Table 3

Journal of the American College of Cardiology
© 2002 by the American College of Cardiology Foundation
Published by Elsevier Science Inc.
Vol. 39, No. 9, 2002
ISSN 0735-1097/02/$22.00
PII S0735-1097(02)01799-0
Physical Activity Attenuates the Effect of Increased
Left Ventricular Mass on the Risk of Ischemic Stroke
The Northern Manhattan Stroke Study
Carlos J. Rodriguez, MD, MPH,* Ralph L. Sacco, MD, MS,†‡§ Robert R. Sciacca, ENGSCD,*
Bernadette Boden-Albala, DRPH,†‡ Shunichi Homma, MD, FACC,* Marco R. Di Tullio, MD*
New York, New York
The goal of this study was to determine whether the risk of ischemic stroke associated with
increased left ventricular mass (LVM) is modified by physical activity (PA).
BACKGROUND Increased LVM is associated with an increased risk for stroke. Physical activity can decrease
the risk of stroke and may have variable effects on LVM.
We used a case-control study design in a multiethnic population in northern Manhattan,
New York, to study 394 case subjects who had a first ischemic stroke and 413 stroke-free
control subjects. All subjects were interviewed and two-dimensional echocardiograms
obtained to determine LVM.
A sharp increase in risk of ischemic stroke was seen in the highest quartile of LVM (odds ratio
[OR]: 6.14 [95% confidence interval {CI}: 3.04 to 12.38]). Thus, increased LVM was defined
by the highest quartile of LVM. In multivariate analysis, the effect of increased LVM on the
risk of stroke was significantly decreased by the presence of any level of PA versus no PA (OR:
1.59 [95% CI: 0.99 to 2.57] p ⬍ 0.07 vs. 3.53 [95% CI: 1.94 to 6.42] p ⬍ 0.0001). Although
PA decreased the risk of stroke in all patients, the effect was stronger in subjects with
increased LVM than among those without increased LVM (p ⫽ 0.033).
CONCLUSIONS Increased LVM is associated with an increased risk of stroke, especially among sedentary
patients. Physical activity decreases the risk of stroke among patients with increased LVM to
a level comparable to that of patients without increased LVM. Recommending PA may be
a nonpharmacologic tool to reduce the stroke risk, especially among patients with increased
LVM. (J Am Coll Cardiol 2002;39:1482– 8) © 2002 by the American College of
Cardiology Foundation
Stroke has tremendous impact on public health. It continues
to rank as the third leading cause of death in the U.S. and
is also the leading cause of serious, long-term disability (1).
Data show that over 700,000 people suffer a new or
recurrent stroke each year (2). Because treatment options are
relatively limited, the modification of risk factors and
prevention of future events is especially important.
Echocardiographically determined left ventricular mass
(LVM) has been shown to be predictive of cardiovascular
disease and mortality (3), as well as being associated with an
increased risk for stroke and transient ischemic attacks (4).
The adverse effects of increased LVM are independent of
age, gender, blood pressure, cholesterol level, smoking and
presence of coronary artery disease (3–7). The mechanisms
responsible for the increased stroke risk associated with
From the *Department of Medicine and †Neurological Institute, Columbia
University, New York, New York; ‡Division of Epidemiology, Columbia University
School of Public Health, New York, New York; and §Sergievsky Center, College of
Physicians & Surgeons, Columbia University, New York, New York. Supported, in
part, by grants from the National Institute of Neurological Disorders and Stroke (R0I
NS 29993, R01-32525, NS 33248, T32 NS 07153). Dr. Rodriguez is a recipient of
the National Institute of Neurological Disorders and Stroke Minority Supplement
Grant. Presented, in part, at the American Heart Association 73rd Scientific Session,
New Orleans, November 2000.
Manuscript received October 4, 2001; revised manuscript received February 1,
2002, accepted February 8, 2002.
increased LVM are unknown (8). Thus, preventive strategies to decrease the risk have been generally lacking.
