Our knowledge about the natural history of untreated hypertension is mainly based on historical information from a relatively short duration of time (1900s to 1970s). During that period, there was a significant evolution in the understanding of the health-related impact of elevated blood pressure (BP) and in developing new antihypertensive medications ( Fig. 18.1 ) that solidified the awareness of the adverse association between untreated hypertension and cardiovascular morbidity and mortality. Therefore, from the early 1970s, it was no longer ethical to conduct observational studies or hypertensive clinical trials that included hypertensive individuals with untreated diastolic hypertension. Similarly, isolated systolic hypertension could not be left untreated starting in the 1990s.
The diagnosis of hypertension in ancient times was based on analysis of arterial pulse and it has been known for millennia that “hard pulse disease” (what we now call hypertension) is a major risk factor for apoplexy or the modern diagnosis of stroke, and usually resulted in untimely individual death. The first scientific reports about the association of hardening pulse with end-organ damage appeared in the early 19th century when Bright published a case series illustrating patients with hardening pulse, elevated blood urea, dropsy with albuminuria, and histological findings of left ventricular hypertrophy and hardening of the kidneys. However, these patients likely comprised a heterogeneous group with various etiologies of renal disease, and elevated blood pressure in many cases was secondary as a result of renal disease itself. Forty years later Mohamed described histologic findings of nephrosclerosis in individuals with hardening pulse that he believed were independent from primary renal disease; this was the first suggestion that hypertension itself can result in kidney damage. Subsequently, Gull and Sutton described hypertensive left ventricular hypertrophy, and Gowers reported retinal hypertensive changes. The development of a method for indirect BP measurement was the next crucial step in understanding of effects of elevated BP. The first sphygmomanometer was invented by Samuel Siegfried Karl Ritter von Basch in 1881 and later improved by Scipione Riva-Rocci in 1896. However, the Riva-Rocci method allowed only measurement of systolic blood pressure (SBP) through palpation of the pulse obliteration pressure. Systolic and diastolic blood pressure (DBP) differentiation and measurements became possible after Nikolai Korotkoff introduced auscultation of sounds into existing sphygmomanometric technique in 1905. This method allowed for routine accurate BP measurement and enabled the development of average age-specific BP charts and understanding of the relationship of different levels of BP and patient-related outcomes.
Early scientific reports from the 1910s to 1930s described hypertension as a disease with two variants. It was noted that many individuals with elevated BP, usually in the outpatient setting, were asymptomatic and had little abnormal physical and laboratory findings. This type of hypertension was considered to be “benign” and not requiring any treatment and many prominent physicians continued to advocate until the early 1950s that elevation of blood pressure was a physiologic response to maintain adequate blood flow to vital organs in the setting of aging vasculature. The second type of hypertension, which was often referred as malignant or accelerated hypertension, was mainly observed in the hospital setting when patients typically presented with markedly elevated BP (SBP in upper 200s and lower 300s, and DBP above 120 mm Hg) and suffering terminal complications of hypertension such as stroke, heart failure, papilledema, and renal failure. Only these extreme elevations in blood pressure were considered to require treatment, which at that time was mostly symptomatic because of a lack of other definitive therapies. The thresholds for abnormal BP were slowly decreasing, but remained much higher than what is accepted today. For example, the recommended BP levels for intervention were suggested as higher than 200/100 and 180/110 in two respected cardiology textbooks in the 1940s. The 1950 edition of Harrison’s Internal Medicine textbook still advocated that asymptomatic hypertension should not be treated.
One of the best case illustrations of natural history of untreated hypertension was written by one of the earliest hypertension treatment advocates, Dr. Marvin Moser, and described the medical history of the 32nd United States President, Franklin Roosevelt. It was first noted that Mr. Roosevelt suffered moderate BP elevations in the mid 1930s. Untreated hypertension in his case progressed from moderate elevations, 160/90s mm Hg, to a higher level (>180/110 mm Hg) over a 7-year period. This was associated with the sharp deterioration of Mr. Roosevelt’s health and the development of progressive heart failure and his untimely death (likely from stroke) in less than 1 year.
In this chapter we will review the important, although generally older, information from epidemiologic studies and clinical trials that led to the clear and inescapable conclusion that elevated BP is associated with adverse cardiovascular and renal outcomes. Fig. 18.2 depicts the framework of the discussion, which broadly characterizes the progression of prehypertension to hypertension to target organ damage to adverse clinical events, and finally to death.
