The successful use of fish oil in food products: improving shelf-life and sensory properties

The application of fish oil in a wide range of different products such as bread, mayonnaise, salad dressing, ice cream spread, milk drink and infant formula has been attempted.  Attempts have also been made to increase n-3 fatty acid contents in animal products such as milk fat, pork meat and hen eggs by adding fish oil to the animal feed. Several n-3 LC PUFA-enriched food products have entered the marketplace. Some of these products have been withdrawn from the market again owing to poor market performance, but other products are still on the market. The reasons for the poor market performance for some of the products are not known, but it may be speculated that some of the reasons could be either that the products had

poorer sensory properties than traditional products;higher prices;poorer shelf-life; orthe fact that health claims on the beneficial effects of n-3 LC PUFA are not yet allowed in many countries.

No matter the reason, the fact is that the industry is still searching for better methods to improve the sensory properties and shelf-life of n-3 enriched foods.

The literature on the oxidative stability of real foods enriched with n-3 PUFA is scarce. Most studies have either been carried out in bulk fish oil or in simple fish oil-in-water emulsion systems. In the following, the major findings will be summarised.

Jafar et al. reported that mayonnaise containing 70 per cent fish oil prepared without antioxidants had a shelf-life of 1 day when stored at room temperature and when evaluated by sensory analysis (only the odour was evaluated). Addition of citric acid or sodium citrate and propyl gallate in the oil phase and EDTA and ascorbic acid in the aqueous phase increased the shelf-life to 49 days. The shelf-life could be further increased to 89 days by refrigeration and to 132 days by the addition of the glucose oxidase—catalase oxygen scavenging system.

Li Hsieh and Regenstein reported that after 14 weeks of storage a sensory panel was not able to distinguish between a traditional soy bean oil mayonnaise and a mayonnaise containing fish oil provided that EDTA (0.075 per cent) and TBHQ (0.02 per cent) were used, the storage temperature was low (2ëC) and that oxygen was excluded.

In Europe, TBHQ is not allowed and therefore this antioxidant cannot be used. Jacobsen et al.41, 56—59 evaluated the antioxidative effect of EDTA, propyl gallate, gallic acid, tocopherol, ascorbic acid or a mixture of ascorbic acid, lecithin and tocopherol (the so-called A/L/T system) by sensory profiling, measurements of lipid hydroperoxides and volatiles and in some cases also by electron spin response (ESR) determination of free radical formation. The volatiles were measured by dynamic headspace gas chromatography mass (GC-MS). Weak pro-oxidative effects of propyl gallate and gallic acid were observed. Tocopherol was inactive as an antioxidant and ascorbic acid and the A/L/T were strong pro-oxidants, probably because ascorbic acid promoted the release of iron from the egg yolk located at the oil—water interface. In contrast, EDTA (0.0075 per cent) was observed to be a strong antioxidant that totally inhibited oxidative flavour deterioration during storage at 20ëC. The strong antioxidative effect of EDTA was proposed to be due to the metal­chelating properties of this antioxidant. In our laboratory, we have subsequently shown that lower concentrations (i.e. 10 ppm) of EDTA is sufficient to retard lipid oxidation in mayonnaise (unpublished data).

Young developed low-calorie (40 per cent fat) spreads of commercially acceptable quality. The spreads had 20 per cent of the fat replaced by fish oil and the oxidative stability was optimised by adding EDTA (150 ppm). Moreover, the fish oil contained 300 ppm of Grindsted 117 (ascorbyl palmitate, propyl gallate and citric acid) and 1000 ppm Toco 50 (50 per cent tocopherol in vegetable oil). In another margarine study, tocopherol (0.2 per cent) in combination with ascorbyl palmitate (0.01 per cent) and propyl gallate (0.01 per cent) was found to be the most effective in retarding lipid oxidation as evaluated by peroxide, carbonyl and anisidine values. No sensory evaluation was performed. TBHQ also retarded lipid oxidation, but it was less efficient than the combination of the above three antioxidants and it gave rise to discoloration. The effect of EDTA was not evaluated in this study. Kolanowski et a1. suggested that spreads could only be enriched with up to 1.5 per cent fish oil (i.e. 0.5 per cent EPA and DHA) without deteriorating the sensory quality of the product. No antioxidants were added to this product. In another study, Kolanowski et a1 concluded that low-calorie spreadable fats (soft margarine and mix of butter and vegetable oil) could be enriched with up to 1 per cent EPA and DHA without significantly affecting the sensory quality. The margarine spread may be stored up to 6 weeks and the spread based on butter and vegetable up to 3 weeks without significant decrease of quality. These spreads contained 55 per cent fat and no antioxidants were added.

