Medical Hyporhem (19w) 31,135~139 0 Longman Group UK Ltd 19%
Dietary
Change and Coronary
Heart Disease
G. P. WALSH 26 Gorse Road, Blackburn,
UK
Abstract - Dietary changes a hundred years ago in Europe, America and Australia were needed to feed their growing industrial populations. By 1909 margarines were first made by the hydrogenation of marine oils and, later, vegetable oils as a substitute for butter, thereby introducing saturated fats. The demise of the highly nutritious herring and its oil’s hydrogenation into margarines seems to have coincided with a big increase in coronary heart disease (CHD). The Japanese escaped such changes to their rice diet by the influence of the inevitable annual monsoon while southern Europe’s CHD immunity may hinge on its permanent olive oil and vitamin C antioxidant staples. Britain’s ‘angina pectoris’ probably arose following the enclosures which changed arable (strip) into animal farming in the middle ages but cases rose only slowly from Heberden’s 100 in 1802 to McKenzie’s 200 by 1923. It seems likely, therefore, that much of the sudden increase to 2000 reported by Maurice Cassidy in 1945, despite the impact of this century’s increased longevity on CHD, will be due to dietary change together with smoking’s increasing impact.
History
Large industrial population increases 100 years ago led to acute butter shortages relieved by mixing beef fat with milk and later with vegetable oils. By 1899, beef itself being scarce, Sabatier and Sanderens’s discovery of the hydrogenation of oils into margarines was timely. This coincided with a large surplus of marine oil from the 30fold increase of the herring catch which followed the introduction of 1800 steam drifters which became necessary by 1900 for the heavy task of supplying fish to the increased population. But margarines in those early days, although derived from ‘unsaturated oils’, were of the hard
saturated variety. The recent use of unsaturated oils such as corn, containing essential fatty acids (EFAs), lead to much safer margarines and shortenings, but those from certain vegetable oils of the non-EFA and cheaper variety such as palm, palm kernel and coconut still render them saturated and could theoretically constitute a health hazard. Indeed, the coronary epidemic may have its roots in our older population’s long use of saturated oils and the epidemic’s demise may need to await the benefit of the filtering through of the safer EFA oils. Among the polyunsaturated fats (PUFAs) are the more important EFAs - so named because
135
136 they cannot be made in the body and have to be ingested. The desaturation of EFA is very slow in nature and relies, as shown in the Figure, on enzymes (desaturase - D6 and D5). These convert the common EFA linoleic acid (LA) found in vegetable oils, of the omega (w or n)-6 series, but also &nolenic acid (ALA) of the omega-3 series, through to their respective long chain EFAs and prostaglandins (PGs). The EFAs are identified by having their double bonds at the 3rd or 6th carbon interval from the end (hence, for instance, omega 6 linoleic acid = 18: 2n-6 where there are ‘18’ carbon atoms, ‘2’ double bonds and the first double bond starts at the ‘n’ or ‘0’ or ‘omega’-‘6th carbon from the end). Prostacyclin (PGI;?) (1) was discovered in the intimal coat of blood vessels in 1976. By its presence it prevents platelets from aggregating and adhering to the intima and eases vasospasm by counteracting platelet thromboxane A2 (TXA*) which, like PG12 itself, is derived from arachidonic acid (AA). PGEl from DGLA and
MEDICAL
HYPOTHESES
recently challenged (5, 6), near 50% fall in primary IDH rates among American physicians. Thus whereas aspirin forms a much better inhibitor of the cycloxygenase enzyme and, therefore, blocker of platelet aggregation than EPA, it seems to suffer the disadvantage, versus EPA, particularly if given at 600 mg daily (7), of increasingly inhibiting vessel wall prostacyclin. Treatment
In 1982, 40 CHD patients in my practice were started on 20 ml pure cod liver oil (8) and, 7 years later, most still take it. The oil averages 1.8 g EPA and, with the antioxidant vitamin E, retails at 1/12th the price, the partially concentrated fish oil. Original drug therapy was continued but after a few weeks, angina and exercise tolerance being eased, many could discontinue their trinitrin. Fasting triglyceride levels fell by an average 29%, but both low density (LDL) and high density (HDL) lipoprotein cholesterol levels changed very little even with 40 ml of the oil. HDL cholesterol levels seem to vary inversely with triglyceride levels whose n-6 EFAs n-3 EFAs reduction may benefit through their antithromLinoleic 16:2n-6 10:3n-3 Alpha-linolenic botic association with factor VllC and tissue (LA) (ALA) /delta-6-desaturase 1 plasminogen reaction (9). But, unfortunately, al+ + Gamma-linolenic 16:3n-6 though fish oils reduce triglyceride levels, like 16:4n-3 (GLA) I beta-blockers, they raise the protein moiety of + + Dihomogamma20:3n-6 20:4n-3 LDL cholesterol lipoprotein (apolipoprotein B3) llnolenlc (DGLA) delta-Sdesaturase (10) which is now known to constitute a further 1 1 CHD risk along with LDL cholesterol. The Arachidonic 20:4n-6 20:5n-3 Eicosapentaenolc therapeutic benefit of the fall in triglyceride (AA) WA) 4 4 levels may, therefore, have to be balanced Adrenic 22:4n-6 22:5n-3 against the rise in apolipoprotein B3 levels. &ta-4-desaturasel Nevertheless, hyperlipidaemic swine models 22:5-n-6 22:6;-3 Docosahexaenolc WA) given cod liver oil (11) showed a remarkable clearance of atherosclerosis without altering their high cholesterol levels (12). Pure cod liver oil PG13 from the dietary marine oil eicosapenwas used because, fortunately, it has only 40% of taenoic acid (EPA) (2) antagonise the aggrethe British Pharmacopeia (BP) variety’s potengatory thromboxanes and may thus help to tially noxious levels of vitamins A and D which prevent thrombosis by supporting the effects of were once needed for rickets but are unwanted PGI*. for heart purposes. There is a recent hint that The Eskimo CHD immunity may stem from a vitamin A, which should be avoided in pregnanfreedom from thrombosis due to EPA in their cy, may help guard against cancer (13). Serum diet consisting of seal, caribou and oily fish. This and 24 hour calcium levels, which were taken to gives them, like the fish-eating Japanese (3) who guard against kidney stones, rarely moved above also suffer much hypertension and smoke heavithe normal range (1 out of 50 patients over 5 ly, an increased bleeding tendency and a surfeit years) even at 40 ml dosage. Taste and weight of cerebral haemorrhage but little CHD. EPA gain problems were minimal in genuine heart may here mimic low dose alternate-day aspirin patients, perhaps because angina and exercise which has shown (4) a remarkable, though tolerance were so often relieved.
