Jan 5, 1985 - Whatever adjective is used, hypertension in women is a suitable, not separate, case for treatment. It is a sad day when female equality.
74
under 70 with mild hypertension sustained over a four month period of repeated blood pressure measurement who might be safely spared drug treatment. IndLed, among smokers, the benefit of drug treatment was actually greater for women with hypertension than for men. By way of extra confusion, Dr Silman has chosen to redefine the terms "mild" and "severe" hypertension. The Australian trial' and the Medical Research Council trial4 defined mild hypertension as a diastolic blood pressure of 95-109 and 90-109 mm Hg respectively. Yet Dr Silman states that severe hypertension is a diastolic blood pressure over 105 mm Hg, although the term is normally reserved for a diastolic blood pressure over 120 mm Hg.' Whatever adjective is used, hypertension in women is a suitable, not separate, case for treatment. It is a sad day when female equality can be embraced only on production of single blind placebo controlled evidence reaching the required significance; sadder still when such evidence is overlooked. NICHOLAS BRADLEY Ide, Exeter, Devon
BRITISH MEDICAL JOURNAL
Increases in platelet and red cell counts, blood viscosity, and arterial pressure during mild surface cooling
SIR,-Professor W R Keatinge and others reported an increase in plasma and blood viscosity after hypothermic exposure and discussed the intriguing possibility of this (and other) changes being a causal factor for the increased incidence of thrombosis during
the cold seasons (24 November, p 1405). These results are remarkable, and a more thorough rheological analysis comprising up to date techniques of the various facets of blood rheology might be indicated.
We have submitted 12 volunteers to acute whole body hypothermia by water bath with an initial temperature of 30°C which was reduced to 20C during the first five minutes and then maintained at this temperature for 20 minutes. Immediately before and after, plasma and blood viscosity (on native blood and at a packed cell volume of 0-45), packed cell volume, red cell deformability, red cell aggregation, and leucocyte count were determined (all at controlled temperature) by methods that are generally accepted and published elsewhere.' We saw significant (p < 0-05) changes in the variables listed in the table. Plasma viscosity, red 1 Management Committee, Australian National Blood cell deformability, red cell aggregation, and high Pressure Study. The Australian therapeutic trial in shear blood viscosity did not change significantly. mild hypertension. Lancet 1980;i:1261-7.
2 Management Committee, Australian National Blood Pressure Study. Prognostic factors in the treatment of mild hypertension. Circulation 1984;69:668-76. 3 Reader R. Australian therapeutic trial in mild hypertension. Med_Y Aust 1984;i:752-4. 4 Medical Research Council Working Party on Mild to Moderate Hypertension. Randomised controlled trial of treatment for mild hypertension: design and pilot trial. Br Med J 1977;i:1437-40. 5 Hart JT. Hypertension. London: Churchill Livingstone, 1980.
***Dr Silman replies below.-ED, BM7. SIR,-The recently published analysis of subgroups from the Australian National Blood Pressure Study became available too late for inclusion in my leading article.' It is worthy of comment but, as the authors themselves point out, definite conclusions cannot be made because of the small number of trial end points in women. They argue, in agreement with my advice, that the question of treatment in women (and other subgroups) awaits answers from the current prospective studies such as that by the Medical Research Council. Interestingly, there was virtually no beneficial effect of hypotensive treatment in non-smoking women, though there appeared to be a large benefit in smokers. The authors caution against drawing conclusions from these data because of the small numbers of terminal events. Nevertheless, their data are consistent with my stated view that hypertensive women who smoke are at greater risk and thus may benefit from a lower threshold for the initiation of treatment. Severity, like beauty, is in the eyes of the beholder, and a cut off point of 105 or 109 mm Hg diastolic pressure as tIe: upper limit for mild to moderate hypertension is normally accepted. Finally, sexual equality in treatment is justified only if there is sexual equality in risk. Unfortunately, for half the population such is not the case. ALAN J SILMAN
Mean (SD) blood variables before and after hypothermia Before
After
Blood viscosity (mPa s) at shear rate of: 0 06 s-' 80 7 (14 7) 88 5 (11-7) 945 s-' 48 (04) 49 (03) 0 06 s ', packed cell volume 0 45 80 4 (5 5) 84-2 (7 7) 6-7 (0 9) 7-7 (1-2) Leucocytes ( x 109/1) 045 (0(029) 0 46 (0-019) Packed cell volume Conversion: SI to traditional units-Blood viscosity: 1 mPa s = 1 cP.
