ACRYLAMIDE IN FOOD - FORMATION OF ACRYLAMIDE AND ITS DAMAGES TO HEALTH

Serkan Ötles, Semih Ötles

ABSTRACT

Acrylamide and its analogues have been widely used since the last century for various chemical and environmental applications and can be formed by heating of biological material derived from plant tissues. This compound, identified previously as a potential industrial hazard, has now been found in many cooked foods.

Key words: acrylamide, food.

ACRYLAMIDE EXPOSURE

The exposure of acrylamide by “chips and comparable products” (approximately 20% of total acrylamide exposure) is attributed to most home cooking processes because acrylamide amounts of unfried frozen chips are relatively low (mean: 62 mg/kg, n=6) compared to the deep-fried product (mean: 436 mg/kg, n=6). Furthermore, acrylamide amount is affected by frying temperature and frying time. So, cooking conditions might influence acrylamide formation. Results of exposure estimations in different countries are summarised in Table 1 [1, 7, 9].

Why is acrylamide present in food?

The usage of acrylamide as a synthetic chemical and the surprising discovery that it occurs naturally in foods treated at high temperatures raise a complex issue. Baking certain types of breads, frying potato chips have presumably been accompanied by the formation of acrylamide since preparation these foods began centuries ago. And there are other chemicals like benzo[a]pyrene, formed during grilling or frying, which have been recognised as potential cancer – causing agents of similar potency. However, reported levels of acrylamide are higher than those for other contaminants, although a quantitative answer to the question is difficult to provide [7].

In order to understand this difficulty, it is first necessary to understand that the concept applied in the risk assessments of compounds like acrylamide is different than for most other chemicals. Following the empirical postulate of Paracelsus that the dose defines the poison, toxicology – specifically food toxicology – has applied the assumption that a certain dose is required to observe an adverse effect, and that above a threshold amount, toxic reactions should be observed within a certain probability range. Consequently, a level where no effects are seen should also exist; this is called the no – observed – adverse – effect level (NOAEL). Below this level, absence of toxic reactions is explained by the capability of human metabolism to detoxify the hazardous chemical. This concept is not merely theoretical, it has been used effectively to describe the majority of chemicals which have been assessed since committees like the Joint FAO/WHO Expert Committee on Food Additives (JECFA ) started to carry out their work [7].

The usefulness of the dose – response curve was challenged by the discovery of the structure of DNA and the understanding of how this large molecule is modified by mutations induced by using “genotoxins” like acrylamide. Hypothetically, very low concentrations of a genotoxin, even one single molecule, could induce a change in DNA that could possibly lead to the development of cancer later. Although this concept has not yet been sufficiently established through observation, the idea that some poisons may not obey Paracelsus’ theory is not easy to defend, because animal tests are not sensitive enough to detect and to quantify induction of cancer at such low levels. When, in an animal study, effects are observed at all levels administered, toxicologists can not quantify safe levels of intake; however, society requests an assessment of the risk in order to be able to understand its extent. The answer provided by toxicologists in such a case is very often a different one: th e “safe level” of intake is not defined; rather, the number of “possible adverse effects” e.g. cases of cancer in a human population, is estimated. The mathematical models used for this purpose derive their figures from animal data. They require a certain set of assumptions: by varying the factors the resulting so – called “quantitative risk assessment” may vary by several orders of magnitude. It is easy to imagine that public reaction will be different depending upon whether the outcome predicted is one, ten, a hundred or a thousand additional cases of cancer in a population of one million [7].

How is acrylamide formed in foods?

Acrylamide is found in certain foods that is cooked or processed at high temperatures. As shown in Table 2 the levels of acrylamide appear to increase with the time of heating (Saulo, 2003). However, the mechanism of acrylamide formation in foods is not well understood. Acrylamide appears to be formed as a by product of the Maillard reaction. The Maillard reaction is best known as a reaction that produces the tasty crust and olden colour in fried and baked foods as shown in Figure 1. Maillard reaction is a type of non-enzymatic browning which involves the reaction of simple sugars (carbonyl groups) and amino acids (free amino groups). They begin to occur at lower temperatures and at higher dilutions than caramelization.

The reaction can occur during baking or frying, when there is a proper combination of carbohydrates, lipids, and proteins in foods. FDA recently proposed that one mechanism may involve the amino acid asparagines which, when heated in the presence of glucose, forms acrylamide. Since the asparagine content of foods within a certain category (e.g., potatoes) varies greatly, this asparagine-dependent reaction may explain the tremendous variability in acrylamide levels even within one food category.

