Blood pressure response during exercise as a prognostic factor

Most of the data used to evaluate blood pressure as a risk factor for future cardiovascular morbility and mortality are based on blood pressure recorded during a rest situation. When necessary precautions are taken, such as a correct cuff size related to the arm circumference and a mercury sphygmomanometer, several studies have shown that the agreement between intra-arterial recordings and cuff method values is excellent, provided that Korotkoff phase V (disappearance of the sound) is used for diastolic pressure [1,2]. Korotkoff phase IV usually gives an overestimation of diastolic pressure by 5–10 mmHg. Although blood pressure during rest in large groups of subjects strongly reflects future cardiovascular risk, it is well known that blood pressure during rest often is very labile and easily influenced by external factors, such as noise, sudden talking, etc., as excellently observed and described by Riva-Rocci in 1896 [3]. Investigators of blood pressure are strongly recommended to read this classic paper, which was translated into English and republished 100 years later by Zanchetti and Mancia in the Journal of Hypertension [4]. They concluded, ‘the detailed recommendations that he gave on how to measure blood pressure are unsurpassed by any of the more recent guidelines provided by scientific societies or organizations and found in specialist's hand books'. Having observed how easily blood pressure is influenced by external stimuli during my 30 years of intra-arterial measurements, I could not agree more! The common concensus is that blood pressure recorded during exercise is less influenced by nervousness and external factors such as noise, talking, etc. When the circulatory system is challenged by a large increase in cardiac output and peripheral blood flow, ‘true’ hypertension could be unveiled and distinguished from a temporary blood pressure increase during rest [2]. In order to obtain accurate systolic and diastolic pressure values during exercise, intra-arterial recordings are necessary [5]. Although systolic blood pressure may be recorded reasonably well during bicycle exercise by the cuff method, diastolic values may be grossly over- or underestimated. This is also true for the modern automatic blood pressure recorders with a cuff around the upper arm, and which rely on detection of Korotkoff sounds or on oscillometry [6]. Several large studies based on ergometer bicycle and non-invasive recordings of blood pressure during different exercise levels have demonstrated that augmented blood pressure increase during exercise predicts increased future cardiovascular risk. Thus, in a study from Norway [7], the outcome of 1999 apparently healthy men, aged 40–59 years, who were investigated from 1972–75, was ascertained after 16 years to determine whether systolic blood pressure measured with the subjects in the sitting position during a bicycle ergometer exercise test added prognostic information on cardiovascular mortality beyond that of casual blood pressure measured after 5 min of rest. During a total follow-up time of 31 984 patient years, 278 patients died, 150 of these due to cardiovascular causes. The results showed that peak systolic blood pressure during 6 min on a starting workload of 600 k.p.m./min was strongly related to cardiovascular mortality (relative risk = 1.5–95% confidence interval) [7]. The influence of blood pressure at 600 k.p.m./min was so strong that the predictive value of resting casual blood pressures became non-significant when these were analysed as continuous variables, also including exercise blood pressure as a covariant. In a French study in 4907 middle-aged men followed for 17 years [8], total mortality was strongly related to the exercise-induced increase in systolic blood pressure, independently of resting blood pressure. Several other studies based on non-invasive methods also concluded that exercise-induced increases in blood pressure during an ergometer bicycle test predict future cardiovascular complications and mortality [9,10]. However, in an 11-year follow-up of 143 male hypertensive patients, Fagard et al. [11] observed that intra-arterially recorded systolic and diastolic blood pressure at 50 W and at 50% of peak exercise did not independently predict mortality and the incidence of cardiovascular events. Fagard points out that the blood pressure response is different when the workload is increased sharply (versus a progressive increase during a multistage exercise test) as conventionally used for clinical purposes. In studies of blood pressure during work, the type of exercise (gradual versus sudden or isometric versus dynamic) is crucial for the blood pressure response [12]. The use of the ergometer bicycle allows standardization of the procedure, such as exercise time and an exact determination of the exercise load. Furthermore, with correct use of the cuff method, the systolic values are reasonably accurate when the arm with the cuff is kept steady [6]. In order to avoid the problems with noise partly caused by arm movements during exercise testing, some investigators have measured blood pressure ‘immediately after exercise'. However, this is a dubious method because, as soon as the exercise is stopped, intra-arterial recording usually shows a very rapid fall in blood pressure in the following seconds and minutes. Measurement of blood pressure during this period (when the pressure is on a falling stage) by a non-invasive technique is very inaccurate and should not be used [2]. The ergometer bicycle test takes time and obviously necessitates appropiate equipment. An alternative, less costly and less time-consuming method is to use the old Master step test, as described by Tsumura et al. [13] in this issue of the journal. In a large study, 6557 Japanese men, aged 35–63 years, with a resting blood pressure < 140 mmHg systolic and < 90 mmHg diastolic, were submitted to Master step test shortly after blood pressure had been recorded during rest. The Master step consists of two podium-like steps, with each step being 23 cm in height. The test lasted 3 min, but the number of trips to be performed depended on the age and weight of the subject. Normotension was defined as a blood pressure < 130/85 mmHg and hypertension as a blood pressure > 160/95 mmHg. Individuals with blood pressures in between these values were classified as ‘high normal'. Trained nurses took all the blood pressure readings, and an average of two readings was used to provide the blood pressure value. Immediately after the test step was performed, the patients rested supine for 4 min and the blood pressure was then recorded. The blood pressure 4 min after exercise was classified in quintiles of systolic and diastolic pressure. The patients were followed for 5–27 years, providing data for 63 696 person/years in total. During the observation time, 660 men became hypertensive. It was demonstrated that subjects with the highest systolic blood pressures 4 min after the Master step test had an increased risk of future hypertension. When those with the highest quintile (130–176 mmHg) were compared with those in the lowest quintile (80–110 mmHg), the relative risk was 11.72 (95% confidence interval 8.17–16.80). For every 10 mmHg increase in systolic blood pressure at 4 min after exercise, the multiple-adjusted relative risk for hypertension was 1.99. This study also included other risk factors, such as age, body mass index, alcohol consumption, physical activity, smoking and glucose levels, and the calculation of risk was adjusted accordingly to allow for these. Somewhat surprisingly, blood lipids were not studied (or at least were not reported in the paper), probably reflecting the observation that, until recent years, blood cholesterol levels in Japan and many other Asian countries have been very low and were not considered to be of great prognostic importance. However, the smoking habits in the study population are most alarming, with 61.5% being current smokers and only 20% being never-smokers. If this trend continues and these Japanese men were also to adopt a Western diet with more sugar and fat (which has unfortunately taken place in other Asian countries such as South Korea), an explosion in the incidence of coronary heart disease will most likely be the result. Subjects becoming hypertensive are particularly at high risk. The Master step test with blood pressure recordings after 4 min during supine rest could possibly help to identify those subjects in the highest risk zone, and vigorous prophylactic measures should be introduced. A new follow-up of this interesting study after 5–10 years should be recommended and encouraged.