Non-Conjugated-Industrially-Produced-Trans Fatty in Lebanese Foods: The Case of Elaidic and Linolelaidic Acids

To determine Industrially-Produced Trans fatty acids (IP-TFAs) distribution of Lebanese traditional foods, especially regarding Elaidic acid (EA; 9t18:1) and Linolelaidic acid (LEA; 9t12t18:2), a mapping exercise was enrolled between January 2019 and April 2021 in which 145 food samples of three categories (traditional dishes, Arabic sweets, and market food products) were analyzed using Gas chromatography methods. Results showed that about 93% of the products tested in Lebanon, between 2019 and 2021, met the World Health Organization recommendations, while about 7% exceeded the limit. The mean level of the IP-TFAs Elaidic and Linolelaidic acid in most Traditional dishes (0.9%), Arabic sweets (0.6%), butter and margarine (1.6%), and market foods (0.52%) were relatively low compared with other countries. Despite that, the relative impact of IP-TFAs on heart diseases mortality in Lebanon is limited but unambiguously still substantial. The persistence of food products with high IP-TFAs levels threatens the health of Lebanese people. Fortunately, this problem is fairly easy to solve in Lebanon via proper legislation.


Introduction
The intake of Industrially-produced-Trans fatty acids (IP-TFAs) is associated with an increased risk of heart attacks and death from coronary heart disease (CHD) [1]. A 2% absolute increase in energy intake from IP-TFAs has been associated with a 23% increase in cardiovascular risk [2]. In 2018, IP-TFAs elimination was identified as one of the priority targets in the World Health Organization (WHO) 13th General Programme of Work, which guides the five-year work of WHO in 2019-2023 [3]. Also, in 2018, the REPLACE action package was launched to help countries removing IP-TFAs from their food supplies [4]. In addition, WHO released additional resources in 2019 to support country actions, including six implementation modules and a live policy tracking map-the TFAs Country Score Card 1-to monitor global progress towards the 2023 target [3]. In 2020, WHO established an indicator that records whether countries have adopted WHO best-practice policies for eliminating IP-TFAs [5]. Around fifty-eight countries have introduced laws to date that will protect more than 3 billion people from TFAs by the end of 2021 [3]. However, more than 100 countries have yet to act to eliminate TFAs from their national food supply and make the world TFAs free by 2023 [3]. The European Region has the largest number of mandatory TFA limits in place and has had the most policy progress of all WHO regions since 2019. Since Denmark's effort (2004), Austria (2009), Iceland (2011), Hungary (2014), Norway (2014), Latvia (2018), Slovenia (2018) [6], and New Zealand (2008) have passed similar bestpractice regulations [7]. Switzerland, one of the first countries in Europe to take legal action using culinary skills [18][19][20]. A total of 30 traditional composite dishes were identified as most frequently consumed and hence were included for analysis. The names of the food dishes were reported in the current analysis considering the most familiar name used for the dish at a national level with respect to its ingredients. The ingredients of these traditional dishes were described in Hoteit et al. [18][19][20], and the food samples were collected from five different central kitchens in the 5 governorates listed above. The central kitchens were randomly chosen based on (1) their specialties in cooking homemade dishes, (2) their popularity in the area, (3) their implications in social entrepreneurship and women empowerment (e.g., household women who cook for these central kitchens). Consequently, the food samples were classified into 5 strata, per governorate area [20]. The samples were identified according to their frequency of consumption [21,22] and selected for IP-TFAs analysis mainly for two non-conjugated fatty acids (EA and LEA). In contrast, the subsequent samples collections 2020, conducted in April, were centrally coordinated at Beirut having the broadest coverage in terms of products selected and had a sample of thirty-five types of Arabic sweets and forty-six types of market food products. The full methodology of food list identifications and food sampling is described elsewhere [18][19][20]. On the other hand, the 2021 sample collections, conducted in March, were nationally coordinated, with a coverage of 34 available types of butter and margarines purchased from all the Lebanese markets. Lot numbers were checked to ensure that each unit belonged to a different lot. The samples were stored, labeled, and analyzed before expiry dates. Samples were selected to include all types of butter and margarines in Lebanon. The analyses were carried out in duplicate for each sample. Thus, a composite sample from each type of food, according to each governorate, was prepared and analyzed. To further interpret current levels of IP-TFAs in Lebanese foods, product categories were compared with similar products found in other countries. A graphical scheme for the whole study is shown in Figure 1. collection, we collected thirty types of traditional dishes. Traditional composite dishes are defined as dishes consumed at main meals (i.e., lunch or dinner), containing ingredients from at least three of the five main food groups and requiring preparation using culinary skills [18][19][20]. A total of 30 traditional composite dishes were identified as most frequently consumed and hence were included for analysis. The names of the food dishes were reported in the current analysis considering the most familiar name used for the dish at a national level with respect to its ingredients. The ingredients of these traditional dishes were described in Hoteit et al. [18][19][20], and the food samples were collected from five different central kitchens in the 5 governorates listed above. The central kitchens were randomly chosen based on (1) their specialties in cooking homemade dishes, (2) their popularity in the area, (3) their implications in social entrepreneurship and women empowerment (e.g., household women who cook for these central kitchens). Consequently, the food samples were classified into 5 strata, per governorate area [20]. The samples were identified according to their frequency of consumption [21,22] and selected for IP-TFAs analysis mainly for two non-conjugated fatty acids (EA and LEA). In contrast, the subsequent samples collections 2020, conducted in April, were centrally coordinated at Beirut having the broadest coverage in terms of products selected and had a sample of thirtyfive types of Arabic sweets and forty-six types of market food products. The full methodology of food list identifications and food sampling is described elsewhere [18][19][20]. On the other hand, the 2021 sample collections, conducted in March, were nationally coordinated, with a coverage of 34 available types of butter and margarines purchased from all the Lebanese markets. Lot numbers were checked to ensure that each unit belonged to a different lot. The samples were stored, labeled, and analyzed before expiry dates. Samples were selected to include all types of butter and margarines in Lebanon. The analyses were carried out in duplicate for each sample. Thus, a composite sample from each type of food, according to each governorate, was prepared and analyzed. To further interpret current levels of IP-TFAs in Lebanese foods, product categories were compared with similar products found in other countries. A graphical scheme for the whole study is shown in Figure  1.

