Accepted Manuscript Title: Acrylamide in commercial potato crisps from spanish market. Trends from 2004-2014 and assessment of the dietary exposure Author: Marta Mesías, Francisco J. Morales PII: DOI: Reference:
S0278-6915(15)00124-6 http://dx.doi.org/doi:10.1016/j.fct.2015.03.031 FCT 8278
To appear in:
Food and Chemical Toxicology
Received date: Accepted date:
2-2-2015 30-3-2015
Please cite this article as: Marta Mesías, Francisco J. Morales, Acrylamide in commercial potato crisps from spanish market. Trends from 2004-2014 and assessment of the dietary exposure, Food and Chemical Toxicology (2015), http://dx.doi.org/doi:10.1016/j.fct.2015.03.031. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACRYLAMIDE
2
MARKET. TRENDS FROM 2004-2014 AND ASSESSMENT OF THE DIETARY
3
EXPOSURE
IN
COMMERCIAL
POTATO
CRISPS
FROM
SPANISH
4 5
Marta Mesías*, Francisco J. Morales
6 7
Institute of Food Science, Technology and Nutrition, ICTAN-CSIC, 28040-Madrid, Spain
8 9 10 11
Keywords: potato crisps, acrylamide, dietary intake, exposure, trend
12 13
* Corresponding author. Institute of Food Science, Technology and Nutrition, ICTAN-
14
CSIC, José Antonio Novais 10, 28040 Madrid, Spain. Tel.: +34 91 549 2300; fax: +34 91
15
549 3627
16
E-mail address: [email protected] (M. Mesías)
17
1
Page 1 of 29
18
Abbreviations: AECOSAN, Spanish Agency of Consumption and Food Security and
19
Nutrition; BMD, Benchmark dose; BMDL, benchmark dose lower confidence limit; BW,
20
body weight; EC, European Commission; EFSA, European Food Safety Authority;
21
EUPPA, European Association for Potato Transformers; FAO, Food and Agriculture
22
Organization; FAPAS, Food Analysis Performance Assessment Scheme; IARC,
23
International Agency for Research on Cancer; JECFA, Joint FAO/WHO Expert Committee
24
on Food Additives; LOQ, limit of quantification; MAGRAMA, Ministry of Agriculture,
25
Food and Environment; MOE, margin of exposure; WHO, World Health Organization.
26
2
Page 2 of 29
27
Highlights
28
-
Acrylamide content in potato crisps ranged from 108 to 2180 µg/kg (mean: 630 µg/kg)
29
-
High variations were observed between batches for certain brands
30
-
Acrylamide content in potato crisps showed a decreased trend from 2004 to 2014
31
-
Exposure to acrylamide from potato crisps was estimated to be 0.035 µg/kg BW/day
32
Abstract
33
This research updates the acrylamide content of commercial potato crisps marketed in
34
Spain with the aim to evaluate its trend since 2004. Two different batches of 40 potato
35
crisps brands from 18 producers were analysed. Acrylamide content ranged from 108 to
36
2180 µg/kg, with an average value of 630 µg/kg and a median of 556 µg/kg. Data revealed
37
a continuous decreased trend from 2004 to 2014. In 2014, potato crisps showed an average
38
acrylamide content 14.6% lower than the previous report in 2009 and 57.6% lower than the
39
first report in 2004. These data confirm the overall effectiveness in the mitigation strategies
40
implemented by the Spanish industrial sector, although variations up to 80% were observed
41
between a number of brands. However, the 17.5% of the samples registered values higher
42
than the indicative value recommended by European Commission for potato crisps. The
43
dietary exposure to this contaminant by Spanish population through the potato crisps intake
44
was estimated to be 0.035 µg/kg body weight/day. Although exposure has decreased over
45
the last ten years, it is necessary to continue efforts to reduce acrylamide content in potato
46
crisps since there is still margin for it.
47 48
3
Page 3 of 29
49
1.
Introduction
50
Acrylamide is a process contaminant generated in several foods during cooking as a
51
consequence of the Maillard reaction, derived from the reaction between the free amino
52
acid asparagine with reducing sugars or other carbonyl compounds (Mottram and
53
Wedzicha, 2002). This chemical reaction mainly occurs when foods are subjected to high
54
temperatures as during frying, roasting or baking (Tareke et al., 2002), and in low moisture
55
conditions (Stadler et al., 2004). According to the European Food Safety Agency (EFSA),
56
processed potatoes together with coffee and cereal based food (cookies, crackers, breakfast
57
cereals and toasted bread), are the main sources of exposure to acrylamide in the diet
58
(EFSA, 2014). Due to the natural presence of large concentration of acrylamide precursors
59
in the potato, when this product is subjected to high temperatures, as during frying,
60
acrylamide is generated (Williams et al., 2005). Levels of this contaminant by up to 3000 or
61
even more than 4000 µg/kg have been detected in potatoes chips and crisps (Friedman et
62
al., 2003; Hilbig et al., 2004; Shamla and Nisha, 2014; Becalski et al., 2010).
63
It is known that dietary habits have changed over the last few years, increasing the
64
consumption of processed foods and decreasing this of natural foods (Delgado-Andrade,
65
2014). Such dietary changes are particularly evident in younger generations whose habits
66
are associated with high levels of snacking and fast food consumption which contributes to
67
high levels of acrylamide intake (Gilbert and Khokhar, 2008; Delgado-Andrade et al.,
68
2012). Among snacks foods, potato chips are one of the most popular foodstuffs and large
69
quantities are consumed in worldwide, especially by young people (Ouhtit et al., 2014;
70
Wyka et al., 2015). According to Katz et al. (2012) French fries and potato chips may
71
contribute 56% of the total acrylamide intake in the Western diet of the adolescents. Due to
72
this fact, recent researches have been focused on different aspects affecting acrylamide 4
Page 4 of 29
73
formation in fried potato products and possible mitigation strategies, which could be
74
summarized in two approaches: 1) to control acrylamide precursors in the fresh product
75
and, 2) to control the thermal treatment conditions applied during processing (Medeiros
76
Vinci et al., 2012). Authorities and producers have invested great efforts in mitigating
77
acrylamide formation and reducing acrylamide concentration in processed potatoes
78
(Biedermann et al., 2010). In this respect, reductions in the acrylamide content in potato
79
chips/crisps as compared with previous results have been reported by several authors
80
(Arribas-Lorenzo et al., 2009; Biederman et al., 2010; Cressey et al., 2012).
81
The aim of the present investigation was to determine the acrylamide content of
82
commercial potato crisps marketed in Spain as a major source of acrylamide in the diet.
83
Results were compared with data previously reported by our research group since 2004, in
84
order to evaluate the trend in the acrylamide content in Spanish potato crisps over the last
85
years and the effectiveness of the mitigation strategies. Finally, it was conducted an
86
assessment of the dietary exposure to this contaminant by Spanish population through the
87
potato crisps intake.
88 89
2.
90
2.1. Chemicals and materials
Materials and methods
91
Acrylamide standard (99%), potassium ferrocianide (Carrez I), and zinc acetate
92
(Carrez-II) and methanol (HPLC grade) were purchased from Sigma-Aldrich (St. Louis,
93
MO, USA). [13C3]-acrylamide (isotopic purity 99%) was from Cambridge Isotope Labs
94
(Andover, MA, USA). Acetic acid was from Merck (Darmstads, Germany) and hexane
95
from Panreac (Madrid, Spain). Milli-Q water used was produced using an Elix3 Millipore 5
Page 5 of 29
96
water purification system coupled to a Milli-Q module (model Advantage10) (Millipore,
97
Molsheim, France). All other chemicals, solvents and reagents were of analytical grade.
98
Reversed-phase Oasis-HLB cartridges (30 mg, 1 mL) were from Waters (Milford, MA,
99
USA). Syringe filter units (0.45 µm, cellulose) were purchased from Análisis Vínicos
100
(Tomelloso, Ciudad Real, Spain).
101
2.2. Samples
102
Two different batches from 40 brands of commercial potato crisps from 18 producers
103
were purchased in different Spanish supermarkets in September 2014. Samples containing
104
different flavourings and/or added spices were excluded for the sampling. Initially, ten
105
slices of potato crisps samples were randomly selected from each container and weighed in
106
order to calculate the mean weight of one slice of potato crisp. After that, samples (200–
107
500 g, according to the size of the bag) were mixed and grounded to assure a homogeneous
108
distribution of potential hotspots. A portion (100-200 g) was distributed in two containers
109
and stored under vacuum and light protected at 4ºC until analysis. Eighty different samples
110
were therefore individually analysed by duplicate.
111
2.3. Colour determination
112
Colour parameters were expressed according to CIE L*a*b* scale (CIE Colorimetric
113
Committee, 1974; McLaren & Riggs, 1976). The measurements were made using a
114
HunterLab Spectrophotometer CM-3500D colorimeter (Hunter Associates laboratory,
115
Stamford, Connecticut, USA). Three independent measurements of a*(redness),
116
b*(yellowness) and L*(lightness) parameters were carried out on different areas of the
117
grounded potato crisps. E index, C index and Y index were calculated according to the 6
Page 6 of 29
118
following equations: E = (L2 + a2 + b2)1/2; C = (a2 + b2)1/2; Y = 142.86b/L, which allows
119
evaluating the colour changes in the potato samples (Giangiacomo and Messina, 1988).
