OUP user menu

Effect of dietary supplementation with ω-3 fatty acids on progression of atherosclerosis in carotid arteries

Peter Angerer, Wolfgang Kothny, Stefan Störk, Clemens von Schacky
DOI: http://dx.doi.org/10.1016/S0008-6363(02)00229-8 183-190 First published online: 1 April 2002

Abstract

Objective: Omega-3 polyunsaturated fatty acids (ω-3 PUFA) from fish oil slow atherosclerosis progression in coronary arteries, as we showed in a randomized double-blind placebo-controlled clinical trial. Embedded in this trial, the present study examined the influence of 2 years of dietary supplementation with 1.65 g ω-3 PUFA per day on progression of carotid atherosclerosis in 223 patients with coronary artery disease. Methods: Coronary angiography, a comprehensive clinical examination, and intima-media thickness measurement by B-mode ultrasound of the carotid arteries (common, internal and bifurcation), were performed at the study start and study end. An expert panel visually evaluated the global change of carotid atherosclerosis on a semiquantitative scale. A second outcome measure was the change of overall mean maximum intima-media thickness. Results: One hundred and seventy-one patients completed the study. In the global change score, 38% of the patients in the fish oil group and 35% in the placebo group showed progression. Global change was not different between intervention groups. Mean maximum intima-media thickness increased by 0.07±0.13 mm and 0.05±0.11 mm in the fish oil and placebo group, respectively (mean±S.D., P=0.24). No correlation was found between the change in carotid and coronary arteries. Conclusions: In this group of selected patients with documented coronary artery disease ω-3 PUFA given for 2 years did not demonstrate an effect on slowing progression of atherosclerosis in carotid arteries as measured by ultrasound.

Keywords
  • Arteries
  • Atherosclerosis
  • Cerebrovascular disorders
  • Lipid metabolism
  • Ultrasound

Time for primary review 22 days.

1 Introduction

An association between high fish intake and high incidence of stroke, mainly of the hemorrhagic subtype, was observed in early epidemiological studies, comparing Greenland Inuit to the Danish population [1,2]. This may be one reason why as yet no randomized trial has been undertaken with the primary objective to clarify the effect of ω-3 polyunsaturated fatty acids (ω-3 PUFA) on cerebral arteries or stroke. Meanwhile several large-scale epidemiological studies in Western populations have either found no association between fish consumption and stroke [3,4] or an inverse relationship [5–7]. The Gruppo Italiano per lo Studio della Sopravivenza nell* Infarto miocardico-Prevenzione (GISSI-P) trial clearly demonstrated a reduced mortality in patients with a recent acute myocardial infarction who were treated by low dose ω-3 PUFA; however, no change in the risk for stroke, a secondary end-point, was observed [8].

Carotid artery intima-media thickness (IMT) predicts stroke [9,10]. Stenosis of internal carotid arteries with >60% lumen narrowing plays a causal role for a large percentage of strokes [11]. In addition, progression in carotid IMT predicts risk for coronary events beyond that predicted by coronary arterial measures of atherosclerosis [12]. Carotid IMT measured by ultrasound on the wall far from the probe reflects the atherosclerotic changes detected by autopsy [13]. Thus, complementary to the information about the vessel lumen obtained by angiography IMT provides information about the vessel wall.

In the Study on Prevention of Coronary Atherosclerosis by Intervention with Marine Omega-3 Fatty Acids (SCIMO) we have shown that ω-3 PUFA slow progression of atherosclerosis in coronary arteries [14]. The study on which we report here investigated the effect of ω-3 PUFA on progression of carotid artery IMT. We hypothesized that in patients with coronary artery disease (CAD) dietary supplementation with 1.65 g/day eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) for 2 years would slow the progressive thickening of the intima-media layer in the carotid arteries.

2 Methods

2.1 Study design

The trial on the effect of ω-3 PUFA on carotid atherosclerosis was embedded in SCIMO, which has been reported in detail previously [14]. In brief, SCIMO was a randomized prospective, double-blind 2-year intervention trial, conducted in accordance with the European Guidelines for Good Clinical Practice (GCP) in one medical center of a tertiary care university hospital. GCP requires among others regular monitoring by an independent external institution. Primary end-point of the present study was the change of atherosclerosis in carotid arteries as expressed by a global change score per patient in an intention-to-treat analysis. This end-point was chosen as primary by analogy with the coronary study [14] where the change of severity of CAD as assessed by an expert panel on a scale (see below) was the primary end-point. Secondary end-point was the change of the overall mean maximum IMT per patient.

