White Coat Hypertension and Target Organ Involvement: the Impact of Different Cut-Off Levels on Albuminuria and Left Ventricular Mass and Geometry

Journal of Human Hypertension (1998) 12, 433–439  1998 Stockton Press. All rights reserved 0950-9240/98 $12.00 http://www.stockton-press.co.uk/jhh ORIGINAL ARTICLE White coat hypertension and target organ involvement: the impact of different cut-off levels on albuminuria and left ventricular mass and geometry

AHøegholm, KS Kristensen, LE Bang and JW Nielsen Department of Internal Medicine, County Central Hospital, N{stved, Denmark

The aim of this cross-sectional study which took place atory daytime BP of 135.6/90.4 mm Hg was found to cor- in a hypertension clinic at a district general hospital in respond to an office BP of 140/90 mm Hg in normal con- Denmark was to make a pragmatic definition of white trols; used as a cut-off level in patients with newly coat hypertension. A total of 420 patients were referred diagnosed hypertension it separated 19% as white coat consecutively from general practice with newly diag- hypertensives. The end-organ involvement of these nosed untreated essential hypertension and 146 normal white coat hypertensives differed significantly from subjects were drawn at random from the Danish those with established hypertension but not from the national register. The following measurements were normal controls. Lower cut-off levels were less efficient taken: office blood pressure; 24-h ambulatory blood in this respect, as was the case when the systolic BP pressure (BP) monitoring; echocardiography with deter- was not taken into account. mination of left ventricular mass index and relative wall In conclusion a pragmatic definition of white coat thickness; and early morning urine albumin/creatinine hypertension should—apart from well-established ratios. hypertensive office measurements—include a cut-off Four different cut-off levels were studied. An ambul- level close to 135/90 mm Hg ambulatory daytime BP.

