Abstract

The drive for this study stemmed from the need to continue to monitor acrylamide ingestion from foods known to contain high loads, such as bread. Three main types of bread were sampled in a survey as part of consumption data using a food frequency questionnaire administered to a cross-section of consumer groups (males, females, adults, and children) in and around a mall. Acrylamide was extracted from the samples using the QuEChERS method, cleaned up, and quantified using HPLC. Palisade @Risk software was used to fit distributions of the acrylamide content and the other elements of exposure (body weight and mass of bread consumed per day). The exposure was subsequently integrated with the US EPA guidelines to determine the hazard quotient (HQ), the margin of exposure (MoE), and cancer risk (R) of the consumer groups. Subsequently, these indices were simulated at 10⁵ iterations. The results showed that acrylamide was detected in 100% of the samples analyzed, ranging from 7.0×10^-5 mg/g to 1.59×10^-3 mg/g, and that “Wheat bran bread” showed the highest levels. These risk indices of the presence of acrylamide indicated that the frequent consumption of bread could be of public health concern for genotoxicity during consumers' life stages. Though there were isolated cases of low public health concerns (MoE >10⁴), a worrying trend of children consumers showing a frequently occurring (modal) higher life-stage cancer risk (5×10^-4), that increases to even more serious risk (4×10^-3) at the 95^th percentile cannot be overlooked.

1. Introduction

Acrylamide is ubiquitous in frequently consumed foods such as potato chips, breakfast cereals, crispbread, biscuits, and crackers ¹. Baking conditions (temperatures and times) required for producing baked products often lead to acrylamide formation ². Acrylamide is a reactive molecule that can react by ionic and free radical mechanisms ³. It is known to have both neurotoxic and carcinogenic effects in animals and humans. It has been classified as “probably carcinogenic to humans” ⁴. After ingestion, absorption is effective through oral means in all species ⁵.

The occurrence of acrylamide in staple foods has drawn concern globally due to its adverse health effects. The European commission in 2010 recommended that acrylamide levels in foods be monitored periodically. Relative to developed countries, acrylamide exposure and risk assessment are not well documented in developing countries. Thus, leaving a considerable gap and heightened public health concerns.

A comprehensive risk assessment includes several steps: hazard identification, hazard evaluation, exposure assessment, and risk characterization. The risk associated with acrylamide consumption can be quantified using the hazard quotient (HQ), the ratio of the chronic human exposures to the reference dose of acrylamide ⁶. HQ estimates the non-carcinogenic risk of a known hazard to consumers ⁷, a policy cut-off that indicates that sensitive populations may begin to exhibit mild adverse responses. The MoE approach employs the benchmark dose limit, lower bound (BMDL₁₀) value, obtained from animal studies to determine whether the estimated human exposure to the hazards is a cause for public health concern ⁸. The wider the gap between the threshold value and the experimentally determined exposure, the greater the assurance of minimum risk and consequently less public health concern. A margin of exposure (MoE) >10⁵ for a carcinogen and genotoxic material indicates low public health concern ^{9, 10}. Relatively, for non-genotoxic adverse effect, a limit greater than 100 indicate less cause for public health concern ¹¹. Carcinogenic risk, often determined as lifetime risk, is the product of the hazard's potency factor (slope factor) and the exposure. The potency factor, derived from the regulatory database, represents the risk estimated to be produced by a lifetime average dose of 1 mg/kg(bw)-d of that particular hazard. For acrylamide, a value of 0.5 mg/kg(bw)-d has been reported ¹².

The Maillard reaction is the most likely mechanism of acrylamide formation in food since acrylamide formation is mainly favored by high temperatures and relatively low moisture content ¹³. The first stage in acrylamide formation is the Schiff's base product between the carbonyl and the R-amino group of asparagine ¹⁴. Studies support the Schiff's base mechanism that explains acrylamide originates from the reaction of the amino acid asparagine with the carbonyl carbon of reducing sugars, such as fructose and glucose ¹⁵. Thus, acrylamide is particularly common in fried and baked foods or products that show high levels of free asparagine ². Acrylamide can also be generated from lipid-rich foods through reactions between ammonia and acrolein.

