Article pubs.acs.org/JAFC
Content and Composition of Fatty Acids in Marine Oil Omega‑3 Supplements Cynthia Tyburczy Srigley* and Jeanne I. Rader† Office of Regulatory Science, Center for Food Safety and Applied Nutrition, Food and Drug Administration, 5100 Paint Branch Parkway, College Park, Maryland 20740, United States S Supporting Information *
ABSTRACT: Marine oil omega-3 supplements are among the most frequently consumed dietary supplements in the United States. However, few studies have evaluated the overall fatty acid composition of these products. We investigated the content and composition of fatty acids in 46 commercially available marine oil omega-3 supplements by gas chromatography with flame ionization detection using the 200 m SLB-IL111 ionic liquid column. Seventy-three fatty acid isomers were quantified, including n-6, n-4, n-3, and n-1 polyunsaturated fatty acids and trans isomers of eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3), the chromatographic separations of which we report for the first time on the 200 m SLB-IL111 column. Contents of EPA and DHA met their respective label declarations in more than 80% of the products examined. Eleven of the products (24%) carried the Food and Drug Administration’s qualified health claim for EPA and DHA omega-3 fatty acids. KEYWORDS: marine oil, SLB-IL111, EPA, DHA, trans PUFA, dietary supplement
■
INTRODUCTION Nutrition advisory committees in the United States (US) and other countries support the dietary intake of n-3 long chain polyunsaturated fatty acids (LCPUFA), specifically eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3), to reduce the risk of chronic disease,1−6 especially cardiovascular disease.7 Sources of EPA and DHA include fatty fish,7 dietary supplements and fortified foods that contain one or more sources of marine oil and for infants, breast milk and fortified infant formula. Dietary supplements containing fish oil (FO)/omega-3 LCPUFA/DHA are reported to be the most common nonvitamin, nonmineral dietary supplements consumed for health reasons among US adults and the second most common dietary supplements consumed by US children.8 However, despite the increasing use of such supplements, few studies have evaluated their overall fatty acid compositions.9−13 The fatty acid composition of marine oils is most commonly determined by gas chromatography with flame ionization detection (GC-FID). Official Method Ce 1i-0714of the American Oil Chemists’ Society (AOCS) has been validated for the determination of saturated fatty acids (SFA), cismonounsaturated fatty acids (MUFA), and cis-PUFA in marine and other oils containing LCPUFA. However, this method,14 which recommends the use of a 30 m polyethylene glycol (PEG) column, is not suitable for the determination of trans fatty acids because the PEG stationary phase offers a limited capacity for resolving geometric (cis/trans) fatty acid methyl ester (FAME) isomers.15 trans-Isomers of EPA and DHA are formed during high temperature processing of marine oils,16,17 and the coelution of these isomers with their corresponding allcis isomers on PEG columns may lead to an overestimation of EPA and DHA contents in the products analyzed.15 In contrast, GC columns containing ionic liquid stationary phases are This article not subject to U.S. Copyright. Published 2014 by the American Chemical Society
increasingly recognized as powerful separation tools in the analysis of, among others, essential oils and fatty acids.18 These columns offer excellent retention selectivity and high thermal stability, making them particularly useful for the analysis of FAME from complex matrices such as milk fat,19 partially hydrogenated oil,19 and FO.20 Fardin-Kia et al.20 recently reported the separation of 125 fatty acid isomers derived from menhaden FO, including isoprenoid and furanoid fatty acids, in a single 90 min GC run on the 200 m SLB-IL111 ionic liquid column. In the present study, 46 marine oil omega-3 supplements were analyzed by GC-FID for the content and composition of fatty acids, including trans isomers of EPA and DHA, using the 200 m SLB-IL111 ionic liquid column. We report, for the first time, the chromatographic separation of the trans isomers of EPA and DHA on this column and verify the accuracy of our method using Standard Reference Material (SRM) 3275 “Omega-3 and Omega-6 Fatty Acids in Fish Oil” from the National Institute of Standards and Technology (NIST). Analyzed contents of EPA and DHA were compared with values declared on the Supplement Facts panels of the products. Estimated daily intakes for EPA and DHA, which were calculated based on analytical determinations of product composition and label declarations for product use, were also evaluated relative to guidance provided in the US Food and Drug Administration’s (FDA) qualified health claim for EPA and DHA omega-3 fatty acids.21,22 This study is the first to evaluate the use of FDA’s qualified health claim for marine oil omega-3 supplements, which states that “supportive but not Received: Revised: Accepted: Published: 7268