Keywords: white coat hypertension; target organ damage

Introduction morbid and mortal events4,8 and intervention studies;9 a value of 140/90 mm Hg is widely accepted for Blood pressure (BP) measured under standardised adults.10,11 In contrast, no similar large scale studies conditions at the office and BP during daily life as on morbidity/mortality and ABP exist, and hence measured by ambulatory monitoring often give dif- cut-off values for ABP have not gained consensus.12 ferent results. The correlation between values from This dilemma had led to a disputable practise in these two measurement modalities is relatively low which cross-sectional studies on ‘normal popu- with reported coefficients of correlation around 0.5– lations’ of limited size have been the substrates for 0.7.1–3 It is unsolved which measure is ‘true’, ie, 4 different types of computations. Most often a per- which best reflects the ‘usual’ BP and prognosis. centile or standardised deviation from the mean of The lack of identity between the two methods of the distribution is computed, in other cases a BP of measurement yields a group of patients with the 140/90 mm Hg at the office has been used to calcu- combination of hypertensive office BP and normal late a corresponding ABP value. Some studies make ambulatory BP (ABP), which is generally referred to conversions for different gender and age groups sep- 5,6 as ‘white coat hypertension’. The literature has arately, which makes the results less easy to handle shown numerous definitions of this condition and, in clinical practice.2 Evidently, the varying cut-off consequently, great variation in the characterisation levels give very large differences in the frequency of 7 of the group has evolved. The primary explanation white coat hypertension, ie, 7–52%.13–16 for these diversities is the lack of widely accepted In the absence of large scale longitudinal studies ABP normalcy limits. The cut-off values between and generally accepted cut-off values we have made normal and hypertensive office BP has been deter- the present analysis on BP and surrogate end-points mined on the basis of major epidemiological longi- with the purpose of examining the effect of different tudinal studies regarding BP level and frequency of cut-off levels for the definition of normalcy and hence white coat hypertension. Our cohort of patients offers a potential for describing a white coat Correspondence: Dr A Høegholm, Medicinsk Afdeling, Centralsy- group with a normal level of target organ involve- gehuset, DK-4700 N{stved, Denmark ment, in that it includes a genuine normal control Received 26 January 1998; revised 7 April 1998; accepted 16 group and has data on both albuminuria and echo- April 1998 cardiography. White coat hypertension and target organ involvement AHøegholm et al 434 Subjects and methods were obtained, but all ambulatory readings were considered acceptable, defined as at least 18 day- In order to estimate the frequency of white coat time readings from outside the office. hypertension we have performed 24-h ABP monitor- The monitorings were edited for supposedly arti- ing in 420 consecutive patients with newly diagnosed factual measurements, ie, single measurements asso- mild-to-moderate hypertension. The inclusion cri- ciated with either an unexplained pulse rate above terion was that their general practioner had planned 120/min, or an unexplained systolic or diastolic to start anti-hypertensive treatment, but had not yet pressure deviation exceeding 30% compared with instituted it. In order to avoid confounding, patients the readings immediately before and after. The sub- with other known reasons for albuminuria, eg, jects did not wear activity monitors or the like, but known diabetes or renal disease, were excluded from according to their diaries almost everybody was the statistical analyses regarding albuminuria data. awake throughout the period between 08.00 am to No patients with major cardiovascular diseases 09.59 pm, which was subsequently defined as the were included. daytime period, and the daytime ABP was the aver- As part of a multicentre study to establish the dis- age of the readings obtained in this period. tribution of ABP in the population,17 we performed 24-h BP monitoring in subjects drawn at random from the Danish national register. There was an Echocardiography intended similar number of subjects in different gen- One hundred and thirty-two (132) of the 146 normal der and age strata from 20–79 years. Subjects with subjects (90.4%) and 333 of the 420 hypertensive known renal or cardiovascular disease, diabetes, patients (79.3%) were examined with echocardio- hypertension or other condition necessitating ther- graphy. The same experienced physician examined apy with drugs with anti-hypertensive effects had all subjects with an Aloka 720 SSD mechanical sec- been excluded, leaving 146 monitorings. tor scanner (Aloka, Tokyo, Japan). All the examin- The study was in accordance with The Second ations were videotaped. In the parasternal short-axis Declaration of Helsinki, and was approved by the view two to four M-mode photographs were taken local ethics committee. All participating subjects at the optimum perpendicular plane just below the gave their informed consent. tips of the mitral leaflets. These photographs were evaluated independently by two observers, who Blood pressure measurements were blind regarding the age, gender, size, and BP of the subjects. The two observers could individually All the referred patients were considered by their inspect the videotapes, if it was felt necessary. All general practioner to have hypertension, and were evaluations were performed in the same batch; scheduled to start anti-hypertensive treatment in the hence, the readings were different from the data pre- near future. No patient had for the 5 weeks prior to viously reported.21 The reason for the blinding pro- the study received any anti-hypertensive drug; no cedure was to assure an estimate free from bias, but patient was taken off medication in order to enter of course at the cost of lower accuracy which, how- the study, but, a third had previously been on such ever, is of less importance due to the relatively high medication (most often because of previous hyper- number of subjects. No patient was excluded from tension, but also large numbers because of palpi- analysis because of difficulties in obtaining images, tations, oedema and tremor) which had been taken although 55 (11.4%) were considered of minor tech- for median 3 years. The BP of the patients who had nical quality. Interventricular septal thickness never been on anti-hypertensive medication had (IVST), posterior wall thickness (PWT), and end- been followed by the general practitioners for diastolic left ventriculum internal diameter (LVID) median 4 months. The practitioners had determined were determined according to the Penn-conven- the BP at least three times (median 4) with at least tion.22 Left ventricular mass was computed as 1.04 weekly intervals. Nearly everybody used a standard [(IVST + LVID + PWT)3 − LVID3] − 13.6 g22 and was size cuff (12 × 35 cm) and in the sitting position; indexed for body surface area (BSA) (left ventricular approximately a third used aneroid sphygmoman- mass index (LVMI)); relative wall thickness (RWT) ometers, the remainder mercury columns. The aver- was computed as 2 × PWT/LVID.23 age of the diastolic BP measured in this way was The between-observer coefficients of variation above 90 mm Hg in all subjects. were 10.8% and 9.5% for LVMI and RWT, respect- The office BP of the normotensive subjects was ively. The mean of the measurements of the two determined as a mean of five measurements with a observers was used for data analysis. Hawksley ‘random zero’ sphygmomanometer (Hawksley and Sons Ltd, Lancing, UK),18 performed Urinalysis in the sitting position after 15-min of rest in the hypertension clinic. All the subjects were asked to bring with them an The subjects wore TM 2420 ABP recorders (A&D, early morning urine sample, defined as the first Tokyo, Japan)19,20 for at least 24 successive hours on voided specimen on arising; 127 of the 146 normal working days, during which period they performed subjects (87.0%) and 302 of the 420 hypertensive their habitual daily activities; the recorders took patients (71.9%) complied in this regard. The urine readings every 15 min 07.00 am to 10.59 pm and samples were stored at −20°C, until analyses were every 30 min 11.00 pm to 06.59 am. In 17 subjects carried out using the same batches of reagents. Albu- (3%) less than 40% of the stipulated measurements min concentrations were determined by an immu- White coat hypertension and target organ involvement AHøegholm et al 435 nonephelometric technique.24 Creatinine concen- ´ trations were measured by the routine Jaffe method (SMA-II, Technicon, Tarrytown, NY, USA). The results from this analysis have previously been published in another context.25 Albuminuria was estimated as the ratio between albumin and creatinine. The day-to-day coefficient of variation for this measurement is 22.8% in normal subjects.25