In a study by Kolanowski et a1., it was not possible to incorporate even low levels of EPA and DHA (0.15 per cent fish oil which equals 0.05 per cent EPA and DHA) into milk without significantly reducing the palatability of the milk. In contrast, the same authors found that enrichment of flavoured yoghurt with up to 0.3 per cent fish oil (i.e. 0.1 per cent EPA and DHA) resulted in a product with acceptable sensory characteristics. In our laboratory, we have shown that the oxidative deterioration of fish oil-enriched milk strongly depends on the quality of the fish oil. Cod liver oil with a relatively low peroxide value (PV) (1.5 meq/kg) was found to give rise to the development of fishy off-flavours when added to milk in a concentration of 1.5 per cent. In contrast, milk with a more unsaturated tuna oil (20.1 per cent in cod liver oil vs 30.2 per cent EPA + DHA in tuna oil) with a lower PV (0.1 meq/kg) oxidised much slower and off-flavours were not detected. On the other hand, it seems to be possible to enrich milk with n-3 LC PUFA and obtain a product of satisfactory sensory quality provided that the fish oil is of a high quality. Moreover, preliminary results in our laboratory also suggest that the process for homogenising the fish oil together with the milk must be optimised in order to reduce lipid oxidation in this product.

Kolonowski and co-workers found that addition of up to 0.3 per cent fish oil (0.15 per cent EPA and DHA) to orange juice resulted in an acceptable product, but after 10 days of storage a strong fishy off-flavour had developed. The effect of antioxidant addition was not evaluated in this study.

The sensory quality of formula concentrates consisting of an instant flavoured powder milk-based protein-carbohydrate formulae enriched with up to 6 per cent microencapsulated fish oil (0.3 per cent EPA and 0.6 per cent DHA) was acceptable after 6 months of storage.

In a study by SatueÂ-Gracia et a1. on traditional infant formula without n-3 LC PUFA, lactoferrin was able to reduce lipid oxidation. The antioxidative effect of lactoferrin was suggested to be due to its ability to chelate metal ions. These findings indicate that lactoferrin could be an important efficient antioxidant in infant formula enriched with n-3 LC PUFA.

Becker and Kyle evaluated the sensory stability of bread baked with either a regular pharmaceutical grade fish oil made from sand eel, a specialty tuna oil or an algal oil. The taste panel indicated that sensory off-flavours were less likely to be detected in the DHA bread made with algal oils compared with those made with fish oils. Based on these data they suggested that the algal source had a better stability than the fish oils. This proposition was later invalidated by Frankel et a1. They showed that the high oxidative stability of the commercial DHA-rich algal oil was lost when the triglycerides were purified to remove tocopherols and other antioxidants. Moreover, an oil-in-water emulsion with the same algal oil had a lower oxidative stability than corresponding fish oils.

n-3 LC PUFA for foods are commercially available both as neat oils (fish oil or algal oil) or as microencapsulated fat powders. The latter are essentially dried, homogenised emulsions of an oil or fat where proteins, modified starches or hydrocolloids are used as emulsifying materials. A non-emulsifying, water-soluble material such as sugar or hydrolysed starch is also used as filler. Different drying techniques, i.e. spray drying or freeze drying, may be used to produce the powders. The advantage of microencapsulation for fish oil is that the shelf-life of the fish oil may be extended by protecting the fish oil from contact with atmospheric oxygen. Keogh et a1. studied the effect of emulsifier type (three different casein types), free fat, surface fat and the air content of the spray-dried fish oil powder on the shelf-life as monitored by sensory evaluation. They concluded that fish oil powder with a low level of off-flavour can be produced with a shelf-life of 31 weeks at 4ëC using dairy ingredients alone as encapsulating ingredients. They also found that the shelf-life increased when the free fat and vacuole volumen of the powder decreased. They did not find any effect of the surface fat. In another study by Heinzelmann et aL, fish oil was microencapsulated using freeze-drying techniques. It was shown that the best shelf-life was obtained when the fish oil contained a combination of ascorbic acid, lecithin and tocopherol (i.e. the A/L/T system) and when the freezing rate was slow. No sensory evaluation was performed on these powders. Based on the authors personal communication with the industry producing microencapsulated n-3 LC PUFA, more research seems to be required to optimise the quality of the n-3 powders.

