Effect of n-3 Fatty Acid Supplementation in Hyperlipidemic Patients Taking Statins

This study is a parallel randomised, double-blind placebo - controlled trial which was conducted in Istanbul, Turkey.

BACKGROUND

Cardiovascular disease (CVD) has been, and still is the number 1 cause of deaths all over the world. According to the World Health Organisation (WHO), 1 of every 3 death was caused by CVD in 2012. Heart disease and stroke were the No. 1 and No. 2 killers worldwide, according to American Heart Association. In Turkey, it is even worse, 40% of all deaths were caused by CVD. Such high prevalence naturally brings a huge disease burden and in many countries there are ongoing research to mitigate CVD risk.

It is essential to highlight that CVD is not a single disease, it is actually a family of diseases that mostly cause by a common pathology named "atherosclerosis". Atherosclerosis is plaque formation in artery walls which leads a reduced lumen diameter, causing organ damage or clinical events like myocardial infarction (MI), angina or stroke. One of the two reasons of atherosclerosis is hyperlipidemia, which is the theme of this study. Atherosclerosis and lipids are evident to be related; the elevations in circulating LDL cholesterol level plays a central role in atherogenesis, given the fact that they cause the formation of fatty plaque. Therefore, in all clinical guidelines, LDL level is the primary target to prevent atherosclerosis.However, more recent studies have revealed that LDL is structurally heterogeneous based on its size and density. There are large, buoyant LDL particles and small, dense LDL particles, which are believed to be more atherogenic. Individuals that have a large amount of small and dense LDL particles are at an increased risk of CV events, regardless of their total LDL level. In studies, small and dense LDL profile was associated with a three-fold to seven-fold increase in risk of CVD as they:1- Have the increased ability to penetrate arterial wall. 2- Are readily oxidized.

The main therapeutic approach to treat small\&dense LDL is using drugs that lower serum triglyceride (TG) levels, as they also cause a change in the LDL size profile to larger, less dense and, therefore, less atherogenic species. Most commonly used such drugs are statins. Apart from statins, there is strong evidence that omega-3 fatty acids also have a beneficial effect in lowering TGs. However, their effect on small, dense LDL has not been studied enough yet. If omega-3 fatty acids have a beneficial effect in changing LDL profile to less dense and larger particles, omega-3 supplementation will be quite effective for patients with hyperlipidemia. Given the fact that statin use is the gold standard in all clinical guidelines, this study will not compare statins with omega-3 fatty acids and instead, compare statin monotherapy to the combined use of statin and omega-3 supplements to investigate whether omega-3 fatty acids have a supportive role to statin.

RECRUITMENT AND SAMPLING:

Eligible participants were men or women from Istanbul, Turkey, aged between 50 and 79 years who had been receiving a stable dose of atorvastatin for the control of LDL-C levels at least for 8 weeks before initial screening.

One hundred patients were selected for initial screening and 44 matched the inclusion/exclusion criteria and were randomized to the intervention using www.randomization.com.

Participants were informed informally via telephone first, and then invited to the Clinic by a Patient Invitation Letter. A structured one-to-one information session was carried out with each participant at the Clinic, during which they were given a Participant Information Sheet. Participants provided written informed consent by signing the a Participant Consent Form.

SAMPLE-SIZE

Total sample size was 44, consisting of two 22-subject groups (control and intervention group).

A targeted sample size of 36 subjects (18 per arm) was expected to provide at least 80% power to detect a difference of 5% or above in primary outcome non-HDL-C and LDL-C III particle concentration when compared with placebo (α=0.05) . Sample size was assigned as 44 to allow for subject attrition and other potential causes of study withdrawal up to 20%.

The power calculation was based on the mean changes in non-HDL-C measurement of the previous COMBOS trial which is a largest randomized placebo controlled trial that has the same design with this study. Data from COMBOS trial is summarized as follows:

Intervention Group:

baseline non-HDL mean + standard deviation (SD): 135.8 mg/dL + 24.5 mg/dL final non-HDL mean + SD: 124.1 mg/dL + 24.7 mg/dL Percent change: -7.9%

Placebo Group:

baseline non-HDL mean +SD: 141.3 mg/dL + 29.3 mg/dL final non-HDL mean + SD: 138.8 mg/dL + 32.0 mg/dL Percent change: -1.5%

METHODS

A randomized, double-blind, placebo controlled, parallel groups study was carried out. Before entry into the intervention phase of the study, 8-weeks of dietary lead-in was carried out. Study participants received dietary counselling for the 8 weeks before the intervention, based on the heart-healthy diet structured in the United Kingdom (UK) National Institute of Health Care Excellence (NICE) guidelines (2014) and were instructed to maintain this diet throughout the study. Adherence to dietary advice was also measured by a 3-day weighed intake food diary at Week 4. Household measures were used to assess the food intake. Nutrient content of the diet was estimated using Nutritics software (Nutritics Ltd, Ireland) using UK: Scientific Advisory Committee on Nutrition (SACN) 2015 food composition tables.

After the dietary lead-in, baseline measurements of fasting blood Triglycerides (TG), Total Cholesterol (TC), High-density Lipoprotein cholesterol (HDL-C), Low-density Lipoprotein cholesterol (LDL-C), and LDL-C sub groups (LDL-C I, LDL-C II, LDL-C III, LDL-C IV, LDL-C V) were carried out for all subjects at 2 clinics separated by 1 week, and the means were used as baseline values. Non-High-density Lipoprotein cholesterol (non-HDL-C) was calculated by subtracting HDL-C from TC.

After baseline measurements, all participants were randomized in equal numbers to receive either 4g/day EPA (75%) \& DHA (25%) ethyl esters or 4g/day olive oil (placebo) for 8 weeks in combination with the same dose of atorvastatin they had been prescribed. For all subjects, atorvastatin dosage was kept constant throughout the trial.

Both groups consumed four 1000mg capsules orally 4 times daily and compliance was measured by the number of capsules consumed relative to the number estimated to be consumed.

Baseline measurements were repeated at the end of week 8. Study participants and investigators remained blinded to all laboratory results until completion of analysis.

BIOCHEMICAL MEASUREMENTS

Laboratory analyses were performed by e-Lab Laboratories (Istanbul, Turkey) on the serum or plasma of 12 hours fasting blood samples. LDL-C, VLDL-C and HDL-C was measured with homogeneous enzymatic colorimetric assay of Roche Diagnostics (USA). TG and TC was measured with enzymatic colorimetric assay of Roche Diagnostics (USA). LDL-C subgroups were analyzed with electrophoresis by Lipoprint (Quantimetrix, USA).

STATISTICAL ANALYSIS

Statistical analyses was be performed with Statistical Package for the Social Sciences (SPSS) version 24 (IBM, USA).

Demographic and baseline analysis was performed for all study participants, while efficacy analysis was performed only on subjects that successfully completed the 8-week study protocol. The primary efficacy end points are non-HDL and LDL-C III particle concentration percentage change from baseline; secondary efficacy end points are changes in TG, TC, LDL-C, VLDL-C and HDL-C.

Normal distribution of the sample was checked with Shapiro-Wilk's test. Mixed model ANOVA test was carried out to compare changes in outcome parameters from baseline to end-of-treatment between the intervention and control groups.

Paired t-tests were performed to analyze the changes from baseline to Week 8 for each outcome parameter, and multiple Independent t-tests were performed to analyze significant differences between groups across time points. A P-value < 0.05 determined statistical significance.