PRACTICAL LOW-FAT DIETS

505 to

be the maintenance of survival after

a

successful initial

resuscitation. 4,5 Nuffield Department of Anaesthetics, Radcliffe Infirmary, Oxford OX2 6HE

R. P. F. SCOTT

REHABILITATION AFTER MYOCARDIAL INFARCTION

SIR,-In interpreting clinical trials which are essentially negative it can be very difficult to distinguish between real effects and artefacts of chance among the plethora of results which the organisers, quite naturally, feel they should report. A recent example is the study by Mayou et al.’ to assess rehabilitation after myocardial infarction. The trial design could have provided an extra facility for making such a distinction, since information was available from three graded treatment groups: (1) the control group given routine incidental advice only (Dr Macdonald has discussed this in detail in his letter in your Jan. 30 issue); (2) exercise and some related advice; (3) wider ranging and fuller advice from the psychiatrist. In such a design an apparent effect, although statistically significant, is more likely to be real when the response is ranked in the order of the intensity of the intervention. Hence in this study it would have been more helpful to standardise the advice given to the two intervention groups so that the extra effect of exercise (if any) could have been estimated. Indeed much more discriminatory power could have been achieved from a 2 x2 factorial design2 with the same number of patients, although this might have been difficult to organise in this case. In trying to interpret the findings, one is constantly asking whether a statistically significant difference is plausible or not. This consideration quickly becomes circular because one cannot guess the expected effect of advice from a psychiatrist, relative to exercise plus a different kind of advice. Hence one is forced back into the familiar anguish of trying to discern a meaningful pattern from a multiplicity of significance tests. One cannot even begin without knowing how tht reported results were selected from these not reported. The possibilities are legion: there are at least ten outcome measures with three time periods, as well as three treatment groups. divided into attenders and non-attenders. Which contrasts were tested and which not, and why? One other problem among many that might be raised concerns randomisation. It appears that the patients were not randomly allocated to the treatment groups, and if this is so then the performance of significance tests may not be justified anyway. Like Dr Macdonald, we think Dr Mayou and colleagues’ data do not permit their conclusion that "exercise training increased confidence during exercise in the early stages of convalescence but the overall results show that rehabilitation is of little benefit to cardiac function, everyday life, or emotional state". Department of Pharmacology, Charing Cross Hospital Medical School, London W6 8RF

Department of Community and General

University of

ANDREW HERXHEIMER

Medicine

Practice, Oxford

KLIM MCPHERSON

PRACTICAL LOW-FAT DIETS

SiR,-Jean Marr and Professor Morris (Jan. 23, p. 217) are unduly pessimistic about the general acceptability of a target for the whole population of 30% of dietary energy from fat, with no more than 10% of the total energy from saturated fatty acids. Their own suggestion as to how this might be achieved may well be "draconian" but theirs is an extreme and unrealistic model for examining acceptability of these dietary recommendations. A maximum of 30% total energy from fat suggests a 25% reduction in total fat in an isoenergetic U.K. average diet. Yet to 4 5

Fisher L. Why outcome of cardiopulmonary resuscitation in general words is poor. Lancet 1982; i: 31-34. Scott RPF. Cardiopulmonary resuscitation in a teaching hospital. Anaesthesia 1981;

Hershey CO,

36: 526-30. Mayou R, MacMahon D, Sleight P, Florencio MJ. Early rehabilitation after myocardial infarction. Lancet 1981: ii. 1399-1401. 2 Peto R. Clinical trial methodology. Biomedicine 1978: 28: 24-36. 1

no more than 10% of energy from saturated fatty acids implies a 5007o reduction in these fatty acids. Different foods have varying proportions of saturated fatty acids in their fats as well as different amounts of fat. Foods which are generally available can be suggested as alternatives to the present dietary pattern, and they provide a practical and palatable way of reducing the fat in the average diet and allow a selective reduction in saturated fatty acids

achieve

the extent proposed, without the need for draconian measures. There is no need to reduce the intake of all other nutrients by attempting a 50% reduction in all foods which supply an appreciable amount of fat. One example of how acceptable change might be achieved is through skimmed milk. Milk has 63% of its fat as saturated fatty acids, and this fat constitutes 15% of all ingested fat. Were Britons to have ready access to skimmed milk they could have the advantages of milk’s nutrients without the disadvantages of its saturated fatty acid. Another means would be to recognise the lower fat cortent of poultry and fish compared with pork, lamb, and beef (see table): by selecting meats it is simple to halve the fatty acid intake without a similar reduction in the consumption of meat as such. to

CONTRIBUTION OF MEAT PRODUCTS TO DIETARY FAT IN U.K.

