Background. Frail health in old age has been conceptualized as a loss of ph
ysiologic reserve associated with loss of lean mass, neuroendocrine dysregu
lation, and immune dysfunction. Little work has been done to define frailty
and describe the underlying pathophysiology.
Methods. Frailty status was defined in participants of the Cardiovascular H
ealth Study (CHS), a cohort of 5,201 community-dwelling older adults, based
on the presence of three out of five clinical criteria. The five criteria
included self-reported weight loss, low grip strength, low energy, slow gai
t speed, and low physical activity. We examined the spectrum of clinical an
d subclinical cardiovascular disease in those who were frail (3/5 criteria)
or of intermediate frailty status (1 or 2/5 criteria), compared to those w
ho were not frail (0/5). We hypothesized that the severity of frailty would
he related to a higher prevalence of reported cardiovascular disease (CVD)
, as well as to a greater extent of CVD, measured by noninvasive testing.
Results. Of 4,735 eligible participants, 2,289 (48%) were not frail, 299 (6
%) were frail, and 2,147 (454) were of intermediate frailty status. Those w
ho were frail were older (77.2 yrs) compared to those who were not frail (7
1.5 yrs) or intermediate (73.4 yrs) (p < .001). Frailty status was associat
ed with clinical CVD and most strongly with congestive heart failure (odds
ratio [OR] = 7.51 (95% confidence interval [CI] = 4.66-12.12). In those wit
hout a history of a CVD event (n = 1,259), frailty was associated with many
noninvasive measures of CVD. Those with carotid stenosis >75% (adjusted OR
= 3.41), ankle-arm index <0.8 (adjusted OR = 3.17) or 0.8-0.9 (adjusted OR
= 2.01), major electrocardiography (ECG) abnormalities (adjusted OR = 1.58
), greater left ventricular (LV) mass by echocardiography (adjusted OR = 1.
16), and higher degree of infarct-like lesions in the brain (adjusted OR =
1.71), were more likely to be frail compared to those who were not frail. T
he overall associations of each of these noninvasive measures of CVD with f
railty level were significant (all p < .05).
Conclusions. Cardiovascular disease was associated with an increased likeli
hood of frail health. In those with no history of CVD, the extent of underl
ying cardiovascular disease measured by carotid ultrasound and ankle-aim in
dex, LV hypertrophy by ECG and echocardiography, was related to frailty. In
farct-like lesions in the brain on magnet resonance imaging were related to
frailty as well.