Prehypertension and Hypertension
Hypertension is typically preceded by a gradual rise in BP from normal values into the prehypertensive range. Prehypertension is defined as SBP 120 to 139 mm Hg and/or DBP 80 to 89 mm Hg. Because hypertension is mostly asymptomatic, unless BP is measured regularly, it is often not possible to detect when these transitions occur. However, there are several indirect and direct observations supporting gradual progression of elevated BP. First, it was known from series of life insurance reports conducted from the 1920s through the 1960s and later confirmed by the National Health Examination Survey (NHES 1960 to 1962) and three separate Health and Nutrition Examination Survey (NHANES 1 to 3) that average SBP and DBP tends to increase with aging. NHANES 1 was conducted in 1971 to 1974, when hypertension treatment was still not uniform and it showed that the mean SBP at age 18 increased by 0.2 mm Hg per year until age 35, and after that the rise in mean SBP accelerates to an average 0.8 mm Hg per year. Although males aged 18 to 44 have higher mean SBP compared with females, with the difference in mean SBP up to 9 mm Hg, the rate of mean SBP rise per year among females in the 18 to 44 years age group exceeds the rates of the mean SBP rise in males, leading to “equalization” of mean SBP at around age 55. After the age of 55, the mean SBP in females starts to exceed the mean SBP in males by as much as 4 to 6 mm Hg. The mean DBP in males increases with aging; however, the rate of DBP increase is less pronounced as compared with SBP. In contrast to the mean SBP, the mean DBP in females rises in ages 18 to 64 and then remains stable thereafter. In addition, the mean DBP in men exceeds the mean of DBP in females until age 54 and becomes similar after age 55. As average BP rises with age, the proportion of individuals with prehypertension and hypertension increases with age as well. The prevalence of SBP 140 or higher and/or DBP 90 or higher mm Hg at age 18 to 34 in males and females is 13.8% and 6.3%, respectively, and increases to 65% and 74%, respectively, in the 65 to 74 age group (NHANES 1971 to 1974).
Several prospective cohort studies looked at rates of progression of normotension and prehypertension to hypertension. The Framingham Heart Study (FHS) was initiated in 1948 and included 5209 men and women aged 30 to 62 years who were subsequently followed for over 30 years. During an average of 26 years of follow-up, 23.6% and 36.2% of men and women with normal BP (defined as DBP <85 mm Hg) at baseline developed hypertension (defined as DBP ≥95 mm Hg) as compared with 54.2% and 60.6% of men and women with prehypertension (defined as DBP 85 to 89 mm Hg) at baseline. In the age-adjusted analysis, the presence of prehypertension was associated with a 3.4-fold increased risk of subsequent hypertension in both men and women, as compared with normal BP at baseline. The metabolic Life Style and Nutrition Assessment in Young Adults prospective cohort study evaluated the development of hypertension in 26,980 adolescents with mean age 17.4 years. Overall, 12.4% of young adults with normal BP (BP <120/80 mm Hg) and 17.1% of young adults with prehypertension (BP 120 to 139/80 to 89 mm Hg) developed hypertension during a maximum of 17 years of follow-up with the incidence of hypertension among men being four-fold higher as compared with women ( Fig. 18.3A ). In a Cox regression analysis adjusted for age and body mass index (BMI), the cumulative incidence of hypertension gradually increased for each 10/5 mm Hg BP increase from baseline BP less than 100/70 mm Hg with no evidence for a threshold ( Fig. 18.3B ). In a more modern investigation, the Trial of Preventing Hypertension (TROPHY), during 2-year and 4-year follow-ups, 40% and 60% of individuals, respectively, with prehypertension developed sustained hypertension. There is a consistent pattern among nonmodifiable (increasing age, African-American race) and modifiable (weight) risks factors that are shown to accelerate the conversion rates of prehypertension to hypertension.