Statistical analyses Data are expressed as mean ± s.d. unless otherwise stated; since the microalbuminuria data are skewed, they were transformed logarithmically before being statistically tested. ⌾2 test was used for group com- parisons regarding categorised data. Group concentrations regarding continuous variables were performed with unpaired t-test or one-way analysis of variance followed by Student–Newman–Keul’s procedure if more than two groups were analysed. In order to define cut-off levels, simple regression analysis between office and ambulatory daytime BPs were performed in the normal controls, and the ambulatory level corresponding to 140/90 mm Hg at the office was determined (later referred to as the regression method). Another cut-off level was produced by determining which percentiles in normal controls correspond to 140/90 mm Hg, the same percentiles were then applied to the ambulatory data (later referred to as equal percentile method). P Ͻ 0.05 was considered significant. All computations were carried out with SPSS/PC+ V2.0 software (SPSS, Chicago, IL, USA).

Results The subjects characteristics are shown in Table 1. Most of the BP data differ significantly, as could be expected, but age and body mass index (BMI) differ as well. The difference regarding age is explained by the fact that the normal controls were recruited with Figure 1 Percentage of patients who have different forms of target organ involvement (level above the 95th percentile of normal an intended even age stratification. controls) according to ambulatory daytime BP cut-off levels, (a), The daytime ABP cut-off level between normoten- systolic and (b) diastolic BPs. (२) albuminuria, (×) left ventricular sion and hypertension has a pronounced effect on mass index, (b) relative wall thickness. the amount of end-organ involvement in the resulting groups. Figure 1 shows the relation between the cut-off level and the percentage of end- group were defined as the levels above which end- organ involvement in the group of newly diagnosed organ involvement was present. For the ratio hypertensives. In this analysis the 95th percentiles between albumin and creatinine (on a molar basis of ecco- and albuminuria-data in the normal control after logarithmic transformation) the 95th percentile

Table 1 Blood pressures and demographic data for normal controls and patients with newly diagnosed hypertension

Characteristics Normal controls Hypertensive P-value (n = 146) (n = 420)

