Effects of fish oil and endorphins on the cold pressor test in hypertension The effects of fish oil and naloxone on blood pressure, catecholamines, and endorphins during the cold pressor test were evaluated in a randomized, double-blind, placebo-controlled, two-way crossover trial of normotensive and medication-free hypertensive men (n = 13 each). Subjects were given 5 gm w-3 fatty acids per day or placebo for 30 days with a 1-month washout between interventions. The cold pressor test (hand in ice water for 5 minutes) was done at the end of the treatment periods. Intravenous naloxone (10 mg) or placebo was given before the cold pressor test. Fish oil-treated, normotensive, or hypertensive groups had similar changes in blood pressure, plasma catecholamine levels, and P-endorphins during the cold pressor test, but naloxone treatment was associated with fivefold and tenfold increases in plasma epinephrine and cortisol levels, respectively. Naloxone may modulate sympathomedullary discharge through blockade of endorphin activity. I t is unlikely that endorphins are involved in the blood pressure increase during the cold pressor test or that fish oil alters this response. (CLINPHARMACOLTHER 1991;50:538-46.)
George S. Hughes, Jr., MD, Thomas V. Ringer, MD, Steven F. Francom, PhD, Kathy C. Caswell, RD, Michael J. DeLoof, BS, and Carol R. Spillers, RN
Fish oil has been reported to reduce blood pressure in both normotensive and hypertensive patients, but the mechanism is not fully understood. The mechanism of action of the antihypertensive effects of fish oil cannot be fully explained by alteration in the synthesis of the vasodilator prostaglandins.4-6 Fish oil may be able to reduce blood pressure by either decreasing sympathomedullary d i ~ c h a r g e reducing ,~ the sensitivity of the vasculature to catecholamines by intercalation of unsaturated fatty acids (0-3 or w-6) into membrane^,^.^ altering transmembrane cation f l ~ x e s , ~ or modulating the interaction of catecholamines and vasoactive endogenous opiates (endorphins) on the v a s ~ u l a t u r e . ~ -Reduced '~ levels of endorphins (or P-endorphin-like activity) may be involved in the pathogenesis of essential hypertension.I0-l2 Endorphins can blunt the release of catecholamines, an effect that is reversible by the opiate antagonist naloxone.13 This study was designed to investigate the role of fish oil and endorphins on blood pressure during a stressful stimulus, the cold pressor test.I4-l6 The cold pressor test acutely raises blood pressure'4-'6 through From The Upjohn Company. Received for publication May 1, 1991; accepted July 2, 1991. Reprint requests: George S. Hughes, Jr., MD, The Upjohn Company, Upjohn Research Clinics-BCIUIJasper (7216-BRN-5), 7000 Portage Rd., Kalamazoo, MI 49001. 13l1132191
enhanced sympathomedullary discharge (largely with increases in epinephrine through a,-adrenergic recep" ~ presumably elevates endorphin tor a ~ t i v a t i o n ) ~and levels (or their effects) in response to this painful stimulus. I4-l6 It is possible that the response of the patient with essential hypertension to the stress of the cold pressor test may be different from that of the normotensive subject because the vasculature of the hypertensive patient may be more sensitive to catecholamines.9.~7,~8 The cold pressor test provides an ideal situation to test the potential interrelationships between plasma levels of catecholamines and endorphins. To rigorously test the hypothesis whether fish oil or the opiate (and endorphin) antagonist naloxone can modify the physiologic responses of the hypertensive patient or normotensive to the cold pressor test,I4 we designed a randomized, doubleblind, placebo-controlled, two-way crossover trial to examine the effects of fish oil or naloxone on vital signs (blood pressure, pulse, and respiration), endorphins (plasma levels of @-endorphin and the physiologic effects of the endorphin blockade with naloxone at opiate receptors), and catecholamines (epinephrine and norepinephrine) in these healthy volunteers.
SUBJECTS AND METHODS Subjects. Thirteen normotensive subjects and 13 patients with essential hypertension participated in this
VOI.UME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 539
Table I. Schedule of events during cold pressor test Event
-45
-10
-3
0
2
4
Time (min) 5 6
8
10
20
30
60
120
-
IV inserted
IV infusion naloxone or placebo Hand in ice Vital signs Visual analog scale Blood samples
X X X X
X* X -X$ X X X X X
Xi X
X
X X X X
X
X
X X X
X
X X X
X X X
X X X
IV, Intravenous. *Time 0 refers to time in which the subject has hand in ice water (time 0 to 5 minutes) ;Heparin locks withdrawn. $Length of infusion of naloxone or placebo infus~on
study. Informed consent was obtained from each subject. The study was approved by the Protocol Review Committee of The Upjohn Company and the Bronson Methodist Hospital Human Use Committee (Kalamazoo, Mich.) and was conducted according to the 1983 Declaration of Helsinki. This was a randomized, double-blind, placebocontrolled, two-way crossover study. There was a lead-in period of 30 days in which patients discontinued all antihypertensive medications and were medication free for at least 2 weeks before the start of the study. Patients with essential hypertension had a diastolic blood pressure of at least 90 mm Hg in the supine position on three separate occasions. There was no evidence of secondary hypertension by history or physical or laboratory examinations. Subjects were randomized to receive fish oil capsules (10 Promega capsules; Parke-Davis, Morris Plains, N. J. ; 350 mg eicosapentaenoic acid and 150 mg docosahexanoic acid), 500 mg w-3 fatty acids per capsule, or 10 wheat germ oil capsules (770 mg oil per capsule; Jewel Osco, Chicago, Ill.), which we refer to as placebo capsules, for 30 days each. There was a washout period of 30 days and then crossover to the other study medication for 1 month. This amount of time has been shown to be optimal for intercalation of lipids and blood pressure response. Subjects were seen weekly in the outpatient clinic for blood pressure measurement, to assess compliance with the study medications (by pill count), and to have biweekly samples of blood drawn for safety studies (serum lipids, liver function tests, and complete blood count). Subjects were given isocaloric diets (i.e., caloric balance to maintain body weight) throughout the study consisting of the following daily intake: sodium (150 mmol), potassium (60 mmol), calcium (800 mg), and phosphorus (1000 to 1200 mg).