In essence these results confirm those reported by Keatinge et al. They also suggest that acute severe hypothermia has similar haemorheological effects to mild prolonged hypothermia. The explanation of increased adrenergic tone being causally involved, as outlined by Keatinge et al, seems likely, especially as similar modifications of blood rheology in other situations of high adrenergic
tone-for example, during psychoemotional stress-have been reported.' It may be worth mentioning that acute whole body hyperthermia induces the reverse changes-namely, "fluidification" of blood.2 The possibility of blood "viscidation" induced by hypothermia having a role in thrombogenesis seems to merit further investigation. E ERNST A MATRAI A SCHERER Hemorheology Research Unit, University Clinics, Munich, FDR 1 Ernst E, Baumann M, Matrai A. Prolonged psychoemotional stress decreases blood fluidity. Clinical Hernorheology 1984;4:423. 2 Ernst E, Scherer A, Magyarosy I, Drexel H. Hyperthermia and blood rheology. International Journal of Aficrocirculation 1984;3:618.
SIR,-I was most interested to read the article by Professor W R Keatinge and his colleagues on increased blood viscosity and other factors Department of Clinical Epidemiology, The London Hospital, in relation to coronary and cerebral thrombosis. London El 1BB Though I realise that this was not part of their I Management Committee of the Australian National study protocol, I wonder whether they should Blood Pressure Study. Prognostic factors in the not consider the effects of localised organ cooltreatment of mild hypertension. Circulation 1984; ing in addition to surface cooling of the body. 69:668-76.
VOLUME 290
5 JANUARY 1985
Those who suffer from angina may become all too familiar with the rapid induction of pain when they go out in cold weather. In such circumstances the inspiration of cold air into the lungs (and possibly into the gullet too) must lead to fairly rapid surface cooling of the heart and mediastinum. Does this not induce platelet sludging and spasm in the cardiac vessels ? While the mechanism in the cerebral circulation is not quite analogous, there must be an element of cooling of the aorta and great vessels (and therefore of the cerebral arterial blood) by cold air in the lungs, as well as cooling of the neck. Moreover, anyone who is bald usually recognises how unwise it is to go out in cold weather with his head uncovered, when he quickly appreciates rapid cooling of the scalp and head. Professor Keatinge and his colleagues very properly suggest that, in the circumstances they describe, there is delayed onset of coronary thrombosis and death therefrom, and a generally longer delay before onset of a stroke. However, from my personal experience of suffering a myocardial infarct in 1977 within half an hour of starting a walk in very cold weather, and of an earlier infarct in 1972, when I was unwise enough to take long swims in cold water, I suggest that this hypothesis of local organ cooling is worthy of further study. I recognise, of course, that the changes I postulate would be more likely to occur in previously diseased vessels or beyond a localised site of narrowing. REGINALD MURLEY Radlett,
Herts WD7 7JN
SIR,-Professor W R Keatinge and his colleagues have shown that surface cooling in young, healthy adults induces a rise in the formed elements of the blood which is associated with an increase in whole blood and plasma viscosity. They suggest that such changes probably explain the increased incidence of coronary and cerebral thrombosis in cold weather. The fact that whole blood viscosity increased much more than plasma viscosity and that this increase exceeded that expected purely on the grounds of the rise in cell numbers strongly suggests that there is a qualitative change in either red cells or platelets, or both, which also enhances viscosity. Possible mechanisms might include an increase in the ability of red cells to aggregate or, as Professor Keatinge and his colleagues indicate, a decrease in red cell deformability consequent on enhanced platelet reactivity. In regard to the latter possibility cooling induces a number of possibly relevant changes in platelet and vessel wall metabolism. Below 37 C there is a progressive decrease both in the rate of production of prostacyclin (PGI,)1 and in its antiaggregatory action on human platelets2 3; the enzymatic degradation of adenosine diphosphate (a stimulator of platelet aggregation) is probably also reduced, while vasospasm induced by 5-hydroxytryptamine (5HT) is accentuated.4 Professor Keatinge and others could not show any changes in plasma thromboxane B2 (TXB2) concentrations between cold and control experiments. Nevertheless, although cooling increases the resistance of platelets to PGI.,2 3 it also impairs the production of TXB, by platelets.2 Platelet 5HT release, however, can, under certain circumstances, be essentially unaltered by cooling.5 Hence, after surface cooling, platelets may release more 5HT, which is a vasoconstrictor and stimulator of platelet aggregation, without necessarily raising plasma TXB2 levels. The suggestion by the authors that increased catecholamine release may be implicated in some of the changes observed is of great interest. One
BRITISH MEDICAL JOURNAL
VOLUME 290
consequence of this would have been a rise in plasma non-esterified fatty acid (NEFA) concentrations due to catecholamine mediated lipolysis. Raised non-esterified fatty acid concentrations promote platelet activation in man and induce thrombosis and vascular damage in animals.6 8 They also impair the synthesis," and decrease the stability,6 of PGI2 and also reduce vascular adenosine diphosphate degrading activity.9 Thus, platelet-prostaglandin interactions are extensively modified by cooling in ways that would not only support the observations of the increased incidence of coronary and cerebral thrombosis in winter but might also account for the concomitant increased incidence of femoral and brachial emboli'° and for the well documented association between cooling and
5 JANUARY 1985
75
infarct of rapid onset after exposure to cold, though additional factors are always possible. The suggestion by Dr Mikhailidis and others that prostacyclin production might be reduced by a fall in body temperature or non-esterified fatty acids is also interesting. We did investigate a possible role for prostacyclin by measuring plasma concentrations of 6-keto-PGF,,, but in most subjects these levels were unrecordable, both after cold exposure and in control conditions. There was no important fall in body core temperature, so that local temperature would have had an opportunity to affect prostacyclin production and destruction only in the limbs.
