Effects of eicosapentaenoic acid and docosahexaenoic acid on low-density lipoprotein cholesterol and other lipids: A review
Article Outline
- Abstract
- Methods
- Results
- Discussion
- Conclusions
- Acknowledgments
- Financial disclosures
- Supplementary data
- References
- Copyright
Abstract
In this exploratory, hypothesis-generating literature review, we evaluated potentially differential effects of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) on low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), and non-HDL-C in published studies of ω-3 fatty acid supplementation or prescription ω-3 fatty acid ethyl esters. Placebo-adjusted changes in mean lipid parameters were compared in randomized, controlled trials in subjects treated for ≥4 weeks with DHA or EPA. Of 22 studies identified, 6 compared DHA with EPA directly, 12 studied DHA alone (including 14 DHA–treated groups), and 4 examined EPA alone. In studies directly comparing EPA with DHA, a net increase in LDL-C of 3.3% was observed with DHA (DHA: +2.6%; EPA: −0.7%). In such head-to-head comparative studies, DHA treatment was associated with a net decrease in TG by 6.8% (DHA: −22.4%; EPA: −15.6%); a net increase in non-HDL-C by 1.7% (DHA: −1.2%; EPA −2.9%); and a net increase in HDL-C by 5.9% (DHA: +7.3%; EPA: +1.4%). Increases in LDL-C were also observed in 71% of DHA-alone groups [with demonstrated statistical significance (P < .05) in 67% (8 of 12) DHA-alone studies] but not in any EPA-alone studies. Changes in LDL-C significantly correlated with baseline TG for DHA-treated groups. The range of HDL-C increases documented in DHA-alone vs EPA-alone studies further supports the fact that HDL-C is increased more substantially by DHA than EPA. In total, these findings suggest that DHA-containing supplements or therapies were associated with more significant increases in LDL-C and HDL-C than were EPA-containing supplements or therapies. Future prospective, randomized trials are warranted to confirm these preliminary findings, determine the potential effects of these fatty acids on other clinical outcomes, and evaluate the generalizability of the data to larger and more heterogeneous patient populations.
Keywords: Apo B, Cardiovascular disease, DHA, EPA, LDL-C, Omega-3 fatty acids
Approximately 33% of U.S. adults have increased levels of serum triglyceride (TG) (≥ 150 mg/dL), which represent an independent risk factor for coronary heart disease (CHD).1 Long-chain ω-3 fatty acids (OM3s)—as either fish oil supplements or prescription ethyl esters (P-OM3s)—are an accepted therapy to lower TG levels.2 These substances, which include eicosapentaenoic acid (EPA; 20:5 [n-3]) and docosahexaenoic acid (DHA; 22:6 [n-3]), are believed to reduce TG levels primarily by promoting fatty-acid degradation (via peroxisomal β-oxidation), inhibiting lipogenesis in the liver, and accelerating clearance of TG from the plasma.3
Beyond their TG-lowering effects, OM3s appear to further lessen the risk of CHD via antithrombotic mechanisms, reducing susceptibility to cardiac arrhythmias, inhibiting atherosclerotic plaque formation (by reducing adhesion molecule expression, platelet-derived growth factor, and inflammation), and decreasing blood pressure by promoting nitric oxide–elicited endothelial relaxation.2 In major randomized controlled trials (RCTs),4, 5 treatment with 1- to 2-g doses of P-OM3 was associated with significant cardioprotective benefits that could not be directly ascribed to changes in lipid levels, which were modest or equivocal across treatment groups. It is not clear whether administration of greater, lipid-altering doses of OM3s (ie, 3–4 g/d) will confer similar cardioprotective benefits.
P-OM3s include Omacor® (Pronova Biocare, Oslo, Norway; marketed by GlaxoSmithKline, Research Triangle Park, NC, in the United States as Lovaza®) and Epadel® in Japan (Mochida Pharmaceutical Co., Ltd., Tokyo). Lovaza and Omacor are highly concentrated fish-oil preparations in the form of OM3 ethyl esters containing approximately 47% EPA and 38% DHA. Epadel consists of >98% ethyl-EPA.2, 3 Lovaza is approved in the United States for adjunctive treatment of severe hypertriglyceridemia (TG ≥ 500 mg/dL) to reduce TG levels, along with dietary modification.6 Omacor is approved in the European Union for adjunctive treatment in the secondary prevention of myocardial infarction, in combination with other standard therapies (eg, statins, antiplatelet medicinal products, β-blockers, angiotensin-converting enzyme inhibitors) and for hypertriglyceridemia. Epadel is approved in Japan for the management of hyperlipidemia and arteriosclerosis obliterans.7
In two pivotal RCTs with Omacor in patients with TG ≥ 500 mg/dL,8, 9 investigators documented median increases in LDL-C of 44.5% in concert with similar-magnitude reductions in TG. Increases in LDL-C (1.9%–11.3%) were also observed in most (5 of 6) studies involving patients with median baseline TG levels between 250 and 300 mg/dL in a statistical review by the Food and Drug Administration. The Food and Drug Administration expressed concerns about such increases in LDL-C among study participants with baseline TG < 500 mg/dL.10 Increases in LDL-C were primarily observed in patients with greater baseline TG and were also associated with greater reductions in TG (Fig. 1).10
Figure 1
Median percent change in TG and LDL-C in Omacor/Lovaza studies pooled by baseline TG inclusion criteria (From CDER-FDA Statistical Review10). CDER, Center for Drug Evaluation and Research; C.I., confidence interval; FDA, Food and Drug Administration; LDL-C, low-density lipoprotein cholesterol; TG, triglyceride. (To convert TG to mmol/L, multiply by 0.0113.)
Further studies of combined P-OM3s and statin treatments in patients with TGs of 200–499 mg/dL in the Combination of prescription Omega-3 with Simvastatin (COMBOS) trial reported increases in median LDL-C of 3.5% greater than placebo at 8 weeks of treatment.11, 12 Increases of up to 10.2% persisted for up to 2 years in the open-label extension study.13 These effects are of some concern because many statin-treated patients with mixed or diabetic dyslipidemia have baseline TG levels between 200 and 500 mg/dL. In the COMBOS trial,11, 12, 13 subjects in the lowest tertile of baseline LDL-C (and/or highest baseline TG levels) had large percentage increases in LDL-C, whereas those in the greatest LDL-C tertile had reductions in LDL-C; however, end-point LDL-C levels were generally <100 mg/dL.
In part because of concerns about increases in LDL-C in patients with TG levels of 200–499 mg/dL, combined treatment with Lovaza and statins was not approved for use in populations with TG levels in this range.14 Ongoing clinical trials (eg, ClinicalTrials.gov identifiers NCT00360217, NCT01047501) are seeking to establish whether treatment with either DHA or EPA may confer superior effects on lipid profiles in patients without profoundly elevated TG. To compare the effects of EPA versus DHA on the lipid profile, we conducted a descriptive, exploratory, hypothesis-generating review of the published clinical literature.
Methods
Literature search and strategy
The primary objective of this review was to assess the impacts of EPA and DHA on the circulating level of LDL-C because this is the most validated and widely accepted lipid cardiovascular risk factor. LDL-C remains the central target of pharmacotherapy to prevent cardiovascular events, and multiple studies have demonstrated a significant, direct relationship between reductions in LDL-C and decreases in cardiovascular risk.15 By primarily elucidating the effects of different OM3s on LDL-C, our review was intended to inform clinical decision making concerning the optimal management of lipid and lipoprotein abnormalities in order to prevent cardiovascular events.