April 9, 2014 June 27, 2014 June 28, 2014 June 28, 2014 dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 1. Gas chromatogram of FAME derived from an SO supplement. in their ingredients lists (reportedly added for the purpose of increasing EPA and DHA potency, species undefined) and thus were classified as FO products. The remaining four supplements contained a combination of plant oil (e.g., borage, olive, and flax) and FO (i.e., plant/FO blends). Supplements were purchased in softgel (also referred to on product labels as capsule or minigel; n = 40) and liquid (n = 6) forms. Softgel products listed gelatin, glycerin, and water as ingredients. Forty-one of the products contained antioxidants (e.g., tocopherols, ascorbyl palmitate, and botanical extracts). Eleven of the products (24%) carried the qualified health claim for EPA and DHA. All products, which were collected for research purposes only, were stored at −80 °C until analysis and were analyzed prior to their expiration dates. Methylation. Methylation was performed in duplicate according to AOCS Official Method Ce 2-66.24 This method involves an alkali hydrolysis with 0.5 M NaOH in MeOH followed by methylation with fresh 14% BF3 in MeOH.24 For supplements, oil from 2−3 softgels was combined, then ∼50 mg was weighed into 16 × 150 mm glass extraction tubes for methylation. Following derivatization, FAMEs were solubilized in heptane (∼12.5 mg/mL) and stored in silanized, amber-colored autosampler vials. Preparation of Isomerized EPA and DHA FAME. Standards for the trans isomers of EPA and DHA were prepared from all-cis EPA and DHA FAME, respectively, by isomerization with p-toluenesulfinic acid according to Snyder and Scholfield.25 Samples were heated for a maximum of 3 min to limit the degree of isomerization. Isomerized EPA and DHA FAMEs were then fractionated by number of trans double bonds on silica gel plates impregnated with 10% silver nitrate (Analtech, Newark, DE). Plates were developed at −20 °C in toluene/ MeOH (85:15, v/v) according to Fournier et al.26 Individual bands
conclusive research shows that consumption of EPA and DHA omega-3 fatty acids may reduce the risk of coronary heart disease”.21,22
■
MATERIALS AND METHODS
Chemicals and Reagents. ACS reagent grade chemicals and solutions were purchased from Sigma-Aldrich (St. Louis, MO). HPLC grade solvents were purchased from Fisher Scientific (Pittsburgh, PA). FAME (EPA and DHA) and triacylglycerol (TAG; C11:0) standards were purchased from Nu-Chek Prep (Elysian, MN). Our choice of C11:0 TAG as internal standard (IS) was based on the fact that the content of C11:0 is negligible in FO and that the C11:0 FAME isomer is chromatographically resolved from other FAME on the 200 m SLBIL111 column.20 SRM 3275. SRM 3275 was purchased from NIST (Gaithersburg, MD). This reference material consists of three fish oils: SRM 3275-1, a concentrate high in DHA; SRM 3275-2, an anchovy oil high in EPA and DHA; and SRM 3275-3, a concentrate containing nominally 60% omega-3 LCPUFA. The SRM was stored at 4 °C until analysis. Details of the development of SRM 3275 have been reported.23 Marine Oil Omega-3 Supplements. Marine oil omega-3 supplements were purchased online between May and July of 2012. Products were included in the study if their ingredient statements listed a source of marine oil and the Supplement Facts panel listed declarations for EPA and DHA. Eighteen of the products listed FO (e.g., anchovy, sardine, and mackerel) as the primary ingredient. Fifteen of the products listed FO concentrate (from anchovy, sardine, mackerel, or species undefined) as the primary ingredient. Nine of the supplements were marketed as salmon oil (SO) supplements. However, two of these products included an additional FO source 7269
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 2. Partial gas chromatograms of FAME derived from an FO concentrate in the regions in which EPA (panel A) and DHA (panel B) elute. For each panel, the top partial gas chromatogram corresponds to FAME derived from the FO concentrate, while the lower partial gas chromatograms correspond to isomerized EPA and DHA FAME fractions (F1−F3). *20:5 contains C20:5n-4 plus t-EPA. were identified by comparison with published thin layer chromatography (TLC) plate images.27 GC Instrumentation. Analysis by GC-FID was performed in split mode on a 6890N GC (Agilent Technologies, Wilmington, DE) equipped with a 200 m SLB-IL111 column, which was formed by connecting two 100 m SLB-IL111 columns (Supelco, Bellefonte, PA) using an Agilent press fit quartz column connector. The oven was maintained at 170 °C for 50 min, then ramped at 6 °C/min to 185 °C, held for 50 min, then ramped at 20 °C/min to 220 °C, and held for 5 min. The oven was maintained at 168 °C for the 5 min postrun. A single GC run time was 114.25 min. The flow program for the H2 carrier gas was as follows: 1.6 mL/min for 35 min, ramped at 0.3 mL/ min/min to 3.0 mL/min, and held for ∼70 min. The inlet and FID temperatures were 300 and 235 °C, respectively. The split ratio was 100:1, and the injection volume was 1 μL. A complementary analysis of FAME, which was performed to resolve FAME coeluting on the 200 m SLB-IL111 column, was carried out on the 30 m Supelcowax 10 column (Supelco, Bellefonte, PA) according to AOCS Official Method Ce 1i-07.14 Calculations for fatty acid content were performed using theoretical correction factors.14 GC-mass spectrometry (MS) analysis was performed on a 7890A GC coupled to a 240 MS ion trap mass spectrometer (Agilent Technologies) and equipped with a 200 m SLB-IL111 column. The oven was maintained at 170 °C for 63 min, then ramped at 6 °C/min to 185 °C, held for 20 min, then ramped at 15 °C/min to 220 °C, and held for 22 min. The oven was maintained at 170 °C for a 5 min postrun. The flow of the He carrier gas was held at 1.6 mL/min for 45.5 min, then ramped at 0.3 mL/min/min to 2.0 mL/min, and held for 63 min. FAMEs were identified by covalent adduct chemical ionization (CACI) tandem MS using acetonitrile as the reagent gas.28 Calculations for Supplements. Single serving contents of individual fatty acids were calculated based on analyzed fatty acid
