Estimation of Aflatoxin M1 Exposure through Consumption of Various Dairy Milk Products in Yogyakarta, Indonesia

This study was conducted to investigate the occurrence of aflatoxin M1 (AFM1) in various milk products marketed in Yogyakarta Province (Indonesia) and to estimate the exposure of aflatoxin through contaminated milk consumption. Fresh milk (n=20), pasteurized milk (n=16), and recombined milk products (n=6) were sampled for AFM1 concentration testing by a competitive ELISA test using ELISA kit for AFM1 assay. A survey was conducted to interview consumers (n=88) on milk consumption habit (milk type and amount of consumption). ELISA assays showed 92.5% of samples werecontaminated with AFM1 in a range of 24-570 ng/L (average: 216 ng/L). The highest average AFM1 concentration was detected in pasteurized milk sample (244 ng/L), fol-lowed by fresh milk (219 ng/L), and the lowest was in recombined milk sample (131 ng/L). How-ever, 100% of recombined milk samples had AFM1 concentration >50-500 ng/L. Thus, recombined milk product was most likely the main source of AFM1 intake due to its high daily consumption in all age groups. Based on AFM1 levels found in milk and consumption of corresponding milk sample, it was estimated that the overall AFM1 exposure ranges from 1.23 ng/kg body weight/day (in 6-15 year-old children) up to 5.26 ng/kg body weight/day (in 3-5 year-old children). In conclusion, this study revealed high occurrences of AFM1 dairy milk marketed in Yogyakarta. Although levels of AFM1 contamination were in Indonesian regulatory limit, high exposure of aflatoxin found in all age groups of consumer. Thus, this preliminary study provides evidence that AFM1 contaminated milk is a serious public health hazard in Indonesia.


INTRODUCTION
Aflatoxin is a secondary metabolite generally produced by toxigenic strains of fungi, Aspergillus flavus and A. parasiticus. Among the group of mycotoxin, aflatoxin B1 (AFB1) is the most toxic and carcinogenic (IARC, 2002). Temperature and moisture in the tropical climate favor the growth of and aflatoxin production by toxigenic strains of molds. A 3-year survey conducted by Rodriques and Naehrer (2012) indicated high aflatoxin contamination in feedstuffs and feed on the equatorial region.
Aflatoxin was present in 82% of corn samples collected from South Asia and in 71% of samples collected from South-East Asia. Surveys on aflatoxin contamination in feedstuffs have been conducted in Indonesia and reveal high occurrence and levels of aflatoxin contamination (Pranowo et al., 2013;Sumantri et al., 2017).
In lactating animals, consumption of AFB1-contaminated feed will result in excretion of aflatoxin metabolites into milk, especially aflatoxin M1 (AFM1), which is similar to AFB1 and is classified as a human carcinogen by IARC (Kos et al., 2014). The carry-over of aflatoxin from feed and metabolite transfer into milk are very worrying because milk constitutes the main food of babies and children, the two age groups that are most vulnerable to the negative impacts of aflatoxin exposure (Mohammadi, 2011). Aflatoxin exposure on human will cause teratogenicity, immu-notoxicity, hepatotoxicity, and mortality. In children, aflatoxin exposure can result in impaired growth (Quintana et al., 2012).
Because of this risk, many countries have imposed regular aflatoxin checks of dairy feed to prevent aflatoxin residues entering the food chain (Volkel et al., 2011). Indonesia National Agency of Drug and Food Control (BPOM) has established the maximum AFM1 concentration in milk, namely 500 ng/L (BPOM, 2009). However, only a few surveys were conducted to investigate aflatoxin contamination in dairy milk products in Indonesia and none assessed on the AFM1 intake.
Expanding of middle class in Indonesia leads to strong growing of dairy milk consumption, which is reported 15% increases (GAIN, 2016). Therefore, current survey was conducted to investigate the presences and levels of AFM1 in dairy milk products marketed in Yogyakarta (Indonesia). Also, to estimate AFM1 intake and describe the consumer risk of aflatoxin exposure through consumption of contaminated milk. Yogyakarta city is important in this study due to dairy milk consumption is significantly influenced by age and education level of the consumer (Widiati et al., 2013). This preliminary study on aflatoxin exposure will provide evidence the potential hazard of aflatoxin contaminated milk on food safety in Indonesia.