Laboratory Analysis Protocol
Around 500 g of each sample was mashed, then analyzed, and the remaining samples were kept frozen at −18 • C for further analysis. The fatty acid profile was measured using gas chromatography. The IP-TFA analysis method was selected considering guidance from the technical committee at the Industrial Research Institute laboratories in Beirut and following standardized protocols. The Association of Official Analytical Chemists (AOAC) methods were used for the analysis of nutrients in food matrices [23].
Soxhlet extraction (Total fatty acids extraction): The Roese-Gotlieb method was used in the investigation of the fat content [24]. Between 1-2 g of the dried food sample was filtered using a piece of filter paper. Later, it was wrapped and introduced into the soxhlet thimble. To avoid sample spilling, a cotton plug was placed at the top of the thimble. The soxhlet apparatus was assembled, and light petroleum, hexane, heptane, diethyl ether, or cyclohexane was used. The extraction was performed overnight. If solids from the thimble or sample were found in the solvent extract, a filtration before evaporation into another tarred flask or beaker was performed. Then it was dried to a constant weight, which was attained when successive 1 h drying periods showed additional loss of less than 0.05 fat. The percentage (%) fat was equal to (g) fat × 100 g sample.
Fatty Acid Profile (saturated, unsaturated, trans): The extracted fat of the sample that was obtained during fat determination was used to analyze the fatty acid profile. Between 200-500 mg of the lipid sample was placed in a boiling flask with chips; then, 5 mL of 0.5 M methanolic KOH was added. Esterification was performed by boiling under reflux for 3-5 min. An addition of 15 mL of esterification reagent (2 g NaNH4 + 60 mL methanol + 3 mL conc sulfuric acid) through the condenser was performed, and then the sample was boiled for 15 min. After cooling, there was an addition of 50 mL of distilled water and 25 mL of the solvent. The organic layer was isolated by means of a separatory funnel. Finally, the solvent layer was washed twice with distilled water.
Chromatographic analysis: Using the chromatograph software Chromquest 4.2.34, USA, an integration of the areas under the peak was detected under the standards peaks. A calculation as percentage areas was done. The sum of trans fatty acids was calculated accordingly [24]. TFAs isomers were later on detected through SP-2560 100 m capillary column (180 • C isothermal, H2 at 1.0 mL/min) [23].
Statistical tests: The study variables were presented as continuous variables and listed as reported values per 100g. Means and standard deviations were calculated for the group of traditional dishes, Arabic sweets, and other market products. T-test was used to compare the mean content of the group of foods in terms of EA and LEA. A p-value less than 0.05 indicates a statistically significant difference. Statistical analysis was conducted on IBM SPSS Statistics for Mac, Version 24, USA.

