Isolated Diastolic Hypertension
A Favorable Finding Among Young and Middle-aged Hypertensive Subjects
Abstract To identify pretreatment characteristics associated with subsequent myocardial infarction in young and middle-aged previously untreated hypertensive individuals, we examined the experience of 1560 participants in a work-site hypertension control program who were younger than 60 years. Subjects were categorized by initial blood pressure as having isolated diastolic hypertension (<160/≥90 mm Hg, n=965) or combined systolic and diastolic hypertension (≥160/≥90 mm Hg, n=595). During 4.5 years of follow-up, there were 24 myocardial infarctions, yielding an overall incidence of 3.89 per 1000 person-years. Subjects with systolic/diastolic hypertension were older, had higher cholesterol and blood sugar levels, and included more smokers and people with left ventricular hypertrophy on electrocardiogram than those with isolated diastolic hypertension. Age-adjusted incidence rates for myocardial infarction were 5.20 and 2.21 per 1000 person-years in systolic/diastolic hypertension and isolated diastolic hypertension, respectively, and the relative risk of systolic/diastolic hypertension was 2.31 (95% confidence interval, 1.29-4.15). Among subjects with isolated diastolic hypertension, no myocardial infarction occurred in those with systolic pressure less than 140 mm Hg. Cox regression analysis including other known risk factors showed that pulse pressure, as a continuous variable (hazards ratio, 1.54; 95% confidence interval, 1.08-2.20), and type of hypertension, ie, systolic/diastolic hypertension versus isolated diastolic hypertension (hazards ratio, 2.11; 95% confidence interval, 1.08-4.13), were independently associated with myocardial infarction. These results suggest that young and middle-aged treated hypertensive individuals with normal pretreatment systolic pressure enjoy a more favorable prognosis than do those with systolic elevation. This data, based on the outcomes of treated subjects, cannot be extrapolated to define what the experience of individuals with isolated diastolic hypertension might be in the absence of drug therapy.
Coronary heart disease morbidity and mortality are directly and linearly related to BP level.1 Among the multiple cardiovascular diseases that are increased in individuals with elevated BP, MI most commonly afflicts American adults of both sexes, all ages, and all socioeconomic and ethnic groups.2 3 4 5
Both high systolic and diastolic pressures are related to MI. In the elderly, ISH is more common and is associated with profoundly increased cardiovascular morbidity and mortality.6 By contrast, ISH is rarely found in younger subjects, and increased diastolic BP is the commonly accepted guide to cardiovascular risk.7
We now report that young and middle-aged (age <60 years) previously untreated hypertensive individuals, whose only BP abnormality was a diastolic elevation, had a better prognosis than did those with both systolic and diastolic pressure elevations. In this prospective study of well-treated young and middle-aged subjects, only high pretreatment systolic pressure and wide pretreatment PP were associated with an increased risk of subsequent MI. Moreover, in the absence of systolic BP elevation (<140 mm Hg), the increased level of pretreatment diastolic pressure was not associated with MI.
Subject Evaluation and Program Description
Study subjects were employee participants in an occupationally based, union-sponsored, systematic hypertension control program who began therapy between May 1973 and December 1992.8 9 Hypertension was defined as an initial untreated systolic BP greater than or equal to 160 mm Hg and/or diastolic BP greater than or equal to 90 mm Hg. In addition, subjects eligible for this study had no previous MI or stroke, were untreated at entry, and were observed for at least 6 months on antihypertensive drugs.
This study was further limited to subjects’ experience through their 60th birthday. Events of interest, therefore, all occurred before age 60. Demographic and behavioral characteristics, medical histories, physical status, routine clinical chemistry measures, and ECG findings were recorded before therapy. Thereafter, routine data on drug treatment and interval morbidity were collected during clinic revisits and at an annual physical examination where entry evaluation was repeated. A modified, stepped-care treatment approach starting with hydrochlorothiazide was later expanded to include propranolol in 1978 as an alternative drug of first choice. After 1986 first-line drugs included newer adrenergic blockers, calcium channel blockers, and angiotensin-converting enzyme inhibitors.