2.2 Patients

All patients who had a diagnostic coronary angiography at our institution between September 1, 1992, and May 19, 1994 were eligible if they had >20% stenosis in ≥1 coronary vessel unless at least one of the following exclusion criteria was present: revascularization by percutaneous coronary intervention or coronary bypass operation planned or performed in the preceeding 6 months in more than one vessel; cardiac transplantation; age under 18 or over 75 years; hemodynamically relevant left main stenosis or proximal stenosis in all three main vessels; biplane left ventricular ejection fraction below 35%; diabetes; and others [14]. Written informed consent was obtained from all patients prior to randomization. The study was approved by the internal review board of the Ethics Committee of the Faculty of Medicine of the Ludwig-Maximilians-University of Munich. The investigation conformed with the principles outlined in the Declaration of Helsinki.

2.3 Intervention

Patients were randomized to fish oil or placebo to be ingested in opaque, identical-looking and -smelling, soft gelatine capsules. Each capsule contained 1 g of a fatty acid mixture. During the first 3 months, six capsules, corresponding to an intake of 3.3 g ω-3 PUFA (EPA plus DHA) per day, and in the following 21 months, three capsules, corresponding to an intake of 1.65 g ω-3 PUFA (EPA plus DHA) per day, were advised. The placebo group received a mixture of fatty acids reflecting the fatty acid composition of the average European diet. Further details of the treatment have been reported [14]. Overweight patients (body mass index >25 kg/m2) were advised to restrict their caloric intake, and all patients were advised to avoid cholesterol-rich food. Other dietary advice was not given. Medical treatment was advised according to current guidelines. Decisions concerning treatment were mainly up to the patients* personal physician.

Compliance was assessed by interrogation, by counting returned capsules and by analysis of red cell phospholipid fatty acid composition which reflects dietary fatty acid composition [14–16]. Patients, whose capsule count suggested ingestion of less than 1658 capsules, (≥2 S.Ds from the mean), were considered non-compliant. Patients whose percentage of EPA in red cell phospholipid fatty acids differed, (≥2 S.Ds from the mean of the respective treatment group), were also considered non-compliant. Non-compliant patients were excluded from the valid case analysis.

The randomization list was computer-generated by the trial monitor in Norway and the medication was prepared, encoded, and sent from there [14]. All patients, study personnel in contact with patients, and analytical personnel, were blinded with respect to study medication. In order to ensure blindness, patients were told that the contents of capsules were differing in composition but not necessarily in taste.

2.4 Measurements

At study start, after the diagnostic coronary angiography had been performed, the clinical status was determined by the investigators. Case report forms were then filled out with predefined criteria relative to history and clinical examination. Blood laboratory work-ups including 38 routine parameters plus erythrocyte phospholipid fatty acid composition, a resting electrocardiogram (ECG), and an exercise stress test were obtained. Patients were seen as outpatients at months 1, 6, 12, 18. At each visit, history, clinical status, and the complete laboratory work-up were repeated. At month 24, all investigations were repeated, including coronary angiography during a 2-day hospital stay. At month 0 and 24, but not at months 1, 6, 12, 18, blood was taken after an overnight fast. Low density lipoprotein-cholesterol (LDL) was calculated according to the Friedewald formula. Myocardial infarction, neurologic deficits (ischemic or hemorrhagic), hypertension, and diabetes mellitus were defined according to World Health Organization [17].

The evaluation process of the coronary angiograms has been reported previously [14]. In brief, three cardiologists who were blinded to all aspects which might identify the coronary angiograms evaluated the films, pairwise per patient, determined the lumen narrowing for each part of the coronary tree, then firstly rated each single change and eventually the overall change of the angiograms on a scale termed global change score. Of the seven possible scores both extreme values (extreme progression and regression) did not occur, resulting in a score with five categories [14].