Trans Fatty Acid Acids in Frequently Consumed Traditional Dishes
The mean levels of total IP-TFAs (total of Elaidic acid and Linolelaidic acid) in the tested traditional dishes was equal to 0.9% ± 0.62 and ranged from less than 0.1 to 2.8 g/100 g of total fat except for the dishes Riz a dajaj and Shawarma Lahma in which total IP-TFA exceeded 2% of the total fat ( Table 1). The comparison between the mean values of the IP-TFA (EA and LEA) in the traditional dishes tested shows that EA was significantly higher than LEA in all traditional dishes (p-value = 0.00). Per each governorate, the mean level of total IP-TFAs in all dishes was 0.78 ± 0.65% in Mount Lebanon (range: <0.1-2.1%), 1.1% ± 0.6 in Beqaa (range: <0.1-3%), 1 ± 2.4% in Beirut (range: <0.1-11.2%), 1.2 ± 0.9% in Tripoli (range: <0.1-3.2%) and 0.8 ± 0.4% in Saida (range: <0.1-2%) (Data not shown).

Trans Fatty Acid Acids in Frequently Consumed Arabic Sweets
The average of the total IP-TFAs in all samples of Arabic sweets was 0.6 ± 0.3%, predominantly from the EA type. (Table 2). Among 35 samples of Arabic sweets, none exceeded 2% as total IP-TFA in 100 g of total fat. The comparison between the mean values of the IP-TFA (EA and LEA) in the Arabic sweets tested shows that EA was significantly higher than LEA in Arabic sweets (p-value = 0.00). Table 2. Total fat in 100 g of edible portions, total IP-TFAs and IP-TFAs (EA and LEA) in 100 g of total fat of frequently consumed Arabic sweets.

IP-TFAs in 100 g of Total Fat
Name

Cereals and Breads Group
In the group of cereals and breads, the mean level of total IP-TFAs was less than 2% of total fat except for pain au lait (total IP-TFAs: 3.8%), which is usually prepared from wheat, milk, and butter or ghee to be consumed frequently by children as a sandwich (Tables 3 and 4). Table 3. Total fat in 100 g of edible portions, total IP-TFAs and IP-TFAS (EA and LEA) in 100 g of total fat of Market food products collected from Lebanese markets.

Butter and Margarines
Particular attention was given to the margarine group as it is used as an ingredient and therefore amongst the main sources of IP-TFAs in processed foods. The average of total IP-TFAs in 18 margarines used frequently in Lebanon was 2.4 ± 0.4% (Table 3) with a range between <0.1% and 11.8% (Data not shown). The dominant IP-TFA was EA in almost all these products (Table 3). Within the group of butter, none of the samples exceeded 2% of total fat. The average of total IP-TFAs in the butter and margarines group was 1.6 ± 0.6% of total fat in which EA predominates in these products. Generally, the level of total IP-TFAs in cooking oils, Halawa and Tahina was negligible (Tables 3 and 4).

