Hypertension is not otitis media! As pediatricians we believe we understand the condition we call otitis media. We have developed the ability to distinguish a pink tympanic membrane from a truly infected ear. We may even take pride in not diagnosing this condition too frequently. We know the bacterial causes of middle ear disease and the appropriate antibiotics to counter those agents. We understand the prognosis of otitis media and know what should (and perhaps should not) be done if the infection or effusion persists after a course of therapy. One might say that otitis media is a pediatrician's disease.
But what about hypertension? Can we say that as pediatricians we are "comfortable" with hypertension? Are we able to list the causes of elevated blood pressure? Is the treatment of hypertension in our pharmaceutical armamentarium? What is our understanding of the prognosis for the children who fail to respond to the prescribed treatment? Might we even be underdiagnosing elevated blood pressure?
Certainly part of the explanation for any discomfort we may have discussing hypertension is its infrequency. Experience is more easily developed with those diseases we see every day. Yet a good part of our discomfort is well justified by the lack of basic information about childhood hypertension. Consider the following statement from the Second Task Force on Blood Pressure Control in Children, 1987: "In 1977, data linking cardiovascular risk with systolic and diastolic BPs in children were not available and, unfortunately, are not yet available."1 In other words, we do not have the information to say that high blood pressure in childhood is a health risk. Is it any wonder that we have not developed a clear approach to the prognosis and treatment of childhood hypertension?
This article offers some practical points about the evaluation of hypertension in childhood. It also conveys some useful concepts and background information that are more important for pediatricians to remember than which diagnostic instrument is best for renovascular hypertension or which drug is best for treatment of elevated blood pressure in childhood. In fact, the pediatrician may prove more effective than a particular medication in treating childhood hypertension.
Several years ago 1 attended a lecture on the treatment of hypertension. The speaker, an internist, greeted the audience: "Fellow Hypertensives!" He explained that he had chosen that introduction because hypertension is relative. A person whose blood pressure is 120/80 mmHg is hypertensive compared with one whose blood pressure is 110/70 mmHg, and that person is hypertensive compared with one whose reading is 100/60 mmHg. Those adults whose blood pressure is 120/80 mmHg have a higher risk of cardiovascular complications than those whose blood pressure is 110/70 mmHg.
Such observations have led experts to call for a halt to the concept of hypertension as a "disease. "2 It is one of several risk factors for the development of cardiovascular, brain, and kidney damage. No level of blood pressure is safe (although some is helpful). An adult with "normal" blood pressure (less than 140/90 mmHg) has a certain statistical risk of stroke. That risk triples for the adult whose blood pressure is 140 to 189 mmHg systolic or 90 to 95 mmHg diastolic. Therefore the definition of hypertension is somewhat arbitrary. We should not tell the adult (or child) whose blood pressure is 125/85 mmHg there is "nothing to worry about." That person may have a lot to worry about, especially if other risk factors - elevated cholesterol, smoking, and a family history of myocardial infarction at age 40 - are present. It is equally true that many individuals live to a full life expectancy with untreated hypertension. Hypertension is a risk factor, not a disease.
A great deal of information has been gathered about hypertension in adults, its risks, and the value of treatment. Much of the epidemiologic data has been valuable in providing risk-benefit information for patients whose blood pressures are termed "elevated." Unfortunately, similar data are not yet available to guide us in a risk-benefit analysis for children. As has been true for other conditions that affect the pediatric patient, it may take 30 years or longer to uncover the morbidity associated with hypertension. Several longitudinal studies are well underway which should provide us with the necessary "facts" to help our patients and their families make sensible and defensible therapeutic decisions.
There are at present two hypotheses which suggest that elevated blood pressure in childhood is unhealthy. Many investigators have pointed to the concept of blood pressure "tracking" in children. This hypothesis suggests that adult hypertensives develop somewhat elevated blood pressure as children. This early tendency can be tracked over time until hypertension is clearly established. If this hypothesis holds true, children who may benefit from treatment can be identified early and the majority of other children can be reassured of their low risk of developing hypertension.
Definition of Hypertension in Children
A second body of evidence used to support identification and treatment of hypertension in childhood is the development of secondary end-organ damage. Studies in adults have shown that left ventricular hypertrophy is a serious and ominous complication of hypertension. Echocardiography studies in children with elevated blood pressure also have shown significant cardiac enlargement suggestive of early hypertensive heart disease.3 Such information is worrisome. However, without more compelling data or results from the long-term studies now in progress, establishing a definition of pediatric hypertension and guidelines for treatment becomes a more practical necessity. The Second Task Force on Blood Pressure Control in Children has developed such a definition (Table 1).
