005; Viklund and others 2008). The total RS, that is, glucose and fructose, in the SP root is similar to that in potatoes (9.58 mg/g) but higher than in rye flour (4.55 mg/g) and wheat flour (0.97 mg/g) (Elmore and others 2005; Viklund and others 2008).

Table 1. Chemical composition of sweet potato roots (% on a fresh weight basis)a
  1. a

    Values are the means of 3 determinations ± SD.

Moisture content65.3 ± 0.97
Carbohydrates24.9 ± 0.66
Crude protein4.51 ± 0.249
Crude fiber2.29 ± 0.086
Ash2.11 ± 0.032
Crude fat0.814 ± 0.0729

Table 2 shows that the highest amino acid concentration in sweet potatoes is glutamic acid, followed by asparagine and aspartic acid. These 3 of AA contributed more than 60% of the total free AA in SP roots. Different authors showed that both asparagine and aspartic acid are among the highest concentrations of free AA in potato, rye, and wheat samples (Elmore and others 2005; Viklund and others 2008). The relative ratio of free asparagine to the total free AA for SP root was 0.21. This ratio is similar to that of rye flour (0.26), higher than that of wheat flour (0.16), but lower than that of potato flakes (0.38) (Elmore and others 2005).

Table 2. Free AA composition of sweet potato roots (mg/g on a dry weight basis)a
Amino Acidmg/g%b
  1. a

    Values are the means of 3 determinations ± SD.

  2. b

    % of the total amino acids.

Glutamic acid2.399 ± 0.00425.6
Asparagine1.976 ± 0.00621.1
Aspartic acid1.633 ± 0.00117.4
Alanine0.601 ± 0.006.43
Serine0.483 ± 0.0025.17
Phenylalanine0.312 ± 0.0013.35
Threonine0.282 ± 0.0013.03
Valine0.261 ± 0.0022.79
4-Hydroxyproline0.225 ± 0.0012.42
Tryptophan0.180 ± 0.0011.93
Leucine0.173 ± 0.0011.85
allo-Isoleucine0.165 ± 0.0011.77
Lysine0.143 ± 0.0011.53
Proline0.137 ± 0.0011.47
Isoleucine0.103 ± 0.0021.1
α-Aminoadipic acid0.102 ± 0.0041.09
Tyrosine0.094 ± 0.0011.01
Glycine0.073 ± 0.0020.78

The comparison of RS and free asparagine content implies that the acrylamide concentration in SP chips, which ranged from 296 to 2019 μg/kg (Table 3), was within the range of the acrylamide concentration in products made from normal potatoes (1800 to 7600 μg/kg) (Viklund and others 2008) and from rye (59 to 3166 μg/kg) (Elmore and others 2005). The presence of RS and free asparagine in SP roots, as precursors of acrylamide formation, suggests that acrylamide formation in SP chips is associated with the Maillard reaction.

Table 3. Acrylamide concentration (μg/kg) of deep-fried sweet potato chips using 4 different types of vegetable oils in 10 consecutive frying sessionsa
 Type of Frying Oil
No of FryingPOCOOCAOSBO
  1. PO, Palm Olein; COO, Coconut Oil; CAO, Canola oil; SBO, Soya Bean Oil

  2. a

    Values are the means of 4 determinations ± SD

  3. b-iValues within the same column with different letters are significantly different (P < 0.05)

  4. j-lMean within the same row with different letters are significantly different (P < 0.05)

1296 ± 44.9b886 ± 22.4b374 ± 98.2b1060 ± 183b
2810 ± 73.7c1113 ± 58.2bc1185 ± 68.9c1356 ± 38.9c
31072 ± 98.2cd1090 ± 38.1bc1054± 146c1754 ± 115d
41098 ± 152cd1352 ± 84.9cd1736 ± 11.4d1859 ± 170de
51349 ± 39.7de1469 ± 44.2d1955 ± 62.4de2036 ± 72ef
61524 ± 153e1808 ± 59.7e1975 ± 115def2180 ± 92.6fg
71591 ± 36.5e2155 ± 180f1925 ± 223def2279 ± 90.6fgh
82167 ± 373f2243 ± 13.0fg2121 ± 36ef2337 ± 91.7gh
91992 ± 225f2460 ± 281gh2208 ± 274f2488 ± 61.2h
102530 ± 96.5fg2643 ± 31.1h2574 ± 148g2849 ± 100.2i
Mean1443 ± 668j1722 ± 626k1711 ± 652k2019 ± 534l

