Furan, 2-methylfuran and 3-methylfuran in Selected Foods – April 1, 2018 to March 31, 2019

Food chemistry – Targeted surveys – Final report

Summary

Targeted surveys provide information on potential food hazards and enhance the Canadian Food Inspection Agency's (CFIA's) routine monitoring programs. These surveys provide evidence regarding the safety of the food supply, identify potential emerging hazards, and contribute new information and data to food categories where it may be limited or non-existent. They are often used by the Agency to focus surveillance on potential areas of higher risk. Surveys can also help to identify trends and provide information about how industry complies with Canadian regulations.

Furan is a chemical that can unintentionally form in foods that undergo thermal treatment such as canning^Reference 1. Precursors to furan that are often present in food include ascorbic acid, polyunsaturated fatty acids, amino acids and sugars^Reference 1 ^Reference 2 ^Reference 3. This compound occasionally coexists with 2-methylfuran and 3-methylfuran. In this report, the term 'furans' refers to the sum of furan, 2-methylfuran and 3-methylfuran, whereas 'furan' refers only to the furan compound. The term "analogue" refers to compounds which have similar but slightly different structures; it is sometimes used in this report to refer to the 3 forms of furan. It should be noted that the furans in this survey are not related to chlorinated dibenzofurans, the environmental contaminants which are often also referred to as "furans".

Furan may pose a health risk to the consumer, as the International Agency for Research on Cancer (IARC) has classified it as 'possibly carcinogenic to humans'^Reference 4. Additionally, 2-methylfuran and 3-methylfuran have been shown to have a similar toxicity to furan^Reference 5. Although preliminary estimates for consumer exposure are well below what would cause harmful effects, limited information is available concerning furan levels in food. Therefore, the goal of this survey was to generate further baseline surveillance data on the presence and levels of furan, 2-methylfuran and 3-methylfuran in selected heat treated foods available on the Canadian retail market.

A total of 250 samples were collected from retail stores in 6 cities across Canada. The samples collected included canned fish, canned meat, canned milk, dried dairy products, dried beverage mix and soy sauce. Furans were detected in 70% of the survey samples and levels ranged from 1.23 ppb (parts per billion) to 179 ppb. The highest average concentrations of furans were found in soy sauce. None of the dried dairy products had detected levels of furans. In 52% of the samples, only 1 of the 3 furans was present. It was also common for products to contain both furan and 2-methylfuran. The results from this survey were comparable to those found in international surveys and a variety of scientific studies.

The levels of furans observed in this survey were evaluated by Health Canada (HC) who determined that none of the samples would pose an unacceptable human health concern, therefore there were no recalls resulting from this survey.

What are targeted surveys

Targeted surveys are used by the CFIA to focus its surveillance activities on areas of highest health risk. The information gained from these surveys provides support for the allocation and prioritization of the agency's activities to areas of greater concern. Originally started as a project under the Food Safety Action Plan (FSAP), targeted surveys have been embedded in our regular surveillance activities since 2013. Targeted surveys are a valuable tool for generating information on certain hazards in foods, identifying and characterizing new and emerging hazards, informing trend analysis, prompting and refining health risk assessments, highlighting potential contamination issues, as well as assessing and promoting compliance with Canadian regulations.

Food safety is a shared responsibility. We work with federal, provincial, territorial and municipal governments and provide regulatory oversight of the food industry to promote safe handling of foods throughout the food production chain. The food industry and retail sectors in Canada are responsible for the food they produce and sell, while individual consumers are responsible for the safe handling of the food they have in their possession.

Why did we conduct this survey

The main objectives of this targeted survey were to generate further baseline surveillance data on the level of furan, 2-methylfuran and 3-methylfuran in domestic and imported products on the Canadian retail market, and to compare the prevalence of furans in foods targeted in this survey with that of similar products in international surveys and to the scientific literature.

Furan may pose a health risk to consumers since the IARC has classified it as 'possibly carcinogenic to humans'^Reference 4. Furan can sometimes form in foods that undergo heat treatments, particularly in foods that contain its precursors such as ascorbic acid, polyunsaturated fatty acids, amino acids and sugars^Reference 1 ^Reference 2 ^Reference 3. In some foods, 2-methylfuran and 3-methylfuran can also form, which have a similar toxicity to furan^Reference 5. Because thermal treatments are widely used for manufacturing shelf-stable food, it is important to establish data on the prevalence of furan, 2-methylfuran and 3-methylfuran in food available on the Canadian retail market.

Maximum Residue Limits (MRLs) for furans levels have not yet been established, as the toxicity of furans in humans is not well known. The U.S. Food and Drug Administration (FDA) and the European Food Safety Authority (EFSA) have studied furan levels in a variety of commodities^Reference 6 ^Reference 7, but limited data is available concerning 2- and 3-methylfuran levels. This survey was initiated in consultation with HC to establish further baseline surveillance data to compliment and expand upon the data collected by other agencies^Reference 1.