2.5 Carotid ultrasound

Ultrasound examination of the carotid arteries was performed by the same investigator at the study start and study end, according to the identical protocol and using the same high resolution ultrasound unit with a 7.5 MHz sector scanner with Doppler facility (Aloka SSD 620, Japan). Instrument performance was checked regularly by means of an RMI 413 tissue mimicking phantom. Patients rested supine in a semi-darkened, air-conditioned room with a small pillow under their neck. Three anatomically predefined segments were examined: the distal 20 mm of the common carotid artery, the carotid bifurcation, and the proximal 10 mm of the internal carotid artery, on both sides of the neck (i.e. six segments in total). Firstly, the entire circumference of each segment was scanned as far as possible in order to detect any hemodynamically relevant stenosis. Then, three angles of interrogation were used: oblique anterior (+45°), lateral (0°) and oblique posterior (−45°). The optimal longitudinal view with a clear image of interfaces between lumen, intima-media, and adventitia was 2-fold magnified and recorded on video and high resolution X-ray film.

For evaluation, only images of the far wall were used since measurements of the near wall underestimate the true histological thickness [13,18]. An expert panel of two experienced sonographers evaluated the films, paired (before and after treatment) per patient, but in random order, and blinded with respect to all patient characteristics like name, randomization status, and temporal order of films. They first decided whether a segment was documented in sufficient quality at both points of time, at least from one angle of interrogation. Then they measured the maximum IMT per segment, to get two absolute values of IMT, on the digitized images, by means of the software NIH-Image (National Institutes of Health, USA), on a Power Macintosh 8100/80 with a high resolution screen. Each measurement was performed twice and the results averaged. The mean of all IMT measurements (maximum of six far wall measurements) per patient and per point of time was defined as overall mean maximum IMT. The mean of left and right side per segment, patient, and point of time was defined as the mean maximum IMT of the common and internal carotid artery and the carotid bifurcation. Finally, by directly comparing the images, the experts decided whether a segment had changed. At the beginning, they worked separately, and eventually reached a consensus for each segment. After the file had been closed, each segment's change was identified as progression or regression, according to the temporal order of the films. The global change was obtained by summing up the changes for each segment per patient. Thus a score resulted ranging from −6 to +6 including 0. This method was adapted from coronary angiography trials and modified according to the conditions of ultrasound images [14,19].

Reproducibility was assessed by repeating ultrasound examination either within 1 month of the examination at study start or at study end in a randomly selected subsample of 33 patients. The correlation coefficient between the first and second measurement was r=0.92. The mean difference of the overall mean maximum IMT and its S.D. were used as measures of reproducibility [20], and were 0.04±0.23 mm.

2.6 Statistical analysis

The estimation of study size has been described previously [14]. In brief, based on the assumption that ω-3 PUFA would slow progression of carotid atherosclerosis as assumed for coronary atherosclerosis, with an α of 5% and a β of 10%, and a drop-out rate of 33%, inclusion of 97 patients per group (73 patients per group for the analysis) was necessary. In the present trial we used an almost identical score as the primary end-point, which differed only from the coronary end-point in that it contained more subdivisions. In addition, a numerical measurement (IMT) was used. Both adaptations are expected to increase power. Therefore it was estimated that the sample size of SCIMO would allow us to detect a relevant treatment effect also in the carotid arteries.

In order to compare changes of coronary arteries with changes of carotid arteries, the change score of carotid arteries was modified, i.e. reduced to contain also only five (instead of originally 13) categories. This was done by integrating (‘collapsing’) scores >2 and <−2 into score 2 and −2, respectively. We termed the resulting score for the carotid arteries ‘modified score’.

All data were entered into the study computers twice independently, using Data Entry (SPSS, USA), verified, and a copy of all relevant data was deposited at the trial monitor in Norway. After confirmation of receipt, the randomization code was broken, and data processed on a Power Mac 7600/120 (Apple, USA) using the software SPSS in version 6.0 for the Macintosh. Student's t-test, chi2 and Wilcoxon's rank sum tests were used according to the nature of the variables.

3 Results

Of 2284 patients screened, 112 and 111 were randomized to fish oil and placebo, respectively. After 2 years, 89 patients in the fish oil group and 86 patients in the placebo group had a complete clinical follow-up. Of these, 87 patients with fish oil and 84 patients with placebo had a second ultrasound examination in the required quality (intention to treat population). Sequential coronary angiograms were performed in 159 of these subjects (in three subjects coronary angiograms but not carotid examinations were complete before and after intervention). Eventually, 80 patients with fish oil and 78 patients with placebo fulfilled the criteria for compliance (valid case population). (Trial profile see Fig. 1).

