Do n-3 Polyunsaturated Fatty Acids Increase or Decrease Lipid Peroxidation in Humans?

Abstract

Background:

Despite many known health benefits of n-3 polyunsaturated fatty acids (PUFA), there is a concern that their high degree of unsaturation may actually increase oxidative stress, lipid peroxidation (LPO), and chronic inflammatory diseases.

Methods:

In this review, we have analyzed results from published human studies regarding the effects of n-3 PUFA supplementation on markers of lipid peroxidation.

Results:

Of the 22 published human studies, nine found no change, eight a decrease, and five an increase in markers of LPO. These inconsistencies may be due to methods, subject characteristics, dose, duration, fatty acid and antioxidant composition of supplements, and basal diets. Methods used for analysis seem to be the most significant factor. Six of eight studies with a decrease in LPO determined F2-isoprostanes produced in vivo, and two determined plasma antioxidant capacity or hydroperoxides. n-3 PUFA can serve as scavengers for free radicals and also modulate expression of genes that determine the balance between oxidative and antioxidative status. Recent studies that monitored oxidation products of cholesterol and fatty acids support the hypothesis that n-3 PUFA decrease LPO. Most of the studies showing no change or increase in LPO determined markers that involved ex vivo sample preparation or oxidation (malondialdehyde, low-density lipoprotein oxidation, lipid hydroperoxides).

Conclusion:

A majority of studies do not indicate that n-3 PUFA increased LPO. Future studies need to investigate the effects of dose, duration, and composition of n-3 PUFA with standardized diets and methods on concentrations and types of LPO products produced.

Introduction

Dietary n-3 polyunsaturated fatty acids (PUFA) have numerous health benefits and help counteract symptoms of several chronic diseases. Diets rich in n-3 PUFA decreased platelet aggregation, plasma triglycerides, inflammation, and the number of total as well as atherogenic small dense low-density lipoprotein (LDL) particles.^1
–3 Despite health benefits, there is concern that n-3 PUFA may increase oxidative stress due to their high degree of unsaturation. However, this is debatable because studies have shown that not only unsaturation but also the position of double bonds are factors.⁴ Thus, it is not entirely clear whether n-3 PUFA reduce or increase oxidative stress.

A main consequence of oxidative stress is lipid peroxidation (LPO), which increases inflammation and the development and progression of a number of chronic inflammatory diseases, including some components of metabolic syndrome (insulin resistance, hypertension, nonalcoholic fatty liver disease), cardiovascular diseases, and diabetes.^5

–8 LPO is a term used to describe the oxidative degradation of lipids typically found in biological membranes such as phospholipids and cholesterol. This process can be achieved via enzymatic, nonenzymatic, nonradical, or nonenzymatic free radical-mediated peroxidation. LPO-mediated degradation compromises membrane integrity and can lead to changes in permeability and fluidity. It also propagates free radicals, which initially act locally with short half-lives. These local insults are followed by formation of more stable secondary breakdown products of lipid peroxides, such as alkanes, aldehydes, ketones, alcohols, and furans. Those products can then disseminate from their site of formation and diffuse across cellular membranes. Most secondary breakdown products of LPO are reactive due to their electrophilic nature. They play powerful roles in cell signaling, mainly in cell cycle regulation.^9

–12 The concentration of the secondary breakdown products can be quantified to assess the degree of in vivo lipid peroxidation. Some of those compounds include malondialdehyde (MDA), 4-hydroxynonenal (HNE), 4-hydroxyhexenal (HHE), F2-isoprostanes (F2-IsoPs)/8-isoprostaglandin F2α, acrolein, hexenal, hydroxyoctadenaoic acid (HODE), hydroxyeicosatetraenoic acid (HETE), 7α-hydroxycholesterol, and 7β-hydroxycholesterol. The in vivo concentrations of those specific LPO biomarkers vary greatly as does the sensitivity of their assay methods. Although some LPO biomarkers indicate the specific lipid substrate used to generate the measured breakdown product, others serve as an overall assessment of LPO.

Because of discrepancies among published results, concerns have been raised regarding whether n-3 PUFA reduce or induce oxidative stress and LPO. From all the publications in English that have focused on the effects of n-3 PUFA on LPO in humans, 27% reported an increase,^13

–18 32% a decrease,^19

–25 and 41% no change in LPO.^{26

–34} Several other studies showed that n-3 PUFA decreased LPO in cultured cells. However, those results from in vitro studies may not be representative of the in vivo milieu and should be interpreted with caution.^35

–39 The discrepancies of published studies need to be reconciled because n-3 PUFA are being recommended as components of healthy diets.

The aim of this review was to analyze results from published human studies that investigated effects of n-3 PUFA on LPO. For that, we compared dose, duration, and form of n-3 PUFA supplemented; antioxidants supplementation; health status of participants; and methods used to evaluate LPO. We analyzed studies published in past 23 years.

Methods

PubMed was used as the primary search engine to obtain publications for this review. The following criteria were used to determine exclusion or inclusion of n-3 PUFA and LPO human studies. All subjects received n-3 PUFA as combined or individual supplements of docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA) or a fish oil mixture. Studies selected were published in 1990–2014.

Results

Biomarkers of LPO

F2-isoprostanes

F2-IsoPs refer to a collection of compounds created by nonenzymatic, free radical attack of the arachidonic acid (AA) components in lipid membranes. Most AA present in animal cells is esterified to phospholipid components of lipid membranes. F2-IsoPs are hydrolyzed and released by the catalytic action of phospholipase A2 and platelet activating factor acetylhydrolase.

F2-IsoPs are usually measured from samples of urine, plasma, serum, and, less commonly, other body fluids. Temporarily, plasma levels of F2-IsoP peak quicker than urinary levels. Hence, urine and plasma samples should not be viewed interchangeably. It is desirable to include 24-hr pooled urine or at least several time points for measurement.⁴⁰ Specimens undergo various purification procedures and then are measured by gas chromatography/mass spectrometry (GC/MS), gas chromatography/tandem mass spectrometry (GC-MS/MS), liquid chromatography/tandem mass spectrometry (LC-MS/MS), or enzyme-linked immunosorbent assay (ELISA).⁴⁰ Mass spectrometry techniques do not measure F2-IsoP directly, but its metabolites. Usually, the metabolite quantified is 8-isoprostaglandin F2.⁴⁰ However, there are other isomers and various mass spectrometry techniques to assay their concentrations.

F2-IsoP is usually considered as the “gold standard” biomarker of lipid peroxidation in vivo and is increasingly used to examine n-3 PUFA and LPO. However, using F2-IsoP as a marker for LPO has several limitations. For example, detection methods such as high-performance liquid chromatography (HPLC)-MS/MS are expensive, time consuming and technically challenging. Currently available ELISA kits to determine F2-IsoP have poor reproducibility. Additionally, F2-IsoP is present in the picomole range in body fluids, which further complicates their quantification and use as biomarkers.