Regular physical activity (PA), even at low levels, has well
established benefits for reducing the risk of premature death
and cardiovascular disease and has been associated with
reduced coronary artery disease incidence (9 –11). In recent
years, evidence has been accumulating that supports the
protective effect of PA on stroke incidence among men and
women (12–15). The benefits are apparent even for lightto-moderate activities, such as walking, with additional
benefits to be gained from increasing the level and duration
of activity (12,15).
The relationship between exercise and left ventricular
structure is variable. Strenuous PA is known to affect LV
structure, inducing hypertrophy and dilation as normal
physiologic adaptations, the relative magnitude of which
depends on the type and intensity of the exercise (16 –19).
Exercise of relatively low-to-moderate intensity and duration has been shown to decrease blood pressure (20) and
possibly cause regression of LVM among hypertensive
subjects (21). However, the effect of PA on the risk of
ischemic stroke associated with increased LVM has not
been extensively studied.
We, therefore, used a case-control study design to inves-
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
and Acronyms
⫽ Activities of Daily Living
⫽ confidence interval
⫽ interventricular septal thickness
⫽ left ventricular diastolic dimension
⫽ left ventricular mass
⫽ metabolic equivalents of the task
⫽ Northern Manhattan Stroke Study
⫽ odds ratio
⫽ physical activity
⫽ posterior wall thickness
⫽ Quality of Well Being
tigate in a multiethnic community whether the risk of
ischemic stroke associated with increased LVM is modified
by the presence of PA and whether this effect is related to
age or gender.
Stroke patients and control subjects for the present report
were obtained from the Northern Manhattan Stroke Study
(NOMASS). NOMASS is a population-based study designed to determine stroke incidence, risk factors and
prognosis in a multiethnic, urban population. Details of
recruitment, study design and data collection have been
previously described (15,22,23).
Study patients. Case subjects met the following criteria: 1)
first ischemic stroke confirmed by head computed tomography or magnetic resonance imaging, 2) age ⬎39 years, and
3) residents of northern Manhattan for at least three
months. Controls were eligible if they: 1) had never been
diagnosed with a stroke, 2) were aged ⬎39 years, and 3)
resided in northern Manhattan for at least three months.
Control subjects were derived though random-digit dialing
and matched to cases by age (within five years), race/
ethnicity and gender.
Index evaluation of cases and control subjects. Subjects
were interviewed by trained research assistants regarding
sociodemographic characteristics, stroke risk factors and
other medical conditions. Standardized questions were
adapted from the Behavioral Risk Factor Surveillance System by the Centers for Disease Control and Prevention (24)
regarding the following conditions: arterial hypertension,
diabetes mellitus, hypercholesterolemia, transient ischemic
attack, cigarette smoking and cardiac conditions such as
myocardial infarction, coronary artery disease, angina, congestive heart failure and atrial fibrillation. Subjects also
completed a comprehensive functional status assessment,
which included Quality of Well Being (QWB) data items
and the Barthel Activities of Daily Living (ADL).
Height and weight were determined by the use of
calibrated scales. Hypertension was defined as a blood
pressure recording ⱖ140/90 mm Hg (based on the average
of two blood pressure measurements) or the patient’s selfreport of a history of hypertension or antihypertensive
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
medication use. Diabetes mellitus was defined by the
patient’s self-report of such a history or use of insulin or
hypoglycemic agent. Coronary artery disease was defined as
a history of myocardial infarction or typical angina, the
presence of a positive diagnostic test (stress test or coronary
angiography) or drug treatment. The presence of atrial
fibrillation had to be documented on a current or past
electrocardiogram or Holter monitoring. Hypercholesterolemia was defined by a fasting total cholesterol ⬎240 mg/dl
or the patient’s self-report of such a history or use of lipid
lowering therapy. For this analysis, smoking was defined as
cigarette or cigar smoking at the time of the initial evaluation.
PA assessment. The measures of PA adopted in this study
have been previously described (15). In brief, a questionnaire administered through a standardized in-person interview and adapted from the National Health Interview
Survey of the National Center for Health Statistics (25) was
used to measure recent recreational PA. The questionnaire
records the frequency and duration of 14 different recreational activities during the two-week period before the
interview. Activities were then grouped as light (⬍4.5
metabolic equivalents of the task [METS]), moderate (5 to
6 METS) and heavy (⬎6 METS). This survey form has
been found to be reliable in evaluating elderly subjects (26).
