|
|
| |||||||||||||||||||||||||||||||||||||||
Background An earlier prospective study reported an association between high levels of plasma renin activity (as measured by the renin-sodium profile) and the incidence of myocardial infarction in patients with hypertension. We have investigated the relation between plasma renin activity and ischemic heart disease in the Northwick Park Heart Study.
Methods The study included 803 white men 40 to 64 years of age selected from industrial workers in London. Plasma renin activity and established risk factors for ischemic heart disease were measured at entry, which was between 1972 and 1978. Ascertainment of the primary clinical end points of fatal or nonfatal myocardial infarction and sudden death from coronary causes was carried out until the end of 1991.
Results In an analysis of the 86 first coronary events, we found an independent relation between higher systolic blood pressure and coronary end points (relative risk per 1 SD increase in blood pressure, 1.47; 95 percent confidence interval, 1.16 to 1.85; P<0.001), but no relation between plasma renin activity and coronary end points (relative risk per 1 SD increase in the level of plasma renin activity, 1.04; 95 percent confidence interval, 0.84 to 1.30). In the 242 men who had hypertension of a degree similar to that of the subjects in the earlier prospective study of the renin profile, and in whom 44 of the 86 coronary events occurred, the relative risk of those in the highest as compared with the lowest third for plasma renin activity was 1.26 (95 percent confidence interval, 0.63 to 2.56).
Conclusions Our results suggest that there is no association between plasma renin activity and myocardial infarction or sudden death from coronary causes, at least in normotensive men.
In a prospective study of patients with hypertension, Alderman et al.3 reported a significant relation between a high renin-sodium profile and the occurrence of myocardial infarction, the renin-sodium profile being defined as high, normal, or low according to the plasma renin value for a given level of urinary sodium excretion. We now report findings that show no relation between plasma renin activity and major ischemic heart disease in a predominantly normotensive population.
Methods
Study Population
The study involved white male day-shift workers 40 to 64 years old, who were recruited into the Northwick Park Heart Study between 1972 and 1978. The main purpose of the study was to investigate the role of hemostatic function in the pathogenesis of ischemic heart disease. Full details are available elsewhere4. All the men in a number of industrial groups in northwest London who had been employed for at least a year (there being no other exclusion criteria) were invited to participate, and about 80 percent accepted4. Men on daytime shifts were studied on two consecutive days, between 8 and 11 a.m. On the first day, personal and medical histories were obtained. Smokers were defined as those who had been smoking at least one cigarette a day for a year, or the equivalent in pipe or cigar smoking. Ex-smokers were all those who had smoked in the past. The body-mass index was defined as the weight in kilograms divided by the square of the height in meters. Using a standard mercury sphygmomanometer, we measured blood pressure twice to the nearest 2 mm Hg. The first reading was taken after the subject had been interviewed for about 20 minutes, and had lain on a couch for about 5 minutes. The second was taken about five minutes after the first, the subject having remained recumbent. Results are based on the first of these readings (the relation between blood pressure and plasma renin activity was very similar, however, for both readings). On the second day, a blood sample was obtained after the fasting subject had been sitting for about five minutes. This sample was used for plasma renin assays, and results were available for 90 percent of those who participated. The salt intake of the study subjects had been unrestricted. A clinical diagnosis of hypertension had previously been made in 51 of the 883 men who were screened, and nearly all were still being treated. They and the 29 men who had had definite (24) or possible (5) myocardial infarctions were excluded from this report, leaving 803 daytime workers who had no diagnosis of hypertension, were receiving no treatment for hypertension, and had no history of myocardial infarction, although 44 of them had angina pectoris.
Plasma Renin Assay
The method we used to measure plasma renin activity has been described in detail elsewhere2. Whole blood was drawn into tubes containing citrate. The plasma was then separated, and aliquots were stored in liquid nitrogen within four hours of venipuncture. A subsidiary study showed no difference in plasma renin activity between samples handled this way and samples separated and frozen immediately after venipuncture. The aliquots were sent in solid carbon dioxide from Northwick Park to St. Mary's Hospital Medical School in 12 separate batches over a two-year period. They were stored at -20 °C as received and were assayed over the next few weeks.
Other Laboratory Measurements
Serum cholesterol levels were measured routinely, as previously described4. During part of the recruitment phase, samples were also analyzed for serum sodium (213 participants), urea nitrogen (212), potassium (215), and uric acid (655). Urinary sodium excretion was not measured.