Randomization was successful with respect to all characteristics relevant for the primary end-point (Table 1). During follow-up, changes in characteristics with potential influence on carotid atherosclerosis were not significantly different in the fish oil group compared to the placebo group (Table 2). However, EPA and DHA in the erythrocyte membrane increased in the fish oil group within the first month (not shown) and stayed stable thereafter, whereas fatty acid composition did not change in the placebo group (Table 2).

View this table:
Table 1

Baseline characteristics of all patients with complete ultrasound follow-up by treatment group

Fish oil (n=87)Placebo (n=84)
Age (years)57.4±9.158.9±8.2
Male72 (82.8%)69 (82.1%)
Prior myocardial infarction43 (49.4%)47 (55.9%)
Prior TIA2 (2.3%)1 (1.2%)
Prior PRIND1 (1.1%)0
Prior ischemic stroke1 (1.1%)2 (2.4%)
Current smokers9 (10.3%)a17 (20.2%)a
Diagnosis of systemic hypertension45 (51.7%)37 (44.1%)
Family history of premature myocardial infarction20 (23.0%)27 (32.1%)
LDL-cholesterol (mmol/l)4.07±1.253.87±1.04
HDL-cholesterol (mmol/l)1.31±0.391.25±0.34
Triglycerides (mmol/l)2.19±1.342.15±1.09
Number of main coronary vessel with stenosis ≥70%1.3±0.81.3±0.9
Number of coronary segments with stenosis ≥50%2.1±1.42.1±1.5
Carotid arteries: mean maximum IMT (mm)1.26±0.411.31±0.41
Common carotid artery: mean maximum IMT (mm)0.86±0.290.91±0.28
Carotid bifurcation: mean maximum IMT (mm)1.54±0.551.65±0.62
Internal carotid artery: mean maximum IMT (mm)1.11±0.541.10±0.59
  • Figures indicate: mean±S.D. or absolute number (%). t-test or chi2-test; all P>0.25 except for the following.

  • a chi2, n.s. (P=0.072).

View this table:
Table 2

Potential influences on carotid atherosclerosis during 24 months follow-up by treatment group

Fish oil (n=87)Placebo (n=84)
Number of patients on aspirin85 (97.7%)82 (97.6)
Number of patients on lipid lowering agents40 (47.6)43 (53.1)
Number of patients on ACE inhibitors31 (36.0%)23 (28.4%)
Number of patients on beta-blockers68 (78.2%)69 (82.1%)
Number of patients smoking at follow-up12 (13.4%)13 (15.7%)
Change in LDL-cholesterol (mmol/l)−0.25±1.20−0.41±1.03
Change in HDL-cholesterol (mmol/l)−0.00±0.25a0.09±0.31a
Change in triglycerides (mmol/l)b−0.16±0.980.09±1.19
Change in systolic blood pressure (mmHg)8.3±19.18.6±24.2
Change in diastolic blood pressure (mmHg)4.2±11.53.4±14.6
Change of eicosapentaenoic acid in erythrocyte membrane (percent points)2.6±0.6c0.1±0.4c
Change of docosahexaenoic acid in erythrocyte membrane (percent points)4.2±0.2d0.1±2d
  • Figures indicate: mean±S.D. or absolute number (%). t-test or chi2-test; all P>0.25 except for the following:

  • a t-test, n.s. (P=0.073).

  • b t-test, n.s. (P=0.20).

  • c t-test, P<0.001.

  • d t-test, P=0.004.

Repeated assessment (score and IMT measurement) could be performed at a mean number of 5.3 (of maximum six) segments per patient. By expert panel evaluation, 37.8% of patients in the fish oil group and 34.5% in the placebo group were rated to show progression of carotid atherosclerosis after 2 years. The global score was not different in the fish oil group compared with the placebo group (Table 3). During follow-up, overall mean maximum IMT per patient increased by 0.07±0.13 mm and 0.05±0.11 mm in the fish oil and placebo group, respectively (mean±S.D., P=0.24). Increase of IMT was also slightly greater in the fish oil group when the change of each of the segments was analysed separately (common and internal carotid artery and carotid bifurcation) (Table 4). Exclusion of non-compliant patients (valid case analysis) yielded almost identical results for change of global score and change of mean maximum IMT (data not shown).