Further, the kind of oil used during deep frying can also affect the process of acrylamide formation ¹⁶. Another study showed that the product thickness is an essential parameter determining the drying rate and, subsequently, the rate of acrylamide formation during baking ¹⁷. Mechanistically, a substantial fraction of the ingested acrylamide is biotransformed to the genotoxic epoxide known as glycidamide. Glycidamide is an even more reactive carcinogen and reportedly 100-1000 times more reactive with DNA than acrylamide ¹⁸. Many adducts of glycidamide with purine bases of DNA have been isolated from the DNA of the lungs, liver, and kidney of mice dosed with acrylamide ¹⁹. Acrylamide and glycidamide are excreted from rats approximately two hours after ingestion ⁸.

The Joint FAO/WHO Expert Committee on Food Additives (JECFA) has presented the results of numerous animal models on the health risks associated with acrylamide exposures. The committee observed a no-observed-adverse-effect level (NOAEL) of 0.2 mg/kg(bw)-d for morphological changes in rats' nerves ²⁰. Experts have also proposed a benchmark dose, lower bound 10%(BMDL₁₀) of 0.31 mg/kg(bw)-d acrylamide as capable of inducing mammary tumors in rats and 0.18 mg/kg(bw)-d to induce Harderian gland tumors in mice ²¹. The metabolism and elimination of acrylamide are varied in different species; therefore, there may be some variations in the sensitivity of humans to acrylamide exposures ²².

A recent study in Europe indicates that breakfast cereals, bread, crackers, and other cereal-based products contributed up to 25% of acrylamide ⁸. A survey of commonly eaten snacks shows that bread contributes 10-30% of acrylamide ingestion in Europeans ²³. Indeed a monitoring report in Europe revealed that acrylamide levels ranged from 3×10^-5 to 4.7×10^-3 mg/g, depending on the product type ²³. In the case of adults, the estimated average ingestions ranged from about 3×10^-5 to 6×10^-5 mg/kg(bw)-d. Infants and children are exposed relatively more than adults because of their lower body weights ²⁴. Studies also indicate that other likely dietary exposures to acrylamide are from foods such as potato chips (1×10^-3 mg/g). The acrylamide concentration is higher in baked samples formulated with ammonium carbonate as a rising agent ²⁵. The European Food Safety Authority (EFSA) also monitored the levels of acrylamide in bread between 2007 and 2010 and reported the levels of acrylamide to range from 3.0×10^-5 to 4.25×10^-4 mg/g ²⁶. There are many bread consumers in developing countries such as Ghana, and this warrants continuous monitoring of acrylamide exposure for policy decisions. Thus, the study's objective was to determine acrylamide exposures in bread consumption and estimate the public health concerns.

2. Materials and Methods

Samples of three types of commonly consumed bread, “Butter bread”, “Sugar bread”, and "Wheat bran bread", were collected from three malls in the Greater Accra region of Ghana. These samples were collected in plastic bags every day and refrigerated for three weeks before analysis.

Acetonitrile and hexane were acquired from Merck (Darmstadt, Germany, and Prolabo VWRI, France. The salts used (NaCl and MgSO₄) were obtained from Sigma Aldrich (Germany), and acrylamide standards were purchased from Acros Organics (USA).

The study area was Accra, and it was chosen because it attracts many people from different socioeconomic backgrounds. The Accra Mall, West Hills Mall, and the Junction Mall were selected as points for the survey since, over 3-4 years, they have become the hub of commercial activities for the middle class. A random sampling approach was used to collect data from consumers from 25^th June to 25^th July 2019. It is only the willing participants who participated in the bread consumption survey. The quantity of bread consumed (taken as cost of bread purchased) plus such biodata as age, body weight, and gender were also collected. To facilitate the completion of the questionnaire, two experienced, and trained assistants who had previously gone through a similar survey were recruited. The frequently used dialects, Twi and Ga, were used to administer 301 questionnaires to respondents.