Malondialdehyde

The bulk of LPO-derived MDA is produced by nonenzymatic free radical peroxidation. However, it can also be produced enzymatically. The free radical degradation of membrane PUFA creates highly unstable and reactive lipid hydroperoxides, which decompose to form more stable aldehyde products, specifically MDA.⁴¹ MDA is generated by peroxidation of PUFA containing at least three double bonds, such as AA or DHA, and serves as a biomarker for both n-3 and n-6 PUFA LPO.⁴¹

MDA levels are most commonly quantified by the thiobarbituric reactive species (TBARS) assay. The TBARS assay is conducted with acidic conditions at high temperatures (95–100°C), and involves the reaction of MDA with thiobarbituric acid. TBARS assay measures not only the quantity of MDA, but also lipid hydroperoxides that yield MDA. Decomposition of those lipid hydroperoxides into MDA can be induced by transition metal ions, such as Fe³⁺ or Cu²⁺, included in the assay. Specificity of the TBARS assay is questionable due to the production of reactive aldehyde species from non-LPO sources, and cross-reactivity resulting from acidic and high temperature conditions of the assay. MDA concentrations assayed by those methods are normally in the nanomole range. MDA concentrations can also be quantified by HPLC; however, it is not commonly used. While it has high reproducibility and precision, because of high volatility of MDA, the concentrations obtained by this method are much lower than those obtained by TBARS. Thus, HPLC analysis under estimates and TBARS over estimates the MDA concentrations.

4-Hydroxy-2(E)-nonenal and 4-hydroxy-2(E)-hexenal

MDA and TBARS can be generated from peroxidation of both n-6 and n-3 PUFA. In contrast, HNE is generated by auto-oxidation of n-6 PUFA⁴² and HHE is produced from autooxidation of n-3 PUFA.⁴³ Alkenal species such as HHE and HNE are reactive substances and can be assayed by the unique adducts formed. For example, HNE readily reacts with cellular proteins forming adducts with cysteine, histidine, lysine, and serine.^42,44
–46 In addition, glutathione (GSH) conjugation is a key step in the metabolism of secondary LPO products such as HNE. The GSH moiety of these by products can be reduced to mercapturic acid, which is the major metabolite of HNE production found in urine.⁴¹ Because multiple metabolites can be formed, there are numerous methods to quantifying these byproducts. Some immunochemical methods, such as ELISA, can be used for quantification because of adducts formed between aldehydes and specific peptides.⁴⁷ Nonimmunological methods are based on the reaction of the aldehyde moiety with dinitrophenylhydrazine to form hydrazone compounds. They can be measured by spectrophotometric methods.⁴⁸ Those methodologies are rather nonspecific, because many carbonyl-modified proteins can be identified along with LPO alkenal byproducts.⁴¹ Thus, quantification methods were developed to determine specific adducts formed with peptides. Those include methods for quantifying carbonyl groups associated with sulfhydryl groups of LPO products that can be used to identify cysteine adducts.⁴⁹ Other peptide adducts formed with histidine and lysine can be quantified with HPLC when a strong reducing agent such as sodium borohydride (NaBH₄) is used.⁵⁰ The advantage of using HHE, HNE, or HHE as LPO biomarkers is the specificity of substrate. However, quantification of HNE and HHE is less sensitive than those of F2-IsoP or MDA.

Degree of LDL oxidation

The degree of LDL oxidation is of particular interest in LPO studies because of its direct link to atherosclerosis. Numerous studies have investigated the ex vivo, metal ion-induced oxidation of LDL lipoproteins, in addition to the quantification of other LPO biomarkers. Inducing agents include metal ions, such as copper and iron, and also other agents such as carbon tetrachloride and 2,2′-azobis(-amidinopropane)dihydrochloride. The time elapsed between introduction of oxidizing agents and the actual oxidation of LDL lipoproteins indicates the susceptibility of the lipoprotein to oxidation. A longer time lapse to oxidation indicates less susceptibility of the LDL to oxidation, and vice versa. Therefore, lag time is a commonly employed technique to assess LPO. However, the ex vivo measurement of LDL oxidizability may not correlate with in vivo oxidation of LDL. Incubation of LDL with several cell types such as monocytes, macrophages, or fibroblasts or metal ions such as copper in the absence of cells can result in the oxidative modification of LDL. Although both routes have similar properties, there are a number of differences, such as sensitivity to inhibition by superoxide dismutase and lipoxygenase inhibitors as well as levels of TBAR produced.

Oxidation products of individual fatty acids, such as HODE from linoleic acid and HETE from arachidonic acid, or α- and β-oxidation products of cholesterol, have also been used to assess lipid peroxidation. Other methods assess oxidative stress by a determination of oxygen radical absorbance capacity, Trolox equivalent antioxidant capacity, total antioxidant status, or enzyme activity, such as catalase and glutathione peroxidase. Because most of the surveyed studies in this review used the methods described above, our discussion will be limited to those methods.

Effects of n-3 PUFA on LPO

We found a total of 22 published reports that investigated the effects of feeding fish, or supplementing diets with fish oil, purified EPA or DHA. Nine studies reported no change,^{26

–34} eight a decrease,^19

–25 and five an increase^13

–18 in markers of LPO.

No change in LPO

Studies showing no change in LPO after n-3 PUFA supplementation are listed in Table 1. The amounts of n-3 PUFA varied from 529 mg/day²⁷ to 5.0 grams/day.^27,29 Intervention time ranged from a single administration to 5 months of supplementation.^32,34 One of those studies used purified DHA (2.14 grams/day)³³ and none used purified EPA. The remaining eight studies used a combined supplement of both EPA and DHA, with one study including docosapentaenoic acid (DPA) as well.³⁰ In most of these studies, concentrations of EPA were greater than DHA, except in one where the DHA concentration was more than three times that of EPA³²; three studies did not indicate the concentrations of individual fatty acids.

Table 1.

