potential, and analytical techniques will begin to be developed in the near future. The evolution of such methodology can only begin with an appreciation and understanding by food scientists of this powerful and flexible technology.

References 1 Martin, A.E. (1966) Infrared Instrumentation and Techniques, Elsevier 2 Manning, D.J. (1972)in Laboratory Methods in Infrared Spectroscopy

(R.J.R. Miller, ed.), pp. 47-63, Hyeden Ltd 3 Biggs, D.A. (1967) J. Dairy ScL 50, 799-803 4 Biggs, D.A. (1972)J. Assoc. Off. Anal, Chem. 55, 488-492 S Association of Official Analytical Chemists (1980) Official Method:; of

Analysis, Association of Official Analytical Chemists 6 Biggs, D.A., Szijarto, L. and van de Voort, F.R. (1984) ]. Dairy 5ci. 67,

3085-3092 7 van de Voort, RR., Kermasha, S., Smith, J.R, Mills, B.L. and Ng-Kwai-

Hang, K.R (1987) J. DairySci. 70, 1515-1523 8 van de Voort, F.R. (1980) J. Assoc. Off. Anal Chem. 63, 973-980 9 Goulden, J.D.S., Shields, J. and Haswell, R. (1964) J. Soc. Dairy

Technol, 17, 28-32 10 Biggs, D.A. (1979)J. Assoc. Off. Anal. Chem. 62, 1202-1211 11 Mills, B.L, van de Voort, F.R. and Kakuda, Y. (1984) ]. Meat 5cL 11,

"1-10 12 Mills, B.L., van de Voort, F.R. and Ushorne, W.R. (1983) J. Assoc. Off.

Anal. Chem. 66, 1048---1050 13 Bjarno, O.C. (1982) J. Assoc. Off. Anal. Chem. 65, 696-701 14 Darwish, G.S., van de Yoort, F.R. and Smith, J.P. (1988) Can. J. Fish.

Aquat. Sci. 46, 644-649 15 Mills, B.L., Alyea, E.C. and van de Voort, F.R. (1986) Spectrosc. Lett.

19, 277-291

16 Mills, B.L. and van de Voart, ER. (1982) J. Assoc. Off. Anal, Chem. 65, 1357-1361

17 van de Voort, F.R., l,aureano, M. and Smith, I.P. (1988) J. Assoc. Off. Anal. Chem. 71, 1104-1110

18 van de Voort, ER., Grunfeld, E., Mills, B.L. and Paquette, G. (1986) I. Assoc. Off. Anal. Chem. 69, 285-303

19 Perkins, W.D. (1987) J. Chem. Educ. 64, A296 20 Masserschmidt, R.G. (1986) Appl, Spectrosc. 4(1, 632-635 21 Compton, D.A.C., Young, I.R., Kollar, R.G., Mooney, J.R. and Grasseli,

J.G. (1987) in Computerized Quantitative Infrared Analysis (McClure, G.I.., ed.), pp. 37-57, American Socieh, for Testing and Materials, PA, USA

22 Griffiths, P.R. and de Haseth, I.A. (1986) Fourier Transform In£'ared Spectroscopy, John Wiley and Sons

23 Coates, J., Rein, A. and Morris, K. (1988) Am. Lab. 2, 117-124 24 Crocombe, R.A., Olson M.L. and Hill, S.L. (1987) in Computerized

Quantitative Infrared Analysis (McClure, G.L., ed.), pp. 95-129, American Society for Testing and Materials, PA, USA

25 Bellamy, LJ. (1958) The tnfra.red Spectra of Complex Molecules, John Wiley and Sons

26 Nicolet lnstruments (1988) Nicolet Technical BuUetin1004/88, Nicolet Instruments, WI, USA

27 Harrick, N.J. (1967) Internal Reflection Spectroscopy, Interscience 28 Fink, D.J. and Chittur, K.K. (1986) Enzyme .Microb. Technol. 9,

568-572 29 van de Vooa, ER., Elkashef, A.A. and Mills, B.l,. (1990) ]. Assoc. Off.

