Prognostic Role of Circadian Blood Pressure Pattern in Octogenarians. Is the Evidence Solid Enough for a Paradigm Shift?

In recent years, the importance of out-of-office blood pressure (BP) monitoring has been acknowledged by international guidelines as an important complementary tool to office BP measurement either to diagnose hypertension, to refine the assessment of cardiovascular (CV) risk associated with BP levels, and to evaluate the daily life coverage by antihypertensive treatment.1 Because of the robust association between adverse CV prognosis and the ability to yield a number of clinically useful BP parameters, such as average 24-hour BP, daytime, and particularly nighttime values, ambulatory BP monitoring (ABPM) is currently regarded as the reference methodology for out-of-office BP measurement.2 Among ABPM-derived indexes, the classification of the degree of nocturnal BP fall and its categorization into groups of dipping, extreme dipping, nondipping, or nocturnal BP rising (also known as reverse BP dipping), has generated much interest because of its independent association with CV risk. A blunted, absent or reverse BP dipping during nighttime was in fact shown to be associated to CV events and mortality above and beyond average 24-hour BP in many large-scale prospective studies and across different age ranges.3,4 In such a way, the CV risk of each single patient could be effectively reclassified based on dipping status.5 However, as compared with the strong and positive association of average ABPM-derived BP values, the magnitude of risk estimation related to dipping status appeared as of relatively minor importance.6,7 By definition, the day–night BP difference depends from daytime and nighttime BP so that, for any given value of daytime BP, a nocturnal BP rising pattern, as compared with normal BP dipping, assumes also 24-hour BP levels to be higher. For such reason, adjustment for 24 hours often significantly attenuates the relationship between nocturnal BP rising and CV outcomes. Therefore, there is an overall belief that, on a rating scale of importance for the assessment of CV risk, the evaluation of 24-hour average BP comes first and dipping status follows closely. Populations are getting progressively older worldwide and the number of individuals aged over 80 years is expected to triple by 2050.8 It is therefore of fundamental importance to evaluate whether the relevance of a number of CV risk predictors derived from large-scale prospective studies could be safely extrapolated also to very elderly subjects or, conversely, whether there are any differences in this regard. It has been reported that the prevalence of nocturnal BP rising pattern increases with aging.9 This is the expected consequence of the increased prevalence of diseases and clinical conditions related to both diurnal hypotension or nocturnal hypertension, such as autonomic dysfunction, in particular a reduced baroreflex sensitivity, type II diabetes mellitus, sleep apnea syndrome, or chronic kidney disease (Table 1).10 It is unknown, however, if the relative impact of average 24-hour BP and that of circadian BP fluctuations on CV risk assessments are comparable in very elderly individuals and in other age classes. One of the reasons of this knowledge gap is the limited amount of ABPM data available in subjects aged more than 80 years, often excluded from randomized controlled interventional trials. In two prospective observational studies conducted in the elderly, the Dublin Outcome Study11 and the Syst-Eur trial,12 average 24-hour BP seemed to confirm its prognostic relevance in predicting CV outcomes, although mean age of population enrolled in that studies was 73 and 70 years, respectively. Interestingly, in a recent study reporting results about age-specific risk ratios related to dipping status, the magnitude of hazard ratio increase associated with a nocturnal BP rising pattern was found to be nearly doubled in octogenarians as compared with other age groups.8 Conditions associated with reverse BP dipping profile in elderly people Abbreviation: BP, blood pressure. Conditions associated with reverse BP dipping profile in elderly people Abbreviation: BP, blood pressure. In the present issue of the journal,13 Fujiwara et al. report data on the prognostic relevance of the degree of nocturnal BP dipping in a cohort of 485 Japanese octogenarians (mean age 83 years). To be included, subjects should have demonstrated ability to attend the clinic or hospital without walking difficulties, capacity to live without assistance and a sufficient degree of cognition to be able to provide informed consent by themselves; furthermore, they should not have had any sign of CV or cerebrovascular disease in the previous 6 months, nor be in dialysis or have any malignant disease at baseline. Needless to say