White coat effect

The occurrence of a transient increase in blood pressure during sphygmomanometric measurements in the clinic environment was first described in 1897 by Scipione Riva-Rocci: ‘… the simple application of the instrument (the sphygmomanometer) can cause a temporary rise in blood pressure. It is therefore necessary to take not just one reading, but several in succession, such as 3 in 3 minutes or 5 in five minutes, until an average constant pressure is obtained …’ 1. However, the quantitative importance of this phenomenon was not fully appreciated until the early 1980s, when, with the advent of intra-arterial ambulatory blood pressure monitoring techniques, the precise assessment of dynamic changes in blood pressure occurring under different behavioural conditions became possible. One of the first attempts to quantitatively describe the pressor effect of the physician's visit was made in 1983 by our group. By using the Oxford intra-arterial ambulatory blood pressure monitoring system 2,3, we observed that the blood pressure rise associated with this phenomenon (i) becomes evident at the very beginning of the physician's visit, often before the time of actual blood pressure measurement, (ii) persists for approximately 10–15 min (i.e. the duration of the visit) and (iii) is accompanied by a parallel rise in heart rate (Fig. 1) 4,5. This observation was interpreted as the haemodynamic result of a patient's alarm reaction to the physician's visit, a reaction that was later referred to as the ‘white coat effect’ 6. The potential clinical importance of this phenomenon was emphasized by the demonstration that (i) it is of considerable magnitude, the maximal increase in intra-arterial systolic and diastolic blood pressure observed in the first 2–4 min of the physician's visit being on average 27/14 mmHg, (ii) its size is largely different in different subjects 4, which makes it hardly predictable in the individual patient and (iii) it does not easily fade with time, as it was of comparable magnitude at the time of four consecutive visits (always performed by the same physician) repeated over a 48-h intra-arterial recording period 5.Fig. 1: Orginal intra-arterial blood pressure recording showing the size and time course of the blood pressure increase during a physician's visit (direct measurement of the white coat effect). Arrows indicate the begining and the end of a 15-min physician's visit. Adapted with permission 4.However, for obvious practical reasons, these invasive techniques are not applicable in clinical studies including large groups of patients. Therefore ‘surrogate’ non-invasive solutions to the quantification of this phenomenon, more easily suitable applicable both to epidemiological research and to a clinical setting, have been proposed. The introduction of non-invasive automated blood pressure monitoring techniques has made it possible to obtain blood pressure values outside of the physician's office that are not influenced by the stressful conditions associated with clinic blood pressure measurement 7. This has led to the suggestion that the difference between clinic and mean daytime ambulatory or home blood pressure values might provide an easier, although indirect, assessment of the white coat effect 6. As a consequence of this assumption, although with limited experimental support, the condition characterized by a persistently high blood pressure in the clinic environment and a persistently normal blood pressure outside of it has been assumed to reflect the persistence over time of an alerting reaction to the physician's visit, being responsible for the repeated finding of elevated office blood pressure readings compared to normal daytime or home blood pressure levels. On this background, such a condition is commonly referred to as ‘white coat hypertension’ 8. Direct and surrogate measures of white coat effect: unresolved issues When considering the quantification and the pathophysiological and clinical implications of the white coat effect, a number of important issues need to be adequately clarified 9. These include the relationship between direct and surrogate measures of the white coat effect, the relationship between the size of the white coat effect and the degree of cardiovascular reactivity to stressful stimuli, and the possible prognostic relevance of these ‘reactive’ blood pressure changes. Relationship between direct and surrogate measures of white coat effect Notwithstanding its diffusion as the most popular approach to indirectly quantify the white coat effect, there is no evidence showing that the difference between clinic and ambulatory blood pressure values reflects the actual blood pressure rise induced by the physician's visit. By contrast, data recently obtained through techniques for non-invasive continuous finger blood pressure monitoring 10, employed during the physician's visit, support the opposite conclusion. In fact, the difference between clinic and ambulatory average daytime blood pressure has been shown to carry no 11 or negligible 12 correlation with the magnitude of the true white coat effect, as assessed directly by this approach (Fig. 2) 11.Fig. 2: Difference and lack of correlation between the directly assessed white coat effect (WCE) and the surrogate measure of this effect obtained by calculating the difference between clinic and daytime average blood pressure. Data are separately shown for systolic blood pressure (SBP), diastolic