Untreated Hypertension and Subclinical Target Organ Damage
Left Ventricular Hypertrophy
The FHS demonstrated substantially higher risk of future clinical coronary heart disease in individuals with electrocardiographic (ECG) or echocardiographic left ventricular hypertrophy (LVH). In early series of hypertensive patients (before the advent of drug therapy), ECG LVH was very commonly found at diagnosis (usually ∼40% to 60% were affected), and a much higher prevalence was found during follow-up. Janeway, in his 1912 report, demonstrated that in patients with median SBP between 200 and 220 mm Hg LVH based on physical examination was present in 75.7% of patients. The majority of patients with LVH (81%) had mild to moderate LVH on physical examination (defined as presence of one or two of the following three findings: enlarged area of cardiac percussion, downward displacement of cardiac apex, and upward lifting of cardiac impulse). However, severe LVH (defined as presence of the all three findings) was found in 22.8% of patients who died during follow-up, as compared with 7.8% of patients who were still alive ( p < 0.001). In a series of 500 consecutive hypertensive patients (mean age 32 years) without target organ damage at baseline, Perera reported in 1955 that during 20 years of average follow-up, 59% to 74% of patients developed LVH (detected by electrocardiogram or chest radiograph, respectively), after which they lived only 6 or 8 more years (on average). LVH was shown to correlate with levels of systolic and diastolic BP. Another strong piece of evidence supporting the relationship between elevated BP and LVH comes from clinical trials, such as the Losartan Intervention For Endpoint (LIFE) reduction study, which showed that antihypertensive drug therapy reduces LVH, which in turn was associated with reduction in cardiovascular events. Overall, LVH is more closely related to systolic rather than diastolic BP. It has been observed that patients with untreated isolated systolic hypertension had similar significant LVH as compared with patients with combined (elevation of both systolic and diastolic BP) hypertension despite 12 mm Hg lower mean BP in patients with isolated systolic hypertension.
Albuminuria
Urinary albumin excretion (UAE) exceeding normal values (≥30 mg of albumin per gram of creatinine or ≥30 mg per 24 hours) in patients with hypertension is considered a marker of widespread endothelial dysfunction and is associated with other asymptomatic organ damage such as LVH, carotid intima thickness, hypertensive retinopathy, and higher risk of symptomatic cardiovascular disease (CVD). Abnormal UAE is also associated with a higher risk of progression of renal dysfunction and the development of end-stage renal disease (ESRD) in patients with hypertension (HTN). UAE, even within normal range, positively correlated with levels of BP. Abnormal UAE is a common finding in patients with untreated hypertension, although prevalence varies across different cohorts. In a study involving 127 patients with untreated stage 1 hypertension (mean BP 150.1 ± 16.9/96.7 ± 8.5 mm Hg), 24.4% patients were found to have microalbuminuria (urine microalbumin ≥ 30 mg per 24 hours). SBP and DBP measured by 24-hour ambulatory blood pressure monitoring (ABPM) best correlated with the presence of microalbuminuria. Another study found even higher prevalence of microalbuminuria in up to 40% of untreated individuals with stage 1 hypertension. However, a study involving a larger number (787) of untreated hypertensive individuals found lower rates of abnormal UAE at 6.7%. Microalbuminuria in hypertensive individuals is associated with a faster rate of glomerular filtration rate (GFR) decline as compared with normal UAE. In a study involving 141 hypertensive individuals followed for 7 years, an adjusted analysis showed the rate of estimated GFR (eGFR) decline was faster in patients with microalbuminuria than was in those with normal UAE (decrease of 12.1 ± 2.77 mL per min versus 7.1 ± 0.88 mL per min, p < 0.03, respectively).
Retinal Microvascular Changes
Among asymptomatic target organ damage, retinal microvascular changes are by far the most common finding in patients with untreated hypertension. Hypertensive retinopathy was evident by nonmydriatic retinography in up to 85% of 437 untreated hypertensive individuals. In comparison, LVH is typically found in up to 44% of patients, followed by carotid intima thickness in 21.8%, microalbuminuria in 14.6%, and elevated serum creatinine (SCr) concentration in 11% of patients with untreated hypertension at the time of diagnosis. Historically, hypertensive retinopathy played a very important role in the assessment of target organ damage in the era before antihypertensive drugs were available, but its incidence and progression have been reduced since the advent of antihypertensive therapy. Papilledema or grade IV hypertensive retinopathy, the hallmark of malignant hypertension, was strongly associated with mortality in individuals with untreated hypertension. However, even the lower grades of hypertensive retinopathy carry important prognostic implications and are associated with increased all-cause mortality and stroke.
Untreated Hypertension and Mortality
As soon as routine BP measurements became available, it was quickly evident that individuals with even a mild elevation in BP and who remain asymptomatic suffered earlier death as compared with individuals with normal BP. Theodore C. Janeway published the earliest description of mortality among 458 individuals with hypertension in 1913. His cohort was restricted to patients with SBP higher than 165 mm Hg; therefore, excluding observations in individuals with milder BP elevations, Janeway noted that 21.8% of patients died within 1 year of diagnosis of hypertension and the 5-year mortality rate was 42.5%. The median SBP in survivors was 200 mm Hg as compared with a median SBP of 220 mm Hg in individuals who died. The most common causes of death were cardiac insufficiency (32.6%), uremia (25%), cerebral apoplexy (15.8%), and angina (5.4%). Janeway analyzed the relationship of BP to the cause of death and found that patients dying from angina had the lowest median SBP, in the 175 to 180 mm Hg range; whereas patients dying from heart failure and uremia had a median SBP in the 215 to 220 mm Hg range. A higher median SBP was associated with death from cerebral apoplexy (225 to 230 mm Hg) and pulmonary edema (245 to 250 mm Hg). The median age of 50 to 59 years was similar in men and women dying from cardiac insufficiency and uremia, whereas median age of death from stroke was 10 years earlier in men (50 to 59 years), as compared with women (60 to 69 years).