In-unit phase. Overnight admissions were on days 29, 30, 89, and 90 (Table I). After a 12-hour overnight fast, an intravenous catheter was inserted into a peripheral vein in each arm. Baseline blood samples for plasma epinephrine, norepinephrine, cortisol, and @-endorphin and platelets for determination of phospholipid and fatty acid composition were drawn 30 minutes later. Baseline blood pressure (systolic, diastolic, and mean) was taken by a Dinamap blood pressure monitor (Critikon Inc., Tampa, Fla.) on the right arm of the volunteer. Infusions of 10 mg naloxone (Narcan; DuPont Pharmaceuticals, Wilmington, Del.) in 50 ml of 5% dextrose in water (naloxone infusion) or 50 ml of 5% dextrose in water (dextrose infusion) were given through a catheter in a vein in the left arm during 3 minutes with an IMED 960 pump (IMED Corp., San Diego, Calif.). The naloxone or dextrose infusions were given randomly on 1 of the 2 days at the end of each phase of the study. Immediately after the infusion of either naloxone or dextrose, the subjects began the cold pressor test. Briefly, the subjects placed the left hand up to the wrist in ice water (at 4" C) for 5 minutes. A visual analog scale was used to measure pain perception. Vital signs, blood samples, and readings for visual analog scale for recording reports of pain levels were obtained as outlined in Table I. There were four possible groups from the randomization: placebo capsules/dextrose infusion, placebo capsuleslnaloxone infusion, fish oil capsulesldextrose infusion, and fish oil capsuleslnaloxone infusion; these terms will be used in the text. Plasma catecholamines. Plasma catecholamines were drawn into prechilled 5 ml glass tubes containing EGTA as an anticoagulant. The technique for separation involved separation on alumina and then analysis with HPLC with an electrochemical d e t e ~ t o r . ~ ' Plasma P-endorphin and serum cortisol. Plasma
CLIN PMARMACOI. ?HER NOVEMBER 1991
540 Hughes et al. Table 11. Characteristics of normotensive subjects and hypertensive patients Normotensives (n = 13)
Hypertensives ( n = 13)
Age (yr) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Weight (kg) Body mass index (kg/m2) Data are mean values +- SD * p < 0.01. ?p < 0.001.
P-endorphin and serum cortisol levels were measured by radioimmunoassay with kits from Incstar, Inc. (Stillwater, Minn.). Platelet phospholipids and fatty acid content. Lipid composition of platelet phospholipids was determined with separation of phospholipid classes with thin-layer chromatography and gas chromatography, as described elsewhere.22 Statistical analysis. For the major response variables discussed in this article, the change from baseline (defined as the time points before administration of the cold pressor test) response was obtained. These responses were then used to estimate, by inspection, and the time to the the maximum response (C,,,) maximum response (t,,,). The area under the response curve (AUC) was obtained by the trapezoidal rule. The values for C,, and AUC, being obtained from change scores, may be negative for situations in which there was little or no change as a result of the cold pressor test (representing random fluctuation). AUC, C,,,, and t,, were analyzed by a repeatedmeasures analysis of variance model that included the between-subject factor blood pressure group, the nested-error term subjects, the within-subject factor associated with treatment groupings, and the interaction term between blood pressure and treatment groupings (PROC, GLM; SAS, Inc., Cary, N.C.). Leastsquares means along with their associated p values were used to interpret selected treatment comparisons. Contrasts between normotensive and hypertensive subject groupings were tested with the betweensubject error term. The within-subject error term was used to test the following contrasts: placebo capsules/dextrose infusion versus fish oil capsules/naloxone infusion, placebo capsules/dextrose infusion versus fish oil capsules/dextrose infusion, and placebo capsules/naloxone infusion versus fish oil capsules/ naloxone infusion. Results are expressed in Systeme International d'Unite as mean SD. The level of significance was p < 0.05.
*
RESULTS Demographics. Clinical characteristics of the volunteers are summarized in Table 11. Blood pressure at entry to the study was greater in patients with hypertension than in normotensive subjects. The hypertensive group was slightly older than the normotensive group ( p < 0.01), but weight and body mass index were similar in both groups. Vital signs. The response of blood pressure in the normotensive subjects and hypertensive patients is shown in Table 111. There were only a few significant differences in AUC, C,,,, or t,, of the blood pressure response to the cold pressor test between normotensive subjects or hypertensive patients given either placebo or fish oil capsules or the naloxone or dextrose infusions. These differences were statistically significant for systolic blood pressure for fish oilldextrose infusion and fish oil/naloxone infusion for C,, (Table 111). Systolic and diastolic blood pressure returned to baseline within 1 minute after the test was completed. There were no changes in pulse or respiration during the cold pressor test or outpatient blood pressure noted during the study (data not shown). Plasma catecholamines. Mean AUC and C,, values for plasma epinephrine and norepinephrine levels increased in both the normotensive and hypertensive groups in response to the cold pressor test (Table IV; Figs. 1 and 2). The mean values for plasma epinephrine were consistently higher for the naloxone-treated groups compared with those given dextrose infusion. There were no statistically significant differences in the magnitude of response of epinephrine or norepinephrine in either the normotensive or hypertensive groups given fish oil or placebo capsules or the naloxone or dextrose infusions. Plasma cortisol. Mean plasma AUC, C,,,, and t,, values for cortisol in both normotensive and hypertensive groups are shown in Table V. Mean AUC and C,, values were greatest in the naloxone-treated groups regardless of whether the normotensive or hy-
VOLUME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 54 1
Table 111. Systolic and diastolic blood pressure (change from baseline) in response to cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose and naloxone infusions PiD
PIN
FOID
FOIN
p Value*
46.0 t 18.3 96.4 t 16.2 0.06
67.3 t 15.0 63.7 t 17.1 0.89
67.5 t 14.6 104.8 2 15.4 0.13
53.5 2 21.6 89.3 2 32.9 0.18
0.45 0.25
17.7 t 2.1 26.7 + 2.7 0.06
18.8 ? 2.3 22.2 t 3.1 0.89
18.3 t 2.2 28.8 2 3.0 0.03
17.7 t 2.2 27.6 t 3.5 0.04
0.75 0.20
2.8 t 0.4 2.8 t 0.4 0.90
3.2 ? 0.3 1.8 t 0.2 0.04
2.8 t 0.3 2.7 2 0.4 0.93
2.8 t 0.4 2.8 t 0.4 0.90
0.38 0.04$
54.2 t 10.9 44.3 t 12.0 0.70
45.7 t 13.5 21.7 ? 13.5 0.34
39.7 t 15.4 44.3 t 6.6 0.76
30.5 2 16.0 14.0 t 21.4 0.51
0.40 0.08
15.5 ? 1.9 14.8 ? 1.6 0.87
14.3 2 2.3 14.4 2 2.5 0.98
12.6 t 1.9 15.9 t 1.6 0.36
12.5 t 2.0 13.1 2 2.1 0.88
0.23 0.20
3.2 2 0.3 2.3 t 0.2 0.08
2.9 2 0.4 2.0 t 0.2 0.06
2.5 t 0.4 2.0 t 0.0 0.31
2.2 ? 0.3 2.8 ? 0.4 0.20
0.07 0.07
Systolic blood pressure
AUC (mm Hglmin) Normotensives Hypertensives p Value? C ,, (mm Hg) Normotensives Hypertensives p Valuet Laxb i n ) Normotensives Hypertensives p Value? Diastolic blood pressure
AUC (mm Hglmin) Normotensives Hypertensives p Value? C,, (mm Hg) Normotensives Hypertensives r, Value? tmax(mid Normotensives Hypertensives p Value?