There may be more than one way that acrylamide forms in foods. An understanding of how acrylamide forms in various foods may lead to the development of methods to prevent or limit its formation [11].

Preliminary analyses from existing limited data indicate that potato and potato products such crisps, chips and other high-temperature cooked potatoes (e.g. roasted, baked) would contribute most to the total mean acrylamide intakes, particularly when considered together. This is observed in data from studies in Nordic, central European and Mediterranean countries (e.g. Spain and France) and other regions in the world (e.g. Australia, USA). However, other food groups with lower acrylamide concentration but consumed on a daily (or more regular) basis (e.g. bread, crisp – bread), and other foods in which levels of acrylamide are currently unknown, may also contribute substantially to the total intakes, with the magnitude varying across countries or study populations [10].

Short – term intake

For intake estimates Monte Carlo statistical techniques are used. The purpose of Monte Carlo Simulation is to evaluate by experiment quantities that would be very difficult or impossible to evaluate analytically. Such experiments typically begin by creating a set of data with known statistical properties. This is achieved by specifying every aspect of data generating process or class of such processes, and replacing the random errors of the DGP by pseudo-random numbers (numbers generated deterministically to mimic a random process with a particular distribution). Unlike analytical studies, Monte Carlo simulations cannot produce exact results. Nonetheless, Monte Carlo results are useful when analytical results are difficult to obtain. In particular, Monte Carlo experiments are often used to investigate the finite sample performance of statistical techniques, the analytical properties of which are known only asymptotically [8, 10, 17].

In an experiment, in order to provide an estimate of likely short-term intakes on the basis of the acrylamide residue data using the available food consumption data for two populations may be helpful (The Netherlands, USA). Although the matching of the residue data to foods consumed and the modelling methods varied slightly, the estimated exposures were similar. The resulting intake estimates ranged from 0.8 ľg/kg bw per day for the average consumer, to 3 ľg/kg bw per day for the 95^th percentile consumer, and 6.0 ľg/kg bw per day for the 98^th percentile consumer [10].

Long – term intake

Estimating of exposure over longer periods of time, including chronic or lifetime exposures, could be assessed given the present state of knowledge for acrylamide. The sparse, unrepresentative data available on acrylamide occurrence in foods, limited the degree to which extrapolations could be made for subsets of populations based on either biologic (e.g. gender, age, ethnic background) or food consumption differences. Nonetheless, the data do allow uncertainty estimates for the typical or median exposures that occur through food for Western European, Australian and North American diets [10].

The general agreement of the several methods used to estimate exposure using well described food consumption data from Australia, Norway, The Netherlands, Sweden, USA and from the IARC EPIC study indicate a lower bound estimate of typical exposures in the range of 0.3 to 0.8 ľg/kg bw per day depending upon whether the average or median exposure is estimated and which age groups are evaluated. Within a population, it is anticipated that children will generally have exposures that are two to three times those of adult consumers when expressed on a body weight basis. Figure 2 shows the contribution of the most important food groups to the dietary exposure of acrylamide [10].

Although there is inadequate data to reliably estimate exposure for high consumers, their exposure could be several times the mean exposure. There is essentially no acrylamide occurrence data applicable to populations where the staple food consumption, or food preparation methods, differs substantially from the Western European or North American diet. Figure 3 shows the frequency of acrylamide intake in North America. Furthermore, the generally poor understanding of the mechanisms of formation of acrylamide in foods does not allow speculation as to the presence of acrylamide in foods that is not sampled.

It is known that acrylamide is well absorbed from inhalation exposure. Also the bio–availability following oral administration in drinking water is good: approximately 50–75%. However, the bio–availability of acrylamide of food matrices is not known. But on the basis of data indicating that non-smoking people not occupationally exposed to acrylamide do have adducts of haemoglobin with acrylamide and its metabolite glycidamide (which are sensitive biomarkers for acrylamide exposure) it must be assumed that acrylamide in foods is at least partially absorbed. Repeated administration of acrylamide to experimental animals result in damage to peripheral nerves (peripheric neuropathy) as the most sensitive effect, while at higher dosages in addition muscular and testicular atrophy and decreases erythrocyt-parameters is observed. Peripheral neuropathy and haemoglobin adduct formation is also seen in occupationally exposed humans. In 1985 WHO derived a TDI of 12 mg/kg bw/day based on neurotoxicity i n sub – chronically exposed rats. With the same data US-EPA concluded to a RfD of 0.2 mg/kg bw/day. In a chronic toxicity and carcinogenicity study with rats, peripheric neuropathy was observed with a LOAEL of 2 and a NOAEL of 0.5 mg/kg bw/day. This LOAEL and NOAEL will be used for the risk evaluation [8].