The secondary objective of our review was to explore the effects of DHA and EPA on other lipids and lipoproteins. These included TG, HDL, and non-HDL-C. Although these biomarkers are important in the prediction of CHD risk, no randomized controlled trials to date have specifically shown that modulating either HDL-C or TG, individually, reduces the risk of CHD.
Searches of MEDLINE (from January 1966 through September 2010) were conducted by including the term EPA or DHA (both abbreviated and as the expanded names) as title terms, along with lipid, cholesterol, or LDL (Medical Subject Heading [MeSH] key words), human, dietary supplement, or drug therapy (MeSH terms), and RCT (publication type). Relevant articles were also identified from the reference lists of these studies and other reviews and meta-analyses.16, 17, 18
Inclusion and exclusion criteria for study reports
Abstracts were reviewed to determine potential eligibility on the basis of the OM3 used in the study (EPA or DHA). By a priori consensus of the authors, eligible studies: 1) included a treatment period of ≥4 weeks; 2) documented OM3 purity (we imposed no purity exclusion criteria but were interested in documenting that each product tested was primarily or exclusively an EPA or DHA monotherapy); 3) had evidence that subjects were primarily or exclusively treated with DHA, EPA, or both (ie, minor product having <4% of total OM3 in product tested); 4) included a placebo or other control group; and 5) reported LDL-C, TG, and HDL-C values both at baseline and study end point for both the treatment and control groups. No preclinical studies; case reports or series; reviews; meta-analyses; editorials; or correspondence were included in our analysis. Observational and other uncontrolled studies were also excluded. RCTs in which participants received other lipid-altering drugs concomitant with OM3s were not excluded. In addition, no minimum OM3 dose was imposed for study inclusion.
Study reports reviewed
The initial literature search identified 494 papers, most of which were excluded after reviewing their abstracts and deeming them ineligible according to the criteria outlined previously. Of 101 full-text articles reviewed, 51 were eliminated because there was insufficient information about OM3 composition and purity (we excluded studies because of either insufficient information about the dose of EPA or DHA received or if we were unable to determine if the preparations were mixtures of EPA and DHA comprising >4% of the other OM3); 13 because lipid values (baseline and end-point) were not reported; 13 because studies did not include adequate control groups; and 2 because of insufficient study duration. Hence, 22 published studies were analyzed in the present review.
Data extraction and synthesis
The chief outcome measure was the placebo- (or control-) adjusted change in LDL-C (and other lipids/lipoproteins) from baseline to end point, by treatment group. These data were presented as Forest plots, including mean values and measures of dispersion (standard deviation [SD], standard error of the mean [SEM], or 95% confidence interval [CI]). Percent changes in mean values were calculated from the provided end-point and baseline values, and, where only one variable along with mean change was provided in the original paper, either the end-point or baseline value was calculated as noted. Any lipid or lipoprotein measures expressed in International System of Units (ie, mmol/L) were converted to mg/dL by dividing by 0.0259 (for cholesterol) and by 0.0113 (for TG), and SEM values were uniformly converted to SDs. Ninety-five percent CIs were estimated from the mean and SD of lipid levels from treated and control/placebo groups at baseline and end point, according to a method reported by Haney and coworkers (Supplemental Appendix).19 If not reported, non-HDL-C was calculated as the difference between total cholesterol and HDL-C.
Unless otherwise stated, all percent changes for treated groups were corrected by subtracting the placebo or control group values. All statistical tests were conducted at a 2-tailed α = 0.05. Pearson correlational analyses were conducted to compare doses, study durations, and/or baseline TG or LDL-C with mean placebo-adjusted percent changes in mean lipid parameters. In one study, where only oxidized LDL-C, HDL-C, TG, and total cholesterol measures were provided, LDL-C was estimated by use of the Friedewald equation (mean TG levels were <400 mg/dL).20 Because of methodological heterogeneity between and among studies—including differences in study designs, patient eligibility criteria, and doses and durations of OM3 treatments—we have limited statistical analyses to relatively few tests of correlations between changes in lipid variables relative to baseline TG or LDL-C levels or relative to end-point lipid values. Because in studies the authors did not consistently statistically compare lipid end points (5 of 22 did not report any statistical comparisons, 5 of 22 reported only end-point vs baseline comparisons, 10 of 22 reported treatment vs placebo comparisons, and 2 of 22 reported both), we were unable to make a summary statement about the statistical significance of changes at end point across studies.
Results
Study characteristics
Of the 22 controlled studies identified, 6 compared the effects on lipids of DHA vs EPA (head-to-head studies),21, 22, 23, 24, 25, 26 4 examined the effects of EPA alone,27, 28, 29, 30 and 12 evaluated the effects of DHA alone31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42 (2 of which examined DHA at 2 doses each). Study parameters and percent changes in mean LDL-C and other lipid end points associated with EPA or DHA treatment are summarized in Table 1.
Table 1. Randomized controlled trials reporting the effects of DHA, EPA, or EPA vs DHA on LDL-C, HDL-C, TG, and non-HDL-C
| Study | Patient population, total n∗ | Treatment, dose† (g), n∗Follow-up | BL to EP % change | Comments | |||