blood pressure (DBP) and heart rate (HR). Adapted with permission 11In this issue of the journal, Palatini et al. 13 reach the same conclusion, and add a further interesting contribution to this debate. In their study, no significant correlation was found between the clinic-daytime average blood pressure difference and the directly measured blood pressure reaction elicited by the physician's visit, regardless of the number of readings averaged to compute the latter blood pressure response. These results, together with those from two previous studies 11,12, are thus far from supporting the use of the difference between clinic and daytime blood pressure as a reliable quantitative index of the white coat effect. The reasons for the discrepancy between these direct and surrogate measures of this phenomenon are likely to be multifold, and partly depend on the different types of blood pressure measurements from which they are derived. In the case of direct assessment, blood pressure reactivity to the physician is precisely quantified both in its size and duration, taking the blood pressure levels recorded under the standardized resting condition preceding the physician's visit as a reference. In the case of the surrogate approach based on the clinic-daytime blood pressure difference, important methodological problems are to be anticipated. On the one hand, clinic blood pressure is quantified from the average of a limited number of readings (usually 1–3), with the risk of poor reproducibility of the values obtained. On the other hand, daytime average blood pressure can hardly be regarded as a reliable ‘reference’ blood pressure level, because it includes the effects on blood pressure of varying degrees of physical activity and of the multiple and different stressful conditions occurring in daily life. The choice of daytime blood pressure as a reference level for calculating the white coat effect is further undermined by the possibility that subjects hyper-reactive to the physician's visit might also be hyper-reactive to daily life stress, leading to higher daytime blood pressure levels and, at variance from the assumptions underlying the adoption of this surrogate method, to a smaller, rather than wider, clinic-daytime blood pressure difference. The study by Palatini et al. 13 provides clear evidence in this direction by showing that subjects with a more pronounced direct white coat effect also display increased daytime ambulatory blood pressure levels, and a higher level of 24-h average urinary epinephrine. Together with the observation that, in different subjects, daytime blood pressure may be differently influenced by the variable occurence of physical activity, challenging emotional conditions and periods of quiet rest during waking hours, this finding is clearly not in favour of using the comparison between clinic and daytime average blood pressure values as a means to indirectly quantify the white coat effect, and suggests that behavioural factors affecting daytime blood pressure might be more important determinants of the size of the clinic-daytime average blood pressure difference than the real white coat effect itself. The indirect assessment of white coat effect based on the difference between clinic and home blood pressure may be free from the above problems, but there are other methodological difficulties, such as the uncertainty on the number of home blood pressure readings to be obtained, the accuracy of the home blood pressure devices employed and the reliability of patients in reporting the measured values. These are the reasons why the practical relevance of this surrogate approach too needs to be investigated further. White coat effect and cardiovascular reactivity to stress An additional interesting finding in the paper by Palatini et al. 13 is the relationship observed between the true white coat effect and the blood pressure response to a stressful condition as represented by public speaking, a response that, conversely, showed no association with the surrogate measure of white coat effect as derived from the clinic-daytime blood pressure difference. These results are in line with those of Lantelme et al. 12,14 who reported a significant correlation between the true, directly measured, white coat effect and the blood pressure response to another laboratory stressor requiring a high degree of coping (i.e. the colour-word Stroop test). Conversely, in our own studies, the directly assessed white coat effect did not show any significant correlation with the blood pressure or heart rate response to a number of classical laboratory stressors, such as hand-grip exercise, cold pressor test or mental arithmetic, while only a marginal relationship was observed with the blood pressure response to the mirror drawing test 15,16. These discrepant results may reflect the variability in the responses to the different laboratory tests employed to evaluate stress reactivity 15, as well as the limited reproducibility of these responses 17, emphasizing that quantification of the cardiovascular effects of stressful stimuli may be influenced by the type of stressor employed because different tests might explore different features of stress reactivity. In fact, the Stroop test and public speaking may elicit a reaction more closely associated with the psychological mechanisms underlying the white coat effect than other more ‘physical’ stressors, such as hand grip and cold exposure, or a variably