120
2.4. LC-ESI-MS-MS determination of acrylamide
121
Sample extraction and clean up were based on the method described by Morales et al.
122
(2007) with some modifications. Grounded sample (0.500 g) was weighed and mixed with
123
9.4 mL of water in polypropylene centrifugal tubes. Two mL of hexane were added in the
124
tubes in order to remove the fat content of the potato crisps. Mixture was spiked with 100
125
µL of a 5 µg/mL [13C3]-acrylamide methanolic solution as internal standard and later
126
homogenized (Ultra Turrax, IKA, Mod-T10 basic, Germany) for 10 min. Afterwards,
127
sample was treated with 250 µL of each Carrez I (15 g potassium ferrocianide/100 mL
128
water) and Carrez II (30 g zinc acetate/100 mL water) solutions and centrifuged (9,000g for
129
10 min) at 4ºC. Hexane was removed and samples were cleaned up by using of Oasis-HLB
130
cartridges. Cartridges were preconditioned with 1 mL methanol and 1 mL water. An aliquot
131
of the clear supernatant (1 mL) was loaded onto the cartridge at a flow rate of 2 mL/min
132
and first drops were discharged and the rest collected. Solution was filtered through a 0.45
133
µm filter into an amberlite LC–MS vial.
134
Sample extracts and calibration standards were analysed on an Agilent 1200 liquid
135
chromatograph coupled to an Agilent Triple Quadrupole MS detector (Agilent
136
Technologies, Palo Alto, CA, USA). Analytical separation was achieved with an Inertsil
137
ODS-3 column (250 x 4.6 mm, 5 µm; GL Sciences Inc., Tokio, Japan) at 30ºC. Isocratic
138
elution was achieved with a mobile phase of formic acid in water (0.2 mL/100 mL) at a
139
flow rate of 0.4 mL/min. The injection volume was 10 μL. Electrospray ionization in the
140
positive ionization mode was used. Under these chromatographic conditions, acrylamide 7
Page 7 of 29
141
eluted at 6.1 min. The needle was set at 1.0 kV. Nitrogen was used as nebulizer gas (12.0
142
L/min) and the source temperature was set at 350ºC. Signals at m/z 72.1 - m/z 55.1 and m/z
143
75.1 - m/z 58.1 were isolated for acrylamide and [13C3]-acrylamide, respectively. For the
144
transitions m/z 72.1 > m/z 55.1 and m/z 75.1 > m/z 58.1 the fragmentation was set at 76 V
145
and the collision energy at 8 V. Masses were recorded using multiple reactions monitoring
146
(MRM). For quantitation the signals at m/z 75.1 and 78.1 were used, while signals at m/z
147
58.1 and 55.1 served for qualification.
148
Acrylamide was quantified with standard solutions (5-100 µg/L) spiked with 50 µg/L
149
[13C3]-acrylamide. The peak identity is confirmed by comparison of the peak area ratios for
150
the transitions m/z 72 > 55 and 75 > 58 from sample extracts and standard solutions,
151
respectively. The limit of the quantitation was set at 20 µg/kg. The accuracy of the results
152
was recently demonstrated for potato crisps in an interlaboratory comparison study
153
launched by the Food Analysis Performance Assessment Scheme (FAPAS) program
154
(2013), yielding a z-score of 0.3. Precision (reproducibility) was lower than 10% and
155
recovery between 84 and 109%. Analysis was performed in duplicate and results were
156
expressed as µg/kg of product.
157
2.5. Food consumption data
158
Dietary intake of acrylamide from potato crisps was estimated by using the national
159
consumption database produced by the Ministry of Agriculture, Food and Environment
160
(MAGRAMA, 2010). For this purpose, annual data are obtained from households, hotels
161
and restaurants, and also from institutions such as schools, hospitals and the armed forces.
162
These sources make it possible to classify significant changes in dietary patterns and
163
constitute the most reliable means available for estimating food consumption in Spain 8
Page 8 of 29
164
(Varela Moreiras et al., 2007). Data collection is carried out by questionnaires that include
165
daily notes on household food purchased, with monthly results for 12000 households. In
166
addition, monthly notes on food purchased in 1500 hotels and restaurants and in 300
167
institutions are carried out by specialist interviewers, with trimester results (MAGRAMA,
168
2010).
169
In the present study, information about processed potatoes consumption data reported
170
for the year 2010 (million kg per year) was used. Data expressed as kg per capita has been
171
calculated, taking into account the population in Spain for this year. An average body
172
weight (BW) of 70 kg was used to estimate the total daily intake of acrylamide from
173
processed potatoes for the total population and expressed as µg/kg BW/day.
174
2.6. Statistical analysis
175
Statistical analyses were performed using Statgraphics Centurion XV (Herndon, VA,
176
USA). Data were expressed as mean ± standard deviation (SD). Analysis of variance
177
(ANOVA) and Student’s t test were applied to determine significant differences between
178
means. Differences were considered to be significant at p < 0.05. Relationships between the
179
different parameters analysed were evaluated by computing Pearson linear correlation
180
coefficients at the p < 0.05 confidence level.
181 182
3.
183
3.1. Acrylamide content in potato crisps
Results and discussion
184
In the present study, two different batches from 40 classical potato crisps brands,
185
without flavour or species added, from 18 different Spanish producers were analysed for 9
Page 9 of 29
186
the determination of acrylamide content (n = 80). Acrylamide content ranged from 108 to
187
2180 µg/kg, with an average value of 630 µg/kg and a median of 556 µg/kg (Table 1).
188
These results are in concordance with those reported by other authors, who found data
189
ranging from 591 to 1999 µg/kg (Arisseto et al., 2009), 430-1100 µg/kg (Wong et al., 2014)
190
or 244-1688 µg/kg (Boroushaki et al., 2010), being lower than other values described in
191
potato chips from Indian (82.0-4245.5 µg/kg, Shamla and Nisha, 2014) or Canadian market
192
(106-4630 µg/kg, Becalski et al., 2010). Similarly, both mean and median are close to the
193
values published by EFSA in 2010 for the category of potato crisps from fresh potatoes
194
(mean: 758 µg/kg; median: 543 µg/kg), and the highest content observed in our dataset was
195
lower than that reported by EFSA (4533 µg/kg) (EFSA, 2011).
196
The acrylamide distribution in the forty brands from potato crisps marketed in Spain is
197
shown in Fig. 1, where each value represents the mean ± standard deviation from the two
198
different batches analysed for the same brand. All the crisps contained acrylamide levels
199
higher than the limit of quantification (LOQ = 20 µg/kg). It is noticeable the identification
200
of two outliers in the distribution. Data distribution shown the high variability in the
201
acrylamide content among samples, which is likely the result of the unsatisfactory raw
202
material used and/or processing conditions applied. It is not the aim of this investigation to
203
describe the mitigation strategies applied for potato products and exhaustively reviewed
204
somewhere (Medeiros Vinci et al., 2012)
205
The application of signal values (P90, percentile 90) is useful for companies to monitor
206
the success during application of the mitigation strategies, if any. This concept can be
207
applied to the whole dataset. Three levels can be defined: acceptable (samples below the
208
P50), under evaluation (samples between P50-P90) and unacceptable (samples above P90).
209
P90 is identified as the signal value of the dataset and points out the reference value from 10
Page 10 of 29
210
which actions might be taken to low the acrylamide levels in these products. In our
211
investigation, 10% of the samples overcome the signal value (1169 µg/kg), as depicted in
212
Fig. 1.
213
Another reference used for quality control is considering the indicative values
214
recommended by the European Commission (EC). EC has recently adopted new indicative
215
acrylamide values based on the EFSA monitoring data from 2007-2012 for the main
216
sources of acrylamide exposure. These recommendations involve reference values for
217
acrylamide content in the major contributors to the intake of this compound. In cases where
218
the level of acrylamide in a foodstuff exceeds the acrylamide indicative value set for the
219
respective food category according to the recommendation, investigations into the
220
production and processing methods used by food producers must be carried out. Thereby,
221
risk associated with the acrylamide intake can be regulated and, consequently, controlled.
222
Indicative acrylamide value for potato chips according to the latest EC Recommendation of
223
8 November 2013 (2013/647/EU) is reported to be 1000 µg/kg (EC, 2013). In the present
224
study seven of the brands analysed exceed this indicative value, which means that
225
additional efforts should be implemented by this companies in order to decrease acrylamide
226
levels in their samples.
227
Additional information about the samples was taken into account to establish the
228
possible relationship with the acrylamide content. In this respect, samples were classified
229
according to (1) the presence or not of oil drops on the container layer after visual
230
inspection (oily, medium oily, not oily), (2) type of frying oil based on the information
231
provided by the manufacturer (sunflower oil, olive oil and other or undefined oil), (3) light
232
protection of the container (light-, partly light- or not light-protected) and (4) packaging
11
Page 11 of 29
233
under protective atmosphere (with or without protective atmosphere), as described in Table
234
1. In addition, price (EUR/100 g) of the potato samples was also considered as variable.
235
Statistically significant differences (p < 0.05) were observed respect to the presence of
236
oil drops on the container, where higher oil drops observed by consumer, higher acrylamide
237
level. Type of frying oil used and protective packaging, however, were not directly related
238
with the acrylamide content in the potato crisps. The price of the container ranged from
239
0.33 to 1.40 EUR/100 g (mean: 0.75 ± 0.27 EUR/100 g). Only one of the batches from one
240
of the 40 brands analysed showed a discordant result if price and acrylamide are correlated.