CARCINOGENICITY

According to WHO the additional carcinogenic risk of a lifelong daily intake of 1 mg per person amounts to 1 case per 100,000 exposed people (equivalent with a unit life time cancer risk at 1 mg/kg bw/day of 0.7 per 1000, or an additional carcinogenic risk of 1 per 10,000 exposed people at a lifelong intake of 0.14 mg/kg bw/day). US-EPA conservatively estimated the carcinogenic risk as a unit lifetime cancer risk at 1 mg/kg bw/day of 4.5 per 1000 (equivalent with an additional carcinogenic risk of 1 per 10,000 exposed people at a lifelong intake of 0.02 mg/kg bw/day). Recently the Scientific Committee of the Norwegian Food Control Authority, considering all available data, estimated a unit lifetime cancer risk at 1 mg/kg bw/day of 1.3 per 1000. This is equivalent with an additional carcinogenic risk of 1 per 10,000 exposed people at a lifelong intake of 0.08 mg/kg bw/day. The estimations of WHO will be used in the risk evaluation. It must be noted however, that although recent evaluations o f international bodies agreed on acrylamide probably being carcinogenic to humans, these bodies also concludes that the theoretical models to calculate the actual human risk are not sufficiently reliable to quantify the risk. One of the important issues why theoretical models cannot be used is the limited knowledge about the bio–availability of acrylamide from foods. The cancer risk estimates in laboratory animals, is determined after administering acrylamide via drinking water. The carcinogenic risk should be utilised with extreme care [5, 17].

Animal data

Acrylamide is carcinogenic in laboratory rats in standard 2 year bioassays, producing increased incidences of a number of benign and malignant tumours is identified in a variety of organs (for example thyroid, adrenals, tunica vaginalis). Two separate, independent studies have confirmed this phenomenon at a dose of 2 mg/kg per day, administered in drinking water. There is also a suggestion of tumours in brain and spinal cord, and in other tissues. In a series of non-standard carcinogenicity bioassays in mice, acrylamide induced lung and skin tumours [17].

Human data

Epidemiological studies have been conducted on a cohort of more than 8 000 workers exposed to acrylamide in monomer and polymer production plants during 1925 – 1976. An evaluation performed in 1983 revealed no statistically significant excess risk of cancer in any organ, and no trend in cancer mortality was seen with increasing cumulative exposure. Data for this cohort were subsequently updated for the period 1984 – 1994, and again no statistically significant excess cancer risks were observed, with the single exception of pancreatic cancer for which a doubling of risk was found in workers most heavily exposed. The statistical power of this study was adequate to have detected a 75% excess incidence of brain cancer, a 40% increase in pancreatic cancer, a 15% increase in lung cancer, or a 9% increase in all cancers combined. All epidemiological studies have limited power to detect small increases in tumour incidence. Therefore, the absence of positive results found in most studies on acrylami de cannot be interpreted as proof that the substance cannot induce cancer in humans [2, 6, 8, 13, 17].

Risk communication

The Consultation would encourage transparent and open risk assessment and risk management processes and recognises the importance of involving interested parties (consumer, industry, retail etc.) in this process at some stages. Risk communication policy could facilitate the crucial communication process between risk assessor and risk manager and among all parties involved [8, 12, 16].

What consumers can do?

Many potentially harmful chemicals are present at extremely low levels in both the environment and our foods. In many cases the levels of these is far below those expected to have an effect on human health. In recent years, analytical methods and instrumentation have advanced considerably, allowing detection of very small levels of chemicals that may or may not have adverse effects on human health. Although the information on acrylamide in foods and its implications for human health is not yet complete, the FAO and WHO have issued interim advice based on current knowledge to minimise existing risks.

1. Foods should not be cooked excessively (for too long or at too high a temperature), but they should be cooked thoroughly enough to destroy foodborne pathogens.

2. People should eat a balanced and varied diet that includes plenty of fruits and vegetables, and should moderate their consumption of fried foods.

Sensitive and reliable methods are available to identify and measure acrylamide in foodstuffs. The measurement uncertainty of the methods is small in relation to the between-sample and the within-lot variability expected for acrylamide levels.

Acrylamide is formed when some foods are cooked or processed at high temperatures. It seems to arise when different food components react together. These may be carbohydrates, proteins and amino acids, lipids, and possibly other minor food components also. The reaction is promoted by heating and increases with the time of heating. It is not yet clear what combinations of food components are involved and it may well be that the situation is complex with many mechanisms operating. The situation is further complicated by the fact that acrylamide is a volatile and reactive substance that could itself be partially lost after formation. With the limited data available so far, it is not possible to identify any specific routes of formation nor exclude any possibilities.