Human Studies in Which n-3 PUFA Did Not Change Lipid Peroxidation

Authors	Participants (age), n	n-3 PUFA dose; intervention time; placebo/comparison; study design	Sample type; biomarker; time of measurement	Antioxidant	Comments
Nenseter et al.²⁶	Normolipidemic men and women (27–63); n=11–12	2.83 grams EPA+1.78 grams DHA/day; 4 months; 6 grams corn oil; randomized placebo; controlled trial	Plasma; ox-LDL; @ 0 and 4 months	None indicated	Low-sensitivity nonspecific LPO quantification methods; broad age range
Frankel et al.²⁷	Hypertriglycemic men and women; n=9	375 mg EPA+154 mg DHA/day; 12 weeks; no placebo; nutritional intervention	Blood; propanal, hexenal, and ox-LDL; @ 6 and 12 weeks	16.75 mg α-tocopherol/day	Low-sensitivity nonspecific LPO quantification methods
Bonanome et al.²⁸	Hypertriacylglycemic hemodialyzed men and women (mean=64±2); n=12	2.5 grams EPA+DHA/day; 2 months, no control; two-phase sequential intervention, supplementation followed by washout	Plasma; ox-LDL; @ 0, 1, and 2 months	0.9 mg DL-α-tocopherol/day	Low-sensitivity nonspecific LPO quantification methods
Brude et al.²⁹	Male smokers (40–60); n=9–11/group	Groups: (1) 5 grams EPA+DHA/day, (2) n-3+antioxidants, (3) antioxidants, (4) control oil; no EPA or DHA; 6 weeks; 8 grams oil; randomized, double-blind, placebo-controlled trial	Plasma; LOOH, ox-LDL, conjugated diene formation; @ 0 and 6 weeks	75 mg vitamin E/day, 150 mg vitamin C/day, 15 mg β-carotene/day, and 30 mg coenzyme Q10/day	Low-sensitivity nonspecific LPO quantification methods; broad age range
Sorensen et al.³⁰	Healthy men (29–60), n=21–24/group	0.91 gram EPA, DPA, and DHA/day; 7 week, 4 grams sunflower oil, two-phase randomized design	Plasma; ox-LDL; @ 3 weeks (after run in period) and 7 weeks	FO margarine contained propylgallate (20%), citric acid monoglyceride ester (80%), and 0.1 gram EDTA/kg	All subjects healthy and nonsmokers except 8 active smokers; broad age range; low-sensitivity nonspecific LPO quantification methods
Higdon et al.³¹	Postmenopausal women (50–75); n (treatment)=15	2.0 grams EPA+1.4 grams DHA/day; 5 weeks; 15 grams/day sunflower or safflower oil; three periods, three treatments, double-blinded crossover trial	Plasma; F2-IsoP and MDA; @ 0 and 5 weeks; blood; LOOH, ox-LDL, total antioxidant activity; @ 0 and 5 weeks	20 mg α-tocopherol/day	Plasma F2-IsoP decreased, differences eliminated after adjusting for AA concentrations; broad age range
Shoji et al.³²	Healthy pregnant women (18–40); n=23/ group	500 mg DHA and 150 mg EPA/day; week 20 gestation to delivery; negligible amounts of DHA and EPA; double-blind randomized control trial	Plasma and urine; MDA, 8-hydroxy-2-deoxyguanosine, α-tocopherol; @ 20, 30 weeks and @ delivery	3.0 mg α-tocopherol/day	Low-sensitivity nonspecific LPO quantification methods; broad age range
Wu et al.³³	Postmenopausal women vegetarians (<60 years); n=11–14/group	2.14 grams DHA/day, 42 days, 6 grams corn oil; single-blind, randomized, placebo-controlled trial	Plasma and urine; F2- isoP and α-tocopherol; @ 0 and 42 days	4 mg α-tocopherol acetate/day	NC in markers of oxidative stress
Hanwell et al.³⁴	Hypertriglycemic overweight and obese men (>45); n=10	2.8 grams EPA and 1.4 grams DHA with small amounts of LA, OA, PA, and ALA; acute admininstration; 7 grams corn oil; randomized double-blind, placebo-controlled crossover	Serum; LOOH, ox-LDL, total antioxidant status; @ 2, 4, and 6 hr postprandially	None indicated	Lack of effect may be due to acute administration of n-3 PUFA; low-sensitivity nonspecific LPO quantification methods
Ottestad et al.⁶³	Healthy men and women 18–50; n=58	8 grams/day oxidized or nonoxidized fish oil (1.6 grams EPA/DHA) or 8 grams/day high-oleic sunflower oil; randomized double-blind, placebo-controlled	Urinary 8-iso PGF2α2, plasma 4-hydroxy-2-hexenal, 4-hytroxy-2-nonenal @ 3 and 7 weeks	Tocopherols measured and standardized within all diets	Oxidized or nonoxidized fish oil did not altered markers of oxidative stress. Whether 4HNE is a reliable marker for in vivo LPO needs to be confirmed

n-3 PUFA, n-3 polyunsaturated fatty acids; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; ox-LDL, oxidized low-density lipoprotein; LPO, lipid peroxidation; LOOH, lipid hydroperoxides; FO, fish oil; F2-IsoPs, F2-isoprostanes; MDA, malondialdehyde; AA, arachidonic acid; NC, no change; LA, linoleic acid; OA, oleic acid; PA, palmitic acid; ALA, α-linolenic acid; PGF2α2, prostaglandin F2 alpha 2; 4HNE, 4-hydroxynonenal.

Because n-3 PUFA are susceptible to oxidation prior to ingestion, a number of different antioxidants were included. Four studies used α-tocopherol ranging from 3.0 to 20 mg/day.^27,31
–33 One study incorporated 0.9 mg/day DL-α-tocopherol.²⁸ Mixtures of other antioxidants were used (Table 1). Two studies did not use any antioxidants.^26,34 The ages of men and women studied ranged from 18 to 75 years. Six studies involved one gender exclusively. Study participants included normolipidemic, obese, hypertriglyercidemic men and women, and postmenopausal and pregnant women.

The methods used for assessing lipid peroxidation varied among these studies: Total anti-oxidant status (two studies^31,34); oxidized LDL and lipid hydroperoxides (seven studies^26–,34); MDA (two studies^31,32); F2-IsoP (two studies^31,33); and 8-hydroxy-2-deoxyguanosine (one study³²). Those biomarkers were assayed from serum, plasma, and urine samples. Considering the wide range of subject characteristics, dose and duration of n-3 PUFA, antioxidants, and methods used, it is difficult to determine why n-3 PUFA failed to increase or decrease LPO in these studies. Other confounding variables such as anti-oxidant nutrients and health status of participants may have contributed to the results reported.

Decrease in LPO

Eight studies demonstrated a decrease in LPO biomarkers after intervention with n-3 PUFA (Table 2). The administered n-3 PUFA dosages varied from 0.2 to 4.8 grams/day.^23,24 Intervention time ranged from 6 weeks to 5 months.^20,21 Two of these studies administered both purified DHA and EPA to separate groups^20,25; one used only purified DHA.²⁴ Remaining studies provided a mixture of EPA and DHA at different ratios with no rationale for specific ratios selected. Various antioxidants were included. Three studies utilized 6.4 to 32 mg/day α-tocopherol^20,23,25; two also included an additional 3.6 mg/day of γ-tocopherol.^20,25 One study used 0.125 mg of synthetic DL-α-tocopherol and 0.125 mg of ascorbic palmitate per 200-mg capsule of DHA.²⁴ Three studies did not use antioxidants.^20
–22 The study participants were men and women who ranged from 18 to 75 years of age and were healthy, overweight, dyslipidemic, hypertensive type 2 diabetic, postmenopausal, or pregnant (Table 2).

Table 2.