241
Surprisingly, a significant negative correlation between both parameters was observed (p =
242
0.0198, r2 = -0.2617), which involves that more expensive samples were associated with
243
lower levels of the process contaminant. This observation probably could be associated
244
with the fact that the largest companies, whose products may be more expensive, applied
245
more effectively the mitigation strategies than the smaller companies. In that sense, the cost
246
for the implementation of the mitigation strategies is reflected in the price for consumers.
247
Regarding colour parameters, means of E index, C index and Y index were calculated
248
to be 68.14 ± 2.11, 26.72 ± 1.84 and 0.75 ± 0.27, respectively. Colour of the samples was
249
also related with the acrylamide content. Thereby, acrylamide concentrations were
250
positively correlated with Y index (p = 0.0036, r2 = 0.3260) and negatively correlated with
251
E index (p = 0.0016, r2 = -0.3468), which would be expected due to more severe frying
252
treatment and the formation of brown pigments in the advance of the Maillard reaction
253
would involve loss of lightness and progress of the colour in the red zone (Pedreschi et al.,
254
2006). However, it is important to highlight that colour parameters are not expressed as
255
increment since the original potato tuber was not available.
12
Page 12 of 29
256
A relevant aspect to be considered is the difference observed between the batches for a
257
same brand of potato crisps. Two different batches from each commercial brand were
258
analysed. In the case of the differences between batches were higher than 20%, analyses
259
were repeated to confirm the results. Fig. 2 shows the distribution in number of cases of
260
percentage of the variation found between acrylamide content of the two different batches
261
for a same commercial brand. The variability among batches, expressed as coefficient of
262
variation (CV), ranged from 1.5% to 83.8%. Most of the samples reported differences
263
among batches lower than 20%. However it must be emphasized that 40% of the samples
264
presented variations higher than 20%, highlighting eight samples displaying variations
265
among batches in the range of 50-80% and another one with batches presenting differences
266
even higher than 80%. Similar observations have been reported by Cressey et al. (2012) in
267
potato chips from New Zealand, where some brands were extremely consistent across
268
batches while concentrations for acrylamide detected in other brands varied considerably
269
(range 371-1460 μg/kg). The results indicate that the products vary considerably in the
270
amount of acrylamide due to differences in the potato tuber and processing/manufacturing
271
conditions during the industrial process.
272
Fig. 3 shows the mean and standard deviation of the acrylamide concentration for the
273
18 producers in the potato crisps analysed in the current study. Even though different
274
number of brands were analysed per producer, it is interesting to mark that the acrylamide
275
content of some potato crisp producers were notably more variable than others. Acrylamide
276
concentrations varied from brands to brands, but also within some producers. For instance,
277
producers 1 and 6, both providing 2 batches x 7 different brands, exhibited totally different
278
results; while results for acrylamide content in potato crisps from the producer 6 ranged
279
from 145 to 1332 µg/kg, values for samples from the producer 1 were very close (range: 13
Page 13 of 29
280
145-432 µg/kg), involving low variability in the levels of acrylamide among both batches
281
and brands. In agreement with Cressey et al. (2012), the small deviations observed for some
282
brands support the statement that it is possible for a crisp manufacturer to produce a
283
consistent product with respect to acrylamide content.
284
3.2. Trends in acrylamide levels
285
Successful mitigation achievements of acrylamide levels in potato crisps and chips
286
have been reported by several authors in some countries during the last few years.
287
Biederman et al. (2010) explored the acrylamide content in potato chips from Switzerland
288
and observed that levels of the contaminant, evaluated from April 2007 to December 2009,
289
were diminished from 6900 µg/kg to 420 µg/kg over the years. In studies carried out in
290
New Zealand the range of acrylamide concentration in potato crisps decreased from 370-
291
2320 µg/kg in 2006 (Love and Grounds, 2006) to 112-1460 µg/kg in 2012 (Cressey et al.,
292
2012), possibly due to the application of mitigation measures by potato chips industries. In
293
a similar way, average acrylamide levels in German potato crisps produced from stored
294
potatoes decreased from the range of 800–1000 μg/kg in 2002–2003 to 400–600 μg/kg in
295
2004–2009 (WHO, 2011). Recently Powers et al. (2013) evaluated the acrylamide
296
concentrations in potato crisps in Europe from 2002 to 2011, observing a decrease in the
297
levels of the contaminant from 763 µg/kg in 2002 to 358 µg/kg in 2011. Overall the study
298
demonstrated the effectiveness of the approaches taken by potato crisp manufactures to
299
reduce acrylamide levels in their products. On the contrary, this downward trend in
300
acrylamide levels in potato crisps was not shown in data compiled by the EFSA (EFSA,
301
2012). In this report, mean acrylamide level of 570 µg/kg in 2007 was slightly increased to
302
758 µg/kg in 2010, suggesting, therefore, the limited evidence of active mitigation efforts. 14
Page 14 of 29
303
In the present study levels of acrylamide in potato crisp marketed in Spain in 2014
304
were compared with data previously reported by our group research in 2004 (Rufián-
305
Henares et al., 2006) and 2009 (Arribas-Lorenzo and Morales, 2009). In the 2004 sampling,
306
potato crisps showed a mean and median of 1484 µg/kg and 1180 µg/kg, respectively
307
(range: 211-5492 µg/kg). Nearly a 50% reduction was reached in the 2009 study, reporting
308
a mean content of 740 µg/kg (range: 81-2622 µg/kg, median: 592 µg/kg). In the current
309
study, levels of acrylamide were reduced again as compared with the previous sampling
310
data. In 2014, mean acrylamide content was 14.6% lower than this from 2009 and 57.6%
311
lower than this from 2004. Values for the first and third quartiles were however very close
312
to those reported in 2009, all of them being diminished when compared with 2004 results.
313
Even though minimum vales detected were close in the three studies, maximum level was
314
substantially reduced over the years, which minimizes the heterogeneity of the distribution
315
(Fig. 4). Among the samples marketed in the present study, ten potato crisps brands were
316
also analysed in the previous 2009 sampling. Seven of these ten samples showed lower
317
acrylamide content in the current study compared with the previous results; the reduction
318
was between 22-66%. The remaining three samples showed practically the same results
319
than those reported in 2009. All these findings illustrate the effectiveness in the mitigation
320
strategies carried out to diminish acrylamide levels in potato products.
321
3.3. Acrylamide exposure from potato crisps
322
Exposure assessment was estimated from the latest national consumption database
323
available, containing data obtained from households, hotels, restaurants, and institutions
324
(MAGRAMA, 2010). During 2010, the total consumption of processed potatoes was 65.78
325
million kg/year, which represents an overall estimation of annual consumption of 1.43 kg 15
Page 15 of 29
326
processed potato products per capita. Assuming an average body weight of 70 kg for adults,
327
the average contribution of potato crisps to dietary acrylamide exposure in Spain was
328
calculated to be 0.035 µg/kg BW/day. If minimum and maximum levels of acrylamide
329
quantified in the potato crisps of this experiment are considered, the range of acrylamide
330
exposure from these products would be 0.006-0.12 µg/kg BW/day. The mean value was
331
similar to those reported in other studies on adults from Latvia (Pugajeva et al., 2014) or the
332
UK (Mills et al., 2008), being higher than these from France (Sirot et al., 2012), Poland
333
(Mojska et al., 2010) or Hong Kong (Wong et al., 2014). With regard to the previous study
334
the annual consumption of processed potato products per capita in 2009 and 2014 are
335
similar but due to the mean of acrylamide levels in potato crisps have been found
336
decreased, the dietary acrylamide exposure of Spanish consumers has been reduced
337
(Arribas-Lorenzo et al., 2009). Table 2 describes the estimation of total dietary exposure to
338
acrylamide and the contribution calculated for potato chips and crisps as reported in
339
different studies. As can be observed the contribution of potato chips and crisps to the
340
dietary acrylamide exposure ranges from 1 to 46%.
341
For dietetics purposes, we consider useful to provide adequate information to
342
professionals in the nutrition area since these professionals use the concept of food
343
portions. Then, the mean weight of one slice of potato crisp was calculated by weighing ten
344
slices of potato crisps samples randomly selected from each container. This mean weight
345
was estimated to be 1.5 g/slice, which would involve an average of acrylamide intake of
346
0.94 µg/slice (range: 0.16-3.27 µg/slice).
347
The risk assessment for acrylamide in potato crisp was conducted according to the
348
harmonized approach of the EFSA (EFSA, 2005), which recommends the use of the margin
349
of exposure (MOE) based on the benchmark dose (BMD) as standardised reference point of 16
Page 16 of 29
350
departures (Benford et al., 2010). Likewise MOE is calculated for the most sensitive
351
carcinogenic effect of acrylamide by using the ratio of the reference point (BMDL10 when
352
using data from experimental animals) and a conservative estimate of human exposure
353
(JECFA, 2010). EFSA Scientific Committee on Contaminants in the Food Chain
354
(CONTAM Panel) adopted that an MOE above 10000 for a compound that is both
355
genotoxic and carcinogenic is considered of low health concern (EFSA, 2005) and would
356
not be a priority for risk management action (Barlow et al., 2006). The MOE calculated for
357
acrylamide in food is low and this may indicate a human health concern (EFSA, 2005).