Influence of the degree of unsaturation of VO on acrylamide formation

All VOs showed that the temperature dropped from initial 180 ± 5 °C to 155 ± 5 °C after SP slices loading for frying, while the temperature increased to 172 ± 5 °C when SP chips were removed from deep fryer. No significant difference (P < 0.05) in those temperatures change among the VOs used in this study. The acrylamide concentration in the SP chips in each batch of frying, using 4 different types of VO were measured. Table 3 shows that there was a significant difference (P < 0.05) in the acrylamide concentration among the frying oils tested. The mean acrylamide concentration in SP chips from 10 fried batches, was the highest for SBO (2019 μg/kg) and the lowest for PO (1443 μg/kg). The mean acrylamide concentrations for COO and CAO are at 1722 and 1711 μg/kg, respectively. The difference among the 4 types of VO is the degree of unsaturation, and the relative degree of unsaturation can be measured by determining the FA composition and IV in the oils (Table 4) (Lalas 2009). Table 4 shows that COO contained the highest percentage of saturated FAs (>80%), whereas SBO and CAO contained more than 80% unsaturated FAs. Table 4 also shows that PO contained 43.3% saturated FAs and 55.9% unsaturated FAs. However, SBO contained higher polyunsaturated FAs than PO, which was dominated by linoleic acid (53.1%). In contrast, CAO contained a higher level of unsaturated FAs, which were dominated by oleic acid (63.7%). Moreover, Table 4 shows that before 10 consecutive frying sessions, the IV for SBO was the highest, followed by CAO, and lastly PO while the IV for COO was the lowest. A high IV indicates a high unsaturation level (Lalas 2009). Therefore, the relative degrees of unsaturation in the VOs used in the present study the lowest to the highest, are as follows that is, COO < PO < CAO < SBO. Oils containing a higher degree of polyunsaturated FAs are more susceptible to oxidative reactions than more saturated oils (Choe and Min 2006; Choe and Min 2007). Various authors found that lipids play a role in the formation of acrylamide in foods and reported that the higher the degree of unsaturation of the lipid, the higher the lipid oxidation rate and thus the higher the formation of acrylamide (Lingnert and others 2002; Ehling and others 2005; Capuano and others 2010). For example, Ehling and others (2005) revealed that oleic and linoleic acids formed a 10-fold higher acrylamide level in a heating model system compared with stearic acid when they were heated with asparagine at 180 °C. This correlation explains our results, which showed a higher acrylamide concentration in SP chips that were fried using SBO and CAO, compared with PO. This is most likely due to the higher content of oleic acid and linoleic acid in these 2 oil types when compared against PO. In addition, linoleic acid is more reactive to oxidation than oleic acid due to the presence of conjugated dienes in linoleic acid (Choe and Min 2006; Lalas 2009).

Table 4. Quality characteristics of vegetable oils before and after 10 consecutive frying sessionsa
Oil propertiesBefore fryingAfter frying
 POCOOCAOSBOPOCOOCAOSBO
  1. Abbreviations: SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; PUFA, polyunsaturated fatty acid; POV, peroxide value; IV, iodine value; PAV, p-anisidine value; FFAs, free fatty acids; FAs, fatty acids; PO, palm olein; COO, coconut oil; CAO, canola oil; SBO, soya bean oil.

  2. a

    Values are means of 4 determinations with SD in parentheses.

  3. b

    Value is significantly different (P < 0.05) compare to value before frying.

POV (meq/kg oil)1.39 (0.004)3.99 (0.197)4.62 (0.056)4.39 (0.006)8.08 (0.09)b9.65 (0.129)b9.74 (0.056)b10.0 (0.056)b
IV (g of I2/100 g oil)57.5 (0.93)14.9 (0.41)108 (0.96)125 (0.74)56.7 (0.64)14 (0.46)107 (0.31)122 (0.28)b
PAV1.26 (0.157)3.17 (0.103)3.67 (0.186)3.01 (0.091)15.4 (0.176)b11.9 (1.52)b15.5 (1.49)b20.4 (1.57)b
FFAs (%)0.09 (0.001)3.33 (0.002)0.06 (0.002)0.10 (0.001)0.11 (0.008)b4.30 (0.003)b0.07 (0.003)b0.15 (0.01)b
FAs distribution (%)        
SFA43.3 (0.05)86.1 (0.30)7.16 (0.015)15.8 (0.17)43.4 (0.19)86.2 (0.70)6.72 (0.064)b14.3 (0.34)b
MUFA45.9 (0.01)11.1 (0.12)64.4 (0.48)24.6 (0.42)45.8 (0.17)10.5 (0.03)b64.9 (0.22)30.8 (0.14)b
PUFA10.6 (0.03)2.79 (0.017)27.4 (0.51)58.7 (0.30)10.5 (0.04)2.58 (0.014)b27.0 (0.36)53.7 (0.35)b

Our findings are consistent with those of Gertz and others (2002), who reported that the lowest concentration of acrylamide was formed in French fries fried at 180 °C with PO (722 μg/kg), followed by sunflower oil (852 μg/kg), and rapeseed/CAO (1060 μg/kg). In addition, Capuano and others (2010) reported that a lower concentration of acrylamide was formed in a model system containing palm oil (approximately 750 μg/kg) compared with sunflower oil (approximately 1000 μg/kg). They concluded that a heating medium with saturated fats is less prone to thermal oxidation and hence produces less acrylamide (Capuano and others 2010).