The life insurance industry was the champion in detecting and reporting the adverse association between elevated levels of BP and mortality on a large population level, although analysis included a disproportionally higher number of middle-aged employed men. The Medical Impairment Study of 1929 involved information about approximately 1,200,000 policyholders and reported that the observed over-expected (O/E) mortality rates in all age groups (24 to ≥65 years) with SBP greater than 5 mm Hg over age-specific average SBP was 1.74, even though the average BP levels were below 140/90 mm Hg. The O/E mortality ratio was further increased to 2.05, 2.65, and 3.84 in individuals with SBP 25 to 34, 34 to 45, and greater than 45 mm Hg above age-specific averages, respectively. The causes of death could not be evaluated in the whole group of individuals with elevated BP; however, a limited sample of 200 individuals with BP above average showed that the incidence of cerebrovascular death and coronary heart disease were 3.5 and 2.75 times higher than average, respectively. The findings of increased mortality with increasing BP levels were corroborated with minor differences in the Build and Blood Pressure Study of 1959 and 1971 to 1974. In addition, the later two studies also demonstrated a gradual increase in O/E mortality with increases above average DBP. It was also possible to more granularly assess causes of death among hypertensive individuals. Individuals with a BP in range of 138/83 to 147/92 and 148/93 to 167/97 mm Hg experienced markedly higher than standard all-cause mortality, coronary artery disease, cerebral hemorrhage, hypertensive heart disease, and renal disease ( Table 18.1 ).
| BLOOD PRESSURE 138/83 TO 147/92 mm Hg | BLOOD PRESSURE 148/93 TO 167/97 mm Hg | BLOOD PRESSURE 168/93 TO 177/102 mm Hg | |
|---|---|---|---|
| All-Cause Mortality (Excess Over the Standard) | |||
| 48% 42% | Not reported 93% | 137% 119% |
| Death From CAD (Excess Over the Standard) | |||
| 61% 51% | Not reported 137% | 140% 59% |
| Death From Cerebral Hemorrhage (Excess Over the Standard) | |||
| 131% 62% | Not reported 140% | 480% 321% |
| Death From Coronary Heart Disease (Excess Over the Standard) | |||
| Not reported 136% | Not reported 312% | Not reported 258% |
| Death From Renal Disease (Excess Over the Standard) | |||
| 160% 21% | Not reported 23% | 350% 250% |
John Fry published similar observations from the cohort of 704 individuals with untreated hypertension who were followed between 1949 and 1969. The diagnosis of hypertension in his study was based on DBP of 100 or higher mm Hg. There was an inverse relation in the O/E death rates and age until age 70. Hypertensive individuals aged 30 to 39, 40 to 49, 50 to 59, 60 to 69, and older than 70 were 7.5, 4.9, 2.2, 1.15, 0.9 times more likely to die, as compared with corresponding ages in normotensive counterparts. Similar to Janeway’s report, the most common causes of death were cardiac death (about 50%) and cerebrovascular death (25%).