358
In this respect, an MOE value of 8857 was obtained when comparing mean acrylamide
359
exposure with the established BMDL10 (310 µg/kg BW/day) for mammary tumours in
360
female rats. Similarly, an MOE value of 5143 was obtained when comparing with the
361
established BMDL10 (180 µg/kg BW/day) for Harderian gland tumours in male mice.
362
Former MOE values are just restricted to the contribution of the consumption of potato
363
crisp and are lower than the value of 10000 categorized by EFSA as low concern from a
364
public health point of view. It is worth commenting that it should be always considered the
365
uncertainty before any risk extrapolation since the confidence of the MOE estimation is
366
very much dependent on the quality of the dose–response and exposure information. The
367
dose–response data is derived from rodent bioassays where differences in species
368
toxicokinetics may have an impact on extrapolation to human exposure (Bolger et al.,
369
2010). These results are in line with data recently revised by EFSA (EFSA, 2014).
370
Therefore it is necessary to continue efforts to reduce acrylamide content in potato crisps
371
and, consequently, the dietary exposure to acrylamide.
372
17
Page 17 of 29
373
4.
Conclusions
374
The present research reports acrylamide content in commercial potato crisps marketed
375
in Spain. Results were compared with data previously reported by our research group, in
376
order to evaluate the trend in the acrylamide content in Spanish potato crisps over the last
377
ten years and, subsequently to assess the effectiveness of the mitigation strategies over the
378
time. Results revealed a decreased trend in acrylamide content from 2004 to 2014, which
379
illustrates the effectiveness in the mitigation strategies carried out to diminish acrylamide
380
levels in potato products. However, it is noticeable the high variability among batches for
381
several brands. The dietary exposure to this contaminant by Spanish population through the
382
potato crisps intake was also estimated, concluding that, although exposure has decreased
383
over the last few years, it is therefore necessary to continue efforts to reduce acrylamide
384
content in potato crisps since it is technically realistic. These data are useful for risk
385
assessment purposes in the Spanish population.
386
Conflict of interest
387 388
The authors declare that there are no conflicts of interest. Acknowledgements
389
This research was partly funded by projects S1013/ABI-3028-AVANSECAL
390
(Comunidad Autónoma of Madrid and FEDER program) and CSIC-201370E027. Authors
391
wish to thank to Ms. H. Moreno for the technical assistance and to Ms. I. Alvarez and Mr.
392
MA. Martinez for their technical assistance with acrylamide analyses.
393
References
18
Page 18 of 29
394
Arisseto, A.P., de Figueiredo Toledo, M.C., Govaert, Y., van Loco, J., Fraselle, S.,
395
Degroodt, J.M., Rosseto Caroba, D.C., 2009. Contribution of selected foods to
396
acrylamide intake by a population of Brazilian adolescents. LWT - Food Sci. Technol.
397
42, 207–211.
398 399
Arribas-Lorenzo, G., Morales, F.J., 2009. Dietary exposure to acrylamide from potato crisps to the Spanish population. Food Addit. Contam. 26, 289–297.
400
Barlow, S., Renwick, A.G., Kleiner, J., Bridges, J.W., Busk, L., Dybing, E., et al., 2006.
401
Risk assessment of substances that are both genotoxic and carcinogenic report of an
402
international conference organized by EFSA and WHO with support of ILSI Europe.
403
Food Chem. Toxicol. 44, 1636–1650.
404
Becalski, A., Stadler, R., Hayward, S., Kotello, S., Krakalovich, T., Lau, B.P.Y., et al.,
405
2010. Antioxidant capacity of potato chips and snapshot trends in acrylamide content
406
in potato chips and cereals on the Canadian market. Food Addit. Contam. 27, 1193–
407
1198.
408
Benford, D., Bolger, P.M., Carthew, P., Coulet, M., DiNovi, M., Leblanc, J.C., et al., 2010.
409
Application of the Margin of Exposure (MOE) approach to substances in food that are
410
genotoxic and carcinogenic. Food Chem. Toxicol. 48, S2–S24.
411
Biedermann, M., Grundböck, F., Fiselier, K., Biedermann, S., Bürgi, C., Grob, K., 2010.
412
Acrylamide monitoring in Switzerland, 2007–2009: results and conclusions. Food
413
Addit. Contam. 27, 1352–1362.
414
Bolger , P.M., Leblanc, J.C., Setzer, R.W., 2010. Application of the Margin of Exposure
415
(MoE) approach to substances in food that are genotoxic and carcinogenic EXAMPLE:
416
Acrylamide (CAS No. 79-06-1). Food Chem. Toxicol. 48, S25-S33. 19
Page 19 of 29
417
Boroushaki, M.T., Nikkhah, E., Kazemi, A., Oskooei, M., Raters, M., 2010. Determination
418
of acrylamide level in popular Iranian brands of potato and corn products. Food Chem.
419
Toxicol. 48, 2581–2584.
420 421
CIE Colorimetry Committee, 1974. Technical notes: working program on colour differences. J. Opt. Soc. Am., 64, 896–897.
422
Claeys, W., Baert, K., Mestdagh, F., Vercammen, J., Daenens, P., De Meulenaer, B., et al.,
423
2010. Assessment of the acrylamide intake of the Belgian population and the effect of
424
mitigation strategies. Food Addit. Contam. 27, 1199–1207.
425
Cressey, P., Thomson, B., Ashworth, M., Grounds, P., McGill, E., 2012. Acrylamide in
426
New Zealand food and updated exposure assessment. MAF Technical Paper No:
427
2011/19.
428
.
430 431
Risk
Assessment
Studies.
Report
No.
43.
Delgado-Andrade, C., 2014. Maillard reaction products: some considerations on their health effects. Clin. Chem. Lab. Med. 52, 53–60.
432
Delgado-Andrade, C., Mesías, M., Morales, F.J., Seiquer, I., Navarro, M.P., 2012.
433
Assessment of acrylamide intake of Spanish boys aged 11e14 years consuming a
434
traditional and balanced diet. LWT - Food Sci. Technol. 46, 16–22.
435
EC, 2013. European Commission. Commission Recommendation of 8 November 2013 on
436
investigations into the levels of acrylamide in food. Official Journal of the European
437
Union 315-317. 12-11-13. L 301:15-17. .
439
EFSA, 2005. Statement on summary report on Acrylamide in food of the 64th meeting of
440
the joint FAO/WHO expert committee on food additives by the Scientific Panel on 20
Page 20 of 29
441
contaminants
442
443
EFSA, 2011. European Food Safety Authority. Results on acrylamide levels in food from
444 445
in
the
food
447
.
450 451 452 453 454 455
Parma.
EFSA, 2012. European Food Safety Authority. Update on acrylamide levels in food from monitoring
449
(CONTAM),
monitoring years 2007-2009. EFSA J. 4, 2133. .
446
448
chain
years
2007
to
2010.
EFSA
J.,
10:
2938.
EFSA, 2014. European Food Safety Authority. Draft Scientific Opinion on Acrylamide in Food. < http://www.efsa.europa.eu/en/consultations/call/140701.pdf> Friedman, M., 2003. Chemistry, biochemistry, and safety of acrylamide. A review. J. Agric. Food Chem. 51, 4504–4526. Giangiacomo, R., Messina, G., 1988. Determinazione oggettiva del colore del latte alimentare mediante colorimetro tristimolo. Sci. Tecn. Latt. Cas. 39, 20–39. Gilbert, P.A., Khokhar, S., 2008. Changing dietary habits of ethnic groups in Europe and implications for health. Nutr. Rev. 66, 203–215.
456
Hilbig, A., Freidank, N., Kersting, M., Wilhelm, M., Wittsiepe, J., 2004. Estimation of the
457
dietary intake of acrylamide in German infants, children and adolescents as calculated
458
from dietary records and available data on acrylamide levels in food groups. Int. J.
459
Hyg. Environ. Health 207, 463–471.
460
IARC, 1994. International Agency for Research on Cancer. Monographs on the evaluation
461
of carcinogenic risks to humans: Some industrial chemicals 60, 389–433. International
462
Agency for Research on Cancer, Lyon, France.
463
JECFA, 2005. Joint FAO/WHO Expert Committee on Food Additives. Summary and
464
conclusions of the Sixty-fourth Meeting of the Joint FAO/WHO Expert Committee on 21
Page 21 of 29
465
Food
466
Additives,
Rome,
Italy,
8–17
February
2005.
467
JECFA, 2010. Joint FAO/WHO Expert Committee on Food Additives. Summary report of
468
the Seventy-second Meeting of the JECFA. Rome, 16–25 February 2010.
469
470
Katz, J.M., Winter, C.K., Buttrey, S.E., Fadel, J.G., 2012. Comparison of acrylamide intake
471
from Western and guideline based diets using probabilistic techniques and linear
472
programming. Food Chem Toxicol 50, 877–883.
473 474 475
Love, J., Grounds, P., 2006. Chemical food safety. Acrylamide in New Zealand food. ESR Client Report FW0545. Christchurch: ESR. MAGRAMA, 2010. Ministry of Agriculture, Food and Environment. Datos de consumo
476
alimentario.