Women/men 74/72 223/197 0.615 Age (years) 52.1 ± 15.8 47.7 ± 12.2 0.001 Weight (kg) 72.4 ± 14.1 75.0 ± 15.5 0.074 Height (cm) 170.4 ± 9.3 170.4 ± 9.3 0.962 Body mass index (kg/m2) 24.8 ± 3.8 25.7 ± 4.2 0.025 Office BP (mm Hg) 129.1 ± 19.3/81.5 ± 11.4 165.4 ± 18.0/103.7 ± 7.4 Ͻ0.001/Ͻ0.001 Average daytime ambulatory BP (mm Hg) 131.6 ± 19.1/80.7 ± 9.5 146.9 ± 18.9 /96.1 ± 10.1 Ͻ0.001/Ͻ0.001 Average daytime heart rate (min−1) 76.1 ± 8.3 80.1 ± 9.4 Ͻ0.001 White coat hypertension and target organ involvement AHøegholm et al 436 was 0.26. For RWT and LVMI it was 0.45 and for diastolic ambulatory daytime BP, the percentage 148.1 g/m2, respectively. These levels are higher of white coat hypertension is 30.7%. If the cut-off than reported by others,22,23 probably due to the level is 135/85 mm Hg ambulatory daytime BP, as efficient blinding procedure causing a larger scatter proposed by some groups,12 the percentage of white about the mean. There was a significant gender dif- coat hypertension is calculated to 9.3%. It is widely ference in echocardiographic data but not in albumi- accepted that 140/90 mm Hg measured at the office nuria. For men the mean LVMI was can be considered normotensive.10,11 This office 110.7 ± 26.7 g/m2 and mean RWT 0.36 ± 0.06; the measurement was transformed into corresponding corresponding values for women were ambulatory daytime measurements in two ways. 85.8 ± 26.0 g/m2 and 0.34 ± 0.06. However, substitut- Firstly, the simple regression method15 gives a cut- ing gender-based 95th percentiles (which are less off level of 135.6/90.4 mm Hg daytime ABP. Sec- reliably defined due to a lower number of subjects) ondly, in the normal controls an office BP of 140/90 for the above mentioned had no significant influence mm Hg correspond to percentile 71.9 for systolic on the results and was therefore not used. and 80.8 for diastolic pressure; when the same The cut-off level has an impact on the resulting pecentiles are applied to the ambulatory data a cut- prevalence of white coat hypertension, too. Figure 2 off level of 139.6/89.4 mm Hg is found. These two shows this interrelation for systolic and diastolic cut-off levels yield percentages of white coat hyper- pressures separately. If the cut-off level is 90 mm Hg tensives amounting to 19.0% and 18.6%, respectively, if applied to the hypertensive group. The influence of the four cut-off levels on end- organ involvement in the resulting groups of white coat hypertensives and established hypertensives is shown in Table 2. The influence of the office BP on the frequency of white coat hypertension can be calculated. Table 3 shows the frequency of white coat hypertension in the group of newly diagnosed hypertensives, according to different levels of office BP; the cut-off level in this table is 135.6/90.4 mm Hg (with a white coat percentage of 19.0%), as derived by the regression method. If the 45 of the 146 normal controls (30.8%) who did not have an office BP below 140/90 mm Hg were excluded from the analyses of target organ damage given in Table 2, the results deviate modestly. The mean values in the normotensive group were 92.0 g/m2 for LVMI, 0.341 for RWT and −0.220 for albuminuria; all four groups of patients with established hypertension (see Table 2) were still significantly different from the normotensive controls, but only by using definition #2 (equal percentile method) and #3 (regression method) the resulting groups of white coat hypertensives ended up with target organ involvement significantly lower than the corresponding established hypertensives but not different from the normotensive controls. The percentage of the normal controls who are not normotensive differ according the selected cut-off level. Definition #1 (cut-off 90 mm Hg diastolic ambulatory daytime BP) makes 18.5% of the nor- mals non-normotensive, whereas the corresponding figures for the other definitions are: definition #2 (equal percentile method with cut-off 139.6/89.4 mm Hg daytime ABP) 34.9%, definition #3 (regression method with cut-off 135.6/90.4 mm Hg daytime ABP) 36.4%, definition #4 (cut-off 135/85 mm Hg daytime ABP) 43.2%.

