267
Letters to the Editor
EXERCISE AND THE HEART
S!R,—In light of the correspondence (Jan. 3, p. 36) on our paper of Dec. 6 (p. 1207) it is plain that we should have written more explicitly about self-selection as a possible explanation of the findings, though to our mind the whole report bears on the issue. First, however, as described in the paper and previously,I,2 our middle-aged civil servants are neither "athletes" nor "marathon runners". Next, it is incorrect of Dr Jarrett to say that "heavy work had little if any association with a reduction of CHD". Men in their 50s, 60s, and early 70s, initially reporting heavy work, suffered fewer heart-attacks; it is only by comparison with the sports that the reduction appears small. Furthermore, while some of the men’s digging in the garden could well have been necessary (just as some stair-climbing will be unavoidable), most of the heavy work reported by the men was every bit as "voluntary" as the vigorous exercise (VE) sports and recreations. Jarrett’s substantive point is that the difference in rates of CHD between those engaging in vigorous exercise and the no-VE men could be an artifact, arising from self selection into the latter category of men who had discontinued VE because of angina that was not induced merely by hurrying or walking uphill, about which the standard questionnaire inquires; such men, it is proposed, would subsequently have a high incidence of heart-attack. Now this kind of process may well occur, but the question is-how often, and could it be responsible for the main result? Simple arithmetic indicates that even with a CHD incidence of 15% among the men with second-degree angina, as high a rate as in those with more easily induced chest pain, some 4500 men (out of a total of 18 000) would have had to make this cross-over from VE sports to produce the difference in’coronary numbers and rates between the no-VE men and VE-sports men that was found. Such a mass migration would imply a trebling of the number initially engaged in VE sports in middle age, and the new epidemic disorder that Jarrett adumbrates would multiply the overall frequency of angina in the population 5 or 6 fold. If those over 50 years alone are considered, as Jarrett intends, the numbers become even more unreal. The main effect of any such change in VE category after entry would be falsely to raise the CHD rates among the VE-sports men. In fact our data indicate that few men give up VE in middle age for any reason; it is in their 20s and 30s that many slip into bad habits. Jarrett suggests also that dyspnoea on vigorous exercise may lead to such selective bias, but this is to add to the confusion. VE-sports men report far more often than the non-VE that they get out of breath-this, after all may be integral to any "training" that occurs. What, on the contrary, we need to know is if such dyspnoea is associated with particularly low rates of CHD; whether merely to perspire does as well; the prognostic significance of dyspnoea with relatively little effort when reported by no-VE men; and so on. Our own treatment of the possibility of selection into the non-VE category by incipient heart disease, as Professor Burch calls it (though he seems to have overlooked what we wrote), is more direct and, we think, goes some way to dispose of this general category of selective process. The proposition is that men who, on entry, were avoiding vigorous exercise because they were sickening (e.g., with myocardial ischaemia and diminished cardiac output) would suffer particularly high CHD incidence early in the follow-up, so producing a spurious overall excess. The figures given in the paper of a long-term continuing effect of VE indicate that this did not happen, at any rate on any scale that was critical. In detail, for example, here are the age-standardised incidence rates of CHD (%) for men aged 50-65 years at entry: 1. Morris JN Uses of epidemiology. London. Churchill Livingstone, 1975. 2 Epstein L, Miller GJ, Stitt FW, Morris JN Vigorous exercise in leisure-time, coronary risk factors, and resting electrocardiogram in middle-aged male civil servants Br Heart J 1976; 38: 403-09
In discussing the ECG study’Jarrett omits to mention that this kind of reasoning was carried further, and when, singly or in combination, men with a clinical history not merely of CHD, but of hypertension, cerebrovascular disease or diabetes, or positive on the standard questionnaires for chest pain, leg pain or dyspnoea-when all these were excluded from the analysis and the residual groups compared-those reporting VE continued to show only half the abnormal ECGs of the no-VE men. The other and more important side of the issue is positive selection : that men with a stronger cardiovascular system, and naturally more active (engaging, for example, in VE sports), will constitute a group with inherently low CHD, so generating the kind of advantage reported in the paper. What we find most interesting is that computation of the experience of men who, plausibly, were likely to be high on such selective fitness factors-men with long-lived parents (this is the only approach to "genetics" we can make), the tall, those who never smoked-demonstrates that they did indeed have low or very low coronary incidence rates. But tables 11, III, and IV of the paper also show that within each such category those who reported VE sports had substantially less CHD than the others. And, conversely, their opposites (men with a bad family history, the small of stature, cigarette-smokers) did indeed have high or very high rates of CHD, but again, engaging in VE-sports, only about half as high as the rest. It is this generality and predictability which suggests that adequate exercise is a natural defence of the body against CHD, and so affects longevity. The main issue is usually expressed as a dichotomy, of selection of men resistant to CHD and physically more active, or protection from CHD by the exercise itself, when commonsense and empirical study 3-6 suggest the useful hypothesis that both are effective. Years ago we used to put it-how much does it matter that different men choose different jobs, to be a bus conductor or driver for example, and how much that they then spend 20, 30, 40 years doing different things? To identify the favourable personal factors is of great theoretical interest and importance; evidence that the exercise itself is beneficial is of practical importance, as well. Departments of Community Health, Human Nutrition, and Medical Statistics and Epidemiology, London School of Hygiene and Tropical Medicine,
J. N. MORRIS
London WC1
M. G. EVERITT R. POLLARD S. P. W. CHAVE
Medical Advisory Service, Civil Service Department, London
A. M. SEMMENCE
BIOCHEMICAL AND MORPHOLOGICAL CHANGES OF HUMAN CERVIX AFTER LOCAL APPLICATION OF PROSTAGLANDIN E, IN PREGNANCY
SIR,-Research on spontaneous and artificially induced labour has focused on agents which stimulate the myometrium to contract. Recently, however, the functional significance of the cervix during pregnancy and delivery has been acknowledged. 1,2 Thus, the control of cervical ripening-i.e., softening, effacement, and dilatation 3. Morris JN, Heady JA, Raffle PAB. Physique of London busmen: Epidemiology of uniforms. Lancet 1956; ii: 569-70. 4. Heady JA, Morris JN, Kagan A, Raffle PAB. Coronary heart disease in London busmen. A progress report with particular reference to physique. Br J Prev Soc
Med 1961; 15: 143-53. 5. Morris JN, Kagan A, Pattison DC, Gardner MJ, Raffle PAB. Incidence and prediction of ischaemic heart-disease in London busmen Lancet 1966; ii: 553-59. 6. Paffenbarger RS Jr. Physical activity and fatal heart attack. protection or selection? In: Amsterdam EA, Wilmore JH, DeMaria AN, eds Exercise in cardiovascular health and disease. New York: Yorke Medical Books, 1977. 34-49. 1. Liggins GC. Ripening ofthe cervix. In: Oliver TK, Kirschbaum TH, eds. Seminars in perinatology; vol II. New York, 1978: 261-71. 2. Editorial. The unripe cervix. Lancet 1979; i: 364-65.