477
comercializacion-y-distribucion-alimentaria/panel-de-consumo-alimentario/ultimos-
478
datos/>.
479 480
S0278-6915(15)00124-6 http://dx.doi.org/doi:10.1016/j.fct.2015.03.031 FCT 8278
To appear in:
Food and Chemical Toxicology
Received date: Accepted date:
2-2-2015 30-3-2015
Please cite this article as: Marta Mesías, Francisco J. Morales, Acrylamide in commercial potato crisps from spanish market. Trends from 2004-2014 and assessment of the dietary exposure, Food and Chemical Toxicology (2015), http://dx.doi.org/doi:10.1016/j.fct.2015.03.031. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
ACRYLAMIDE
2
MARKET. TRENDS FROM 2004-2014 AND ASSESSMENT OF THE DIETARY
3
EXPOSURE
IN
COMMERCIAL
POTATO
CRISPS
FROM
SPANISH
4 5
Marta Mesías*, Francisco J. Morales
6 7
Institute of Food Science, Technology and Nutrition, ICTAN-CSIC, 28040-Madrid, Spain
8 9 10 11
Keywords: potato crisps, acrylamide, dietary intake, exposure, trend
12 13
* Corresponding author. Institute of Food Science, Technology and Nutrition, ICTAN-
14
CSIC, José Antonio Novais 10, 28040 Madrid, Spain. Tel.: +34 91 549 2300; fax: +34 91
15
549 3627
16
E-mail address: [email protected] (M. Mesías)
17
1
Page 1 of 29
18
Abbreviations: AECOSAN, Spanish Agency of Consumption and Food Security and
19
Nutrition; BMD, Benchmark dose; BMDL, benchmark dose lower confidence limit; BW,
20
body weight; EC, European Commission; EFSA, European Food Safety Authority;
21
EUPPA, European Association for Potato Transformers; FAO, Food and Agriculture
22
Organization; FAPAS, Food Analysis Performance Assessment Scheme; IARC,
23
International Agency for Research on Cancer; JECFA, Joint FAO/WHO Expert Committee
24
on Food Additives; LOQ, limit of quantification; MAGRAMA, Ministry of Agriculture,
25
Food and Environment; MOE, margin of exposure; WHO, World Health Organization.
26
2
Page 2 of 29
27
Highlights
28
-
Acrylamide content in potato crisps ranged from 108 to 2180 µg/kg (mean: 630 µg/kg)
29
-
High variations were observed between batches for certain brands
30
-
Acrylamide content in potato crisps showed a decreased trend from 2004 to 2014
31
-
Exposure to acrylamide from potato crisps was estimated to be 0.035 µg/kg BW/day
32
Abstract
33
This research updates the acrylamide content of commercial potato crisps marketed in
34
Spain with the aim to evaluate its trend since 2004. Two different batches of 40 potato
35
crisps brands from 18 producers were analysed. Acrylamide content ranged from 108 to
36
2180 µg/kg, with an average value of 630 µg/kg and a median of 556 µg/kg. Data revealed
37
a continuous decreased trend from 2004 to 2014. In 2014, potato crisps showed an average
38
acrylamide content 14.6% lower than the previous report in 2009 and 57.6% lower than the
39
first report in 2004. These data confirm the overall effectiveness in the mitigation strategies
40
implemented by the Spanish industrial sector, although variations up to 80% were observed
41
between a number of brands. However, the 17.5% of the samples registered values higher
42
than the indicative value recommended by European Commission for potato crisps. The
43
dietary exposure to this contaminant by Spanish population through the potato crisps intake
44
was estimated to be 0.035 µg/kg body weight/day. Although exposure has decreased over
45
the last ten years, it is necessary to continue efforts to reduce acrylamide content in potato
46
crisps since there is still margin for it.
47 48
3
Page 3 of 29
49
1.
Introduction
50
Acrylamide is a process contaminant generated in several foods during cooking as a
51
consequence of the Maillard reaction, derived from the reaction between the free amino
52
acid asparagine with reducing sugars or other carbonyl compounds (Mottram and
53
Wedzicha, 2002). This chemical reaction mainly occurs when foods are subjected to high
54
temperatures as during frying, roasting or baking (Tareke et al., 2002), and in low moisture
55
conditions (Stadler et al., 2004). According to the European Food Safety Agency (EFSA),
56
processed potatoes together with coffee and cereal based food (cookies, crackers, breakfast
57
cereals and toasted bread), are the main sources of exposure to acrylamide in the diet
58
(EFSA, 2014). Due to the natural presence of large concentration of acrylamide precursors
59
in the potato, when this product is subjected to high temperatures, as during frying,
60
acrylamide is generated (Williams et al., 2005). Levels of this contaminant by up to 3000 or
61
even more than 4000 µg/kg have been detected in potatoes chips and crisps (Friedman et
62
al., 2003; Hilbig et al., 2004; Shamla and Nisha, 2014; Becalski et al., 2010).
63
It is known that dietary habits have changed over the last few years, increasing the
64
consumption of processed foods and decreasing this of natural foods (Delgado-Andrade,
65
2014). Such dietary changes are particularly evident in younger generations whose habits
66
are associated with high levels of snacking and fast food consumption which contributes to
67
high levels of acrylamide intake (Gilbert and Khokhar, 2008; Delgado-Andrade et al.,
68
2012). Among snacks foods, potato chips are one of the most popular foodstuffs and large
69
quantities are consumed in worldwide, especially by young people (Ouhtit et al., 2014;
70
Wyka et al., 2015). According to Katz et al. (2012) French fries and potato chips may
71
contribute 56% of the total acrylamide intake in the Western diet of the adolescents. Due to
72
this fact, recent researches have been focused on different aspects affecting acrylamide 4
Page 4 of 29
73
formation in fried potato products and possible mitigation strategies, which could be
74
summarized in two approaches: 1) to control acrylamide precursors in the fresh product
75
and, 2) to control the thermal treatment conditions applied during processing (Medeiros
76
Vinci et al., 2012). Authorities and producers have invested great efforts in mitigating
77
acrylamide formation and reducing acrylamide concentration in processed potatoes
78
(Biedermann et al., 2010). In this respect, reductions in the acrylamide content in potato
79
chips/crisps as compared with previous results have been reported by several authors
80
(Arribas-Lorenzo et al., 2009; Biederman et al., 2010; Cressey et al., 2012).
81
The aim of the present investigation was to determine the acrylamide content of
82
commercial potato crisps marketed in Spain as a major source of acrylamide in the diet.
83
Results were compared with data previously reported by our research group since 2004, in
84
order to evaluate the trend in the acrylamide content in Spanish potato crisps over the last
85
years and the effectiveness of the mitigation strategies. Finally, it was conducted an
86
assessment of the dietary exposure to this contaminant by Spanish population through the
87
potato crisps intake.
88 89
2.
90
2.1. Chemicals and materials
Materials and methods
91
Acrylamide standard (99%), potassium ferrocianide (Carrez I), and zinc acetate
92
(Carrez-II) and methanol (HPLC grade) were purchased from Sigma-Aldrich (St. Louis,
93
MO, USA). [13C3]-acrylamide (isotopic purity 99%) was from Cambridge Isotope Labs
94
(Andover, MA, USA). Acetic acid was from Merck (Darmstads, Germany) and hexane
95
from Panreac (Madrid, Spain). Milli-Q water used was produced using an Elix3 Millipore 5
Page 5 of 29
96
water purification system coupled to a Milli-Q module (model Advantage10) (Millipore,
97
Molsheim, France). All other chemicals, solvents and reagents were of analytical grade.
98
Reversed-phase Oasis-HLB cartridges (30 mg, 1 mL) were from Waters (Milford, MA,
99
USA). Syringe filter units (0.45 µm, cellulose) were purchased from Análisis Vínicos
100
(Tomelloso, Ciudad Real, Spain).
101
2.2. Samples
102
Two different batches from 40 brands of commercial potato crisps from 18 producers
103
were purchased in different Spanish supermarkets in September 2014. Samples containing
104
different flavourings and/or added spices were excluded for the sampling. Initially, ten
105
slices of potato crisps samples were randomly selected from each container and weighed in
106
order to calculate the mean weight of one slice of potato crisp. After that, samples (200–
107
500 g, according to the size of the bag) were mixed and grounded to assure a homogeneous
108
distribution of potential hotspots. A portion (100-200 g) was distributed in two containers
109
and stored under vacuum and light protected at 4ºC until analysis. Eighty different samples
110
were therefore individually analysed by duplicate.
111
2.3. Colour determination
112
Colour parameters were expressed according to CIE L*a*b* scale (CIE Colorimetric
113
Committee, 1974; McLaren & Riggs, 1976). The measurements were made using a
114
HunterLab Spectrophotometer CM-3500D colorimeter (Hunter Associates laboratory,
115
Stamford, Connecticut, USA). Three independent measurements of a*(redness),
116
b*(yellowness) and L*(lightness) parameters were carried out on different areas of the
117
grounded potato crisps. E index, C index and Y index were calculated according to the 6
Page 6 of 29
118
following equations: E = (L2 + a2 + b2)1/2; C = (a2 + b2)1/2; Y = 142.86b/L, which allows
119
evaluating the colour changes in the potato samples (Giangiacomo and Messina, 1988).
120
2.4. LC-ESI-MS-MS determination of acrylamide
121
Sample extraction and clean up were based on the method described by Morales et al.