Discussion We have aimed at describing a ‘real’ white coat Figure 2 Frequency of white coat hypertension as a function of hypertensive group as individuals with the highest ambulatory daytime BP cut-off levels, (a) systolic and (b) diastolic BPs. Eg, if a cut-off level of 90 mm Hg diastolic ambulatory day- ambulatory systolic and diastolic values but without time BP is applied to our group of newly diagnosed hypertensive a greater extent of target organ damage than a normal patients 30.5% will be deﬁned as white coat hypertensives. control group. Our study was conducted on con- White coat hypertension and target organ involvement AHøegholm et al 437 Table 2 End-organ involvement in normotensives, white coat hypertensives and established hypertensives, where the latter two groups are differently deﬁned

Cut-off level Cut-off level Cut-off level Cut-off level 90 mm Hg diastolic 139.6/89.4 mm Hg 135.6/90.4 mm 135/85 mm Hg (def. #1) (def. #2) Hg (def. #3) (def. #4)

Normal White coat Hypertens. White coat Hypertens. White coat Hypertens. White coat Hypertens. controls (n = 129) (n = 291) (n = 78) (n = 342) (n = 80) (n = 340) (n = 39) (n = 31) (n = 146) (30.7%) (69.3%) (18.6%) (81.4%) (19.0%) (81.0%) (9.3%) (90.7%)

Albuminuria −0.161 0.038† 0.083 −0.110† 0.085* −0.097† 0.081* 0.090† 0.061* log mol/mol (± 0.357) (± 0.374) (± 0.440) (± 0.360) (± 0.430) (± 0.360) (± 0.431) (± 0.371) (± 0.427) creatinine Left 98.2 98.8† 106.7* 92.8† 106.8* 90.5† 107.4* 86.0† 106.0* ventricular (± 29.1) (± 27.9) (± 27.6) (± 24.6) (± 28.0) (± 20.4) (± 28.4) (± 17.3) (± 28.1) mass index g/m2 Relative 0.352 0.352† 0.373* 0.350 0.371* 0.349† 0.371* 0.348 0.369* wall (± 0.058) (± 0.066) (± 0.072) (± 0.061) (± 0.073) (± 0.061) (± 0.073) (± 0.068) (± 0.071) thickness

*P Ͻ 0.05 by Student–Newman–Keul’s procedure compared with normotensives; †P Ͻ 0.05 compared with corresponding hypertensives. secutive patients referred per protocol from general no obvious threshold for the diastolic pressures. It practice with newly diagnosed hypertension and could be argued that this lack of a clearcut threshold with the practitioner intending to start pharmacol- is explained by the significant gender difference in ogic anti-hypertensive treatment. Furthermore, for RWT and LVMI, but also when analysing the data comparison, we included a control group of non- for men and women separately no definite threshold treated healthy individuals drawn at random from level can be deduced. the national register. Figure 2 shows that apart from influencing the The main finding from our study is that in the extent of end-organ damage the frequency of white hypertensive population daytime ABP below levels coat hypertension in the referred (hypertensive) in the range of 135.6–139.6/89.4–90.4 mm Hg is individuals is also determined by the applied sys- associated with the lowest prevalence of end-organ tolic and diastolic cut-off values. We have focused damage, and that white coat hypertension defined on four different cut-off levels derived from the on this basis is not significantly different from the literature, the tradition and two different ways of normal control group in this regard (Table 2). transforming the generally accepted WHO-limits The association between BP cut-off level and the from office into ABP. The definitions and the calcu- presence of different forms of end-organ is illus- lated resulting frequencies of white coat hyperten- trated in Figure 1. It should be noted that only a few sion among the patients with newly diagnosed patients have very low BP (27 were below 120 hypertension were as follows: mm Hg systolic and 15 below 80 mm Hg diastolic), thus explaining the fluctuation of the curves at the (1) A daytime ABP cut-off level of 90 mm Hg and lower end of BP levels; furthermore, regarding not taking systolic BP into consideration was microalbuminuria the J-shaped curve could be used in several earlier papers6,26 and results in partly explained by the presence of three patients 30.7% white coat hypertensives. (all younger women) with microalbuminuria in spite (2) Our own equal percentile method, where 140/90 of low ABPs. It is interesting to note that the pres- mm Hg at the office is transformed into ambulat- ence of end-organ damage seems to be largely unre- ory daytime BP, gives a cut-off level of lated to the diastolic BPs, whereas the impact of sys- 139.6/89.4 mm Hg, thus rendering 18.6% white tolic ABP is much stronger. The threshold where the coat hypertensive. prevalence of end-organ damage increases seems to (3) Baumgart et al27 advocate a simple regression be in the BP range of 130–145 mm Hg systolic with method to transform 140/90 mm Hg at the office

