This report presents an exploratory secondary analysis of the double-blinded VALUE outcome trial including 15,313 participants. The investigators divided the study population into an exploratory cohort (two thirds of participants) and a validation cohort (one third) to develop and test predictive models. The analysis focused on whether combinations of established patient-level predictors could identify groups with different probabilities of achieving blood pressure (BP) control and different cardiovascular outcomes.
Four previously recognized predictors of antihypertensive response were selected and combined to form composite patient profiles: age, severity of hypertension, presence of comorbidity, and previous treatment status. The authors reasoned that these factors, when considered together, might offer more discriminative power than when evaluated singly.
The analysis used logistic regression to evaluate differences in BP control between profiles and Cox proportional hazards models to assess associations between profiles and cardiovascular event rates. The exploratory cohort was used to construct and refine composite profiles; the validation cohort tested reproducibility of findings. The key statistical comparisons included testing whether profiling added predictive information beyond individual predictors and whether profile-specific event rates differed.
BP control rates varied significantly between the composite profiles, independent of randomized treatment assignment (likelihood ratio test P<0.001). Profiles developed in the exploratory cohort reproduced in the validation cohort, indicating internal consistency within the VALUE dataset. The composite profiles predicted BP control significantly better than any single factor alone (P<0.001 for all comparisons reported by the authors).
The composite profiles were significantly associated with incidence of cardiovascular events according to the Cox model analyses. The authors report that adverse events were not related to patient profiles, suggesting the profiles were informative for efficacy (BP control and event incidence) but not for treatment tolerability as measured by adverse events in this dataset.
Analyses showed that combining the four predictors into composite profiles offered superior prediction of BP control compared with each individual predictor on its own. The authors used likelihood-ratio testing to demonstrate the incremental predictive value of profiling and reported statistically significant improvements (P<0.001) for all comparisons versus single-factor models.
The authors conclude that composite patient profiles—integrating age, hypertension severity, comorbidity, and prior treatment status—may more accurately predict BP responses to antihypertensive therapy than single factors. They propose that this approach has the potential to support better individualization of antihypertensive treatment, although the analysis is exploratory and based on secondary reanalysis of trial data.
This manuscript is a preprint and has not undergone peer review; the authors explicitly state that the findings should not be used to guide clinical practice at this stage. The VALUE trial was conducted before routine trial registry use; therefore no trial registry identifier is available. The authors note that data will be made available upon reasonable request after publication. As an exploratory reanalysis, the findings require external validation and peer review before clinical implementation.
Several authors disclosed relationships with pharmaceutical and device companies. Specific competing interests are listed by author in the source: financial interactions and honoraria from multiple industry entities were reported for some contributors; other authors reported no conflicts. The authors confirm ethical oversight and participant consent procedures were followed as per the original VALUE investigators' documentation.
The VALUE dataset (n=15,313) was reanalysed to test whether combining established predictors into composite profiles improves prediction of BP control and cardiovascular outcomes.
Composite profiles incorporating age, hypertension severity, comorbidity, and prior treatment status performed better than any single predictor in this exploratory analysis.
Profiles were associated with cardiovascular event incidence but not with adverse event rates in the dataset.
As this is a preprint and an exploratory secondary analysis, the results should be interpreted cautiously and require peer review and external validation before influencing treatment decisions.
Acknowledgement: All factual statements in this summary derive from the preprint report of this exploratory analysis of the VALUE trial. No new data or recommendations beyond those reported by the authors are provided here.