Data are mean values 2 SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertensives. $PIN versus PID and FOIN.
pertensive groups received fish oil or placebo capsules. The tmax generally occurred later in the naloxonetreated groups than in the dextrose-treated groups (Table V). Plasma endorphins. Mean plasma P-endorphin values for AUC, Cmax, and t,, were not significantly different between normotensive and hypertensive groups (Table VI). The tmaXvalues generally peaked by 10 to 17 minutes in each group and returned to baseline by 30 minutes after the cold pressor test was begun. Peak plasma P-endorphin (Cmax)levels were generally higher in all naloxone-treated periods regardless of blood pressure (normotensive or hypertensive) or treatment (i.e., fish oil or placebo capsules). Platelet phospholipids and fatty acid content. Levels of w-3 fatty acids (eicosapentaenoic acid) increased (up to fiftyfold) significantly @ < 0.0001) in the lecithin and phosphatidylethanolamine fractions in
both the normotensive and hypertensive groups. The results are published el~ewhere.'~ Pain measurements. The mean values of AUC, Cmax,and tmaxpain scores across time for each treatment intervention(s) for normotensive and hypertensive groups are shown in Table VII. AUC and tmax were generally similar in the normotensive and hypertensive groups. Pain values reached a similar maximum intensity (between 2 and 4 minutes) in both normotensive subjects and hypertensive patients during the cold pressor test, regardless of whether fish oil or placebo capsules had been taken.
DISCUSSION Our results indicate that the physiologic response of normotensive and hypertensive men to the cold pressor test was similar to that of other ~ t u d i e s . ' ~ .The '~ magnitude of change in blood pressure with concomi-
CLIN PHAhMACOI. THER NOVEMRER 1991
542 Hughes et al.
Table IV. Plasma epinephrine and norepinephrine levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil o r placebo capsules o r dextrose or naloxone infusion PID
PIN
FOID
FOIN
p Value*
Epinephrine
AUC (nmollLlmin) Normotensives Hypertensives p Values C,, (nmollL) Normotensives Hypertensives p Value§ Laxb i n Norrnotensives Hypertensives p Values Norepinephrine
AUC (nmollLimin) Norrnotensives Hypertensives p Values C,,, (nmol1L) Normotensives Hypertensives p Values Laxb i n ) Norrnotensives Hypertensives p Value§ Data are mean values I SD. P. Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tPlN versus PID and FOIN. SPID versus PIN. PNormotensives versus hypertensives. I PID versus FOID.
Table V. Serum cortisol levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose o r naloxone infusions PID AUC (nmollLlmin) Normotensives Hypertensives p Values C,, (nmollL) Normotensives Hypertensives p Values tmax(min) Normotensives Hypertensives p Values Data are mean values ? SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone. *Treatment comparisons. +PIN versus PID and FOIN. $PID versus PIN. INormotensives versus hypertensives.
PIN
FOID
FOIN
p Value*
VOLUME SO NUMBER 5, PART 1
Naloxone and the cold pressor test 543
TIME (MINUTES) Fig. 1. Plasma epinephrine levels during cold pressor test in hypertensive patients: placebo/naloxone (0) or placeboldextrose (m), fish oillnaloxone (A), and fish oilldextrose (A); in normotensive fish oilldextrose ( J r ) , or fish oillnaloxsubjects: placeboldextrose (a),placebolnaloxone (O), one (b).
tant elevation of plasma catecholamine levels also confirms other observation^.^^'^^'^ In addition, naloxone exacerbated the rise in plasma epinephrine levels but did not substantially affect vital signs, plasma P-endorphin and norepinephrine levels, or pain scores. Naloxone has been shown to cause disinhibition of P-endorphins at the opiate-receptor level to al-
ter sympathomedullary discharge."I2 Naloxone has been used as a pharmacologic tool to investigate the role of endorphins in blood pressure regulation in patients with essential hypertension, but with mixed res u l t ~ . ' ~ - 'In * some cases the putative P-endorphinmediated reduction of blood pressure with clonidine is blocked by naloxone.I0 However, infusion of nalox-
CLIN PHARMACOL THER NOVEMBER 1991
544 Hughes et al.
TIME (MINUTES) Fig. 2. Norepinephrine levels during cold pressor test in hypertensive patients: placebolnaloxone (0) or placeboldextrose (m), fish oillnaloxone (A), and fish oilldextrose (A); in normotensive subjects: placebo/dextrose (a),placebolnaloxone (O), fish oilldextrose (*), or fish oillnaloxone (6).