Anal. Chem. 73, 688-692 30 van de Vooa, ER., Elkashef, A.A. and Blais, J.5. J. Assoc. Off. Anal.

Chem. (in press) 31 Youden, W.J. and Steiner, E.H. (1975) Statistical Manual of the AOAC

Association of Official Analytical Chemists 32 Barnett, S. and Ismail, A.A. (1989) Appl. Spectrosc. 43, 1494-1496

Viewpoint

Epidemiological and experimental evidence suggests that a

modest intake of o i ly fish confers some protection against

ischaemic heart disease. However, the mode of action of fish

oi l is still uncertain. Fish oi l has a favourable effect on serum

triglyceride levels, blood pressure, clott ing mechanisms and

heart rhythm but, in most experimental situations, the requi-

site dose far exceeds the amount that would be provided by a

typical amount of dietary fish. Further work is needed, particu-

larly on human subjects, to explain the observed protective

effects.

Current interest in the beneficial effects of the consump- tion of oily fish arose largely from observations of Greenland Eskimos. It was noticed that Eskimos rarely suffered from ischaemic heart disease (IHD), and thi~ observation was linked with their high intake of seafoods, which are rich in n-3 polyunsaturated fatty

Michael t. Burr is with the MRC Epidemiology Unit, Llandough Hospital, Penarth CF6 1XX, U K.

is oiHy fish the heart.

good for

Michael L. Burr

acids such as eicosapentaenoic acid (EPA) 1'2. Similar observations were made in Japan, where the population ~ exceptional among those of developed countries in having a high intake of EPA from fresh fish, and also in having a low incidence of cardiovascular thrombotic disease 3.

More direct evidence for the benefits of fish con- sumption was sought from long-term studies of diet and mortality. The Zutphen study 4 involved a 20-year follow-up of 852 men, initially free of IHD, whose diets had been recorded. It was found that mortality from IHD was substantially lower in those who regularly ate any type of fish than in those who did not, even after allowing for various other factors 4. Data from other cohort studies have been examined in a similar way, and

Trends in Food Science & Technology January 1991 01991, Elsevier Science Publishers Dd, (UK) 0924-2244/91/302.00 17

the results suggest that people who eat fish once or twice a week have a lower mortality from IHD than people who do not, although there is little or no advan- tage in more frequent consumption of fish ~-s.

The change in living conditions in Norway during the second world war presented the opportunity for a retro- spective epidemiological study of the effect of fish intake on heart disease. During the war, the national diet changed radically, with a substantial increase (about threefold) in the intake of fish (particularly of oily fish). Within a year, the incidence of myocardial infarc- tion and mortality from cardiovascular disease had decreased, to rise again at the end of the war when the diet revered to its former pattern 9.

The most reliable evidence for specific effects of diet on health comes from planned intervention studies. One randomized controlled trial involving 2033 men recov- ering from a myocardial infarct has been conducted to investigate the effect of fish on mortality I~. Half of the subjects were advised to eat oily fish (herring, mackerel, pilchard, salmon, sardine or trout) two or three times a week, while the others were used as controls. Dietary compliance was monitored by means of a detailed ques- tionnaire and by seven-day weighed dietary intake records. Subjects who could not eat the stipulated amount of oily fish were given fish oil capsules as a par- tial or total substitute. On average, the intervention group took 2.3 g EPA per week; the control group took 0.7 g EPA per week. After two years, morality from IHD was 29% lower in the men advised to eat fish; there was no decrease in the incidence of nonfatal re- infarction. Although the results of a single trial should never be regarded as definitive, these results, together with those from previous studies, provide strong evi- dence that oily fish is good for the heart.