Table 3 Prevalence of white coat hypertensives among newly diagnosed hypertensive patients with different levels of ofﬁce BP

Diastolic (mm Hg) Systolic (mm Hg)

Ͻ150 150–160 160–170 Ͼ170

90–100 45.5% (15/33) 54.5% (12/22) 36.0% (9/25) 18.2% (4/22) 100–105 25.0% (3/12) 25.6% (10/39) 11.4% (4/35) 13.2% (5/38) Ͼ 105 25.0% (2/8) 13.3% (4/30) 10.9% (6/55) 6.0% (6/100)

White coat hypertension defined by an ambulatory daytime BP below 135.6/90.4 mm Hg. White coat hypertension and target organ involvement AHøegholm et al 438 into an ambulatory value, which in our normal level of 95 mm Hg diastolic ABP yielding a very control group shows that a daytime ABP value high white coat percentage in spite of the tertiary of 135.6/90.4 mm Hg corresponds to an office setting; the normotensive controls were recruited by value of 140/90 mm Hg. The resulting white coat virtue of their referral to the hypertension clinic (!); prevalence is then 19.0%. ecco-doppler data insinuated incipient diastolic (4) Finally, the present literature often regards dysfunction in the white coat group, but are prob- 135/85 mm Hg as a reasonable cut-off level,28 ably the result of erroneous calculations.31 Our own giving a prevalence of 9.3% white coat hyperten- previous study on albuminuria25 supported the view sion. that white coat hypertensives were at intermediate risk; however, these data were produced in the era If the systolic BP was not taken into account at all of minor emphasis on systolic BP, in that we had (definition #1), the albumin excretion in the defined white coat hypertension only on the basis resulting white coat group was higher than in the of daytime diastolic ABP being below 90 mm Hg (as normal controls, as we have published previously.25 is the case in definition #1 in Table 2). In the present When the cut-off criterion was sharpened by includ- study it is shown that when daytime systolic ABP ing the systolic BP as in definition #2 (around below 135–140 mm Hg is included in the definition 140/90 mm Hg) the white coat group had an albu- there is no significant difference between white coat min excretion and an LVMI similar to the normal hypertensives and the normal group, neither regard- controls but an RWT not significantly different from ing microalbuminuria nor echocardiographic data. the established hypertensives. When the cut-off cri- At present there are a number of cross-sectional terion was made more rigorous (around 135/90 studies,14,32,33 but only two longitudinal studies34,35 mm Hg as in definition #3) the differentiation implying the benign nature of white coat hyperten- seemed optimum. By sharpening the cut-off cri- sion; of course, further prognostic studies are terion even further to 135/85 mm Hg (definition #4) awaited with great interest. Meanwhile the present the differentiating power was lost regarding echocar- cross-sectional study supports the view that white diographic data. Thus, our data indicate that a day- coat hypertensives, when properly defined, are at time ABP of around 135/90 mm Hg is the proper dis- the same risk as normal controls. criminating level for white coat hypertension. This The main limitation of our study is its cross- limit is somewhat higher than reported by others; sectional nature. Another limitation is that a large the explanation for this can—apart from the differ- fraction of the normal controls had an office BP ent methodologies—rely on the different popu- exceeding 140/90 mm Hg. This does not imply that lations studied, eg, unlike in mediterranean coun- they are hypertensive, mainly because such a diag- tries2 the ’siesta effect’ is not discernible in the 24- nosis has to be based on several measurements with h BP profiles in the northern European countries.17 adequate time intervals. However, this way of meas- If one accepts the cut-off criterion as suggested uring BP is the same as the majority of the epidemi- from our data, the prevalence of white coat hyper- ological studies have used, and therefore it can be tension is approximately 19% among patients with assumed that a BP measured in this way which is newly diagnosed mild-to-moderate hypertension, a lower than 140/90 mm Hg is associated with very somewhat lower figure