trans fatty acid content in malaysian supermarket foods: a field-to-laboratory...

Trans fatty acid content in Malaysian supermarket foods: a field-to-laboratory approach inassessing food risk

Tilakavati Karupaiaha*, Hui Kuen Tana, Wei Wen Onga, Choon Heen Tana and Kalyana Sundramb

aFaculty of Health Sciences, National University of Malaysia, Kuala Lumpur 50300, Malaysia; bMalaysian Palm Oil Council, WismaSawit, 47301 Kelana Jaya, Malaysia

(Received 23 March 2014; accepted 25 May 2014)

The extent of industrial trans fatty acids (TFA) in the food supply is unknown in Malaysia, whilst TFA disclosure on foodlabels is not mandatory by Malaysian food standards. Supermarket foods such as dairy products, fats and oils, meatproducts, snack foods, soups, and confectionery are commonly cited to be major contributors of TFA in the diet. Aconsumer survey (n = 622) was used to develop a food listing of these ‘high risk’ foods. TFA content of high-risk foodswere analysed by gas chromatography. Food samples (n = 158) were analysed and their total TFA content were comparedwith Malaysian Food Standards. A wide variation in TFA content within food categories was indicated. Of the foodscontaining TFA, many food labels did not cite TFA content or the use of partially hydrogenated vegetable oils (PHVO) asan ingredient. Hypothesised estimates of TFA intake from these supermarket foods in a sample day’s menu providing2000 kcal projected a minimum intake of 0.5 g and a maximum intake of 5.2 g TFA. This study found there was novoluntary disclosure of TFA content on food labels or identifying PHVO as an ingredient. It appears that health educationtargeting consumers to minimise TFA consumption is required supported by mandatory PHVO disclosure on the food label.

Keywords: trans fatty acids; survey; gas chromatography; PHVO; supermarket foods; labelling; high-risk foods

Introduction

Trans fatty acids (TFA) became a perceived threat topublic health when a causal link between TFA and cardi-ovascular disease (CVD) became established (Micha &Mozaffarian 2009). Applying the rationale of Kjaergårdet al. (2013), TFA existence in the food supply is anexample of a constraint on a sustainable public healthinitiative (an industry solution to replace saturated fat) tosolving a health problem (elevated blood cholesterollevels) leading to a sustainability problem (achievingshelf-life) which caused an undesired outcome (TFA gen-eration) and unforeseen health problems (CVD). Forexample, the switch from the use of tallow to partiallyhydrogenated vegetables oils (PHVO) for frying purposesby the food industry in the United States was said toincrease TFA availability from 0.3 to 8.4 g day–1 perperson (Hunter & Applewhite 1991). Cited potentialharm from TFA are adverse effects on serum lipids, sys-temic inflammation, positive association with risk ofCVD, sudden death from cardiac causes, diabetes andeven overt aggressive behaviour (Micha & Mozaffarian2009; Golomb et al. 2012). These adverse effects areconsidered to be worse for consumer health than the threatof food contaminants or pesticide residues in food (Micha& Mozaffarian 2009).

Three major processes give rise to TFA in the foodchain: (1) naturally occurring such as cis-9, trans-11 and

trans-10, cis-12 conjugated linoleic acid (CLA) isomersthrough ruminant digestion and biohydrogenation; (2)industrial partial hydrogenation of oils and fats; and (3)trans-polyunsaturated fatty acids (PUFA) formed duringthe deodorisation or frying processes. The source ofhuman TFA consumption is primarily from industriallyproduced PHVOs available through fast food chains, res-taurants and school food services as well as from indust-rially processed foods sold through the retail sector (Micha& Mozaffarian 2009). Further, household sources of TFAinclude salad and cooking oils, shortenings, margarinesand spreads for home food preparation. The trans-mono-unsaturates namely trans-9-octadecenoic acid (t18:1n9) orelaidic acid are the most dominant form of TFA in theWestern consumer diet, as was shown from early surveil-lance studies in Europe and the United States (Aro,Amaral et al. 1998; Aro, Antoine et al. 1998; van Erp-Baart et al. 1998; Hulshof et al. 1999). Regulatoryapproaches for tackling the TFA issue in developed coun-tries have included a maximum allowable limit for TFAcontent in processed food, mandatory labelling on front-on-pack information, and even a TFA ban from restaurantmenus with some countries such as the Netherlands,Australia and New Zealand encouraging industry self-reg-ulation (L’Abbé et al. 2009; Angell et al. 2012). Stringentregulatory initiatives to reduce TFA in the food chain haveborne benefits. For example, the mean dietary intake of

*Corresponding author. Email: [email protected]

Food Additives & Contaminants: Part A, 2014Vol. 31, No. 8, 1375–1384, http://dx.doi.org/10.1080/19440049.2014.929183

© 2014 Taylor & Francis

industrially produced TFA in the United States hasdecreased from 4.6 g/person/day in 2003 to 1.3 g/person/day by 2009 for adults, as cited by the USFDA (Doellet al. 2012). Major contributors to artificial trans fat intakein the Western diet were considered to be ready-to-eatFrench fries, soups and snacks, bakery products, dairyand meat products, fast food, frozen pizzas, margarinesand spreads, ready-to-use frosting, and coffee creamers(Aro, Amaral et al. 1998; Aro, Antoine et al. 1998; vanErp-Baart et al. 1998; Hulshof et al. 1999).

