Abstract
Background
Short-term systolic and diastolic blood pressure variability have been associated with adverse cardiovascular (CVD) outcomes, especially in conjunction with traditional CVD risk factors. However, there are limited data on the relationship between short-term pulse pressure (systolic minus diastolic blood pressure) variability (PPV) and mortality.
Purpose
We examined the association between PPV and death due to cardiovascular causes.
Methods
Data from the United States National Health and Nutrition Examination Surveys (NHANES-III, 1988–1994) were linked to death certificates from the National Death Index until December 31, 2015. A total of 6,340 adults (2,981 men and 3,359 non-pregnant women) aged ≥20 years who were normotensive (BP<140/90, without history of hypertension and not taking antihypertensive medication) were followed for an average of 22.3 years. Individuals with any self-reported history of CVD (heart failure, myocardial infarction, stroke) were excluded. PPV was calculated as the standard deviation of six pulse pressure measurements across two visits less than two weeks apart. PPV was categorized into quartiles: Q1: ≤4, Q2: 4.1–6, Q3: 6.1–8, Q4: >8 mmHg. The primary outcome was CVD mortality. Cox proportional hazards models were used to determine hazard ratios, adjusting for demographics & sociobehavioral factors (age, race, ethnicity, poverty-income ratio, smoking status), cardiometabolic factors (waist circumference, HDL, triglycerides, microalbuminuria, diabetes status), BP-related factors (mean pulse pressure, between-visit variability), and accounting for the complex survey design.
Results
Whereas the proportion of CVD death among men was similar across quartiles, women in Q4 had significantly higher proportion of CVD death (Figure 1; p=0.0055). Women in Q4 of PPV had significantly higher risk of CVD mortality (unadjusted HR 3.63, 95% CI 1.66, 7.90) compared to Q1, even after 1) adjustment for demographics & sociobehavioral factors (HR 2.80, CI 1.40, 5.60), 2) additional adjustment for cardiometabolic factors (HR 2.59, CI 1.33, 5.05), and 3) additional adjustment for mean pulse pressure and between visit variability (HR 2.71, CI 1.42, 5.17). Men in Q4 also had increased, but insignificant, risk of CVD mortality (adjusted HR 1.06, CI 0.38, 2.96). Gender significantly modified the effect of PPV on CVD mortality (p=0.036 for interaction term). When looking at the first visit alone, every 1 mmHg increase in PPV was associated with a 11% significant increase in risk of cardiovascular mortality in females (1.11 [1.03, 1.19]), but only a 1% insignificant increase in males (1.01 [0.91, 1.12]).
Conclusions
These NHANES data with an average 22.3 years of follow-up indicate that two visit pulse pressure variability is associated with cardiovascular death and that this effect is more pronounced in women.
FUNDunding Acknowledgement
Type of funding sources: None. Figure 1. Pulse Pressure Variability and Mortality
Association between short-term pulse pressure variability and cardiovascular death among normotensive adults
