Use of Censored Data Methods to Estimate Blood Pressure Percentiles in US Adults

Area of Science:

• Cardiovascular Health
• Epidemiology
• Biostatistics

Background:

• Blood pressure (BP) is a key modifiable risk factor for cardiovascular disease (CVD).
• Anti-hypertensive medications (MEDS) lower observed BP but are markers for higher CVD risk.
• Accurately estimating underlying BP (BP without medication) is essential for precise risk assessment.

Purpose of the Study:

• To propose and validate a novel time-to-event method for estimating underlying BP.
• To quantify the treatment effect of anti-hypertensive medications on BP.
• To derive updated, age-sex-specific BP percentiles for underlying BP.

Main Methods:

• Utilized a time-to-event analysis, treating BP measurements on MEDS as censored data.
• Applied the method to a representative sample of U.S. adults (NHANES 1999-2018).
• Calculated median treatment effects and analyzed modifications by age, BMI, and race.

Main Results:

• The median treatment effect (Underlying BP - Observed BP) was 12.0/7.7 mm Hg for SBP/DBP in men and 16.9/6.5 mm Hg in women.
• Treatment effects were significantly modified by age, body mass index (BMI), and Black race.
• Derived age-sex-specific percentile charts for underlying BP.

Conclusions:

• The novel time-to-event approach provides a more accurate estimation of underlying BP compared to fixed increments.
• Underlying BP estimation and its derived percentiles offer improved tools for CVD risk stratification.
• Findings highlight the need for personalized risk assessment considering medication effects and individual factors.