The QuEChERS method, which was used to extract acrylamide ²⁷, involved homogenizing a 2 g bread sample with 4000 mg of MgSO₄ and 1000 mg of NaCl in a 50 mL centrifuge tubes. Five (5) milliliters of hexane were added and vortexed, followed by adding 10 mL of acetonitrile and 10 mL of distilled and further vortexed for 1 min. The mixture was centrifuged for 5 min at 3000 rpm, and 1 ml of the resulting supernatant was taken and treated with another round of 1500 mg MgSO₄ and 500 mg NaCl. The mixture was vortexed for 5 min and centrifuged at 4000 rpm. Two (2) milliliters of the supernatant were then siphoned for HPLC analysis.

Following the recommended protocol ²⁸, acrylamide was quantified using an HPLC system made of Cecil-Adept binary pump coupled to an absorbance detector. The Agilent eclipse plus C18 column was housed in an oven set at 25°C. Water and acetonitrile in a 20:80 v/v ratio were employed as the mobile phase and adjusted to pH 3.5 with orthophosphoric acid. A 60 µL of the extracted sample was then injected into the HPLC. The mobile phase had a flow rate of 1 mL/min, and the analytes were detected at 223 nm. Detection and quantification were done by comparing the observed peaks with standard retention time. The Cecil-Adept PowerStream (CE 4300, the UK) automatically integrated the area under the peaks expressing it as acrylamide concentration in the samples. Quality control was ensured using 2 g each of different concentrations of analytical starch (20, 50, and 100 µg) containing standardized acrylamide in recovery studies. Recoveries of 97% were obtained, indicating the accuracy and sufficiency of the method used ²⁹. The limits of detection (LOD) were 0.03 µg/g (min) and 0.1 µg/g (max). The curve had an r² of 0.998 and was linear.

The data obtained from the consumer survey were captured into a spreadsheet in Microsoft Excel and grouped according to inputs such as acrylamide concentration (C), the mass of bread consumed (M_c), body weights (B_W), gender, infants and children (≤19 years), and adults (≥20 years). Palisade @Risk software was then used to fit the distributions, and the average acrylamide exposure was quantified using Equation 1.

(1)

Hazard quotient (HQ) was quantified as the ratio of the average exposure of acrylamide to its reference dose (R_fD) of 2.0×10^-3 mg/kg(bw)-d per the protocols of regional screening levels ⁶ and estimated using Equation 2.

(2)

The margin of exposure (MoE) of acrylamide exposure was also determined (Equation 3) per EFSA protocols ³⁰. As proposed by regulators, respective BMDL₁₀ values of 0.17 mg/kg(bw)-d for genotoxicity and carcinogenicity and 0.43 mg/kg(bw)-d for risk of neurotoxicity of acrylamide were used ⁸.

(3)

Life-stage cancer risk (R) of acrylamide was quantified as in Equation 4 following regulation ⁶, where a potency factor (PF) of acrylamide (0.5 mg/kg(bw)-d) was used as a secondary data ¹².

(4)

All the risk indices were iterated 10⁵ times using version 7.6 of Palisade @risk software ³¹.

3. Results and Discussion

Acrylamide was detected in all 106 samples of bread collected, of which 38 were “Sugar bread”, 39 were “Wheat bran bread”, and the remaining 28 were “Butter bread”. The acrylamide content ranged from a minimum of 7.0×10^-5 mg/g to a maximum of 1.59×10^-3 mg/g (Table 1). The sampled “Wheat bran bread” had the highest acrylamide content relative to "Butter bread", presenting the least acrylamide concentrations. A study in Sweden indicated that regular bread products contained between 5×10^-5 mg/g and 1.7×10^-3 mg/g of acrylamide ³². Relatively, other studies indicate different bread types presenting acrylamide concentrations ranging from 3 ×10^-5 to 2.5×10^-3 mg/g ³³. These findings are similar to what was quantified in this study.

