Blood Pressure Numbers, Not Diagnosis Labels, Predict Death Risk Across Two Nations

High blood pressure kills quietly, and public health campaigns have long assumed that the first step to saving lives is simply telling people they have the condition. A new study challenges the comfort of that assumption. Analyzing tens of thousands of adults in China and the United States, researchers found that what actually predicted who died during follow-up was not whether a person knew they had hypertension, but how severe their blood pressure was and whether it was under control. The findings, published in BMC Public Health, suggest that measured blood pressure severity carries more consistent prognostic weight than self-reported awareness of a diagnosis.

The research team, led by Yan Gao, Yingjie Peng, and Yuanyuan Wang of Weifang Maternal and Child Health Hospital and collaborating institutions in Shandong, China, set out to answer a deceptively simple question: after accounting for how high someone’s blood pressure actually is, does knowing about a hypertension diagnosis matter for survival? The question matters because awareness is often treated as a key performance indicator in hypertension care. National surveys track the percentage of hypertensive adults who know their status, and screening programs celebrate rising awareness figures. But if awareness itself does not translate into lower mortality, the metric may be a weaker proxy for health benefit than commonly believed.

To investigate, the researchers turned to two of the most respected population-based cohort studies in the world. The first was the China Health and Retirement Longitudinal Study, known as CHARLS, which enrolled adults aged 45 and older with a baseline survey conducted in 2011 and 2012. The second was the US National Health and Nutrition Examination Survey, or NHANES, drawing on cycles conducted between 1999 and 2018. Both studies combine household interviews with physical examinations, meaning participants had their blood pressure actually measured by trained staff rather than merely self-reported. That combination of subjective awareness data and objective measurements is what made the analysis possible.

The analytic samples were substantial: 12,868 participants in CHARLS, of whom 12,640 contributed mortality follow-up, and 26,351 participants in NHANES. Hypertension awareness was defined strictly as a self-reported previous diagnosis of hypertension made by a physician or other health professional. Participants were then sorted into five mutually exclusive categories based on their awareness status and their baseline measured blood pressure: normotension, aware controlled hypertension, aware uncontrolled hypertension, unaware grade 1 hypertension, and unaware grade 2 to 3 hypertension. This classification scheme allowed the team to compare, for example, someone with severe but undiagnosed hypertension against someone with equally severe diagnosed hypertension, isolating the effect of awareness itself.

The two cohorts were analyzed independently using a broadly common analytical framework, a deliberate design choice that guards against pooling artifacts and allows each country’s data to speak with its own weighting and covariate adjustments. The researchers employed cohort-specific Cox proportional hazards models, the standard statistical tool for survival analysis, incorporating appropriate survey weights and adjusting for demographic, lifestyle, and clinical covariates such as age, sex, body mass index, smoking status, and relevant clinical markers. They also compared aware and unaware hypertension within the same blood pressure severity categories and modeled continuous systolic blood pressure in relation to mortality, providing both categorical and continuous views of the same underlying biology.

The headline results were striking in their consistency across two very different health systems. Compared with normotension, aware but uncontrolled hypertension was associated with higher all-cause mortality in both cohorts. In CHARLS, the hazard ratio was 1.53, with a 95 percent confidence interval of 1.26 to 1.85, meaning a roughly 53 percent higher risk of death during follow-up. In NHANES, the corresponding hazard ratio was 1.39, with a confidence interval of 1.28 to 1.52. Unaware severe hypertension, grade 2 to 3, showed similarly elevated risks: a hazard ratio of 1.58 in CHARLS and 1.50 in NHANES. In other words, a person walking around with dangerously high blood pressure faced substantially elevated mortality whether or not a doctor had ever told them they had the condition.

The more provocative finding emerged when the researchers compared aware and unaware participants within the same broad blood pressure severity categories. If awareness itself conferred a survival advantage, aware individuals should have fared better than unaware individuals with equivalent blood pressure. They did not, at least not consistently. Awareness status was not reliably associated with mortality across the two cohorts, and the pattern of results differed between China and the United States. When the team modeled systolic blood pressure as a continuous variable, mortality rose steadily with increasing systolic pressure in both countries, but evidence of an interaction between systolic pressure and awareness appeared in NHANES and not in CHARLS. Notably, the main phenotype associations did not differ significantly between the cohorts, underscoring that the core relationship between blood pressure severity and death was robust on both sides of the Pacific.

What might explain the inconsistency of awareness as a predictor? The authors are careful to frame their conclusions as observational, describing associations among baseline-defined phenotypes rather than causal effects of knowing one’s diagnosis. Awareness, after all, is only meaningful if it leads to treatment, and treatment only matters if it achieves control. A person diagnosed with hypertension who never fills a prescription, or whose medication fails to bring readings down, is physiologically little different from someone who was never diagnosed at all. The data appear to reflect that reality: the aware uncontrolled group fared about as badly as the unaware severe group, while the measured blood pressure categories told a coherent story in both nations. Awareness without control is, from the body’s perspective, a label rather than a protection.

The authors explicitly warn against a misreading of their work. These findings should not be interpreted as evidence that hypertension awareness itself is unimportant, they conclude. Detection remains the essential gateway to treatment, and untreated severe hypertension carried some of the highest mortality risks in the study. The correct lesson is more nuanced: screening programs that stop at diagnosis, without ensuring continued clinical follow-up and sustained blood pressure control, may be capturing the credit for awareness statistics while leaving the underlying lethal physiology untouched. The researchers emphasize that hypertension detection remains important and should be followed by continued clinical follow-up and sustained blood pressure control.

For policymakers, the study arrives at a moment when hypertension burdens are climbing in both aging China and the United States, and when digital screening initiatives promise to push awareness figures ever higher. The message embedded in these two cohorts is that the numbers on the cuff, not the words on a chart, are what the body registers. Public health systems that measure success by diagnosis rates alone may be optimizing the wrong variable. The true target, this research suggests, is sustained control, verified by measurement, for everyone whose blood pressure sits in the danger zone, whether or not they have ever heard a doctor speak the word hypertension.

Subject of Research: Association of hypertension awareness and blood pressure severity with all-cause mortality in Chinese and US population cohorts

Article Title: Hypertension awareness, blood pressure severity, and all-cause mortality: independent analyses of population-based cohorts in China and the United States

Article References: Gao, Y., Peng, Y., & Wang, Y. (2026). Hypertension awareness, blood pressure severity, and all-cause mortality: independent analyses of population-based cohorts in China and the United States. BMC Public Health. https://doi.org/10.1186/s12889-026-29908-3

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29908-3

Keywords: hypertension, blood pressure, all-cause mortality, awareness, CHARLS, NHANES, systolic blood pressure, blood pressure control, Cox proportional hazards, epidemiology, public health, cohort study