The utility of long-term blood pressure variability for cardiovascular risk prediction in primary care

Area of Science:

• Cardiovascular Medicine
• Epidemiology
• Biostatistics

Background:

• Blood pressure (BP) is a known risk factor for cardiovascular disease (CVD).
• Standard cardiovascular risk scores utilize systolic blood pressure (SBP) for clinical decision-making.
• Recent research highlights within-person BP variability as a significant predictor of CVD.

Purpose of the Study:

• To investigate if incorporating BP variability into existing cardiovascular risk scores enhances their predictive accuracy.
• To assess the added value of BP variability alongside standard risk factors in predicting CVD.

Main Methods:

• Utilized cohort data from English general practices (Clinical Practice Research Datalink) linked to hospital and mortality records.
• Calculated SBP variability independent of the mean across multiple clinic visits for patients aged 40-74.
• Developed and validated predictive models, comparing a reference model with standard risk factors to an index model including BP variability.

Main Results:

• BP variability demonstrated a statistically significant association with CVD independently of other risk factors in the derivation dataset (P=0.005).
• In the validation dataset, both the reference model and the index model incorporating BP variability showed comparable performance.
• Key validation statistics, including c-statistic, R, and calibration ratio, were nearly identical between the two models.

Conclusions:

• While BP variability is statistically associated with CVD in large datasets, its inclusion does not significantly enhance the performance of current cardiovascular risk scores.
• Further research may be needed to explore alternative methods for integrating BP variability into risk prediction models.