In contrast to these developments about industrial TFAs(emergence, concern, detection and elimination) in Westerncountries, the availability of TFA content in foods has beenscarcely reported in Asian countries. A limited analysis of 26food items collected from Korean grocery stores in 2005 and2008 indicated an increased TFA content in biscuits, friedsnacks and cookies compared with crackers, processed cho-colate and ice cream (Adhikari et al. 2010). Based on theoccurrence of TFA in PHVO and blended fats in Pakistan,reported TFA in vanaspati samples were in the range 14.2–34.3%, 7.3–31.7% in shortenings, and 1.6–23.1% in hard-type margarines, whereas soft margarines contained less than4.1% TFA (Bhanger & Anwar 2004). Another study evalu-ating 12 biscuit brands in Pakistan noted the variation of TFAto be 9.3–34.9% (Kandhro et al. 2008). In India, wherevanaspati accounts for 60% of use in the bakery industryalone and is also widely used in cooking, commercially friedfoods, street foods and snacks, sweets and savouries, the 18:1trans content is reported to be 5%–38.1% (L’Abbé et al.2009). In contrast in Malaysia there is lack of informationon the availability of TFA in the food supply. Mandatorydeclaration of TFA is not required by the Malaysian foodauthorities. There is now keen interest from government andnon-governmental agencies to understand the status ofindustrially produced TFAs in Malaysia. Further, in contrast

to Western populations with a greater consumption of fastfoods, which are a major source of TFA, such habits are lessof a hazard in Malaysia. The likelihood for TFA content inAsian foods exists depending on cultural food use, but thereis a paucity of data in relation to the wider regiospecificityand scope of foods consumed given the ethnic and agricul-tural diversity of this region. Our approach, therefore, was toanalyse frequently purchased food products from likely TFArisk categories available in large supermarket chains.Therefore, the main objective of this study was to assessTFA content of selected ‘risk’ foods sold at supermarketsusing a ‘field to laboratory’ approach in our design of thestudy. The research questions focused on the following:

● Identifying ‘high risk’ food brands sold in super-markets likely to contain TFA.

● To determine the TFA content of food productsfrequently purchased by consumers.

● To determine the minimum and maximum TFAcontent in a standard 2000 kcal diet plan con-structed from analysed foods.

Materials and methods

The study flow from field-to-laboratory analyses is pre-sented in Figure 1.

Food listing

The identification of food products started with a literaturesearch for ‘high risk’ foods sold through the retail sector.A food listing was then developed with local food brandson supermarket shelves grouped under the following foodrisk categories for TFA:

Sampling and Laboratory Analysis

FATS & OILS

Margarine (n = 6)

Butter (n = 8)

Butter blends (n = 12)

Salad dressing (n = 6)

Vegetable shortening (n = 2)

Ghee (n = 6)

Vanaspati (n = 3)

Peanut Butter (n = 9)

DAIRY PRODUCTS

Milk powder/ Adult (n = 4)

Milk powder/ Children (n = 6)

Cheese (n = 5)

Ice cream (n = 6)

SNACKS

Potato chips (n = 13)

Ethnic breads (n = 7)

Frozen dough snacks (n = 13)

French fries (n = 4)

SOUPS

Canned (n = 3)

Concentrate (n = 3)

CONFECTIONARY (Chocolate Products)

Cooking chocolate (n = 3)

Wafers & Pralines (n = 21)

Candies (n = 3)

MEAT & MEAT

PRODUCTS

Sausage (n = 5)

Nugget (n = 7)

Beef patty/ burger (n = 2)

Food Listing (FL)

development for product

risk categorization

[Based on literature

review, identification of

products on supermarket

shelves and label reading]

Consumer

Survey

by convenience

sampling using FL

questionnaire at

4 major

supermarkets

Figure 1. Study flow diagram for the ‘field to laboratory’ approach.

1376 T. Karupaiah et al.

● Dairy products including ice cream, cheese and full-cream milk powder.

● Fats and oils including fat spreads, margarine, but-ter, salad dressings, vegetable shortenings, ghee andvanaspati (vegetable ghee).

● Meat and meat products including local beef patties,chicken sausages and nuggets.

● Snack foods such as ready-to-eat potato chips andfrozen partially cooked French fries and dough-based snacks.

● Soups including canned liquids and concentrates.● Confectionary such as chocolate products.

Inclusive criteria for selection of these products werebased on front-of-pack information indicating the signifi-cant amount of fat and/or type of fat such as vegetable oil,vegetable shortening, hydrogenated fat, partially hydroge-nated soya bean oil, and partially hydrogenated fat. Low-fat products such as low-fat yogurt, ice cream and soups orskim milk products were excluded from the food list.

Sampling

A consumer survey (n = 622; men/women = 224/398) usingthe food listing was carried out at four major supermarketoutlets in Klang Valley in Selangor, Malaysia. Eligibilitycriteria were adult householders shopping for their familiesand willing to fill in the forms. A total of 158 frequentlypurchased food brands on the food list were ranked by theseconsumers and these products were then sampled in dupli-cate for TFA analysis from these supermarkets.

Laboratory analyses

Sample preparation

Two units of each food brand were purchased from differentretail sources and samples for each food brand were pooledfor homogenisation. Given the diversity of the food groups,approaches specific to the food matrix were adopted asdescribed by the AOCS official methods (AOCS 2009).Fats and oils were directly extracted with solvents. Highmoisture-content samples such as canned soups and icecreams required freeze-drying prior to lipid extraction(Heto-Lyolab 3000, Allerød, Denmark). Chocolates andpotato chips were pulverised with liquid nitrogen in a mortarand pestle. Meat products were homogenised with fixedvolumes of distilled water before drying overnight in a con-vection oven at 105°C. Aliquots were taken in duplicatefrom pooled samples and analysed separately from lipidextraction, to methylation to chromatography.

Lipid extraction and FAME preparation

Total lipid extraction from dried food samples was carriedout by an automated Soxhlet (Soxtec Avanti 2055 ManualSystem, Foss North America, Eden Prairie, MN, USA)procedure using chloroform:methanol (2:1 v/v) with 3–5 gof food sample, as described in AOAC official methods960.39 and 933.05 (AOAC 2005). For products such asice cream, lipid extraction was carried out by the proce-dure of Roese-Gottlieb and for full-cream milk powderproducts by Monjonnier’s method, as described in AOACofficial methods 952.06 and 989.05 (AOAC 2005). Lipidswere dried under nitrogen until a constant weight wasachieved. Fat content was expressed as %w/w of the wetproduct. Dried lipids were methylated with toluene:2%H2SO4 in methanol (1:2 v/v) to yield fatty acid methylesters (FAME) (Hitchcock & Hammond 1980). All sol-vents and reagents used were of analytical grade.