The results show that “Wheat bran bread” was the most frequent bread type that had the highest levels of acrylamide (modal: 0.36×10^-3 mg/g; 95th percentile: 1.50×10^-3 mg/g). This observation was similar to what was detected in some bread types in Sweden (0.79×10^-3 mg/g) and Germany (0.92×10^-3 mg/g) ³². The high acrylamide content of “Wheat bran bread” can be attributed to one of the ingredients for making the dough; the bran. The observation is based on studies that confirmed the presence of considerable amounts of asparagine, an amino acid which is a significant precursor of acrylamide formation in baked cereal products ³⁴.

Based on data collected from 2007 to 2010 in Europe, it has been revealed that the acrylamide content of diets ranges from 3×10^-5 to 4.7×10^-3 mg/g ²⁶. The results from the current study fall within what was published in the European study, showing that acrylamide concentration in diets may proportionally vary according to processing methods.

The statistical distributions resulting from fitting the elements of exposure to acrylamide in bread are indicated in Table 2. These distributions are simply mathematical expressions of data collection, and their scores on variables are usually arranged in order from smallest to largest. The masses of bread consumed varied from a minimum of 20 g to a maximum of 365 g among respondents classified as males, females, and children. The most frequently consumed(modal) mass, ranged between 50 g and 60 g. Generally, the top 5% (95^th percentile) of adult respondents consumed 300 g of bread relative to children respondents that consumed 200 g. Among the groups of consumers (Table 2), the statistical distribution of bread consumption in adults registered as Kumaraswamy, and presented the highest modal mass of 60 g and a 95^th percentile mass of 300 g of bread consumed. In children, however, the distribution of the mass of bread consumed presented as Invgaus, giving a low modal value (50 g) but relatively high 95^th percentile (200 g) consumption. Between gender, however, there were no significant differences. This observation emphasizes the variability among eating behaviors conceptualized as biological factors, peer pressure, neighborhood, institutional settings, food culture, and economic trends ³⁵.

The simulated acrylamide exposure to consumers through bread (Table 3) ranged from a low of 1.8×10^-4 mg/g (5^th percentile) to a high of 1.17×10^-3 mg/g (95^th percentile). The mean acrylamide concentration in the sampled bread was 0.47×10^-3 mg/g. However, the most prevalent (modal) acrylamide concentration was lower (0.31×10^-3 mg/g). These varying concentrations yielded acrylamide exposure, from a low of 2.4×10^-4 mg/kg-d (5^th percentile), in male consumers up to 1.10×10^-2 mg/kg-d in female consumers (95^th percentile). A study in Finland presented a higher mean exposure to acrylamide through bread as 6.7×10^-4 mg/kg(bw)-d ³⁶. The mean exposure obtained in this current study ranged from a high of 3.1×10^-3 mg/kg(bw)-d in female consumers down to 1.5×10^-3 mg/kg(bw)-d in adults. Thus, the exposure to acrylamide in the current study is relatively higher than the mean concentrations (3.4-0.4 µg/kg(bw)-d) ⁸, reported across consumer groups in Europe. This difference may likely be due to different baking processes. It has been explained that this high exposure is because of the traditional female consumption of more bread-based breakfast relative to their male counterparts, who instead prefer other traditional foods not based on bread ³⁷. The levels of acrylamide exposure in our female consumers are consistently higher relative to a study in Norway. In this particular study, a mean dietary acrylamide exposure for the entire population stood at 5.3×10^-4 mg/kg(bw)-d, relative to a lower value of 5.0×10^-4 mg/kg(bw)-d for female consumers ³⁸.

In other study areas for consumers aged between 4 and 65 years, acrylamide exposures compared with the WHO acceptable limit. Among such areas include Germany (5.7×10^-4 mg/kg(bw)-d), Finland (4.0×10^-4 mg/kg(bw)-d) and Poland (4.3×10^-4 mg/kg(bw)-d) ^{5, 39}. The outcome of a consultative meeting organized by WHO and FAO in 2002, with data available from Switzerland, Sweden, Norway, the USA, and the United Kingdom, indicate that children in developed countries were more exposed to acrylamide ²¹. It was also suggested that children would generally have exposures two to three times those of adults when the total sum of consumed foods is considered. The findings in this study (Table 3) compare with this observation.

