LIPID QUALITY AND HEALTH BENEFIT OF COMMONLY CONSUMED FRESH WATER AND SEA WATER FISH SPECIES IN SERBIAN POPULATION LIPIDNI PROFIL I ZDRAVSTVENI ZNAĈAJ NAJĈEŠĆE KONZUMIRANIH REĈNIH I MORSKIH RIBA U SRPSKOJ POPULACIJI

Background/Aim. Dietary intake of n-3 long-chain polyunsaturated fatty acids (LC-PUFA) is important in prevention and treatment of different diseases. In general population the average intake of n-3 LC-PUFA is often significantly lower than recommended levels. Fish lipids are rich sources of these fatty acids, of which the most important are eicosapentaenoic (20:5 n-3, EPA) and docosahexaenoic (22:6 n-3, DHA) fatty acids. This study was designed to determine and compare fat, fatty acids and lipid quality indexes in 10 commercial fish species available on the Serbian market, as well as relation between their price and nutritional value. Methods. Freshwater fish originated from the Danube river in the Belgrade region, while seawater fish were mostly from the Adriatic Sea. A gas chromatography method was used to define fatty acids in 40 fish samples after lipid extraction. Cost-minimization analysis was conducted to assess the economic utility. Results. Seawater fish had a significantly higher value of flash lipid quality compared to the freshwater fish, (p<0.05). Hypercholesterolaemic fatty acids (OFA) for the freshwater group was 18.70 (17.40-21.30) while the seawater group had a similar range of values 18.90 (17.55-22.75). Hypocholesterolaemic fatty acids (DFA) also showed similar ranges for both groups 68.80 (66-70.20) for freshwater and 68.40 (64.85-73.05) for seawater group. The ratio of DHA/EPA ranged from 1.8 for sardine samples and up to 10 for tuna samples, indicating that the amount of DHA in natural samples exceeds the amount of EPA in many cases. The values of atherogenic (AI) and thrombogenic index (TI) were lower than 1 for all analysed samples. Conclusion. Sardine and mackerel had the highest content of n-3 LC-PUFA and presented the least expensive sources of EPA and DHA. The low values of AI and TI obtained from studied fish indicate its benefits from a health point of view.


Introduction
Fish and seafood are the only significant sources of n-3 long-chain polyunsaturated fatty acids (LC-PUFA) in the modern human diet. Alfa-linoleic acid (ALA) is also found in foods of plant origin, but the process of converting this fatty acid into eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) in humans is not effective enough. Due to the importance of fish as a source of n-3 LC-PUFA in the past decades, numerous investigations have been carried out on the health impact of their consumption. The Danish scientists established a link between the high intake of sea fish and the low incidence of cardiovascular diseases in the Greenland Inuit 1 . Further research has shown that the consumption of fish in everyday diet has numerous beneficial effects. A study including subjects from 36 countries, found a significant correlation between fish consumption and a reduced risk of general mortality, as well as the mortality caused by ischemic heart disease and cerebrovascular disease 2 . The Hordaland Health Study among middle-aged men and women confirmed an inverse relation between total fish intake and parameters of metabolic syndrome 3 . Also, very recent review revealed results from interventional and prospective cohort studies, with mostly beneficial effects of fish consumption on cardiovascular disease and metabolic syndrome 4 . In addition to the protective effect on the cardiovascular system, it is considered that the intake of fish can positively influence the development and proper functioning of the central nervous system, especially in children. Fish intake and n-3 LC-PUFA are associated with favourable effects in neurodegenerative and retinodegenerative diseases 5,6 . In 2012, European Food Safety Authority (EFSA) established dietary recommendations for the intake of 250-500 mg/day of EPA+DHA based on cardiovascular risk for European adults (EFSA, 2012) 7 . Fish consumption of at least 2 times a week (recommended fatty fish -salmon, herring and mackerel) and the use of vegetable oils with ALA for preparing food (soybean oil) has been recommended by the American Heart Association (AHA). For the purpose of secondary preventions, AHA recommends 1 g EPA+DHA/day 8 . According to other governing bodies, at least 2 servings of fish per week is set by Australian Dietary Guidelines 9 , while 280-525 g of fish is the appropriate weekly intake recommended by Dietary Guidelines for Chinese Residents 10 .
Intake of n-3 LC-PUFA in a population can vary significantly according to dietary and supplementation habits, age group, and gender, but in general, the average intake is often significantly lower than recommended levels [11][12][13] . For the Serbian population, there are 6 still no published data on EPA and DHA intake, but it is deemed that dietary fish intake is low and infrequent, and our previous study showed that n-3 LC-PUFA status in middleaged healthy individuals was inadequate, confirming low fish consumption 14,15 .
Additionally, patients with cardiometabolic risk factors in Serbia have abnormal n-3 LC-PUFA profile compared to healthy subjects 16 .
When increased fish intake is recommended, attention should be paid to the fact that both the total fat and PUFA content vary considerably among species. The fat content and the fatty acid composition of fish can vary significantly due to species, variations in their diet, life cycle, temperature, location, gender and environmental conditions 17,18 .
Moreover, fish can be an important source of aquatic contaminants such as heavy metals and polychlorinated organic compounds with high persistence, bio-accumulative properties and potential harmful human health effects 19 .
Thus, the aim of this study was to determine and compare the fat, fatty acids and lipid quality indexes in common freshwater and seawater fish species available on the Serbian market and to evaluate the most economical fish species as sources of dietary longchain n-3 fatty acids. In addition, the potential health risks due to the toxicity of contaminants present in fish were also taken into account.