Fatty acid composition (FAC)

Methyl esters of dried lipid samples were dissolved insmall volumes of hexane prior to GC analysis using aShimadzu GC-2010 (Shimadzu Corporation, Tokyo,Japan) equipped with a flame ionisation detector, aShimadzu AOC-20i auto-injector and an SP™-2560 capil-lary column (0.25 mm ID × 100 m, 20 μm film thickness)from Supelco Inc. (Belleforte, PA, USA). The carrier gaswas pure helium with linear velocity of 21.7 cm s–1 and ahead pressure of 45 psi with column temperatures startingat 150°C for 5 min followed by 4°C min–1 increments upto 240°C. Auto-injection of samples was done in a splitratio of 1:100.

Fatty acids in chromatographic spectra were identified incomparison with relative retention times of commercial stan-dards for FAME with 99% purity (Supelco™). To assist inconfirming the cis/trans identifications, the Supelco 37-com-ponent FAME Mix (47885-U) containing methyl esters ran-ging from C4 to C24, including key monounsaturated fattyacid (MUFA) and PFA, was run as an external standard alongwith linoleic acid methyl ester (C18:2) isomer mix (Cat. No.47791) and linolenic acid methyl ester (C18:3) isomer mix(Cat. No. 47792) standards under the same instrumentalconditions. Running these three standards provided for theidentification of trans-monoenes, dienes and trienes. Theelution of the cis/trans isomers of TFAwere in concordancewith the manufacturer’s guidelines, i.e. C18:1 n9t beforeC18:1 cis, the trans-isomers of C18:2 before and the transisomers of C18:3 after the elution of C18:2 cis-fatty acid.Retention times for the various fatty acids were calculatedrelative to the elution time for C18:1 cis. The chromatogrampeak areas for methyl esters of identified fatty acids werereported as %FAC.

Food Additives & Contaminants: Part A 1377

Data interpretation

Duplicate analyses were carried out and values arereported as their average. These values were corrected ona fresh weight basis. The sum of all trans, cis/trans,geometric and positional isomers, namely TFA, were con-sidered when determining total TFA content: 18:1 n9t;18:2 n6t; cis-9 t-12; t-9, cis-12; 18:3t1; 18:3t2; 18:3t4;and 18:3t5. The TFA content of sampled food brands inthis study are presented as total TFA, %FAC, g/servingsize on a fresh weight basis as referred to on the front-of-pack label, and TFA g kg−1 fresh weight basis. Finally, thetotal TFA content (TFA g kg−1) of sampled foods wascompared with levels indicated for a TFA nutrient labelclaim as per the Malaysian Food Standards which were (1)trans-free if TFA < 0.1 g kg−1; (2) low TFA if < 1.5 g kg−1

portion; and (3) high TFA if > 1.5 g kg−1 portion (Ministryof Health Malaysia 2003).

A simulation diet providing 2000 kcal was plannedbased on selections of analysed food brands containing theminimum and maximum TFA content calculated on thebasis of serving size as recommended by front-on-packinformation (g/serving size). For example, 60 g ice cream(a single scoop) for six brands could range between aminimum of 1 g to a maximum of 5 g TFA per serving.If two scoops of ice cream is reported as the usual con-sumption, then the minimum TFA content would be 2 gand the maximum 10 g TFA in the simulated diet plan.

Statistical methods

Quantitative data for the analysed TFA content of foodwere described as a mean for food categories with therange within the category provided. Data were alsodescribed categorically for comparisons with MalaysianFood Standards (Ministry of Health Malaysia 2003).

Results and discussion

Fatty acid analyses were completed for 158 food brands.Data for total fat content (g/100 g), total TFA (g kg−1),TFA as per cent total FAC (%TFA) and TFA content (g)per serving size of the food brands sampled in the variousfood categories are reported in Table 1. It was found thatmost front-of-pack labels for the food brands did not statethe fat source or TFA content. PHVO was stated on labelsfor only some brands that declared TFA content. For thosefood brands that did not state PHVO presence, the pre-sence of PHVO was concluded if a higher linoleic acidcontent (Skeaff 2009) was indicated in their FAC analyses(data not shown). For food brands that included the sourceof oils and fats on their front-of-pack information, thecommon oils stated were non-hydrogenated coconut,palm, canola and soybean oils and hydrogenated palmkernel oil (PKO).

Through FAC analyses, we found considerable TFAcontent variation amongst brands within food categories,reflecting the different vegetable oils used in the manufac-turing process. Most food manufacturers use single hydro-genated or non-hydrogenated fats and/or oils orcombinations of both hydrogenated and non-hydrogenatedfats and oils to achieve desirable rheological and texturalcharacteristics (Ratnayake et al. 2008; L’Abbé et al. 2009).For example, though chocolate products with a higherstearic acid content are a preferred characteristic, wefound that some cooking chocolate labels indicated alter-nate fat sources to be blends of PKO, illipe and/or sheabutter, and this was reflected in their varying content ofstearic or palmitic acid content in their FAC analyses (datanot shown). Additionally, total saturated fatty acid (SFA)and PUFA content in foods, as indicated by their FACcomposition data shown in Table 1, reflected the type offats and/oils used in their formulation. This is expected asPHVO-containing products will contain a lower SFA andhigher PUFA content compared with natural vegetable oilswith a higher solid content (Skeaff 2009).