According to Table 3, all respondents that fell in the bottom 5% of consumers of bread (5^th percentile) do not seem to present risk (HQ<1). All top 5% consumers (95^th percentile), however, presented risk (HQ>1).

The result is similar to an acrylamide exposure study of general food consumption in Ghana, which recorded HQ values, showing a spread of safe (HQ<1) to unsafe (HQ>1) acrylamide exposures ⁴⁰. It confirms FAO/WHO the position that there is an urgent need to control the presence of acrylamide in foods, especially starting from sources that produce excessive quantities ⁴¹.

The most frequently (modal) occurring MoE indices of public concern for genotoxicity/carcinogenicity in the study area (Table 3), presented all consumers to be at risk since all the MoE indices were low(<10⁵) according to regulation ¹². This observation is similar to the mean MoE reported in Poland, which showed severe adverse public concern for infants and children consumer groups ⁴². Perhaps there would be the need to finally compel bakeries to seek measures to control acrylamide if this trend is not controlled.

Neurotoxicity, however, presented a more moderate outlook. Across the subgroup of consumers (adults/children and male/female), there were isolated consumers in the study area that seemed to show low public health concerns because of their max MoE >10⁵ indices (Table 3). Thus, the public health concern of neurotoxicity of acrylamide, based on the modal MoE indices between adults and children on the one hand, and males and females, on the other hand, show a marginally low safety concern.

In fact, according to regulation ¹², MoE indicators for neurological effect must be greater than 1075 for high adult consumers and 126 for high infant and children consumers before they can be considered to have moderate to low public health concerns. However, this should not imply that consumers in the study area are safe from the adverse effects of acrylamide because the total diet might contain other sources of acrylamide apart from bread ⁴³.

The frequently occurring(modal) cancer risks per ten thousand consumers, as indicated among all the consumers' groups (Table 3), followed the order: children (5×10^-4)> adult (3×10^-4)> gender (2×10^-4). Though the modal likely cancer risks ranged from a high of 5×10^-4 to a low of 2×10^-4, the risk still appears to be moderate according to the US EPA’s generally acceptable risk range (10^-6-10^-4) ⁴⁴. However, the acceptance of this range of risk by the US EPA is in the context that the study area must be low risk for the hazard in question but not when there are other multiple sources of acrylamide ⁴³; therefore, the level of the modal risk may be regarded as unacceptable. Also, the results show consistent exposure of children at a higher modal (5×10^-4) and 95^th percentile (4×10^-3) cancer risks, though, between gender (male/female), the risks were the same (Table 3). The higher risk in children may be attributed to their body weights relative to the amount of acrylamide ingested ⁸. This observation is similar to what was reported in similar in a study in China, where the risks were the same among gender ²⁹.

4. Conclusion

Bread consumption in the study area was relatively high in both adults and children however, consumption was higher in adults. Thus, a significant amount of acrylamide is ingested, though giving a low modal concentration (0.31×10^-3 mg/g). However, these low concentrations of acrylamide translate into a moderately low to high neurotoxicity risk outlook across the subgroup of consumers. There are also cases of determined MoE values exceeding the maximum limit, 10⁴, which sometimes shows low public health concerns. Comparatively, children were at a higher modal life-stage cancer risk relative to adults. Since these indices present significant risks, there is a justified cause for public health attention, warranting mitigation measures to control acrylamide since other sources of the hazard may be ingested in a total diet.

Conflict of Interest

The authors declare that there is no conflict of interest.

Funding

This research received no specific grant from any public, commercial, private, or not-for-profit funding agency.

Contribution of Authors

David O. Asamoa-Agayre collected the data and drafted the manuscript.Naa Kwarley-Aba Quartey contributed significantly to editing the final manuscript. Isaac W. Ofosu designed the study, worked on the final manuscript and made significant corrections before submission.

References