contents (as % of total weight) and the weight of oil in a single serving. For products for which serving size was reported by volume (e.g., 5 mL), 10 servings were pipetted into tared weighing boats and weighed to the nearest 0.0001 g on an AE100 analytical balance (MettlerToledo, Columbus, OH). For products for which serving size was reported by quantity (e.g., two softgels), the weights of 20 servings or a minimum of 1/5 of the contents of the bottle were recorded. Softgel capsules were then punctured, rinsed with hexane, and the weights of the dry capsule shells recorded. The weight of oil in a single serving was calculated as the mean serving weight (capsule shell plus oil) minus the mean capsule shell weight. Statistical Analysis. Statistical analysis was performed using JMP (version 9.0.0. 2010, SAS Institute, Cary, NC). One way analysis of variance was used to compare mean values, and Student’s t-test was applied when significant differences were observed (F < 0.05). Values are reported as the means ± standard deviations (SD). Statistical outliers were determined using Grubbs’ test.29
■
RESULTS Chromatography. The chromatographic separation of FAME derived from an SO supplement is presented in Figure 1. Seventy-three FAME isomers, accounting for 91.9 ± 3.4% of total fat, were included in this analysis. Chromatographic conditions presented in this article are similar to those reported by Fardin-Kia et al.20 but included a final temperature ramp to 220 °C, maintained for 5 min. The extended temperature program ensured complete elution of analytes in a single GC run and avoided the occurrence of ghost peaks in subsequent chromatograms due to analyte carryover. 7270
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 3. Partial GC-CACI-MS chromatograms of FAME derived from an FO concentrate product in the regions in which EPA (panel A) and DHA (panel B) elute. For each panel, the top partial GC-MS chromatogram corresponds to the TIC. The lower partial GC-MS chromatogram is a reconstructed ion chromatogram specific for ions attributed to C20:5 and C22:6 FAME (molecular ion, [MH]+; m/z 54 adduct ion, [M+54]+).28
Figure 4. Gas chromatograms of FAME derived from the three fish oils of SRM 3275. FAME for which certified and reference mass fraction values are reported in the Certificate of Analysis are labeled.
silver ion TLC plates according to the number of trans double bonds. Plates were sprayed with 2′,7′-dicholorfluorescein solution (0.1% in isopropanol) and bands identified by
Standards for the trans isomers of EPA and DHA FAME, which were prepared from all-cis EPA and DHA FAME by isomerization with p-toluenesulfinic acid, were fractionated on 7271
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Table 1. Recoveries of Individual FAME from the Three Fish Oils of NIST SRM 3275a SRM 3275-1 FAME C12:0 C14:0 C14:1cis-9 C16:0 C16:1cis-9 C18:0 C18:1cis-9 C18:1cis-11 C18:2n-6 C18:3n-6 C18:3n-3 C20:0 C20:1cis-11 C20:4n-6 C20:5n-3 (EPA) C22:5n-3 C22:6n-3 (DHA) C22:0 C22:1cis-13 C24:0 C24:1cis-15
certificate value (mg/g) 1.094 ± 0.053 5.25 7.43 4.22 11.25 5.33 2.31 0.344 1.21 1.910
± 0.35 ± 0.24 ± 0.13 ± 0.93 ± 0.35 ± 0.19 ± 0.025 ± 0.05 ± 0.071
5.69 113 70.2 429 4.02 4.76
± 0.19 ± 12 ± 1.1 ± 15 ± 0.24 ± 0.22
SRM 3275-2
analyzed value (mg/g)b
recovery (%)c
certificate value (mg/g)
1.9 ± 0.0
174
3.45 ± 0.40
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
106 102 101 114 102 103 98 143 154
5.6 7.6 4.3 12.8 5.4 2.4 0.3 1.7 3.0 11.7 6.4 112.0 75.5 436.0 5.2 7.7
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0
112 99 107 102 129 162
8.01 5.83 12.94 22.1 9.24 3.00 0.507 1.42 0.357 6.66 22.9 394 67.6 187 1.396 3.43 0.618
± 0.44 ± 0.45 ± 0.62 ± 1.6 ± 0.77 ± 0.42 ± 0.043 ± 0.12 ± 0.027 ± 0.69 ± 1.0 ± 17 ± 2.3 ±8 ± 0.046 ± 0.32 ± 0.028
SRM 3275-3
analyzed value (mg/g)
recovery (%)
0.1 ± 0.0 4.1 ± 0.0
120
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
109 103 108 109 102 109 90 128 1182 252 116 103 109 103 209 106 136
8.8 6.0 13.9 24.2 9.5 3.3 0.5 1.8 4.2 16.8 26.6 407.5 73.9 192.4 2.9 3.7 0.8
0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.4 0.1 0.1 0.0 0.0 0.0
certificate value (mg/g) 0.95 67.9 0.964 186.9 85.7 38.0 112.3 38.5 13.49 1.771 6.61 1.14 2.92
± 0.12 ± 1.5 ± 0.043 ± 9.4 ± 3.1 ± 5.7 ± 2.6 ± 2.2 ± 0.45 ± 0.099 ± 0.31 ± 0.26 ± 0.14
154 27.0 104 0.502 1.61 0.441 3.78
±9 ± 1.1 ±5 ± 0.047 ± 0.11 ± 0.013 ± 0.29
analyzed value (mg/g) 1.1 61.0 0.5 155.4 68.4 36.4 106.1 31.8 13.3 2.0 7.4 1.9 9.3 14.1 163.4 29.0 106.5 0.8 1.7 0.5 4.4
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.0 0.4 0.0 1.3 0.5 0.4 1.0 0.3 0.1 0.0 0.1 0.0 0.1 0.2 2.3 0.5 1.3 0.0 0.0 0.0 0.0
recovery (%) 113 90 49 83 80 96 94 83 99 113 112 164 320 106 107 102 164 104 103 116
a
Mass fraction values for individual FAME are reported in the Certificate of Analysis for SRM 3275. Value assignments are based on two independent methods and the reported mean value represents the mean of the mean values from each method. Certified mass fraction values (bold font) are reported to represent the highest level of confidence in accuracy and account for all known or suspected sources of bias. The associated uncertainty is expressed at a 95% level of confidence and represents the combined standard uncertainty multiplied by a coverage factor k. Reference mass fraction values (not bold font) are reported to represent the best estimate of true value based on available data, and the associated uncertainties may reflect only measurement reproducibility. bAnalyzed values represent the means ± SD of two independent determinations. cRecovery calculated as analyzed/certificate value × 100.
