Characterizing Frailty Status in the Systolic Blood Pressure Intervention Trial

Abstract

Background: The Systolic Blood Pressure Intervention Trial (SPRINT) is testing whether a lower systolic blood pressure (BP) target of 120 mm Hg leads to a reduction in cardiovascular morbidity and mortality among hypertensive, nondiabetic adults. Because there may be detrimental effects of intensive BP control, particularly in older, frail adults, we sought to characterize frailty within SPRINT to address ongoing questions about the ability of large-scale trials to enroll representative samples of noninstitutionalized, community-dwelling, older adults.

Methods: We constructed a 36-item frailty index (FI) in 9,306 SPRINT participants, classifying participants as fit (FI ≤ 0.10), less fit (0.10 < FI ≤ 0.21), or frail (FI > 0.21). Recurrent event models were used to evaluate the association of the FI with the incidence of self-reported falls, injurious falls, and all-cause hospitalizations.

Results: The distribution of the FI was comparable with what has been observed in population studies, with 2,570 (27.6%) participants classified as frail. The median FI was 0.18 (interquartile range = 0.14 to 0.24) in participants aged 80 years and older (N = 1,159), similar to the median FI of 0.17 reported for participants in the Hypertension in the Very Elderly Trial. In multivariable analyses, a 1% increase in the FI was associated with increased risk for self-reported falls (hazard ratio [HR] = 1.030), injurious falls (HR = 1.035), and all-cause hospitalizations (HR = 1.038) (all p values < .0001).

Conclusions: Large clinical trials assessing treatments to reduce cardiovascular disease risk, such as SPRINT, can enroll heterogeneous populations of older adults, including the frail elderly, comparable with general population cohorts.

Keywords: Falls; Frailty; Multimorbidities.

Figure 1.

Distribution of frailty index (FI) in SPRINT participants at randomization. Solid line denotes distribution of FI in entire SPRINT cohort. Dashed line denotes distribution of FI in 1,159 SPRINT participants 80 years of age and older at baseline. Gray line represents distribution of FI from the HYpertension in the Very Elderly Trial (HYVET) (13)

Figure 2.

Sex-specific relationship of frailty index with age in SPRINT participants. Solid line denotes estimated regression fit in SPRINT based on local polynomial regression with 95% point-wise confidence intervals (shaded areas). Single dashed lines represent regression estimates from the National Long Term Care Survey (24), whereas the double dashed lines reflect 10-year (50s, 60s, etc.) mean FI values from the Survey of Health, Ageing and Retirement in Europe (SHARE) (25).

Figure 3.

Estimated mean cumulative count curves for self-reported falls (left), injurious falls (center), and all-cause hospitalizations (right) by baseline frailty status. Shaded areas denote 95% point-wise confidence intervals based on bootstrap resampling.