Human Studies in Which n-3 PUFA Decreased Lipid Peroxidation

Authors	Participants (age); n	n-3 PUFA dose; intervention duration; placebo/comparison; study design	Sample type; biomarker; time of measurement	Antioxidant	Results	Comments
Mori et al.⁶⁴	Nonsmoking men and women with NIDDM (30–65); n=11–14	Treatment groups: (1) 3.6 grams EPA+DHA/day (1 fish meal/day), 2) n-3+moderate exercise, (3) low-fat diet, (4) low-fat diet+moderate exercise; 8 weeks; no placebo; two-way factorial intervention, parallel design	Urine; F2-IsoP @ baseline and at 8 weeks	None indicated	F2-IsoPs decreased by 20% in fish fed and fish fed+moderate exercise groups	Increased oxidative stress due to NIDDM; high sensitivity LPO detection methods may have contributed to the results
Mori et al.²⁰	Nonsmoking, treated hypertensive, type 2 diabetic men and postmenopausal women (40–75); n=16–17/group	4 grams/day of EPA or DHA; 6 weeks; 4 grams/day OO placebo; double-blind placebo-controlled	Urine; F2-IsoP @ baseline and at 6 weeks	EPA: 6.8 mg α-tocopherol/day, 3.6 mg γ-tocopherol/day; DHA: 6.4 mg α-tocopherol/day, 3.6 mg γ-tocopherol/day; OO: 4.0 mg α-tocopherol/day, 3.6 mg γ-tocopherol/day	F2-IsoP decreased 20% by DHA and 19% by EPA relative to OO	Increased oxidative stress due to HT and NIDDM; high-sensitivity LPO detection methods may have contributed to the results
Barden et al.²¹	Pregnant women (mean=31); n=40–43/group	1.1 grams EPA+2.2 grams DHA/day; week 20 gestation to term; 4 grams/day OO; double-blind placebo- controlled parallel intervention	Cord blood, plasma and urine F2-IsoP @ delivery and 1 week postpartum	None indicated	Both plasma and urinary F2-IsoP decreased by n-3 PUFA, but significant only for plasma	Potentially increased oxidative stress due to pregnancy; high-sensitivity LPO detection methods, may have contributed to the results
Nalsen et al.²²	Healthy men and women (30–65); n=39–40/group	Groups: (1) 2.4 grams/day EPA+DHA+SFA diet, (2) n-3+MUFA, (3) SFA+placebo, (4) MUFA+placebo; 3 months; 3.6 grams/day OO; randomized placebo controlled	Plasma; F2-IsoP @ baseline and 3 months	None indicated	Plasma F2-IsoP decreased significantly in n-3 group	Decrease detected despite healthy subjects; may be due to the high sensitivity of LPO detection methods
Cazzola et al.²³	Healthy young (18–42 years) and old (53–70) males; n=93 and 62	Treatment groups: (1) 1.35 grams EPA+0.27 gram DHA/day, (2) 2.7 grams EPA+0.54 gram DHA/day, (3) 4.05 EPA+0.81 DHA/day, (4) placebo; 12 weeks; (1) 6 grams/day, (2) 3 grams/day, (3) 0 grams/day, (4) 9 grams/day corn oil; successive randomized placebo-controlled intervention	Plasma; LOOH, ox-LDL, GSSH; @ baseline and 12 weeks	32 mg α-tocopherol/day	Plasma lipid hydroperoxides decreased by ∼10% in both young and older subjects in EPA group	Broad age range; low-sensitivity nonspecific LPO quantification methods
Parra et al.¹⁸	Healthy men and women (20–40 years); n=67–74/group	EPA+DHA/day intake: (1) Cod diet=227±29 mg/day; (2) salmon diet=1418±34 mg/day; (3) fish oil supplement=3003±128 mg/day; 8 weeks; daily EPA+DHA intake for SFA placebo: 5.6±0.2 mg/day; parallel randomized placebo-controlled, intervention	Plasma; MDA and total antioxidant capacity (colorimetric assay); @ baseline and 56 days	None indicated	Total plasma antioxidant capacity increased in all dietary groups except control, but increase was significant only in the cod-based diet. MDA significantly decreased only in cod based diet	Source of dietary n-3 PUFA impacted oxidative or antioxidative effects. This may be due to the interaction of components of diet other than n-3 PUFA
Guillot et al.²⁴	Healthy males (53–65); n=12	200, 400, 800, and 1600 mg/day DHA; 2 weeks each dose; no placebo; successive dose intervention	Urine; F2-IsoP @ baseline and end of each 2-week dose, and 8 weeks after end of supplement. Plasma; ox-LDL, MDA, 4HHE, and 4HNE @ baseline and at end of each 2-week dose, and 8 weeks after end of supplement	Each capsule contained 200 mg DHA and 0.125 mg DL-α-tocopherol and 0.125 mg ascorbic palmitate (1–8 capsules/day)	Urinary F2-IsoP significantly lower after 200 mg/day DHA, increased significantly in the 1600 mg/day group. MDA: Concentrations significantly decreased after 200,400, and 800, not 1600 mg DHA. 4HHE: No changes after 200 and 400 mg DHA, but increased with 800 and 1600 mg DHA 4HNE: No change. ox-LDL: Significant increased lag time to oxidation in 200- to 800-mg group	Effect observed despite healthy subjects, short duration and high concentrations of antioxidants. Because the dose was increased every 2 weeks, effects seen at higher doses may be due to the dose and increased duration. Study used high-sensitivity specific LPO quantification methods. Increase in HHE expected as longer supplementation should lead to inclusion of n-3 PUFA in membranes
Mas et al.²⁵	Study A: Overweight, dyslipidemic men (20–65 years); n=17–20/group. Study B: Hypertensive type 2 diabetic and postmenopausal women (40–75); n=16–18	Groups: (1) 4 grams/day EPA, (2) 4 grams/day DHA, (3) placebo; 6 week; 4 grams/day OO; both studies, the same three groups	Plasma; F2-IsoP @ baseline and at 6 weeks	EPA: 6.8 mg α-tocopherol/day, 3.6 mg γ-tocopherol/day DHA: 6.4 mg α-tocopherol/day, 3.6 mg γ-tocopherol/day	Study A: Plasma F2-IsoP decreased 24% by EPA and 14% by DHA relative to OO. Study B: Plasma F2-IsoP decreased 23% by DHA and 19% by EPA relative to OO	Dyslipidemia and HT may have compromised oxidative stress; LPO was decreased by n-3 PUFA and not by OO in both studies. Potentially increased oxidative stress prior to joining the study, high concentration of antioxidants, and high-sensitivity. LPO quantification method may have contributed to the results

n-3 PUFA, n-3 polyunsaturated fatty acids; NIDDM, non-insulin-dependent diabetes mellitus; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; F2-IsoP, F2- isoprostanes; LPO, lipid peroxidation; OO, olive oil; HT, hypertension; SFA, saturated fatty acid; MUFA, monounsaturated fatty acid; LOOH, lipid hydroperoxides; ox-LDL, oxidized low-density lipoprotein; GSSH, oxidized glutathione; MDA, malondialdehyde; HHE, 4-hydroxyhexenal; 4HNE, 4-hydroxynonenal.

Markers of LPO investigated were plasma or urine F2-IsoP^20
–22,24 in all studies except for two, which examined plasma oxidized LDL, lipid hydroperoxides, MDA, HHE, and HNE.^23,24 While other factors may have contributed to the observed decreases in markers of LPO in these studies, the most likely reason was the assay system used to quantify LPO products. Plasma or urinary F2-IsoP was monitored in six of eight studies. The concentrations of the antioxidant nutrients used overlapped among the studies showing no change or decrease in markers of LPO following supplements with n-3 PUFA. Generally their concentrations were higher in those showing decrease than those showing no change.

Increase in LPO

Five studies reported an increase of LPO biomarkers after administration of n-3 PUFA for 1–6 months (Table 3). The amount of n-3 PUFA administered ranged from 0.2 to 10 grams/day. All studies used a mixture of DHA and EPA in supplement form, except one, which compared supplements versus diets containing salmon and cod.¹⁸ All studies used α-tocopherol, ranging from 2 to 400 mg/day as an antioxidant.^13
–15,17 Thus, the increase in LPO in these studies could not be attributed to the lack of antioxidant nutrients because the concentrations of α-tocopherol supplemented were comparable or higher than used in studies that showed no change or a decrease in LPO.

Table 3.