Snacks and Processed Foods
As for the group of biscuits, doughnuts, and cakes group, negligible amounts of IP-TFAs were found in these products (Average: 0.5% ± 0.2) ( Table 4). On the other hand, the unlabeled English cake (chocolate flavor) had an apparently high amount of total IP-TFAs (2.6% in the total fat) in which EA was dominantly available (Table 3). Despite being unable to discuss the fat type used in unlabeled samples, based on this data, partially hydrogenated fats were certainly present in high amounts.
The data on chocolate products presented an amount of 1.3% ± 0.3 as total IP-TFAs (Figure 1), except for the case of wafer-coated chocolate originally manufactured in Lebanon which contains a level of 6.5% (Table 3).
According to Tables 3 and 4, it appears that all samples of potato chips, nuts, seeds, coffee, instant coffee, and packed tuna contained low amounts of total IP-TFAs that are below 2% of total fat. When comparing the mean values of the IP-TFA (EA and LEA) in the market foods, EA and LEA didn't show any significant difference (p-value = 0.16).

Discussion
Industrially-Produced Trans fatty acid content in frequently consumed foods in Lebanon compared with different countries.
The available data, the first of its kind in Lebanon, demonstrate that categories with the highest IP-TFAs levels included Riz a dajaj, Shawarma Lahma, Pain au lait, English cake, Chocolate wafers, and margarines. About 93% of the products tested in Lebanon, between 2019 and 2021, met the WHO recommendations (less than 2% of Trans fatty acid in total fat), while about 7% exceeded the limit. As per Tables 1-3, all in all, EA was dominant in almost all the analyzed samples and its higher amount indicates that hydrogenated oils were a major contributor in the processing of food products or baking and cooking meals. In comparison to other countries all over the globe, a broad range of EA was observed in many food products (Table 4). For instance, the mean level of EA in Baklava (0.2%) was relatively low in our study in comparison with the content of EA in Baklava in Iran (2.5%) [25]. Furthermore, our findings showed that the mean levels of EA in cakes (2.6%) was much lower than the content of EA found in cakes in France (18.5-25.6%) [26], Iran (6.95-18%) [25][26][27], Poland (7.95%) [28], India (1.92-3.93%) [29], and higher than EA cake content tested in Lebanon in 2015 (1.7%) [30], Korea (1.36%) [31], Turkey (0.37-1.43%) [32], New Zealand (0.9%) [7], and Malaysia (<0.001%) [33] (Table 4). In addition, the mean levels of EA in biscuits in Iran (9-12.86%) [27], Lebanon 2015 (3.7%) [30], Poland (2.81%) [28], Korea (2.4%) [31], New Zealand (0.9%) [7], and Germany (0.18%) [34] were higher than our results (0.1%), except for Malaysia (<0.001%) [33] and India (0.01%) [29] (Table 4). As for the breakfast cereals, the mean level of EA in our study (0.1%) was much lower than in France (28.9-32.4%) [26] and Korea (0.5-6.75%) [31], and higher than in the UK (0.03%) [35] and Malaysia (<0.001%) [33] (Table 4). Moreover, our findings showed that the mean level of EA in chocolate wafers were six times more than EA content in chocolate wafers in Malaysia [33]. As for the butter, the New Zealand [7] and Costa Rican butter [36] contained five times more EA, and the Pakistani butter [37] contained three more times EA, compared with our results (Table 4). However, the butter in UK, Germany, and Iran contained 0.22% [35], 0.23% [34], and 0.3% [27] EA respectively; this is lower than the content of EA tested in our study (0.6%) ( Table 4). Also, Table 4 showed that the margarines in Slovenia contained the highest content of EA (34.63%) [38] compared to our findings (2.2%) and other countries. As for the EA content in chips, Iranian chips showed the highest level of EA (10%) compared to our results (0.1-0.3%) and other countries [27] (Table 4). On the other hand, the results of LEA in the food products tested in our study and those in other countries are available in Table 4.

Comparison between Lebanese Market Basket Investigation and Other Global and Regional Market Investigations
According to many studies, there was an impact of TFAs labeling on reducing the burden of CVDs due to TFAs [39]. According to an unpublished study conducted by our team, 32% only of the products available in the Lebanese markets reported TFAs on their labels (Data not shown). Our finding came to hand by hand with Kamel et al. [40], in which 181 food products were sampled from local supermarkets in Saudi Arabia and showed that one-third of the products mentioned TFAs on the nutrition label. Moreover, while the majority of the investigated samples in our project had low levels of TFAs, up to 14 g of TFAs per 100 g of food was observed in certain oils and fats sold at the Lebanese markets. Our findings, concerning the range of TFAs in-market products, were relatively low compared with the market investigations published in Stender et al. (2019Stender et al. ( -2020 [41,42].

