Elderly hypertensive patients, particularly those aged 75 years or above, should be carefully treated because they are at higher risk for both cardiovascular and adverse drug events than younger patients. A number of trials showed the efficacy of lowering systolic blood pressure (SBP) to some extent in the elderly with SBP >160 mm Hg irrespective of drug classes. Recently, the Hypertension in the Very Elderly Trial1 demonstrated the benefits of antihypertensive treatment even in patients over the age of 80 years with SBP ⩾160 mm Hg. In this study, the target SBP was <150 mm Hg with the achieved SBP of 144 mm Hg after the mean follow-up period of 2 years. These results rationalize to consider that SBP should be maintained below 150 mm Hg in elderly patients including those over 75 years old.

It is still controversial whether SBP should be lowered below 140 mm Hg in elderly patients, although epidemiological studies and the meta-analysis of 147 randomized trials2 suggest a proportional reduction in cardiovascular events according to BP level. In fact, no previous trials have achieved SBP <140 mm Hg. Conversely, excessive BP lowering in the elderly may cause adverse reactions, such as light-headedness and falls, and has been associated with the J-curve phenomenon.3

The Japanese trial to assess optimal systolic blood pressure in elderly hypertensive patients (JATOS) was the first study that was specially designed to compare the strict (<140 mm Hg) with the mild (140–159 mm Hg) target of SBP for 2 years in the elderly aged 65–85 years. Principal results of JATOS by intention-to-treat analysis revealed that the outcomes were similar between the strict and mild treatment groups.4 However, a large amount of subjects failed to achieve the target SBP, resulting in a weak statistical power of JATOS.

In this issue of Hypertension Research, Rakugi et al.5 reported a per-protocol analysis of JATOS to evaluate the outcomes among the target SBP-achieved subjects. In JATOS, 54% (1192 of 2212 subjects) in the strict treatment group and 69% (1531 of 2206 subjects) in the mild treatment group achieved their target SBPs by use of efonidipine, a long-acting calcium antagonist, as the first-line drug. Although average SBP and DBP were different by 14.3 and 4.3 mm Hg, respectively, the incidence of the primary end points, a composite of cardiovascular disease and renal failure, was similar between the two groups. There was no difference in each of end point components or the incidence of adverse events between the strict target-achieved group and the mild target-achieved group.

These results are consistent with the principal intention-to-treat analysis of JATOS4 and with the recently published Valsartan in Elderly Isolated Systolic Hypertension (VALISH) Study6 as well. VALISH study compared the strict (<140 mm Hg) with the moderate (140–149 mm Hg) target of SBP for ⩾2 years in 3260 hypertensive patients aged 70–84 years on valsartan-based treatment. Both intention-to-treat and per-protocol analyses showed that a composite of end points and adverse events were similar between the two groups.6 By contrast, an Italian study7 demonstrated that the aggressive target of SBP <130 mm Hg (achieved SBP of 131.9 mm Hg) was superior to the less aggressive target of SBP <140 mm Hg (achieved SBP of 135.6 mm Hg) in non-diabetic hypertensive patients. Reduced end points of this study, however, were left ventricular hypertrophy, coronary revascularization and new-onset atrial fibrillation, most of which were not included in JATOS and VALISH study. In addition, the subjects were younger (⩾55 years of age, mean age of 67 years) than those of JATOS and VALISH, and the event rate was remarkably higher than those of the two Japanese studies. The summary of the three studies is shown in Table 1. These points along with ethnicity may explain the difference in the main results.

Table 1 Summary of JATOS, VALISH and Cardio-Sis that were specifically designed to compare the target systolic blood pressure (SBP)
Full size table

Finally, what should we do in clinical practice? Although JATOS targeted SBP <140 vs. 140–159 mm Hg, it may be commonly accepted that SBP should be kept <150 mm Hg in the elderly as HYVET showed.1 This view can be strengthened by the finding of JATOS that target-unachieved patients had worse prognosis than target-achieved patients, despite the study groups.5 Then, should we reduce SBP below 140 mm Hg or maintain SBP between 140 and 150 mm Hg in elderly patients? At present, no clinical trial has confirmed the benefits of lowering SBP below 140 mm Hg in the elderly. Obviously, however, cardiovascular disease risk is higher in elderly patients than younger ones. Accordingly, one might expect the benefits of reducing SBP <140 mm Hg or lower, which have been shown in younger populations such as Cardio-Sis.7 Statistical power might have been insufficient in JATOS and VALISH to detect a small difference between the groups, if present. Furthermore, targeting SBP <140 mm Hg was not associated with the increase in adverse events in JATOS and VALISH. Taken together, strict control of SBP <140 mm Hg may be of little clinical importance for the prevention of cardiovascular and renal events in the elderly. This may not be applicable to patients with cardiovascular disease or non-Asian populations. Conversely, it may not be necessary to withdraw antihypertensive therapy once SBP is safely maintained below 140 mm Hg. Pending future trials and meta-analyses determining the optimal SBP level for elderly patients, we should follow the JSH 2009 guidelines8 that are compatible with the above-mentioned points.