comparison with published TLC images.27 Fractions corresponding to all-cis FAME, mono-trans FAME, and FAME with two or more trans double bonds were collected and analyzed by GC-FID for retention time. Figure 2 presents partial gas chromatograms for the trans EPA and DHA FAME standards and for FAME derived from an FO concentrate product. Peaks for five mono-trans EPA isomers were observed in both the mono-trans EPA standard and the marine oil omega-3 supplements (Figure 2A). The mono-trans DHA standard showed five mono-trans DHA peaks, but only four mono-trans DHA peaks were observed in the supplements (Figure 2B). The first eluting mono-trans DHA isomer coeluted with C22:5n-3 in the supplements (Figure 2B). FAMEs with two or more trans double bonds, if present in the supplements, were found at or near the level of detection. The presence of trans isomers of EPA and DHA in the supplements was also verified by GC-CACI-MS analysis (Figure 3). The top chromatogram in Figure 3A corresponds to the total ion chromatogram (TIC) for FAMEs derived from an FO concentrate product. The lower chromatogram corresponds to the reconstructed ion chromatogram for m/z 317 ([MH]+) and m/z 370 ([M+54]+) ions, which are attributed to C20:5 FAME in the FO concentrate product. Peaks in the reconstructed ion chromatogram at 80.3, 82.3, and 85.8 min (Figure 3A) are consistent with those observed in the partial gas chromatogram for the mono-trans EPA standard (Figure 2A), providing further confirmation of the presence of mono-trans EPA in the supplements. Similarly, in Figure 3B, peaks in the reconstructed ion chromatogram, which correspond to m/z 343 ([MH]+) and m/z 396 ([M+54]+)
ions that are specific to C22:6 FAME in the FO concentrate product, are consistent with those in the partial gas chromatogram for the mono-trans DHA FAME standard (Figure 2B). In addition, the first eluting mono-trans DHA FAME was found to coelute with C22:5n-3 (Figures 2B and 3B). GC-MS analysis indicated that the area of the first eluting mono-trans DHA isomer accounted for 20% of the total peak area for mono-trans DHA in the two products with the highest concentrations of trans DHA (Figure 3B). Therefore, for all samples, the quantitation of trans DHA included a correction factor (i.e., integrated area of mono-trans DHA/0.8) to account for the area of the mono-trans DHA FAME that coeluted with C22:5n3. Analysis of SRM 3275. GC-FID chromatograms for the separation of FAMEs derived from the three SRM fish oils are presented in Figure 4. Analyzed values for 21 FAMEs are presented in Table 1 along with corresponding values reported in the Certificate of Analysis. Recoveries for EPA and DHA, which were calculated as analyzed/certificate values × 100, varied from 99% to 106% and from 102% to 103%, respectively, among the three SRM fish oils. Analyzed values for EPA and DHA were all within the range of mass fraction values covered by the expanded uncertainty. For the remaining FAME, the majority of recoveries were within the range of 80% to 120%, whereas for a few FAMEs, recoveries varied from 49% to 1182%. Analyzed values for three FAMEs (C20:0, C20:1cis-11, and C22:0) were significantly greater than the corresponding certificate values, indicating the presence of coeluting peaks on the 200 m SLB-IL111 column. 7272
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Table 2. Concentrations of Fatty Acids in the Marine Oil Omega-3 Supplement Groups fatty acid (% of total fat)a
FO
FO concentrate
Σ SFA Σ MUFA n-6 PUFA C18:2n-6 C18:3n-6 C20:2n-6 C20:3n-6 C20:4n-6 C22:5n-6 Σ n-6 PUFA n-4 PUFA C16:2n-4 C16:3n-4 C18:2n-4 C18:3n-4 Σ n-4 PUFA n-3 PUFA (all-cis) C18:3n-3 C18:4n-3 C20:4n-3 C20:5n-3 (EPA) C21:5n-3 C22:5n-3 C22:6n-3 (DHA) Σ n-3 PUFA n-1 PUFA C16:4n-1 C18:4n-1 Σ n-1 PUFA Σ t-EPA + t-DHA
17.4 ± 9.2ab 17.0 ± 4.4b
11.7 ± 11.6b 13.3 ± 7.3b
SO
Pb
23.0 ± 1.8a 30.6 ± 6.1a
0.03
Content and Composition of Fatty Acids in Marine Oil Omega‑3 Supplements Cynthia Tyburczy Srigley* and Jeanne I. Rader† Office of Regulatory Science, Center for Food Safety and Applied Nutrition, Food and Drug Administration, 5100 Paint Branch Parkway, College Park, Maryland 20740, United States S Supporting Information *
ABSTRACT: Marine oil omega-3 supplements are among the most frequently consumed dietary supplements in the United States. However, few studies have evaluated the overall fatty acid composition of these products. We investigated the content and composition of fatty acids in 46 commercially available marine oil omega-3 supplements by gas chromatography with flame ionization detection using the 200 m SLB-IL111 ionic liquid column. Seventy-three fatty acid isomers were quantified, including n-6, n-4, n-3, and n-1 polyunsaturated fatty acids and trans isomers of eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3), the chromatographic separations of which we report for the first time on the 200 m SLB-IL111 column. Contents of EPA and DHA met their respective label declarations in more than 80% of the products examined. Eleven of the products (24%) carried the Food and Drug Administration’s qualified health claim for EPA and DHA omega-3 fatty acids. KEYWORDS: marine oil, SLB-IL111, EPA, DHA, trans PUFA, dietary supplement