one in suprapharmacologic doses in other studies has not been shown to substantially affect blood pressure in normal subjects or hypertensive patients.l4,I9 The rapid increase in catecholamines (with only small changes in pulse or respiration, especially epinephrine after naloxone) during the cold pressor test has been documented pre~iously.'~-"However, the actual measurements of circulating P-endorphin levels during this test have not been reported. The response
of the P-endorphin/cortisol axis is usually slower (peak response by 30 to 60 minutes)"12 than the catecholamine elevation (within minutes). It should be noted that although endorphins were elevated in response to the cold pressor test (and even more so with naloxone), fish oil did not modify this response, despite substantially increased levels of 0-3 fatty acids in platelet membranes. There was no effect of fish oil on resting supine
VOLUME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 545
Table VI. Plasma P-endorphin levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose or naloxone infusions PID
PIN
FOID
FOiN
p Value*
AUC (prnol/L/rnin) Norrnotensives Hypertensives p Valuet C,, (pmolIL) Norrnotensives Hypertensives p Valuet Lax( m i 4 Normotensives Hypertensives p Valuet Data are mean values + SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertensives.
Table VII. Pain scores in response to cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose and naloxone infusions PID
PIN
FOID
FOIN
p Value*
AUC (pain unitsimin) Norrnotensives Hypertensives p Valuet C,, (pain units) Norrnotensives Hypertensives p Valuet tmvxb i n ) Norrnotensives Hypertensives p Valuet
*
Data are mean values SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertenswes. $PIN versus PID and FOIN.
blood pressure or the physiologic response to the cold pressor test. Other studies have lauded the beneficial effects of fish oil (and presumably the 0-3 fatty acids) on numerous cardiovascular parameters, whereas more recent studies have failed to show the benefits of these substances at the doses used in this study or at higher doses. In summary, fish oil failed to ameliorate the physiologic increase in blood pressure or other biochemical parameters (e.g., plasma catecholamines or Pendorphins) in response to the cold pressor test in normotensive and hypertensive men. Naloxone was used as a pharmacologic probe to investigate the possible
role of endorphins on the increase in blood pressure during the stress of the cold pressor test. The magnitude of blood pressure increase was the same in both normotensive and hypertensive groups (despite treatment with fish oil or naloxone). From our results, it is unlikely that endorphins are involved in the blood pressure increase during the cold pressor test or that fish oil can modify this physiologic response. We thank the nurses and staff of the Upjohn Research Clinics-Bronson Clinical Investigation UnitIJasper for their expert technical assistance.
CLIN 1'HARMACOL THEK NOVEMBER 1991
546 Htghes et al. References 1. Von Shacky C. Prophylaxis of atherosclerosis with marine omega-3 fatty acids. Ann Intern Med 1987; 107:890-9. 2. Yetiv JZ. Clinical applications of fish oils. JAMA 1988;260:665-70. 3. Leaf A, Weber PC. Cardiovascular effects of n-3 fatty acids. N Engl J Med 1988;318:549-57. 4. Knapp HR, Fitzgerald GA. The antihypertensive effects of fish oil. N Engl J Med 1989;320:1037-43. 5. Lorenz R, Spengler U, Fischer S, Duhm J, Weber PC. Platelet function, thromboxane formation and blood pressure control during supplementation of the western diet with cod liver oil. Circulation 1983;67:504-11. 6. Oates JA, Fitzgerald GA, Branch R, et al. Clinical implications of prostaglandin and thromboxane A, formation (second of two parts). N Engl J Med 1988;319:761-7. 7. De Jong W, Petty MA, Sitsen JM. Role of opioid peptides in brain mechanisms regulating blood pressure. Chest 1983;83(suppl):306-8. 8. Grossman A. Opioids and stress in man. J Endocrinol 1988;119:377-81. 9. Buhler FR, Bolli P, Hulthen UL, Amann FW, Kliowski W. Alpha-adrenoreceptors, adrenaline, and exaggerated vasoconstrictor response to stress in essential hypertension. Chest 1983;83(suppl):304-5. 10. Kraft K, Theobald R, Kolloch R, Stumpe KO. Normalization of blood pressure and plasma concentrations of beta-endorphin and leucine-enkephalin in patients with primary hypertension after treatment with clonidine. J Cardiovasc Pharmacol 1987;lO(supp1):S147-51. 11. Giles TD, Sander GE, Rice JC, Quiroz AC. Systemic methionine-enkephalin evokes cardiostimulatory responses in the human. Peptides 1987;8:609-12. 12. Holaday JW. Cardiovascular effects of endogenous opiate systems. Annu Rev Pharmacol Toxic01 1983; 23541.4. 13. Hughes GS. Naloxone and methylprednisolone sodium
14.
15.
16.
17.
18.
19.
20. 21.
22.
23.
succinate enhance sympathomedullary discharge in patients with septic shock. Life Sci 1984;35:2319-26. Bouloux P, Grossman A, Al-Damluji S, Bailey T, Besser M. Enhancement of sympathoadrenal response to the cold pressor test by naloxone in man. Clin Sci 1985;69:365-8. Robertson D, Johnson GA, Robertson RM, Nies AS, Shand DG, Oates JA. Comparative assessment of stimuli that release neuronal and adrenomedullary catecholamines in man. Circulation 1979;59:637-43. Musgrave IF, Bachmann AW, Saar N, Gordon RD. A comparison of cardiovascular and catecholamine responses to three stimuli in mild hypertension. Metabolism l984;33:7 18-23, Philipp T, Distler A, Cordes U. Sympathetic nervous system and blood-pressure control in essential hypertension. Lancet 1978;i:959-63. De Champlain J, Cousineau D, Lapointe L, Lavallee M, Nadeau R, Denis G. Sympathetic abnormalities in human hypertension. Clin Exp Hypertens 1981; 3:417-38. Fuenmayer N, Cubeddu L. Cardiovascular and endocrine effects of naloxone compared in normotensive and hypertensive patients. Eur J Clin Pharmacol 1986; 126:189-97. Rubin PC. Opioid peptides in blood pressure regulation in man. Clin Sci 1984;66:625-30. De Lavelle E. Mayo Clinic Laboratories interpretive handbook. Rochester and Minneapolis, Minnesota: Mayo Clinic Laboratories, 1990. Nagawa Y, Orimo H, Harasawa M, Morita I, Yashiro K, Nurota S. Effect of eicosapentaenoic acid on the platelet aggregation and composition of fatty acid in man. Atherosclerosis 1983;47:7 1-5. Hughes GS, Ringer TV, Watts KC, DeLoof MJ, Francom SF, Spillers CF. Fish oil produces an atherogenic lipid profile in hypertensive men. Atherosclerosis 1990;84:229-37.