Investigation of the Country of Origin of Imported Food Products in Lebanon
Lebanon imported its food products from France ($107,957), Germany ($98,250), Turkey ($97,015), United Kingdome ($75,571), Italy ($70, 571), Argentina ($69,989), Saudi Arabia ($64, 332) and United States ($57,785). In addition, the main importation sources of butter, oils, and fats are Denmark, Netherlands, France, Belgium, Ukraine, New Zealand, United Kingdom, and Argentina [43]. According to the nutrition labels of tested butter and margarines, the country of origin from which all the butter and margarines were imported to Lebanon were Turkey (n = 5), Egypt (n = 4), Malaysia (n = 3), Saudi Arabia (n = 1), Sri Lanka (n = 3), UAE (n = 1), Netherland (n = 2), Belgium (n = 3), France (n = 4), Italy (n = 1), Ukraine (n = 1), Germany (n = 2), and Denmark (n = 2). Among all these countries, 33 percent (five countries over 15) are implementing mandatory national limits and adopting monitoring mechanisms for mandatory of TFAs limits. On the other hand, in the remaining countries, the bestpractice TFAs policy passed but was not yet in effect [5]. Lebanon, long considered a middle-income country, is rapidly sinking into poverty as it faces a triple shock from the unprecedented economic crisis, the impact of COVID-19 on employment and public health, and the consequences of Beirut port explosions. Despite that, the actual relative impact of IP-TFAs exposure on heart disease mortality in Lebanon is limited, but unambiguously still considerable. The findings in our report highlight the importance of controlling the importation of food products from countries controlling IP-TFAs levels in food to avoid sinking Lebanese markets with IP-TFAs rich food products [44], both of which are often ultra-processed, unhealthy, and rich in IP-TFAs. Therefore, this population group is at higher risk of IP-TFAs-attributable CVDs.

Limits, Advantages, and Future Directions
This study presents some limitations. First, there are many challenges facing the laboratories in Lebanon concerning the testing of IP-TFA, and the lack of standards limits testing other forms of isomers. Second, the food products compared between regions were compared in terms of food groups and not in terms of brands. Moreover, the comparison between traditional dishes or Arabic sweets omits the cooking preparations and ingredients. Third, in the current study, the WHO technique was followed to test the IP-TFA levels in foods tested, however, this was not always reported in many other countries.
Despite these limitations, this study, the first of its kind in Lebanon, should provide the impetus for continuous comprehensive analysis of IP-TFA levels in foods in the regional and national kitchens and markets and the adaptation of the approaches for curbing the health hazards associated with IP-TFA consumption.

Conclusions
For the first time in Lebanon, a database on IP-TFA, mainly EA and LEA content in traditional dishes and market products is available and ready to be used by health care providers. There is more than enough convincing evidence that a high IP-TFAs, mainly EA and LEA intake is detrimental to cardiovascular health. Fortunately, this problem in Lebanon is fairly easy to solve via proper legislation. Despite the poorness of Lebanese dishes in IP-TFAs, however, the persistence of food products with high IP-TFAs levels in Lebanon means that subgroups of the Lebanese population, mainly vulnerable and foodinsecure people, are threatened by high levels of IP-TFAs due to frequent consumption of risky products. The inauguration and implementation of policies to curtail IP-TFAs in Lebanon may therefore be legitimized, and such efforts should underline added fats and packaged foods. The economic crises in Lebanon pushed the Lebanese people to select cheap oils, including butter and margarines instead of vegetable oils. Thus, it appears reasonable that the Lebanese government and ministries should strive to raise public awareness about the issue and lobby for implementing anti-IP-TFAs laws either on the level of national industries or, on the level of food products importation.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data underlying the results are available as part of the article and no additional source data are required.