■
INTRODUCTION Nutrition advisory committees in the United States (US) and other countries support the dietary intake of n-3 long chain polyunsaturated fatty acids (LCPUFA), specifically eicosapentaenoic acid (EPA; C20:5n-3) and docosahexaenoic acid (DHA; C22:6n-3), to reduce the risk of chronic disease,1−6 especially cardiovascular disease.7 Sources of EPA and DHA include fatty fish,7 dietary supplements and fortified foods that contain one or more sources of marine oil and for infants, breast milk and fortified infant formula. Dietary supplements containing fish oil (FO)/omega-3 LCPUFA/DHA are reported to be the most common nonvitamin, nonmineral dietary supplements consumed for health reasons among US adults and the second most common dietary supplements consumed by US children.8 However, despite the increasing use of such supplements, few studies have evaluated their overall fatty acid compositions.9−13 The fatty acid composition of marine oils is most commonly determined by gas chromatography with flame ionization detection (GC-FID). Official Method Ce 1i-0714of the American Oil Chemists’ Society (AOCS) has been validated for the determination of saturated fatty acids (SFA), cismonounsaturated fatty acids (MUFA), and cis-PUFA in marine and other oils containing LCPUFA. However, this method,14 which recommends the use of a 30 m polyethylene glycol (PEG) column, is not suitable for the determination of trans fatty acids because the PEG stationary phase offers a limited capacity for resolving geometric (cis/trans) fatty acid methyl ester (FAME) isomers.15 trans-Isomers of EPA and DHA are formed during high temperature processing of marine oils,16,17 and the coelution of these isomers with their corresponding allcis isomers on PEG columns may lead to an overestimation of EPA and DHA contents in the products analyzed.15 In contrast, GC columns containing ionic liquid stationary phases are This article not subject to U.S. Copyright. Published 2014 by the American Chemical Society
increasingly recognized as powerful separation tools in the analysis of, among others, essential oils and fatty acids.18 These columns offer excellent retention selectivity and high thermal stability, making them particularly useful for the analysis of FAME from complex matrices such as milk fat,19 partially hydrogenated oil,19 and FO.20 Fardin-Kia et al.20 recently reported the separation of 125 fatty acid isomers derived from menhaden FO, including isoprenoid and furanoid fatty acids, in a single 90 min GC run on the 200 m SLB-IL111 ionic liquid column. In the present study, 46 marine oil omega-3 supplements were analyzed by GC-FID for the content and composition of fatty acids, including trans isomers of EPA and DHA, using the 200 m SLB-IL111 ionic liquid column. We report, for the first time, the chromatographic separation of the trans isomers of EPA and DHA on this column and verify the accuracy of our method using Standard Reference Material (SRM) 3275 “Omega-3 and Omega-6 Fatty Acids in Fish Oil” from the National Institute of Standards and Technology (NIST). Analyzed contents of EPA and DHA were compared with values declared on the Supplement Facts panels of the products. Estimated daily intakes for EPA and DHA, which were calculated based on analytical determinations of product composition and label declarations for product use, were also evaluated relative to guidance provided in the US Food and Drug Administration’s (FDA) qualified health claim for EPA and DHA omega-3 fatty acids.21,22 This study is the first to evaluate the use of FDA’s qualified health claim for marine oil omega-3 supplements, which states that “supportive but not Received: Revised: Accepted: Published: 7268
April 9, 2014 June 27, 2014 June 28, 2014 June 28, 2014 dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 1. Gas chromatogram of FAME derived from an SO supplement. in their ingredients lists (reportedly added for the purpose of increasing EPA and DHA potency, species undefined) and thus were classified as FO products. The remaining four supplements contained a combination of plant oil (e.g., borage, olive, and flax) and FO (i.e., plant/FO blends). Supplements were purchased in softgel (also referred to on product labels as capsule or minigel; n = 40) and liquid (n = 6) forms. Softgel products listed gelatin, glycerin, and water as ingredients. Forty-one of the products contained antioxidants (e.g., tocopherols, ascorbyl palmitate, and botanical extracts). Eleven of the products (24%) carried the qualified health claim for EPA and DHA. All products, which were collected for research purposes only, were stored at −80 °C until analysis and were analyzed prior to their expiration dates. Methylation. Methylation was performed in duplicate according to AOCS Official Method Ce 2-66.24 This method involves an alkali hydrolysis with 0.5 M NaOH in MeOH followed by methylation with fresh 14% BF3 in MeOH.24 For supplements, oil from 2−3 softgels was combined, then ∼50 mg was weighed into 16 × 150 mm glass extraction tubes for methylation. Following derivatization, FAMEs were solubilized in heptane (∼12.5 mg/mL) and stored in silanized, amber-colored autosampler vials. Preparation of Isomerized EPA and DHA FAME. Standards for the trans isomers of EPA and DHA were prepared from all-cis EPA and DHA FAME, respectively, by isomerization with p-toluenesulfinic acid according to Snyder and Scholfield.25 Samples were heated for a maximum of 3 min to limit the degree of isomerization. Isomerized EPA and DHA FAMEs were then fractionated by number of trans double bonds on silica gel plates impregnated with 10% silver nitrate (Analtech, Newark, DE). Plates were developed at −20 °C in toluene/ MeOH (85:15, v/v) according to Fournier et al.26 Individual bands