AFM1 Contamination in Various Milk Products
Results indicated 92.5% of milk samples are contaminated with AFM1, with the content ranging from 24 to 570 ng/L (average: 216 ng/L) (Table 1). This demonstrates the high occurrence of aflatoxin residue in dairy milk marketed in Yogyakarta. According to Indonesian national standard, only 2.5% of sample surpassed the maximum regulatory limit (500 ng/L). Mean AFM1 concentration in milk in this study (216 ng/L) was higher than that reported in a previous survey by Nuryono et al. (2009), who found the mean AFM1 concentration in fresh milk samples collected from Yogyakarta to be 8.5 ng/L, with no samples with AFM1 levels >50 ng/ L.
Although recombined milk products were produced in Indonesia, the dried milk for recombined milk production is most likely imported milk product because there is no local nonfat dry milk production in Indonesia (GAIN, 2017). Refer to the European Union Commision Regulation (EC), 97.5% of the positive sample contained AFM1 > 25 ng/L and 92.5% of positive samples contained AFM1 >50 ng/L ( Table 2).
The concentration of AFM1 in milk shows variations according to geographical region, season, environmental conditions, level of development, dairy farming system, and consumption of feed concentrates (Becker-Algeri et al., 2016). In Iran, AFM1 levels in milk were decreased at the end of summer due to more consumption of grass than concentrate feed (Darsanaki et al., 2013). In Thailand, AFM1 concentration in pasteurized milk collected in rainy season is higher than milk samples collected in summer (Ruangwises and Ruangwises, 2009). Study of Widiastuti and Anastasia (2018) on small holder dairy farmer in Lampung and West Java Provinces of Indonesia found that the occurences of AFM1 contaminations in fresh milk are higher in dry season (40%) than in rainy season (2%) because the use of concentrate feed when fresh grass is limited in dry season. combined milk is poured into Erlenmeyer then shaken gently. Five millilitres of homogenized samples were poured into 5 mL tubes and stored in the freezer.

Quantification of AFM1 in Milk Sample
AFM1 concentrations in milk were analysed using Enzyme-Linked Immunosorbent Assay (ELISA) test. Previously, milk samples were defatted by centrifugation at 3000 x g for 10 min at 4 o C. As much as 0.4 mL of skim milk was mixed with 0.1 mL of 100% methanol. ELISA assay for AFM1 was performed using ELISA kit AgraQuant® Aflatoxin M1 Sensitive 25/500 (Romer Labs, Singapore).
ELISA was performed as described in the ELISA kit protocols. The absorbance values for standards and samples were measured using an ELISA reader, and AFM1 concentrations were calculated based on a semilogarithmic equation derived from the standard curve.

Estimation of AFM1 Exposure through Milk Consumption
To estimate the mean intake of AFM1 (ng/kg body weight/day) through milk consumption, an interview using questionnaire was done. The consumers (n=88) were interviewed to obtain information on milk preference, amount of milk consumption per day, body weight, and age. AFM1 intake was estimated by multiplying daily milk consumption with AFM1 content in the milk sample consumed by the respondent. Respondents were between 3-55 years old and selected randomly from consumers in milk bar and milk retailer who agree to be interviewed in this study.