DIETARY
CHANGE
AND CORONARY
137
HEART DISEASE
There is also a hidden advantage in that EPA competes with arachidonate for a second EFA pathway - the lipoxygenase - which introduces the powerful leukotrienes (LTs) (14). The reduction in arachidonate-derived leukotrienes has a bearing on inflammatory (rheumatoid) disease: a CHD patient, whose additional complaint of rheumatoid arthritis had failed to respond to the rheumatologist’s therapy for the previous 8 years, found after taking as little as 5 ml cod liver oil for 10 months that her morning stiffness and then her wrist and elbow pain and swelling receded. Over the counter sales of cod liver oil for rheumatoid arthritis are, I find by asking country-wide chemists, quite considerable. The EFAs in fish oils produce the non-inflammatory leukotriene B5 (LTBs) which supplants the LTB4 derived from dietary AA and thus reduces by up to 100 times LTB4’s chemotaxic effect which attracts monocytes, smooth muscle and damaged endothelial cells to the vessel wall. EPA may thus help reduce the damaging platelet growth factor (15). Further history
The Brixham fish museum in Devon shows dramatically the change from oily to white fish at the turn of the century (16). Oily fish such as the herring that feeds on surface plankton, had been a European staple for centuries, mainly because of the ease of catching, the use of the drift net and methods of preserving the fish. This had secured the Dutch monopoly from 1400 to 1700, but with the burgeoning populations of the industrial revolution the introduction of 1800 steam drifters between 1900 and 1910 made it much easier. But this had the unfortunate consequence of so increasing the catch (30-fold) that the market was broken and the highly oxidisable fish had, to be fed to cattle and their oils hydrogenated into margarines. The extra stream power allowed trawling for white fish, giving rise to the substitution of the deep, white, sea fishing industry for the surface herring industry, and the taste for oily fish has now been mainly lost in Europe and particularly in Scotland, but not, interestingly, among the Japanese. The hydrogenation of marine and later vegetable oils seems to have coincided with an increase in CHD early this century. Ian Hill as a houseman and his chief (17, 18) saw their first coronary in 1928, and he compares the plethora of CHD cases (2000) recorded by M. Cassidy (19) in 1945 with the 100 of W. Herberden in
1802 (20), the 200 of J. McKenzie in 1923 (21) and J. B. Herrick’s 200 in 1919 (22). Although this is challenged (23) on the grounds of the increased longevity bringing more into the coronary age group, together with the 20th century’s better diagnostic ability, intuitively it seems that an easily recognisable condition such as angina pectoris should have been obvious even in 1900. Dietary changes, therefore, seem likely to have brought about a real increase in CHD from 1900 onwards. By the 1950s (24) it became evident that dietary saturated fat and LDL cholesterol were probably major causes of ischaemic heart disease (IHD), and that polyunsaturated fats (PUFAs) providing a high P/S ratio (25) might ameliorate the attacks. Although trials of low saturated fat (26) or even low fats with reduced cigarette smoking and blood pressure in healthy people (27) initially failed to reduce CHD levels, more recently cholestyramine and lovastatin drugs used in longterm trials in hypercholesterolaemic men have reduced both cholesterol and CHD levels. The impact of PUFA on CHD was initially thought to be small and to depend on the degree of unsaturation, making a high P/S ratio, as with the Eskimos, important. But this was because jthere was a failure to realise that only certain vegetable and fish oils (the EFA in PUFA) which could not be made in the body and had therefore to be ingested could ameliorate the attacks. Edinburgh
studies
The main EFA in most PUFA sources other than fish oils is linoleic acid (LA) which seems to help by lowering LDL cholesterol levels. Wood and Oliver (28) and their Edinburgh team set up the Edinburgh-Stockholm study to investigate diet and CHD. After 10 years, this found Edinburgh’s 3-fold excess CHD death toll difference might be largely due to their low dietary intake of LA. This was later confirmed by finding low adipose tissue and platelet LA in samples of victims seen soon after an attack. Similar low levels were found in 28 unsuspecting sufferers from CHD out of 448 seemingly healthy middleaged east Scotland men. Recently, smoking has been found to be linked with reduced LA levels (29), almost as though it impairs the taste for the EFA, and a similar but lesser linkage has been found between LA and the antioxidants vitamins C and E and a tendency to ventricular fibrillation in humans (30) and in rat experiments when LA