What did we sample

A variety of domestic and imported canned fish, canned meats, canned milks, dried beverage mixes, dried dairy products and soy sauces were sampled between April 1, 2018 and March 31, 2019. Samples of products were collected from local/regional retail locations located in 6 major cities across Canada. These cities encompassed 4 Canadian geographical areas: Atlantic (Halifax), Quebec (Montreal), Ontario (Toronto, Ottawa) and the West (Vancouver, and Calgary). The number of samples collected from these cities was in proportion to the relative population of the respective areas. The shelf life, storage conditions, and the cost of the food on the open market were not considered in this survey.

Table 1. Distribution of samples based on product type and origin
Product type	Number of domestic samples	Number of imported samples	Number of samples of unspecified^{Table Note a} origin	Total number of samples
Canned fish	4	45	1	50
Canned meat	14	35	1	50
Canned milk	14	0	2	16
Dried beverage mix	10	7	25	42
Dried dairy products	20	5	17	42
Soy sauce	13	33	4	50
Grand total	75	125	50	250

Table Note

Table Note a

Unspecified refers to those samples for which the country of origin could not be assigned from the product label or available sample information

Return to table note a referrer

How were samples analyzed and assessed

Samples were analyzed by an ISO 17025 accredited CFIA food testing laboratory. The results presented represent finished food products as sold and not as they would be consumed, whether the product sampled is considered an ingredient or requires preparation prior to consumption.

In the absence of established tolerances or standards for furans in foods, elevated levels of in specific foods may be assessed by HC on a case-by-case basis using the most current scientific data available.

What were the survey results

Of the 250 samples tested, 176 (70%) had detected levels of furans. Table 1 shows that the detection rate of furans varied greatly between product types. Among all product types included in this survey, the average concentration was highest in soy sauce and lowest in dried dairy products. Average concentrations of the analogues were comparable, with furan having the highest average concentration and 3-methylfuran having the lowest. In 38% of products included in this survey, furan and 2-methylfuran were both detected, and more than half of the samples contained only a single furan analogue.

Table 2. Summary of targeted survey results on furans in selected foods
Product type	Number of samples	Number of samples (%) with detected levels	Minimum (ppb)	Maximum (ppb)	Average^{Table Note b} (ppb)
Canned fish	50	50 (100)	16.3	146	46.6
Canned meat	50	50 (100)	10.8	64	31.1
Canned milk	16	10 (63)	13.9	33.2	19.8
Dried beverage mix	42	15 (36)	5.1	22.5	9.7
Dried dairy products	42	0 (0)	0	0	0
Soy sauce	50	50 (100)	1.2	179	73.8
Grand total	250	176 (70)	1.2	179	45.5

Table Note

Table Note b

Only positive results were used to calculate the average levels

Return to table note b referrer

Canned fish

All canned fish samples had detected levels of furans. As the chart in Figure 1 shows, most canned fish samples contained between 16 ppb and 84 ppb furans, with a few samples containing above 110 ppb.

Canned meat

All canned meat samples had detected levels of furans. Of the 12 samples included in this survey that contained all 3 furan analogues, 11 were canned meat samples.

Canned milk

A majority (63%) of canned milk samples had detected levels of furans. In canned milk products, the average 2-methylfuran level was nearly twice that of furan, whereas for all other product types, average furan and 2-methylfuran levels were comparable.

Dried beverages

A majority (64%) of sampled dried beverage mixes did not have detected levels of furans and dried beverages contained the lowest positive concentration of all product types included in this survey. Dried beverages included tea-based drink mixes as well as hot and cold chocolate-based drink mixes.

Dried dairy products

None of the 42 dried dairy products sampled in this survey contained detected levels of furans. This category included skim milk powder and coffee whitener.

Soy sauce

All of the 50 soy sauces sampled contained detected levels of furans. The average concentration of furans in soy sauces was higher than that of other product types included in this survey.

Figure 1 shows that the levels of furans varied considerably between individual soy sauce samples. For some brands, the entire range of observed levels was above or below the average concentration, although some brands had only a few samples. Furans levels of different lots within the same brand were generally comparable.

Figure 1. Distribution of furans concentrations for each product type in 2018 to 2019. Description follows. — Figure 1. Distribution of furans concentrations for each product type in 2018 to 2019

What do the survey results mean

For all product types, furan levels found in this survey were comparable to the data in the scientific literature and previous targeted surveys^Reference 6 ^Reference 7 ^Reference 8 ^Reference 9 ^{Reference 10} ^{Reference 11}. Table 3 only compares levels for furan, as limited data is available concerning 2- and 3-methylfuran levels in foods and therefore no comparison on the presence of 2- and 3-methylfuran could be made.