Data Analysis
Milk sampling and respondent interviewes were randomly obtained. Data were statistically analysed using IBM SPSS Statistics software version 21.0 (IBM Corp., USA). Mean concentrations of AFM1 between milk types and also mean of AFM1 intakes and AFM1 exposures between age groups were compared using Kruskal-Wallis test considering the samples are heterogenous.
This study revealed that AFM1 concentrations in pasteurized milk served in milk bar and the restaurant was higher than in fresh milk collected from dairy cooperatives. Although recombined milk had the lowest AFM1 concentration compare to fresh and pasteurised milk samples, recombined milk was the most likely source of AFM1 for the consumer as the AFM1 level in all recombined milk samples exceeded 50 ng/L and higher consumption of recombined milk than pasteurised milk in all of respondent groups.
This finding indicated there were no significant different of AFM1 concentration between milk type. Highlight was should be addressed to recombined milk (branded milk products) due to high frequent of AFM1 contamination. However, only six samples of branded milk products were analyzed in this study that is not sufficient to give general conclusion. High frequent of AFM1 contamination in branded milk products was also reported by Wijaya et al. (2017) who found AFM1 contaminations in 85% of powdered milk samples and 68% of sweetened condensed milk samples that were collected in several cities in Java. Their survey indicated 55% of powdered milk samples contain AFM1 between 50-200 ng/kg. This evidence was worrying because the main consumer of recombined milk in this survey was children, an age group that is most vulnerable to aflatoxin exposure (Mohammadi, 2011).

Estimation of Aflatoxin Exposure
Here in, 88 respondents belonged to different age groups, between 3 to 55 years old. The consumers were interviewed to obtain information on milk preference, amount of milk consumption per day, body weight, and age. Estimation of AFM1 exposure for each group of the consumer was calculated based on the amount of daily milk consumption (Table 3) multiplied with an AFM1 concentration in the corresponding milk samples and divided by body weight (Table 4.). Calculations indicated high aflatoxin exposure to all of the age groups of milk consumer, especially in age groups 3-5 years old, which had AFM1 exposure more than 5 ng/kg body weight/d (Table 4).
Our calculations indicated high AFM1 exposure to all of the age groups of milk consumer, especially in age groups 3-5 years old, which had AFM1 exposure more than 5 ng/kg body weight/ d. High AFM1 intakes were found due to high recombined milk consumption, especially in the age groups of 3-5 year-old. Several works have suggested that dairy milk could be regarded as a hazardous source of AFM1 for children due to the high regular intake of milk in this age group (El-Tras et al., 2011). Because of the carcinogenic potential of aflatoxins, no admissible aflatoxin daily intake has been specified by international expert committees. However, it was concluded that daily aflatoxin exposure of as little as <1 ng/kg body weight contributed to an increased risk of liver cancer (Duarte et al., 2013). Regular intake of a low dose of AFB1 was reported as responsible for the development of liver cancer, kwashiorkor, and impairment of child growth (Bhat et al., 2010).
Exposure to AFM1 poses a threat to human health because its carcinogenic potential is similar to that of AFB1 (IARC, 2002). AFM1 is resistant to heat treatment and acidic conditions in fermented milk; thus AFM1 is found in dairy products (yogurt, cheese, cream or butter) made from contaminated milk (Kos et al., 2014). Several ways have been developed to reduce aflatoxin transfer from feed into the milk. Maki et al. (2016) reported feeding adsorbent could reduce AFM1 excretion and concentration in milk.
Annual milk consumption per capita in Indonesia is low (14.3 liters), but considering to income growth and increasing middle class, milk consumption in Indonesia is increasing higher than other regional countries (GAIN, 2016). Therefore, aflatoxin exposure could be higher than this preliminary study if there is no strict inspection and control in a dairy production system to prevent a high level of AFM1 concentration in milk.

CONCLUSIONS
The frequent occurrence of AFM1 contamination was found in this study, especially in recombined milk which containing AFM1 >50 ng/L. The survey also indicated high consumer exposure to aflatoxin through the consumption of contaminated milk in Yogyakarta. This study could be evidence that contaminated milk is a substantial hazardous source of carcinogenic substance in Indonesia. Therefore, increased effort should be directed toward preventing the AFB1 transfer from dairy feed into milk, as well government and consumer awareness of the aflatoxin risk.