138
MEDICAL HYPOTHESES
levels are low. Further down the desaturation chain of LA is dihomogammalinolenic acid (DGLA), which converts to the anti-thrombotic PGEl; low levels of DGLA showed an even stronger association with CHD in the adipose tissue results (31). Adipose tissue levels of EPA, unlike those for LA, unfortunately are too variable for valid statistical predictions (32), and Scottish and European levels are far too low for these purposes (33, 34). Although, therefore, there i$ evidence that cod liver oil relieves angina and exercise tolerance, prolongs autogenous vein grafts in mongrel dogs (35), and produces a remarkable clearing of atherogenic coronary vessels in swine (ll), we still await proof of benefit in longer human trials (36). But by analogy with the successful saving of some 50% IHD using cheap alternate day aspirin (4), perhaps it is reasonable to speculate that EPA would achieve a not dissimilar saving. Moreover 3.2 g EPA added to the previous use of aspirin and dipyramidole has, remarkably, more than halved the previous restenosis level found 6 months after angioplasty (37). EPA has the advantage of being given as a corrective essence (20 ml cod liver oil or perhaps a few oily fish weekly) without invoking great changes in diet. Gammalinolenic acid (GLA), which is present at 9% in evening primrose oil and immediately precedes DGLA, is reputed to have substantially greater power than LA to lower cholesterol levels. A combination of EPA and DGLA may give Eskimos their CHD protection compared to
Table
18: 20: 20: 20: 22:
zn-6 3n-6 4n-6 S-3 6n-3
Phospholipid
fatty
acid
levels
the Danes. The Eskimos may have an unusual EFA metabolism, rather like cats who are obligate carnivores (38). They may show reduced activity of the D5 enzyme (Figure), with the desaturation process stopping at DGLA instead of passing on to AA and PGE2. DGLA levels are thus higher in Eskimo groups than in Danes (see Table) and their AA levels are exceptionally lower than in Danes irrespective of whether they are on a traditional or Western diet. This shows that the low levels of arachidonic acid in the Greenland Eskimos are not due to their high EPA intake. It may be the DGLA, therefore, which Oliver’s group has shown to be a strong inhibitor of CHD (28), that, allied to EPA, gives the Eskimos their CHD immunity (39). This suspicion is strengthened by the abrupt fall in DHA levels compared to the Danes when Eskimos adopt the Danish diet, and suggests that copying the Eskimo diet might not be enough without the addition of DGLA, or its precursor GLA. Discussion
CHD’s incidence may result from historic accidents such as the change from arable to animal husbandry, the overfishing of the herring, the hydrogenation of lipids into saturated fatty acids, the Eskimo loss of the D5 enzyme, or smoking’s increasing impact. Japan’s heavy smoking habit seems offset by their rice and fish diet, just as Italians and Greeks benefit from olive oil and
(m&00 mg Lipid) in Eskimos
Linoleic (LA) Dihomogammalinolenic (DGLA) Arachidonic (AA) Eicosapentaenoic (EPA) Docosahexaenoic (DHA)
and Danes
Eskimos in Greenland
Dunes in Denmark
Eskimos in Denmark
6.6 2.3 0.8 7.1 3.9
21.0 2.0 8.0 0.2 3.0
22.6 3.8 1.3 0.7 1.0
The decreased blood viscosity found in Japanese fishing villagers. whose extra fish oil (2.5 g EPA) diet compared to farmers (0.9 g EPA) allows their erythrocytes better to alter shape and mould their way through capilliaries. may partly explain the angina and exercise tolerance relief in patients using cod liver oil at 1.8 g/day (4). This level may suffice in a Britain which is almost devoid of EPA in the rest of the diet. The relative CHD immunity found in Mediterranean countries may be part explained by their olive oil diet. Oleic acid, which behaves similarly to the saturated stearic acid (40). both lowers LDL cholesterol and maintains HDL cholesterol levels. In constrast. margarines and shortenings derived from the cheaper tropical oils such as palm, palm kernel and coconut introduce the possibly atherogenic palmitic acid (41). But Mediterranean CHD immunity may also arise from antioxidants such as vitamin C (42). available more in the citrus growing regions of southern than northern Europe. These seem to prevent LDL cholesterol from penetrating the vessel wall. and Edinburgh scientists confirm this by finding an association between both low antioxidant and low linoleic acid levels and CHD.
DIETARY
CHANGE AND CORONARY
139
HEART DISEASE
from citrus fruit antioxidants. EPA and GLAcontaining oils may allow us to benefit by copying the Eskimo metabolism. Stearic acid’s analogy with oleic acid may allow a cheap and safe basis for margarines and shortenings.