George S. Hughes, Jr., MD, Thomas V. Ringer, MD, Steven F. Francom, PhD, Kathy C. Caswell, RD, Michael J. DeLoof, BS, and Carol R. Spillers, RN
Fish oil has been reported to reduce blood pressure in both normotensive and hypertensive patients, but the mechanism is not fully understood. The mechanism of action of the antihypertensive effects of fish oil cannot be fully explained by alteration in the synthesis of the vasodilator prostaglandins.4-6 Fish oil may be able to reduce blood pressure by either decreasing sympathomedullary d i ~ c h a r g e reducing ,~ the sensitivity of the vasculature to catecholamines by intercalation of unsaturated fatty acids (0-3 or w-6) into membrane^,^.^ altering transmembrane cation f l ~ x e s , ~ or modulating the interaction of catecholamines and vasoactive endogenous opiates (endorphins) on the v a s ~ u l a t u r e . ~ -Reduced '~ levels of endorphins (or P-endorphin-like activity) may be involved in the pathogenesis of essential hypertension.I0-l2 Endorphins can blunt the release of catecholamines, an effect that is reversible by the opiate antagonist naloxone.13 This study was designed to investigate the role of fish oil and endorphins on blood pressure during a stressful stimulus, the cold pressor test.I4-l6 The cold pressor test acutely raises blood pressure'4-'6 through From The Upjohn Company. Received for publication May 1, 1991; accepted July 2, 1991. Reprint requests: George S. Hughes, Jr., MD, The Upjohn Company, Upjohn Research Clinics-BCIUIJasper (7216-BRN-5), 7000 Portage Rd., Kalamazoo, MI 49001. 13l1132191
enhanced sympathomedullary discharge (largely with increases in epinephrine through a,-adrenergic recep" ~ presumably elevates endorphin tor a ~ t i v a t i o n ) ~and levels (or their effects) in response to this painful stimulus. I4-l6 It is possible that the response of the patient with essential hypertension to the stress of the cold pressor test may be different from that of the normotensive subject because the vasculature of the hypertensive patient may be more sensitive to catecholamines.9.~7,~8 The cold pressor test provides an ideal situation to test the potential interrelationships between plasma levels of catecholamines and endorphins. To rigorously test the hypothesis whether fish oil or the opiate (and endorphin) antagonist naloxone can modify the physiologic responses of the hypertensive patient or normotensive to the cold pressor test,I4 we designed a randomized, doubleblind, placebo-controlled, two-way crossover trial to examine the effects of fish oil or naloxone on vital signs (blood pressure, pulse, and respiration), endorphins (plasma levels of @-endorphin and the physiologic effects of the endorphin blockade with naloxone at opiate receptors), and catecholamines (epinephrine and norepinephrine) in these healthy volunteers.
SUBJECTS AND METHODS Subjects. Thirteen normotensive subjects and 13 patients with essential hypertension participated in this
VOI.UME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 539
Table I. Schedule of events during cold pressor test Event
-45
-10
-3
0
2
4
Time (min) 5 6
8
10
20
30
60
120
-
IV inserted
IV infusion naloxone or placebo Hand in ice Vital signs Visual analog scale Blood samples
X X X X
X* X -X$ X X X X X
Xi X
X
X X X X
X
X
X X X
X
X X X
X X X
X X X
IV, Intravenous. *Time 0 refers to time in which the subject has hand in ice water (time 0 to 5 minutes) ;Heparin locks withdrawn. $Length of infusion of naloxone or placebo infus~on
study. Informed consent was obtained from each subject. The study was approved by the Protocol Review Committee of The Upjohn Company and the Bronson Methodist Hospital Human Use Committee (Kalamazoo, Mich.) and was conducted according to the 1983 Declaration of Helsinki. This was a randomized, double-blind, placebocontrolled, two-way crossover study. There was a lead-in period of 30 days in which patients discontinued all antihypertensive medications and were medication free for at least 2 weeks before the start of the study. Patients with essential hypertension had a diastolic blood pressure of at least 90 mm Hg in the supine position on three separate occasions. There was no evidence of secondary hypertension by history or physical or laboratory examinations. Subjects were randomized to receive fish oil capsules (10 Promega capsules; Parke-Davis, Morris Plains, N. J. ; 350 mg eicosapentaenoic acid and 150 mg docosahexanoic acid), 500 mg w-3 fatty acids per capsule, or 10 wheat germ oil capsules (770 mg oil per capsule; Jewel Osco, Chicago, Ill.), which we refer to as placebo capsules, for 30 days each. There was a washout period of 30 days and then crossover to the other study medication for 1 month. This amount of time has been shown to be optimal for intercalation of lipids and blood pressure response. Subjects were seen weekly in the outpatient clinic for blood pressure measurement, to assess compliance with the study medications (by pill count), and to have biweekly samples of blood drawn for safety studies (serum lipids, liver function tests, and complete blood count). Subjects were given isocaloric diets (i.e., caloric balance to maintain body weight) throughout the study consisting of the following daily intake: sodium (150 mmol), potassium (60 mmol), calcium (800 mg), and phosphorus (1000 to 1200 mg).