I

Source’ 1979.

Ministry

I

I

of Agriculture, Fisheries and Food National Food

I

Survey Commmee,

Another source of fatty acids ameniable to change is biscuits, cakes, and pastries. Although these foods contribute less than 10% of total fat, the fat content of biscuits can vary from 12 to 28% with the fat often being derived from hydrogenated coconut or palm kernel oils which have a saturated fatty acid content of over 90%. Additional measures could include a halving of the use of fatty table spreads and cooking fat. But, similarly, the choice of suitable products (e.g., corn oil rather than either lard or "blended vegetable oils") can readily allow a selected fall in saturated fatty acids. This more flexible approach still allows a completely acceptable diet. Furthermore, the recommendation to increase the consumption of cereals, fruit, and vegetables, together with choosing low fat alternative foods, has allowed the energy loss from fat reduction to be retrieved more easily. Marr and Morris suggest the reduction of 131 kj (550 kcal) which includes a proportion of non-fat energy which need not be lost from the diet. The drinking of semi skimmed or skimmed milk and eating leaner meats can mean a reduction of only about 71 - 7 kj (300 kcal). The increased consumption of bread, root vegetables, and fruit, can easily replace this energy if desired and bring additional nutritional advantages to the diet. The action to be taken about the level of polyunsaturated fatty acids (PUFAs) in the diet is a source of continual confusion. Marr and Morris designed their first diet with the constraint that there should be "no action" on PUFAs, yet their halving of dietary fat reduces PUFAs by 3-77g or 34% of the current level. The usual concern about the use of PUFAs when supporting dietary change is to avoid a "radical increase". However, it is often forgotten that reducing total dietary fat will usually reduce PUFA intake as well (27% of PUFA in the U.K. diet comes from meat and dairy produce). The controversy that surrounds the role of PUFA in preventing the development of coronary heart disease leads to needless acrimony. We are unaware of any authority which actually advocates a reduction in PUFA ingestion by the British, who already have a low intake. Therefore, it is reasonable to see that PUFA intakes are maintained. In view of the distribution and concentration of PUFAs in food, maintenance of these fatty acids will require either the addition of small amounts of concentrated PUFA sources or judicious substitution from among the different

506 alternatives from certain foods, the most suitable of which are spreads and cooking or salad oils. This is not a surreptitious attempt to drive up PUFA intakes by recommending the exclusive use of highPUFA-containing products. The practical problem, however, is to maintain PUFA levels. To do this by the increased consumption of fish, poultry, cereals, and vegetables alone would require more than a doubling of consumption of these foods. Most of these foods are bulky and have a low fat content, even if it is one rich in PUFAs. Many dietary recommendations include a slight increase in dietary PUFA above the current level. Marr and Morris are right to emphasise the "modest" extent of their 7 g per day increase to no more than 6 - 707o of total energy. This increase, it is suggested, could come from half an ounce (14 g) of margarine and a smaller amount of PUFA rich vegetable oil per day, which is well within the existing range of distribution of consumption of such foods. A diet with 3007o total energy from fat with no more than 10% from saturated fatty acids has been advocated widely for nearly twenty years. It is based on acceptable diets and not a theoretical optimum for blood lipids. Marr and Morris’s rejection of this diet on grounds of unacceptability is a subjective judgment which is difficult to support. While their faith in the likelihood of action by Government and industry is encouraging, it is important to recognise that, in the past, industry has tended to defend existing market patterns, and we have already witnessed the opposition from the agriculture and food industries to any suggestion of even a slow adjustment to the fat content of liquid milk available in the U.K. Any recommendation for reducing fat consumption will be resisted, especially with the strong fat producing interest in the U.K. agricultural and food industries. The dietary recommendations are widely supported and the health professions in the U.K. should seek their implementation. Industry should be encouraged to identify and exploit the many food-marketing opportunities which are in line with the dietary goals-for example, low fat dairy and meat products, fruit and vegetables, cereals. A healthy diet for Britain and a secure food

industry are

not

incompatible goals.

