An outcome-driven threshold for pulse pressure amplification

Pulse pressure (PP), the difference between systolic and diastolic blood pressure, amplifies from central to peripheral arteries. This amplification (PPA), often represented by the brachial-to-aortic PP ratio, decreases with age and factors like hypertension, dyslipidemia, diabetes, and chronic kidney disease. Arterial stiffening, leading to earlier return of reflected waves, increases central systolic pressure and thus reduces PPA. Studies have linked lower PPA to higher cardiovascular risk, coronary artery disease, renal function decline, and mortality. However, prospective studies on major cardiovascular complications related to PPA and a corresponding outcome-driven threshold are lacking. This individual-participant meta-analysis utilizes the International Database of Central Arterial Properties for Risk Stratification (IDCARS) to address this gap, aiming to define, calibrate, and validate an outcome-driven threshold for PPA to alert clinicians to the risk of cardiovascular complications associated with falling PPA.

Existing literature demonstrates a strong correlation between pulse pressure amplification (PPA) and cardiovascular risk. Studies have shown that reduced PPA, indicative of arterial stiffening, is associated with increased risk of cardiovascular events, coronary artery disease, and mortality. However, a consistent and outcome-driven threshold for PPA to guide clinical practice remains elusive. The current study addresses this gap by employing a large dataset and rigorous statistical methods to establish such a threshold.

This individual-participant meta-analysis utilized data from the International Database of Central Arterial Properties for Risk Stratification (IDCARS), comprising 5608 participants (after exclusions) aged 30–96 years. Brachial blood pressure was measured, and central aortic pressure was derived from radial arterial waveforms using SphygmoCor software. The primary endpoint was a composite cardiovascular outcome (cardiovascular mortality, sudden death, myocardial infarction, heart failure, stroke, coronary revascularization). The coronary endpoint included sudden death and fatal/non-fatal myocardial infarction and coronary revascularization. The study population was randomly split into discovery (n=3945) and replication (n=1663) samples. Cox regression models were used to assess the association between PPA and endpoints, adjusting for age, sex, heart rate, BMI, smoking, alcohol consumption, cholesterol ratio, eGFR, antihypertensive drug use, history of CVD, and diabetes. An outcome-driven PPA threshold was determined in the discovery sample using a two-pronged strategy: (1) examining HRs for 0.1 PPA increments and identifying where the upper 95% CI crossed unity, and (2) identifying PPA levels yielding a 5-year risk equivalent to risks associated with specific brachial systolic BP levels. Model calibration was evaluated, and the performance of the PPA threshold in risk stratification was assessed using measures like AUC, IDI, and NRI. Subgroup analyses were conducted based on sex, age, systolic BP, and antihypertensive treatment. A sensitivity analysis for the coronary endpoint additionally adjusted for diastolic BP.

The analysis of 5608 participants revealed a rounded PPA threshold of 1.3. In the discovery dataset, multivariable-adjusted hazard ratios for cardiovascular and coronary endpoints comparing PPA <1.3 to ≥1.3 were 1.54 (95% CI: 1.00–2.36) and 2.45 (95% CI: 1.20–5.01), respectively. These findings were replicated in the independent test dataset, although significance for the coronary endpoint was not reached in the continuous analysis of the replication dataset due to smaller sample size. The models showed good calibration. Net reclassification improvement (NRI) was significant for both endpoints, indicating improved risk stratification using the PPA threshold. Subgroup analyses showed a stronger association between low PPA and cardiovascular events in individuals under 60 years of age. In this younger group, a significantly higher proportion of women were identified as high-risk (PPA<1.3) compared to older individuals. Sensitivity analysis including diastolic blood pressure adjustment showed similar results.

This study provides an outcome-driven threshold of <1.3 for brachial-to-aortic pulse pressure ratio (PPA) as a predictor of cardiovascular and coronary events. The finding that this threshold is a significant risk factor, especially in women aged 30-60 years, highlights an underestimated risk factor in this population. The significant NRI values demonstrate that incorporating PPA into risk prediction models can improve risk stratification beyond conventional risk factors. The age-dependent effect suggests that PPA may be a particularly relevant risk marker in younger individuals. Further research might explore the underlying mechanisms of the age-related differences and the role of PPA in women. Future research should also consider validating this threshold in diverse populations.

This large individual-participant meta-analysis establishes a clinically relevant threshold of <1.3 for pulse pressure amplification (PPA) as a predictor of cardiovascular and coronary events. This threshold adds value in risk stratification, particularly among women aged 30-60 years, highlighting the utility of pulse wave analysis in cardiovascular risk assessment. Future studies should explore the biological mechanisms underlying these findings and evaluate the applicability of this threshold across diverse populations and clinical settings.

While the study included a large sample size, the generalizability of the findings may be limited by the specific characteristics of the IDCARS database, potential biases inherent in observational studies, and the reliance on SphygmoCor measurements. The relatively short median follow-up period might limit the detection of rare events. Further studies with longer follow-up are needed to validate these results and refine our understanding of PPA's role in long-term cardiovascular risk.