JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY

VOL. 65, NO. 6, 2015

2015 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

ISSN 0735-1097/$36.00

PUBLISHED BY ELSEVIER INC.

Letters
The mean age was 49.6 ! 11 years of age, with

No Evidence of an Upper
Threshold for Mortality
Benet at High Levels of
Cardiorespiratory Fitness

63.7% males. The mean cardiorespiratory tness level

for the high-tness population was 11.2 ! 1.6 METs,

with 62% of patients in the 10- to 11-METs group; 33%

in the 12- to 13-METs group; and 5% in the group
at $14 METs.
Figure 1A shows a graded decrease in mortality
across

Although the importance of achieving a healthy

increasing

high

cardiorespiratory

tness

groups after the age of 30 years. Kaplan-Meier anal-

level of cardiorespiratory tness is well established

ysis demonstrated a statistically signicant progres-

(1), some experts have suggested that individuals

sive improvement in survival across increasingly

achieving extreme levels of tness are at an increased

high-tness groups (p < 0.001) that persisted for up

risk of death resulting from cardiovascular disease

to 20 years of follow-up.

(CVD) (2). However, it may not be accurate to

In multivariate models, there was a signicant

conclude that all such individuals achieving elite

reduction in mortality for the most highly t groups,

levels of tness are at increased risk (3). To inform

independent of all study covariates. Compared with

this debate, we sought to determine whether there

individuals in the reference, 10- to 11-METs group,

is a threshold where higher levels of tness were

those in the 12- to 13- and >14-METs groups had a

no longer associated with a lower risk for all-cause

mortality reduction of 51% (95% CI: 0.44 to 0.54) and

mortality.

79% (95% CI: 0.15 to 0.28), respectively, in an unad-

Of 69,855 patients, we studied a subgroup of 37,855

justed model; 36% (95% CI: 0.57 to 0.71) and 65%

patients free of known CVD, with high tness levels

(95% CI: 0.25 to 0.47) in Model 1; and 32% (95% CI:

of $10 metabolic equivalents (METs) from the Henry

0.61 to 0.75) and 61% (95% CI: 0.28 to 0.53) in Model 2,

Ford Exercise Testing Project (The FIT Project) (4).

respectively. In a graphic analysis of predicted mor-

Exercise capacity, expressed in METs, was estimated

tality across the range of commonly attained MET

from the peak treadmill workload. All patients were

values, we observed a consistent decrease in mortal-

followed for all-cause mortality by using linkage with

ity with higher tness with no evidence of a threshold

the death master le.

effect (Figure 1B).

Follow-up was 11.5 ! 4.5 years for all-cause mor-

In summary, in a large study of cardiorespiratory

tality. First, we calculated crude mortality stratied

tness testing, we observed no upper threshold for

by age and high-tness groups (METs 10 to 11, 12 to 13,

mortality benet with increasing tness, even in the

and $14). Kaplan-Meier curves were constructed

most highly t subjects. Our data appear to contrast

using time-to-death data, with log rank testing for

with recent reports in the mainstream media sig-

signicant differences. We then performed hierar-

naling possible cardiovascular harm at the highest

chical multivariable Cox regression analysis to cal-

exercise levels (2,3) and support the public health

culate hazard ratios and 95% condence intervals

message that all patients should engage in regular,

(CI) associated with increasing levels of high cardio-

moderate to vigorous exercise (3).

respiratory tness: unadjusted; adjusted for age,

In 2013, OKeefe et al. (3) argued that mortality

sex, and race (Model 1); and further adjustment for

reductions persist until 13 METs, at which point there

traditional risk factors including smoking, diabetes,

may be no additional gains and possible harms in

hypertension, antihypertensive medication, hyper-

cardiovascular health and longevity. In fact, some

lipidemia, and family history (Model 2). Finally, using

experts have advanced the concept of cardiotox-

standard

fully

icity due to the acute abnormalities associated with

adjusted regression models, we plotted estimated

extreme endurance exercise (EEE) (2). However, there

mortality relative to the complete range of common

is also recognition that the changes associated with

METs values (1 to 16) and constructed a graphic

EEE may resolve in most people and may ultimately

assessment for a threshold effect.

represent no major harm. For example, in 2012,

post-estimation

techniques

from

Letters

630

JACC VOL. 65, NO. 6, 2015

FEBRUARY 17, 2015:62934

Kim et al. (5) reported that the risk of sudden cardiac

In conclusion, our study suggests a continuous,

death in all U.S. marathoners from 2000 to 2010 was

graded reduction in mortality extending to in-

0.0005%, suggesting that cardiotoxicity concerns

dividuals with an estimated exercise capacity of >14

may be overstated.