than previously reported,6 low risk. If the 30.8% of the normal subjects with but similar to the figures reported by others.29 office BPs above 140/90 mm Hg are excluded from It is obvious, though that the frequency of white the analysis of end-organ damage shown in Table 2 coat hypertension is declining with more severe we still find that criterion #3 (the regression method) office BP elevations at admission, as is illustrated defines white coat hypertensives with significantly in Table 3, where 135.6/90.4 mm Hg is the cut-off less end-organ involvement than established hyper- criterion (definition #3 from Table 2). It is seen that tensives but without significant differences from the if the office BP is above 160/100 mm Hg the prob- normotensive controls. A final limitation of our ability of white coat hypertension is materially study is that the key computations rely on one defi- reduced. Verdecchia et al29 have reported a similar nition of a normal amount of end-organ damage, relationship. thus disregarding the independent gender- and age- The opinions on the white coat problem have related differences in echocardiographic variables. shown remarkable fluctuation during the last dec- The main reason for this choice of method is that ade. From being a condition regarded as potentially the sample size of our normal subjects is too small to benign and a source of overtreatment with anti- allow such subdivisions, but control computations hypertensive medications, the tide has now shifted where the gender-related differences were taken into towards sceptiscism and a fear that the condition is account showed that this had no significant effect a prehypertensive state30 that actually has cardio- on the results. vascular dysfunction13 and signs of target organ Conclusively, our findings suggest that in round damage25 leaving anti-hypertensive treatment con- numbers a cut-off level of 135/90 mm Hg daytime siderable after all. However we find the evidence ABP is considered for the definition of normalcy, in supporting these viewpoints still rather weak. that white coat hypertensives defined in this way Although the study by Bidlingmeyer et al30 is longi- have a similar level of target organ involvement as tudinal, is is essentially a re-examination of prob- have normal controls. Such a cut-off level yields ably very selected patients with a very low follow- approximately 19% white coat hypertensives among up rate. The study by Glen et al13 is flawed in several newly diagnosed patients with mild-to-moderate aspects. The authors applied a very high cut-off hypertension. White coat hypertension and target organ involvement AHøegholm et al 439 References line hypertension in Tecumseh, Michigan. Hyperten- sion 1990; 16: 617–623. 1Høegholm A, Wiinberg N, Kristensen KS. The effect of 17 Wiinberg N et al. 24-hour ambulatory blood pressure antihypertensive treatment with dihydropyridine cal- in 352 normal Danish subjects related to age and gen- cium antagonists on white coat hypertension. Blood der. Am J Hypertens 1995; 8: 978–986. Press Monit 1996; 1: 375–380. 18 Wright BM, Dorsey F. Random-zero-sphygmoman- 2 Mancia G et al. Ambulatory blood pressure normality: ometer. Lancet 1970; i: 337–338. 19 Høegholm A et al. Clinical evaluation of the Takeda results of the PAMELA study. J Hypertens 1995; 13: Medical (A&D) TM 2420 ambulatory blood pressure 1377–1390. monitor. Practical experience and comparison with 3 Sokolow M. Werdegar D, Kain HK, Hinman AT. direct and indirect measurements. Scand J Clin Lab Relationship between level of blood pressure meas- Invest 1992; 52: 261–268. ured casually and by portable recorders and severity 20 Clark S et al. Ambulatory blood pressure monitoring: of complications in essential hypertension. Circulation validation of the accuracy and reliability of the TM- 1966; 34: 279–298. 2420 according to the AAMI recommendations. J Hum 4 McMahon S et al. Blood pressure, stroke, and coronary Hypertens 1991; 5: 77–82. artery disease. Part 1, prolonged differences in blood 21 Høegholm A et al. Left ventricular mass and geometry pressure: prospective observational studies corrected in patients with established hypertension and white for the regression dilution bias. Lancet 1990; 335: coat hypertension. Am J Hypertens 1993; 6: 282–286. 