conclusive research shows that consumption of EPA and DHA omega-3 fatty acids may reduce the risk of coronary heart disease”.21,22
■
MATERIALS AND METHODS
Chemicals and Reagents. ACS reagent grade chemicals and solutions were purchased from Sigma-Aldrich (St. Louis, MO). HPLC grade solvents were purchased from Fisher Scientific (Pittsburgh, PA). FAME (EPA and DHA) and triacylglycerol (TAG; C11:0) standards were purchased from Nu-Chek Prep (Elysian, MN). Our choice of C11:0 TAG as internal standard (IS) was based on the fact that the content of C11:0 is negligible in FO and that the C11:0 FAME isomer is chromatographically resolved from other FAME on the 200 m SLBIL111 column.20 SRM 3275. SRM 3275 was purchased from NIST (Gaithersburg, MD). This reference material consists of three fish oils: SRM 3275-1, a concentrate high in DHA; SRM 3275-2, an anchovy oil high in EPA and DHA; and SRM 3275-3, a concentrate containing nominally 60% omega-3 LCPUFA. The SRM was stored at 4 °C until analysis. Details of the development of SRM 3275 have been reported.23 Marine Oil Omega-3 Supplements. Marine oil omega-3 supplements were purchased online between May and July of 2012. Products were included in the study if their ingredient statements listed a source of marine oil and the Supplement Facts panel listed declarations for EPA and DHA. Eighteen of the products listed FO (e.g., anchovy, sardine, and mackerel) as the primary ingredient. Fifteen of the products listed FO concentrate (from anchovy, sardine, mackerel, or species undefined) as the primary ingredient. Nine of the supplements were marketed as salmon oil (SO) supplements. However, two of these products included an additional FO source 7269
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 2. Partial gas chromatograms of FAME derived from an FO concentrate in the regions in which EPA (panel A) and DHA (panel B) elute. For each panel, the top partial gas chromatogram corresponds to FAME derived from the FO concentrate, while the lower partial gas chromatograms correspond to isomerized EPA and DHA FAME fractions (F1−F3). *20:5 contains C20:5n-4 plus t-EPA. were identified by comparison with published thin layer chromatography (TLC) plate images.27 GC Instrumentation. Analysis by GC-FID was performed in split mode on a 6890N GC (Agilent Technologies, Wilmington, DE) equipped with a 200 m SLB-IL111 column, which was formed by connecting two 100 m SLB-IL111 columns (Supelco, Bellefonte, PA) using an Agilent press fit quartz column connector. The oven was maintained at 170 °C for 50 min, then ramped at 6 °C/min to 185 °C, held for 50 min, then ramped at 20 °C/min to 220 °C, and held for 5 min. The oven was maintained at 168 °C for the 5 min postrun. A single GC run time was 114.25 min. The flow program for the H2 carrier gas was as follows: 1.6 mL/min for 35 min, ramped at 0.3 mL/ min/min to 3.0 mL/min, and held for ∼70 min. The inlet and FID temperatures were 300 and 235 °C, respectively. The split ratio was 100:1, and the injection volume was 1 μL. A complementary analysis of FAME, which was performed to resolve FAME coeluting on the 200 m SLB-IL111 column, was carried out on the 30 m Supelcowax 10 column (Supelco, Bellefonte, PA) according to AOCS Official Method Ce 1i-07.14 Calculations for fatty acid content were performed using theoretical correction factors.14 GC-mass spectrometry (MS) analysis was performed on a 7890A GC coupled to a 240 MS ion trap mass spectrometer (Agilent Technologies) and equipped with a 200 m SLB-IL111 column. The oven was maintained at 170 °C for 63 min, then ramped at 6 °C/min to 185 °C, held for 20 min, then ramped at 15 °C/min to 220 °C, and held for 22 min. The oven was maintained at 170 °C for a 5 min postrun. The flow of the He carrier gas was held at 1.6 mL/min for 45.5 min, then ramped at 0.3 mL/min/min to 2.0 mL/min, and held for 63 min. FAMEs were identified by covalent adduct chemical ionization (CACI) tandem MS using acetonitrile as the reagent gas.28 Calculations for Supplements. Single serving contents of individual fatty acids were calculated based on analyzed fatty acid