122
(2007) with some modifications. Grounded sample (0.500 g) was weighed and mixed with
123
9.4 mL of water in polypropylene centrifugal tubes. Two mL of hexane were added in the
124
tubes in order to remove the fat content of the potato crisps. Mixture was spiked with 100
125
µL of a 5 µg/mL [13C3]-acrylamide methanolic solution as internal standard and later
126
homogenized (Ultra Turrax, IKA, Mod-T10 basic, Germany) for 10 min. Afterwards,
127
sample was treated with 250 µL of each Carrez I (15 g potassium ferrocianide/100 mL
128
water) and Carrez II (30 g zinc acetate/100 mL water) solutions and centrifuged (9,000g for
129
10 min) at 4ºC. Hexane was removed and samples were cleaned up by using of Oasis-HLB
130
cartridges. Cartridges were preconditioned with 1 mL methanol and 1 mL water. An aliquot
131
of the clear supernatant (1 mL) was loaded onto the cartridge at a flow rate of 2 mL/min
132
and first drops were discharged and the rest collected. Solution was filtered through a 0.45
133
µm filter into an amberlite LC–MS vial.
134
Sample extracts and calibration standards were analysed on an Agilent 1200 liquid
135
chromatograph coupled to an Agilent Triple Quadrupole MS detector (Agilent
136
Technologies, Palo Alto, CA, USA). Analytical separation was achieved with an Inertsil
137
ODS-3 column (250 x 4.6 mm, 5 µm; GL Sciences Inc., Tokio, Japan) at 30ºC. Isocratic
138
elution was achieved with a mobile phase of formic acid in water (0.2 mL/100 mL) at a
139
flow rate of 0.4 mL/min. The injection volume was 10 μL. Electrospray ionization in the
140
positive ionization mode was used. Under these chromatographic conditions, acrylamide 7
Page 7 of 29
141
eluted at 6.1 min. The needle was set at 1.0 kV. Nitrogen was used as nebulizer gas (12.0
142
L/min) and the source temperature was set at 350ºC. Signals at m/z 72.1 - m/z 55.1 and m/z
143
75.1 - m/z 58.1 were isolated for acrylamide and [13C3]-acrylamide, respectively. For the
144
transitions m/z 72.1 > m/z 55.1 and m/z 75.1 > m/z 58.1 the fragmentation was set at 76 V
145
and the collision energy at 8 V. Masses were recorded using multiple reactions monitoring
146
(MRM). For quantitation the signals at m/z 75.1 and 78.1 were used, while signals at m/z
147
58.1 and 55.1 served for qualification.
148
Acrylamide was quantified with standard solutions (5-100 µg/L) spiked with 50 µg/L
149
[13C3]-acrylamide. The peak identity is confirmed by comparison of the peak area ratios for
150
the transitions m/z 72 > 55 and 75 > 58 from sample extracts and standard solutions,
151
respectively. The limit of the quantitation was set at 20 µg/kg. The accuracy of the results
152
was recently demonstrated for potato crisps in an interlaboratory comparison study
153
launched by the Food Analysis Performance Assessment Scheme (FAPAS) program
154
(2013), yielding a z-score of 0.3. Precision (reproducibility) was lower than 10% and
155
recovery between 84 and 109%. Analysis was performed in duplicate and results were
156
expressed as µg/kg of product.
157
2.5. Food consumption data
158
Dietary intake of acrylamide from potato crisps was estimated by using the national
159
consumption database produced by the Ministry of Agriculture, Food and Environment
160
(MAGRAMA, 2010). For this purpose, annual data are obtained from households, hotels
161
and restaurants, and also from institutions such as schools, hospitals and the armed forces.
162
These sources make it possible to classify significant changes in dietary patterns and
163
constitute the most reliable means available for estimating food consumption in Spain 8
Page 8 of 29
164
(Varela Moreiras et al., 2007). Data collection is carried out by questionnaires that include
165
daily notes on household food purchased, with monthly results for 12000 households. In
166
addition, monthly notes on food purchased in 1500 hotels and restaurants and in 300
167
institutions are carried out by specialist interviewers, with trimester results (MAGRAMA,
168
2010).
169
In the present study, information about processed potatoes consumption data reported
170
for the year 2010 (million kg per year) was used. Data expressed as kg per capita has been
171
calculated, taking into account the population in Spain for this year. An average body
172
weight (BW) of 70 kg was used to estimate the total daily intake of acrylamide from
173
processed potatoes for the total population and expressed as µg/kg BW/day.
174
2.6. Statistical analysis
175
Statistical analyses were performed using Statgraphics Centurion XV (Herndon, VA,
176
USA). Data were expressed as mean ± standard deviation (SD). Analysis of variance
177
(ANOVA) and Student’s t test were applied to determine significant differences between
178
means. Differences were considered to be significant at p < 0.05. Relationships between the
179
different parameters analysed were evaluated by computing Pearson linear correlation
180
coefficients at the p < 0.05 confidence level.
181 182
3.
183
3.1. Acrylamide content in potato crisps
Results and discussion
184
In the present study, two different batches from 40 classical potato crisps brands,
185
without flavour or species added, from 18 different Spanish producers were analysed for 9
Page 9 of 29
186
the determination of acrylamide content (n = 80). Acrylamide content ranged from 108 to
187
2180 µg/kg, with an average value of 630 µg/kg and a median of 556 µg/kg (Table 1).
188
These results are in concordance with those reported by other authors, who found data
189
ranging from 591 to 1999 µg/kg (Arisseto et al., 2009), 430-1100 µg/kg (Wong et al., 2014)
190
or 244-1688 µg/kg (Boroushaki et al., 2010), being lower than other values described in
191
potato chips from Indian (82.0-4245.5 µg/kg, Shamla and Nisha, 2014) or Canadian market
192
(106-4630 µg/kg, Becalski et al., 2010). Similarly, both mean and median are close to the
193
values published by EFSA in 2010 for the category of potato crisps from fresh potatoes
194
(mean: 758 µg/kg; median: 543 µg/kg), and the highest content observed in our dataset was
195
lower than that reported by EFSA (4533 µg/kg) (EFSA, 2011).
196
The acrylamide distribution in the forty brands from potato crisps marketed in Spain is
197
shown in Fig. 1, where each value represents the mean ± standard deviation from the two
198
different batches analysed for the same brand. All the crisps contained acrylamide levels
199
higher than the limit of quantification (LOQ = 20 µg/kg). It is noticeable the identification
200
of two outliers in the distribution. Data distribution shown the high variability in the
201
acrylamide content among samples, which is likely the result of the unsatisfactory raw
202
material used and/or processing conditions applied. It is not the aim of this investigation to
203
describe the mitigation strategies applied for potato products and exhaustively reviewed
204
somewhere (Medeiros Vinci et al., 2012)
205
The application of signal values (P90, percentile 90) is useful for companies to monitor
206
the success during application of the mitigation strategies, if any. This concept can be
207
applied to the whole dataset. Three levels can be defined: acceptable (samples below the
208
P50), under evaluation (samples between P50-P90) and unacceptable (samples above P90).
209
P90 is identified as the signal value of the dataset and points out the reference value from 10
Page 10 of 29
210
which actions might be taken to low the acrylamide levels in these products. In our
211
investigation, 10% of the samples overcome the signal value (1169 µg/kg), as depicted in
212
Fig. 1.
213
Another reference used for quality control is considering the indicative values
214
recommended by the European Commission (EC). EC has recently adopted new indicative
215
acrylamide values based on the EFSA monitoring data from 2007-2012 for the main
216
sources of acrylamide exposure. These recommendations involve reference values for
217
acrylamide content in the major contributors to the intake of this compound. In cases where
218
the level of acrylamide in a foodstuff exceeds the acrylamide indicative value set for the
219
respective food category according to the recommendation, investigations into the
220
production and processing methods used by food producers must be carried out. Thereby,
221
risk associated with the acrylamide intake can be regulated and, consequently, controlled.
222
Indicative acrylamide value for potato chips according to the latest EC Recommendation of
223
8 November 2013 (2013/647/EU) is reported to be 1000 µg/kg (EC, 2013). In the present
224
study seven of the brands analysed exceed this indicative value, which means that
225
additional efforts should be implemented by this companies in order to decrease acrylamide
226
levels in their samples.
227
Additional information about the samples was taken into account to establish the
228
possible relationship with the acrylamide content. In this respect, samples were classified
229
according to (1) the presence or not of oil drops on the container layer after visual
230
inspection (oily, medium oily, not oily), (2) type of frying oil based on the information
231
provided by the manufacturer (sunflower oil, olive oil and other or undefined oil), (3) light
232
protection of the container (light-, partly light- or not light-protected) and (4) packaging
11
Page 11 of 29
233
under protective atmosphere (with or without protective atmosphere), as described in Table
234
1. In addition, price (EUR/100 g) of the potato samples was also considered as variable.
235
Statistically significant differences (p < 0.05) were observed respect to the presence of
236
oil drops on the container, where higher oil drops observed by consumer, higher acrylamide
237
level. Type of frying oil used and protective packaging, however, were not directly related
238
with the acrylamide content in the potato crisps. The price of the container ranged from
239
0.33 to 1.40 EUR/100 g (mean: 0.75 ± 0.27 EUR/100 g). Only one of the batches from one
240
of the 40 brands analysed showed a discordant result if price and acrylamide are correlated.
