It has been 250 years since the clergyman Stephen Hales, using a glass tube as a crude manometer, demonstrated the existence of blood pressure within the neck artery of a horse. This important observation was followed by the landmark development of the forerunner of the stethoscope by Laennec in the first part of the 18th century. By the close ofthat century Riva-Rocci, using a sphygmomanometer remarkably similar to those in use today, described the systolic blood pressure as that equal to the point of appearance of a pulse as an arm cuff was deflated. In 1906, by combining the use of the sphygmomanometer and stethoscope, Korotkoff observed that the audible pulse disappeared with decreasing cuff pressure at a point roughly equivalent to the resting phase of the heart.
Subsequent to these notable advances, numerous investigators have made substantial contributions to our understanding of the physiology of blood pressure regulation. Soon, conditions associated with abnormal blood pressure would be observed and their pathophysiology explored as well. Thus, the "disease" hypertension was born. Inherent in the description of this abnormal condition is the implication that there is a definable state of blood pressure disease. This implication has become a significant source of confusion about normal and abnormal blood pressure in humans.
DEFINING THE PROBLEM
In 1987 the Second Task Force on Blood Pressure Control in Children1 reaffirmed the importance of the current recommendation of the American Academy of Pediatrics regarding blood pressure screening. The Academy's recommendation calls for the routine measurement of blood pressure beginning at the 3-year examination and continuing annually thereafter. Despite this recommendation, many physicians caring for children have questioned the value of repeated blood pressure measurements in the office setting. Although the commitment to routine blood pressure screening may appear trivial, the time involved in complying with the Academy's recommendation shows that the commitment is a serious one. If half of 30 patients seen on a routine day in the office are 3 years of age or older, approximately 15 to 20 minutes per day will be required to obtain and record single, casual blood pressure measurements for all patients. If, however, an effort is made to obtain duplicate basal measurements taken 2 minutes apart, more than 60 minutes of the working day will be taken up by blood pressure measurements alone. Is this use of a health professional's time justifiable?
Figure 1A. Mean and 95th percentile values of systolic and diastolic blood pressure for boys 1 to 13 years of age.
Figure 1B. Mean and 95th percentile values of systolic and diastolic blood pressure for girls I to 13 years of age.
No one can dispute the importance of the early detection of severely elevated blood pressure; the incidence of secondary causes of high blood pressure increases with diminishing age.2,5 Although the importance of identifying and treating these potentially injurious conditions is clear, they are rarely seen in the office setting; milder degrees of elevated blood pressure are being identified with increasing frequency. 5 What, then, is the basis for routine and repetitive blood pressure measurement?
Studies over the past several decades, predominantly in adult patients, clearly demonstrate a relationship between elevated blood pressure and significant end-organ damage, particularly cardiac, cerebrovascular, and renal.4'6 The successful reduction of elevated blood pressure leads to diminished morbidity and mortality in most study populations.7'9 Although no long-term follow-up data exist for individuals developing elevated blood pressure during childhood, short-term studies in children and adolescents have demonstrated vascular and cardiac findings which are the likely forerunners of adult end-organ damage.1012 Based on these inferential and incomplete data, many specialists believe that early detection and intervention in children at risk for hypertensive morbidity in later life is warranted.
The practical clinical problem arises from defining the population at risk. Actuarial statistics support the idea that morbidity is measurable even at levels of blood pressure considered "normal." The acculturated societies from which "norms" are derived have a significant prevalence of hypertension; acculturated norms are higher than those obtained from nonacculturated populations. The use of percentiles and the arbitrary assignment of abnormality to those above the 95th percentile is potentially flawed; this assignment assumes that 5% of the population is abnormal. Whether this assumption is accurate or either overestimates or underestimates the actual problem is yet to be determined.
In the pediatric population, the problem of defining high blood pressure is compounded by several additional factors. North American and European screening studies designed to define normal childhood values have demonstrated gradual increases of blood pressure with increasing age, reaching adult values during adolescence. In addition, there are significant differences in age-related percentile values for blood pressure reported from various studies.13 The Task Force has attempted to reconcile some of these differences in their most recent report, which provides age- and sexrelated percentile data derived from previously reported values in over 17,000 children, including whites, blacks, and Hispanics (Figure I).1 No racespecific effect on blood pressure was noted, and the data were therefore aggregated. Although the range of acceptable values at a given age remains broad, the observation of a linear relationship between height and blood pressure13 has enabled better interpretation of individual values; the taller child whose blood pressure is in the higher percentiles is likely to be normal.
Figure 1C. Mean and 95th percentile values of systolic and diastolic blood pressure for boys 13 to 18 years of age.