|---|---|---|---|---|---|---|---|
| LDL-C | HDL-C | TG | Non-HDL-C‡ | ||||
| EPA vs DHA | |||||||
 Egert et al,21 2009 | ♀♂; TC < 300 mg/dL; TG < 200 mg/dL; BMI < 28 kg/m2 (n = 74¶) | EPA 2.2 (n = 25) DHA 2.3 (n = 25) ALA 3.4 (n = 24) 6 weeks | EPA: −4.3 DHA: −0.8 ALA: +0.4 (NS DHA or EPA vs BL) | EPA: +3.6 DHA: +13.1 (P < .001) ALA: +2.5 | EPA: −15.1 (P < .001) DHA: −30.6 (P < .001) ALA: −16.8 (P < .05) | EPA: −6.6 DHA: −4.2 ALA: −1.2 | Pronova Biocare (EPA); Nichimen Europe (DHA); administered via margarines EPA: 17.7% SFA, DHA: 18.1% SFA ALA: 19.4% SFA TG (median and 25th, 75th percentiles) |
| Grimsgaard et al,22 1997 | ♂; TC < 367 mg/dL; TG < 443 mg/dL (n = 224¶) | EPA 4.0 (n = 75) DHA 4.0 (n = 72) PBO (n = 77) 7 weeks | EPA: −2.0 DHA: +1.7 PBO: +1.5 (NS DHA or EPA vs PBO) | EPA: +0.8 DHA: +4.4 (P < .001) PBO: −0.7 (P < .01 DHA vs PBO or EPA) | EPA: −12.2 (P < .01) DHA: −17.7 (P < .001) PBO: +9.0 (P < .01) (P < .0001 DHA or EPA vs PBO) | EPA: −3.4 DHA: −0.6 PBO: +2.4 | Pronova Biocare Corn oil PBO |
| Mori et al,23 2000 | ♂; TC > 232 mg/dL; TG > 159 mg/dL; BMI 25–30 kg/m2 (n = 56¶) | EPA 4 (n = 19) DHA 4 (n = 17) PBO 4 (n = 20) 6 weeks | EPA: +4.2 DHA: +8.7 (P = .019 vs PBO) PBO: −2.3 | EPA: −1.0 DHA: +9.4 PBO: −8.9 | EPA: −21.4 (P = .012) DHA: −33.3 (P = .003) PBO: −4.4 (NR) | EPA: −0.6 DHA: +1.3 PBO: −2.8 | Source not listed Olive oil PBO Exclusions from number randomized well documented. |
| Nestel et al,24 2002 | ♀♂; ♀ postmenopausal; dyslipidemia; TC > 212 mg/dL; TG > 177 mg/dL; HDL-C < 39 mg/dL (♂) or <46 mg/dL (♀) (n = 38) | EPA 3 (n = 12) DHA 3 (n = 12) PBO (n = 14) 7 weeks | EPA: −0.2 DHA: +3.5 PBO: +2.9 | EPA: +4.7 DHA: +9.9 PBO: +7.0 | EPA: −22.9 (P = .026) DHA: −31.8 (P = .026) PBO: +11.9 (P = .013 for group × time [ANOVA]) | EPA: −3.0 DHA: −1.5 PBO: +2.3 | F Hoffmann-La Roche; small amount of either EPA or DHA in each preparation and other OM3 (EPA + 3.1% 20:4n-3; DHA + 13.2% 22:5n-3) Olive oil PBO |
| Park and Harris,25 2003 | ♀♂; LDL-C < 160 mg/dL; HDL-C > 35 mg/dL; TG < 200 mg/dL; BMI 22–30 kg/m2 (n = 33) | EPA 4 (n = 11) DHA 4 (n = 11) PBO (n = 11) 4 weeks | EPA: +4.9 DHA: +5.2 PBO: +2.8 | EPA: 0 DHA: +4.0 PBO: 0 | EPA: −9.5 DHA: −8.8 PBO: −0.8 | EPA: +2.6 DHA: +1.0 PBO: +2.4 | Fish Oil Test Material Program, NIH Dept. of Commerce Safflower oil PBO |
| Woodman et al,26 2002 | ♀♂; T2D; FG > 126 mg/dL (NFG > 200 mg/dL); HbA1C < 9%; hypertensive; BMI < 35 kg/m2; TC and TG < 664 mg/dL (n = 51¶) | EPA 4 (n = 17) DHA 4 (n = 18) PBO (n = 16) 6 weeks | EPA: +0.8 DHA: +5.1 PBO: +2.2‖ (NS EPA or DHA vs PBO) | EPA: +0.8 DHA: +4.1 PBO: +0.9‖ (NS EPA or DHA vs PBO) | EPA: −17.2 DHA: −16.7 PBO: −3.4 (P < .05 EPA or DHA vs PBO) | EPA: −2.1 DHA: +0.6 PBO: +0.9 | Fish Oil Test Materials Program, NIH Dept. of Commerce Olive oil PBO 8 randomized subjects excluded for well-documented reasons. |
| EPA alone | |||||||
| Ando et al,27 1999 | ♀♂; dialysis patients (n = 38) | EPA 1.8 (n = 19) PBO (n = 19) 12 weeks | EPA: −8.8 PBO: +7.0 | EPA: −1.2 PBO: −1.2 | EPA: −42.1 (P < .01) PBO: −1.6 | EPA: −19.2 PBO: +4.1 | Epadel/Mochida Pharmaceutical. LDL-C estimated with Friedewald formula. SD back-calculated from SEM |
| Kurabayashi et al,28 2000 | ♀; menopausal; HL; TC 220–280 mg/dL; TG 150–398 mg/dL (n = 141) | EPA + estriol 1.8 (n = 69) CTRL + estriol (n = 72) 48 weeks | EPA: −8.0 (P = .02) CTRL: −8.3 (P = .001) | EPA: −1.7 CTRL: +1.5 | EPA: −16.2 (P = .009) CTRL: +8.6 (P = .003 EPA vs CTRL) | EPA: −8.1 CTRL: −5.5 | Epadel/Mochida Pharmaceutical. % change is as reported for mg/dL |
| Satoh et al,29 2007 | ♀♂; obese; T2D; TG ≥ 150 mg/dL; HDL-C < 39 mg/dL; hypertensive; FG ≥ 110 mg/dL (n = 44) | EPA 1.8 (n = 22) Diet (n = 22) 12 weeks | EPA: −7.6 (P = .004) Diet: −6.6 | EPA: −2.1 Diet: −5.5 | EPA: −18.7 (P = .047) Diet: −14.3 | EPA: −6.9 Diet: 0 | Source not listed. Significance obtained with two-tailed, paired t test; lost with repeated-measures ANOVA |
| Tanaka et al,30 2008 | ♀♂; ♀ postmenopausal; HC; TC ≥ 251 mg/dL; LDL-C ≥ 167 mg/dL (n = 18,645) | EPA + statin‖ 1.8 (n = 9326) CTRL (n = 9319) 5 years | EPA: −23.3 CTRL: −23.6 (NS between groups) | EPA: +3.3 CTRL: +4.5 (NS between groups) | EPA: −7.3 CTRL: −2.6 (P < .0001 between groups) | EPA: −23.1 CTRL: −22.3 | Mochida Pharmaceutical. JELIS. TG (median and 25th, 75th percentiles) |
| DHA alone | |||||||
| Agren et al,31 1996 | ♂; healthy (n = 28)∗∗ | DHA 1.68 (n = 14) CTRL (n = 14) 15 weeks | DHA: −2.8 (NS vs BL or CTRL) CTRL: −2.6 | DHA: +6.9 CTRL: −16.7 (P < .01 DHA vs CTRL) | DHA: −17.1 (P < .05 vs BL and CTRL) CTRL: +8.4 | DHA: +0.3 CTRL: +3.4 | Martek |
| Conquer and Holub,32 1996 | ♀♂; healthy, vegetarian (n = 24) | DHA 1.62 (n = 12) PBO (n = 12) 6 weeks | DHA: −6.5 PBO: +9.0 (P < .05 vs DHA and BL) | DHA: +16.7 (P < .05) PBO: +7.4 (P < .05) | DHA: −16.7 (P < .05) PBO: +13.6 (P < .05) | DHA: −8.2 PBO: +9.3 | DHASCO/Martek SD back-calculated from SEM |
| Conquer and Holub,33 1998 | ♀♂; healthy (n = 19) | DHA-L 0.75 (n = 6) DHA-H 1.5 (n = 7) PBO (n = 6) 6 weeks | DHA-L: −9.0 DHA-H: +3.4 PBO: −8.9 | DHA-L: +5.4 DHA-H: −4.5 PBO: +1.7 | DHA-L: −8.5 DHA-H: −5.8 PBO: +22.5 | DHA-L: +10.0 DHA-H: +1.0 PBO: −3.8 | DHASCO/Martek Unclear if study was randomized SD back-calculated from SEM |
| Davidson et al,34 1997 | ♀♂; combined HL (n = 26)∗∗∗ | DHA-L 1.25 (n = 9) DHA-H 2.5 (n = 9) PBO (n = 8) 6 weeks | DHA-L: +9.3 DHA-H: +13.6 (P < .001) PBO: −2.4 | DHA-L: +5.9 (P < .02) DHA-H: +6.2 (P < .03) PBO: +5.6 | DHA-L: −20.9 (P < .01) DHA-H: −17.6 (P < .01) PBO: +3.5 | DHA-L: +1.6 DHA-H: +5.7 (P < .04) PBO: +1.4 | DHASCO/Martek SD back-calculated from SEM |
| Engler et al,35 2004 | ♀♂; children (9–19 yrs); FH (LDL-C > 131 mg/dL); FCH (LDL-C > 131 mg/dL and/or TG >150 mg/dL); crossover (n = 20) | DHA 1.2 + diet (n = 20) Diet (n = 20) PBO + Diet (n = 20) 6 weeks | DHA + diet: +7.4 Diet: −5.6 PBO: +1.3 (P < .001 DHA vs diet) | DHA + diet: +6.5 Diet: −8.1 PBO: +8.9 (P < .001 DHA vs diet) | DHA + diet: −14.7 Diet: −5.1 PBO: −5.7 | DHA + diet: +5.1 Diet: −6.4 PBO: 0 | DHASCO/Martek Diet = National Cholesterol Education Program Step II (NCEP II) & food guide pyramid dietary guidelines. |