In-unit phase. Overnight admissions were on days 29, 30, 89, and 90 (Table I). After a 12-hour overnight fast, an intravenous catheter was inserted into a peripheral vein in each arm. Baseline blood samples for plasma epinephrine, norepinephrine, cortisol, and @-endorphin and platelets for determination of phospholipid and fatty acid composition were drawn 30 minutes later. Baseline blood pressure (systolic, diastolic, and mean) was taken by a Dinamap blood pressure monitor (Critikon Inc., Tampa, Fla.) on the right arm of the volunteer. Infusions of 10 mg naloxone (Narcan; DuPont Pharmaceuticals, Wilmington, Del.) in 50 ml of 5% dextrose in water (naloxone infusion) or 50 ml of 5% dextrose in water (dextrose infusion) were given through a catheter in a vein in the left arm during 3 minutes with an IMED 960 pump (IMED Corp., San Diego, Calif.). The naloxone or dextrose infusions were given randomly on 1 of the 2 days at the end of each phase of the study. Immediately after the infusion of either naloxone or dextrose, the subjects began the cold pressor test. Briefly, the subjects placed the left hand up to the wrist in ice water (at 4" C) for 5 minutes. A visual analog scale was used to measure pain perception. Vital signs, blood samples, and readings for visual analog scale for recording reports of pain levels were obtained as outlined in Table I. There were four possible groups from the randomization: placebo capsules/dextrose infusion, placebo capsuleslnaloxone infusion, fish oil capsulesldextrose infusion, and fish oil capsuleslnaloxone infusion; these terms will be used in the text. Plasma catecholamines. Plasma catecholamines were drawn into prechilled 5 ml glass tubes containing EGTA as an anticoagulant. The technique for separation involved separation on alumina and then analysis with HPLC with an electrochemical d e t e ~ t o r . ~ ' Plasma P-endorphin and serum cortisol. Plasma
CLIN PMARMACOI. ?HER NOVEMBER 1991
540 Hughes et al. Table 11. Characteristics of normotensive subjects and hypertensive patients Normotensives (n = 13)
Hypertensives ( n = 13)
Age (yr) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Weight (kg) Body mass index (kg/m2) Data are mean values +- SD * p < 0.01. ?p < 0.001.
P-endorphin and serum cortisol levels were measured by radioimmunoassay with kits from Incstar, Inc. (Stillwater, Minn.). Platelet phospholipids and fatty acid content. Lipid composition of platelet phospholipids was determined with separation of phospholipid classes with thin-layer chromatography and gas chromatography, as described elsewhere.22 Statistical analysis. For the major response variables discussed in this article, the change from baseline (defined as the time points before administration of the cold pressor test) response was obtained. These responses were then used to estimate, by inspection, and the time to the the maximum response (C,,,) maximum response (t,,,). The area under the response curve (AUC) was obtained by the trapezoidal rule. The values for C,, and AUC, being obtained from change scores, may be negative for situations in which there was little or no change as a result of the cold pressor test (representing random fluctuation). AUC, C,,,, and t,, were analyzed by a repeatedmeasures analysis of variance model that included the between-subject factor blood pressure group, the nested-error term subjects, the within-subject factor associated with treatment groupings, and the interaction term between blood pressure and treatment groupings (PROC, GLM; SAS, Inc., Cary, N.C.). Leastsquares means along with their associated p values were used to interpret selected treatment comparisons. Contrasts between normotensive and hypertensive subject groupings were tested with the betweensubject error term. The within-subject error term was used to test the following contrasts: placebo capsules/dextrose infusion versus fish oil capsules/naloxone infusion, placebo capsules/dextrose infusion versus fish oil capsules/dextrose infusion, and placebo capsules/naloxone infusion versus fish oil capsules/ naloxone infusion. Results are expressed in Systeme International d'Unite as mean SD. The level of significance was p < 0.05.
*
RESULTS Demographics. Clinical characteristics of the volunteers are summarized in Table 11. Blood pressure at entry to the study was greater in patients with hypertension than in normotensive subjects. The hypertensive group was slightly older than the normotensive group ( p < 0.01), but weight and body mass index were similar in both groups. Vital signs. The response of blood pressure in the normotensive subjects and hypertensive patients is shown in Table 111. There were only a few significant differences in AUC, C,,,, or t,, of the blood pressure response to the cold pressor test between normotensive subjects or hypertensive patients given either placebo or fish oil capsules or the naloxone or dextrose infusions. These differences were statistically significant for systolic blood pressure for fish oilldextrose infusion and fish oil/naloxone infusion for C,, (Table 111). Systolic and diastolic blood pressure returned to baseline within 1 minute after the test was completed. There were no changes in pulse or respiration during the cold pressor test or outpatient blood pressure noted during the study (data not shown). Plasma catecholamines. Mean AUC and C,, values for plasma epinephrine and norepinephrine levels increased in both the normotensive and hypertensive groups in response to the cold pressor test (Table IV; Figs. 1 and 2). The mean values for plasma epinephrine were consistently higher for the naloxone-treated groups compared with those given dextrose infusion. There were no statistically significant differences in the magnitude of response of epinephrine or norepinephrine in either the normotensive or hypertensive groups given fish oil or placebo capsules or the naloxone or dextrose infusions. Plasma cortisol. Mean plasma AUC, C,,,, and t,, values for cortisol in both normotensive and hypertensive groups are shown in Table V. Mean AUC and C,, values were greatest in the naloxone-treated groups regardless of whether the normotensive or hy-
VOLUME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 54 1
Table 111. Systolic and diastolic blood pressure (change from baseline) in response to cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose and naloxone infusions PiD
PIN
FOID
FOIN
p Value*
46.0 t 18.3 96.4 t 16.2 0.06
67.3 t 15.0 63.7 t 17.1 0.89
67.5 t 14.6 104.8 2 15.4 0.13
53.5 2 21.6 89.3 2 32.9 0.18
0.45 0.25
17.7 t 2.1 26.7 + 2.7 0.06
18.8 ? 2.3 22.2 t 3.1 0.89
18.3 t 2.2 28.8 2 3.0 0.03
17.7 t 2.2 27.6 t 3.5 0.04
0.75 0.20
2.8 t 0.4 2.8 t 0.4 0.90
3.2 ? 0.3 1.8 t 0.2 0.04
2.8 t 0.3 2.7 2 0.4 0.93
2.8 t 0.4 2.8 t 0.4 0.90
0.38 0.04$
54.2 t 10.9 44.3 t 12.0 0.70
45.7 t 13.5 21.7 ? 13.5 0.34
39.7 t 15.4 44.3 t 6.6 0.76
30.5 2 16.0 14.0 t 21.4 0.51
0.40 0.08
15.5 ? 1.9 14.8 ? 1.6 0.87
14.3 2 2.3 14.4 2 2.5 0.98
12.6 t 1.9 15.9 t 1.6 0.36
12.5 t 2.0 13.1 2 2.1 0.88
0.23 0.20
3.2 2 0.3 2.3 t 0.2 0.08
2.9 2 0.4 2.0 t 0.2 0.06
2.5 t 0.4 2.0 t 0.0 0.31
2.2 ? 0.3 2.8 ? 0.4 0.20
0.07 0.07
Systolic blood pressure
AUC (mm Hglmin) Normotensives Hypertensives p Value? C ,, (mm Hg) Normotensives Hypertensives p Valuet Laxb i n ) Normotensives Hypertensives p Value? Diastolic blood pressure
AUC (mm Hglmin) Normotensives Hypertensives p Value? C,, (mm Hg) Normotensives Hypertensives r, Value? tmax(mid Normotensives Hypertensives p Value?