Foods were categorised as trans-free (TFA < 0.1 g kg−1),low TFA (< 1.5 g kg−1) or high TFA (> 1.5 g kg−1) accordingto Malaysian Food Standards (Ministry of Health Malaysia2003). In the chocolate category, two brands of chocolatewafers were categorised as ‘high TFA’ (Figure 2A). Eitherpartially hydrogenated or hydrogenated shortening couldhave been added to raise the solid fat content, as indicatedby their higher TFA content (category range = 0–7.8 TFA g kg−1; Figure 2A), and concurrently lower SFAcontent (category range = 28.7–97.6% FAC; Table 1).

A higher content of TFA in dairy-based products hasbeen noted and attributed to the natural production of con-jugated linoleic acid in the bovine stomach (Bauman &Griinari 2003). However, most of the dairy-based brandsexamined were identified as ‘low TFA’, whilst the remain-der were ‘trans-free’ (Figure 2B). Of the ‘trans-free’ dairyproducts, six labels stated they contained either PKO orpalm oil, four labels did not identify the vegetable fat, twolabels did not state the source of fat and one brand labelstated it contained milk fat (Table 1). The usual vegetablefat replacer options in dairy-based products are PKO toincrease solid content and palm oil, whilst other vegetableoils such as soya bean, canola, safflower and sunflower oilsare also optional fat fillers. Mozzarella cheese, for instance,may be manufactured with a blend of palm oil and PKO(Berger & Idris 2005). In the 1990s, when PHVO fillerswere used in dairy products, a higher elaidic acid contentwas observed (Aro, Amaral, et al. 1998).

In the edible oils and fats category, margarine and fatspread formulations were classified as either ‘low trans’ or‘trans free’ (Figure 3A). The type of vegetable oils/fatsused would have been PKO-based formulations in locallymanufactured margarines blended with palm oil or soy-bean oil (Berger & Idris 2005). Front-of-label information


Table1.

TFA

contentin

variou

sfood

catego

ries.

Total

fatty

acid

compo

sitio

n(%

)a

Foo

dcatego

ries

Totalfat(g/100

g)a

TFA

(gkg

–1)a

TFA

gserving

size

bTFA

SFA

MUFA

PUFA

Labellin

gadherence

Con

fectiona

ryCoo

king

chocolate(n

=3)

33.1

(2.3–4

1.1)

0.5(0.04–

1.34

)0.3(0–0

.7)

1.27

(0.17–

–3.26)

80.7

(59.8–

92.6)

15.64(3.6–3

5.3)

2.0(0.5–4

.0)

Non

eCho

colate

wafer

(n=22

)27

.3(7.0–4

4.3)

0.73

(0–7

.83)

0.26

(0–2

.74)

2.72

(0–2

8.2)

62.3

(28.7–

97.6)

27.9

(0–5

9.4)

6.4(0.48–

17.6)

8Cho

colate

cand

y(n

=3)

20.1

(17.0–

23.0)

0.01

(0–0

.01)

00.03

(0–0

.06)

54.2

(40.0–

62.2)

35.1

(32.8–

39.1)

8.08

(2.9–1

7.6)

Non

eFats,oils,spreads,d

ressing

Ghee(n

=6)

99.8

(99.8–

99.8)

1.04

(0.20–

1.92

)0.16

(0.03–

0.29

)1.04

(0.20–

1.93

)61

.5(60.9–

62.6)

29.7

(29.1–

30.9)

3.3(1.2–6

.8)

Non

eSho

rtening(n

=2)

99.8

(99.5–

100)

0.20

(0.19–

0.21

)0.20

(0.19–

0.21

)0.20

(0.19–

0.21

)57

.0(52.2–

61.8)

33.6

(30.2–

37.0)

8.8(7.3–1

0.2)

Non

eButter(n

=8)

80.6

(79.0–

81.5)

0.99

(0.07–

3.23

)0.15

(0.01–

0.49

)1.32

(0.11–3.85

)57

.8(36.4–

66.9)

31.7

(25.5–

44.8)

5.9(1.9–1

6.1)

2Margarine

(n=6)

77.0

(70.0–

82.0)

0.27

(0.04–

0.38

)0.04

(0.01–

0.05

)0.36

(0.05–

0.54

)46

.5(20.9–

61.0)

36.3

(29.1–

36.3)

16.8

(8.3–2

8.8)

Onlyon

eSalad

dressing

(n=6)

45.0

(21.7–

72.2)

0.09

(0.02–

0.20

)0.03

(0.01–

0.06

)0.18

(0.05–

0.27

)14

.5(13.6–

15.8)

22.7

(21.4–

24.0)

61.0

(58.6–

62.5)

Non

eVanaspati(n

=3)

99.8

(99.8–

99.8)

0.43

(0.17–

0.88

)0.06

(0.03–

0.13

)0.43

(0.18–

0.89

)50

.6(46.1–

54.0)

37.9

(35.7–

40.2)

10.7

(9.7–1

2.5)

Non

eFat

spread

(n=12

)73

.4(65.0–

82.0)

0.16

(0.01–

0.63

)0.02

(0–0

.09)

0.22

(0.01–

0.89

)36

.3(17.2–

61.5)

39.4

(30.3–

57.2)

23.2

(7.2–4

7.3)

Onlyon

ePeanu

tbu

tter(n

=9)

41.97(30.0–

53.7)

0.22

(0.02–

0.67

)0.07

(0.01–

0.21

)0.52

(0.04–

1.42

)20

.33(12.6–

41.0)48

.56(41.9–

52.0)26

.85(13.9–

37.0)Only6

Dairy-based

prod

ucts

Icecream

(n=6)

11.0

(8.4–1

8.14

)0.21

(0–0

.61)

0.13

(0–0

.37)

2.09

(0.03–

5.47

)68

.3(49.9–

83.0)

23.4

(10.9–

38.4)

4.8(0.27–

11.0)

Onlyon

eCheese(n

=5)

21.5

(13.7–

24.7)

0.18

(0.05–

0.63

)0.04

(0.01–

0.16

)0.78

(0.33–

2.57

)59

.8(50.7–

65.4)

