Effects of Candesartan vs Lisinopril on Neurocognitive Function in Older Adults With Executive Mild Cognitive Impairment: A Randomized Clinical Trial

Abstract

Importance: Observational studies have suggested that angiotensin receptor blockers are associated with a unique cognitive protection. It is unclear if this is due to reduced blood pressure (BP) or angiotensin receptors type 1 blockade.

Objective: To determine neurocognitive effects of candesartan vs lisinopril in older adults with mild cognitive impairment (MCI).

Design, setting, and participants: This randomized clinical trial included participants aged 55 years or older with MCI and hypertension. Individuals were withdrawn from prior antihypertensive therapy and randomized in a 1 to 1 ratio to candesartan or lisinopril from June 2014 to December 2018. Participants underwent cognitive assessments at baseline and at 6 and 12 months. Brain magnetic resonance images were obtained at baseline and 12 months. This intent-to-treat study was double-blind and powered for a sample size accounting for 20% dropout. Data were analyzed from May to October 2019.

Interventions: Escalating doses of oral candesartan (up to 32 mg) or lisinopril (up to 40 mg) once daily. Open-label antihypertensive drug treatments were added as needed to achieve BP less than 140/90 mm Hg.

Main outcomes and measures: The primary outcome was executive function (measured using the Trail Making Test, Executive Abilities: Measures and Instruments for Neurobehavioral Evaluation and Research tool) and secondary outcomes were episodic memory (measured using the Hopkins Verbal Learning Test-Revised) and microvascular brain injury reflected by magnetic resonance images of white matter lesions.

Results: Among 176 randomized participants (mean [SD] age, 66.0 [7.8] years; 101 [57.4%] women; 113 [64.2%] African American), 87 were assigned to candesartan and 89 were assigned to lisinopril. Among these, 141 participants completed the trial, including 77 in the candesartan group and 64 in the lisinopril group. Although the lisinopril vs candesartan groups achieved similar BP (12-month mean [SD] systolic BP: 130 [17] mm Hg vs 134 [20] mm Hg; P = .20; 12-month mean [SD] diastolic BP: 77 [10] mm Hg vs 78 [11] mm Hg; P = .52), candesartan was superior to lisinopril on the primary outcome of executive function measured by Trail Making Test Part B (effect size [ES] = -12.8 [95% CI, -22.5 to -3.1]) but not Executive Abilities: Measures and Instruments for Neurobehavioral Evaluation and Research score (ES = -0.03 [95% CI, -0.08 to 0.03]). Candesartan was also superior to lisinopril on the secondary outcome of Hopkins Verbal Learning Test-Revised delayed recall (ES = 0.4 [95% CI, 0.02 to 0.8]) and retention (ES = 5.1 [95% CI, 0.7 to 9.5]).

Conclusions and relevance: These findings suggest that in older adults with MCI, 1-year treatment with candesartan had superior neurocognitive outcomes compared with lisinopril. These effects are likely independent of the BP-lowering effect of candesartan.

Trial registration: ClinicalTrials.gov Identifier: NCT01984164.

Figure 1.. Flow Diagram of Study Recruitment

AE indicates adverse event; ITT, intention-to-treat; and LTF, loss to follow-up. ^aPersonal reasons for withdrawing from candesartan group included: 1 owing to depression, 2 owing to insufficient study compensation to continue, and 1 owing to high home blood pressure readings. ^bPersonal reasons for withdrawing from lisinopril group included 1 owing to relocation, 1 owing to busy life schedule, and 1 owing to insufficient study compensation. ^cTwo participants discontinued study intervention at 7 months owing to cough and owing to hospitalization for lung tumor resection. The latter was receiving lisinopril as part of their clinical treatment at 12 months. Both participants completed 12-month follow-up visit.

Figure 2.. Mean Sitting and Standing Systolic and Diastolic Blood Pressure (BP) Over Study Period, by Treatment Group

P values are obtained from mixed model repeated measure for the interaction of treatment effect over time. Whiskers indicate SDs.

Figure 3.. Change in Trail Making Test Parts B and B − A, and Hopkins Verbal Learning Test-Revised, Delayed Recall and Retention During 12 Months Stratified by Treatment Group

Values are model-derived least-square means and SEs (whiskers) from mixed model repeated measure and are adjusted for systolic blood pressure and stratification variables. Trail Making Test analyses are also adjusted for baseline Trail Making Test scores. P values for the treatment effect are obtained from the mixed model repeated measure models comparing change over the study period (visit as a continuous measure: 1 = baseline, 2 = 6 months, and 3 = 12 months) between the 2 groups: A, P = .004; B, P = .01; C, P = .04; D, P = .02.

Figure 4.. Change in White Matter Lesion Volumes From Baseline to 12 Months Stratified by Treatment Group

Adjusted values are least-square means and SEs (whiskers) obtained from mixed model repeated measure and are adjusted for systolic blood pressure, participant race/ethnicity, and prestudy antihypertensive medications.