Data are mean values 2 SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertensives. $PIN versus PID and FOIN.
pertensive groups received fish oil or placebo capsules. The tmax generally occurred later in the naloxonetreated groups than in the dextrose-treated groups (Table V). Plasma endorphins. Mean plasma P-endorphin values for AUC, Cmax, and t,, were not significantly different between normotensive and hypertensive groups (Table VI). The tmaXvalues generally peaked by 10 to 17 minutes in each group and returned to baseline by 30 minutes after the cold pressor test was begun. Peak plasma P-endorphin (Cmax)levels were generally higher in all naloxone-treated periods regardless of blood pressure (normotensive or hypertensive) or treatment (i.e., fish oil or placebo capsules). Platelet phospholipids and fatty acid content. Levels of w-3 fatty acids (eicosapentaenoic acid) increased (up to fiftyfold) significantly @ < 0.0001) in the lecithin and phosphatidylethanolamine fractions in
both the normotensive and hypertensive groups. The results are published el~ewhere.'~ Pain measurements. The mean values of AUC, Cmax,and tmaxpain scores across time for each treatment intervention(s) for normotensive and hypertensive groups are shown in Table VII. AUC and tmax were generally similar in the normotensive and hypertensive groups. Pain values reached a similar maximum intensity (between 2 and 4 minutes) in both normotensive subjects and hypertensive patients during the cold pressor test, regardless of whether fish oil or placebo capsules had been taken.
DISCUSSION Our results indicate that the physiologic response of normotensive and hypertensive men to the cold pressor test was similar to that of other ~ t u d i e s . ' ~ .The '~ magnitude of change in blood pressure with concomi-
CLIN PHAhMACOI. THER NOVEMRER 1991
542 Hughes et al.
Table IV. Plasma epinephrine and norepinephrine levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil o r placebo capsules o r dextrose or naloxone infusion PID
PIN
FOID
FOIN
p Value*
Epinephrine
AUC (nmollLlmin) Normotensives Hypertensives p Values C,, (nmollL) Normotensives Hypertensives p Value§ Laxb i n Norrnotensives Hypertensives p Values Norepinephrine
AUC (nmollLimin) Norrnotensives Hypertensives p Values C,,, (nmol1L) Normotensives Hypertensives p Values Laxb i n ) Norrnotensives Hypertensives p Value§ Data are mean values I SD. P. Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tPlN versus PID and FOIN. SPID versus PIN. PNormotensives versus hypertensives. I PID versus FOID.
Table V. Serum cortisol levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose o r naloxone infusions PID AUC (nmollLlmin) Normotensives Hypertensives p Values C,, (nmollL) Normotensives Hypertensives p Values tmax(min) Normotensives Hypertensives p Values Data are mean values ? SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone. *Treatment comparisons. +PIN versus PID and FOIN. $PID versus PIN. INormotensives versus hypertensives.
PIN
FOID
FOIN
p Value*
VOLUME SO NUMBER 5, PART 1
Naloxone and the cold pressor test 543
TIME (MINUTES) Fig. 1. Plasma epinephrine levels during cold pressor test in hypertensive patients: placebo/naloxone (0) or placeboldextrose (m), fish oillnaloxone (A), and fish oilldextrose (A); in normotensive fish oilldextrose ( J r ) , or fish oillnaloxsubjects: placeboldextrose (a),placebolnaloxone (O), one (b).
tant elevation of plasma catecholamine levels also confirms other observation^.^^'^^'^ In addition, naloxone exacerbated the rise in plasma epinephrine levels but did not substantially affect vital signs, plasma P-endorphin and norepinephrine levels, or pain scores. Naloxone has been shown to cause disinhibition of P-endorphins at the opiate-receptor level to al-
ter sympathomedullary discharge."I2 Naloxone has been used as a pharmacologic tool to investigate the role of endorphins in blood pressure regulation in patients with essential hypertension, but with mixed res u l t ~ . ' ~ - 'In * some cases the putative P-endorphinmediated reduction of blood pressure with clonidine is blocked by naloxone.I0 However, infusion of nalox-
CLIN PHARMACOL THER NOVEMBER 1991
544 Hughes et al.
TIME (MINUTES) Fig. 2. Norepinephrine levels during cold pressor test in hypertensive patients: placebolnaloxone (0) or placeboldextrose (m), fish oillnaloxone (A), and fish oilldextrose (A); in normotensive subjects: placebo/dextrose (a),placebolnaloxone (O), fish oilldextrose (*), or fish oillnaloxone (6).
one in suprapharmacologic doses in other studies has not been shown to substantially affect blood pressure in normal subjects or hypertensive patients.l4,I9 The rapid increase in catecholamines (with only small changes in pulse or respiration, especially epinephrine after naloxone) during the cold pressor test has been documented pre~iously.'~-"However, the actual measurements of circulating P-endorphin levels during this test have not been reported. The response
of the P-endorphin/cortisol axis is usually slower (peak response by 30 to 60 minutes)"12 than the catecholamine elevation (within minutes). It should be noted that although endorphins were elevated in response to the cold pressor test (and even more so with naloxone), fish oil did not modify this response, despite substantially increased levels of 0-3 fatty acids in platelet membranes. There was no effect of fish oil on resting supine
VOLUME 50 NUMBER 5, PART 1
Naloxone and the cold pressor test 545
Table VI. Plasma P-endorphin levels in response to the cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose or naloxone infusions PID
PIN
FOID
FOiN
p Value*
AUC (prnol/L/rnin) Norrnotensives Hypertensives p Valuet C,, (pmolIL) Norrnotensives Hypertensives p Valuet Lax( m i 4 Normotensives Hypertensives p Valuet Data are mean values + SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertensives.