symptoms with that of subjects who did not have symptoms is feasible. A study of the relation between the incidence of hysterectomy in a given country and the decrease in incidence of cervical cancer might uncover a striking result. The reduction in death rates from cervical cancer is probably largely related to screening programmes. It is equally probable that many women have been harmed as a consequence of the screening process. The major need is not a reorganisation of the records and computerisation but a careful objective review of the potential risks and benefits of cervical screening based on at least limited clinical trials, preferably on an interW R KEATINGE national basis, to approximate the balance Department of Physiology, between the two. A potential benefit of the The London Hospital Medical College, current revival of interest in cervical screening London El 2AD might be a searching examination of its basic risks and benefits. EUGENE D ROBIN Failure of the cervical cytology screening programme Stanford University School of Medicine,
episodes of Raynaud's phenomenon, angina pectoris, and asthma. These interactions may also play a part in the development of multiple thrombotic lesions and platelet trapping associated with systemic cooling (hypothermia).2 3 D P MIKHAILIDIS J Y JEREMY SIR,-Current articles and correspondence M A BARRADAS (6 October, 3 November, 24 November 1984) P DANDONA concerning cervical Papanicolaou screening do not grapple with the most fundamental Department of Chemical Pathology and Human Metabolism problems. These problems include, firstly, R A HUTTON the failure to accumulate data on the number Haemophilia and Haemostasis Unit, and fate of subjects with false positive smears. Department of Haematology, Royal Free Hospital, This is critical. The patient with a false London positive cervical smear is the principal (and 1 Jeremy JY, Mikhailidis DP, Hutton RA, Dandona P. usually hidden) victim.' All such patients Effect of cooling on prostacyclin-induced increase in intraplatelet cyclic AMP and on prostacyclin suffer emotional trauma. Each follow up production. ClGn Sci 1982;62:42P. 2 Hutton RA, Mikhailidis DP, Bernstein RM, Jeremy diagnostic approach is associated with its own JY, Hughes GRV, Dandona P. Assessment of risks, as is each therapeutic approach. For platelet function in patients with Raynaud's example, hysterectomy, a common form of syndrome. J Clin Pathol 1984;37:182-7. 3 Mikhailidis DP, Hutton RA, Jeremy JY, Dandona P. management, is associated with a mortality Cooling decreases the efficiency of prostaglandin of two deaths/10002 and a morbidity of inhibitors of platelet aggregation-a factor of possible relevance in cold-induced pathology. 350/1000 operations.3 The specificity of Microcirculation 1983 ;2 :413-23. 4 Vanhoutte PM, Shepherd JT. Effect of temperature Papanicolaou smears may be as low as 0-01, on reactivity of isolated cutaneous veins of the dog. so that extraordinarily large numbers of AmJ Physiol 1970;218:187-90. 5 Valdorf-Hansen JF, Zucker MB. Effect of tempera- normal women are being subjected to needless ture and inhibitors on serotonin-'4C release from risks. human platelets. Am J Physiol 1971;220:105-11. Secondly, there is no adequate database for 6 Mikhailidis DP, Mikhailidis AM, Barradas MA, Dandona P. Effect of non-esterified fatty acids on establishing the risks versus benefits of the the stability of prostacyclin activity. Metabolism various diagnostic and therapeutic alternatives 1983 ;32:717-21. 7 Jeremy JY, Mikhailidis DP, Dandona P. Simulating pursued after an abnormal smear is reported. the diabetic environment modifies in vitro prosta- A host of diagnostic and therapeutic apcyclin synthesis. Diabetes 1983;32:217-21. 8 Mikhailidis DP, Hutton RA, Jeremy JY, Dandona P. proaches is used by individual physicians Hypothermia and pancreatitis. 7 Cln Pathol largely on the basis of opinion. Effective 1983 ;36 :483-4. 9 Barradas MA, Mikhailidis DP, Dandona P. The screening requires that effective management effect of non-esterified fatty acids on vascular for the disease be available.4 This is simply ADP-degrading enzyme (ADPase) activity. Clin not true for subjects with abnormal smears. Sci 1984;66:72P. 10 Clark CV. Seasonal variation in incidence of brachial The result is that, at least in the United States, and fenmoral emboli. Br MedJ 1983;287:1109. an abnormal smear subjects the patient to the ** *Professor Keatinge replies below.