The Veterans Administration (VA) Cooperative Trial investigating the role of lowering BP with antihypertensive medications in never-treated individuals with DBP 90 or higher mm Hg, had a control arm with placebo and no active treatment received, which provided important information about the natural history of untreated hypertension. The results of that study were reported for two separate cohorts. The first manuscript described outcomes among 143 individuals with DBP between 115 and 129 mm Hg and randomized to antihypertensive treatment versus a placebo control group. At the baseline, the average age of participants was 51 years and mean BP was 187/121 mm Hg. During an average of 15.7 months of follow-up, 27 major adverse CVD events occurred in untreated patients, including 4 deaths (3 from abdominal aortic aneurism catastrophe and 1 sudden death). The remaining events included accelerated hypertension with grade 3 and 4 hypertensive retinopathy, congestive heart failure (CHF), cerebrovascular accident (CVA), coronary artery disease (CAD), and 2 cases of renal failure. In contrast, in the treated group BP dropped from baseline 186/121 mm Hg to 143/91 mm Hg during 20.7 months of follow-up. There were no deaths and only 1 CVA in the treated group. The second cohort included 380 individuals with DBP between 90 and 114 mm Hg who were randomized to the active treatment arm (186 patients) or placebo (194 patients). The average age of patients in the control group was 50.5 years and the mean BP was 162/104 mm Hg. During 3.9 years of follow-up, SBP and DBP increased by 4.2 mm Hg and 1.2 mm Hg, respectively, on placebo. Additionally, 20 patients (10.3%) developed persistently elevated DBP greater than 124 mm Hg. A total of 19 patients died in the control group. There were a total of 56 morbid events in the control group: 20 (10.3%) patients experienced cerebrovascular events, 11 (5.7%) patients had congestive heart failure, 13 (6.7%) patients had coronary artery disease events, and 3 (1.6%) patients experienced progressive renal disease. In contrast, BP fell in the treated patients from a mean baseline BP of 165/105 mm Hg by 27.2/17.4 mm Hg; 8 deaths were observed during an average of 3.7 years of follow-up (relative risk [RR] 0.44, p < 0.001). The number of morbid events was also significantly reduced with hypertension treatment. CVA events occurred in 5 (2.7%) (RR 0.26, p = 0.003) and CAD in 5 (2.7%) (RR 0.40, p = 0.066) of treated patients, and no CHF or renal events were observed. Overall, the treatment of hypertension resulted in a 70% reduction of combined outcomes of all-cause mortality, uncontrolled hypertension, and morbid events among individuals with baseline DBP between 90 and 114 mm Hg. Of note, adverse events were more pronounced in untreated patients with higher baseline BP. For example, adverse events occurred in 15.3% of patients with SBP lower than 165 mm Hg, as compared with 42.7% in patients with baseline SBP 165 or higher mm Hg. Similarly, adverse events were higher in patients with higher baseline DBP: 25% of patients with baseline DBP 90 to 104 mm Hg had morbid events, compared with 31.8% of patients with DBP 105 to 114 mm Hg. It is not surprising, therefore, that the effect of BP reduction was also more pronounced in those with a higher baseline BP. For example, in patients with baseline SBP less than 165 mm Hg or DBP 90 to 104 mm Hg, the reduction in morbid events with hypertension treatment was 40% and 35%, respectively. A more pronounced effect of hypertension treatment was seen in patients with baseline SBP 165 or higher mm Hg or DBP between 105 and 114 mm Hg, where the reduction in morbid events with hypertension treatment was 64% and 75%, respectively. These data strongly support that high cardiovascular and renal complications and mortality are directly attributed to the elevated BP. Additionally, the outcomes of untreated hypertension varied with age in the VA Cooperative Trial. Patients aged older than 60 years were more likely to die, have CVA or CHF; whereas patients younger than 50 years were more likely to develop progressive hypertension with DBP greater than 124 mm Hg or renal failure. However, the incidence of CAD did not appear to vary with age.
The Multiple Risk Factor Intervention Trial (MRFIT) evaluated the relationship between levels of systolic and diastolic BP and fatal coronary heart disease (CHD) among 356,222 men aged 35 to 57 years screened for, but not entered into, the randomized trial. During 6 years of follow-up there was a strong graded relationship between levels of SBP from less than 115 mm Hg to 175 or higher mm Hg and of DBP from less than 75 mm Hg to 115 or higher mm Hg. The finding of an adverse association between isolated systolic HTN and CHD mortality is also an important contribution of this study.
The latest and perhaps one of the most powerful evidence of the relationship of elevated blood pressure and vascular mortality came with the publication of the Prospective Study Collaboration in 2002, which included information about causes of death among close to 1 million participants (958,074) from 61 individual prospective observational studies of BP and mortality. Using time-dependent correlation, this meta-analysis demonstrated that usual (or long-term average) SBP and DBP strongly and directly correlated with stroke, ischemic heart disease, and other vascular-related mortality rates ( Fig.18.4 ). In ages 40 to 69 years (irrespective of gender) each difference of 20 mm Hg in SBP or 10 mm Hg in DBP over BP 115/75 mm Hg was associated with over a two-fold increase in stroke death rates, two-fold increase in ischemic heart disease death rates, and other vascular death rates. Although proportional differences in death rates are lower in persons aged older than 80, given the higher incidence of vascular events, the annual difference in absolute risks of vascular death are greater in older age.