All BP measurements were taken by a nurse with a standard sphygmomanometer. At each visit the average of the second and third readings from a set of three seated, right-arm measures (5 to 10 minutes apart) was that day’s record. Pretreatment BP (initial BP) was recorded at screening. Final BP (intreatment BP) was the average of all available BP readings for 6 months (maximum) preceding a cardiovascular event or, in those subjects who did not experience an event, for the 6 months before and including the last visit. PP was the difference between systolic BP and diastolic BP. MAP was calculated as diastolic BP plus one third PP.
Of 6506 subjects, 1586 met eligibility criteria. These 1586 subjects were categorized by initial BP level into three groups: IDH (BP <160/≥90 mm Hg, n=965), SDH (BP ≥160/≥90 mm Hg, n=595), and ISH (BP ≥160/<90 mm Hg, n=26). Because only 26 (1.6%) of these young and middle-aged subjects met criteria for ISH, they were excluded from further analysis. Thus, the final number of study subjects was 1560. IDH subjects were further divided into BSDH (BP 140-159/≥90 mm Hg, n=698) and NSDH (BP<140/≥90 mm Hg, n=267).
Morbidity and Mortality
Morbidity and mortality were ascertained by nurses who systematically monitored reports by subjects, family members, or friends of any hospitalizations or deaths. Hospital and medical charts and death certificates were reviewed by the research staff for confirmation of events.
Illnesses and deaths were classified with the use of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM). Clinical events of interest were MI (ICD-9-CM code 410), angioplasty (ICD-9-CM code 414.22), and coronary bypass surgery (ICD-9-CM code 414.21). The other cardiovascular events were also coded (ICD-9-CM codes 390-459). For subjects with more than one event during follow-up, only the first event was included in the analysis. Events that occurred within 6 months after a dropout were collected and counted. Events that occurred more than 6 months after a nonmedical dropout were not included.
Among the 1560 study subjects there were 22 MIs and 2 cases of bypass/revascularization. For the 22 MI subjects, confirmation of events by hospital record or death certificate (or both) was made in 20 of the 22. For the remaining 2 subjects, information was obtained from treating physicians, family members, or union records.
We computed descriptive statistics based on initial subject characteristics to compare the study subjects categorized by type of hypertension. Means and SDs for BP measurements in continuous variables were obtained for each type of hypertension, and differences between types were tested for statistical significance with Student’s t test.
The length of follow-up for each person was from the first screening BP to the date of first event or date of last visit for nonevent cases. Years of follow-up and numbers of subjects in each group were used to calculate the person-years of follow-up. Age-adjusted incidence rates of MI (per 1000 person-years) were computed according to type of hypertension. The relative risk and 95% CI of different types of hypertension were calculated and tested for statistical significance. Cox proportional-hazards models were applied to assess the independent association of pretreatment BP measures and type of hypertension with the occurrence of MI during the treatment after other potential confounding or competing factors had been accounted for. In this case, different models were constructed for different types of BP measures (systolic BP, diastolic BP, PP, and MAP).
Overall, subjects were predominantly male (73.8%), with the following pretreatment means: age, 45.4 years; BP, 156/104 mm Hg; PP, 53 mm Hg; MAP, 121 mm Hg; and body mass index, 28.8 kg/m2. Of these subjects, 34.9%, 32.3%, and 30.4% were white, black, and Hispanic, respectively; 28.7% were current smokers; 10.1% had LVH on ECG; and 4.2% had a history of diabetes.
Subjects With MI and Nonevent Subjects
Subjects with MI (Table 1⇓) were significantly older and more likely to smoke and report diabetes than nonevent subjects. Body mass index, cholesterol, blood glucose, blood urea nitrogen, creatinine, race, and LVH on ECG were similar between the two groups.
Comparison of BP of MI and nonevent subjects showed that those with MI had significantly higher initial and final systolic pressures and initial MAP and sharply wider initial (10 mm Hg or 20%) and final PP. Initial and final diastolic pressures and final MAP did not differ.