Although COO contained the highest saturated FA concentration (86.1%) while CAO contained the lowest saturated FA concentration (7.16%, Table 4), no significant difference was found in the mean concentration of acrylamide in SP chips that were deep-fried in these 2 respective oils (P > 0.05). This result is possibly due to the higher FFA content in COO compared with CAO (Table 4). The high FFA content is well known in tropical copra extracts due to the poor refining process of the oil. Lingnert and others (2002) suggested that the higher glycerol and FFA levels were generated during frying using COO compared with palm oil due to the difference in the smoke point between these 2 oils. The remaining oxidized compounds (that is, FFAs) upon oxidation produced three-carbon-unit carbonyl compounds (that is, glycerols) that later form acrolein, which reacts with the ammonia from asparagine to form acrylamide (Yasuhara and others 2003; Ehling and others 2005).

Influence of consecutive frying sessions on acrylamide formation

In the present study, we found that consecutive frying significantly (P < 0.05) influenced the formation of acrylamide for all types of VOs. The Pearson correlation was used to study the relationship between the increase of consecutive frying sessions to acrylamide concentration from Table 3 and found a strong and significant (P < 0.05) correlation for all 4 types of VOs for the increase of consecutive frying sessions and acrylamide concentration. The Pearson correlation coefficient (r) was determined to be 0.941 for PO, 0.976 for COO, 0.92 for CAO, and 0.96 for SBO. The acrylamide concentrations were gradually increased as the number of consecutive frying sessions increased (Table 3). Table 3 shows that the initial concentration of acrylamide in SP chips that were fried using PO was significantly increased from 296 to 810 μg/kg in the second frying session, and increased to 1072 μg/kg in the third frying session. However, the concentration of acrylamide from the third to the 4th frying sessions was not significantly different (P > 0.05) (Table 3). Then a significant (P <0.05) increase of acrylamide concentration from the 4th (1098 μg/kg) to the 5th (1349 μg/kg) frying sessions. Following, there was an acrylamide increment from the 6th to the 9th frying sessions. Then the acrylamide concentration increased again from the 9th (1992 μg/kg) to the 10th frying session (2530 μg/kg). In comparison, the initial concentration of acrylamide in SP chips that were fried using SBO significantly (P < 0.05) increased from 1060 to 1754 μg/kg in the third frying session. Subsequently, the concentration of acrylamide gradually, but significantly (P < 0.05) increased from the 4th to the 10th frying sessions (Table 3). Additionally, CAO performed a similar trend when compared to SBO. These results demonstrate that PO is an oxidative stable VO as suggested by Matthäus (2007). Furthermore, Matthäus (2007) highlighted that palm oil has the highest oxidative stability when compared with other animal fats and highly unsaturated oils. Our results are in agreement with those of Capuano and others (2010), who reported that the thermal stability of palm oil, contributed to the lower formation of acrylamide in a sugar-free model system compared with a sunflower oil model.

Table 3 shows that the mean concentrations of acrylamide in SP chips, which were fried using COO, were not significantly (P > 0.05) different from the first (886 μg/kg), second (1113 μg/kg), and third (1091 μg/kg) frying sessions. There is an increase of acrylamide concentration from the third to the 4th (1352 μg/kg) frying sessions. Additionally, the mean concentrations of acrylamide between the 4th, and the 5th (1469 μg/kg) frying sessions were not significantly different (P > 0.05). The acrylamide formation appears to be slowing down, with a declining trend being observed in the first 5 frying sessions. The high composition of saturated FAs (86.1%, Table 4) in COO, possibly contributed to the lower or less prone thermal oxidation during the above mentioned period. However, the concentration of acrylamide slowly increased from the 6th to 10th frying sessions (Table 3). The increased acrylamide concentration could be due to the increased FFA composition of COO (4.30% after the 10 consecutive frying sessions compared with 3.33% before frying, Table 4). As discussed previously, these oxidized compounds (FFAs) or the precursor of three-carbon-unit carbonyl compounds (glycerols) upon oxidation will generate acrylamide formation (Yasuhara and others 2003; Ehling and others 2005).