Follow-up and End Points
Major coronary events were defined as fatal or nonfatal myocardial infarction and sudden death from coronary causes. A total of 13 men who had nonfatal events and later died of myocardial infarction have been classified among the men with fatal events (in the Cox proportional-hazards analyses, virtually identical results were obtained when their previous nonfatal events were considered). The follow-up period for mortality was until the end of 1991. The records of all 803 men were flagged in the National Health Service Central Register. This system ensures the automatic provision of copies of death certificates for all study subjects. Further information was then obtained from general practitioners, hospitals, and coroners. Autopsy findings were available for most of those who died. To ascertain nonfatal myocardial infarctions, we invited the surviving men to undergo reexamination between 1978 and 1984; 91 percent were seen. For those not reexamined, questionnaires were sent both to the subjects themselves and to their general practitioners, as a result of which further details were obtained from any hospitals they used. In only 2.8 percent was follow-up information not available. After the reexaminations from 1978 to 1984, a second round of ascertainment of nonfatal events was carried out between 1986 and 1991 through further questionnaires to the subjects and their general practitioners, and informative replies were received for 89 percent of the men. In the remaining 11 percent, plasma renin activity at recruitment was similar to that in the subjects for whom follow-up information was available. Since the ascertainment of nonfatal events did not extend to the end of 1991 for all study subjects, the period of observation was somewhat shorter than that for fatal events. When all the available clinical and pathological information had been collected, decisions about the occurrence of myocardial infarction or sudden death from coronary causes were made according to the criteria of the World Health Organization5 by independent assessors to whom the results of the laboratory investigations were not available.
Statistical Analysis
We used log-transformed values in all our analyses, since these provided a more nearly normal distribution than the untransformed values and also tended to equalize the error variance at different levels of plasma renin activity. All values for plasma renin activity were adjusted for age. The means reported are geometric, and 95 percent ranges (rather than standard deviations) are given as measures of dispersion. Unpaired t-tests were used to test for differences in means, and chi-square tests for differences in proportions6. The Cox proportional-hazards method was used for multivariate analyses to take account of significant correlations between plasma renin activity and other independent variables and of varying intervals between recruitment and end points7. All P values are two-tailed.
Results
Table 1 summarizes the characteristics of the 803 subjects at entry, according to clinical outcome. There were 86 events, of which 35 were nonfatal and 51 fatal. Eight of the events (four nonfatal and four fatal) occurred in the 44 men with angina. The data confirm that those in whom events (particularly fatal events) subsequently occurred had higher mean body-mass indexes, serum cholesterol levels, and blood pressures than those who did not4. The relation of smoking to the occurrence of events was also as expected, but it was not statistically significant. There were no significant differences between the groups in serum sodium, urea nitrogen, or uric acid levels. Serum potassium levels were higher in those who had nonfatal events than in those who remained event-free.
|
|
|
The results shown in Table 3 were not materially different when the analysis was confined to fatal episodes. Thus, event rates per 1000 years of observation in the lowest, middle, and highest thirds of the blood-pressure distribution were 3.10, 2.99, and 7.48, respectively, and for plasma renin activity they were 3.99, 3.83, and 4.90.
Among the subjects in the Northwick Park Heart Study group, 242 (30 percent) satisfied the blood-pressure criteria of the study by Alderman et al.3 -- i.e., systolic pressure of more than 160 mm Hg or diastolic pressure of more than 95 mm Hg -- and they had 44 of the 86 major coronary events. The relative risk of an event in those in the highest as compared with the lowest third of the plasma-renin-activity distribution was 1.26 (95 percent confidence interval, 0.63 to 2.56). Comparing the top 12 percent in the plasma-renin-activity distribution with the bottom 32 percent (as in the corresponding analysis of those in the high- and low-renin-profile groups in the study by Alderman et al3.), we found a relative risk of 1.88, but the 95 percent confidence interval, 0.78 to 4.54, did not overlap the point estimate of 5.3 reported by Alderman et al3.
Table 4 summarizes the Cox proportional-hazards analysis. The independent relation between blood pressure and coronary events was significant, whereas there was no evidence of a relation in the case of plasma renin activity. For diastolic blood pressure, the relative risk was 1.43 (95 percent confidence interval, 1.13 to 1.81).
|
Our results reveal no relation between plasma renin activity and the incidence of major coronary events in normotensive men. A relation in those whose blood pressures were in the highest third of the distribution at entry cannot be excluded but is not clearly supported by our findings. It is therefore necessary to consider possible explanations for the apparent difference between our study and the study of Alderman et al.,3 which found a higher risk of myocardial infarction in patients with hypertension and high renin-sodium profiles.