Subjects With IDH and SDH
Subjects with SDH were older and had higher cholesterol and blood glucose than those with IDH (Table 2⇓). They were more likely to be female, have LVH on ECG, and be current smokers than IDH subjects. By definition, SDH subjects had higher systolic BP, but they also had higher diastolic BP (108.0 versus 101.3 mm Hg), wider PP (66.7 versus 43.0 mm Hg), and higher MAP (130.3 versus 115.6 mm Hg) than IDH subjects. The reductions of BP from the initial to final were 30.5 versus 11.3 mm Hg for systolic BP and 16.7 versus 11.2 mm Hg for diastolic BP among SDH and IDH subjects, respectively, and they were both significantly different.
Among subjects with IDH, those with BSDH (systolic BP 140 to 159 mm Hg) were slightly older (44.6 versus 43.2 years) and had significantly higher cholesterol (5.78 versus 5.59 mmol/L) and blood glucose (5.75 versus 5.54 mmol/L) than NSDH subjects (systolic BP <140 mm Hg). In addition to higher initial systolic BP (148.8 versus 132.6 mm Hg), BSDH subjects also had significantly higher initial diastolic BP (102.1 versus 99.1 mm Hg) and MAP (117.7 versus 110.3 mm Hg) and wider PP (46.7 versus 33.4 mm Hg) than NSDH subjects.
Morbidity and Mortality
During an average of 4.54 years of follow-up (range, 6 months to 17 years) there were 24 coronary heart disease events (22 MIs and 2 coronary angioplasty/bypasses), for an incidence rate of 3.89 per 1000 person-years. The follow-up years for SDH and IDH subjects were 4.46 and 4.59, respectively, and among IDH subjects the follow-up years for NSDH and BSDH subjects were 4.67 and 4.57, respectively. No statistically significant difference in number of years of follow-up were detected between these groups. The age-adjusted incidence rate of MI in SDH subjects was 5.20 per 1000 person-years, which was significantly higher than that in IDH subjects (2.21 per 1000 person-years, Table 3⇓). Of note is the fact that all MIs in the IDH group occurred among those 698 subjects with systolic pressure greater than or equal to 140 mm Hg, with an age-adjusted incidence rate of 3.11 per 1000 person-years. There were no MI events among the 267 subjects with systolic pressure less than 140 mm Hg.
During the follow-up period there were 11 strokes (ICD 430-438), 1 episode of congestive heart failure (ICD 428), 1 case of unstable angina (ICD 411.1), 15 with other cardiovascular diseases, 9 with cancer, and 77 with all other diseases. There were 1422 nonevent subjects. The age-adjusted incidence rate of total cardiovascular diseases in IDH and SDH subjects were 5.32 and 10.20 per 1000 person-years, respectively. Compared with IDH subjects, the relative risk of total cardiovascular events in SDH subjects was 1.92 (95% CI=1.30-2.84).
We assessed the independent association of different BP measures and types of hypertension (IDH and SDH) with MI by multivariate Cox proportional-hazards regression analysis after adjusting for age, sex, cigarette smoking, LVH on ECG, cholesterol, and blood sugar. PP, as a continuous variable, was independently associated with the occurrence of MI (β=0.029; hazards ratio=1.54; 95% CI=1.08-2.20). By contrast, other measures of BP (systolic BP, diastolic BP, and MAP), as continuous variables, when entered into the model instead of PP all failed to be associated with coronary heart disease. In addition, hypertension type, ie, IDH or SDH, as categorical variables when entered into the model did predict MI. Subjects with SDH were at significantly higher risk of MI (β=0.747; hazards ratio=2.11; 95% CI=1.08-4.13) compared with those with IDH (Table 4⇓).
Among this large group of young and middle-aged treated hypertensive subjects, IDH proved to be a favorable finding. By contrast, a wide pretreatment PP was strongly associated with the occurrence of MI despite successful treatment. Among subjects who had MI before age 60, pretreatment diastolic pressure was similar to that of subjects who did not experience events.
Moreover, when subjects with IDH were further stratified by systolic pressure above or below 140 but still less than 160 mm Hg, it was found that MIs occurred exclusively among those with borderline systolic hypertension. No subject with a systolic BP below 140 mm Hg experienced an MI regardless of the level of diastolic pressure (90 to 115 mm Hg).
Subjects with SDH were older and had higher cholesterol and blood glucose levels and were more likely to be female, smoke cigarettes, and have LVH on ECG than IDH subjects. All BP measures in the SDH group were higher than in the IDH group. Clearly, these factors help to explain the observed association of systolic BP with the outcomes noted. However, through multivariate analysis it was possible to adjust for these factors and still detect a strong independent association of hypertension type with the occurrence of MI.