To understand completely the formation and fate of acrylamide in heated foods it will be necessary to conduct hypothesis-driven model studies coupled with a systematic examination of the relation between acrylamide levels and processing/cooking conditions. This understanding would allow processing and cooking conditions to be optimised to minimise and possibly eliminate acrylamide levels in heated foods.

• The relation between acrylamide levels and processing/cooking conditions should be systematically examined.

• Hypothesis-driven model studies are needed to elucidate sources, mechanism(s) of formation and fate of acrylamide in heated foodstuffs.

• Optimisation of formulation, processing and cooking conditions to minimise and possibly eliminate acrylamide levels in foods prepared industrially and at home should be investigated.

• The range of foods investigated needs to be extended to include staple foods from different regions and diets.

Despite the uncertainties in the extrapolation of tumour incidences from experimental animals to humans, such an extrapolation does provide some insight into the (maximum) risk for humans that might be contributable to acrylamide exposure in foods. From the exposure estimations it appears that the additional cancer risk for the average population (aged 1 – 97 years), for the children aged 1 – 6 years, and for the youngsters aged 7 – 18 years might not be negligible.

REFERENCES

1. Acrylamide. 2002. In: Guidelines for drinking-water quality. vol. 2: Health criteria and other supporting information. International Programme on Chemical Safety, World Health Organization, Geneva.

2. Chakrabarti T., Ungeheuer P., 2002. Health implications of acrylamide in food, Food safety. Report of a Joint FAO/WHO Consultation. FAO, Geneva, 1-34.

3. Exploratory data on acrylamide in foods. U. S. F. D. A., CFSAN, www.cfsan.fda.gov/~dms/acrydata.html.

4. Giese J., 2002. Acrylamide in foods. Food technol. 56(10), 71-72.

5. Johnson K., Gorzinski S., Bodner K., Campbell R., Wolf C., Friedman M., Mast R., 1986. Chronic toxicity and oncogenicity study on acrylamide incorporated in the drinking water of Fischer 344 rats. Toxicol. Appl. Pharmacol. 85, 154-168.

6. Kistemaker C., Bouman M., Hulshof K., 1998. The consumption of seperate food products by Dutch populations. Dutch National Food Consumption Survey 1997-1998, TNO - Rep. V98.812.

7. Konings E., Baars A., van Klaveren D., Spanjer M., Rensen P., Hiemstra M., van Kooij J., Peters P., 2003. Acrylamide exposure from foods of the dutch population and an assessment of the consequent risks. Food Chem. Toxicol. 41, 1569-1579.

8. Richmond P., Borrow R., 2003. Acrylamide in food. Lancet 361(2), 361-362.

9. Risk assessment of acrylamide intake from foods with special emphasis on cancer risk. 2002. Report from the Scientific Committee of the Norwegian Food Control Authority, Oslo.

10. Rosen J., 2002. Acrylamide in food: Is It a real treat to public health? A position paper of Am. Council Sci. Health 12, 1-17.

11. Saulo A., 2003. Acrylamide in hoods. Food Safety Technol. 13, 1-2.

12. Simonne H., Archer L., 2002. Acrylamide in foods: A review and Update. Univ. Florida Extension 10, 1-3.

13. Stadler R.H., Blank I., Natalia V., Robert F., Hau J., Guy P.A., Robert M., Riediker S., 2002. Acrylamide from Maillard reaction products. Nature 419, 449 - 450.

14. Svensson K., Abramsson L., Becker W., Glynn A., Hellenas K., Lind Y., Rosen J., 2003. Dietary intake of acrylamide in Sweden. Food Chem. Toxicol. 41, 1581-1586.

15. Tareke E., Rydberg P., Karlsson P., Ericksson S., Törnquist M., 2000 Acrylamide: A cooking carcinogen? Chem. Res. Toxicol. 13, 517-522.

16. Tyl R., Crump K., 2003. Acrylamide in food. Food Stand. Agen. 5, 215-222.

17. Vattem A., Shetty K., 2003. Acrylamide in food: a model for mechanism of formation and its reduction. Inn. Food Sci. Emerg. Technol 4, 331-338.

Responses to this article, comments are invited and should be submitted within three months of the publication of the article. If accepted for publication, they will be published in the chapter headed ‘Discussions’ in each series and hyperlinked to the article.