The literature shows that 2-methylfuran and 3-methylfuran can form alongside furan from precursors found in foods, although limited data is available concerning specific precursors or reaction pathways^{Reference 12} ^{Reference 13}. The variety of 2- and 3-methylfuran levels in foods sampled in this survey is possibly due to differences in ingredients and processing, which may favour the formation of different furan analogues.

Table 3. Minimum, maximum and average concentration of furan across various studies
Product type	Study	Number of samples	Minimum (ppb)	Maximum (ppb)	Average (ppb)
Canned fish	CFIA survey, 2018 to 2019	50	1.62	64.4	16.9^{Table Note c}
Canned fish	Crews et al., 2012^{Table Note g}	37	<0.3	70	26
Canned fish	Kim et al., 2009^{Table Note h}	29	0.4	212	39.1^{Table Note c}
Heat treated fish	EFSA survey, 2004 to 2010^{Table Note i}	47	<5	172	17
Canned fish	FDA survey, 2004^{Table Note j}	6	3.25^{Table Note d}	7.1	4.93
Canned meat	CFIA survey, 2018 to 2019	50	1.19	40	14^{Table Note c}
Canned meat	Kim et al., 2009^{Table Note h}	19	2.8	194	24.3^{Table Note c}
Meat products	EFSA survey, 2004 to 2010^{Table Note i}	147	<8	160	13-17^{Table Note e}
Meat products	FDA survey, 2004^{Table Note j}	9	3.25^{Table Note d}	39.2	15.4^{Table Note c}
Canned milk	CFIA survey, 2018 to 2019	16	1.27	20.8	6.87^{Table Note c}
Canned milk	FDA survey, 2004^{Table Note j}	3	10.9	15.3	12.5
Milk products	EFSA survey, 2004 to 2010^{Table Note i}	64	<0.5	80	5-5.6^{Table Note e}
Dried beverage	CFIA survey, 2018 to 2019	43	2.7	8.22	5.43^{Table Note c}
Dried chocolate beverage^{Table Note f}	FDA survey, 2004^{Table Note j}	7	0.5^{Table Note f}	0.8^{Table Note d} ^{Table Note f}	0.76^{Table Note f}
Dried dairy products	CFIA survey, 2018 to 2019	42	0	0	0
Baby formula powdered milk	Liu et al., 2018^{Table Note k}	10	2.4	26.1	8.93
Soy sauce	CFIA survey, 2018 to 2019	50	1.09	159	38.5^{Table Note c}
Soy sauce	Arisseto et al., 2009 to 2011^{Table Note l}	46	6	138	28.2
Soy sauce	EFSA survey, 2004 to 2010^{Table Note i}	94	<0.07	78	27
Soy sauce	FDA survey, 2004^{Table Note j}	5	17.2	75.6	52.1

Table Notes

Table note c

Only positive results were used to calculate the average (hazard) levels.

Return to table note c referrer

Table note d

Recorded value was between LOD and LOQ, so an average of LOD and LOQ is used.

Return to table note d referrer

Table note e

A range indicates different results for the lower and upper bounds.

Return to table note e referrer

Table note f

Commodity tested as it is consumed.

Return to table note f referrer

Table note g

Crews, C., Pye, C. (2013). Furan in canned sardines and other fish. Food Additives & Contaminants: Part B, 7(1), pp. 43-45.

Return to table note g referrer

Table note h

Kim, Y. M., Her, J.-Y., Kim, M.K., Lee, K.-G. (2015). Formation and reduction of furan in a soy sauce model system. Food Chemistry, 189, pp. 114-119.

Return to table note h referrer

Table note i

Update on furan levels in food from monitoring years 2004-2010 and exposure assessment. (2011) Italy. European Food Safety Authority.

Return to table note i referrer

Table note j

Exploratory Data on furan in Food. (2005). United States of America. U.S. Food and Drug Administration.

Return to table note j referrer

Table note k

Liu, Y.-T., Tsai, S.-W. (2010). Assessment of dietary furan exposures from heat processed foods in Taiwan. Chemosphere, 79(1), pp. 54-59.

Return to table note k referrer

Table note l

Arisseto, A.P., Furlani, M.S., Pereira, A.L.D., Toledo, M.C.F., Ueno, M.S., Vincente, E. (2012). Occurrence of furan in commercial processed foods in Brazil. Food Additives and Contaminants: Part A, 29(12), pp. 1832-1839.

