[Skip to Navigation]
Views 120
Citations 0
Editor's Correspondence
Aug 13/27, 2012

Do Angiotensin Receptor Blockers Really Hold Promise for the Improvement of Cognitive Functioning?—Reply

Author Affiliations

Author Affiliations: Division of Geriatric, Hospital & General Internal Medicine, Department of Medicine (Dr Hajjar) and Department of Biostatistics (Dr Mack), University of Southern California, Los Angeles; and Institute for Aging Research, Hebrew SeniorLife, Harvard School of Medicine, and Division of Gerontology, Beth Israel Deaconess Medical Center, Boston, Massachusetts (Dr Lipsitz).

Arch Intern Med. 2012;172(15):1191-1192. doi:10.1001/archinternmed.2012.3885

In reply

We appreciate the interest by Drs Andrade and Fernandes in our recent report.1 As noted, this is a small pilot study intended to provide preliminary data in a randomized setting. Nonetheless, correcting for multiple testing (8 cognitive tests) using the False Discovery Rate method,2 a P value less than .01 would be significant. Therefore, our findings for Trails B and Trails B-A remain significant with adjustment for multiple comparisons. After 1 year of treatment, the lisinopril group showed a worsening of 14.44 seconds, while the candesartan and hydrochlorothiazide groups showed improvements of 17.11 seconds and 4.24 seconds, respectively. These correspond to improvements in the candesartan group of 31.55 seconds relative to the lisinopril group and 12.87 seconds relative to the hydrochlorothiazide group (effect sizes [difference in group means divided by pooled standard deviation] of 1.05 for the candesartan vs lisinopril groups and 0.43 for the candesartan vs hydrochlorothiazide groups). These are medium to large effects.3 We did not include a cross-over or a discontinuation phase because these approaches were not included in the study protocol and were beyond the scope of this study. Finally, as shown in Figure 1 in our report,1 the effects of angiotensin receptor blockers on Trails B and Trails B-A were progressive at the 6- and 12-month assessments. These data suggest an increasing effect over the 1-year period. However, we cannot rule out a ceiling or nonenduring effect beyond this time frame.