765–774. 22 Devereux RB, Reichek N. Echocardiographic determi- 5 Pickering TG et al. How common is white coat hyper- nation of left ventricular mass in man: anatomic vali- tension? JAMA 1988; 259: 225–228. ø dation of the method. Circulation 1977; 55: 613–618. 6Hegholm A, Kristensen KS, Madsen NH, Svendsen 23 Reichek N, Devereux RB. Reliable estimation of peak TL. White coat hypertension diagnosed by 24-hour left ventricular pressure by M-mode echocardio- ambulatory monitoring. Examination of 159 newly graphic-determined end-diastolic relative wall thick- diagnosed hypertensives. Am J Hypertens 1992; 5: ness: identification of severe valvular aortic stenosis in 64–70. adult patients. Am Heart J 1982; 103: 202–209. 7 Mancia G, Zanchetti A. Editors’ corner: White-coat 24 Irgens-Møller L, Hemmingsen L, Holm J. Diagnostic hypertension: misnomers, misconceptions and misun- value of microalbuminuria in pre-eclampsia. Clin derstandings. What should we do next? J Hypertens Chim Acta 1986; 157: 295–298. 1996; 14: 1049–1052. 25 Høegholm A et al. Microalbuminuria in 411 untreated 8 The Swedish Council on technology assessment in individuals with established hypertension, white coat health care. Hypertension and risk. J Internal Med hypertension and normotension. Hypertension 1994; 1995; 238 (Suppl 737): 35–45. 24: 101–105. 9 Collins R et al. Blood pressure, stroke, and coronary 26 Waeber B et al. Are some hypertensive patients over- disease. Part 2, short-term reductions in blood press- treated? Lancet 1987; ii: 732–734. ¨ ure: overview of randomised drug trials in their epide- 27 Baumgart P, Walger P, Jurgens U, Rahn KH. Reference miological context. Lancet 1990; 335: 827–838. data for ambulatory blood pressure monitoring: What 10 Subcommittee of WHO/ISH Mild Hypertension Liai- results are equivalent to the established limits of office son Committee. Summary of 1993 World Health blood pressure? Klin Wochenschrift 1990; 68: 723–727. Organisation-International Society of Hypertension 28 O’Brien E, Staessen J. Normotension and hypertension guidelines for the management of mild hypertension. as defined by 24-hour ambulatory blood pressure Br Med J 1993; 307: 1541–1546. monitoring. Blood Pressure 1995; 4: 266–282. 11 Joint National Committee on Detection, Evaluation, 29 Verdecchia P et al. White coat hypertension and white and Treatment of High Blood Pressure. The fifth report coat effect. Similarities and differences. Am J Hyper- of the Joint National Committee on Detection, Evalu- tens 1995; 8: 790–798. ation, and Treatment of High Blood Pressure (JNC V). 30 Bidlingmeyer I et al. Isolated office hypertension: a Arch Intern Med 1993; 153: 154–183. prehypertensive state? J Hypertens 1996; 14: 327–332. 12 Pickering TG. A review of national guidelines on the 31 Kristensen KS, Høegholm A. White coat hypertension clinical use of ambulatory blood pressure monitoring. [letter]. Lancet 1996; 348: 1444. Blood Press Monit 1996; 1: 151–156. 32 White WB, Schulman P, McCabe EJ, Dey HM. Average 13 Glen SK et al. White-coat hypertension as a cause of daily blood pressure, not office blood pressure, deter- cardiovascular dysfunction. Lancet 1996; 348: 654– mines cardiac function in patients with hypertension. 657. JAMA 1989; 261: 873–877. 14 Verdecchia P et al. Variability between current defi- 33 Cavallini MC et al. Is white coat hypertension associa- nitions of ‘normal’ ambulatory blood pressure. Impli- ted with arterial disease or left ventricular hypertro- cation in the assessment of white coat hypertension. phy? Hypertension 1995; 26: 413–419. Hypertension 1992; 20: 555–562. 34 Verdecchia P et al. Ambulatory blood pressure: an ¨ 15 Baumgart P, Walger P, Jurgens U, Rahn KH. Reference independent predictor of prognosis in essential hyper- data for ambulatory blood pressure monitoring: What tension. Hypertension 1994; 24: 793–801. results are equivalent to the established limits of office 35 Perloff D, Sokolow M, Cowan R. The prognostic value blood pressure? Klin Wochenschrift 1990; 68: 723–727. of ambulatory monitoring. JAMA 1983; 249: 2793– 16 Julius S et al. ‘White coat’ versus ‘sustained’ border- 2798.