The Task Force has also developed percentile curves by age, sex, and body size, it should be noted that the blood pressure readings that are now considered hypertensive are lower than those given in the First Task Force Report (1977). A classification of hypertension by age group is shown -in Table 2. The Second Task Force has added several qualifications that should be kept in mind when interpreting blood pressure readings in children. A single elevated blood pressure reading does not establish a diagnosis of hypertension. Several readings may be necessary, and extra effort and time may be required to calm an anxious patient. Be certain the cuff size is appropriate; the bladder should be long enough to encircle the arm and wide enough to cover about 75% of the upper arm. The child should be seated with the arm resting on a supportive surface at the level of the child's heart. The bell of the stethoscope should be used to auscultate for Korotkoffs first phase (onset of clear tapping) and the fourth (fading, low pitched sounds) or fifth phase (disappearance of all sounds). It is important that the stethoscope be pressed lightly over the brachial artery. Finally, height and weight should be used in assessing the significance of an elevated blood pressure reading in a child; larger children (taller or heavier) have higher normal blood pressure than smaller children of the same age.
Criteria for Classification of Hypertension by Age Group
Because increasing blood pressure is as much a part of the developmental continuum of childhood as is growth or motor development, it makes sense to plot blood pressure readings annually as we do for height, weight, and head circumference. A blood pressure graph can be created on top of the growth charts we already employ. If that seems impractical, percentile graphs developed by the Task Force can be used to plot each child's blood pressure sequentially on the appropriate curve. I also suggest that you record the parents' blood pressures on the child's data sheet. Such family information may be helpful in a prognostic sense, just as parental heights are fairly predictive of the child's adult stature. This information may also act as a reminder to discuss cardiovascular health issues as part of preventive counseling.
EVALUATION OF THE HYPERTENSIVE CHILD
When a child with elevated blood pressure is diagnosed as having hypertension (at least 3 different readings; proper cuff size; 2= 95 percentile for age, sex, height, and weight), the cause of hypertension should be evaluated. A detailed medical history of the patient and the family is helpful in some instances in suggesting the diagnosis and guiding (or limiting) the laboratory and radiographic studies. Questions about relatives with hypertension, myocardial infarction, stroke, renal disease (eg, polycystic kidney disease), and endocrine conditions should be explored. The age of onset of any of these conditions should be ascertained. If a family history of essential hypertension is present, the child will likely have a similar diagnosis.
The age of the child will of course direct the types of historical information to be obtained. The neonatal course may have been complicated, and an umbilical artery catheter may have been used. Headaches are neither specific nor sensitive as an indicator of hypertension. Most children with hypertension do not have headaches. Symptoms of pheochromocytoma - sweating, flushing, palpitations - are usually associated with hypertension. Abdominal trauma is a possible cause of hypertension, secondary to renovascular or renal parenchymal injury. Ingestion of prescription and overthe-counter drugs - stimulants and vasoconstrictors - may also cause hypertension. An adolescent girl should be asked specifically whether she is taking oral contraceptives, as she may be unwilling to offer such information or may not consider birth control pills to be a medication. Licorice is an unusual but possible cause of hypertension because of its potent mineralocorticoid property.4 Another more commonly used mineralocorticoid substance, chewing tobacco, may cause elevated blood pressure, even severe levels of hypertension.5 A great deal of time, expense, and discomfort can be saved if a few extra minutes are spent on exploring the history and habits of the child with hypertension. In my experience, the history has been the single most valuable "test" in the evaluation of childhood hypertension.
The most important part of the physical examination is proper measurement of the blood pressure. Was the cuff size correct? Was the child anxious? This problem usually arises with the adolescent girl, but one should remember that pheochromocytoma may produce tachycardia and other symptoms that mimic anxiety. I find that leaving the blood pressure cuff on the arm while I do the rest of the examination lessens the anxiety for the child and allows me to repeat the blood pressure readings several times during the exam.
When I perform a physical examination of a hypertensive patient, I look for the etiology of the hypertension by region.
Head. The funduscopic examination detects papilledema (increased intracranial pressure) as well as evidence of chronic hypertensive vascular changes - arteriolar narrowing, arteriovenous nicking, or, rarely, hemorrhages and exudates. The face may appear full and rounded (Cushing's syndrome).
