¹Department of Preventive Medicine, Rush University Medical Center, Chicago, IL, USA

Correspondence: Dr HR Black, Department of Preventive Medicine, Rush University Medical Center, Chicago, IL 60612, USA

Abstract

Since the middle of the 20th century, most physicians and epidemiologists assessed the risks associated with hypertension based on the level of diastolic blood pressure (DBP). In a classic paper in 1971, the Framingham Heart Study clearly showed that systolic BP more accurately described the risk of all the complications we attribute to hypertension. It took 22 years until JNC V in 1993 also used systolic blood pressure (SBP) to define hypertension in US national guidelines. Since then, the paradigm has shifted dramatically. In JNC VI (1997) and JNC VII (2003), SBP has become the primary focus of risk stratification and treatment goals. This shift is a result of the Framingham results being confirmed by many others analyses, the most compelling of which is the recently published report of the Prospective Collaborative Study Group, which pooled 61 observational studies in more than 1 million volunteers with a collective experience of more than 12 million person-years. This group showed that the SBP level at baseline was a significantly more informative reading than DBP for predicting strokes and coronary heart disease (CHD). Furthermore, three trials of older individuals with isolated systolic hypertension, SHEP, SYST-Eur, and SYST-China, unambiguously demonstrated that effective antihypertensive therapy lowered the rate of strokes, heart failure, CHD, and even all-cause mortality. Finally, the World Health Organization (WHO)/International Society of Hypertension (ISH) Hypertension Trialists also showed that the level of SBP achieved in clinical trials comparing different antihypertensives with placebo and with each other was the strongest determinant of how effectively strokes and CHD events were reduced, although a similar relationship was not evident for heart failure. A recent metaregression analysis using new trials, many of which were used by the Trialists, and older studies not included in their analysis also showed that small differences in SBP can have a dramatic impact on cardiovascular outcomes. If there is one thing we have learned in the recent past, it is the need for us to focus on lowering SBP and getting it down to a reasonable goal. We have also learned that to do so, we will need to combine a variety of lifestyle and pharmacological approaches, always with combinations of drugs that will usually contain a low-dose thiazide-type diuretic with other antihypertensives.

Reverend Stephen Hales first measured blood pressure (BP) by inserting a glass tube in the carotid artery of a restrained horse. BP measurements in humans began in the late 19th century when Riva-Rocci developed the technology to measure BP noninvasively. In 1906, Korotkoff described the still-used K-sounds, by which indirect measurements could be easily performed. Over the next several decades, systolic blood pressure (SBP) was used to define hypertension and describe the risks associated with having an elevated BP and stratifying risk by the degree of that elevation. Diastolic blood pressure (DBP) elevations were notable but not considered to be as relevant as elevations in SBP.

By the middle of the twentieth century, most clinicians and epidemiologists began to appreciate the risks associated with DBP. Although it was generally agreed that indirect measurements of DBP were not as accurate as indirect measurements of SBP, the medical community's attention was directed toward diastolic hypertension. Elevations in SBP were considered to be a natural consequence of ageing and the concept arose that a normal SBP was '100+your age'.

In the late 1950s and early 1960s, the Epidemiologic Pooling Project, which included the Framingham Heart Study, combined the results of several observational studies and concluded that while both SBP and DBP contributed to risk, DBP was the more definitive and the target to which therapy should be directed.¹ Those studies were relatively small and short by today's standards and enrolled primarily younger people.

As a result, early clinical trials recruited subjects with elevated DBP and the investigators did not concern themselves with systolic elevations or with isolated systolic hypertension.^{2, 3, 4} Regulatory agencies, such as the United States Food and Drug Administration, required pharmaceutical companies only to demonstrate that their antihypertensive drug was able to lower DBP, although all effective drugs always lower SBP as well, usually in a ratio of about 3 mmHg for each 2 mmHg reduction in DBP. The National High Blood Pressure Education Program of the National Heart, Lung and Blood Institute of the National Institutes of Health, which was responsible for the periodic reports of the Joint National Committees (the JNCs), defined hypertension by the level of DBP and did not give SBP much attention.^{5, 6, 7, 8}

Yet the Framingham Heart Study clearly showed, in a classic paper in 1971, that it was SBP that more accurately described the risk of all the complications we attribute to hypertension, than did DBP.⁹ It took 22 years until the Fifth Report of the Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure (JNC V) in 1993 also used SBP to define hypertension in US national guidelines.¹⁰ Since then, the paradigm has dramatically shifted. In JNC VI (1997) and JNC 7 (2003), SBP has become the primary focus of risk stratification and treatment goals.^{11, 12}

This shift resulted from the Framingham finding having been confirmed by many other analyses, the most compelling of which is the recently published report of the Prospective Collaborative Study Group.¹³ This group pooled 61 observational studies in more than 1 million volunteers with a collective experience of more than 12 million person-years. They showed that the SBP level at baseline (entry into the observational study) was a significantly more informative reading than was DBP for predicting strokes and coronary heart disease (Table 1). The results of this analysis solidified the place of SBP as the best widely used predictor of hypertension-related cardiovascular outcomes although the 'mid-blood pressure,' which is the sum of ½ the SBP and ½ the DBP, was the most informative (predicted outcomes) of the many ways to examine BP data.

Table 1 - Blood pressure parameters informativeness of coronary heart disease (CHD) and stroke. SBP was more accurate than DBP at predicting cardiovascular events but 'mid-blood pressure' was the most informative (from Lewington et al¹³).