View this table:
Table 3

Change of carotid atherosclerosis per patient within 24 months follow-up as evaluated by expert panel

Fish oil n=87Placebo n=84
Mean number of sites compared5.4±1.15.3±1.1
Number of patients with:a
Score −2 (regression)1 (1.1%)0
Score −1 (regression)1 (1.1%)5 (6%)
Score 0 (no change)52 (59.8%)50 (59.5%)
Score 1 (progression)25 (28.7%)18 (21.4%)
Score 2 (progression)6 (6.9%)9 (10.7%)
Score 3 (progression)1 (1.1%)2 (2.4%)
Score 4 (progression)1 (1.1%)0
  • Figures indicate: number (%).

  • a Wilcoxon, n.s. (P=0.62).

View this table:
Table 4

Absolute change of IMT in carotid artery far walls during 24 months follow-up by treatment group

Fish oil n=87Placebo n=84
Mean number of sites compared5.4±1.15.3±1.1
Overall mean maximum IMT (mm)a0.07±0.130.05±0.11
Common carotid mean maximum IMT (mm)0.05±0.160.03±0.10
Bifurcation mean maximum IMT (mm)b0.06±0.130.03±0.18
Internal carotid mean maximum IMT (mm)0.11±0.290.09±0.23
  • Figures indicate: mean±S.D. t-test; all P>0.25 except for the following.

  • a t-test, n.s. (P=0.24).

  • b t-test, n.s. (P=0.12).

No stroke or prolonged reversible ischemic neurologic deficit occurred during the study. Of the 175 patients with complete clinical follow-up, one patient in the fish oil group and three patients in the placebo group reported symptoms that were rated as transient ischemic attack.

Data for the analysis of correlation between the modified global change score of the carotid arteries (five categories) and the coronary global change score (five categories) were complete for 159 patients. There was no correlation between the two scores in the entire population (Spearman's coefficient, r=−0.04, P=0.67) and no correlation when the fish oil and the placebo group were analysed separately.

4 Discussion

In this randomized controlled double-blind trial, patients with CAD received dietary supplementation with 1.65 g ω-3-PUFA per day (EPA plus DHA) or placebo, for 2 years, in addition to standard treatment. Active treatment did not slow progression of carotid atherosclerosis. Change of atherosclerosis in carotid arteries and in coronary arteries in these patients during the same period were not correlated.

Previously we have shown that ω-3 PUFA mitigate progression of atherosclerosis in coronary arteries of patients with CAD [14]. Occlusion of venous aorto–coronary bypass grafts occurred less frequently when patients were treated with ω-3 PUFA for 1 year compared to placebo [21]. Experimental studies in dog, swine and non-human primates, demonstrated a reduction of the atherosclerotic narrowing of the arterial lumen by dietary ω-3-PUFA [22]. DHA and EHA influence the atherosclerotic process by several mechanisms [23,24]. They intervene in early stages of atherogenesis, e.g., by specific downregulation of gene expression for PDGF-A, PDGF-B, and MCP-1 [25]. Epidemiological studies demonstrated an inverse correlation between fish consumption and incident non-fatal myocardial infarction [1,26,27]. Recently, this has been confirmed by a 30-year follow-up in men who were free of overt cardiovascular disease at baseline and who consumed ≥35 g fish per day [28]. In this context, our present negative finding suggests that the anti-atherogenic effect of ω-3 PUFA is specific for certain vascular beds such as the coronary circulation, or that this effect is at least materially smaller in carotid than in coronary arteries. Compared to myocardial infarction, stroke usually occurs in patients at a much older age. This indicates that the cerebral vascular bed is affected at a different pace. Moreover, the relative importance of risk factors on atherosclerosis progression seems to vary for the cerebral and the coronary circulation, e.g., there is a lower impact of hyperlipidemia on the first compared to the latter [29].

In the present study, the absence of a correlation between change of carotid artery wall thickness assessed by ultrasound and coronary lumen narrowing assessed by angiography, emphasizes that both methods examine different aspects of the same disease and that these aspects may be influenced by interventions in a specific way. These findings also suggest that the effect of any intervention on carotid IMT must be interpreted with caution.

