Ambulatory blood pressure monitoring and dipping status in predicting left ventricular hypertrophy

Ambulatory blood pressure (BP) monitoring is guideline recommended for the exclusion of white coat hypertension, clarification of hypertension burden and response to antihypertensive therapy [1]. Since its inception, however, the utility of several parameters derived from the 24-h recording has been debated, including whether nocturnal average BP carries greater prognostic significance than daytime BP and whether loss of normal diurnal variation (nondipping pattern) adds incremental prognostic value. Clarification of these issues has been clouded, as others in the field have commented, by a lack of adjustment for confounders, use of surrogate endpoints and retrospective study design [2]. Most importantly, the definition of nondipping status using a less than 10% cutoff, although well established in the literature, is arbitrary [2]. Although it is clear that the principal metric derived from the 24-h study is the cumulative average BP, the extent to which dipping status relates to other markers of cardiovascular risk and outcomes has not been unequivocally established. Large prospective observational studies and individual level meta-analyses powered for clinical endpoints have demonstrated that the most informative measure obtained from a 24-h BP study is the weighted 24-h average BP, following in prognostic utility is the nocturnal average and finally the daytime average BP [2–4]. Importantly, dipping status, quantified by the night : day BP ratio and by the traditional less than 10% decrease in nocturnal BP cutoff, demonstrated prognostic utility in determining total mortality and cardiovascular events [2]. Importantly in this analysis, dipping status was adjusted for the 24-h BP. This is highly pertinent to the hypothesis that dipping status per se is an important determinant of outcome and not the confounding influence of mean 24-h BP (which is often higher in nondippers). Despite being statistically significant, the night : day ratio only explained 0.1% of the estimated variance in this meta-analysis [2]. Therefore, given this finding in a cohort of 9641 patients, it is not entirely surprising that dipping status failed to predict LVH in the analysis reported by Yi et al.[5] in a substantially smaller cohort of patients. The principal finding of the Korean-Ambulatory Blood Pressure study [5], that nocturnal BP is a far more powerful determinant of LVH is, however, consistent with the previously published data [3,6]. Previous studies have demonstrated that dipping status is an important predictor of LVH [7–11] – why then was this effect not evident in the present analysis? The critical factor may relate to the confounding effect of antihypertensive drug therapy. In the five studies reported previously [7–11], patients were not on antihypertensive drug therapy, whereas in the study reported by Yi et al.[5] in this edition of the journal, 39% of the cohort was receiving antihypertensive therapy at the time of ambulatory BP measurement. Antihypertensive therapy may lower daytime BP to a greater extent than nocturnal BP, thereby artefactually inducing ‘nondipping’ in a proportion of patients. It is instructive that nondipping status as defined by a less than 10 mmHg drop in BP from daytime to night-time is typically present in 50–60% of studies including treated patients but only 25–30% of studies with untreated patients [3,4,8,9,12,13]. It may be expected that this confounding effect could dilute the power of any study to detect the true relationship between dipping status (arising de novo) and left ventricular hypertrophy (LVH). Indeed, in a study performed in treated essential hypertensive patients, no association between dipping status and LVH was demonstrated [12]. Finally, in an analysis performed by the International Database of Ambulatory blood pressure in relation to Cardiovascular Outcome investigators, the night-to-day SBP ratio was only a weak predictor of fatal and nonfatal cardiovascular events and daytime BP was not predictive of the cardiovascular events in those patients receiving antihypertensive therapy [3]. An additional explanatory factor for the discordant results presented by Yi and colleagues may relate to the particularly strong association between cardiovascular outcomes and nondipping status in women [14]. In the present analysis, women only constituted 60% of the population with LVH, and a stratified analysis between the sexes was not performed [5]. As a unifying theme, however, all studies regardless of the treatment emphasize the importance of nocturnal BP as an important predictor of LVH [7]. Taken together, several pertinent conclusions may be made: firstly that nondipper status may lose its prognostic utility for cardiovascular end-organ damage in treated hypertensive patients; secondly, that the 24-h weighted average BP may represent the single best predictor of both end-organ damage and cardiovascular mortality and finally, that the nocturnal BP is typically superior to daytime BP in the prediction of cardiovascular risk. With these conclusions in mind, several questions regarding ambulatory BP monitoring remain: firstly, the role of ‘chronotherapy’ in patients with de novo nondipping to restore the normal diurnal BP variation and secondly whether drugs with a longer duration of antihypertensive effect as demonstrated on 24-h BP monitoring are superior to shorter acting agents (with identical absolute BP lowing efficacy) regarding the outcomes and presence of end-organ damage. With these important issues in mind, efforts to conduct adequately controlled, prospective, randomized studies, rather than the more convenient, retrospective, case–control design, will be critical. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.