On the basis of the above summary of how lipid oxidation has been retarded in products enriched with n-3 LC PUFA the following strategies to avoid lipid oxidation are suggested:

Exclude oxygen from the system, for example by packaging under vacuum.Store the enriched products at chilled temperatures.Ensure that ingredients have a low content of hydroperoxides, transition metals and other pro-oxidants. It seems to be especially important that fish oil has a low PV. Therefore, marine oils should be stored at low temperatures (<0 ëC), in the dark, with reduced oxygen and the fish oils should be used as fast as possible after deodorisation as hydroperoxides will form even at temperatures below 0 ëC.The choice of emulsifier may significantly affect lipid oxidation rates. Therefore, when applying n-3 oils in a new food product it may be necessary to reformulate the conventional recipe to include other emulsifier types. In turn, the recipe may also need to be changed with respect to addition of thickening agents in order to obtain the same rheological properties as the traditional product.Use metal chelators such as EDTA, citric acid, proteins, polysaccharides and metallic chelating plant polyphenols to prevent lipid hydroperoxide decomposition.Addition of free radical chain breaking antioxidants may further reduce lipid oxidation. Select antioxidants that will be located where they are required, i.e. normally near the oil±water interface where the decomposition of lipid hydroperoxides takes place.Optimise the processing conditions. In some food systems the particle/droplet size will affect the oxidation rates, in other foods they may not. In addition, the emulsification process may disrupt natural membranes that may protect the fish oil from protein-bound metals. Emulsification processes should be optimised to minimise lipid oxidation.

Public awareness of the beneficial effects of the n-3 LC PUFA seems to have increased since the late 1990s. Likewise, the industrial interest in exploiting health effects of the n-3 oils has apparently also increased. Therefore, it is expected that more new food products enriched with n-3 PUFA will enter the marketplace in the coming years. A promising new area is baby food. As previously mentioned, the infant’s requirements for DHA has recently received substantial attention, and several formulas enriched with DHA are now on the market. It is therefore likely that efforts will also be made to develop other types of baby foods enriched with n-3 LC PUFA in the coming years.

Dairy products are the fastest growing product within the functional food area. So far, most dairy products in this category have been `functional’ due to the addition of probiotic bacteria, and consumers already perceive low-fat dairy products as being healthy. Therefore, milk drinks and yoghurts may be a good vehicle for n-3 LC PUFA enrichment and we may see a number of new products in this category in the future.

Ice cream producers are now also targeting healthy consumers. In the recent years, new fat reduced ice cream formulations have entered the market, and recently calcium-enriched ice creams have been marketed. Efforts are currently being made to develop ice cream enriched with n-3 LC PUFA.

This postr has mainly dealt with EPA and DHA from marine sources. However, with the increased focus on the beneficial effects of n-3 LC PUFA in general, it can be expected that products enriched with the 18:3 PUFA will also receive more attention from the industry. Traditionally, the industry has tried to protect food products from lipid oxidation by the addition of free radical chain-breaking antioxidants. However, this strategy does not seem to be very efficient in preventing lipid oxidation in emulsified food systems, especially when they are enriched with n-3 LC PUFA. With our increased understanding of the important role of trace metals, emulsifiers and processing conditions in the lipid oxidation processes, more efforts will be dedicated to use this knowledge to develop alternative strategies to retard lipid oxidation in real foods with n-3 oils.

The aquaculture industry is the main customer to fish oil, which is almost entirely produced from different types of industrial fish species such as sand eel, menhaden, etc. With the growing aquaculture industry, the demand for fish oil for fish feed production is increasing. This could lead to a situation where the demand for fish oil for human consumption and fish feed exceeds the production. However, sources of n-3 LC PUFA other than the traditional industrial fish species are available. Thus, efforts are being made to use the oil from waste products from fish species used for human consumption. In addition, n-3 LC PUFA can also be produced from algae and commercial products are already available.