31.6

(27.1–

38.8)

4.6(1.7–9

.4)

2Children’smilk

>3years

(n=4)

17.8

(13.0–

27.2)

0.18

(0.08–

0.39

)0.07

(0.03–

0.13

)0.93

(0.60–

1.42

)44

.7(18.0–

58.4)

36.8

(27.4–

44.6)

16.4

(1.9–3

7.6)

Non

e

Children’smilk

<1year

(n=2)

27.4

(26.7–

28.1)

0.04

(0.01–

0.06

)0.01

(0–0

.02)

0.14

(0.04–

0.24

)39

.6(37.0–

42.2)

40.7

(36.0–

45.4)

18.3

(16.1–

20.5)

Non

e

Adu

ltmilk

powder(n

=4)

25.6

(19.9–

28.7)

0.44

(0.06–

0.82

)0.14

(0.02–

0.26

)1.65

(0.28–

3.17

)58

.9(51.2–

66.4)

30.8

(24.5–

35.6)

5.2(1.3–1

0.0)

Onlyon

eSo

ups

Sou

p,concentrates

(n=3)

16.98(11.0–

22.3)

0.37

(0.02–

0.59

)0.07

(0–0

.11)

1.94

(0.17–

3.37

)52

.0(49.3–

55.1)

36.28(35.7–

37.1)

9.0(6.0–11.5)

Onlyon

eSou

ps,cann

ed(n

=3)

45.8

(40.1–

49.0)

0.00

4(0.002

–0.01)

0.01

(0–0

.01)

0.09

(0.03–

0.18

)10

.7(10.3–

11.3)

54.6

(54.3–

55.5)

32.3

(30.9–

33.1)

none

Snacks

Potatochips(n

=13

)32

.7(25.4–

36.1)

0.08

(0.01–

0.14

)0.02

(0–0

.03)

0.24

(0.05–

0.50

)38

.3(6.8–4

5.5)

45.3

(37.4–

75.7)

15.1

(11.5–

38.0)

Non

eFrenchfries(n

=4)

2.55

(1.99–

3.08

)0.01

(0–0

.01)

0.01

(0–0

.01)

0.26

(0.09–

0.51

)51

.3(47.5–

54.5)

36.9

(34.4–

38.5)

10.8

(9.3–1

3.1)

2Ethnicbreads

(n=7)

9.1(6.15–

14.7)

0.06

(0–0

.15)

0.04

(0–0

.12)

0.64

(0.02–

1.45

)52

.1(42.0–

60.1)

34.3

(30.1–

43.7)

12.2

(9.1–1

6.2)

Non

eFrozendo

ugh(n

=13

)5.5(2.36–

12.1)

0.01

(0–0

.06)

0(0–0

.02)

0.28

(0.06–

1.0)

48.9

(42.2–

58.0)

37.8

(28.5–

43.0)

12.2

(10.5–

13.9)

Non

eMeatprod

ucts

Nug

gets(n

=7)

15.04(12.5–

16.8)

0.02

(0.01–

0.03

)0.02

(0.01–

0.03

)0.18

(0.07–

0.36

)43

.5(40.4–

46.5)

42.1

(40.6–

45.0)

13.1

(9.4–1

4.0)

Non

eBurgerpatties

(n=2)

13.0

(13.0–

13.0)

0.01

(0.01–

0.01

)0.01

(0.01–

0.01

)0.08

(0.06–

0.09

)40

.9(39.2–

42.5)

43.4

(42.9–

43.9)

12.3

(8.0–1

6.7)

Non

eSausages(n

=5)

13.75(10.0–

19.0)

0.01

(0–0

.02)

0.01

(0–0

.02)

0.15

(0.04–

0.36

)31

.1(27.4–

39.6)

45.9

(39.1–

50.4)

21.3

(16.3–

28.0)

Non

e

Notes:TFA

,tran

sfatty

acid;MUFA

,monounsaturated

fatty

acid.

a Valuesaremean(m

inim

umto

maxim

um)values

forbrands

with

incategories.

bg/servingsize

onafreshweightbasisas

referred

toon

thefront-of-packinform

ation.

Labellin

gadherenceisdeterm

ined

asadeclarationof

TFA

contentas

pertherisk

category

definedby

theFoo

dAct

and/or

declarationof

partially

hydrog

enated

faton

theingredient

label.


on imported margarine products indicated product ingre-dients to be vegetable oil combinations with canola, soy-bean, palm and/or sunflower oils or PKO. Since theassociation of the unfavourable effects of TFA with cardi-ovascular risk has been made, many European manufac-turers have reformulated margarine with alternativeprocesses such as inter-esterification or blending of softwith saturated oils to be trans-free by using palm oil orfully hydrogenated vegetable oils (Pedersen & Kirkhus

2008; L’Abbé et al. 2009; Skeaff 2009). In contrast vanas-pati, a common substitute for ghee in Indian recipes, isformulated with PHVO (Bhanger & Anwar 2004; Berger& Idris 2005). However, all vanaspati brands analysed inthis study were categorised as ‘low TFA’ (category range= 0.17–0.88 kg–1) whereas some butter and ghee brandshad detectable amounts of TFA (Figure 3A). Heat isomer-isation of fatty acids may be a possible source for TFA inthese products (Johnson & Saikia 2009).