challenging condition, such as mental arithmetic 18. A more general problem in this field is whether the size of the white coat effect might also reflect the features of a more generic blood pressure and heart rate responsiveness to daily life stress, outside of the laboratory environment. Also in this case, the answer is not easily provided. As mentioned above, Palatini et al. 13 show an association between the size of the true, directly assessed, white coat effect and reactivity to daily life stress, at least when the latter is quantified as average daytime blood pressure levels and the 24-h average urinary output of epinephrine. However, when we assessed the response to daily life stress in our previous studies through the quantification of daytime or 24-h blood pressure variability, we did not find any significant association between the true white coat effect and quantification of blood pressure variability provided by the standard deviation of 24-h daytime average ambulatory blood pressure values, recorded on a beat-by-beat basis through an invasive procedure 14,15. Evidence against an association between the white coat effect and response to daily stress is also provided by the routine evaluation of commonly used non-invasive discontinuous ambulatory blood pressure recordings. In several cases, a pronounced rise in blood pressure observed during the first and/or the last hour of a 24-h ambulatory blood pressure monitoring (i.e. during those hours when patients are in the hospital environment for fitting or removal of the blood pressure recorder) is associated with a smooth blood pressure profile throughout the remaining hours of the day and night, with no enhancement in the degree of blood pressure fluctuations (Fig. 3).Fig. 3: Non-invasive discontinuous ambulatory blood pressure recording showing pronounced blood pressure reactivity in the first hours (hospital environment) and normal blood pressure behaviour throughout the remaining portion of the recording. BP, blood pressure; HR, heart rate. (Parati G., personal observation).It therefore appears that blood pressure and heart rate reactivity to the physician's visit are variably associated with different patterns of daily life blood pressure or heart rate changes, which prevents any definitive conclusions on this issue 19. Clinical relevance of the white coat effect Very limited evidence is available on the possible prognostic relevance of the true white coat effect. Neither in the study by Palatini et al. 13, nor in our own database 10, is any evidence available for an association between reactivity to the physician's visit and the cardiovascular complications of hypertension. In addition, in the study by Lantelme et al. 12, no association was found between direct or surrogate measures of the white coat effect and quantification of hypertension-induced target organ damage as represented by left ventricular hypertrophy. Thus, the real clinical relevance of blood pressure reactivity to the physician's visit remains to be clarified, and larger studies are needed to gather more solid evidence on this issue. Another aspect still requiring a more definite assessment is the possible association between the true white coat effect and the clinical condition characterized by persistently elevated blood pressure in the physician's office and by a persistently normal blood pressure at other times. This condition is commonly referred to as ‘white coat hypertension’ for which evidence both in favour and against its clinical relevance is available. Because the few studies investigating this association showed no correspondence between the actual ‘white coat’ phenomenon and the clinic-daytime or clinic-home blood pressure difference, the last WHO/ISH Guidelines recommended naming this condition ‘isolated office hypertension’ instead of the otherwise appealing term ‘white coat hypertension', a term that appears to be misleading based on the above evidence 20–22. The implication underlying such a recommendation is that the many studies now available reporting on the diagnostic and prognostic value of a persistent discrepancy between clinic and daytime or home blood pressure should not be interpreted as necessarily providing information on the prognostic value of the real white coat effect, and this issue needs to be assessed through different and more specific approaches, as demonstrated by Palatini et al. 13. Conclusions The possible prognostic value of cardiovascular hyper-reactivity to stressful stimuli has been investigated by a large number of studies (16), leading to discrepant conclusions. The same uncertainty appears to characterize the clinical value of blood pressure and heart rate reactivity to the physician's visit; however, this issue should not be simplistically disregarded, also suggested by the data provided by Palatini et al. 13 on its possible association with neurohormonal changes and with higher ambulatory blood pressure levels. Given the inadequacy of surrogate methods in offering a proper quantification of the white coat effect, as confirmed by Palatini et al. 13, the direct assessment of this phenomenon not only has the merit of preventing misunderstandings and misconceptions in this field 23, but also may help in the more precise identification of those subjects whose enhanced blood pressure reactivity to specific challenges might carry a higher risk of cardiovascular complications and require a closer follow-up 24–26.