Coronary Prevention Group, Central Middlesex Hospital, London NW10 7NS

CHRISTOPHER ROBBINS

Dunn Clinical Nutritional Centre,

Cambridge CB2 1QE

CAROLINE WALKER

INTERACTION BETWEEN PARACETAMOL AND COUMARIN ANTICOAGULANTS

SIR,-Investigations of the effects of paracetamol (acetaminophen) on anticoagulation with coumarin derivatives have yielded differing results. Antlitz et al. studied fifty patients on longterm medication with warfarin, dicoumarol, and phenprocoumon who took 2600 mg paracetamol daily for 2 weeks. The PTT (prbthrombin time) rose slightly but significantly compared with values in a placebo group. By contrast, Uda112 showed that PTT values did not change in ten patients on warfarin medication who took 3250 mg paracetamol daily for 14 days. Antlitz and Awalt33 demonstrated that two doses of 650 mg given for one day did not influence the PTT. Because of these inconclusive results and because of the enormous consumption of paracetamol with and without prescription we mounted a double-blind investigation. Twenty volunteer patients on coumarin therapy were selected from patients controlled by the Amsterdam Thrombosis Service. They were allocated at random to paracetamol (ten patients) or placebo (ten patients). The patients took four tablets daily at 0900, 1300, 1800, and 2300 hours containing 500 mg paracetamol or placebo for 3 weeks. Their ’Thrombotest’ times4,5 and AM, Mead JA, Tolentino MA. Potentiation of oral anti-coagulant therapy by acetaminophen. Curr Ther Res 1968; 10: 501-06. 2. Udall JA. Drug interference with warfarin therapy. Clin Med 1970; 77: 20-25. 3. Antlitz AM, Await LF. A double blind study of acetaminophen used in conjunction with oral anticoagulant therapy. Curr Ther Res 1969; 11: 360-61. 4. Owren PA. Critical studies of tests for control of anticoagulant therapy. Thromb Diath Haemorrh 1962; suppl 1: 294-306. 5. Owern PA. Control of anticoagulant therapy. Arch Intern Med 1963; 3: 248-58.

Effect of paracetamol and

placebo on thrombotest times (TT). Standard deviations shown below figure.

anticoagulant dosages were recorded once a week for 3 weeks before the trial, during the trial, and for 2 weeks afterwards. Coumarin medication was reduced for 2 days in patients whose thrombotest values exceeded 200 s. Patients taking paracetamol showed a significant (p<0 05, t test) increase in thrombotest times (figure). Dosage of anticoagulant had to be reduced in five patients in the paracetamol group and in one of the controls. The mechanism underlying this interaction is not clear, but interference with hepatic synthesis of factors II, VII, IX, and X is possible. This study shows that high daily doses of paracetamol significantly prolong thrombotest times in patients on coumarin

therapy. JOHN J. BOEIJINGA Department of Pharmacology, Subfaculty of Pharmacy, University of Leiden, 2300 RA

Leiden, Netherlands

ETTIEN E. BOERSTRA PETER RIS DOUWE D. BREIMER ARNIE JELETICH-BASTIAANSE

MODE OF ACTION OF SULPHASALAZINE: AN ALTERNATIVE VIEW

SIR,-Recent Lancet correspondence’ has suggested that sulphasalazine (SPZ) is effective in ulcerative colitis by altering prostaglandin turnover. However, whilst prostaglandins can mimic several facets of the inflammatory response it seems unlikely that cyclo-oxygenase products of arachidonic acid metabolism mediate all aspects of the inflammatory processes in ulcerative colitis. Potent inhibitors of PG synthesis have not been effective as therapeutic agents in this disease.2 Does SPZ affect other aspects of the inflammatory response? In ulcerative colitis, as in many forms of chronic inflammation, there is lymphocytic involvement, with persistent infiltration of the colonic mucosa by lymphocytes, of which a large proportion are in an activated phase.3 Drugs used in

1. Antlitz

1. Hoult JRS, Page H. 5-aminosalicylic acid, a co-factor for colonic prostacyclin synthesis? Lancet 1981; ii: 255. 2. Campieri M, Lanfranchi GA, Bazzocchi G, Brignola C, Benatti A, Boccia S, Labo G Prostaglandins, indomethacin and ulcerative colitis. Gastroenterol 1980; 78: 193 3. Otto HF. The interepithelial lymphocytes of the intestinum: Morphological observations and immunologic aspect of intestinal enteropathy. Curr Topics Pathol

(Berlin) 1973; 57:

81.