References 1. Moncada S, Higgs E A, Vane J R. Human arterial and venous tissue generate prostacyclin (prostacyclin X) a potent inhibitor of platelet aggregation. Lancet 1977, 1: 18-20. Dyerberg J. Bang H 0, Stofferson E, Moncada S. Vane J R. EPA and the prevention of thromboxis and atherosclerosis. Lancet 1978, 11: 117-119. Walsh G P. Epidemic rise of coronary heart disease. Lancet 1987. 11: 562. Relman A S. Aspirin for the primary prevention of mvocardial infarction. N Enal J Med 1988, 318: 246-8. De Gaetano G. Primary prevention of vascular disease by aspirin. Lancet 1988, 1: 1093-8. Doll R. Gray R. Wheatley K. Randomised trial of prophylactic daily aspirin in British male doctors. Br Med J 1988. 296: 1193. 7. Rosmanis G, Green K, Vesterquist 0. Edhad 0. Hendriksson P. Aspirin all round. Br Med J 1988, 290: 1003. 8. Walsh G P. A GPs use of omega-3 lipids in coronary heart disease. Br J Clin Pratt (2nd Maxepa Conference) 2 1983: 75-7. 9. Oliver M F. Secondary prevention of coronary heart disease. Lancet 1,. 1989: ?i9.. 10. Barbir M. Wale D, Travner I, Aber V R. Thompson G R. High prevalence. of hypertrigliceridaemia and apolipoprotein abnormalities in coronary artery disease. Br Heart J 1988, 60: 397-403. 11. Weiner Bonni H, Ockene Ira S, Revene Pater H, et al. Inhibition of atherosclerosis by cod liver oil in a hyperlipidaemic swine model. N Engl J Med 1986, Vol 315, No. 14: 841-6. 12. Leaf A, Weber P. Cardiovascular effects of n-3 fatty acids. N Engl J Med 1988, Vol 318, No. 9: 549-57. 13. Isles C G. Holed D J, Gillis C R, Hawthorne V M, Lever A F. Plasma cholesterol, coronary heart disease, and cancer in the Renfew and Paisley survey. Br Med J 298. 1989: 91. 14. Terano T, Salmon J A. Moncada S. Effect of orally admind. EPA on the formation of leukotriene B4 and leukotriene B5 by rat leukocytes. Bioch Pharm 1984. Vol 33, No. 19: 3071-76. 15. Rubin K. et al. The induction of B-type receptors for platelet-derived growth factor in vascular inflammation: Possible impiicastions for development of vascular proliferative lesions. Lancet 1988. 1: 1353. 16. Walsh G P. The history of the herring and with its decline the significance to health. Medical Hypotheses 1986,20: 133-7. 17. Hill I. The wind of change in cardiology. Centenary Practitioner 1967, Vol 201: 44-56.
18. I-$ Ij3Treatment of heart disease. Practitioner
1980, VOI
19. Cassidy M. Coronary disease. Lancet 1946, 11: 587-590. 20. Herberden W. Commentaries on the history of the cure of disease. Br Med Asso Library, London 1802. 21. McKenzie J. Angina uectoris 1923; London. 22. Herrick J B. Thrombosis of the coronary arteries. JAMA 1919. Vol 72. No. 6: 387-90. 23. Stehbens W E. An appraisal of the epidemic rise of CHD and its decline. Lancet 1987, 1: 606-11. 24. Keys A. The cholesterol problem. Voeding 1952: 539-66. 25. Mass field trials of the diet-heart question. Report of the Diet-Heart Review Panel of the National Heart Institute. America. Monograph 28, 1969. 26. Multiple Risk Factor Intervention Trial Research Group. Rick Factor Changes and Mortality Result. JAMA 1982, 1248: 1465-77. 27. Bronte-Stewart B. Effects of feeding dietary fat on serum cholesterol in man. Am J of Clin Nutrition 1965; Lovastatin Studv Group: 17: 281-9533. 28. Wood D A. Butler S, Riemersma R A. Thompson M, Oliver M F. Adipose tissue and platelet fatty acids and coronary heart disease in Scottish men. Lancet 1984. 11: 117-21. 29. Fulton M, Oliver M F. et al. Cigarette smoking, social class and nutrient intake: Relevance to coronary heart disease. European J of Clin Nutr 1988. 30. Oliver M F. Dietary fat and CHD. Br Heart J 1987, 58: 423. 31. Horrobin D F. Genetically low AA and high DGLA levels in Eskimos may contribute to low incidences of CHD. psoriasis. arthritis and asthma. Proc. ACOS 1987: 41’3-15. 32. Sinclair H. Gale M. Eicosapentaenoic acid in fat. Lancet 1987, 1, 1202: 797-803. 33. Akos Kiss. Eicosapentaenoic acid. Lancet 1987. 1: 1083. 34. Walsh G P. EPA and dihomogammalinolenic acids. Lancet 1987, i, 1502. 35. Landvmore R W, et al. Cod liver oil in the prevention of intimal hyperplasia in autogenous vein grafts used for arterial bvoass. J Thor Card Sum 1985. 89: 351-5. 36. Sanders ? A B Fish and coro;ary artery disease. Br Heart J 1987, 57: 214-8. 37. Dehmer G J, Popma J J. et al. Reduction in the early rate of restenosis after coronary angiopolasty by a diet supplemented with n-3 fatty acids. N Engl J Med 1988. 319: 733-740. 38. Gibson R A, Sinclair A J. Are Eskimos obligate carnivores? Lancet 1981, 1: 1100. 39. Horrobin D F. Geneticallv low AA and hiah DGLA levels in Eskimos may contribute to low incidences of CHD. osoriasis. arthritis and asthma. Proc ACOS 1987. 413-5. L 40. Sachs F M. Effect of dietary stearic acid on plasma cholesterol levels. N Enel J Med 1988, 319: 1090. 41. Bonanome A. Grundy S-M. Effect of dietary stearic acid on plasma cholesterol and lipoprotein levels. N Engl J Med 1988, 318: 1244-8. 42. Steinberg D. Metabolism of lipoproteins and their role in the pathogenesis of atherosclerosis. Atherosclerosis Review 1988, Vol 18: l-2.