One epidemiological study examined the relationship between intake of fish and ω-3 PUFA and the risk of stroke by stroke subtype [7]. The association of fish intake with a reduced risk of thrombotic infarction was due to a reduced risk in lacunar (small cerebral) artery but not in large (carotid) artery occlusive infarction [7]. Of note, the inverse association between ω-3 PUFA intake and stroke was not observed in the subgroup on aspirin medication [7]. In our study, more than 97% of the patients were on treatment with aspirin. Thus, if favorably altered platelet function and hemostasis explains some of the effects of ω-3 PUFA [30], treatment with aspirin which strongly affects platelet aggregation might have obscured these effects. Moreover, the beneficial impact of ω-3 PUFA on endothelial function of small arteries [31], endothelial modulation [32], arterial compliance [33], blood viscosity [34], inflammation [35], blood pressure [36], a shift in the eicosanoid system towards anti-aggregation and vasodilation [23], and other mechanisms [24] alone or in concert, may reduce risk for vascular disease morbidity and mortality. Thus, we consider the lack of an effect on carotid atherosclerosis not contradictory to the beneficial clinical effect found in other studies. Our study was not powered to detect differences in morbidity and mortality. Yet, one TIA and three TIAs in the fish oil and placebo-treated groups were observed. Possibly, a study including patients with a substantially higher risk for cerebrovascular events (or a much larger study) would be able to clarify the issue of risk reduction for stroke.

Several issues must be discussed that potentially limit our conclusions. The annual progression of carotid IMT was smaller than in another study which included patients with CAD, an IMT of ≥1.3 mm, and elevated cholesterol [37], but it was comparable to that of a more healthy population with elevated LDL [38]. This may be due to the medical treatment patients received (see Table 2), especially the use of lipid lowering drugs in half of the patients at some time during the study. This represents the standard situation for patients with high risk for coronary or cerebrovascular events. We observed an increase of blood pressure in both the fish oil and the placebo group. At the study start and during the follow-up the identical protocol for measuring blood pressure was used. Yet, whereas at the study start patients underwent heart catheterization due to an acute medical problem, the majority of them was relatively well during follow-up. As a consequence of the well-being, there is a possibility that (non-reported) compliance with medication other than the study medication deteriorated during the study. In addition, the reported use of beta-blockers decreased and the body mass index increased. Reproducibility of the carotid IMT measurements in this study is within the range of several other trials which published the respective data [39]. Since the effect of fish oil on atherosclerosis progression in coronary circulation in the main study was smaller than expected, we performed a secondary power estimation for the present study. It showed that a minimal difference of 0.02 mm in the annual progression rate between treatment groups was detectable with a power of 80%. Thus, it is possible that we missed a small effect. Yet, as the IMT progression rate in the fish oil group was even slightly higher and the global score was almost identical in both intervention groups, the risk for this type of error is small. A review of epidemiological and intervention studies suggests that 0.5 to 2.0 g ω-3 PUFA per day are effective in reducing clinical end-points [24]. By contrast, higher doses yielded no effect [40]. Thus, we would not expect a stronger effect with higher doses. Blinding may be a problem in studies using fish oil because of the intense taste of the encapsuled content. As shown previously [14], patients* guess on what they had received was not significantly different between groups, and the majority of patients (70% on fish oil, 76% on placebo in the present study) was undecided.

In conclusion, the present randomized trial aimed at exploring whether ω-3 PUFA from fish oil could beneficially influence carotid atherosclerosis. The evaluation was based on ultrasound measurements of IMT and for global scoring of the change of carotid vessel wall. In our patients with coronary artery disease, ω-3 PUFA given over 2 years in a dose shown to be effective for coronary arteries did not slow progression of atherosclerosis of the carotid arteries compared to conventional treatment. Due to the power of the study a small beneficial effect could have been missed. However, in the context of other epidemiological and randomized trials our data indicate that in patients with CAD ω-3 PUFA do not substantially slow carotid atherosclerosis.

Acknowledgments

Supported in large parts by the Bundesministerium für Forschung und Technologie, Germany, through Gesellschaft für Strahlenforschung (GSF, 07ERG03) and Deutsche Forschungsanstalt für Luft- und Raumfahrt (DLR, 01 EA 9501/7), and Wilhelm Sander-Stiftung (93.032). Other support was from Fundacion Federico and the Deutsche Forschungsgemeinschaft (Scha 398). Pronova, A.S., Lysaker, Norway provided the capsules, and funds for monitoring.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
View Abstract