Human Studies in Which n-3 PUFA Increased Lipid Peroxidation

Authors	Participants (Age); n	n-3 PUFA dose; intervention duration; placebo/comparison; study design	Sample type; biomarker; time of measurement	Antioxidant	Results	Comments
Meydani et al.¹³	Young females (22–35 years) and older females (51–71 years); n=14 (young) and 9 (older)	1680 mg EPA+720 mg DHA/day (600 mg other fatty acids); 3 months; no placebo; randomized intervention with before and after comparison	Plasma; MDA and LOOH @ baseline & 1, 2, and 3 months	4 mg α-tocopherol/day	Plasma MDA increased in both groups up to 2 months of supplementation then declined, but remained higher than baseline levels	Effect seemed to be biphasic
Harats et al.¹⁴	Male and female smokers and non-smokers; 3 study groups: mean age(s): Group A (smokers)=37.4, Group B (smokers)=29.1 and 35.2, Group C (nonsmokers)=41.0 and 38.8); Group A: n=5–6, Group B: n=3–4 Group C: n=6–8	Groups: A=FO, B=FO+antioxidants, C=FO and FO+antioxidants; Group A: 10 grams FO/day, Group B: same as A+400 mg/day α-tocopherol, Group C: 10 grams/day FO and 10 grams/day FO+400 mg/day α-tocopherol (FO supplement contained 16.9% EPA+11 % DHA); no placebo; 4 weeks; randomized controlled intervention	Plasma; MDA and LOOH; @ baseline and 4 weeks	Indicated in the n-3 PUFA dosage; groups were created based partially on antioxidants	TBARS; Study A: 50% increase; Study B: 38% increase; Study C: 33%. Individuals receiving α-tocopherol, no increase in plasma TBARS	Low-sensitivity nonspecific LPO quantification methods; intervention includes lifestyles characterized by elevated levels of oxidative stress
Suzukawa et al.¹⁵	Stable uncomplicated hypertensive adults treated with blood pressure–lowering medication (mean=60±10); n=10/group	1.642 grams EPA+1.210 grams DHA/day; 6 weeks; 3.4 grams corn oil/day; randomized double-blind crossover	Plasma; MDA and ox-LDL;@ baseline and at 6 weeks	2.0 mg α-tocopherol/mL FO and 2.2 mg α-tocopherol /mL CO	Lag time of LDL oxidation significantly reduced compared to baseline. TBARS production from LDL significantly increased	LPO significantly increased by n-3 PUFA despite increased intake of antioxidants; may be due to the health status of subjects
Palozza et al.¹⁶	Healthy men and women (25–46 years); n=10/group	Groups: (1) 1.4 grams EPA+1.1 grams DHA/day, (2) 2.7 grams EPA+2.4 grams DHA/day, (3) 4.1 grams EPA+3.6 grams DHA/day; 180 days; placebo, 7.2 grams blend of 25% saturated, 65% oleic and 10% LA; placebo-controlled parallel study	Plasma; LOOH and MDA @ baseline, 30 days, and 180 days	0.9 mg DL-α-tocopherol, 1.5 mg DL-α-tocopherol, 2.1 mg DL-α-tocopherol/day. No α-tocopherol included in placebo capsules	30 days slight, dose-dependent decrease in AAPH-induced lipid peroxidation in n-3 PUFA groups. At 180 days, n-3 PUFA treatment=dose-dependent increase in MDA	LPO increased with increased doses of n-3 PUFA; effect was biphasic with an initial decrease and a later increase in LPO
Egert et al.¹⁷	Healthy men and women (18–45 years); n=15–17/group	Groups: (1) 2.8 grams/day EPA±0.6 gram; (2) 2.9 grams/day DHA±0.5 gram; 5 weeks; 6.0 grams/day ALA±1.1 gram; parallel randomized controlled dietary intervention	Plasma; ox-LDL; @ baseline and 5 weeks	2.0 mg α-tocopherol/day	Both EPA and DHA decreased lag time for LDL peroxidation compared to ALA group; both also increased the amounts of conjugated dienes	Both EPA and DHA increased LPO compared to ALA
Filaire et al.⁶⁵	Male judo Competitors (22–23 years); n=8–10/group	0.6 EPA+0.4 DHA grams/day alone or with additional antioxidants; placebo-controlled parallel study	Plasma-conjugated dienes, MDA, LOOH; baseline and 6 weels, pre- and postexercise	30 mg vitamin E/day	n-3 FA increased oxidative stress; did not reduce exercise-induced oxidative stress	Antioxidants+n-3 FA decreased some LPO after judo training
Egert et al.⁶⁶	Healthy men and women (19–45 years); n=24–25/group	Maragine with 10% ALA (4.4 grams/day), EPA (2.2 grams/day), or DHA (2.3 grams/day) ethyl esters; randomized	Plasma and RBC LPO, MDA and antioxidant capacity; baseline and 6 weeks	Mixed tocopherols added	No change in antioxidant capacity; MDA increased in plasma but not RBCs	No placebo group; DHA superior to EPA with EPA+DHA tissue incorporation

n-3 PUFA, n-3 polyunsaturated fatty acids; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid; MDA, malondialdehyde; LOOH, lipid hydroperoxides; FO, fish oil; TBARS, thiobarbituric reactive species; LPO, lipid peroxidation; ox-LDL, oxidized low-density lipoprotein; CO, corn oil; AAPH, 2,2′-azobis-(2-amidinopropane); ALA, α-linolenic acid; FA, fatty acid; RBC, red blood cells.

The subjects studied were both men and women and ranged in age from 18 to 71 years. Only one study used a single gender with young and elderly women.¹³ Other studies used both genders with subjects who were cigarette smokers, hypertensive, or healthy. All studies showing an increase in LPO following n-3 PUFA administration utilized combinations of MDA, lipid hydroperoxides, and oxidation of LDL to assess LPO.^13

–18 Because all of these methods involve ex vivo sample preparation, oxidation, or both, it is possible that the increases in LPO were introduced ex vivo and did not reflect in vivo LPO.

Reasons for the discrepancies

Whereas all 22 studies examined the effects of n-3 PUFA on markers of lipid peroxidation, there were differences in individual characteristics of each study, and therefore, the results obtained. Some possible explanations center on the quantification methods of LPO, study subject characteristics, n-3 PUFA composition, dosages and duration, and the concentrations of other pro- or anti-inflammatory nutrients in diet.

Biomarkers and quantification methods

The most notable differences among all studies were the biomarkers and quantification methods used. Most studies demonstrating no change or an increase in LPO relied on nonspecific quantification methods of LPO, such as MDA, lipid hydroperoxides, and oxidized LDL. The use of thiobarbituric acid to assay MDA and lipid hydroperoxides presents specificity issues because some of these species can result from nonlipid peroxidation origins. Unwarranted increases in reactive aldehyde species alone are not the only concern. Chemical influences that perturb TBAR values can result in either a perceived increase or decrease in quantified MDA.

Due to the complex milieu of biological samples and large number of reactive species, the reliability of MDA as an indicator of lipid peroxidation is questionable. Some investigators have taken this into account when interpreting their results. It is difficult to exclude all reactive species that appear naturally in biological samples and could influence the results of MDA quantification. In addition, studies using HPLC to quantify MDA may also lack reliability due to the volatility of MDA, which can impede accurate detection. Of the studies surveyed in this review that used MDA as an LPO biomarker, only three^24,31,32 quantified MDA via HPLC, whereas the remainder used the TBARS method for analysis. HPLC methods were used in studies that found no change or a decrease in LPO. However, more studies that incorporate HPLC detection methods need to be published before a definitive conclusion can be made.

Other classical methods used to assess lipid peroxidation, such as LDL oxidation susceptibility, reduced GSH levels, and total antioxidant status, produced results that are questionable due to concerns regarding specificity and sensitivity. Each of those methods is better classified as a general indicator of oxidative stress and not necessarily as specific assessments of lipid peroxidation. Therefore, it is preferable to use biomarkers when investigators can be certain of the lipid substrate precursors. For example, AA containing phospholipids are common membrane components, and they can be modified via oxidative and enzymatic pathways. F2-IsoP is specifically generated from the free radical–mediated oxidative degradation of membrane-bound AA. Thus, F2-IsoP is a valuable biomarker because it only represents the oxidative degradation of a specific membrane component in vivo. Of the eight studies incorporating F2-IsoPs as LPO biomarkers after n-3 PUFA administration, six reported a decrease and two reported no change. This high level of consistency among the studies was unique. All other biomarkers produced variable results, including no change, an increase, or a decrease. Besides of their high level of consistency, F2-IsoP measurements were much more sensitive than classical methods of LPO detection.