Figure 1D. Mean and 95th percentile values of systolic and diastolic blood pressure for girls 13 to 18 years of age.
BLOOD PRESSURE MEASUREMENT
Although the measurement of blood pressure is one of the first technical tasks learned by medical and nursing students, measurements are not as reliable as generally believed. Numerous screening studies of blood pressure in childhood have demonstrated a number of sources of significant measurement error. For example, the use of a single blood pressure reading in the Bogalusa Heart Study would have generated a 95% confidence interval for both systolic and diastolic blood pressures of greater than 20 mmHg above or below a given reading. Other studies have demonstrated significant measurement variance with multiple observations, even when trained personnel were used to record blood pressure values. If time-consuming blood pressure measurements are to be obtained in the office, procedures should ensure a reasonable degree of reliability.
There are three major obstacles to accurate blood pressure measurement: (1) selecting the appropriate sized cuff; (2) generating a nonthreatening and unhurried setting; and 3) accurately identifying the proper Korotkoff phase.
Accurate indirect blood pressure readings depend on relatively symmetric compression of the artery.14 Thus, selection of a blood pressure cuff the bladder of which encircles at least 75% of the circumference of the arm is mandatory. An alternative method requires that at least 75% of the upper arm be covered. Use of a cuff size smaller than that recommended risks significant overestimation of blood pressure. If in doubt, use the larger cuff, as this will lead to fewer factitious results. Because of the relatively low frequency of the transmitted sounds, the bell of the stethoscope should be used for auscultating Korotkoff sounds.
Obtaining blood pressure readings in a basal state - that is, under conditions that attempt to control a number of biologic factors affecting blood pressure - significantly contributes to the minimization of measurement artifact. Practical guidelines such as those listed in the table should be followed.
In the 80 years since Korotkoff described what would ultimately be designated as the fifth Korotkoff (K) sound, the clinical use of the Korotkoff sounds remains inconsistent. The first K sound, Kl, closely correlates with systolic pressure, and is generally considered to be that point when two consecutive tapping sounds are audible. As cuff pressure decreases, at a recommended rate of 2 to 3 mmHg/sec, an abrupt muffling of the sound is heard (K4) followed by complete disappearance (K5). In older children and adults, K5 is generally considered to reflect the true intraarterial diastolic pressure. By contrast K4, which usually occurs 5 to 10 mmHg above the diastolic pressure, is currently recommended for diastolic measurements in younger children. This convention was established in part because the disappearance of the audible phase (K5) may occur far below the true diastolic pressure in infants and small children.14 Londe15 has argued that this phenomenon is not a fundamental characteristic of young children, but is a result of excessive pressure of the stethoscope head over the artery. This pressure leads to artifactual sounds from inadvertent partial occlusion of the artery, which may occur in all age groups. He has demonstrated the importance of bell placement so that a seal with the skin is achieved with minimal downward force. Nonetheless, the percentile data from the Task Force uses K4 for diastolic pressure in children up to 13 years, and K5 thereafter. This inconsistency can lead to a great deal of confusion when blood pressure measurements are recorded and subsequently compared with normal values. To minimize confusion, I recommend that all three measurements^ - K1/K4/K5 - be obtained and recorded. If muffling and disappearance of the sounds occur virtually simultaneously, the K4 and K5 values should be recorded as the same number.
WHO IS AT RISK?
The point of the preceding discussion is to emphasize several facts:
1. Blood pressure norms are relative, not absolute.
2. The forerunner of high blood pressure-associated end-organ damage occurs in childhood.
3. Consistency of measurement is essential to formulate a reasonable assessment of risk.
As statistically defined, approximately 5% of all patients should have a value greater than the 95th percentile at initial blood pressure screening measurement. Although the Task Force norms are based on single values, they recommend that an increased blood pressure measurement be confirmed by repeated measures. With repeated measures, however, the prevalence of high blood pressure falls to fewer than 1% in most studies. Who, then, should be screened, counseled, and followed most closely?
Factors Potentially Confounding Blood Pressure Measurements
The risks of cardiac and cerebrovascular morbidity and mortality in adults relate to the presence of sustained blood pressure elevations of long duration. The challenge to the pediatrician or family physician is the proper identification of the child who has, or is at risk for, sustained blood pressure elevations. Numerous studies have examined the predictive value of single and serial blood pressure measurements. Although a single, casual blood pressure value greater than the 95th percentile is inadequate for diagnosing hypertension, I3,11,17 many studies have examined the phenomenon of "tracking," or the tendency of patients to maintain blood pressure levels over time.13,18,19 These studies generally demonstrate that at relatively early ages blood pressure percentile has the statistical power to predict the percentiles that will be measured some years later. The probability of accurately predicting future high blood pressure improves with multiple measurements. However, although the correlations are significant, particularly in older children and adolescents, they are of limited use in making diagnostic or treatment decisions in individual patients. Data also indicate that patients whose blood pressure is in the normal range at the time of initial assessment have a high probability of remaining within that range on subsequent evaluation in later childhood.18 Based on these data, multiple reliable blood pressure measurements (perhaps a minimum of 3 or 4 over a period of 5 to 8 years) may be valuable in judging future risk.