–0.5

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

6.5

7

7.5

8

8.5

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

TF

A g

/ 1

00

g p

ort

ion

Chocolate products

Wafers & Pralines 1 Wafers & Pralines 2 Wafers & Pralines 3 Wafers & Pralines 4 Wafers & Pralines 5 Wafers & Pralines 6



Wafers & Pralines 19 Wafers & Pralines 20 Wafers & Pralines 21 Wafers & Pralines 22 Candy 23 Candy 24

Candy 25 Cooking Chocolate 26 Cooking Chocolate 27 Cooking Chocolate 28

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

TF

A g

/ 1

00

g p

ort

ion

Dairy products

Cheese 1Cheese 2 Cheese 3 Cheese 4Cheese 5Ice cream 6 Ice cream 7 Ice cream 8Ice cream 9Ice cream 10 Ice cream 11 Infant follow–up milk 12Infant follow–up milk 13Children's milk powder 14 Children's milk powder 15 Children's milk powder 16Children's milk powder 17Adult milk powder 18 Adult milk powder 19 Adult milk powder 20Adult milk powder 21

A

B

Figure 2. Chocolate (A) and dairy product (B) distribution as per risk labelling category. For chocolate products (A) two brandscategorised as ‘high trans’ [▲] and 12 brands categorised as ‘low trans’ [▪] did not state hydrogenated fat or TFA on their labels. Onechocolate brand (#18, □) was declared hydrogenated fat on the label. For dairy-based products (B) one of five cheese brands, two of sixice cream brands, two of four children’s milk powder, and two of four adult milk brands were classified as ‘low TFA’ (▪). None of thesebrands stated hydrogenated fat or TFA on their labels.


Most peanut butter brands fell within a range of 0.02–0.67 g TFA kg−1 and identified as ‘low TFA’ (Figure 3B).In comparison, a mean value of 0.4 g TFA kg−1 wasreported in a combined analysis of top brands for peanutbutter in the United States (Mozaffarian et al. 2006),whilst a range of 1.6–6.7 g TFA kg−1 was reported inCanadian brands (Elias & Innis 2002). In most peanutbutter formulations, PHVO or even unhydrogenatedpalm oil are added to peanut butters at levels of 1−2% toprevent oil separation resulting in non-detectable levels ofTFA (Hinds et al. 1994; Sanders 2001). Locally producedpeanut butter brands in Malaysia contain palm oil in theirformulation, thereby avoiding or reducing the final transcontent (Berger & Idris 2005).

In the soup category, soup concentrates were mostly‘low TFA’, whilst canned soup brands were ‘trans-free’(Figure 4A). Essentially all products in the meat categorywere also ‘trans free’ (Figure 4B). Food brands in thesnack food category were classified as ‘low TFA’ or‘trans-free’ (Figure 4C). Potato chip brands were mostlyformulated with palm oil, as stated on their labels, whilstPHVO was stated on the labels of frozen French fries. Thepotato chip food category is an example of how Malaysianmanufacturers have striven to reduce TFA content (Berger& Idris 2005).

Overall, of the 158 food brands analysed, only fivefoods were categorised as ‘high TFA’ with a content> 1.5 g kg−1, whilst a good proportion of the foods were

–0.5

0

0.5

1

1.5

2

2.5

3

3.5

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

TF

A g

/ 100g

po

rtio

ns

Oils and fats products

Ghee 1 Ghee 2 Ghee 3 Ghee 4 Ghee 5 Ghee 6 Vanaspati 7 Vanaspati 8

Vanaspati 9 Shortening 10 Shortening 11 Butter 12 Butter 13 Butter 14 Butter 15 Butter 16

Butter 17 Butter 18 Butter 19 Margarine 20 Margarine 21 Margarine 22 Margarine 23 Margarine 24

Margarine 25 Fatspread 26 Fatspread 27 Fatspread 28 Fatspread 29 Fatspread 30 Fatspread 31 Fatspread 32

Fatspread 33 Fatspread 34 Fatspread 35 Fatspread 36 Fatspread 37

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

TF

A g

/ 100g

po

rtio

n

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Fat–based products

Salad Dressing 1 Salad Dressing 2 Salad Dressing 3 Salad Dressing 4 Salad Dressing 5

Salad Dressing 6 Peanut Butter 7 Peanut Butter 8 Peanut Butter 9 Peanut Butter 10

Peanut Butter 11 Peanut Butter 12 Peanut Butter 13 Peanut Butter 14 Peanut Butter 15

A

B

Figure 3. Fat product distribution as per risk labelling category. For fats and oils category (A) two brands (#25 and 36, □) stated thepresence of hydrogenated fat and one brand the presence of TFA (#18, 4) on the label. The fat source was not stated on 16 labels and thevegetable fat was not clearly identified on five labels. For fat-based products (B) the use of hydrogenated fat sourced from rapeseed,cottonseed and/or soybean oils was declared on six of nine peanut butter brands (□, �) whilst two salad dressings and two peanut butterbrands classified as ‘low TFA’ (▪) did not state hydrogenated fat or TFA on their labels. ¤, Trans-free = < 0.1 g TFA kg−1 portion; ▪, lowTFA = < 1.5 g TFA kg−1 portion; and ▲, high TFA = > 1.5 g TFA kg−1 portion.


categorised as ‘low trans’ giving an impression that indus-trial TFA risk would be low in Malaysia compared withreports from India, Pakistan and Korea (Bhanger & Anwar2004; Kandhro et al. 2008; L’Abbé et al. 2009; Adhikariet al. 2010). One explanation would also be the trickle-down effect of industrial TFA regulation in Western coun-tries (Ratnayake et al. 2008; L’Abbé et al. 2009; Skeaff2009). Replacement fats and oils to minimise industrialTFA content in formulations include modification of thehydrogenation process, inter-esterification of different fatsto produce desirable FAC profiles, fractionation of SFA-rich vegetable oils such as palm and coconut oils, oilblends such as palm oil-palm stearin-palm olein, as wellas production from seeds with novel FAC through plant-breeding techniques or genetic modification (Skeaff 2009).In order to explore the influence of variability in TFA

content across similar foods based on hypothecated esti-mates of TFA intake, a sample day’s menu providing2000 kcal was calculated (Table 2). Projections of a mini-mum intake of 0.5 g TFA and a maximum intake of 5.2 gTFA were associated with a 2000 kcal day–1 diet based onselections from the food brands sampled in this study. Themaximum risk was close to the highest tertile of TFAintake reported in the Nurse’s Health Study or even the5.8 g TFA intake per day previously reported in theaverage American diet before TFA regulation was insti-tuted in the United States (Lichtenstein et al. 2006). Thedaily intake of 5 g of TFA that we calculated from themaximum risk of the supermarket foods is associated witha 25% increased risk of ischaemic heart disease as shownin the Zutphen Study (Oomen et al. 2001). In comparisonwith other fats and nutrients, TFA is also considered to