Table 4 also shows that there was a significant difference in the mean POV and PAV before and after 10 consecutive frying sessions for all types of VOs (P < 0.05). Increased POVs indicate that the level of the primary lipid oxidation products increased, which resulted in the formation of hydroperoxides (Lalas 2009). Meanwhile, increased PAVs imply that the level of secondary lipid oxidation increased and resulted in the formation of aldehydes, generally 2-alkenals and 2,4-alkadienals (Shahidi and Zhong 2005). Zamora and Hidalgo (2008) revealed that different lipid oxidation products are able to convert asparagine to acrylamide. The reactivity of the top 3 assayed compounds was in the order of 2,4-decadienal, 4,5-epoxy-2-decenal, and methyl 13-oxooctadeca-9,11-dienoate, which were all derived from linoleic acid. They claimed that α,β,γ,δ-diunsaturated carbonyl compounds (that is, products of secondary lipid oxidation) were relatively more reactive than hydroperoxide compounds (that is, products of primary lipid oxidation) to form acrylamide (Zamora and Hidalgo 2008). These reports again explained why SBO produced higher acrylamide than other oils, probably due to linoleic acid as the dominant FA in SBO.

In addition to the POV and PAVs, Table 4 shows that there was a significant difference in the mean of FFA content before and after 10 consecutive frying sessions for all types of VOs (P < 0.05). The mean FFA contents after 10 consecutive frying sessions for PO, COO, CAO, and SBO were significantly (P < 0.05) higher than those before the 10 consecutive frying sessions. These results are in agreement with those of Chung and others (2004), who reported that the FFAs content in VO increased (from 0.017% to 0.062%) as the number of frying sessions increased (from the 1st to the 13th frying session). The FFA contents are expected to increase because water is introduced into the frying system by the raw SP chips, which accelerates the rate of hydrolysis. It is well understood that water can promote the hydrolysis of triacylglycerols to form a mixture of mono- and diacylglycerols, glycerol and FFAs (Lalas 2009). The dehydration of glycerol generates acrolein, which has a 3 carbon unit, which later forms acrylamide (Lingnert and others 2002; Yasuhara and others 2003). Moreover, repeated hydrolysis occurs during consecutive frying sessions. The water content gradually increased during consecutive frying sessions and enhanced the heat transfer of the deep frying oil; therefore, more interaction among the acrylamide precursors could occur to generate more acrylamide (Gertz and others 2003). Our findings are in similar agreement with those of Gertz and others (2003), who found a higher concentration of acrylamide in partially hydrogenated rapeseed oil upon the consecutive frying of potato French fries. They suggested that lipid oxidation upon prolonged heating could promote the formation of acrylamide. However, Mestdagh and others (2007) reported that acrylamide formation is independent of oil oxidation in a consecutive frying experiment, in which French fries were deep fried in corn oil and in palm fat.

Correlation of acrylamide concentration with color

Pearson correlation analysis was used to study the relationship of the color parameter (L*, a*, and b*) with the formation of acrylamide. Although the CIE color space is composed of 3 parameters, L*, a*, and b*, that are linked to each other, only b* showed a strong correlation with the acrylamide concentration (r = 0.642, n = 160, P < 0.05) (Figure 1). The color parameters L* and a* showed no significant correlation (P > 0.05) with the acrylamide concentration. Different studies have shown a good correlation between the L*, a*, and b* color parameters and acrylamide concentration in potato chips (Serpen and Gökmen 2009; Mulla and others 2011). However, literatures which indicated that SP chips’ color is correlated with acrylamide formation have not been found.

image

Figure 1. Acrylamide content versa color parameters b* (yellowness) for deep fried sweet potato chips for all treatments tested.

Download figure to PowerPoint

Conclusion

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

Our present study showed that the formation of acrylamide in SP chips can be associated with the Maillard reaction due to the presence of both free asparagine and RS in SP roots as precursors. Frying oil with lower degree of unsaturation such as PO appeared to be superior and safer to be used in deep fat frying. However, COO which contained highest saturated FA did not produce lowest acrylamide; this is probably due to its highest FFA values, which may lead to acrylamide formation. Even though VOs, that is SBO and CAO contain polyunsaturated which may lead to higher acrylamide formation, their nutritious values are still equally important. Meanwhile, it is certainly advisable not to reuse or perform consecutive frying with the same oil to avoid a high concentration of acrylamide in foods. Future research should focus on the quantitative measurement of the levels of the major primary and secondary lipid oxidation products after consecutive frying sessions, mainly because these oxidation products are thought to contribute to acrylamide formation.

Acknowledgment

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

This research was performed as part of project E/5172–1 supported by the International Foundation for Science and the Ministry of Science, Technology and Innovation of Malaysia for the scholarship awarded. All the authors declare no conflict of interest.

Author Contributions

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References

S. Jinap designed and planned the study, interpreted and drafted the manuscript. C.P. Tan and A. Khatib designed the study and interpreted the results. P.K. Lim designed and planned the study, collected test data, interpreted and drafted the manuscript. M. Sanny drafted the manuscript.

References

  1. Top of page
  2. AbstractPractical Application
  3. Introduction
  4. Materials and Methods
  5. Results and Discussion
  6. Conclusion
  7. Acknowledgment
  8. Author Contributions
  9. References