In formulating a risk assessment in individual patients, the physician must consider other factors associated with a higher likelihood of acquiring sustained hypertension. These include the contributory roles of obesity, sympathetic nervous system activity, and the characteristics of sodium, chloride, and potassium regulation in children with elevated blood pressure. These factors have environmental as well as genetic bases, and the index of suspicion should increase in the child with a family history of hypertension.
Many investigators have provided the primary care clinician with data that profile the pediatric patient at risk for sustained high blood pressure and the attendant morbidity. This is the science of hypertension. However, the level of certainty of this diagnosis ultimately relies heavily on the judgment of the physician who must consider the relative importance of all potential contributing factors. This is the art.
Blood pressure data in Figures IA-D from the 1987 Task Force on Blood Pressure Control in Children.1
1. Report of the Second Task Force on Blood Pressure Control in Children- 1981. Pedkuna 1987; 79:1-25.
2. Rocchini AP: Childhood hypertension: Etiology, diagnosi. and treatment. Peduitr Clin North Am 1984; 31:1259-1273.
3 Feld LG. Sprintate JE: Hypertension in childtcn. Curr Pr Pediatr 1988; 18:317-373.
4. Pickering Ci: High Blood Pressure (ed 2). New York, Crune. 1968.
5. Kannel WB, Wolf PA. Verter J. et al: Epidemiologic assessment of the role of blood pressure in stroke. The Framingham Study. JAMA 1970; 214:301.
6. Kannel WB. Castelli WP, McNamara PM. et al: Role of blood pressure in the development of congestive heart failure. N Engl f Med 1972; 287:781-787.
7. Veteran's Administration Cooperative Study Group on Antihypertensive Agents: Effect of treatment on morbidity and mortality: Results in patients with diastolic blood pressures averaging 1 1 6 through 129 mmHg. JAMA 1967; 202:1028-1034.
8. Veteran's Administration Cooperative Study Group on Antihypertensive Agents: I]. Results in patients with diastolic blood pressure averaging 90 through 114 mmHg. JAMA 1970; 213:1143-1152.
9. Hypertension Detection and Follow-up Program Cooperative Group: The effect of treatment on mortality m "mild" hypertension. Results of the Hypertension Detection and Follow-up Program. N Engl) Med 1982; 307:976-980.
10. Culpepper WS, Sodt K:. Messeri i FH. et al: Cardiac status in juvenile borderline hypertension.
11. Newman WP. Freedman DS. Voors AW. et al: Relation of serum lipoprotein levels and systolic blood pressure to early atherosclerosis. N Engl) Med 1986; 314:138-144.
12. Anstimuno GG, Foster TA. Berenson GS. et al: Subtle electrocardiographic changes in children with high levels of blood pressure. Am J Cardiol 1984: 54:1272-1276.
13 Berenson GS. Cresanta JL. Webber LS: High blood pressure in the young. Annual Rev of Medicine 1984. 35.535-560.
14. K.rkendall WM. Feinleib M. Freís ED. et al: American Heart Association recommendation, for human blood pressure determination by sphygmomanometers. Circulation 1980; 62:I146A-I155A.
15 Londe S: Fifth versus fourth Korotkotf phase. Pediatrics 1985. 76:460-461.
16. Rosner B. Polk BF: Predictive values of routine blood pressure measurements in screening for hypertension. AmJ Epidemiol 1983; 117:429-442.
17. Sinatko AR. Gomez-Mann O. PrineasRJ: "Significant" diastolic hypertension in prehigh school black and white children. Am J Hvpertens 1988; 1:178-180.
18. Shear CL. Burke GL. Freedman DS. et al: Value of childhood blood pte.ssure measurements and family history in predicting future blood pressure status: Results from S vears of follow-up in the Bogalusa heart study. Pedtatncs 1986; 77:862-869.
19. Lauer RM. Clarke WR. Beaglehole R Level, trend and variability of blood pressure during childhood: The Muscatine study. Circulation 1984: 69:242-249.
Factors Potentially Confounding Blood Pressure Measurements