241
Surprisingly, a significant negative correlation between both parameters was observed (p =
242
0.0198, r2 = -0.2617), which involves that more expensive samples were associated with
243
lower levels of the process contaminant. This observation probably could be associated
244
with the fact that the largest companies, whose products may be more expensive, applied
245
more effectively the mitigation strategies than the smaller companies. In that sense, the cost
246
for the implementation of the mitigation strategies is reflected in the price for consumers.
247
Regarding colour parameters, means of E index, C index and Y index were calculated
248
to be 68.14 ± 2.11, 26.72 ± 1.84 and 0.75 ± 0.27, respectively. Colour of the samples was
249
also related with the acrylamide content. Thereby, acrylamide concentrations were
250
positively correlated with Y index (p = 0.0036, r2 = 0.3260) and negatively correlated with
251
E index (p = 0.0016, r2 = -0.3468), which would be expected due to more severe frying
252
treatment and the formation of brown pigments in the advance of the Maillard reaction
253
would involve loss of lightness and progress of the colour in the red zone (Pedreschi et al.,
254
2006). However, it is important to highlight that colour parameters are not expressed as
255
increment since the original potato tuber was not available.
12
Page 12 of 29
256
A relevant aspect to be considered is the difference observed between the batches for a
257
same brand of potato crisps. Two different batches from each commercial brand were
258
analysed. In the case of the differences between batches were higher than 20%, analyses
259
were repeated to confirm the results. Fig. 2 shows the distribution in number of cases of
260
percentage of the variation found between acrylamide content of the two different batches
261
for a same commercial brand. The variability among batches, expressed as coefficient of
262
variation (CV), ranged from 1.5% to 83.8%. Most of the samples reported differences
263
among batches lower than 20%. However it must be emphasized that 40% of the samples
264
presented variations higher than 20%, highlighting eight samples displaying variations
265
among batches in the range of 50-80% and another one with batches presenting differences
266
even higher than 80%. Similar observations have been reported by Cressey et al. (2012) in
267
potato chips from New Zealand, where some brands were extremely consistent across
268
batches while concentrations for acrylamide detected in other brands varied considerably
269
(range 371-1460 μg/kg). The results indicate that the products vary considerably in the
270
amount of acrylamide due to differences in the potato tuber and processing/manufacturing
271
conditions during the industrial process.
272
Fig. 3 shows the mean and standard deviation of the acrylamide concentration for the
273
18 producers in the potato crisps analysed in the current study. Even though different
274
number of brands were analysed per producer, it is interesting to mark that the acrylamide
275
content of some potato crisp producers were notably more variable than others. Acrylamide
276
concentrations varied from brands to brands, but also within some producers. For instance,
277
producers 1 and 6, both providing 2 batches x 7 different brands, exhibited totally different
278
results; while results for acrylamide content in potato crisps from the producer 6 ranged
279
from 145 to 1332 µg/kg, values for samples from the producer 1 were very close (range: 13
Page 13 of 29
280
145-432 µg/kg), involving low variability in the levels of acrylamide among both batches
281
and brands. In agreement with Cressey et al. (2012), the small deviations observed for some
282
brands support the statement that it is possible for a crisp manufacturer to produce a
283
consistent product with respect to acrylamide content.
284
3.2. Trends in acrylamide levels
285
Successful mitigation achievements of acrylamide levels in potato crisps and chips
286
have been reported by several authors in some countries during the last few years.
287
Biederman et al. (2010) explored the acrylamide content in potato chips from Switzerland
288
and observed that levels of the contaminant, evaluated from April 2007 to December 2009,
289
were diminished from 6900 µg/kg to 420 µg/kg over the years. In studies carried out in
290
New Zealand the range of acrylamide concentration in potato crisps decreased from 370-
291
2320 µg/kg in 2006 (Love and Grounds, 2006) to 112-1460 µg/kg in 2012 (Cressey et al.,
292
2012), possibly due to the application of mitigation measures by potato chips industries. In
293
a similar way, average acrylamide levels in German potato crisps produced from stored
294
potatoes decreased from the range of 800–1000 μg/kg in 2002–2003 to 400–600 μg/kg in
295
2004–2009 (WHO, 2011). Recently Powers et al. (2013) evaluated the acrylamide
296
concentrations in potato crisps in Europe from 2002 to 2011, observing a decrease in the
297
levels of the contaminant from 763 µg/kg in 2002 to 358 µg/kg in 2011. Overall the study
298
demonstrated the effectiveness of the approaches taken by potato crisp manufactures to
299
reduce acrylamide levels in their products. On the contrary, this downward trend in
300
acrylamide levels in potato crisps was not shown in data compiled by the EFSA (EFSA,
301
2012). In this report, mean acrylamide level of 570 µg/kg in 2007 was slightly increased to
302
758 µg/kg in 2010, suggesting, therefore, the limited evidence of active mitigation efforts. 14
Page 14 of 29
303
In the present study levels of acrylamide in potato crisp marketed in Spain in 2014
304
were compared with data previously reported by our group research in 2004 (Rufián-
305
Henares et al., 2006) and 2009 (Arribas-Lorenzo and Morales, 2009). In the 2004 sampling,
306
potato crisps showed a mean and median of 1484 µg/kg and 1180 µg/kg, respectively
307
(range: 211-5492 µg/kg). Nearly a 50% reduction was reached in the 2009 study, reporting
308
a mean content of 740 µg/kg (range: 81-2622 µg/kg, median: 592 µg/kg). In the current
309
study, levels of acrylamide were reduced again as compared with the previous sampling
310
data. In 2014, mean acrylamide content was 14.6% lower than this from 2009 and 57.6%
311
lower than this from 2004. Values for the first and third quartiles were however very close
312
to those reported in 2009, all of them being diminished when compared with 2004 results.
313
Even though minimum vales detected were close in the three studies, maximum level was
314
substantially reduced over the years, which minimizes the heterogeneity of the distribution
315
(Fig. 4). Among the samples marketed in the present study, ten potato crisps brands were
316
also analysed in the previous 2009 sampling. Seven of these ten samples showed lower
317
acrylamide content in the current study compared with the previous results; the reduction
318
was between 22-66%. The remaining three samples showed practically the same results
319
than those reported in 2009. All these findings illustrate the effectiveness in the mitigation
320
strategies carried out to diminish acrylamide levels in potato products.
321
3.3. Acrylamide exposure from potato crisps
322
Exposure assessment was estimated from the latest national consumption database
323
available, containing data obtained from households, hotels, restaurants, and institutions
324
(MAGRAMA, 2010). During 2010, the total consumption of processed potatoes was 65.78
325
million kg/year, which represents an overall estimation of annual consumption of 1.43 kg 15
Page 15 of 29
326
processed potato products per capita. Assuming an average body weight of 70 kg for adults,
327
the average contribution of potato crisps to dietary acrylamide exposure in Spain was
328
calculated to be 0.035 µg/kg BW/day. If minimum and maximum levels of acrylamide
329
quantified in the potato crisps of this experiment are considered, the range of acrylamide
330
exposure from these products would be 0.006-0.12 µg/kg BW/day. The mean value was
331
similar to those reported in other studies on adults from Latvia (Pugajeva et al., 2014) or the
332
UK (Mills et al., 2008), being higher than these from France (Sirot et al., 2012), Poland
333
(Mojska et al., 2010) or Hong Kong (Wong et al., 2014). With regard to the previous study
334
the annual consumption of processed potato products per capita in 2009 and 2014 are
335
similar but due to the mean of acrylamide levels in potato crisps have been found
336
decreased, the dietary acrylamide exposure of Spanish consumers has been reduced
337
(Arribas-Lorenzo et al., 2009). Table 2 describes the estimation of total dietary exposure to
338
acrylamide and the contribution calculated for potato chips and crisps as reported in
339
different studies. As can be observed the contribution of potato chips and crisps to the
340
dietary acrylamide exposure ranges from 1 to 46%.
341
For dietetics purposes, we consider useful to provide adequate information to
342
professionals in the nutrition area since these professionals use the concept of food
343
portions. Then, the mean weight of one slice of potato crisp was calculated by weighing ten
344
slices of potato crisps samples randomly selected from each container. This mean weight
345
was estimated to be 1.5 g/slice, which would involve an average of acrylamide intake of
346
0.94 µg/slice (range: 0.16-3.27 µg/slice).
347
The risk assessment for acrylamide in potato crisp was conducted according to the
348
harmonized approach of the EFSA (EFSA, 2005), which recommends the use of the margin
349
of exposure (MOE) based on the benchmark dose (BMD) as standardised reference point of 16
Page 16 of 29
350
departures (Benford et al., 2010). Likewise MOE is calculated for the most sensitive
351
carcinogenic effect of acrylamide by using the ratio of the reference point (BMDL10 when
352
using data from experimental animals) and a conservative estimate of human exposure
353
(JECFA, 2010). EFSA Scientific Committee on Contaminants in the Food Chain
354
(CONTAM Panel) adopted that an MOE above 10000 for a compound that is both
355
genotoxic and carcinogenic is considered of low health concern (EFSA, 2005) and would
356
not be a priority for risk management action (Barlow et al., 2006). The MOE calculated for
357
acrylamide in food is low and this may indicate a human health concern (EFSA, 2005).