Dietary
Change and Coronary
Heart Disease
G. P. WALSH 26 Gorse Road, Blackburn,
UK
Abstract - Dietary changes a hundred years ago in Europe, America and Australia were needed to feed their growing industrial populations. By 1909 margarines were first made by the hydrogenation of marine oils and, later, vegetable oils as a substitute for butter, thereby introducing saturated fats. The demise of the highly nutritious herring and its oil’s hydrogenation into margarines seems to have coincided with a big increase in coronary heart disease (CHD). The Japanese escaped such changes to their rice diet by the influence of the inevitable annual monsoon while southern Europe’s CHD immunity may hinge on its permanent olive oil and vitamin C antioxidant staples. Britain’s ‘angina pectoris’ probably arose following the enclosures which changed arable (strip) into animal farming in the middle ages but cases rose only slowly from Heberden’s 100 in 1802 to McKenzie’s 200 by 1923. It seems likely, therefore, that much of the sudden increase to 2000 reported by Maurice Cassidy in 1945, despite the impact of this century’s increased longevity on CHD, will be due to dietary change together with smoking’s increasing impact.
History
Large industrial population increases 100 years ago led to acute butter shortages relieved by mixing beef fat with milk and later with vegetable oils. By 1899, beef itself being scarce, Sabatier and Sanderens’s discovery of the hydrogenation of oils into margarines was timely. This coincided with a large surplus of marine oil from the 30fold increase of the herring catch which followed the introduction of 1800 steam drifters which became necessary by 1900 for the heavy task of supplying fish to the increased population. But margarines in those early days, although derived from ‘unsaturated oils’, were of the hard
saturated variety. The recent use of unsaturated oils such as corn, containing essential fatty acids (EFAs), lead to much safer margarines and shortenings, but those from certain vegetable oils of the non-EFA and cheaper variety such as palm, palm kernel and coconut still render them saturated and could theoretically constitute a health hazard. Indeed, the coronary epidemic may have its roots in our older population’s long use of saturated oils and the epidemic’s demise may need to await the benefit of the filtering through of the safer EFA oils. Among the polyunsaturated fats (PUFAs) are the more important EFAs - so named because
135
136 they cannot be made in the body and have to be ingested. The desaturation of EFA is very slow in nature and relies, as shown in the Figure, on enzymes (desaturase - D6 and D5). These convert the common EFA linoleic acid (LA) found in vegetable oils, of the omega (w or n)-6 series, but also &nolenic acid (ALA) of the omega-3 series, through to their respective long chain EFAs and prostaglandins (PGs). The EFAs are identified by having their double bonds at the 3rd or 6th carbon interval from the end (hence, for instance, omega 6 linoleic acid = 18: 2n-6 where there are ‘18’ carbon atoms, ‘2’ double bonds and the first double bond starts at the ‘n’ or ‘0’ or ‘omega’-‘6th carbon from the end). Prostacyclin (PGI;?) (1) was discovered in the intimal coat of blood vessels in 1976. By its presence it prevents platelets from aggregating and adhering to the intima and eases vasospasm by counteracting platelet thromboxane A2 (TXA*) which, like PG12 itself, is derived from arachidonic acid (AA). PGEl from DGLA and
MEDICAL
HYPOTHESES
recently challenged (5, 6), near 50% fall in primary IDH rates among American physicians. Thus whereas aspirin forms a much better inhibitor of the cycloxygenase enzyme and, therefore, blocker of platelet aggregation than EPA, it seems to suffer the disadvantage, versus EPA, particularly if given at 600 mg daily (7), of increasingly inhibiting vessel wall prostacyclin. Treatment
In 1982, 40 CHD patients in my practice were started on 20 ml pure cod liver oil (8) and, 7 years later, most still take it. The oil averages 1.8 g EPA and, with the antioxidant vitamin E, retails at 1/12th the price, the partially concentrated fish oil. Original drug therapy was continued but after a few weeks, angina and exercise tolerance being eased, many could discontinue their trinitrin. Fasting triglyceride levels fell by an average 29%, but both low density (LDL) and high density (HDL) lipoprotein cholesterol levels changed very little even with 40 ml of the oil. HDL cholesterol levels seem to vary inversely with triglyceride levels whose n-6 EFAs n-3 EFAs reduction may benefit through their antithromLinoleic 16:2n-6 10:3n-3 Alpha-linolenic botic association with factor VllC and tissue (LA) (ALA) /delta-6-desaturase 1 plasminogen reaction (9). But, unfortunately, al+ + Gamma-linolenic 16:3n-6 though fish oils reduce triglyceride levels, like 16:4n-3 (GLA) I beta-blockers, they raise the protein moiety of + + Dihomogamma20:3n-6 20:4n-3 LDL cholesterol lipoprotein (apolipoprotein B3) llnolenlc (DGLA) delta-Sdesaturase (10) which is now known to constitute a further 1 1 CHD risk along with LDL cholesterol. The Arachidonic 20:4n-6 20:5n-3 Eicosapentaenolc therapeutic benefit of the fall in triglyceride (AA) WA) 4 4 levels may, therefore, have to be balanced Adrenic 22:4n-6 22:5n-3 against the rise in apolipoprotein B3 levels. &ta-4-desaturasel Nevertheless, hyperlipidaemic swine models 22:5-n-6 22:6;-3 Docosahexaenolc WA) given cod liver oil (11) showed a remarkable clearance of atherosclerosis without altering their high cholesterol levels (12). Pure cod liver oil PG13 from the dietary marine oil eicosapenwas used because, fortunately, it has only 40% of taenoic acid (EPA) (2) antagonise the aggrethe British Pharmacopeia (BP) variety’s potengatory thromboxanes and may thus help to tially noxious levels of vitamins A and D which prevent thrombosis by supporting the effects of were once needed for rickets but are unwanted PGI*. for heart purposes. There is a recent hint that The Eskimo CHD immunity may stem from a vitamin A, which should be avoided in pregnanfreedom from thrombosis due to EPA in their cy, may help guard against cancer (13). Serum diet consisting of seal, caribou and oily fish. This and 24 hour calcium levels, which were taken to gives them, like the fish-eating Japanese (3) who guard against kidney stones, rarely moved above also suffer much hypertension and smoke heavithe normal range (1 out of 50 patients over 5 ly, an increased bleeding tendency and a surfeit years) even at 40 ml dosage. Taste and weight of cerebral haemorrhage but little CHD. EPA gain problems were minimal in genuine heart may here mimic low dose alternate-day aspirin patients, perhaps because angina and exercise which has shown (4) a remarkable, though tolerance were so often relieved.