-ED, hazards of an untracked jungle, often for a lifetime. BMY. Thirdly, screening programmes have failed to change with the emergence of new knowSIR,-Sir Reginald Murley raises the question ledge. For example, current views of cancer of the cause of myocardial infarct or angina biology suggest that not every cancer clone induced within a few minutes by exposure to will develop to established cancer. It is cold. The usual explanation given for cold difficult to escape the conclusion that many induced angina is that reflex sympathetic drive more patients are treated because of abnormal to the heart and vasoconstriction increase smears than would have developed invasive cardiac work to the point at which a narrowed cancer.5 coronary circulation can no longer meet the Fourthly, there has never been an adequate increased metabolic demands of the myo- prospective clinical trial to evaluate the risks cardium. Sir Reginald's alternative suggestion versus the benefits of widespread screening in that cooling of the heart and brain may lead to normal women. As a result, recommendations local arterial thrombosis is interesting but are based on inferential and inadequate would probably call for greater local cooling considerations. of these organs than usually occurs during It may be too late to develop such trials, brief local cooling of the chest or head by cold but it is not too late for obtaining important air or during 30 minutes' general exposure to data. A comparison of the fate of women even very cold air. I suspect that in most who did not undergo Papanicolaou testing instances reflex factors, together with in- with those who did is feasible. A comparison creased circulating platelets and red cells, of the outcome of subjects whose cancer was account for cases of angina and myocardial detected after the development of uterine
Stanford, California 94305, USA
1 Robin ED. Matters of life and death: risks versus benefits of medical care. New York: W H Freeman, 1984. 2 Howkins J, Williams D. Total abdominal hysterectomy: 1000 consecutive unselected operations.
J7ournal of Obstetrics and Gynaecology of the British Commonwealth 1963 ;70 :20-8. 3 White SC, Wartel LJ, Wade ME. Comparison of abdominal and vaginal hysterectomies: a review of 600 operations. Obstet Gynecol 1971;37:530-7. 4 World Health Organisation. Mass health examination. Public Health Pap 1971;No 45. 5 Cochrane AL, Holland WW. Validation of screening procedures. Br Med Bull 1971;27:3-8.
SIR,-The DHSS recommendations may appear limited, as claimed by Dr Ann McPherson (24 November, p 1452) but, if properly implemented, would be a great improvement on current practice. Paradoxically the failures reported by Dr Jocelyn Chamberlain (6 October, p 853) will probably be exacerbated, unless laboratory staffing is improved. An inquiry by the British Society for Clinical Cytology provided the following information from over 150 laboratories in England, Wales, Scotland, and Belfast. The average increase in the 1984 gynaecological workload in January to April was 15% over that in 1983. In Wales, Yorkshire, and the Oxford region it was 20%. In one Yorkshire laboratory it was 79%. In 30 laboratories there is a backlog of more than four weeks, seven having a backlog of more than 10 weeks. Many other laboratories with a backlog of three weeks or less are under increasing stress from the rising workload. Some have had to curtail greatly the number of cervical smears from their screening programmes. At present the number of women over the age of 35 years who are being screened diminishes with increasing age. In Cardiff we found that the numbers dropped from 92% of the female population at the age of 25-29 years to 26% at the age of 65-69 years.' That it is possible to improve on this has been shown by Standing and Mercer, who screened 940 of women aged 36-64 years in their practice.2 This produced an initial increased workload of 459 smears, which, with repeats on abnormal smears, would give a total of about 500 smears. At such a rate South Glamorgan would provide 20 000 smears in addition to the current annual load of about 30 000 smears. If spread over five years this additional load would be 4000 smears a year. It is doubtful whether the screening of women under 35 years could be reduced sufficiently to compensate for this extra load.