Hypertension has usually been explored according to systolic and/or diastolic BP. Initially, the hazard of hypertension to the cardiovascular system was felt to derive principally from its diastolic component.10 However, the Framingham Study suggested that this was principally true for subjects younger than 45 years.7 11 For older subjects systolic BP was superior to diastolic as a predictor of cardiovascular mortality or morbidity. Data from other epidemiological studies seemed to confirm that systolic BP was more strongly associated with coronary heart disease death than was diastolic, but this association was generally demonstrated in older subjects. More recently, it has been found that in all age groups at the same level of diastolic pressure a higher systolic BP is associated with an increased risk of cardiovascular morbidity and mortality.5 12 13
It has also been noted recently that other reflections of BP measures, particularly PP, are related to cardiovascular outcome. A wide PP has recently been shown to be independently associated with an increased risk of MI morbidity and mortality.14 15 16 17
The importance of PP and high systolic pressure has not been explored previously in younger hypertensive subjects. Our present findings suggest that as in older hypertensive subjects, a wide PP and high systolic pressure are ominous signs in young and middle-aged individuals. The notable observation here is that a lack of elevated systolic pressure is a favorable finding.
It is logical to expect that all forms of cardiovascular disease would be more frequent in the presence of elevated systolic pressure. Higher systolic BP is a reflection of rigid, atherosclerotic arteries. Reduced arterial compliance is associated with declining elasticity of the aorta and other large arteries and characterizes atherosclerosis.6 18 19 20 This is reflected by an increase in systolic BP and a reduction in diastolic BP, thus widening PP.18 19 21 The increased systolic BP and lack of compliance tend to generate LVH, whereas a reduced diastolic pressure tends to reduce coronary artery blood flow. Both effects increase the risk for MI and are common in the elderly. Perhaps in young and middle-aged individuals an elevated systolic BP marks the presence of already developed, premature arterial atherosclerosis.
By contrast, subjects with IDH with normal systolic BP might be expected to have normal arterial compliance. Presumably, this would signify the absence of atherosclerosis and therefore a decreased likelihood of MI. Coronary artery blood flow may be improved with higher diastolic pressure. This could possibly contribute to the apparent cardiac protection enjoyed by these subjects. If this is the case, it is possible that in the absence of systolic elevation, diastolic BP reduction might not confer benefit. Of course, the experience in the present study sheds no light on this issue because no events occurred during treatment in the NSDH group.
Another explanation for the observed findings is that the recorded diastolic elevation was an artifact of BP measurement.22 23 If the auditory method ever overestimated actual diastolic pressure, the result would be a misclassification of subjects whose true intra-arterial pressure was within the normal range. Although suggested elsewhere,24 no data are available for examination of this possibility. However, in the present study the antihypertensive therapy had different hypotensive effects on the diastolic pressure of IDH subjects, as it did on those with SDH. This tends to support the notion that misreading of diastolic pressure may have occurred in some IDH subjects. Alternatively, misclassification may have occurred because BP recordings were all made at one visit. It is certainly possible that recordings made on multiple occasions might have produced different average or final BP values.
The higher cholesterol and more common LVH on ECG undoubtedly increased the risk of cardiovascular disease in SDH subjects.25 26 In fact, in multivariate analysis, these proven risk factors were not independently associated with MI. This might be explained by the high correlation of these phenomena with systolic BP in the young population.27 Perhaps in the face of an elevated systolic pressure, the incremental risk of elevated cholesterol might be less prominent. In fact, only age and cigarette smoking were additional independent risk factors among these young and middle-aged subjects.
Although isolated diastolic pressure elevation is common and apparently confers a much lower risk than does systolic elevation, the data here should not be taken to imply that treatment is unnecessary. Indeed, it may well be that such subjects simply respond particularly well to therapy. However, the entire clinical profile of these subjects suggests low risk. For such subjects the value of antihypertensive therapy may be quite low.28 The issue is not trivial. More than 16% of subjects (<60 years) drawn from this general working population met the strictest definition of IDH. They had no events during 5 years of follow-up. If a randomized clinical trial were to show that antihypertensive therapy was not of benefit (hazards equal to or exceed benefits) in these circumstances, many millions throughout the world could be spared long-term drug therapy.