DIETARY
CHANGE
AND CORONARY
137
HEART DISEASE
There is also a hidden advantage in that EPA competes with arachidonate for a second EFA pathway - the lipoxygenase - which introduces the powerful leukotrienes (LTs) (14). The reduction in arachidonate-derived leukotrienes has a bearing on inflammatory (rheumatoid) disease: a CHD patient, whose additional complaint of rheumatoid arthritis had failed to respond to the rheumatologist’s therapy for the previous 8 years, found after taking as little as 5 ml cod liver oil for 10 months that her morning stiffness and then her wrist and elbow pain and swelling receded. Over the counter sales of cod liver oil for rheumatoid arthritis are, I find by asking country-wide chemists, quite considerable. The EFAs in fish oils produce the non-inflammatory leukotriene B5 (LTBs) which supplants the LTB4 derived from dietary AA and thus reduces by up to 100 times LTB4’s chemotaxic effect which attracts monocytes, smooth muscle and damaged endothelial cells to the vessel wall. EPA may thus help reduce the damaging platelet growth factor (15). Further history
The Brixham fish museum in Devon shows dramatically the change from oily to white fish at the turn of the century (16). Oily fish such as the herring that feeds on surface plankton, had been a European staple for centuries, mainly because of the ease of catching, the use of the drift net and methods of preserving the fish. This had secured the Dutch monopoly from 1400 to 1700, but with the burgeoning populations of the industrial revolution the introduction of 1800 steam drifters between 1900 and 1910 made it much easier. But this had the unfortunate consequence of so increasing the catch (30-fold) that the market was broken and the highly oxidisable fish had, to be fed to cattle and their oils hydrogenated into margarines. The extra stream power allowed trawling for white fish, giving rise to the substitution of the deep, white, sea fishing industry for the surface herring industry, and the taste for oily fish has now been mainly lost in Europe and particularly in Scotland, but not, interestingly, among the Japanese. The hydrogenation of marine and later vegetable oils seems to have coincided with an increase in CHD early this century. Ian Hill as a houseman and his chief (17, 18) saw their first coronary in 1928, and he compares the plethora of CHD cases (2000) recorded by M. Cassidy (19) in 1945 with the 100 of W. Herberden in
1802 (20), the 200 of J. McKenzie in 1923 (21) and J. B. Herrick’s 200 in 1919 (22). Although this is challenged (23) on the grounds of the increased longevity bringing more into the coronary age group, together with the 20th century’s better diagnostic ability, intuitively it seems that an easily recognisable condition such as angina pectoris should have been obvious even in 1900. Dietary changes, therefore, seem likely to have brought about a real increase in CHD from 1900 onwards. By the 1950s (24) it became evident that dietary saturated fat and LDL cholesterol were probably major causes of ischaemic heart disease (IHD), and that polyunsaturated fats (PUFAs) providing a high P/S ratio (25) might ameliorate the attacks. Although trials of low saturated fat (26) or even low fats with reduced cigarette smoking and blood pressure in healthy people (27) initially failed to reduce CHD levels, more recently cholestyramine and lovastatin drugs used in longterm trials in hypercholesterolaemic men have reduced both cholesterol and CHD levels. The impact of PUFA on CHD was initially thought to be small and to depend on the degree of unsaturation, making a high P/S ratio, as with the Eskimos, important. But this was because jthere was a failure to realise that only certain vegetable and fish oils (the EFA in PUFA) which could not be made in the body and had therefore to be ingested could ameliorate the attacks. Edinburgh
studies
The main EFA in most PUFA sources other than fish oils is linoleic acid (LA) which seems to help by lowering LDL cholesterol levels. Wood and Oliver (28) and their Edinburgh team set up the Edinburgh-Stockholm study to investigate diet and CHD. After 10 years, this found Edinburgh’s 3-fold excess CHD death toll difference might be largely due to their low dietary intake of LA. This was later confirmed by finding low adipose tissue and platelet LA in samples of victims seen soon after an attack. Similar low levels were found in 28 unsuspecting sufferers from CHD out of 448 seemingly healthy middleaged east Scotland men. Recently, smoking has been found to be linked with reduced LA levels (29), almost as though it impairs the taste for the EFA, and a similar but lesser linkage has been found between LA and the antioxidants vitamins C and E and a tendency to ventricular fibrillation in humans (30) and in rat experiments when LA