Sample preparation
Five seawater fish species commonly consumed in the Serbian diet were chosen and purchased from the local fish market (salmon, tuna, mackerel, sardines, and hake). The origin of salmon was the North Sea (Norway), while other four species originated from the Adriatic Sea around Montenegro and Croatia. On the other hand, freshwater fish species were caught on the Sava and Danube rivers in the vicinity of Belgrade. These are carp, catfish, perch and bream. Trout was the only freshwater cultivated species analysed in this study. The season chosen for investigation was winter. Skin, head, gills, fins, and bones 7 from each fish species were removed; meat was cut into pieces and homogenized. Three to five specimens from each species were used for the analyses.

Total fat content
The total lipids content from each homogenized sample was determined by the gravimetric method after Soxhlet extraction. Total fats were extracted according to Soxhlet method with chloroform as an organic solvent 20 . About 10 g of the homogenized sample was extracted for 4-5 hours. Fat content per 100 g of sample is expressed in %.

Fatty acid methyl esters (FAMEs) preparation and determination
Approximately 0.2 g of extracted lipids from each specimen was transferred into glass cuvettes and 1.

The Lipid Quality Indexes
From the data on the fatty acid (FA) composition, the atherogenic and thrombogenic index were calculated using modified equations by Ulbricht and Southgate 22 and Garaffo et al 23 .
Myristic acid (C14:0) is considered to be 4-times more atherogenic than the other FA; thus a coefficient of 4 has been assigned to it. MUFA and n-6 PUFA have been assigned coefficients of 0.5 since they are less antiatherogenic than n-3 PUFA, which has been assigned a coefficient of 3 24 . (2) Flesh-lipid quality (FLQ) was calculated following the formulae by Abrami et al. 25 and Senso et al. 26 :

Statistical analysis
All descriptive variables were shown as mean ± standard error (SE). Shapiro-Wilk test was used for testing the distribution. Asymmetrically distributed variables were shown as median (interquartile range). Continuous variables were compared by Mann-Whitney Utest. All data were analyzed using IBM® SPSS® Statistics version 22 software. A p-value less than 0.05 was considered statistically significant. Cost-minimization analysis was conducted to assess the economic utility. Table 1 shows the total lipid content of homogenized tissues of freshwater and seawater species. Among freshwater fish, it was observed that fat content varied the most in  (Table 3). Results in Table 4  Committee on Nutrition, which is ~ 150 g. The evaluation of the costs of EPA+DHA using average prices of raw fish from the Serbian market was also presented.

Results
Median values (interquartile range) for lipid quality indexes are shown in Figure 2.
TI, OFA, and DFA within the freshwater and seawater fish groups did not differ

Discussion
It was noted that all species of freshwater fish (except trout) had a similar content of lipids and the obtained values were very low compared to tested species of seawater fish.
Ljubojević et al. 27 reported high variability (6.3-15%) in lipid content of common carp in Serbia. On the other hand, Łuczyńska et al. 28 shown lower values of the fat content in common carp found on Polish market (0.21-1.47%). However, according to the USDA database 29 , where data come from a variety of sources, the mean value for common carp lipid content is 5.6%. Ljubojević et al. 27 and Stancheva et al. 30 found about 4% of total lipid in catfish, while USDA database records value of 2.8%. These results are higher compared to our results for catfish. Previous study on fat quality of marketable freshwater fish species in Serbia have indicated that total lipids obtained for perch were in range from 1.5 up to 2.2% 27 , which is in agreement with our results as well as with results given by USDA database. Lipid content for bream obtained in this study is similar to previous data published by Łuczyńska et al. 31 . The muscles of trout contained more lipids than the other freshwater fish species. Our results are in accordance with the data obtained by Łuczynska et al. 31 and with the USDA database value (6.6%). In a recent study conducted by Bandarra et al. 32 , similar results for sardine fat content were obtained as in our research (14±2.9, SD), while in the study conducted by Zorica et al. 33 , the fluctuations in fat content were also reported for the same species of sardine from the middle eastern Adriatic Sea region. Wood et al. 37 have suggested that the ratio of PUFA/SFA in food should be above 0.4, and according to that all the fish species examined revealed a favourable PUFA/SFA ratio from 1.4 to 3.5. The DHA/EPA ratio for freshwater fish and most seawater fish analysed in this study are in agreement with the literature data 27,28,[30][31][32][33]35 . Only in carp, the pattern n-3/n-6 ratio was less than 1 in present study, while in all other fish samples, the content of n-3 fatty acids was several times higher than n-6 fatty acids. According to the research conducted by Buchtova et al. 38 and Ćirković et al. 39 , the carp grown on natural food had a high content of both n-6 and n-3 fatty acids. Ljubojević et al. 27 also reported n-3/n-6 ratio lower than 1 (0.48±0.18) for common carp. Sardine had the highest ratio of these fatty acids at 10.2. The n-6/n-3 ratio should not exceed 4 for the prevention of cardiovascular, heart, and certain chronic diseases 40 . All studied species meet this suggestion.
As it was expected, the content of n-3 fatty acids, EPA, and DHA was significantly When considering the optimal fish intake there is one more aspect that should be taken into account. Due to the presence of various contaminants in the aquatic environment, fish may be contaminated with persistent chemicals resulting in potential risks to human health. Health benefit and health risk of fish consumption is generally known as the nutritional-toxicological conflict 43 . Contaminant concentrations in fish species depend strongly on the species itself, its metabolism and feeding habits,   Table 3. Comparison of fatty acid content between freshwater and seawater fish groups (g/100 g).

Data shown as median values (interquartile range).
Bold values indicate statistical significance. P-value of less than 0.05 was considered statistically significant.