Psychological behaviour, pulse pressure and left ventricular hypertrophy

Target organ damage is a major issue in hypertension research and treatment. Various phenotypes of target organ damage are increasingly used as intermediate criteria for predicting cardiovascular events in patients with borderline or sustained hypertension. Among them, left ventricular hypertrophy (LVH) [1] and arterial stiffness [2] have been shown to be independent predictors of cardiovascular events, including death and coronary events. Recently, the LIFE study [3], which showed that the regression of LVH under treatment predicted the reduction of cardiovascular events, conferred a status of ‘surrogate end point’ to LVH [4]. Both LVH and arterial stiffness are related to pulse pressure: the stiffening of large arteries, by modifying the timing of the reflected waves, raises systolic blood pressure, favouring the development of LVH [5,6], and lowers diastolic blood pressure, impairing coronary perfusion during diastole. These mechanisms are generally accepted to explain the predictive value of pulse pressure for cardiovascular events [7]. Taken together, these data suggest that it would be possible to early detect patients prone to develop LVH (i.e. at high cardiovascular risk) in order to best prevent cardiovascular complications. In this issue of the journal, Jokiniitty et al. [8] address this issue. These authors investigated, in 65 untreated men, aged 35–45 years, the role of casual blood pressure measurements and blood pressure responses to psychological tasks in the prediction of future left ventricular mass index (LVMI). Forty-nine of these subjects remained unmedicated for hypertension at 10 years of follow-up. The predictive value of baseline LVMI on future LVMI significantly improved when the model incorporated blood pressure responses to psychological tasks at baseline. Particularly, a higher pulse pressure at the end of an habituation task or a relaxation task was associated with greater future LVMI. By contrast, casual blood pressure measurements did not improve the predictive value of the model. The authors conclude that ‘the PP, which reflects the properties of the arterial wall, is the most significant BP variable in predicting future LVMI'. The article by Jokiniitty et al. [8] has several strengths. First, this is the longest prospective study to show that blood pressure responses to psychological tasks improve the prediction of LVMI compared to casual measurements. Second, the authors used a large battery of different standardized behavioural tasks in the same study. Third, a methodological ‘gold standard’ was used, with blood pressure recorded intra-arterially via the Oxford technique. Finally, it was possible to separately analyse subjects who remained untreated because, during the follow-up, antihypertensive medication was started in only a small percentage (22%) of subjects who had LVMI measurements at baseline. This study, although conclusive, exhibits several limits. First, the blood pressure value which is analysed here is not truly the ‘response’ to the test (i.e. the blood pressure changes) but rather the mean intra-arterial pulse pressure of the final minute of each task. When the changes in pulse pressure are calculated as ‘after’ minus ‘before’ the task, and tested as significant covariate, they did not correlate significantly with future LVMI, and did not improve the predictive value of baseline LVMI on future LVMI. Similarly, the changes in casual pulse pressure (baseline – 10 years) did not improve the predictive value of the model. Second, pulse pressure may not only reflect the properties of the arterial wall, but also the intervention of other mechanisms. Psychological tasks increase heart rate and left ventricular ejection time (LVET), which are additional determinants of pulse pressure, particularly in a middle-aged population such as the present one. To rule out a significant role of cardiac ejection, it would have been useful to determine whether the predictive value of pulse pressure during the psychological task remained predictive of future LVMI after adjustment on heart rate and LVET. The calculation of left ventricular ejection time requires a continuous recording of electrocardiogram and carotid pressure waveform, which may not have been available at study baseline. In addition, the level of pulse pressure at the end of each test may be a marker of various neurohumoral systems (including the sympathetic nervous system) activated during the psychological task, and exerting their growth-promoting effect on the cardiac myocytes independently of blood pressure, during recurrent stress exposures. Third, this group previously reported, in the same population, that ambulatory pulse pressure was the most significant predictor of LVMI [9]. Thus, it would have been interesting to investigate whether the predictive value of pulse pressure during psychological tasks on future LVMI remained significant after adjustment to ambulatory pulse pressure (i.e. whether an ambulatory recording would be representative enough of pulse pressure changes during standardized stress tests or, on the contrary, whether it is necessary to perform blood pressure measurement during psychological tasks). In conclusion, the findings of the present study by Jokiniitty et al. [8] may have important clinical implications for the assessment of cardiovascular risk and the therapeutic approach of hypertension, through early detection of high-risk patients. Confirmation of these results by additional studies measuring various indices of large artery elastic properties and cardiac ejection would also strengthen mechanistic hypotheses concerning the role of arterial stiffness on the development of LVH.