138
MEDICAL HYPOTHESES
levels are low. Further down the desaturation chain of LA is dihomogammalinolenic acid (DGLA), which converts to the anti-thrombotic PGEl; low levels of DGLA showed an even stronger association with CHD in the adipose tissue results (31). Adipose tissue levels of EPA, unlike those for LA, unfortunately are too variable for valid statistical predictions (32), and Scottish and European levels are far too low for these purposes (33, 34). Although, therefore, there i$ evidence that cod liver oil relieves angina and exercise tolerance, prolongs autogenous vein grafts in mongrel dogs (35), and produces a remarkable clearing of atherogenic coronary vessels in swine (ll), we still await proof of benefit in longer human trials (36). But by analogy with the successful saving of some 50% IHD using cheap alternate day aspirin (4), perhaps it is reasonable to speculate that EPA would achieve a not dissimilar saving. Moreover 3.2 g EPA added to the previous use of aspirin and dipyramidole has, remarkably, more than halved the previous restenosis level found 6 months after angioplasty (37). EPA has the advantage of being given as a corrective essence (20 ml cod liver oil or perhaps a few oily fish weekly) without invoking great changes in diet. Gammalinolenic acid (GLA), which is present at 9% in evening primrose oil and immediately precedes DGLA, is reputed to have substantially greater power than LA to lower cholesterol levels. A combination of EPA and DGLA may give Eskimos their CHD protection compared to
Table
18: 20: 20: 20: 22:
zn-6 3n-6 4n-6 S-3 6n-3
Phospholipid
fatty
acid
levels
the Danes. The Eskimos may have an unusual EFA metabolism, rather like cats who are obligate carnivores (38). They may show reduced activity of the D5 enzyme (Figure), with the desaturation process stopping at DGLA instead of passing on to AA and PGE2. DGLA levels are thus higher in Eskimo groups than in Danes (see Table) and their AA levels are exceptionally lower than in Danes irrespective of whether they are on a traditional or Western diet. This shows that the low levels of arachidonic acid in the Greenland Eskimos are not due to their high EPA intake. It may be the DGLA, therefore, which Oliver’s group has shown to be a strong inhibitor of CHD (28), that, allied to EPA, gives the Eskimos their CHD immunity (39). This suspicion is strengthened by the abrupt fall in DHA levels compared to the Danes when Eskimos adopt the Danish diet, and suggests that copying the Eskimo diet might not be enough without the addition of DGLA, or its precursor GLA. Discussion
CHD’s incidence may result from historic accidents such as the change from arable to animal husbandry, the overfishing of the herring, the hydrogenation of lipids into saturated fatty acids, the Eskimo loss of the D5 enzyme, or smoking’s increasing impact. Japan’s heavy smoking habit seems offset by their rice and fish diet, just as Italians and Greeks benefit from olive oil and
(m&00 mg Lipid) in Eskimos
Linoleic (LA) Dihomogammalinolenic (DGLA) Arachidonic (AA) Eicosapentaenoic (EPA) Docosahexaenoic (DHA)
and Danes
Eskimos in Greenland
Dunes in Denmark
Eskimos in Denmark
6.6 2.3 0.8 7.1 3.9
21.0 2.0 8.0 0.2 3.0
22.6 3.8 1.3 0.7 1.0
The decreased blood viscosity found in Japanese fishing villagers. whose extra fish oil (2.5 g EPA) diet compared to farmers (0.9 g EPA) allows their erythrocytes better to alter shape and mould their way through capilliaries. may partly explain the angina and exercise tolerance relief in patients using cod liver oil at 1.8 g/day (4). This level may suffice in a Britain which is almost devoid of EPA in the rest of the diet. The relative CHD immunity found in Mediterranean countries may be part explained by their olive oil diet. Oleic acid, which behaves similarly to the saturated stearic acid (40). both lowers LDL cholesterol and maintains HDL cholesterol levels. In constrast. margarines and shortenings derived from the cheaper tropical oils such as palm, palm kernel and coconut introduce the possibly atherogenic palmitic acid (41). But Mediterranean CHD immunity may also arise from antioxidants such as vitamin C (42). available more in the citrus growing regions of southern than northern Europe. These seem to prevent LDL cholesterol from penetrating the vessel wall. and Edinburgh scientists confirm this by finding an association between both low antioxidant and low linoleic acid levels and CHD.
DIETARY
CHANGE AND CORONARY
139
HEART DISEASE
from citrus fruit antioxidants. EPA and GLAcontaining oils may allow us to benefit by copying the Eskimo metabolism. Stearic acid’s analogy with oleic acid may allow a cheap and safe basis for margarines and shortenings.