Effects on serum lipids Why should fish be good for the heart? Early papers

that drew attention to the rarity of IHD in Eskimos focused on their low plasma concentrations of choles- terol and, especially, triglycerides'. Experimental studies have repeatedly shown that fish oil reduces serum triglyceride levels, even when it is ingested in fair!~ small amounts (corresponding to levels found in the amounts of fish that are commonly eaten), but has no effect on serum cholesterol levels unless very large amounts of fish are ingested II. Fish oil does, however, reduce the serum concentration of very-low-density lipoproteins (VLDL) 12-14.

Such effects may be relevant to the progression of atherosclerosis. However, they are likely to be long- term effects. Consequently, while they might explain the protective effect of fish in cohort studies (where the sub- jects presumably eat a similar diet over many years), they could hardly account for a reduction in mortality within a year or two.

Effects on haemostasis and platelets It has long been known that Eskimos tend to bleed

excessively after minimal injury. This observation has

led to the investigation of their haemostatic function, and it has been found that Eskimos have a prolonged bleeding time and that their platelets aggregate less readily than those of Danish subjects 15. A Japanese study showed that the tendency of platelets to aggregate was significantly lower in people living in a fishing vil- lage than in people in a farming village; this was ascribed to the difference (more than 21/2 times) in their fish intakes 3.

Observations of the effects of fish consumption on haemostasis and platelet function have led to a number of studies of the effects of fish oil on blood-clotting mechanisms in patients and volunteers. Again, the most reliable evidence has been derived from randomized, controlled trials. Several of these have been conducted, and they appear to show favourable effects. Fish oil seems to reduce thromboxane B2 production 16-L9, the tendency for platelets to aggregate '6, platelet factor 4 concentration ~° and blood viscosity 21, and to prolong bleeding time '2 and heparin/thrombin clotting time 22. The daily doses of EPA in these studies (1.4 g or more) were larger than would usually be provided by diet alone, but most of the studies were rather small; a lower dose might require a larger trial for an effect to be apparent. One dietary trial involved the consumption of 135 g of mackerel paste or meat paste supplements daily for six weeks 23. Platelet aggregation was inhibited by the mackerel supplement, showing that changes in diet can be suffici~:;t to produce a detectable effect on clot- ting mechanisms. Further evidence of an effect of fish consumption on haemostasis or platelet function has been provided by non-randomized or uncontrolled studies.

It seems reasonable to conclude that fish oil reduces the tendency of blood to clot, probably due to an effect on platelets. This effect is likely to be a more important component of the apparent protection conferred against heart disease than is the effect on blood lipids, since it acts over shorter periods of time (weeks, rather than years or even decades).

Effects on cardiac arrhythmias Another possible mechanism of action of fish oil is

the suppression of cardiac arrhythmia. Acute myocar- dial ischaemia is often accompanied by disturbances of rhythm, including ventricular ectopic beats and, some- times, fatal ventricular fibrillation. The reperfusion of a temporarily occluded coronary artery is also liable to provoke arrhythmia.

Controlled trials have investigated the possibility that fish oil has a protective effect in rats subjected to exper- imental coronary occlusion and reperfusion. The results show that a high intake of fish oil (12%, w/w, of the diet) confers protection against arrhythmia during both occlusion and reperfusion, and reduces mortality from ventficular fibrillation 24.25. It must be acknowledged that this intake of fish oil does not correspond to amounts that would normally be obtained in the diet of humans. Nevertheless, the findings are very interesting in view of the observation that oily fish appears to reduce mor- tality, but not the incidence of reinfarction, in human

18 Trends in Food Science & Technology January 1991

patients ~°. It is possible that fish protects, not from the occurrence of myocardial infarction, but rather from the risk of dying should infarction occur; an antiarrhythmic action wou~d explain this effect.