The study by Alderman et al.3 included black and white men and women, so their population was substantially more heterogeneous than ours. In their study group, 40 percent of the black men had low renin profiles, as compared with 18 percent of the white men. It seems possible that blacks and women, with lower rates of myocardial infarction, are overrepresented in the group with low renin profiles and that those with vascular damage are overrepresented in the group with high renin profiles. Alderman et al. corrected as far as possible for the first of these imbalances through their statistical model, but they may not have corrected for the second. The contrast in the proportions of black men and white men in the different renin-profile groups was due almost entirely to differences in plasma renin activity; the mean sodium excretion values were very similar. This observation emphasizes that plasma renin activity is considerably more important than the sodium excretion value in establishing the renin-sodium profile.
As Alderman et al.3 acknowledged, the degree to which the later treatment of hypertension may have influenced the onset of myocardial infarction (and other events) after recruitment to their study is unknown. Only 58 subjects in the Northwick Park Heart Study (7 percent) were treated for hypertension between recruitment (when, by definition, they had not received antihypertensive agents) and the first follow-up examination.
The plasma renin assays in the two studies differed to some extent. It has been suggested that acid cryoactivation in the assay used in our study may result in unpredictable and varying conversion of prorenin to renin3. Even if this occurred, however, it is unlikely to have done so to different degrees in those who went on to have major coronary events and those who did not. Moreover, our assay is a standard method for measuring plasma renin activity, and the relation between a range of characteristics (such as age, sex, and ethnic group) and plasma renin activity is very similar for our assay and the assay used by Alderman et al.3. It is therefore unlikely that differences between the assays account for the difference between the results of the two studies.
In the Northwick Park Heart Study, urinary sodium excretion was not measured, but we doubt that this explains the difference in findings. First, it is unlikely that more than a small number of the men in the study had urinary sodium levels outside the range within which the relation between plasma renin activity and urinary sodium is linear -- i.e., about 50 to 250 mmol per day. Second, it is probable that changes in sodium excretion become important only in more severe hypertension,8 which the men in the Northwick Park Heart Study did not have. Third, we have already commented that sodium excretion appears to play little part in determining the renin profile in a population of mixed ethnic groups.
There were considerably more events for analysis in the Northwick Park Heart Study (86) than in the study by Alderman et al (27). For the most reliable end point, fatal ischemic heart disease, our study included 51 events, as compared with only 9 in the study by Alderman et al3.
The most obvious clinical difference between the two studies is that Alderman et al.3 recruited subjects with known hypertension, whereas we recruited from working populations characterized by a full range of blood pressures. Furthermore, those with known or treated hypertension at entry were deliberately excluded from the analyses. However, our results, even in those who met the blood-pressure criteria of Alderman et al.3 (and still involving a substantially larger number of events than their study), did not establish a clear relation between plasma renin activity and coronary events, and the incidence of such events increased in the highest third of the blood-pressure distribution (Table 3) at all levels of plasma renin activity. On the other hand, it was among men in the Northwick Park Heart Study whose systolic pressures were in the highest third of the distribution that there may have been an association between plasma renin activity and later events (Table 3) (although there was no evidence of an interaction [P = 0.93] between blood pressure and plasma renin in their relation to coronary events). The results of the two studies are thus not necessarily incompatible. Given the potential hazards of subgroup analyses, it may be that there is no relation between plasma renin activity and the incidence of ischemic heart disease among normotensive subjects. Among some of those with high blood pressure (those in the highest third of the distribution in the Northwick Park Heart Study), however, the usual regulatory, inverse relation between blood pressure and plasma renin activity suggested by our previous report2 may be weakened. Renal vascular damage involving the afferent arteriole may then cause the level of plasma renin activity to rise,9 thus establishing a relation between increasing levels of plasma renin activity and the risk of a major coronary event. The absence of any correlation between systolic blood pressure and plasma renin activity in the men in the highest third of the blood-pressure distribution in the Northwick Park Heart Study is compatible with this explanation, as is the tendency for those who had events within five years of entry to have elevated levels of plasma renin activity as well as elevated blood pressure. Our results show, however, that it is the independent relation between elevated blood pressure and ischemic heart disease that predominates.
We are indebted to Dr. Katharine Garrow for assessing the episodes of heart disease ascertained since our previous report.
Source Information
From the Medical Research Council Epidemiology and Medical Care Unit, Wolfson Institute of Preventive Medicine, Medical College of St. Bartholomew's Hospital, Charterhouse Sq., London EC1M 6BQ, United Kingdom, where reprint requests should be addressed to Professor Meade.
References
| |||||||||||||||||||||||||||||||||||||||
Related Letters:
Plasma Renin Activity and Ischemic Heart Disease
Alderman M. H., Sealey J. E., Laragh J. H., Meade T.W., Cooper J.A., Peart W.S.
Extract |
Full Text
N Engl J Med 1994;
330:506-507, Feb 17, 1994.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