References 1. Moncada S, Higgs E A, Vane J R. Human arterial and venous tissue generate prostacyclin (prostacyclin X) a potent inhibitor of platelet aggregation. Lancet 1977, 1: 18-20. Dyerberg J. Bang H 0, Stofferson E, Moncada S. Vane J R. EPA and the prevention of thromboxis and atherosclerosis. Lancet 1978, 11: 117-119. Walsh G P. Epidemic rise of coronary heart disease. Lancet 1987. 11: 562. Relman A S. Aspirin for the primary prevention of mvocardial infarction. N Enal J Med 1988, 318: 246-8. De Gaetano G. Primary prevention of vascular disease by aspirin. Lancet 1988, 1: 1093-8. Doll R. Gray R. Wheatley K. Randomised trial of prophylactic daily aspirin in British male doctors. Br Med J 1988. 296: 1193. 7. Rosmanis G, Green K, Vesterquist 0. Edhad 0. Hendriksson P. Aspirin all round. Br Med J 1988, 290: 1003. 8. Walsh G P. A GPs use of omega-3 lipids in coronary heart disease. Br J Clin Pratt (2nd Maxepa Conference) 2 1983: 75-7. 9. Oliver M F. Secondary prevention of coronary heart disease. Lancet 1,. 1989: ?i9.. 10. Barbir M. Wale D, Travner I, Aber V R. Thompson G R. High prevalence. of hypertrigliceridaemia and apolipoprotein abnormalities in coronary artery disease. Br Heart J 1988, 60: 397-403. 11. Weiner Bonni H, Ockene Ira S, Revene Pater H, et al. Inhibition of atherosclerosis by cod liver oil in a hyperlipidaemic swine model. N Engl J Med 1986, Vol 315, No. 14: 841-6. 12. Leaf A, Weber P. Cardiovascular effects of n-3 fatty acids. N Engl J Med 1988, Vol 318, No. 9: 549-57. 13. Isles C G. Holed D J, Gillis C R, Hawthorne V M, Lever A F. Plasma cholesterol, coronary heart disease, and cancer in the Renfew and Paisley survey. Br Med J 298. 1989: 91. 14. Terano T, Salmon J A. Moncada S. Effect of orally admind. EPA on the formation of leukotriene B4 and leukotriene B5 by rat leukocytes. Bioch Pharm 1984. Vol 33, No. 19: 3071-76. 15. Rubin K. et al. The induction of B-type receptors for platelet-derived growth factor in vascular inflammation: Possible impiicastions for development of vascular proliferative lesions. Lancet 1988. 1: 1353. 16. Walsh G P. The history of the herring and with its decline the significance to health. Medical Hypotheses 1986,20: 133-7. 17. Hill I. The wind of change in cardiology. Centenary Practitioner 1967, Vol 201: 44-56.
18. I-$ Ij3Treatment of heart disease. Practitioner
1980, VOI
19. Cassidy M. Coronary disease. Lancet 1946, 11: 587-590. 20. Herberden W. Commentaries on the history of the cure of disease. Br Med Asso Library, London 1802. 21. McKenzie J. Angina uectoris 1923; London. 22. Herrick J B. Thrombosis of the coronary arteries. JAMA 1919. Vol 72. No. 6: 387-90. 23. Stehbens W E. An appraisal of the epidemic rise of CHD and its decline. Lancet 1987, 1: 606-11. 24. Keys A. The cholesterol problem. Voeding 1952: 539-66. 25. Mass field trials of the diet-heart question. Report of the Diet-Heart Review Panel of the National Heart Institute. America. Monograph 28, 1969. 26. Multiple Risk Factor Intervention Trial Research Group. Rick Factor Changes and Mortality Result. JAMA 1982, 1248: 1465-77. 27. Bronte-Stewart B. Effects of feeding dietary fat on serum cholesterol in man. Am J of Clin Nutrition 1965; Lovastatin Studv Group: 17: 281-9533. 28. Wood D A. Butler S, Riemersma R A. Thompson M, Oliver M F. Adipose tissue and platelet fatty acids and coronary heart disease in Scottish men. Lancet 1984. 11: 117-21. 29. Fulton M, Oliver M F. et al. Cigarette smoking, social class and nutrient intake: Relevance to coronary heart disease. European J of Clin Nutr 1988. 30. Oliver M F. Dietary fat and CHD. Br Heart J 1987, 58: 423. 31. Horrobin D F. Genetically low AA and high DGLA levels in Eskimos may contribute to low incidences of CHD. psoriasis. arthritis and asthma. Proc. ACOS 1987: 41’3-15. 32. Sinclair H. Gale M. Eicosapentaenoic acid in fat. Lancet 1987, 1, 1202: 797-803. 33. Akos Kiss. Eicosapentaenoic acid. Lancet 1987. 1: 1083. 34. Walsh G P. EPA and dihomogammalinolenic acids. Lancet 1987, i, 1502. 35. Landvmore R W, et al. Cod liver oil in the prevention of intimal hyperplasia in autogenous vein grafts used for arterial bvoass. J Thor Card Sum 1985. 89: 351-5. 36. Sanders ? A B Fish and coro;ary artery disease. Br Heart J 1987, 57: 214-8. 37. Dehmer G J, Popma J J. et al. Reduction in the early rate of restenosis after coronary angiopolasty by a diet supplemented with n-3 fatty acids. N Engl J Med 1988. 319: 733-740. 38. Gibson R A, Sinclair A J. Are Eskimos obligate carnivores? Lancet 1981, 1: 1100. 39. Horrobin D F. Geneticallv low AA and hiah DGLA levels in Eskimos may contribute to low incidences of CHD. osoriasis. arthritis and asthma. Proc ACOS 1987. 413-5. L 40. Sachs F M. Effect of dietary stearic acid on plasma cholesterol levels. N Enel J Med 1988, 319: 1090. 41. Bonanome A. Grundy S-M. Effect of dietary stearic acid on plasma cholesterol and lipoprotein levels. N Engl J Med 1988, 318: 1244-8. 42. Steinberg D. Metabolism of lipoproteins and their role in the pathogenesis of atherosclerosis. Atherosclerosis Review 1988, Vol 18: l-2.