A series of controlled experiments has recently been conducted in marmoset monkeys 26. Being primates, monkeys are physiologically closer to man than is the rat, particularly in having a predominantly ~-adrenergic cardiac receptor system. Administration of fish oil for eight months led to increased cardiac ejection fraction, increased end diastolic volume, reduced heart rate and reduced pressure rate index - all changes associated with a reduced risk of cardiac dysfunction in man. Further studies showed that fish oil conferred protection against ventricular fibrillation induced by programmed electrical stimulation; the effect was even greater when the marmoset hearts were rendered partially ischaemic by coronary artery ligation. A diet containing 6% (w/w) fish oil fed for nearly three years appeared to abolish the risk of sustained ventricular fibrillation on electrical stimulation.

The possible mechanism of action of fish oil in mar- moset monkeys was also explored. There is evidence that thromboxane A2 is strongly arrhythmogenic, while prostacyclin is probably anti-arrhythmogenic; the bal- ance between these actions determines the degree of arrhythmia. After supplementation of the diet with fish oil, the urinary excretion of thromboxane metabolites declines and that of prostacyclin metabolites increases, in both marmosets and humans. The production of thromboxane by the isolated myocardium was differen- tially reduced with respect to prostacyclin production after the animals had been given fish oil supplements for two years: this did not occur in animals that had received supplements of saturated fat. k may be that thromboxane synthesis is specifically inhibited in the presence of cellular polyunsaturated fatty acids of the n-3 series 2~.

Obviously, it is too early to evaluate these findings in terms of their relevance to human disease. However, they suggest mechanisms that can be further explored in experimental studies on humans. One controlled triaP of the effects of cod-liver oil on humans failed to show any reduction m the frequency of ventricular extra- systoles after myocardial infarction. However, this study involved only 18 subjects, and the period of adminis- tration (six weeks) may have been too short. In the absence of a larger and longer trial, this study does not negate the possibility that fish oil has an antiarrhythmic effect on humans.

Effects on blood pressure Several controlled trials have been conducted to see

whether consumption of fish oil reduces blood pressure. Some of these studies have reported significant re- ductions in blood pressure~'lT'~8'22'zg; others have shown little or no effect3°'3L The closes used have been in excess of amounts that would normally be obtained from dietary sources. Non-randomized studies of fish supplementation have suggested that blood pressure can

be reduced by eating mackerel 3z,33. Howev~,; no change in blood pressure was attributable to oily fish in the 'DART' (diet and reinfarction trial) study I°, which used randomized, controls and a lower intake of fish.

It seems, therefore, that the beneficial effects of mod- erate intakes of oily fish (-300 g per week) are unlikely to be mediated to any important extent by a reduction in blood pressure.

Conclusions There is reason to believe that oily fish may reduce

mortality from ischaemic heart disease. The amount needed is -300 g oily fish per week; the average weekly intake of adults in the I.JK is 47g (Ref. 34). Fish oil appears to have a beneficial action on several mech- anisms known to operate in ischaemic heart disease. Much of the experimental work has employed large doses of oil that would not be readily obtained in the diet; these experiments cannot explain the observed protection con- ferret by a moderate intake of oily fish. Further studies of the effects of realistic intakes of fish are needed.

References 1 Bang, H.O., Dyerberg, J. and Nielsen, A.B. (19711Lanceti, 1143-1146 2 Bang, H.O., Dyerberg, J. and Hjorne, N. (1976) Acta Med. Scant. 200,

69-73 3 Hirai, A., Hamazaki, T., Terano, T., Nishikawa, T., Tamura, Y., Kumagai,

A. and S~jiki, I. (1980) Lancet ii, 1132-1133 4 Kromhout, D., Bosschieter, E.B. and Coulander, C. de L. (1985) N. Engl.

]. Med. 312, 1205-1209 5 Shekelle, R.B., Missell, L., Paul, O., Shryock, A.M. and Slamler, J.