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0 1 2 3 4 5 6

TF

A g

/ 100g

po

rtio

n

Soup products

Soup concentrate 1 Soup concentrate 2Soup concentrate 3 Canned soup 4Canned soup 5 Canned soup 6

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

TF

Ag

/ 100g

po

rtio

n

Meat products

#1–7 = chicken nuggets; #8–9 = beef patty;#10–14 = sausages/frankfurters;

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

TF

A g

/100g

po

rtio

n

Snack-food products

Ethnic bread 1 Ethnic bread 2 Ethnic bread 3 Ethnic bread 4 Ethnic bread 5 Ethnic bread 6Ethnic bread 7 Frozen Dough 8 Frozen Dough 9 Frozen Dough 10 Frozen Dough 11 Frozen Dough 12Frozen Dough 13 Frozen Dough 14 Frozen Dough 15 Frozen Dough 16 Frozen Dough 17 Frozen Dough 18Frozen Dough 19 Frozen Dough 20 Potato chips 21 Potato chips 22 Potato chips 23 Potato chips 24Potato chips 25 Potato chips 26 Potato chips 27 Potato chips 28 Potato chips 29 Potato chips 30Potato chips 31 Potato chips 32 Potato chips 33 Frenchfries 34 Frenchfries 35 Frenchfries 36

A B

C

Figure 4. Soups, meat and snack product distribution as per risk labelling category. For soups (A) two of three soup concentrates werecategorised as ‘low TFA’ (▪) whereas canned soups were ‘trans-free’ (▲). Neither hydrogenated fats nor TFA content were stated onproduct labels categorised as ‘low TFA’ (▪). All meat products (B) tested free of TFA (¤). For snack-food products (C) hydrogenated fatcontent was stated on three food labels (□, �). TFA content was not stated for six snacks categorised as ‘low TFA’ (▪). ¤, Trans-free = < 0.1 g TFA kg−1 portion; ▪, low TFA = < 1.5 g TFA kg−1 portion; and ▲, high TFA = > 1.5 g TFA kg−1 portion.


have a unique cardiometabolic effect by modulating insu-lin resistance and metabolic syndrome pathways (Micha &Mozaffarian 2009).

Given the wide number of food products available forconsumption, our study was limited to the restricted num-ber of retail foods surveyed and analysed. Nevertheless,we focused on a ‘field to laboratory’ approach to narrowthe gaps which were to understand consumer food choicebehaviour and selecting food brands for analysis based onthis perspective. A majority of these food products did notstate the TFA content as this is not mandatory in Malaysia.Therefore, the strength of this study is that we provided anindependent analysis of TFA content and developed adatabase from which mock planned diets were simulatedwith the lowest TFA and highest TFA content from food-choice alternatives. Of course, the projected scenario ofTFA intake does not replicate the actual reported TFAintake and this is the second limitation of this study.Despite these limitations, this study serves to highlightthe issue of food safety based on the risk exposure toTFA in foods. A future research direction would be toassess actual intake patterns of TFA by the public in orderto determine the true risk in terms of minimum and max-imum exposure.

Conclusions

Processed foods sold in the supermarket were found to bea potential source of industrial TFAs. It appears that healtheducation targeting consumers to minimise TFA consump-tion is required, supported by mandatory PHVO disclosureon the food label.

FundingThis study was funded under grant number NN-025-2009/UKM1.21/244 provided by the Food Claims and Safety Division ofthe Ministry of Health, Malaysia. The authors disclose no con-flict of interest.

ReferencesAdhikari P, Yu F, Lee J-H, Park HK, Kim JW, Lee EJ, Lee K-T.

2010. Comparative study of trans fatty acid content in 2005and 2008 processed foods from Korean market. Food Sci &Biotech. 19:335–341.

American Oil Chemists’ Society (AOCS). 2009. Official MethodCe56-89. In: Firestone D, editor. Official methods andrecommended practices of the American Oil Chemists’Society. 6th ed. Champaign (IL): AOCS Press.

Angell SY, Cobb LK, Curtis CJ, Konty KJ, Silver LD. 2012.Change in trans fatty acid content of fast-food purchasesassociated with New York City’s restaurant regulation: apre-post study. Ann Intern Med. 157:81–86.

Aro A, Amaral E, Kesteloot H, Rimestad A, Thamm M, vanPoppel G. 1998. Trans fatty acids in french fries, soups andsnacks from 14 European countries: the TRANSFAIR study.J Food Comp Anal. 11:170–177.

Aro A, Antoine JM, Pizzoferrato L, Reykdal O, van Poppel G.1998. Trans fatty acids in dairy and meat products from 14European countries: the TRANSFAIR study. J Food CompAnal. 11:150–160.

[AOAC] Association of Official Analytical Chemists. 2005.Official methods of analysis. 18th ed. Gaithersburg (MD):AOAC International.

Bauman DE, Griinari JM. 2003. Nutritional regulation of milkfat synthesis. Ann Rev Nutr. 23:203–227.

Berger KG, Idris NA. 2005. Formulation of zero-trans acidshortenings and margarines and other food fats with productsof the oil palm. J Am Oil Chem Soc. 82:775–782.

Table 2. Case scenario for minimum and maximum risk for TFA intake in a 2000 kcal diet.

