Effects of Antihypertensive Deprescribing Strategies on Blood Pressure, Adverse Events, and Orthostatic Symptoms in Older Adults: Results From TONE

Abstract

Background: The Trial of Nonpharmacologic Interventions in the Elderly (TONE) demonstrated the efficacy of weight loss and sodium reduction to reduce hypertension medication use in older adults. However, the longer-term effects of drug withdrawal (DW) on blood pressure (BP), adverse events, and orthostatic symptoms were not reported.

Methods: TONE enrolled adults, ages 60-80 years, receiving treatment with a single antihypertensive and systolic BP (SBP)/diastolic BP <145/<85 mm Hg. Participants were randomized to weight loss, sodium reduction, both, or neither (usual care) and followed up to 36 months; ~3 months postrandomization, the antihypertensive was withdrawn and only restored if needed for uncontrolled hypertension. BP and orthostatic symptoms (lightheadedness, feeling faint, imbalance) were assessed at randomization and throughout the study. Two physicians independently adjudicated adverse events, masked to intervention, classifying symptomatic (lightheadedness, dizziness, vertigo), or clinical events (fall, fracture, syncope).

Results: Among the 975 participants (mean age 66 years, 48% women, 24% black), mean (±SD) BP was 128 ± 9/71 ± 7 mm Hg. Independent of assignment, DW increased SBP by 4.59 mm Hg (95% confidence interval [CI]: 3.89, 5.28) compared with baseline. There were 113 adverse events (84 symptomatic, 29 clinical), primarily during DW. Compared with usual care, combined weight loss and sodium reduction mitigated the effects of DW on BP (β = -4.33 mm Hg; 95% CI: -6.48, -2.17) and reduced orthostatic symptoms long term (odds ratio = 0.62; 95% CI: 0.41, 0.92), without affecting adverse events (hazard ratio = 1.81; 95% CI: 0.90, 3.65). In contrast, sodium reduction alone increased risk of adverse events (hazard ratio = 1.75; 95% CI: 1.04, 2.95), mainly during DW.

Conclusions: In older adults, antihypertensive DW may increase risk of symptomatic adverse events, highlighting the need for caution in withdrawing their antihypertensive medications.

Clinical trials registration: Trial Number NCT00000535.

Keywords: blood pressure; deprescription; falls; hypertension; sodium reduction; weight loss.

© American Journal of Hypertension, Ltd 2021. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Figures

Graphical abstract

Figure 1.

Mean mm Hg (95% CI) for (a) systolic blood pressure (SBP) and (b) diastolic blood pressure (DBP) according to randomized intervention throughout the TONE trial: usual care (UC, circle), weight loss (WL, diamond), sodium reduction (SR, triangle), and both weight loss and sodium reduction (Comb, square). Point estimates are shown by study visit: RV (randomization visit), P (visits prior to drug withdrawal), D (drug withdrawal visits, 6-month visits, and 9-month visits), 12–18 (12-, 15-, and 18-month visits), and 21–36 (21-, 24-, 27-, 30-, 33-, and 36-month visits). Point estimates are positioned according to the median follow-time corresponding to each cluster of visits. Number of visits contributing to each point estimate is found in Supplement Table ST3. Note that participants were censored if they experienced the primary outcome (which included high blood pressure). All means are adjusted for obesity status and clinic center. Abbreviations: CI, confidence interval; TONE, Trial of Nonpharmacologic Interventions in the Elderly.

Figure 2.

Cumulative incidence curves portraying risk of (a) any adverse events, (b) orthostatic symptoms, (c) falls, fracture, or syncope, or (d) falls with the number at risk according to the 4 randomized assignments over the 36-month trial surveillance period: usual care (UC), weight loss (WL), sodium reduction (SR), or combined weight loss and sodium reduction (Combined). Participants were censored if they reported an event or administratively (as a result of developing the primary outcome in the study or at 36 months). N is the number of recurrent events reported during each study phase. Pre-drug withdrawal (DW) is the study period after randomization and up to and including the visit initiating DW. DW includes all DW visits after drug withdrawal began up to and including the 12-month visit. In the event that study phases were not clearly labeled (6 orthostatic symptoms or adverse events, 2 orthostatic symptoms, 4 adverse events, 3 falls), we relied on the mean follow-time for first DW (3.5 months) and last DW (5.1 months).