(1985) N. Engl. J. Med. 313, 820 6 Vollset, S.,Heuch, l.andBjelke, E.(1985)N. EngI.J. Med. 313,

820-821 7 Curb, J.D. andReed, D.M.(1985]N. EngLJ. Med. 313,821 8 Norell, S E., Ahlbom, A., Feychting, M. and Pedersen, N.L. (1986)

gr. Meal. J. 293, 426 9 Bang, H.O. and Dye~erg, J. (1981) Acta Ivied. Scand. 210, 245-248

10 Burr, M.L., Fehily, A.M., Gilbert, I.F., Rogers, S., Holliday, R.M., Sweetnam, P.M., Elwood, PC. and Deadman, N.M. (1989) Lancetii, 757-761

11 Harris, W.S. (1989) J. Lipid Res. 30, 785-807 12 Mortensen, LZ., Schmidt, E.B., Nielsen, A.H. and Dyerberg, J. (1983)

Thromb. Haemostasis 50, 543-546 13 Schmidt, E.B., Kristensen, S.D. and Dyerberg, J. (1988) Artery 15,

316-329 14 Sanders, T.A.B., Hinds, A. and Pereira, C.C. (1989) J. Intern. Med. 225

(Suppl. 1 ), 99-104 15 Dyerb~rg, J. and Bang, H.G. (1 .~79) Lancet ii, 433-435 16 Brox, J.H., Killie, J-E., Gunnes, S. and Nordoy, A. (1981) Thromb.

Haemostasis 46, 604-611 17 Mehta, J.L., Lopez, L.M., Lawson, D., Wargovich, T.I. and Williams,

L.L (1988) Am. J. Met. 84, 45-52 18 Knapp, H.R. and Fitzgerald, G.A. (1989) N. Engl. J. Med. 320,

1037-1043 19 Hansen, J-B., Olsen, J.O., Wilsgard, L. and Osterud, B. (1989) J. Intern.

Met. 225 (Suppl. 1), 133-139 20 Grace, A.A., Curtis, L.D., Vella, M.A., Ling, K.L.E., Dickson, A.C.,

Wilson, A.P. and Betteridge, D.J. (1987) Oiabetologia 30, A525 21 Woodcock, B.E., Smith, E., Lambert, W.H., Jones, W.M., Galloway,

J.H., Greaves, M. and Preston, I.E. (1984) Br. Med. I. 288, 592-594 22 Rogers, S., James, K.S., Butland, B.K., Etherington, M.D., O'Brien, J.R.

and Jones, I.G. (1987) Atherosclcrosis 63, 137-143

Trends in Food Science & Technology January 1991 19

23 van Houwelingen, A.C., Kester, A.D.M. and Hornstra, G. (1989) Nutr. Res. 9, 1187-1196

24 McLennan, EL., Abeywardena, M.Y. and Charnock, I.S. (1988) Am. Heart 1. 116, 709-717

25 McLennan, RL., Abeywardena, M.Y. and Charnock, J.S. (1990) Am. J. Clin. Nutr. 51, 53-58

26 Chamock, I.S. World Rev. Nu~ Diet (in press) 27 Abeywardena, M.Y., McLennan, P.L. and Chamock, I.S. Am. I. Physiol.

(in press) 28 Hardarson, T., Kristinsson, A., Skuladottir, G., Asraldsdottir, H. and

Snorrason, S.P. (1989) 1. intern. Meal. 226, 33-37 29 Norris, P.G., Jones, CJ.H. and Weston, M.]. (1986) Br. Meal. ). 293,

104-105

30 Miller, I.R, Heath, I.D., Choraria, S.K., Shephard, N.W., Gajendragadkar, R.V., Harcus, A.W., Bat.son, G.A., Smith, D.W. and Saynor, R. (1988) Ctin. Chim. Acta 178, 251-260

31 Demke, D.M., Peters, G.R., Linet, 0.1., Melzler, C.M. and Kloff, K.A. (1988) Atherosclemsis 70, 73-80

32 Singer, P., laeger, W., Wirth, M., Voigt, S., Naumann, E., Zimontkowski, S., Haidu, I. and GOdicke, W. (1983) Atherosclerosis 49, 99-108