In summary, these data suggest that in young and middle-aged hypertensive subjects, an isolated diastolic pressure elevation before treatment implies a more favorable outcome than is the case when elevations of systolic and diastolic pressure coexist. Indeed, in the current study no subject whose systolic pressure was below 140 mm Hg experienced an MI regardless of diastolic pressure level. By contrast, those with elevated systolic pressure and a wide PP were at substantially increased risk of MI, despite equivalent and satisfactory BP control.
Selected Abbreviations and Acronyms
|BSDH||=||borderline systolic with high diastolic hypertension|
|IDH||=||isolated diastolic hypertension|
|ISH||=||isolated systolic hypertension|
|LVH||=||left ventricular hypertrophy|
|MAP||=||mean arterial pressure|
|NSDH||=||normal systolic with high diastolic hypertension|
|SDH||=||systolic and diastolic hypertension|
Supported in part by grant 2p50 HL-18323 from the National Heart, Lung, and Blood Institute, Bethesda, Md.
Reprint requests to Dr Jing Fang, Department of Epidemiology and Social Medicine, Albert Einstein College of Medicine, 1300 Morris Park Ave, Bronx, NY 10461.
- Received February 4, 1995.
- Revision received March 2, 1995.
- Accepted May 2, 1995.
MacMahon S, Peto R, Cutler J, Collins R, Sorlie P, Neaton J, Abbott R, Godwin J, Dyer A, Stamler J. Blood pressure, stroke, and coronary heart disease, part 1: prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet. 1990;335:765-774.
Final Report of the Working Group on Risk and High Blood Pressure. An epidemiological approach to describing risk associated with blood pressure levels. Hypertension. 1985;7:641-651.
Stamler J. Blood pressure and high blood pressure: aspects of risk. Hypertension. 1991;18(suppl I):I-95-I-107.
Stokes J III, Kannel WB, Wolf PA, D’Agostino RB, Cupples LA. Blood pressure as a risk factor for cardiovascular disease: The Framingham study: 30 years of follow-up. Hypertension. 1989;13(suppl I):I-13-I-18.
Systolic Hypertension in the Elderly Program Cooperative Research Group. Prevention of stroke by anti-hypertensive drug treatment in older persons with isolated systolic hypertension: final results of the Systolic Hypertension in the Elderly Program (SHEP). JAMA. 1991;265:3255-3264.
Kannel WB, Stokes J III. Hypertension as a cardiovascular risk factor. In: Bulpitt CJ, ed. Handbook of Hypertension, Volume 6: Epidemiology of Hypertension. Amsterdam, Netherlands: Elsevier Science Publishers; 1985:15-34.
Rutan GH, Kuller LH, Neaton JD, Wentworth DN, McDonald RH, Smith WM. Mortality associated with diastolic hypertension and isolated systolic hypertension among men screened for Multiple Risk Factor Intervention Trial. Circulation. 1988;77:504-514.
Darne B, Girerd X, Safar M, Cambien F, Guize L. Pulsatile versus steady component of blood pressure: a cross-sectional analysis and a prospective analysis on cardiovascular mortality. Hypertension. 1989;13:392-400.
Madhavan S, Ooi WL, Cohen H, Alderman MH. The relationship of pulse pressure and blood pressure reduction to the incidence of myocardial infarction. Hypertension. 1994;23:395-401.
Colandrea MA, Friedman GD, Nichaman MZ, Lynd CN. Systolic hypertension in the elderly: an epidemiologic assessment. Circulation. 1970;41:239-245.
Blank SG, Helseth G, Pickering TG, West JE, August P. How should diastolic blood pressure be defined during pregnancy? Hypertension. 1994;24:234-240.
Working Group on Management of Patients with Hypertension and High Blood Cholesterol. National Education Programs Working Group Report on the management of patients with hypertension and high blood cholesterol. Ann Intern Med. 1991;114:224-237.
Bonaa KH, Thelle DS. Association between blood pressure and serum lipids in a population: The Tromso Study. Circulation. 1991;83:1305-1314.