Meals Food choices per portion Minimum TFA intake (g) Maximum TFA intake (g)

Breakfast Bread with butter (15 g) 0.01 0.49Milk (full cream milk powder) (32 g) 0.05 0.26

Mid-morning Roti canaia/paratha (80 g) 0.01 0.12Ghee (20 g) 0.04 0.38

Lunch Nasi minyakb + margarine (14 g) 0.005 0.05Soup concentrate (one sachet) 0.003 0.11Salad with dressing (30 g) 0.01 0.06Ice cream (120 g) 0 0.61

Snacks Potato chips (160 g) 0.02 0.22Peanut butter sandwich (32 g) 0.006 0.21Chocolate cake (fat spread + cooking chocolate) 0.001 0.09

0.06 0.65Dinner Canned soup (120 g) 0.002 0.01

Milk (full cream milk powder) (32 g) 0.05 0.26Chocolate wafer (20 g) 0 1.57Total daily intake 0.267 5.21

Notes: The 2000 kcal diet plan is composed from a selection of food brands commonly purchased by surveyed consumers from supermarket shelves.Calculations are based on the lowest and highest TFA content (g) per serving size provided in Table 1. TFA values shown are calculated for usual portionsize consumption. For example, ice cream values represent double the serving size reported in Table 1.aFlattened bread.bRice seasoned with vegetable fat.


Bhanger MI, Anwar F. 2004. Fatty acid (FA) composition andcontents of trans unsaturated FA in hydrogenated vegetableoils and blended fats from Pakistan. J Am Oil Chem Soc.81:129–134.

Doell D, Folmer D, Lee H, Honigfort M, Carberry S. 2012.Updated estimate of trans fat intake by the US population.Food Addit Contam: Part A. 29:861–874.

Elias SL, Innis SM. 2002. Bakery foods are the major dietarysource of trans-fatty acids among pregnant women with dietsproviding 30 percent energy from fat. J Am Diet Assoc.102:46–51.

Golomb BA, Evans MA, White HL, Dimsdale JE. 2012. Transfat consumption and aggression. PLoS ONE. 7:e32175.

Hinds MJ, Chinnan MS, Beuchat LR. 1994. Unhydrogenatedpalm oil as a stabilizer for peanut butter. J Food Sci.59:816–820.

Hitchcock C, Hammond EW. 1980. The determination of lipidsin foods. In: King RD, editor. Developments in food analysistechniques. Volume 2. London: Applied Science Pub. Ltd.

Hulshof K, van Erp-Baart MA, Anttolainen M, Becker W, ChurchSM, Couet C, Hermann-Kunz E, Kesteloot H, Leth T, MartinsI, et al. 1999. Intake of fatty acids in Western Europe withemphasis on trans fatty acids: the TRANSFAIR study. Eur JClin Nutr. 53:143–157.

Hunter JE, Applewhite TH. 1991. Reassessment of trans fattyacid availability in the US diet. Am J Clin Nutr. 54:363–369.

Johnson S, Saikia N. 2009. Fatty acid profiles of edible oils andfats in India. New Delhi: Centre for Science & Environment,Pollution Monitoring Laboratory; 48p.

Kandhro A, Sherazi STH, Mahesar SA, Bhanger MI, Talpur MY,Arain S. 2008.Monitoring of fat content, free fatty acid and fattyacid profile including trans fat in Pakistani biscuits. J Am OilChem Soc. 85:1057-1061.

Kjaergård B, Land B, Pedersen KB. 2013. Health and sustain-ability. Health Promot Internat.

L’Abbé MR, Stender S, Skeaff M, Ghafoorunissa TM. 2009.Approaches to removing trans fats from the food supply in

industrialiazed and developing countries. Eur J Clin Nutr.63:550–567.

Lichtenstein AH, Appel LJ, Brands M, Carnethon M, Daniels S,Franch HA, Franklin B, Kris-Etherton P, Harris WS, HowardB, et al. 2006. Diet and lifestyle recommendations revision2006: a scientific statement from the American HeartAssociation Nutrition Committee. Circulation. 114:82–96.

Micha R, Mozaffarian D. 2009. Trans fatty acids: effects onmetabolic syndrome, heart disease and diabetes. Nat RevEndocrinol. 5:335–344.

Ministry of Health Malaysia. 2003. Food Act 1983 and FoodRegulations 1985. In Amendment 2003, Government gaz-ette: Putra Jaya.

Mozaffarian D, Katan MB, Ascherio A, Stampfer MJ, WilletWC. 2006. Trans fatty acids and cardiovascular disease.New Eng J Med. 354:1601-1613.

Oomen CM, Ocké MC, Feskens EJ, van Erp-Baart MA, Kok FJ,Kromhout D. 2001. Association between trans fatty acidintake and 10-year risk of coronary heart disease in theZutphen Elderly Study: a prospective population-basedstudy. Lancet. 357:746–751.

Pedersen JI, Kirkhus B. 2008. Fatty acid composition of posttrans margarines and their health implications. Lipid Tech.20:132–135.

Ratnayake WMN, L’Abbe MR, Mozaffarian D. 2008.Nationwide product reformulations to reduce trans fattyacids in Canada: when trans fat goes out, what goes in?Eur J Clin Nutr. 39:1–4.

Sanders TH. 2001. Non-detectable levels of trans-fatty acids inpeanut butter. J Agric Food Chem. 49:2349-2351.

Skeaff CM. 2009. Feasibility of recommending certain repla-cement or alternative fats. Eur J Clin Nutr. 63:SS34–SS49.

van Erp-Baart M-A, Couet C, Cuadrado C, Kafatos A, Stanley J,van Poppel G. 1998. Trans fatty acids in bakery productsfrom 14 European countries: the TRANSFAIR study. J FoodComp Anal. 11:161–169.


Copyright of Food Additives & Contaminants. Part A: Chemistry, Analysis, Control,Exposure & Risk Assessment is the property of Taylor & Francis Ltd and its content may notbe copied or emailed to multiple sites or posted to a listserv without the copyright holder'sexpress written permission. However, users may print, download, or email articles forindividual use.

trans fatty acid content in malaysian supermarket foods: a field-to-laboratory...

Documents