33 Singer, E, Berger, I., Luck, K., Taube, C., Naumann, E. and G~dicke, W. (1986) Atherosclerosis 62, 259-265

34 Office of Population Censuses and Surveys (1990) The Dietary and Nutritional Survey of British Adults, Her Majesty's Stationery Office, London, UK

Letters

Detecting irradiated foods

As general coordinator of the Community Bureau of Reference (BCR) program on the identification of irradiated foods sponsored by the Commission of the European Communities, I would like to address some aspects of the recent review by Grootveld et aL 1

The authors state that 'the formation of unique radiolysis products with unknown toxicological properties in irradiated foodstuffs is currently a subject of considerable debate'. However, I think the reference cited to support this (the proceedings of a WHO workshop held in November, 1986 in Neuherberg, FRG, which I attended) z disagrees with the statement; there is no 'unique r~diolytic product', Moreover, I was disappointed that the conclusions of the WHO/FAO/IAEA Expert Committee of November 1980 ('the irradiation of any food commodity up

to an overall average dose of 10 kGy presents no toxicological hazard; hence, toxicological testing of foods so treated is no longer required') were not mentioned.

Although the authors' statement that 'determination of such al~canes, alkenes or aldehydes by gas chro~natography may have application as a diagnostic test' is not incorrect, it should be emphasized that lipic; composition has to be determined before the level of alkanes or alkenes can be used to establish whether or not lood has been irradiated 3-s.

Finally, the authors suggest that ESR spectroscopy has been successfully used to detect irradiated foodstuffs. However, the reference cited (to work conducted in my laboratory) did not assert this, but studied the kinetics of the decay of radicals in irradiated starches 6.

lacques I. Baffi D~partement de Physiologie V~8~tale et

Ecosyst~mes, Centre d'Etudes Nucl~aires de Cadarache, Commissarial ~ I'Energie Atomique,

13108 Saint Paul lez Durance, Cadarache, France.

References 1 Grootveld, M., lain, R., Claxson,

A.W.D., Naughton , D. and Blake, D.R. (1990) Trends Food Sci. Technol. 1, 7-14

2 Karam, L.R. and Simic, M.G. (1986) ifJ Proceedings of the WHO Working Group on Health Impact and Control of Irradiated Foods, p. 297, Neuherberg, FRG

3 Nawar, W.W. and Balboni, I.J. (197 ,~) J. Assoc. Off. Anal. Chem. 53, 726-729

4 Vajdi, M. and Nawar, W.W. (197 Ji ). Am. Oil Chem. Soc. 56, 611-~A 5

S Vajdi, M., Nawar, W.W. and M(;rritt, C. (1983) J. Am. Oil Chem. Soc. f !, 978-986

6 Raffi, J.J. and Agnel, J.P. (1983't L Phys. Chem. 87, 2369-2373

In spite of putting it in quotation marks, the use of the term 'unnatural' by Grootveld et al. 1 to refer to radiolysis products is misleading. In fact, no 'unique' radiolytic products have been discovered after 40 years of intensive research; only compounds that also exist as food constituents have been produced by irradiation.

Furthermore, the production of volatile compounds from lipids is the result of a sequence of events much

more complicated than the attack of a single species such as e-caq. ), and no e-~a q) can be Woduced in a lipid moiety, which is hydrophobic.

Finally, I believe the inclusion of the NMR spectrum of irradiated synovial fluid was inappropriate given the main focus of the article. The space devoted to the application of NMR spectroscopy to the detection of irradiated foodstuffs was inappropriate, especially since the successful

application of the technique to foodstuffs remains to be seen.

Branka Kalu~n-R~m 'Ruder Bo~,kovi~' Institute, Bijeni~ka 54,

41 001 Zagreb, Croatia, Yugosia,~ia.

Reference 1 Gmotveld, M., lain, R., Claxson,

A.W.D., Naughton, D. and Blake, D.R. (1990) Trends Food Sci. Technol. 1, 7-14

20 Trends in Food Science & Technology January 1991