| Geppert et al,36 2006 | ♀♂; healthy, vegetarian (n = 114) | DHA 0.94 (n = 53)¶ PBO (n = 53)¶ 8 weeks | DHA: +10.6 (P < .001) PBO: −0.8 (P = .003 DHA vs PBO) | DHA: +7.3 (P = .002) PBO: −0.6 (P = .002 DHA vs PBO) | DHA: −23.1 (P < .001) PBO: 0 (P = .03 DHA vs PBO) | DHA: +5.1 PBO: −0.7 | Nutrinova Exclusions from number randomized well documented. Olive oil PBO SD back-calculated from SEM |
| Kelley et al,37 2007 | ♂; HL; TC < 301 mg/dL; TG 150–398 mg/dL; LDL-C < 220 mg/dL; BMI 22–35 kg/m2 (n = 34) | DHA 3 (n = 17) PBO (n = 17) 12 weeks | DHA∗∗ +18.90 (P < .05) PBO +4.39 (NS DHA vs PBO)∗∗∗ | DHA: +7.5 PBO: +3.2 | DHA:∗∗ −13.82 (P < .05) PBO: −2.15 (NS DHA vs PBO)∗∗∗ | DHA: +4.1 PBO: −2.7 | Martek Olive oil PBO |
| Maki et al,38 2005 | ♀♂; low HDL-C (♂: ≤ 42 mg/dL; ♀ ≤ 54 mg/dL) (n = 57) | DHA 1.5 (n = 27) PBO (n = 30) 6 weeks | DHA: +11.7 PBO: +2.7 (P = .001 DHA vs PBO) | DHA: +9.2 PBO: +5.5 | DHA: −23.8 PBO: −8.5 (P = .015 DHA vs PBO) | DHA: +4.8 PBO: +0.2 | Martek Olive oil PBO |
| Sanders et al,39 2006 | ♀♂; healthy (n = 79)∗∗∗ | DHA 1.5 (n = 40) PBO (n = 39) 4 weeks | DHA: +6.7 PBO: −3.2 (P = .0002 DHA vs PBO; CI: 0.13, 0.40) | DHA: +6.6 PBO: −2.0 (P = .001 DHA vs PBO; CI: 0.05, 0.21) | DHA: −13.9 (P = .002) PBO: −2.2 (NS; CI: −0.21, 0.03) | DHA: +2.9 PBO: −3.1 | Source not listed; DHA formulation included 0.6 g DPA Did not specify randomization |
| Stark and Holub,40 2004 | ♀; postmenopausal (n = 38) | DHA 2.8 (n = 32) PBO (n = 32) 4 weeks | DHA: +8.5 PBO: +3.2 (NS DHA vs PBO) | DHA: +7.9 PBO: −3.1 (NS DHA vs PBO) | DHA: −19.9 (P < .05) PBO: −3.1 (P < .05 DHA vs PBO) | DHA: +42.1 PBO: +1.3 | Martek/Neuromins |
| Theobald et al,41 2004 | ♀♂, middle-aged; crossover (n = 38)∗∗ | DHA 0.7 (n = 38) PBO (n = 38) 12 weeks | DHA: +10.1 PBO: +2.8 (P = .004 DHA vs PBO; 95% CI: 0.08, 0.38) | DHA: +5.4 PBO: +1.4 (between group 95% CI: 0.005, 0.14) | DHA: −1.9 PBO: +12.3 (between group 95% CI: −0.37, 0.05) | DHA: +5.7 PBO: +3.3 | DHASCO/Martek Olive oil PBO 2 randomized subjects excluded for well-documented reasons. |
| Wu et al,42 2006 | ♀; postmenopausal, vegetarian (n = 27) | DHA 2.14 (n = 14) PBO (n = 13) 6 weeks | DHA: −3.9 PBO: +4.4 | DHA: +7.3 PBO: +3.8 | DHA: −17.1 PBO: +2.6 | DHA: −7.0 PBO: +7.4 | Westar Nutrition Corp. |
∗Number randomized, unless noted by ¶(number analyzed). |
†Estimated or provided dose of EPA or DHA administered daily. |
‡No study except Davidson et al, 1997,34 reported statistical comparisons for non–HDL-C. |
¶Number analyzed. |
‖Total of 10 mg pravastatin or 5 mg simvastatin once daily. |
∗∗Percentage change from baseline to end point calculated from raw data provided by study authors. To convert TC, LDL-C, or HDL-C values to mmol/L, multiply by 0.0259; to convert TG values to mmol/L, multiply by 0.0113; to convert FG to mmol/L, multiply by 0.0555. |
∗∗∗All study groups not listed. |
All participants enrolled in DHA/EPA head-to-head trials and studies involving EPA alone had lipid disorders. Healthy subjects were evaluated in 8 of the 12 trials involving DHA alone31, 32, 33, 36, 39, 40, 41, 42; in 3 of these, all subjects were vegetarians32, 36, 42 and in 2, all were postmenopausal women.40, 42 Men were exclusively enrolled in 422, 23, 31, 37 and women in 328, 40, 42 of 22 studies. In head-to-head trials,21, 22, 23, 24, 25, 26 DHA doses ranged from 2.3 to 4 g/day, and EPA doses ranged from 2.2 to 4.0 g/day. In the EPA vs DHA studies,21, 22, 23, 24, 25, 26 the average dose of each OM3 was 3.5–3.6 g/day. In the DHA-alone studies,31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42 the DHA dose ranged from 0.7 to 3.0 g/day and averaged 1.7 ± 0.7 g/day. EPA was always administered at 1.8 g/day in the EPA-alone studies.27, 28, 29, 30
Study durations ranged from 4 to 7 weeks in studies comparing DHA with EPA,21, 22, 23, 24, 25, 26 3 months to 5 years in studies of EPA alone,27, 28, 29, 30 and 6 weeks to 3 months in trials of DHA alone.31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42 Excluding a subgroup analysis from Tanaka’s group30 (a 5-year study), we found that the average duration of DHA studies was 7 weeks, and the average duration of EPA studies was 12 weeks. The mean or median baseline TG was ≤ 150 mg/dL in 14 studies, 151–200 mg/dL in 5 studies,24, 25, 29, 30, 38 and 201–300 mg/dL in 327,34,37 studies. The mean or median baseline TG did not exceed 300 mg/dL in any study. The mean baseline LDL-C was <200 mg/dL in all studies except 1, in which the average baseline LDL-C was 213 mg/dL.35
Effects of DHA and EPA on lipids and lipoproteins
LDL-COverall, DHA-treated subjects exhibited a slight rightward shift in Forest plots of the placebo-adjusted percent changes in mean LDL-C from baseline compared with EPA-treated subjects in each of the 6 controlled head-to-head studies (Fig. 2). Whereas EPA treatment resulted in reductions in LDL-C vs control in 4 of these 6 studies21, 22, 24, 26 (by 1.5%–4.7% from baseline), LDL-C levels decreased in only 1 DHA–treated group21 (by 1.2%). In the other 5 studies,22, 23, 24, 25, 26 the mean LDL-C increased from baseline to end point by 0.2%−10.9% with DHA (Table 1). Increases in LDL-C (by 2.0% and 6.5%) were observed in 2 of 6 studies of EPA treatment.23, 25 On average, LDL-C increased by 2.6 ± 4.3% with DHA treatment and decreased by 0.7 ± 4.2% with EPA treatment in the head-to-head studies (net effect = 3.3% decrease with EPA vs DHA). In head-to-head trials, 4 of 621, 22, 23, 26 performed statistical analyses on LDL-C after OM3 treatment. In 3 trials,21, 22, 26 no difference was found in LDL-C levels at end point with either treatment compared with baseline21 or control,22, 26 but in one study23 authors demonstrated a significant (P = .019) increase in LDL-C with DHA (but not EPA) compared with control.