Neck. The thyroid gland should be palpated for enlargement (hyperthyroidism). If the thyroid gland is overly active, a concomitant tachycardia should be present.
Chest. Coarctation of the aorta is an important diagnosis to consider. It is tempting to overlook this condition in an older child or adolescent in the belief that prior routine examinations would have uncovered it. However, several studies of coarctation of the aorta indicate how late this problem may be discovered. In one series, the mean age at diagnosis was 10 years; the oldest patient was 36. 3 Blood pressure should be measured in all extremities and if the pressure in the thigh is lower than that in the arm, the diagnosis of coarctation is strongly suggested. The presence of a loud second heart sound or a gallop rhythm are indicative of longer-standing and more severe hypertension.
Abdomen. A palpable mass in the abdomen should raise several diagnostic possibilities. If the mass is in the kidney, is it Wilms' tumor, a polycystic kidney, hematoma, or might the mass be a neuroblastoma or pheochromocytoma? An abdominal bruit may be detected by listening carefully just above the umbilicus and over each flank. The presence of an abdominal bruit does not verify the diagnosis of renovascular hypertension.6 Some individuals with a bruit have essential hypertension and others have normal blood pressure.
Skin. The diagnosis of neurofibromatosis is made by the finding of characteristic cafe-au-lait skin lesions. The etiology of hypertension in this condition is renovascular. The neurofibromas may narrow the main renal artery or one of its branches, giving rise to a renin-mediated hypertension. The skin may also suggest other causes of hypertension: striae, truncal obesity, and hirsutism are suggestive of Cushing's syndrome; webbing of the neck and widely spaced nipples of Turner's syndrome; and vasculitis oí collagen diseases or arteritis.
Genitalia. The rare child with virilizing adrenogenital syndrome may be detected by the appearance of the genitalia. A delay in sexual development (Tanner staging) may be a clue to a more longstanding cause of hypertension (eg, renal disease).
The diagnostic evaluation of hypertension in a child should be "individualized and tempered by an awareness that the natural history of childhood hypertension is unknown, and that hypertensive morbidity relates both to its severity and duration as well as to the presence of other cardiovascular disease risk factors."3 Rarely is there any need to evaluate a hypertensive child quickly. It may, in fact, be preferable to "drag one's feet," repeat the blood pressure measurements on subsequent visits, re-examine the child, and review the medical history. In the extreme case a child will present with hypertensive encephalopathy, congestive heart failure, or renal impairment. Such children will require hospitalization and more urgent evaluation and treatment
There is no standard regimen of tests for evaluating the hypertensive child. Each child must have an individualized prescription of diagnostic tests based on age, severity of hypertension, and physical examination. Not too many years ago, we were advised to think of essential hypertension in children as a diagnosis of exclusion. This view committed the physician and the child to a comprehensive workup, including renal arteriography. Fortunately, today we are aware that essential hypertension is the most common "cause" of childhood hypertension.7 All other possible causes need not be ruled out before a presumptive diagnosis of essential hypertension is made.
A few general guidelines are useful when beginning to formulate a diagnostic evaluation of a child with hypertension.
1. Essential hypertension is far more common than renovascular hypertension in children.8
2. The younger the child and the more severe the hypertension, the more likely it is that the hypertension is secondary to another disorder.9
3. A normal physical examination and negative personal/family history should be sufficient to eliminate several causes of hypertension from further evaluation: central nervous system and thyroid disorders, coarctation of the aorta, Cushing's syndrome, pheochromocytoma, virilizing adrenogenital syndrome, and chemically induced hypertension.
4. Obese patients rarely have reasons for their hypertension other than excess weight.1
5. Urinalysis is an insensitive tool for diagnosing many renal causes of hypertension (reflux nephropathy, Wilms' tumor, renovascular and polycystic diseases).
6. A diagnosis of essential hypertension should always be reconsidered and further evaluation undertaken if the clinical status of the patient changes (eg, blood pressure becomes refractory to therapy or new symptoms appear suggesting another etiology).
Having considered these guidelines, I recommend the following tests as a baseline screen: CBC, urinalysis, blood urea nitrogen, creatinine, sodium, potassium, chloride, bicarbonate, fasting cholesterol and triglycerides, echocardiogram, and renal ultrasonography. These tests are helpful in uncovering some of the more silent causes of childhood hypertension (eg, primary aldosteronism, chronic glomerulonephritis, and reflux nephropathy). The cholesterol and triglyceride measurements are important in assessing additional risk factors the child may have. The echocardiogram is a useful and sensitive test for assessing end-organ damage due to the hypertension. Abnormal findings on the echocardiogram should suggest the need for pharmacologic intervention.