METs. Although more research is needed on the car-

There are several limitations to our study. The FIT

diovascular effects of EEE, our data caution against

Project describes a measurement of cardiorespiratory

any public health message that might dissuade pa-

tness in everyday clinical practice with few extreme

tients from routine vigorous physical activity with the

athletes and cannot be considered a study of car-

goal of reaching the highest levels of cardiorespira-

diotoxicity attributable to EEE. Also, patients in The

tory tness.

FIT Project might have been more likely to have other

traditional cardiovascular risk factors than ultramarathoners, which may have accentuated the benets of increasing physical tness. Finally, although
the population was large, our study still lacked sufcient power to model patients with >16 METs,
although exploratory analysis suggested a continued
mortality decrease in this group.

David I. Feldman, BS
Mouaz H. Al-Mallah, MD
Steven J. Keteyian, PhD
Clinton A. Brawner, PhD
Theodore Feldman, MD
Roger S. Blumenthal, MD
*Michael J. Blaha, MD, MPH
*Johns Hopkins Ciccarone Center
for the Prevention of Heart Disease
Carnegie 565A

F I G U R E 1 Cardiorespiratory Fitness and Mortality in the FIT Project

Probability of Death

600 North Wolfe Street

20
18
16
14
12
10
8
6
4
2
0

Baltimore, Maryland 21287

METS
10 - 11

E-mail: MBlaha1@jhmi.edu

12 - 13

http://dx.doi.org/10.1016/j.jacc.2014.11.030

14

REFERENCES
1. Roger VL, Jacobsen SJ, Pellikka PA, et al. Prognostic value of treadmill
exercise testing: a population-based study in Olmsted County, Minnesota.
Circulation 1998;98:283641.

Age < 30

Age 30 - 40

Age 40 - 50

Age 50 - 60

Age > 60

10 - 11

2.9

3.2

3.8

6.0

17.6

12 - 13

0.8

1.9

2.9

5.4

12.7

14

1.4

1.7

1.6

3.0

11.3

2. OKeefe JH, Patil HR, Lavie CJ, et al. Potential adverse cardiovascular
effects from excessive endurance exercise. Mayo Clin Proc 2012;87:58795.
3. OKeefe JH, Lavie CJ. Run for your life . at a comfortable speed and not
too far. Heart 2013;99:5169.
4. Al-Mallah MH, Keteyian SJ, Brawner CA, et al. Rationale and design of
the Henry Ford Exercise Testing Project (the FIT project). Clinical Cardiology
2014;37:45661.

0.08

0.8

Probability of Death

Mortality (%)

Johns Hopkins Hospital

0.6

0.4

5. Kim JH, Malhotra R, Chiampas G, et al. Cardiac arrest during long-distance

running races. N Engl J Med 2012;366:13040.

0.06
0.04
0.02
0

10

11

12

13
14
METS

15

16

0.2

0
0

10
METS

12

14

16

18

(A) Graded improvement in survival is shown across the increasingly high-tness groups in
patients $30 years of age. (B) Multivariate-adjusted estimated survival as a function of
METs. Graphic assessments reveal no upper threshold of survival among even the most
highly t individuals. FIT Henry Ford Exercise Testing Project; METS metabolic

equivalents.

Cholesterol Crystals
Associate With Coronary
Plaque Vulnerability In Vivo
The high local concentration of cholesterol in foam
cells has been reported to formulate cholesterol
crystals, which trigger a local inammatory response
(1). Intracellular crystals also induce apoptosis of
foam cells, leading to further attraction of macrophages and development of a lipid-rich necrotic core
(2). These effects may suggest a potential contribution of cholesterol crystals to plaque destabilization.
Frequency-domain optical coherence tomography