contents (as % of total weight) and the weight of oil in a single serving. For products for which serving size was reported by volume (e.g., 5 mL), 10 servings were pipetted into tared weighing boats and weighed to the nearest 0.0001 g on an AE100 analytical balance (MettlerToledo, Columbus, OH). For products for which serving size was reported by quantity (e.g., two softgels), the weights of 20 servings or a minimum of 1/5 of the contents of the bottle were recorded. Softgel capsules were then punctured, rinsed with hexane, and the weights of the dry capsule shells recorded. The weight of oil in a single serving was calculated as the mean serving weight (capsule shell plus oil) minus the mean capsule shell weight. Statistical Analysis. Statistical analysis was performed using JMP (version 9.0.0. 2010, SAS Institute, Cary, NC). One way analysis of variance was used to compare mean values, and Student’s t-test was applied when significant differences were observed (F < 0.05). Values are reported as the means ± standard deviations (SD). Statistical outliers were determined using Grubbs’ test.29
■
RESULTS Chromatography. The chromatographic separation of FAME derived from an SO supplement is presented in Figure 1. Seventy-three FAME isomers, accounting for 91.9 ± 3.4% of total fat, were included in this analysis. Chromatographic conditions presented in this article are similar to those reported by Fardin-Kia et al.20 but included a final temperature ramp to 220 °C, maintained for 5 min. The extended temperature program ensured complete elution of analytes in a single GC run and avoided the occurrence of ghost peaks in subsequent chromatograms due to analyte carryover. 7270
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Figure 3. Partial GC-CACI-MS chromatograms of FAME derived from an FO concentrate product in the regions in which EPA (panel A) and DHA (panel B) elute. For each panel, the top partial GC-MS chromatogram corresponds to the TIC. The lower partial GC-MS chromatogram is a reconstructed ion chromatogram specific for ions attributed to C20:5 and C22:6 FAME (molecular ion, [MH]+; m/z 54 adduct ion, [M+54]+).28
Figure 4. Gas chromatograms of FAME derived from the three fish oils of SRM 3275. FAME for which certified and reference mass fraction values are reported in the Certificate of Analysis are labeled.
silver ion TLC plates according to the number of trans double bonds. Plates were sprayed with 2′,7′-dicholorfluorescein solution (0.1% in isopropanol) and bands identified by
Standards for the trans isomers of EPA and DHA FAME, which were prepared from all-cis EPA and DHA FAME by isomerization with p-toluenesulfinic acid, were fractionated on 7271
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Table 1. Recoveries of Individual FAME from the Three Fish Oils of NIST SRM 3275a SRM 3275-1 FAME C12:0 C14:0 C14:1cis-9 C16:0 C16:1cis-9 C18:0 C18:1cis-9 C18:1cis-11 C18:2n-6 C18:3n-6 C18:3n-3 C20:0 C20:1cis-11 C20:4n-6 C20:5n-3 (EPA) C22:5n-3 C22:6n-3 (DHA) C22:0 C22:1cis-13 C24:0 C24:1cis-15
certificate value (mg/g) 1.094 ± 0.053 5.25 7.43 4.22 11.25 5.33 2.31 0.344 1.21 1.910
± 0.35 ± 0.24 ± 0.13 ± 0.93 ± 0.35 ± 0.19 ± 0.025 ± 0.05 ± 0.071
5.69 113 70.2 429 4.02 4.76
± 0.19 ± 12 ± 1.1 ± 15 ± 0.24 ± 0.22
SRM 3275-2
analyzed value (mg/g)b
recovery (%)c
certificate value (mg/g)
1.9 ± 0.0
174
3.45 ± 0.40
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
106 102 101 114 102 103 98 143 154
5.6 7.6 4.3 12.8 5.4 2.4 0.3 1.7 3.0 11.7 6.4 112.0 75.5 436.0 5.2 7.7
0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0
112 99 107 102 129 162
8.01 5.83 12.94 22.1 9.24 3.00 0.507 1.42 0.357 6.66 22.9 394 67.6 187 1.396 3.43 0.618
± 0.44 ± 0.45 ± 0.62 ± 1.6 ± 0.77 ± 0.42 ± 0.043 ± 0.12 ± 0.027 ± 0.69 ± 1.0 ± 17 ± 2.3 ±8 ± 0.046 ± 0.32 ± 0.028
SRM 3275-3
analyzed value (mg/g)
recovery (%)
0.1 ± 0.0 4.1 ± 0.0
120
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
109 103 108 109 102 109 90 128 1182 252 116 103 109 103 209 106 136
8.8 6.0 13.9 24.2 9.5 3.3 0.5 1.8 4.2 16.8 26.6 407.5 73.9 192.4 2.9 3.7 0.8
0.0 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.4 0.1 0.1 0.0 0.0 0.0
certificate value (mg/g) 0.95 67.9 0.964 186.9 85.7 38.0 112.3 38.5 13.49 1.771 6.61 1.14 2.92
± 0.12 ± 1.5 ± 0.043 ± 9.4 ± 3.1 ± 5.7 ± 2.6 ± 2.2 ± 0.45 ± 0.099 ± 0.31 ± 0.26 ± 0.14
154 27.0 104 0.502 1.61 0.441 3.78
±9 ± 1.1 ±5 ± 0.047 ± 0.11 ± 0.013 ± 0.29
analyzed value (mg/g) 1.1 61.0 0.5 155.4 68.4 36.4 106.1 31.8 13.3 2.0 7.4 1.9 9.3 14.1 163.4 29.0 106.5 0.8 1.7 0.5 4.4
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
0.0 0.4 0.0 1.3 0.5 0.4 1.0 0.3 0.1 0.0 0.1 0.0 0.1 0.2 2.3 0.5 1.3 0.0 0.0 0.0 0.0
recovery (%) 113 90 49 83 80 96 94 83 99 113 112 164 320 106 107 102 164 104 103 116
a
Mass fraction values for individual FAME are reported in the Certificate of Analysis for SRM 3275. Value assignments are based on two independent methods and the reported mean value represents the mean of the mean values from each method. Certified mass fraction values (bold font) are reported to represent the highest level of confidence in accuracy and account for all known or suspected sources of bias. The associated uncertainty is expressed at a 95% level of confidence and represents the combined standard uncertainty multiplied by a coverage factor k. Reference mass fraction values (not bold font) are reported to represent the best estimate of true value based on available data, and the associated uncertainties may reflect only measurement reproducibility. bAnalyzed values represent the means ± SD of two independent determinations. cRecovery calculated as analyzed/certificate value × 100.