Return to table note l referrer

The differences in canned fish furan levels between this survey and the other studies in Table 3 as well as the variety of reported levels within this survey might be caused by the differing levels of polyunsaturated fatty acids (a precursor to furan) in fish, which has been shown to change depending on type and diet of fish^{Reference 14}. Among canned meats sampled in this and other surveys, differences in ingredients and degree of thermal processing may have contributed to the variety of observed furan, 2-methylfuran and 3-methylfuran concentrations.

Dried beverage products sampled in other surveys were tested as-consumed whereas the CFIA tested dried beverages as-purchased. When a dilution factor is considered, furan levels in dried beverages from this survey are comparable to dried beverages included in other surveys.

Canned milk and soy sauce samples tested in this survey reported comparable furan levels to those reported in literature. Additionally, in this survey and in the literature, soy sauce samples generally contained more furan than other product types tested. This could be due to a process called "Maillard browning" which is often used in soy sauce manufacturing to enhance flavour and extend shelf-life^{Reference 15} ^{Reference 16}. Maillard browning involves the Maillard reaction, which has been shown to be a formation pathway for furan^{Reference 17}.

HC's Bureau of Chemical Safety determined the levels of furan, 2-methylfuran and 3-methylfuran in food observed in this survey are not expected to pose a concern to human health; therefore no follow-up actions were required.

References

Reference 1

Food and Nutrition – furan. (2016). Canada. Health Canada.

Return to reference 1 referrer

Reference 2

Blank, I., Conde-Petit, B., Kerler, J., Limacher, A. (2007). Formation of furan and methylfuran from ascorbic acid in model systems and food. Food Additives and Contaminants, 24(S1), pp. 122-135.

Return to reference 2 referrer

Reference 3

Locas, C.P., Yaylayan, V.A. (2004). Origin and Mechanistic Pathways of Formation of the Parent furans – A Food Toxicant. Journal of Agricultural and Food Chemistry, 52(22), pp. 6830-6836.

Return to reference 3 referrer

Reference 4

Dry Cleaning, Some Chlorinated Solvents and Other Industrial Chemicals. (1995). IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, 63, pp. 393-407.

Return to reference 4 referrer

Reference 5

Risks for public health related to the presence of furan and methylfurans. (2017). EFSA Journal, 15(10).

Return to reference 5 referrer

Reference 6

Exploratory Data on furan in Food. (2005). United States of America. U.S. Food and Drug Administration.

Return to reference 6 referrer

Reference 7

Update on furan levels in food from monitoring years 2004-2010 and exposure assessment. (2011) Italy. European Food Safety Authority.

Return to reference 7 referrer

Reference 8

Crews, C., Pye, C. (2013). Furan in canned sardines and other fish. Food Additives & Contaminants: Part B, 7(1), pp. 43-45.

Return to reference 8 referrer

Reference 9

Kim, S., Kim, T.-K., Lee, K.-G., Lee, Y.-K., Park, Y.S. (2009). Furan in Commercially Processed Foods: Four-Year Field Monitoring and Risk Assessment Study in Korea. Journal of Toxicology and Environmental Health, Part A, 72(21-22), pp. 1304-1310.

Return to reference 9 referrer

Reference 10

Liu, Y.-T., Tsai, S.-W. (2010). Assessment of dietary furan exposures from heat processed foods in Taiwan. Chemosphere, 79(1), pp. 54-59.

Return to reference 10 referrer

Reference 11

Return to reference 11 referrer

Reference 12

Limacher, A., Kerler, J., Conde-Petit, B., Blank, I. (2007). Formation of furan and methylfuran from ascorbic acid in model systems and food. Food Additives & Contaminants, 24, pp. 122-135.

Return to reference 12 referrer

Reference 13

Becalski, A., Seamen, S. (2005). Furan precursors in food: a model study and development of a simple headspace method for determination of furan. Journal of AOAC International, 88(1), pp. 102-106.

Return to reference 13 referrer

Reference 14

Jahreis, G., Kuhnt, K., Strobel, C. (2012). Survey of n-3 and n-6 polyunsaturated fatty acids in fish and fish products. Lipids in Health and Disease, 11(144).

Return to reference 14 referrer

Reference 15

Lertsiri, S., Maungma, R., Assavanig, A., Bhumiratana, A. (2000). Roles of the Maillard reaction in browning during moromi process of Thai soy sauce. Journal of Food Processing and Preservation, 25(2), pp. 149-162.

Return to reference 15 referrer

Reference 16

Gao, X., Cui, G., Ren, J., Zhao, H., Zhao, Q., Zhao, M. (2011). Changes in the chemical composition of traditional Chinese-type soy sauce at different stages of manufacture and its relation to taste. International Journal of Food Science and Technology, 46(2), pp. 243-249.

Return to reference 16 referrer

Reference 17