358
In this respect, an MOE value of 8857 was obtained when comparing mean acrylamide
359
exposure with the established BMDL10 (310 µg/kg BW/day) for mammary tumours in
360
female rats. Similarly, an MOE value of 5143 was obtained when comparing with the
361
established BMDL10 (180 µg/kg BW/day) for Harderian gland tumours in male mice.
362
Former MOE values are just restricted to the contribution of the consumption of potato
363
crisp and are lower than the value of 10000 categorized by EFSA as low concern from a
364
public health point of view. It is worth commenting that it should be always considered the
365
uncertainty before any risk extrapolation since the confidence of the MOE estimation is
366
very much dependent on the quality of the dose–response and exposure information. The
367
dose–response data is derived from rodent bioassays where differences in species
368
toxicokinetics may have an impact on extrapolation to human exposure (Bolger et al.,
369
2010). These results are in line with data recently revised by EFSA (EFSA, 2014).
370
Therefore it is necessary to continue efforts to reduce acrylamide content in potato crisps
371
and, consequently, the dietary exposure to acrylamide.
372
17
Page 17 of 29
373
4.
Conclusions
374
The present research reports acrylamide content in commercial potato crisps marketed
375
in Spain. Results were compared with data previously reported by our research group, in
376
order to evaluate the trend in the acrylamide content in Spanish potato crisps over the last
377
ten years and, subsequently to assess the effectiveness of the mitigation strategies over the
378
time. Results revealed a decreased trend in acrylamide content from 2004 to 2014, which
379
illustrates the effectiveness in the mitigation strategies carried out to diminish acrylamide
380
levels in potato products. However, it is noticeable the high variability among batches for
381
several brands. The dietary exposure to this contaminant by Spanish population through the
382
potato crisps intake was also estimated, concluding that, although exposure has decreased
383
over the last few years, it is therefore necessary to continue efforts to reduce acrylamide
384
content in potato crisps since it is technically realistic. These data are useful for risk
385
assessment purposes in the Spanish population.
386
Conflict of interest
387 388
The authors declare that there are no conflicts of interest. Acknowledgements
389
This research was partly funded by projects S1013/ABI-3028-AVANSECAL
390
(Comunidad Autónoma of Madrid and FEDER program) and CSIC-201370E027. Authors
391
wish to thank to Ms. H. Moreno for the technical assistance and to Ms. I. Alvarez and Mr.
392
MA. Martinez for their technical assistance with acrylamide analyses.
393
References
18
Page 18 of 29
394
Arisseto, A.P., de Figueiredo Toledo, M.C., Govaert, Y., van Loco, J., Fraselle, S.,
395
Degroodt, J.M., Rosseto Caroba, D.C., 2009. Contribution of selected foods to
396
acrylamide intake by a population of Brazilian adolescents. LWT - Food Sci. Technol.
397
42, 207–211.
398 399
Arribas-Lorenzo, G., Morales, F.J., 2009. Dietary exposure to acrylamide from potato crisps to the Spanish population. Food Addit. Contam. 26, 289–297.
400
Barlow, S., Renwick, A.G., Kleiner, J., Bridges, J.W., Busk, L., Dybing, E., et al., 2006.
401
Risk assessment of substances that are both genotoxic and carcinogenic report of an
402
international conference organized by EFSA and WHO with support of ILSI Europe.
403
Food Chem. Toxicol. 44, 1636–1650.
404
Becalski, A., Stadler, R., Hayward, S., Kotello, S., Krakalovich, T., Lau, B.P.Y., et al.,
405
2010. Antioxidant capacity of potato chips and snapshot trends in acrylamide content
406
in potato chips and cereals on the Canadian market. Food Addit. Contam. 27, 1193–
407
1198.
408
Benford, D., Bolger, P.M., Carthew, P., Coulet, M., DiNovi, M., Leblanc, J.C., et al., 2010.
409
Application of the Margin of Exposure (MOE) approach to substances in food that are
410
genotoxic and carcinogenic. Food Chem. Toxicol. 48, S2–S24.
411
Biedermann, M., Grundböck, F., Fiselier, K., Biedermann, S., Bürgi, C., Grob, K., 2010.
412
Acrylamide monitoring in Switzerland, 2007–2009: results and conclusions. Food
413
Addit. Contam. 27, 1352–1362.
414
Bolger , P.M., Leblanc, J.C., Setzer, R.W., 2010. Application of the Margin of Exposure
415
(MoE) approach to substances in food that are genotoxic and carcinogenic EXAMPLE:
416
Acrylamide (CAS No. 79-06-1). Food Chem. Toxicol. 48, S25-S33. 19
Page 19 of 29
417
Boroushaki, M.T., Nikkhah, E., Kazemi, A., Oskooei, M., Raters, M., 2010. Determination
418
of acrylamide level in popular Iranian brands of potato and corn products. Food Chem.
419
Toxicol. 48, 2581–2584.
420 421
CIE Colorimetry Committee, 1974. Technical notes: working program on colour differences. J. Opt. Soc. Am., 64, 896–897.
422
Claeys, W., Baert, K., Mestdagh, F., Vercammen, J., Daenens, P., De Meulenaer, B., et al.,
423
2010. Assessment of the acrylamide intake of the Belgian population and the effect of
424
mitigation strategies. Food Addit. Contam. 27, 1199–1207.
425
Cressey, P., Thomson, B., Ashworth, M., Grounds, P., McGill, E., 2012. Acrylamide in
426
New Zealand food and updated exposure assessment. MAF Technical Paper No:
427
2011/19.
428
.
430 431
Risk
Assessment
Studies.
Report
No.
43.
Delgado-Andrade, C., 2014. Maillard reaction products: some considerations on their health effects. Clin. Chem. Lab. Med. 52, 53–60.
432
Delgado-Andrade, C., Mesías, M., Morales, F.J., Seiquer, I., Navarro, M.P., 2012.
433
Assessment of acrylamide intake of Spanish boys aged 11e14 years consuming a
434
traditional and balanced diet. LWT - Food Sci. Technol. 46, 16–22.
435
EC, 2013. European Commission. Commission Recommendation of 8 November 2013 on
436
investigations into the levels of acrylamide in food. Official Journal of the European
437
Union 315-317. 12-11-13. L 301:15-17. .
439
EFSA, 2005. Statement on summary report on Acrylamide in food of the 64th meeting of
440
the joint FAO/WHO expert committee on food additives by the Scientific Panel on 20
Page 20 of 29
441
contaminants
442
443
EFSA, 2011. European Food Safety Authority. Results on acrylamide levels in food from
444 445
in
the
food
447
.
450 451 452 453 454 455
Parma.
EFSA, 2012. European Food Safety Authority. Update on acrylamide levels in food from monitoring
449
(CONTAM),
monitoring years 2007-2009. EFSA J. 4, 2133. .
446
448
chain
years
2007
to
2010.
EFSA
J.,
10:
2938.
EFSA, 2014. European Food Safety Authority. Draft Scientific Opinion on Acrylamide in Food. < http://www.efsa.europa.eu/en/consultations/call/140701.pdf> Friedman, M., 2003. Chemistry, biochemistry, and safety of acrylamide. A review. J. Agric. Food Chem. 51, 4504–4526. Giangiacomo, R., Messina, G., 1988. Determinazione oggettiva del colore del latte alimentare mediante colorimetro tristimolo. Sci. Tecn. Latt. Cas. 39, 20–39. Gilbert, P.A., Khokhar, S., 2008. Changing dietary habits of ethnic groups in Europe and implications for health. Nutr. Rev. 66, 203–215.
456
Hilbig, A., Freidank, N., Kersting, M., Wilhelm, M., Wittsiepe, J., 2004. Estimation of the
457
dietary intake of acrylamide in German infants, children and adolescents as calculated
458
from dietary records and available data on acrylamide levels in food groups. Int. J.
459
Hyg. Environ. Health 207, 463–471.
460
IARC, 1994. International Agency for Research on Cancer. Monographs on the evaluation
461
of carcinogenic risks to humans: Some industrial chemicals 60, 389–433. International
462
Agency for Research on Cancer, Lyon, France.
463
JECFA, 2005. Joint FAO/WHO Expert Committee on Food Additives. Summary and
464
conclusions of the Sixty-fourth Meeting of the Joint FAO/WHO Expert Committee on 21
Page 21 of 29
465
Food
466
Additives,
Rome,
Italy,
8–17
February
2005.
467
JECFA, 2010. Joint FAO/WHO Expert Committee on Food Additives. Summary report of
468
the Seventy-second Meeting of the JECFA. Rome, 16–25 February 2010.
469
470
Katz, J.M., Winter, C.K., Buttrey, S.E., Fadel, J.G., 2012. Comparison of acrylamide intake
471
from Western and guideline based diets using probabilistic techniques and linear
472
programming. Food Chem Toxicol 50, 877–883.
473 474 475
Love, J., Grounds, P., 2006. Chemical food safety. Acrylamide in New Zealand food. ESR Client Report FW0545. Christchurch: ESR. MAGRAMA, 2010. Ministry of Agriculture, Food and Environment. Datos de consumo
476
alimentario.
477
comercializacion-y-distribucion-alimentaria/panel-de-consumo-alimentario/ultimos-
478
datos/>.
479 480