Figure 2
Comparison of percent changes in mean values from baseline to end point for EPA or DHA treatment. Percent changes in mean TG or LDL-C, adjusted by placebo or control group values for DHA (filled circles) and EPA (open squares), and corresponding 95% CI (dashed lines) are presented. aThe 95% CI was not included where large data dispersion at end point precluded reasonable estimation. ∗N represents the total number of crossover study participants. Egert et al21 and Tanaka et al30 reported median values. BL, baseline; CNTRL, control; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; Exp, experimental group; LDL-C, low-density lipoprotein cholesterol; PBO, placebo; TG, triglyceride. (To convert LDL-C to mmol/L, multiply by 0.0259; to convert TG to mmol/L, multiply by 0.0113.)
In the remaining controlled studies in which the authors assessed the lipid effects of DHA alone, 10 (71%) of 14 found net increases in LDL-C, which ranged from +5.4% to +16.0%. These increases were statistically analyzed in 10 of 12 studies31, 32, 34, 35, 36, 37, 38, 39, 40, 41 and demonstrated to be significant (P < .05) from baseline,34, 37 control,35, 38, 39, 41 or both32, 36 in 8 of these studies. LDL-C levels were not increased vs control in any of the 4 studies in which EPA alone was used. Statistical comparisons were performed in 3 of these studies28, 29, 30 and significant differences (P < .05) with EPA compared with baseline were demonstrated in 2 studies.28, 29 LDL-C was reduced from baseline in 1 of the 4 EPA studies, by 15.9%, and was essentially unchanged in the remaining studies (<1% change in LDL-C). A total of 10 of the 12 DHA-alone groups had baseline TG levels <200 mg/dL; of these, 7 showed an increase in LDL-C with DHA.
Across all studies, LDL-C increased from baseline in 75% of DHA-treated groups (by +0.2 to +16%) as compared with 40% of EPA-treated groups (by +0.3 to +6.5%). The administered dose of DHA or EPA was not significantly associated with changes in LDL-C. However, increases in LDL-C did significantly correlate with baseline TG measures in DHA-treated, but not EPA-treated, subjects. There was no significant association between study duration and change in LDL-C with either treatment.
Percent changes (placebo- or control-adjusted) in mean LDL-C from baseline to end point on DHA treatment significantly correlated with baseline TG (Pearson r = 0.53; P = .02) but not baseline LDL-C levels (r = 0.38; P = .10). Percent changes (placebo- or control-adjusted) in mean LDL-C on EPA treatment were not significantly correlated with baseline TG (r = −0.40; P = .25) or LDL-C (r = 0.35; P = .32).
TGAmong head-to-head studies,21, 22, 23, 24, 25, 26 both EPA and DHA reduced TG levels from baseline. DHA treatment resulted in decreased mean TG from baseline in all groups (decreases of 8.0%–43.7%; Fig. 2; Table 1). Percent changes in mean (or median) TG measures in matched EPA-treated groups ranged from +1.8% to −34.9%. TG was reduced in all (6 of 6) DHA-treated groups, and in 5 of 6 EPA-treated groups. Across all head-to-head studies, average TG levels were reduced by both OM3s (average = –22.4 ± 13.3% [DHA] vs –15.6 ± 12.3% [EPA]), but to a greater degree by DHA than EPA (net effect = 6.8% decrease with DHA vs EPA).
In the 12 studies assessing the lipid effects of DHA alone, TG levels were lowered by 8.9%–30.3%. The range of reductions in TG in these studies was comparable in healthy volunteers (decreases of 11.7%–31.2%) or patients with dyslipidemia (decreases of 8.9%–25.5%). In studies evaluating the lipid effects of EPA monotherapy, TG measures decreased from baseline to end point in all treated groups (decreases of 4.4%–40.5%).
Across all 22 studies, the range of percent changes in mean TG was similar between treatments and, although we did not have sufficient data to examine a dose–response relationship, there were no apparent associations between OM3 doses and changes in TG for DHA studies. There was no significant association of the change in TG with the change in LDL-C for either OM3 treatment (DHA: r = 0.21, P = .36; EPA: r = 0.50, P = .14) or when both treatments were pooled (r = 0.19, P = .31). Changes in TG were also not significantly correlated with the duration of treatment with either OM3.
Non-HDL-CUnlike TG, non-HDL-C was not altered to a great extent by DHA or EPA treatment in head-to-head studies. Changes in levels ranged from –3.8% to +4.1% in DHA-treated subjects and –5.8% to +2.2% in EPA-treated subjects (Table 1). Non-HDL-C responses varied widely after either treatment alone (−17.5% to +40.8% [DHA]; −22.3% to +0.2% [EPA]). EPA treatment led to greater average reductions in non-HDL-C than did DHA treatment (−1.2% ± 2.9% [DHA] vs −2.9% ± 3.3% [EPA]; net effect = 1.7% decrease with EPA vs DHA). Across all studies, non-HDL-C values decreased in 80% of EPA-treated groups compared with 40% of DHA-treated groups. Reductions in non-HDL-C were positively correlated with changes in LDL-C in both DHA- and EPA-treated groups in head-to-head trials (r = 0.97, P = .002 [DHA], and r = 0.98, P = .001 [EPA]).
HDL-CIn head-to-head trials, HDL-C increased with DHA treatment, from +2.9% to +18.3%, but less with EPA treatment, with which the changes in HDL-C ranged from −2.3% to +7.9% (Table 1). Among the 6 head-to-head studies, HDL-C increased in all DHA-treated groups and, in EPA-treated groups, increased in 3 studies, decreased in 1 study, and was unchanged in 2 studies. On average, HDL-C levels increased to a greater degree with DHA than with EPA (+7.3 ± 6.1% [DHA] vs +1.4 ± 3.5% [EPA]; net effect = 5.9% increase with DHA). HDL-C levels increased in 12 of 14 DHA-alone–treated groups (–6.2% to +23.6%) and were decreased or unchanged in 3 of 4 EPA-alone–treated groups (–3.2% to +3.4%). There was no correlation between the dose of DHA and changes in HDL-C nor between changes in HDL-C and changes in TG in either group.
Discussion
We evaluated 22 RCTs to determine whether there was a differential effect of EPA or DHA on LDL-C and other lipid parameters in healthy study participants and individuals with lipid disorders. We identified many more studies meeting our criteria that examined DHA-alone treatments compared with EPA alone, probably because of the availability of algal oils in which DHA is the predominant component. In head-to-head studies, DHA treatment was typically associated with increases (or modest decreases) in LDL-C as compared with reductions or smaller increases with EPA. These trends were maintained in studies involving DHA or EPA alone. Across all study groups (ie, combining all DHA- or EPA-treated groups examined here), LDL-C increased from baseline in 75% (15 of 20) of DHA-treated groups as compared with 40% (4 of 10) of EPA-treated groups. These increases in LDL-C after DHA, but not EPA, treatment were directly and significantly associated with baseline TG levels.
