Comment on “Efficacy and safety of antihypertensive drugs deprescribing in older adults: A systematic review and meta-analysis of randomized controlled trials” Authors: Sambit Kumar Parida, Mohit Kotnala, Suraj Mandal Published: 2026 Dear Editor, We read with great interest the meta-analysis by Alsubaiei et al. [1], examining the clinical outcomes of antihypertensive deprescribing in older adults. What four trials and 2173 patients can and cannot establish The pooled point estimates showed no significant increase in all-cause mortality, cardiovascular mortality, MACE, or falls. The authors acknowledge that the evidence is limited, but the practical implications of this limitation deserve sharper articulation. For most secondary outcomes, including stroke, MI, and falls, the GRADE certainty was very low, driven by wide confidence intervals and low event counts. The confidence interval for cardiovascular mortality (RR 0.80–1.55) was compatible with a 55% increase in risk. For stroke, it spanned from a 34% reduction to an 89% increase. These intervals do not rule out clinically meaningful harm; they simply lack the precision to detect it [2]. Describing these findings as “reassuring” risks overstating what the data support at this sample size suggests. Clinical heterogeneity is the central interpretive problem The four included trials differ in ways that matter clinically: nursing home residents with moderate-to-severe dementia in DANTON, community-dwelling older adults with mild cognitive impairment in DANTE, primary care patients aged over 80 years in OPTIMISE, and frail nursing home residents in RETREAT-FRAIL. These populations have fundamentally different frailty burdens, cardiovascular risk profiles, and physiological reserves. Pooling them into a single risk ratio treats deprescribing as a uniform intervention in a uniform population, which it is not. Statistical heterogeneity was low across most outcomes, but a low I2 in a small set of trials with heterogeneous populations can reflect insufficient power to detect heterogeneity rather than genuine comparability [3]. The clinical implication is direct: a deprescribing strategy that is safe in community-dwelling older adults with mild cognitive impairment may carry a different risk-benefit balance in nursing home residents with severe dementia [4], which is exactly what DANTON's early termination for safety concerns suggested. The DANTON signal should not be absorbed into the pooled estimate without discussion DANTON was stopped by its data safety monitoring board because serious adverse events occurred in 36% of the deprescribing group versus 24% in the usual care group. The pooled SAE risk ratio of 1.08 (0.90–1.30) was numerically dominated by larger trials and effectively diluted this signal. The authors note this but do not fully confront what it means: the subgroup in which deprescribing was most aggressively pursued nursing home residents with moderate-to-severe dementia is the subgroup in which it may be most harmful. A meta-analysis that pools this result with trials enrolling far less vulnerable populations and concludes that deprescribing carries no increased SAE risk is methodologically accurate but clinically misleading for patients most likely to be considered for deprescribing in practice. Deprescribing intensity and the step-down versus cessation distinction The review appropriately notes that OPTIMISE removed one drug, while DANTON and DANTE targeted complete discontinuation. These are pharmacologically distinct strategies with different hemodynamic consequences [5]. A clinician reading that deprescribing is safe may reasonably infer that complete discontinuation is an option when the available evidence is weighted toward single-drug reduction in relatively robust older adults. The distinction between step-down and complete cessation should be foregrounded in the conclusions rather than mentioned in the limitations section. Performance bias and its direction All four trials were open-label studies. The authors flagged this as a concern for performance bias but did not discuss which direction the bias likely ran. Open-label deprescribing studies risk differential co-intervention, clinicians aware that a patient is having antihypertensives withdrawn may increase monitoring intensity, lifestyle counselling, or non-pharmacological support, which would tend to attenuate any adverse outcome signal in the intervention arm. If performance bias operates in this direction, the true safety profile of unsupervised deprescribing in routine clinical practice may be less favorable than these trial results suggest. Alsubaiei et al. provided the most comprehensive RCT synthesis available for this question. The clinical message that supervised, individualized step-down deprescribing appears feasible in carefully selected older adults is supported. The message that deprescribing is broadly safe across the spectrum of frail older patients is not true. Ethical approval Not applicable. Generative AI disclosure The authors declare that no generative artificial intelligence (AI) tools were used in the preparation, writing, analysis, or editing of this manuscript. Funding No funding was received for this work. CRediT authorship contribution statement Sambit Kumar Parida: Conceptualization, Data curation. Mohit Kotnala: Conceptualization, Writing – original draft, Writing – review & editing. Suraj Mandal: Conceptualization, Data curation, Writing – original draft, Writing – review & editing. Declaration of competing interest The authors declare no conflicts of interest. Acknowledgment None. Contributor Information Sambit Kumar Parida, Email: dr.sambit@yahoo.com. Mohit Kotnala, Email: mohitkotnala1988@gmail.com. Suraj Mandal, Email: dr.surajpharma@gmail.com. Data availability No datasets were generated or analyzed during the current study. References Associated Data Data Availability Statement No datasets were generated or analyzed during the current study.