Full table

While it had appeared that pulse pressure (the arithmetic difference between SBP and DBP), which is a convenient way to incorporate both SBP and DBP, would be an ideal way to determine risk more accurately than either alone, the Prospective Collaborative Study Group found that pulse pressure was not as 'informative' as either one at predicting both stroke and coronary heart disease.¹³ This finding was disappointing, since many longitudinal population-based surveys have defined the natural history of hypertension and its relationship to ageing.

The National Health and Nutrition Surveys have demonstrated that both SBP and DBP rise as we age but the patterns are different¹⁴ (Figure 1). The DBP rises from ages 18 to 29 years until the middle of the sixth decade and then declines. This pattern is seen in men and women and in all ethnicities studied and in those with all levels of baseline BP (Figure 2). SBP, on the other hand, rises from adolescence until old age in all demographic groups. Consequently, pulse pressure widens with advancing age.

Figure 1.

Mean SBP and DBP by age and race/ethnicity for men and women. SBP rises with age from teen age years until past 80 years of age but diastolic blood peaks at about age 50–55 years. (From Burt et al¹⁴).

Full figure and legend (109K)

Figure 2.

Age-related changes in blood pressure parameters. Regardless of the blood pressure at entry into this observational study, SBP rises with age and DBP tends to peak in the sixth decade of life. Pulse pressure rises throughout life (from Franklin SS et al¹⁵).

Full figure and legend (121K)

This natural history would predict that most hypertension in older people would be isolated systolic hypertension or predominantly systolic hypertension, a pattern confirmed by Franklin and colleagues using the National Health and Nutrition Third Survey.¹⁶ In their 1990 estimate, 73% of 47 million hypertensives in America were reckoned to be over 50 years of age. The predictions for 2020 are that 80% of hypertensive Americans (now estimated to be more than 70 million strong) will be over 50 and will have predominately SBP elevations (Figure 3).

Figure 3.

Age distribution of hypertension in the US population in 1990 and in 2020 (projection) (from Franklin¹⁶).

Full figure and legend (124K)

For many years, it was felt that isolated systolic hypertension was not only a natural condition of ageing but that treating older individuals with only elevated SBP would not reduce cardiovascular events and might even be harmful. However, three trials of older individuals with isolated systolic hypertension—the Systolic Hypertension in the Elderly Program (SHEP), SYSTolic hypertension in Europe (SYST-Eur) study, and SYSTolic hypertension in China (SYST-China)—unambiguously demonstrated that effective antihypertensive therapy lowered the rate of strokes, heart failure, coronary heart disease, and even all-cause mortality^{17, 18, 19, 20} (Figure 4). The reduction in events in older persons with isolated systolic hypertension is comparable to the reductions seen in subjects of the same age with systolic and diastolic hypertension.

Figure 4.

A meta-analysis of the results of the three large randomized or allocated, placebo-controlled clinical trials which evaluated the benefits or risk of treating isolated systolic hypertension (from Staessen et al²⁰).

Full figure and legend (58K)

Finally, the World Health Organization (WHO)/International Society of Hypertension (ISH) Hypertension Trialists also showed that the level of SBP achieved in clinical trials comparing different antihypertensives with placebo and with each other was the strongest determinant of how effectively strokes and coronary heart disease events were reduced, although a similar relationship was not evident for heart failure²¹ (Figure 5). And a recent metaregression by Staessen et al²² using both new trials, many of which were used by the Trialists, and older studies that were not included in their analysis, also showed that small differences in SBP can have a dramatic impact on cardiovascular outcomes.

Figure 5.

The relationship of the difference between achieved SBP in the experimental group and the comparator on the occurrence of stroke and coronary heart disease (CHD) and heart failure (HF) in the trials analysed by the BP-Lowering Trialists' Collaboration (from Blood Pressure Lowering Treatment Trialists' Collaboration²¹).

Full figure and legend (90K)

Although we currently have more than 125 different antihypertensive drugs and fixed-dose combinations, treating SBP to goal (<140 mmHg for uncomplicated hypertensives and <130 mmHg for hypertensives with diabetes mellitus or chronic renal failure) is still very difficult to achieve.^{12, 23, 24} Even in our Hypertension Specialist Clinic in which 'goal-oriented management' has enabled us to successfully reduce DBP to <90 mmHg in uncomplicated hypertensives and to <85 mmHg in those with renal disease and/or diabetes mellitus in nearly 90% of our patients, we are only able to get SBP to goal in less than 60% of those patients. This finding that controlling DBP to goal is considerably easier than controlling SBP to goal has also been demonstrated in clinical trials which follow forced-titration protocols.^{25, 26} In both Controlled ONset Verapamil INvestigation of Cardiovascular End points (CONVINCE) and Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT), two long and large clinical trials, more than 60% of volunteers were on two or more antihypertensives, with some receiving four or more drugs.^{25, 26}

If there is one thing we have learned in the recent past, it is the need for us to focus on lowering SBP and treating it to a goal. We have also learned that in order to do so, we will need to combine a variety of lifestyle and pharmacologic approaches and use a combination of drugs. Monotherapy alone will rarely be successful. Until such time as we have new antihypertensives that are specifically able to reduce SBP, we will need to learn how to use the multiple agents we currently have and to target the elevations of SBP so our patients can get the maximum benefit of treatment.