comparison with published TLC images.27 Fractions corresponding to all-cis FAME, mono-trans FAME, and FAME with two or more trans double bonds were collected and analyzed by GC-FID for retention time. Figure 2 presents partial gas chromatograms for the trans EPA and DHA FAME standards and for FAME derived from an FO concentrate product. Peaks for five mono-trans EPA isomers were observed in both the mono-trans EPA standard and the marine oil omega-3 supplements (Figure 2A). The mono-trans DHA standard showed five mono-trans DHA peaks, but only four mono-trans DHA peaks were observed in the supplements (Figure 2B). The first eluting mono-trans DHA isomer coeluted with C22:5n-3 in the supplements (Figure 2B). FAMEs with two or more trans double bonds, if present in the supplements, were found at or near the level of detection. The presence of trans isomers of EPA and DHA in the supplements was also verified by GC-CACI-MS analysis (Figure 3). The top chromatogram in Figure 3A corresponds to the total ion chromatogram (TIC) for FAMEs derived from an FO concentrate product. The lower chromatogram corresponds to the reconstructed ion chromatogram for m/z 317 ([MH]+) and m/z 370 ([M+54]+) ions, which are attributed to C20:5 FAME in the FO concentrate product. Peaks in the reconstructed ion chromatogram at 80.3, 82.3, and 85.8 min (Figure 3A) are consistent with those observed in the partial gas chromatogram for the mono-trans EPA standard (Figure 2A), providing further confirmation of the presence of mono-trans EPA in the supplements. Similarly, in Figure 3B, peaks in the reconstructed ion chromatogram, which correspond to m/z 343 ([MH]+) and m/z 396 ([M+54]+)
ions that are specific to C22:6 FAME in the FO concentrate product, are consistent with those in the partial gas chromatogram for the mono-trans DHA FAME standard (Figure 2B). In addition, the first eluting mono-trans DHA FAME was found to coelute with C22:5n-3 (Figures 2B and 3B). GC-MS analysis indicated that the area of the first eluting mono-trans DHA isomer accounted for 20% of the total peak area for mono-trans DHA in the two products with the highest concentrations of trans DHA (Figure 3B). Therefore, for all samples, the quantitation of trans DHA included a correction factor (i.e., integrated area of mono-trans DHA/0.8) to account for the area of the mono-trans DHA FAME that coeluted with C22:5n3. Analysis of SRM 3275. GC-FID chromatograms for the separation of FAMEs derived from the three SRM fish oils are presented in Figure 4. Analyzed values for 21 FAMEs are presented in Table 1 along with corresponding values reported in the Certificate of Analysis. Recoveries for EPA and DHA, which were calculated as analyzed/certificate values × 100, varied from 99% to 106% and from 102% to 103%, respectively, among the three SRM fish oils. Analyzed values for EPA and DHA were all within the range of mass fraction values covered by the expanded uncertainty. For the remaining FAME, the majority of recoveries were within the range of 80% to 120%, whereas for a few FAMEs, recoveries varied from 49% to 1182%. Analyzed values for three FAMEs (C20:0, C20:1cis-11, and C22:0) were significantly greater than the corresponding certificate values, indicating the presence of coeluting peaks on the 200 m SLB-IL111 column. 7272
dx.doi.org/10.1021/jf5016973 | J. Agric. Food Chem. 2014, 62, 7268−7278
Journal of Agricultural and Food Chemistry
Article
Table 2. Concentrations of Fatty Acids in the Marine Oil Omega-3 Supplement Groups fatty acid (% of total fat)a
FO
FO concentrate
Σ SFA Σ MUFA n-6 PUFA C18:2n-6 C18:3n-6 C20:2n-6 C20:3n-6 C20:4n-6 C22:5n-6 Σ n-6 PUFA n-4 PUFA C16:2n-4 C16:3n-4 C18:2n-4 C18:3n-4 Σ n-4 PUFA n-3 PUFA (all-cis) C18:3n-3 C18:4n-3 C20:4n-3 C20:5n-3 (EPA) C21:5n-3 C22:5n-3 C22:6n-3 (DHA) Σ n-3 PUFA n-1 PUFA C16:4n-1 C18:4n-1 Σ n-1 PUFA Σ t-EPA + t-DHA
17.4 ± 9.2ab 17.0 ± 4.4b
11.7 ± 11.6b 13.3 ± 7.3b
SO
Pb
23.0 ± 1.8a 30.6 ± 6.1a
0.03