Because study participants analyzed herein typically had normal or borderline-elevated TG levels, and hence included patients for whom P-OM3s are not indicated and who would not have been anticipated to exhibit increased LDL-C levels on OM3 treatment, our findings somewhat unexpectedly suggested that DHA increases LDL-C levels. In head-to-head comparative studies, the net increase in LDL-C with DHA (vs EPA) was 3.3%. In such studies, both EPA (−15.6%) and DHA (−22.4%) lowered TG (net decrease with DHA = 6.8%); increased HDL-C levels (by 7.3% with DHA vs +1.4% with EPA; net increase = 5.9% with DHA); and lowered non-HDL-C (by 1.2% with DHA vs 2.9% with EPA; net increase = 1.7% with DHA). Changes in TG, LDL-C, non-HDL-C, and HDL-C were not associated with the dose or duration of either OM3 treatment.
Although of modest magnitude, the net percent decreases in LDL-C (3.3%) and non-HDL-C (1.7%) with EPA compared with DHA are arguably of greater consequence in terms of cardiovascular risk status as against the larger average percent decrease in TG (net decrease = 6.8% with DHA), and the reciprocally larger, average percent increase in HDL-C (net increase = 5.9% with DHA), with DHA compared with EPA. Convergent findings from meta-analyses involving >100,000 patients enrolled in RCTs involving lipid-altering pharmacotherapies and other interventions (eg, dietary modification, ileal bypass) indicated an approximately 1:1 (or greater) relationship between percent reductions in LDL-C or non-HDL-C and percent decreases in CHD risk.15, 43, 44, 45 Although net decreases in TG and increases in HDL-C with DHA might be expected to somewhat offset the beneficial effects of net reductions in LDL-C and non-HDL-C with EPA, similarly robust, 1:1 (or greater) direct or inverse relationships between on-treatment changes in TG or HDL-C and CHD risk have not been demonstrated in clinical trials. However, it is important not to minimize the potential impact of raising HDL-C on cardiovascular risk; in an analysis of 4 prospective epidemiologic studies (N > 10,000), there was a 2%–3% reduction in CHD risk for every 1-mg/dL increase in HDL-C.46
To date, few reported systematic reviews have contrasted the effects of EPA and DHA on individual lipids and other parameters.47 Our findings are consistent with data reported by Ryan et al,48 which demonstrated that DHA is associated with significant increases in both LDL-C and HDL-C in patients with and without hypertriglyceridemia. The magnitudes of changes in TG that we observed were also compatible with data from the analysis by Ryan et al,48 which demonstrated significant reductions in TG (by up to 26%) with DHA treatment. In other reviews,16, 17, 18 authors have not evaluated the impact of EPA or DHA in isolation. A recent study further supports our findings: Schaefer and coworkers49 compared the lipid-altering effects of 6 weeks of treatment with EPA (0.6 or 1.8 g/day) or DHA (0.6 g/day) with control in 120 healthy volunteers. In this study, treatment with even this small dose of DHA (but not EPA) was associated with a significant increase in LDL-C (P < .05).
Potential mechanisms for the effects of EPA and DHA on Apo B–containing lipoproteins were reviewed by Bays and colleagues.50 Reductions in TG may result from increased clearance, reduced synthesis, and/or enhanced degradation of fatty acids in the liver. OM3s appear to increase the activity of lipoprotein lipase,25 potentially accelerating chylomicron TG clearance and promoting conversion of very-low-density lipoprotein (VLDL) to LDL. In vitro, VLDL particles enriched with OM3s were more readily converted to LDL than control particles.51 Reductions in TG and increases in LDL-C observed with DHA treatment may partly result from conversion of VLDL to LDL, which was significantly correlated with baseline TG levels in our analysis.
Preliminary preclinical and clinical studies have begun to suggest differential effects of EPA and DHA on TG clearance and metabolism via lipoprotein lipase or peroxisome proliferator–activated receptors (PPARs),25, 52 but the potential relationship of these data to our findings is unclear. Remodeling of LDL particle size and consequent changes in lipoprotein concentrations may result from inhibition of hepatic lipase and cholesteryl ester transfer protein, which exchanges cholesteryl esters in cholesterol-rich lipoproteins (LDL and HDL) for TGs in TG-rich lipoprotein particles (VLDL). In patients with hypertriglyceridemia, cholesteryl ester transfer protein has augmented acceptor activity for VLDL particles, which are increased in this population.53 Certain evidence from studies in vitro suggests that cholesteryl ester transfer protein is inhibited to different extents by DHA and EPA,54 and such disparities may help to explain different (greater) increases in HDL-C and reciprocal decreases in TG observed with DHA compared to EPA.47
An increase in LDL particle size may help to account for the increase in LDL-C observed with DHA in some studies.23 However, several studies on EPA have also shown reductions in small, dense LDL-C without accompanying increases in overall LDL levels.29, 49, 55 Although an increase in LDL particle size has been observed with DHA,23, 47 no study to date has examined whether DHA affects total LDL particle number. As far as large LDL particles being less atherogenic than small, dense LDL, several studies now demonstrate that all LDL particles are atherogenic and that LDL particle number better explains risk than the cholesterol content of different LDL subfractions (eg, in the Cholesterol and Recurrent Events [CARE] and Multi-Ethnic Study of Atherosclerosis [MESA] trials).56, 57
Largely on the basis of previously reported in vitro, in vivo, and clinical data, one plausible mechanism to help explain the dissimilar effects of DHA and EPA on LDL-C levels across studies of varying designs (and in individuals without profound TG elevations) is that dietary supplementation with DHA-containing long-chain fatty acids down-regulates receptor-mediated LDL-C clearance, in part via reduced expression of LDL receptors by hepatocytes.41, 58, 59, 60, 61, 62, 63 This effect may in turn be mediated by the influences of DHA on sterol regulatory element–binding protein or liver X receptor (secondary to increased generation of hepatic oxysterols).41 Distinct effects of DHA and EPA on LDL-receptor expression may parallel a difference between effects of DHA and EPA on lipid transfer protein-mediated cholesteryl ester transfer from HDL to LDL, with attendant reductions in HDL-C levels.64, 65 In addition, the LDL-C–raising effects of DHA might be ascribed to increased conversion of VLDL to LDL via increased lipoprotein lipase activity.60 Increased lipoprotein lipase activity may in turn be partially explained by up-regulated PPAR- γ and/or PPAR-α gene expression.
Potential limitations
This review was of a descriptive and inherently exploratory and hypothesis-generating nature, not a rigorous statistical data assessment. No patients in the present analysis had profoundly elevated levels of TG (>500 mg/dL), and only 3 of 22 studies involved patients with TG >200 mg/dL. We would expect these current conditions to result in relatively conservative estimates of the effects of EPA and DHA on lipids and lipoproteins. Hence, it is also not clear whether our observed effects are generalizable to dyslipidemic patients, including patients with greater levels of TG; the increases in LDL-C and HDL-C from baseline with DHA observed in our analysis might actually underestimate the magnitudes of such increases in patients with more pronounced elevations in TG. Moreover, there was a greater variability in dosing for DHA–alone studies compared with EPA–only studies (all of which assessed 1.8 g/day), of which 1 of 4 studies required the use of the Friedewald equation to back-calculate LDL-C. Another limitation was that 2 of the 12 DHA-alone studies had <10 participants per treatment group. Additional studies are warranted to more methodically and rigorously discriminate effects of DHA or EPA on a broader spectrum of lipids and lipoproteins in a larger and more heterogeneous patient population.
