Blood Pressure Tables

To the Editor.—The Working Group of the National High Blood Pressure Education Program recently published updated guidelines on the diagnosis, evaluation, and treatment of high blood pressure (BP) in children and adolescents.1 This is a timely update, and the working group has done a commendable job in most areas of their recommendations. As a long-time student of children's BP measurement, however, I have some concerns about the current BP standards recommended by the working group,1 in particular, the validity of recommending age and height percentiles to express children's BP standards.In a recently published study from the San Antonio Children's Blood Pressure Study,2 we obtained triplicate BP readings by both the auscultatory2 and oscillometric (Dinamap Model 8100)3 methods in >8000 schoolchildren enrolled in kindergarten through 12th grade. We used the currently recommended BP-measuring technique (BP cuff width: 40–50% of the arm circumference; sitting position: K5 level at which the Korotkoff sounds disappear as the diastolic pressure; averaging multiple BP readings). In that study, auscultatory systolic pressures were correlated more closely with weight (r = 0.677) than with height (r = 0.626), as has been shown by many other studies. More importantly, we found that when age and weight were used in a partial correlation analysis, the effect of height on auscultatory systolic pressure level virtually disappeared (r = 0.068 for boys and 0.072 for girls). However, when adjusted for age and height, the correlation of systolic pressure with weight remained high (r = 0.343 for boys and 0.294 for girls), indicating that weight contributed much more importantly to systolic pressure than height. Although we do not believe it is wise to use a second variable in addition to age in children's BP standards, if one decides to use another measure of body size, weight seems to be a better choice than height. Gain and loss of weight correlate well with increases and decreases in BP levels, respectively, in children and adolescents.4We also reported normative BP standards separately for the oscillometric method3 because BP measurements by the 2 different methods are not interchangeable, although practitioners often use them interchangeably. The oscillometric systolic BP averaged ∼10 mm Hg higher and the oscillometric diastolic pressure averaged 5 mm Hg higher than the auscultatory BP readings.5 The correlations relative to weight and age as normative standards for oscillometric BP were very close to that reported for the auscultatory method described above.3A second reservation is for the working group's recommendation of a relatively complex BP standard requiring additional computational steps to classify the level of BP, a highly variable measurement. In a busy practice, following such guidelines is impractical relative to what is to be gained by such an approach. Assuming that time is not crucial, calculating body mass index is probably justified, because in this case the variables (weight and height) are highly reproducible. Unlike height or weight, children's BP levels are dynamic and vary continually, which is especially true for the single BP readings obtained in health care facilities. The first (or single) readings are usually higher than the average of multiple readings.6,7 The average of multiple BP readings is closer to the basal BP levels and is more reproducible, and its use is recommended by many national studies or committees for children8–11 as well as adults.9,12 The cumbersome part of BP measurement is preparing for it. Once the child realizes the painless and benign nature of the measurement, in most cases very little additional effort is required to obtain 2 more readings. The BP standards recommended by the working group are from single measurements (at least, most studies from which the working group derived their data are from single measurements). In the past, K4 (the level at which there is sudden muffling) was used for diastolic pressure for children and K5 for adolescents,9,11 and K5 values may not be available for small children. The committee should not recommend BP standards obtained by using methods not recommended by the committee. Also, national guidelines should be practical for primary care physicians who are encouraged to take BP as a routine procedure.BP standards from the San Antonio Children's Blood Pressure Study are presented according to age for boys and girls for both the auscultatory2 and oscillometric3 methods. To the best of my knowledge, this is the only set of normative BP standards obtained according to the recommendations of the American Heart Association Task Force9 and the previous working group10 in terms of the BP cuff-selection method, the number of BP measurements, and the use of K5 as the diastolic pressure. In using our normative BP standards,2 we recommended the same approach as used by the National Institutes of Health (NIH) Task Force—1987.11 The NIH normative data,11 however, are no longer acceptable because they have not been obtained in accordance with the current recommendations. The NIH Task Force approach takes into consideration the fact that BP levels are correlated with body size (weight, height, or body mass index) but does not require additional steps in the assessment of children's BP levels. If the average of multiple BP measurements is >90th percentile for the age of the child, consideration should be given to the possibility that the apparent abnormality is a result of obesity before any presumption that hypertensive disease is present.1,10,11 The classification recommended by the current working group1 can then be applied and the remainder of the recommendations followed. In many cases, the high BP may be due to body size and not true essential hypertension. In such cases, weight control would be the preferred therapeutic approach. Also, the diagnosis of hypertension should not be made on a single visit. The finding of abnormally high BP readings on at least 3 different visits is recommended for such a diagnosis.1,10,11 I do not believe the new way of presenting BP standards is an improvement over the approach used earlier by the NIH Task Force,11 because it is not only impractical but inaccurate.In summary, I believe that national standards of children's BP should be developed by prospective studies in accordance with the currently recommended methods. Older, inappropriately obtained data cannot be made appropriate by statistical manipulations. The data should be presented as a function of age separately for boys and girls. Although the San Antonio Children's Blood Pressure Study was a relatively small, regional study, auscultatory BP standards reported from that study2 could serve as interim standards until a larger national database becomes available.