Study participants

The age and health status of subjects studied varied greatly. The age could have influenced results because older subjects are expected to have reduced antioxidant capacity. Also, several studies investigated subjects with either lifestyles or disease states characterized by elevated levels of oxidative stress. These studies included cigarette smokers, a lifestyle known to increase oxidative stress and directly promote the oxidation of LDL. In terms of the health status, studies incorporated individuals that were hypertensive, type 2 diabetic, pregnant, postmenopausal, hypertriglyceridemic, or obese; some were also healthy. Thus, the extensive subject variety and respective levels of oxidative stress could have contributed to the discrepancy in the results.

Administered dosages

Large variations in the concentration of n-3 PUFA were used; their specific fatty acid composition and the duration of the supplements may have also influenced results. All studies included in this review used less than 5 grams of EPA+DHA per day. Although one study¹⁴ used 10 grams/day fish oil, the EPA (16.9%) and DHA (11%) amounts were less than 5 grams/day. Most studies used fish oil supplements containing both EPA and DHA with varying amounts and ratios of each fatty acid. One study used purified samples of DHA or EPA. A few studies administered n-3 PUFA as part of the diet, mainly through fish or enriched foodstuffs such as margarine. The composition and amounts of fatty acids may vary extensively among different types of fish oil. Moreover, the major n-3 fatty acids have different protective effects. To compare results among studies with extensive variation of fatty acid sources, we reported the EPA and DHA concentrations. One study tested DHA doses of 200, 400, 800, and 1600 mg/day, and different methods to determine markers of LPO.²⁴ Urinary F2-IsoP concentrations decreased with 200 mg/day DHA but increased with 1600 mg/day, whereas plasma MDA decreased with 200–800 mg/day. Thus, it appears that the overall fatty acid and antioxidant composition of n-3 PUFA supplements and diets may have influenced the generation of markers of LPO.

Duration of supplementation

There was also considerable variation in the duration of supplementation of n-3 PUFA. One study used single administration, whereas others spanned 5 months. An acute administration of EPA and DHA did not alter total antioxidant status, oxidized LDL, or lipid hydroperoxides.³⁴ Alternatively, the apparent rationale for using much longer periods of n-3 PUFA supplementation in many of the other studies was to allow time for the incorporation of n-3 PUFA into membrane phospholipids. One study incorporated a stepwise increased concentration of n-3 PUFA every 2 weeks.²⁴ Thus, differences in a number of factors probably contributed to the discrepancies in results among different studies.

Mechanisms by which n-3 may decrease lipid peroxidation

n-3 PUFA can serve as antioxidants either as scavengers for reactive oxygen species (ROS), or by modulating the metabolism of n-6 PUFA. The latter would be through the expression of a variety of enzymes that regulate a balance between the oxidative and antioxidative forces. Superoxide scavenging capacity depends on the number of double bonds in a fatty acid.³⁶ Because the most common n-3 PUFA (ALA, EPA, DHA) have more double bonds than the most common n-6 PUFA [linoleic acid (LA) and AA], the n-3 PUFA would have a greater scavenging capacity than n-6 PUFA of the same chain length. In addition to the number of double bonds, the position of double bonds seems to be important in determining peroxidation of PUFA. Thus, in HT-29 human colorectal tumor cells cultured with different 22C fatty acids and less than 2% oxygen, delta 7, 10, 13, and 16 were the most sensitive to peroxidation, whereas delta 4 and 19 were involved in assimilation of those 22C fatty acids into phospholipids.⁵¹ Thus, both n-6 and n-3 PUFA can act as scavenger molecules, but n-3 seems to be more effective than n-6 PUFA.

EPA and DHA are incorporated into cellular membranes by replacing AA. This fatty acid replacement in the phoshopholipids can occur during cellular maturation and also by direct plasma exchange via transfer of serum albumin-associated DHA and EPA containing lysophosphatidylcholine.⁵² This decreases the amount of membrane AA, which serves as a precursor for the F2-IsoPs. Thus, in a human study, supplementing diets with DHA (3 grams/day for 90 days) decreased the red blood cell-AA content from 14.7% to 10.3%, and increased DHA from 2.4% to 8.1% and EPA from 0.44% to 0.96%.⁵³ Such replacement of AA with n-3 PUFA decreased the plasma F2-IsoPs, but this may simply be due to the decreased availability of AA and the percent of oxidation of AA may not have decreased. Most of the published data are plasma or urinary isoprostanes that are expressed on the basis of milligrams of lipids or milliliters of plasma or urine, not on the basis of AA concentration. To have a better understanding of the LPO, results should be expressed on the basis of AA; LPO products of EPA and DHA should also be monitored.^54,55 In addition to limiting the incorporation of AA into cell membranes, DHA also inhibits phospholipase A2 (PLA2), which releases free AA to be used by the cyclooxygenase, lipoxygenase, and P450 enzymes. Furthermore, EPA competes with AA for these enzymes.⁵² The oxygenation products of AA are much more proinflammatory than those produced from EPA and DHA. Thus, a decrease in inflammation may be another mechanism by which n-3 PUFA reduce LPO.

EPA and DHA also increased cellular concentrations of reduced glutathione, glutathione peroxidase, and superoxide dismutase and decreased that of catalase, which improved the antioxidant status.^56

–59 Studies conducted in mice and HepG2-8 cells indicated that DHA decreased the expression of NRF2 which increased expression of the above enzymes that in turn reduced oxidative stress.^60,61 Global gene arrays of the white blood cells from human subjects who supplemented their diets with DHA revealed that DHA supplementation decreased the expression of several genes found in white blood cells that are involved in lipid peroxidation, including cytochrome p450, oxidized LDL receptor 1, prostaglandin E synthase, cathepsin L, and several others.⁶² These findings support the conclusion that DHA decreased LPO. This is further supported by the decrease in plasma concentration of oxidation products of cholesterol and fatty acids caused by DHA. Collectively, these findings suggest that n-3 PUFA decrease lipid peroxidation; however, this remains controversial because of the effects of a number of confounding factors.

Conclusion and Future Research Directions

On the basis of the relevant literature, the main cause for variations in assessments of lipid peroxidation was the methods used. Studies incorporating nonspecific methods that measured MDA, lipid hydroperoxides, or oxidized (ox)-LDL resulted in extensive variation, including no change as well as increased or decreased LPO. Most investigators who used methods specific for in vivo LPO products, such as F2-IsoP, concluded that n-3 PUFA supplementation decreased LPO. On the basis of the results with F2-IsoP, n-3 PUFA supplementation did not increase LPO and could be considered as safe for human consumption. It is possible that doses greater than 5 grams/day may increase LPO; thus, it is hard to justify doses of EPA+DHA higher than 5 grams/day at this time. While appearing safe, additional work must still be conducted to definitively demonstrate that n-3 PUFA supplementation does not increase LPO. For those studies, specific improvements in the assessment of LPO must be incorporated. For example, instead of nonspecific methods, investigators should employ a panel of substrate-specific, high-sensitivity tests that determine the in vivo concentrations of LPO products. Similarly, tests for other metabolites produced should be specific for the oxidative degradation products from membrane components. Examples of metabolites that should be considered in future studies include HODE, produced from LA, HETE from AA, as well as α and β LPO products from cholesterol. Those metabolites, as part of a comprehensive panel including biomarkers such as F2-IsoP, HNE, and HHE, should provide a more thorough and specific assessment of in vivo lipid peroxidation.

The basal diet is likely to play a critical role in the production of LPO products before and after n-3 PUFA supplementation. Several nutrients, including minerals (such as Cu, Zn, Se, and Fe), vitamins (like E, C, and A), amino acids, as well as fatty acids, can all affect the oxidative status in the cells and the resulting LPO. Therefore, to compare results among different studies, it is important that standardized diets be used with a defined fatty acid and antioxidant composition. It is debatable whether DHA is more protective than EPA, therefore studies should be done with the two fatty acids individually as well as mixtures to determine their additive, synergistic, or antagonistic effects. Dose responses, which have generally not been done, are necessary to determine the efficacy, adverse effects, or the duration of the alterations observed. Study subject heterogeneity has made results difficult to interpret. Mixing subjects with a broad range of ages and health status adds a confounding variable and may skew the results for individual groups.