that these characteristics are not that common among community-dwelling octogenarians. The primary outcome was the occurrence of CV events and mortality during a period of about 4 years (n = 41); the secondary outcome was primary outcome + non-CV mortality (n = 72). Results showed that, as compared with normal dippers, subjects with a nocturnal BP rising pattern were at increased risk of CV events and composite outcome. Although limited by the sample size, the statistical significance unmet at the unadjusted analysis (hazard ratio 2.33 for CV events, 95% IC 0.94–5.79), became surprisingly so after adjustment for age, sex, and other clinical conditions such as previous CV disease, chronic kidney disease, and traditional CV risk factors. Adjustment for nighttime BP further strengthened the relationship between rising pattern and CV risk (hazard ratio 3.11 for CV events, 95% IC 1.11–8.88). When the authors explored the relation of other BP parameters measured by ABPM and adverse outcomes, they did not find any significant association for any of them. The hazard ratio associated with nocturnal BP rising status was in fact largely higher than that associated with an increase of 10 mm Hg systolic blood pressure or 5 mm Hg diastolic blood pressure, either for daytime, nighttime, or 24-hour BP. Overall, these results emphasize the importance of assessing circadian BP variations by ABPM in octogenarians as the most relevant ABPM-derived parameter whereas, quite unexpectedly, the evaluation of average 24-hour BP in this particular age group seems to lose its prognostic relevance. At a first glance, results seem to indicate that, in subjects aged 80 years or over, the prognostic significance of 24-hour BP is progressively overtaken by the importance of displaying an abnormal circadian BP pattern. However, as pointed out by the authors, some confounding factors might have influenced these results, suggesting the need of a cautious interpretation of these findings. In fact, quite unexpectedly, adjustment for classical CV risk factors strengthened the association between nocturnal BP rising and CV outcomes rather than “diluting” it. This trend was also observed in the multivariate models after adjustment for nighttime BP. Taken together these results suggest a possible reverse causality because of the presence of conditions carrying themselves an adverse outcome and characterized by lower daytime BP levels, which contributed to an increased night–day BP difference. In our view, many specific clinical conditions known to be associated with an adverse prognosis such as orthostatic hypotension, left ventricular systolic dysfunction, other systemic diseases such as anemia, hypovolemia, comorbidities, or other residual confounding factors commonly responsible also for lower standing and daytime BP levels, and thus potentially related to increased rates of reverse dipping, were not taken into account in the present study. In line with this, the cohort displaying a nocturnal BP rising pattern was significantly more frequently treated with α-blockers (19% vs. 7%, P = 0.01), a drug class that favors orthostatic hypotension (accompanied by lower daytime BP levels) and might directly influence the altered circadian pattern of these patients. Noteworthy, orthostatic hypotension, independently from its determinants, has a greater impact on CV and all-cause serious events and mortality in the elderly.14–16 Moreover, given that, in contrast to 24-hour average BP values, the definition of day–night BP intervals might be affected by lower reproducibility especially when assessed by self-reporting,17 it could be hypothesized that automated methods (e.g., actigraphy) to define the nocturnal period would be more suitable for future ABPM studies in elderly population where some degree of cognitive impairment may nevertheless be expected. In conclusion, while the results from the study by Fujiwara et al. can be considered as interesting hypothesis-generating data, emphasizing the need of a better understanding of the prognostic role of 24-hour average BP levels and of circadian BP behavior in very elderly people, they might not appear solid enough to lead to a paradigm shift in this context. Indeed, if confirmed, these findings could have a dramatic impact in determining specific choices for treatment thresholds and goals in this particular age group. It is therefore important to have this relevant issue explored again in future studies that should however appropriately address the concomitant role of comorbidities, ongoing treatment, frailty, orthostatic hypotension, sleep duration and quality, and other age-specific conditions which could significantly influence a circadian BP pattern and, at the same time, portend unfavorable health consequences in octogenarians. The authors declared no conflict of interest.