Another potential limitation relates to disparities in constituents of the fatty-acid vehicles of DHA sources in head-to-head trials, compared with DHA-alone studies. DHA sources in head-to-head studies comprised mainly ethyl esters in 5 of 6 studies, whereas vehicles in the DHA-alone studies comprised well-defined algal oils. In contrast to the head-to-head studies including DHA ethyl esters, which contained highly pure DHA (>90% pure in 4 of 6 studies and >85% pure in 1 of 6 studies), the algal oils contained only 38% DHA, as well as 30% saturated fatty acids (mainly myristic and palmitic acids), which can, in theory, raise LDL-C independently of the OM3. The control treatment in most (7 of 12) DHA-alone studies contained 12.5% saturated fatty acids, and these studies did not rigorously correct for any potential impact of saturated fatty acids on LDL-C. Although potential differences in vehicle and/or control fatty-acid constituents should certainly not be overlooked in future studies, the introduction of a small number of grams of saturated fats in OM3 vehicles within the trials we reviewed would not be expected to markedly or systematically alter calculations of DHA’s effects on LDL-C.
Our literature review strategy excluded studies that were not randomized or controlled and also excluded case reports and series. Taken together, these criteria might have contributed to some degree of selection bias, given that more substantial lipid changes may be required for publication of nonrandomized trials and case reports or series. Many of the study reports included in our analysis did not provide measures of dispersion along with measures of central tendency. Although an appropriate, statistically rigorous method was employed to estimate these values, it was in general not possible to confirm 95% CI values with authors of the original papers. The present analysis evaluated the effects of DHA or EPA on LDL-C and other lipid risk factors or markers but not on other surrogate outcome variables of interest, including LDL particle size,23 inflammatory markers (eg, high-sensitivity C-reactive protein) or other pathophysiologic indices (eg, measures of coagulation function) or clinical events (e.g. coronary morbidity or mortality). Finally, the present analysis included limited information on patients’ dietary behaviors and apolipoprotein levels, in part because these data were not reported in many of the original studies reviewed herein.
Conclusions
Treatment with EPA or DHA is associated with different effects on LDL-C and HDL-C. Whereas treatment with either agent lowered the TG level, DHA treatment was more often associated with increases in LDL-C and in HDL-C. These effects across studies are noteworthy given that most patient populations did not have major TG elevations. Future long-term, prospective RCTs are warranted to further explore the impact of OM3s on a broader spectrum of lipids and lipoproteins (both their concentrations and architectures), and, if possible, on coronary-event or angiographic end points, in larger and more heterogeneous patient populations, including greater proportions of individuals with more severe dyslipidemia.
Acknowledgments
Assistance in manuscript preparation was provided by Lauren Baker, PhD, and Stephen W. Gutkin, Rete Biomedical Communications Corp. (Wyckoff, NJ), with financial support from the sponsor. We thank William S. Harris, PhD (Sioux Falls, SD), for reviewing earlier drafts of the manuscript and making intellectual contributions to its content.
Financial disclosures
Dr. Jacobson discloses that he has served as a consultant for Abbott, Amarin, AstraZeneca, GlaxoSmithKline, Kowa, and Merck. Dr. Glickstein discloses that she is a paid consultant to Rete Biomedical Communications Corp., which is, in turn, a paid consultancy to the sponsor. Dr. Soni is an employee of, and minor shareholder in, the study sponsor. This review was supported by Amarin Corporation. Dr. Jacobson is a consultant for Amarin but did not receive compensation for his role in the design of the review, interpretation of data, or drafting of the final manuscript.
Supplementary data
Supplemental Appendix.
References
- 30-year trends in serum lipids among United States adults: results from the National Health and Nutrition Examination Surveys II, III, and 1999-2006. Am J Cardiol. 2010;106:969–975
- . Fish consumption, fish oil, omega-3 fatty acids, and cardiovascular disease. Circulation. 2002;106:2747–2757
- . Omega-3 fatty acids and coronary heart disease risk: clinical and mechanistic perspectives. Atherosclerosis. 2008;197:12–24
- Effects of eicosapentaenoic acid on major coronary events in hypercholesterolaemic patients (JELIS): a randomised open-label, blinded endpoint analysis. Lancet. 2007;369:1090–1098
- Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto miocardico. Lancet. 1999;354:447–455
- Lovaza® (omega-3-acid ethyl esters) [US full prescribing information]. Research Triangle Park, NC: GlaxoSmithKline; 2009. Available at: http://us.gsk.com/products/assets/us_lovaza.pdf. Accessed March 23, 2011.
- Epadel® [Complete prescribing information]. Update (Version 5). Tokyo, Japan: Mochida Pharmaceutical; January 2007.
- Safety and efficacy of Omacor in severe hypertriglyceridemia. J Cardiovasc Risk. 1997;4:385–391
- Correlation of serum triglyceride and its reduction by omega-3 fatty acids with lipid transfer activity and the neutral lipid compositions of high-density and low-density lipoproteins. Atherosclerosis. 1999;143:285–297
- Center for Drug Evaluation and Research. Approval Package for: 21-654 (Omacor/Lovaza). Statistical Review; 2004. Available at: http://www.accessdata.fda.gov/drugsatfda_docs/nda/2004/21-654_Omacor_AdminCorres_P1.pdf. Accessed June 24, 2011.
- . Baseline lipoprotein lipids and low-density lipoprotein cholesterol response to prescription omega-3 acid ethyl ester added to simvastatin therapy. Am J Cardiol. 2010;105:1409–1412
- Efficacy and tolerability of adding prescription omega-3 fatty acids 4 g/d to simvastatin 40 mg/d in hypertriglyceridemic patients: an 8-week, randomized, double-blind, placebo-controlled study. Clin Ther. 2007;29:1354–1367
- Long-term up to 24-month efficacy and safety of concomitant prescription omega-3-acid ethyl esters and simvastatin in hypertriglyceridemic patients. Curr Med Res Opin. 2010;26:907–915
- Center for Drug Evaluation and Research. Application Number: 21-853, 21654s016, (Omacor). Statistical Review and Evaluation: Clinical Studies, Omacor (omega-3 acid ethyl ester) Capsules, 4 grams/day; 2007. Available at: http://www.accessdata.fda.gov/drugsatfda_docs/nda/2007/021853s000;%20021654s016_StatR.pdf. Accessed June 24, 2011.
- Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267–1278
- . Effects of omega-3 fatty acids on serum markers of cardiovascular disease risk: a systematic review. Atherosclerosis. 2006;189:19–30
- . Potential impact of omega-3 treatment on cardiovascular disease in type 2 diabetes. Curr Opin Lipidol. 2009;20:30–38
- . Toxicology and safety of DHA. Prostaglandins Leukot Essent Fatty Acids. 2009;81:125–132
- Haney EM, Huffman LH, Bougatsos C, et al. Screening for lipid disorders in children and adolescents; Systematic evidence review for the U.S. Preventive Services Task Force (evidence synthesis). No. 47. Rockville, MD: Agency for Healthcare Research and Quality, US Department of Health and Human Services; AHRQ Publication No. 07-0598-EF-1; July 2007. Available at: http://www.uspreventiveservicestaskforce.org/uspstf07/chlipid/chlipidsyn.pdf. Accessed March 23, 2011.
- . Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499–502
- . Dietary alpha-linolenic acid, EPA, and DHA have differential effects on LDL fatty acid composition but similar effects on serum lipid profiles in normolipidemic humans. J Nutr. 2009;139:861–868
- . Highly purified eicosapentaenoic acid and docosahexaenoic acid in humans have similar triacylglycerol-lowering effects but divergent effects on serum fatty acids. Am J Clin Nutr. 1997;66:649–659
- . Purified eicosapentaenoic and docosahexaenoic acids have differential effects on serum lipids and lipoproteins, LDL particle size, glucose, and insulin in mildly hyperlipidemic men. Am J Clin Nutr. 2000;71:1085–1094
- . The n-3 fatty acids eicosapentaenoic acid and docosahexaenoic acid increase systemic arterial compliance in humans. Am J Clin Nutr. 2002;76:326–330
- . Omega-3 fatty acid supplementation accelerates chylomicron triglyceride clearance. J Lipid Res. 2003;44:455–463
- . Effects of purified eicosapentaenoic and docosahexaenoic acids on glycemic control, blood pressure, and serum lipids in type 2 diabetic patients with treated hypertension. Am J Clin Nutr. 2002;76:1007–1015
- . Eicosapentanoic acid reduces plasma levels of remnant lipoproteins and prevents in vivo peroxidation of LDL in dialysis patients. J Am Soc Nephrol. 1999;10:2177–2184
- . Eicosapentaenoic acid effect on hyperlipidemia in menopausal Japanese women. The Niigata Epadel Study Group. Obstet Gynecol. 2000;96:521–528
- Purified eicosapentaenoic acid reduces small dense LDL, remnant lipoprotein particles, and C-reactive protein in metabolic syndrome. Diabetes Care. 2007;30:144–146
- Reduction in the recurrence of stroke by eicosapentaenoic acid for hypercholesterolemic patients: subanalysis of the JELIS trial. Stroke. 2008;39:2052–2058
- Fish diet, fish oil and docosahexaenoic acid rich oil lower fasting and postprandial plasma lipid levels. Eur J Clin Nutr. 1996;50:765–771
- . Supplementation with an algae source of docosahexaenoic acid increases (n-3) fatty acid status and alters selected risk factors for heart disease in vegetarian subjects. J Nutr. 1996;126:3032–3039
- . Effect of supplementation with different doses of DHA on the levels of circulating DHA as non-esterified fatty acid in subjects of Asian Indian background. J Lipid Res. 1998;39:286–292
- . Effects of docosahexaenoic acid on serum lipoproteins in patients with combined hyperlipidemia: a randomized, double-blind, placebo-controlled trial. J Am Coll Nutr. 1997;16:236–243
- Docosahexaenoic acid restores endothelial function in children with hyperlipidemia: results from the EARLY study. Int J Clin Pharmacol Ther. 2004;42:672–679
- . Microalgal docosahexaenoic acid decreases plasma triacylglycerol in normolipidaemic vegetarians: a randomised trial. Br J Nutr. 2006;95:779–786
- . Docosahexaenoic acid supplementation improves fasting and postprandial lipid profiles in hypertriglyceridemic men. Am J Clin Nutr. 2007;86:324–333
- Lipid responses to a dietary docosahexaenoic acid supplement in men and women with below average levels of high density lipoprotein cholesterol. J Am Coll Nutr. 2005;24:189–199
- . Influence of an algal triacylglycerol containing docosahexaenoic acid (22:6 n-3) and docosapentaenoic acid (22:5 n-6) on cardiovascular risk factors in healthy men and women. Br J Nutr. 2006;95:525–531
- . Differential eicosapentaenoic acid elevations and altered cardiovascular disease risk factor responses after supplementation with docosahexaenoic acid in postmenopausal women receiving and not receiving hormone replacement therapy. Am J Clin Nutr. 2004;79:765–773
- . LDL cholesterol-raising effect of low-dose docosahexaenoic acid in middle-aged men and women. Am J Clin Nutr. 2004;79:558–563
- . 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
- . Meta-analysis of the relationship between non-high-density lipoprotein cholesterol reduction and coronary heart risk. J Am Coll Cardiol. 2009;53:316–322
- . Quantifying effect of statins on low density lipoprotein cholesterol, ischaemic heart disease, and stroke: systematic review and meta-analysis. Br Med J. 2003;326:1423–1427
- . Cholesterol reduction yields clinical benefit: impact of statin trials. Circulation. 1998;97:946–952
- High density lipoprotein cholesterol and cardiovascular disease: four prospective American studies. Circulation. 1989;79:8–15
- . The independent effects of eicosapentaenoic acid and docosahexaenoic acid on cardiovascular risk factors in humans. Curr Opin Clin Nutr Metab Care. 2006;9:95–104
- . Clinical overview of algal-docosahexaenoic acid: effects on triglyceride levels and other cardiovascular risk factors. Am J Ther. 2009;16:183–192
- Effects of eicosapentaenoic acid, docosahexaenoic acid, and olive oil on cardiovascular disease risk factors [abstract 20007]. Circulation. 2010;122:A20007
- . Prescription omega-3 fatty acids and their lipid effects: physiologic mechanisms of action and clinical implications. Expert Rev Cardiovasc Ther. 2008;6:391–409
- . Omega-3 fatty acids alter lipoprotein subfraction distributions and the in vitro conversion of very low density lipoproteins to low density lipoproteins. J Nutr Biochem. 1999;10:151–158
- . In contrast with docosahexaenoic acid, eicosapentaenoic acid and hypolipidaemic derivatives decrease hepatic synthesis and secretion of triacylglycerol by decreased diacylglycerol acyltransferase activity and stimulation of fatty acid oxidation. Biochem J. 1999;343(Pt 1):191–197
- . Cholesteryl ester transfer protein: at the heart of the action of lipid-modulating therapy with statins, fibrates, niacin, and cholesteryl ester transfer protein inhibitors. Eur Heart J. 2010;31:149–164
- . Regulation by long-chain fatty acids of the expression of cholesteryl ester transfer protein in HepG2 cells. Lipids. 2001;36:401–406
- . Effects of purified eicosapentaenoic acid ethyl ester on plasma lipoproteins in primary hypercholesterolemia. Int J Vitam Nutr Res. 1992;62:256–260
- . Low-density lipoprotein size, pravastatin treatment, and coronary events. JAMA. 2001;286:1468–1474
- LDL particle subclasses, LDL particle size, and carotid atherosclerosis in the Multi-Ethnic Study of Atherosclerosis (MESA). Atherosclerosis. 2007;192:211–217
- . Effect of dietary fatty acids on LDL binding. Ann N Y Acad Sci. 1993;683:178–182
- . Low density lipoprotein from humans supplemented with n-3 fatty acids depresses both LDL receptor activity and LDLr mRNA abundance in HepG2 cells. J Lipid Res. 1992;33:647–658
- . Eicosapentaenoic acid is primarily responsible for hypotriglyceridemic effect of fish oil in humans. Lipids. 1996;31(Suppl):S45–S49
- . The effects of dietary fish oil on hepatic high density and low density lipoprotein receptor activities in the rat. FEBS Lett. 1987;222:159–162
- . Dietary fish oil decreases low-density-lipoprotein clearance in nonhuman primates. Am J Clin Nutr. 1996;64:215–221
- . Dependence on dietary cholesterol for n-3 polyunsaturated fatty acid-induced changes in plasma cholesterol in the Syrian hamster. J Lipid Res. 1992;33:263–271
- . Specificity of lipid transfer protein for molecular species of cholesteryl ester. J Lipid Res. 1986;27:523–529
- . Effects of N-3 fatty acids on lipid metabolism. Ann Rev Nutr. 1990;10:149–167
PII: S1933-2874(11)00745-8
doi:10.1016/j.jacl.2011.10.018
© 2012 National Lipid Association. Published by Elsevier Inc. All rights reserved.