Footnotes

Author Disclosure Statement

No conflicting financial interests exist.

N.S.K. and K.L.E. designed the project; N.S.K. collected the data; N.S.K., Y.Y., and K.L.E. analyzed the data; N.S.K., Y.Y., and K.L.E. wrote the paper; K.L.E. had primary responsibility for final content. All authors read and approved the final manuscript

References

Harris

, Von Schacky

. The Omega-3 Index: A new risk factor for death from coronary heart disease?. Prev Med, 2004; 39:212–220.

Calder

. n-3 Fatty acids and cardiovascular disease: evidence explained and mechanisms explored. Clin Sci (Lond), 2004; 107:1–11.

Breslow

. n-3 fatty acids and cardiovascular disease. Am J Clin Nutr, 2006; 83:1477S–1482S.

Yazu

, Yamamoto

, Ukegawa

, et al. Mechanism of lower oxidizability of eicosapentaenoate than linoleate in aqueous micelles. Lipids, 1996; 31:337–340.

Yoshida

, Niki

. Bio-markers of lipid peroxidation in vivo: hydroxyoctadecadienoic acid and hydroxycholesterol. Biofactors, 2006; 27:195–202.

Kaikkonen

, Tuomainen

, Nyyssonen

, et al. C18 hydroxy fatty acids as markers of lipid peroxidation ex vivo and in vivo. Scand J Clin Lab Invest, 2004; 64:457–468.

Spiteller

. The relation of lipid peroxidation processes with atherogenesis: A new theory on atherogenesis. Mol Nutr Food Res, 2005; 49:999–1013.

Spiteller

. The important role of lipid peroxidation processes in aging and age dependent diseases. Mol Biotechnol, 2007; 37:5–12.

Yoshida

, Hayakawa

, Habuchi

, et al. Evaluation of lipophilic antioxidant efficacy in vivo by the biomarkers hydroxyoctadecadienoic acid and isoprostane. Lipids, 2007; 42:463–472.

10.

Leonarduzzi

, Arkan

, Basaga

, et al. Lipid oxidation products in cell signaling. Free Radic Biol Med, 2000; 28:1370–1378.

11.

Niki

, Yoshida

, Saito

, et al. Lipid peroxidation: mechanisms, inhibition, and biological effects. Biochem Biophys Res Commun, 2005; 338:668–676.

12.

Tang

, La

, Kern

, et al. Fatty acid oxidation and signaling in apoptosis. Biol Chem, 2002; 383:425–442.

13.

Meydani

, Natiello

, Goldin

, et al. Effect of long-term fish oil supplementation on vitamin E status and lipid peroxidation in women. J Nutr, 1991; 121:484–491.

14.

Harats

, Dabach

, Hollander

, et al. Fish oil ingestion in smokers and nonsmokers enhances peroxidation of plasma lipoproteins. Atherosclerosis, 1991; 90:127–139.

15.

Suzukawa

, Abbey

, Howe

, et al. Effects of fish oil fatty acids on low density lipoprotein size, oxidizability, and uptake by macrophages. J Lipid Res, 1995; 36:47–484.

16.

Palozza

, Sgarlata

, Luberto

, et al. n-3 fatty acids induce oxidative modifications in human erythrocytes depending on dose and duration of dietary supplementation. Am J Clin Nutr, 1996; 64:297–304.

17.

Egert

, Somoza

, Kannenberg

, et al. Influence of three rapeseed oil-rich diets, fortified with alpha-linolenic acid, eicosapentaenoic acid or docosahexaenoic acid on the composition and oxidizability of low-density lipoproteins: Results of a controlled study in healthy volunteers. Eur J Clin Nutr, 2007; 61:314–325.

18.

Parra

, Bandarra

, Kiely

, et al. Impact of fish intake on oxidative stress when included into a moderate energy-restricted program to treat obesity. Eur J Nutr, 2007; 46:460–467.

19.

Mori

, Dunstan

, Burke

, et al. Effect of dietary fish and exercise training on urinary F2-isoprostane excretion in non-insulin-dependent diabetic patients. Metabolism, 1999; 48:1402–1408.

20.

Mori

, Woodman

, Burke

, et al. Effect of eicosapentaenoic acid and docosahexaenoic acid on oxidative stress and inflammatory markers in treated-hypertensive type 2 diabetic subjects. Free Radic Biol Med, 2003; 35:772–781.

21.

Barden

, Mori

, Dunstan

, et al. Fish oil supplementation in pregnancy lowers F2-isoprostanes in neonates at high risk of atopy. Free Radic Res, 2004; 38:233–239.

22.

Nalsen

, Vessby

, Berglund

, et al. Dietary (n-3) fatty acids reduce plasma F2-isoprostanes but not prostaglandin F2alpha in healthy humans. J Nutr, 2006:136:1222–1228.

23.

Cazzola

, Russo-Volpe

, Miles

, et al. Age- and dose-dependent effects of an eicosapentaenoic acid-rich oil on cardiovascular risk factors in healthy male subjects. Atherosclerosis 200; 193:159–167.

24.

Guillot

, Caillet

, Laville

, et al. Increasing intakes of the long-chain omega-3 docosahexaenoic acid: Effects on platelet functions and redox status in healthy men. FASEB J, 2009; 23:2909–2916.

25.

Mas

, Woodman

, Burke

, et al. The omega-3 fatty acids EPA and DHA decrease plasma F(2)-isoprostanes: Results from two placebo-controlled interventions. Free Radic Res, 2010; 44:983–990.

26.

Nenseter

, Rustan

, Lund-Katz

, et al. Effect of dietary supplementation with n-3 polyunsaturated fatty acids on physical properties and metabolism of low density lipoprotein in humans. Arterioscler Thromb, 1992; 12:369–379.

27.

Frankel

, Parks

, Xu

, et al. Effect of n-3 fatty acid-rich fish oil supplementation on the oxidation of low density lipoproteins. Lipids, 1994; 29:233–236.

28.

Bonanome

, Biasia

, De Luca

, et al. n-3 fatty acids do not enhance LDL susceptibility to oxidation in hypertriacylglycerolemic hemodialyzed subjects. Am J Clin Nutr, 1996; 63:261–266.

29.

Brude

, Drevon

, Hjermann

, et al. Peroxidation of LDL from combined-hyperlipidemic male smokers supplied with omega-3 fatty acids and antioxidants. Arterioscler Thromb Vasc Biol, 1997; 17:2576–2588.

30.

Sorensen

, Marckmann

, Hoy

, et al. Effect of fish-oil-enriched margarine on plasma lipids, low-density-lipoprotein particle composition, size, and susceptibility to oxidation. Am J Clin Nutr, 1998; 68:235–241.

31.

Higdon

, Liu

, Du

, et al. Supplementation of postmenopausal women with fish oil rich in eicosapentaenoic acid and docosahexaenoic acid is not associated with greater in vivo lipid peroxidation compared with oils rich in oleate and linoleate as assessed by plasma malondialdehyde and F(2)-isoprostanes. Am J Clin Nutr, 2000; 72:714–722.

32.

Shoji

, Franke

, Campoy

, et al. Effect of docosahexaenoic acid and eicosapentaenoic acid supplementation on oxidative stress levels during pregnancy. Free Radic Res, 2006; 40:379–384.