If a diagnosis cannot be established following physical examination and the tests described above, more extensive and invasive tests are needed. The quantitation of catecholamines from a 24-hour urine collection is a reliable test for pheochromocytoma. Computerized tomography of the abdomen as well as the 131I MIBG scan are additional aids for localizing the tumor. Before proceeding with these tests, you may want to review the child's workup with a nephrologist.
Renovascular and renal parenchymal causes of hypertension generally require more specific tests as well. Peripheral serum renin and aldosterone tests, along with a spot urine sample for sodium, are recommended.10 These samples can be collected in the office or at the hospital outpatient laboratory. Record should be made as to whether samples were collected after a period of recumbency (30 minutes or more) or while the child was ambulatory. Elevated serum aldosterone with low renin supports the diagnosis of aldosteronism. An elevated peripheral renin should be helpful in determining the need for renal arteriography. However, most investigators have found the peripheral renin too insensitive a measure to be the sole determinant.8 More recently, Captopril (an inhibitor of the enzyme that converts angiotensin I to angiotensin II) has been proposed as a noninvasive diagnostic screen for renovascular hypertension. It has been used as an oral challenge test (peripheral renin measured before and after Captopril) and as part of the radioisotope renogram (renal scan before and after oral Captopril).6,11 Although these newer techniques are promising, the present evidence dictates the need for arteriography and renal vein renin sampling for those children suspected of having renovascular hypertension.12
If the urinalysis or abdominal ultrasound findings are suggestive of renal parenchymal disease, an intravenous pyelogram, radioisotope renogram, and voiding cystourethrogram should be considered. Because each of these tests has a different sensitivity for a given diagnosis, I suggest obtaining the advice of a radiologist on which test to order first. For example, if the renal ultrasound shows small kidneys with presumed scarring, a voiding cystourethrogram is likely to be the test of choice (reflux nephropathy).
I believe the pediatrician is best able to evaluate the child with hypertension. The pediatrician is qualified to decide whether a child has elevated blood pressure and is in the best position to determine the level of evaluation needed for a given child. The pediatrician also enjoys the confidence of both the child and the family, which makes the recommendations given by the pediatrician more likely to be followed. Let us not lose sight of our role as advocates for the child when we approach the challenge of evaluating the hypertensive child.
1. Report of the Second Task Force on Blood Pressure Control in Children - 1987. Pediatrics 1987; 19:1-25.
2. Burke W, Motulky AG: Hypertension: Some unanswered questions. JAMA 1985; 253:2260-2261.
3. Feld LG. Springate JE: Hypertension in children. Current Problems in Pediatrics 1988; 18:317-373.
4. Koster M. David GK: Reversible severe hypertension due to licotice ingestion. N Engl } Med 1968; 278:1381-1383.
5. Adelman RD: Smokeless tobacco and hypertension in an adolescent. Pediatrics 1987; 79:837-838.
6. Idrissi A. Fournier A. Bondailliez B. et al: The Captopril challenge test as a screening test for renovascular hypertension. Kidney /nt 1988; 34(suppl 25):5138-5141.
7. Ogbom MR. Crocker JFS: Investigation of pediatric hypertension. Am ] Dis Child 1987; 141:1205-1209.
8. Daniels SR. Loggie JMH, McEnery PT. et al: Clinical spectrum of intrinsic renovascular hypertension in children. Pediatrics 1987; 80:698-704.
9. Falkner B: Management of hypertension in children and adolescents. Am Earn Physician 1986; 34:101-109.
10. Luetscher JA. Weinberger MH. Dowdy AJ, et al: Effects of sodium loading, sodium depletion and posture on plasma aldosterone concentration and renin activity in hypertensive patients. J Clin Endocrinol Metofe 1969; 29:1310-1318.
11. Geyskes GG, Oei HY, Puylaert CB, et al: Renovascular hypertension identified by captopril-induced changes in the renogram. Hypertension 1987; 9:451-458.
12. Diament MJ, Stanley P, Boechat MI, et al: Pediatnc hypertension: An approach to imaging. Pediorr Radiol 1986; 16:461-467.
Definition of Hypertension in Children
Criteria for Classification of Hypertension by Age Group