Table VII. Pain scores in response to cold pressor test in normotensive subjects and patients with hypertension given oral fish oil or placebo capsules or dextrose and naloxone infusions PID
PIN
FOID
FOIN
p Value*
AUC (pain unitsimin) Norrnotensives Hypertensives p Valuet C,, (pain units) Norrnotensives Hypertensives p Valuet tmvxb i n ) Norrnotensives Hypertensives p Valuet
*
Data are mean values SD. P, Placebo; D, dextrose; FO, fish oil; N, naloxone *Treatment comparisons. tNormotensives versus hypertenswes. $PIN versus PID and FOIN.
blood pressure or the physiologic response to the cold pressor test. Other studies have lauded the beneficial effects of fish oil (and presumably the 0-3 fatty acids) on numerous cardiovascular parameters, whereas more recent studies have failed to show the benefits of these substances at the doses used in this study or at higher doses. In summary, fish oil failed to ameliorate the physiologic increase in blood pressure or other biochemical parameters (e.g., plasma catecholamines or Pendorphins) in response to the cold pressor test in normotensive and hypertensive men. Naloxone was used as a pharmacologic probe to investigate the possible
role of endorphins on the increase in blood pressure during the stress of the cold pressor test. The magnitude of blood pressure increase was the same in both normotensive and hypertensive groups (despite treatment with fish oil or naloxone). From our results, it is unlikely that endorphins are involved in the blood pressure increase during the cold pressor test or that fish oil can modify this physiologic response. We thank the nurses and staff of the Upjohn Research Clinics-Bronson Clinical Investigation UnitIJasper for their expert technical assistance.
CLIN 1'HARMACOL THEK NOVEMBER 1991
546 Htghes et al. References 1. Von Shacky C. Prophylaxis of atherosclerosis with marine omega-3 fatty acids. Ann Intern Med 1987; 107:890-9. 2. Yetiv JZ. Clinical applications of fish oils. JAMA 1988;260:665-70. 3. Leaf A, Weber PC. Cardiovascular effects of n-3 fatty acids. N Engl J Med 1988;318:549-57. 4. Knapp HR, Fitzgerald GA. The antihypertensive effects of fish oil. N Engl J Med 1989;320:1037-43. 5. Lorenz R, Spengler U, Fischer S, Duhm J, Weber PC. Platelet function, thromboxane formation and blood pressure control during supplementation of the western diet with cod liver oil. Circulation 1983;67:504-11. 6. Oates JA, Fitzgerald GA, Branch R, et al. Clinical implications of prostaglandin and thromboxane A, formation (second of two parts). N Engl J Med 1988;319:761-7. 7. De Jong W, Petty MA, Sitsen JM. Role of opioid peptides in brain mechanisms regulating blood pressure. Chest 1983;83(suppl):306-8. 8. Grossman A. Opioids and stress in man. J Endocrinol 1988;119:377-81. 9. Buhler FR, Bolli P, Hulthen UL, Amann FW, Kliowski W. Alpha-adrenoreceptors, adrenaline, and exaggerated vasoconstrictor response to stress in essential hypertension. Chest 1983;83(suppl):304-5. 10. Kraft K, Theobald R, Kolloch R, Stumpe KO. Normalization of blood pressure and plasma concentrations of beta-endorphin and leucine-enkephalin in patients with primary hypertension after treatment with clonidine. J Cardiovasc Pharmacol 1987;lO(supp1):S147-51. 11. Giles TD, Sander GE, Rice JC, Quiroz AC. Systemic methionine-enkephalin evokes cardiostimulatory responses in the human. Peptides 1987;8:609-12. 12. Holaday JW. Cardiovascular effects of endogenous opiate systems. Annu Rev Pharmacol Toxic01 1983; 23541.4. 13. Hughes GS. Naloxone and methylprednisolone sodium
14.
15.
16.
17.
18.
19.
20. 21.
22.
23.
succinate enhance sympathomedullary discharge in patients with septic shock. Life Sci 1984;35:2319-26. Bouloux P, Grossman A, Al-Damluji S, Bailey T, Besser M. Enhancement of sympathoadrenal response to the cold pressor test by naloxone in man. Clin Sci 1985;69:365-8. Robertson D, Johnson GA, Robertson RM, Nies AS, Shand DG, Oates JA. Comparative assessment of stimuli that release neuronal and adrenomedullary catecholamines in man. Circulation 1979;59:637-43. Musgrave IF, Bachmann AW, Saar N, Gordon RD. A comparison of cardiovascular and catecholamine responses to three stimuli in mild hypertension. Metabolism l984;33:7 18-23, Philipp T, Distler A, Cordes U. Sympathetic nervous system and blood-pressure control in essential hypertension. Lancet 1978;i:959-63. De Champlain J, Cousineau D, Lapointe L, Lavallee M, Nadeau R, Denis G. Sympathetic abnormalities in human hypertension. Clin Exp Hypertens 1981; 3:417-38. Fuenmayer N, Cubeddu L. Cardiovascular and endocrine effects of naloxone compared in normotensive and hypertensive patients. Eur J Clin Pharmacol 1986; 126:189-97. Rubin PC. Opioid peptides in blood pressure regulation in man. Clin Sci 1984;66:625-30. De Lavelle E. Mayo Clinic Laboratories interpretive handbook. Rochester and Minneapolis, Minnesota: Mayo Clinic Laboratories, 1990. Nagawa Y, Orimo H, Harasawa M, Morita I, Yashiro K, Nurota S. Effect of eicosapentaenoic acid on the platelet aggregation and composition of fatty acid in man. Atherosclerosis 1983;47:7 1-5. Hughes GS, Ringer TV, Watts KC, DeLoof MJ, Francom SF, Spillers CF. Fish oil produces an atherogenic lipid profile in hypertensive men. Atherosclerosis 1990;84:229-37.