33.

, Lu

, Wang

, et al. Effects of docosahexaenoic acid supplementation on blood lipids, estrogen metabolism, and in vivo oxidative stress in postmenopausal vegetarian women. Eur J Clin Nutr, 2006; 60:386–392.

34.

Hanwell

, Kay

, Lampe

, et al. Acute fish oil and soy isoflavone supplementation increase postprandial serum (n-3) polyunsaturated fatty acids and isoflavones but do not affect triacylglycerols or biomarkers of oxidative stress in overweight and obese hypertriglyceridemic men. J Nutr, 2009; 139:1128–1134.

35.

Bechoua

, Dubois

, Dominguez

, et al. Protective effect of docosahexaenoic acid against hydrogen peroxide-induced oxidative stress in human lymphocytes. Biochem Pharmacol, 1999; 57:1021–1030.

36.

Richard

, Kefi

, Barbe

, et al. Polyunsaturated fatty acids as antioxidants. Pharmacol Res, 2008; 57:451–455.

37.

Brand

, Schonfeld

, Isharel

, et al. Docosahexaenoic acid-dependent iron accumulation in oligodendroglia cells protects from hydrogen peroxide-induced damage. J Neurochem, 2008; 105:1325–1335.

38.

Shimazawa

, Nakajima

, Mashima

, et al. Docosahexaenoic acid (DHA) has neuroprotective effects against oxidative stress in retinal ganglion cells. Brain Res, 2009; 1251:269–275.

39.

Shoji

, Franke

, Demmelmair

, et al. Effect of docosahexaenoic acid on oxidative stress in placental trophoblast cells. Early Hum Dev, 2009; 85:433–437.

40.

Halliwell

, Lee

. Using isoprostanes as biomarkers of oxidative stress: some rarely considered issues. Antioxid Redox Signal, 2010; 13:145–156.

41.

Spickett

, Wiswedel

, Siems

, et al. Advances in methods for the determination of biologically relevant lipid peroxidation products. Free Radic Res, 2010; 44:1172–1202.

42.

Zarkovic

. 4-Hydroxynonenal as a bioactive marker of pathophysiological processes. Mol Aspects Med, 2003; 24:281–291.

43.

Long

, Picklo

Sr . Trans-4-hydroxy-2-hexenal, a product of n-3 fatty acid peroxidation: Make some room HNE. Free Radic Biol Med, 2010; 49:1–8.

44.

Esterbauer

, Schaur

, Zollner

. Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic Biol Med, 1991; 11:81–128.

45.

Poli

, Schaur

. 4-Hydroxynonenal in the pathomechanisms of oxidative stress. IUBMB Life, 2000; 50:315–321.

46.

Zarkovic

, Zarkovic

, Schaur

, et al. 4-Hydroxynonenal as a second messenger of free radicals and growth modifying factor. Life Sci, 1999; 65:1901–1904.

47.

Borovic

, Rabuzin

, Waeg

, et al. Enzyme-linked immunosorbent assay for 4-hydroxynonenal-histidine conjugates. Free Radic Res, 2006; 40:809–820.

48.

Levine

, Garland

, Oliver

, et al. Determination of carbonyl content in oxidatively modified proteins. Methods Enzymol, 1990; 186:464–478.

49.

Uchida

, Stadtman

. Selective cleavage of thioether linkage in proteins modified with 4-hydroxynonenal. Proc Natl Acad Sci USA, 1992; 89:5611–5615.

50.

Uchida

, Toyokuni

, Nishikawa

, et al. Michael addition-type 4-hydroxy-2-nonenal adducts in modified low-density lipoproteins: Markers for atherosclerosis. Biochemistry, 1994; 33:12487–12494.

51.

Sato

, Sato

, Kawamoto

, et al. Differential roles of internal and terminal double bonds in docosahexaenoic acid: Comparative study of cytotoxicity of polyunsaturated fatty acids to HT-29 human colorectal tumor cell line. Prostaglandins Leukot Essent Fatty Acids, 2011; 84:31–37.

52.

Adkins

, Kelley

. Mechanisms underlying the cardioprotective effects of omega-3 polyunsaturated fatty acids. J Nutr Biochem, 2010; 21:781–792.

53.

Kelley

, Siegel

, Fedor

, et al. DHA supplementation decreases Serum C-reactive protein and other markers of inflammation in hypertriglyceridemic men. J Nutr, 2009; 139:495–501.

54.

Niki

. Biomarkers of lipid peroxidation in clinical material. Biochim Biophys Acta, 2014; 1840:809–817.

55.

Giordano

, Visioli

. Long-chain omega 3 fatty acids: molecular bases of potential antioxidant actions. Prostaglandins Leukot Essent Fatty Acids, 2014; 90:1–4.

56.

de Assis

, Rech

, Longoni

, et al. Omega3-Polyunsaturated fatty acids prevent lipoperoxidation, modulate antioxidant enzymes, and reduce lipid content but do not alter glycogen metabolism in the livers of diabetic rats fed on a high fat thermolyzed diet. Mol Cell Biochem, 2012; 361:151–160.

57.

Hajjaji

, Besson

, Bougnoux

. Tumor and non-tumor tissues differential oxidative stress response to supplemental DHA and chemotherapy in rats. Cancer Chemother Pharmacol, 2012; 70:17–23.

58.

Avramovic

, Dragutinovic

, Krstic

, et al. The effects of omega 3 fatty acid supplementation on brain tissue oxidative status in aged wistar rats. Hippokratia, 2012; 16:241–245.

59.

Lluis

, Taltavull

, Munoz-Cortes

, et al. Protective effect of the omega-3 polyunsaturated fatty acids: Eicosapentaenoic acid/docosahexaenoic acid 1:1 ratio on cardiovascular disease risk markers in rats. Lipids Health Dis, 2013; 12:140.

60.

Depner

, Philbrick

, Jump

. Docosahexaenoic acid attenuates hepatic inflammation, oxidative stress, and fibrosis without decreasing hepatosteatosis in a Ldlr−/− mouse model of western diet-induced nonalcoholic steatohepatitis. J Nutr, 2013; 143:315–323.

61.

Saw

, Yang

, Guo

, et al. Astaxanthin and omega-3 fatty acids individually and in combination protect against oxidative stress via the Nrf2-ARE pathway. Food Chem Toxicol, 2013; 62:869–875.

62.

Dawson

, Zhao

, Adkins

, et al. Modulation of blood cell gene expression by DHA supplementation in hypertriglyceridemic men. J Nutr Biochem, 2012; 23:616–621.

63.

Ottestad

, Vogt

, Retterstol

, et al. Oxidised fish oil does not influence established markers of oxidative stress in healthy human subjects: a randomised controlled trial. Br J Nutr., 2012; 108:315–326.

64.

Mori

, Bao

, Burke

, et al. Dietary fish as a major component of a weight-loss diet: Effect on serum lipids, glucose, and insulin metabolism in overweight hypertensive subjects. Am J Clin Nutr, 1999; 70:817–825.

65.

Filaire

, Massart

, Rouveix

, et al. Effects of 6 weeks of n-3 fatty acids and antioxidant mixture on lipid peroxidation at rest and postexercise. Eur J Appl Physiol, 2011; 111:1829–1839.

66.

Egert

, Lindenmeier

, Harnack

, et al. Margarines fortified with alpha-linolenic acid, eicosapentaenoic acid, or docosahexaenoic acid alter the fatty acid composition of erythrocytes but do not affect the antioxidant status of healthy adults. J Nutr, 2012; 142:1638–1644.