Echocardiographic evaluation. Transthoracic echocardiography was performed within two to three days of stroke
presentation for the cases and on the day of initial assessment for the controls. Hewlett-Packard Sonos 1000 and
2500 ultrasound equipment (Hewlett-Packard Division,
Andover, Massachusetts) was used in all study subjects.
Studies were performed and measurements taken according
to the guidelines of the American Society of Echocardiography (27).
Left ventricular diastolic dimension (LVDD), interventricular septal thickness (IVS) and posterior wall thickness
(PWT) were measured in all patients. Left ventricular mass
was calculated with the use of the anatomically validated
formula (28) (corrected American Society of Echocardiography method):
0.8 䡠 关 1.04 䡠 共 IVS ⫹ LVDD ⫹ PWT兲 3 ⫺ LVDD3兴 ⫹ 0.6
Left ventricular mass was divided by body surface area to
obtain the LVM index used in the data analysis.
The interpretation of the echocardiographic studies was
blinded to case-control status and other clinical characteristics. Interobserver variability in our laboratory for the
variables measured in the study is 8% to 10%.
Statistical analysis. Data are reported as mean ⫾ SD for
continuous variables and as frequency for categorical variables. Differences between mean values were assessed by
unpaired Student t test. Differences between proportions
were assessed by the chi-square test. A two-tailed p value of
⬍0.05 was considered significant. One-way analysis of
variance was used to compare LVM as a continuous variable
across PA categories.
Left ventricular mass was divided into quartiles using
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
Table 1. Baseline Characteristics of the Total Cohort
Mean ⫾ SD
⬍60 yrs
ⱖ60 yrs
Risk factors
Coronary artery disease
Atrial fibrillation
Diabetes mellitus
Cigarette smoking
Body mass index (Mean ⫾ SD)
n ⴝ 394
n ⴝ 413
109 (28)
67 (17)
211 (54)
116 (28)
70 (17)
220 (53)
182 (46)
211 (54)
184 (45)
228 (55)
68.3 ⫾ 11.5
107 (27)
287 (73)
68.6 ⫾ 12.2
91 (22)
322 (78)
307 (78)
119 (31)
138 (35)
40 (10)
123 (31)
75 (20)
26.5 ⫾ 5.8
277 (67)
156 (38)
83 (20)
10 (2)
67 (16)
64 (16)
27.2 ⫾ 5.1
p Value
cutoff points obtained from normal controls: women, quartile 1: ⱕ74 g/m2, quartile 2: 75 to 90 g/m2, quartile 3: 91
to 110 g/m2, quartile 4: ⬎110 g/m2; men, quartile 1:
ⱕ77 g/m2, quartile 2: 78 to 94 g/m2, quartile 3: 95 to
117 g/m2, quartile 4: ⬎117 g/m2.
Using the lowest quartile of LVM as the reference group,
a sharp increase in risk of ischemic stroke was observed in
the highest quartile. Thus, increased LVM was defined by
the highest quartile of LVM and compared with the three
lowest quartiles combined for categorical analyses. Unadjusted odds ratios (OR) and 95% confidence intervals (CI)
were calculated for the entire study group and for age and
gender subgroups by univariate analysis. Variables significantly associated with ischemic stroke at univariate analysis
(arterial hypertension, diabetes mellitus, coronary artery
disease, hypercholesterolemia and atrial fibrillation) were
entered as independent variables in the multivariate conditional logistic regression analysis. Cigarette smoking and
body mass index were found not to be univariate cofactors
significantly associated with stroke (p ⫽ 0.16 and p ⫽ 0.08,
respectively) in this cohort. Multivariate conditional logistic
regression analysis for was used to calculate the ORs and
95% CIs for LVM, PA and ischemic stroke after entering
variables significantly associated with ischemic stroke at
univariate analysis as potential confounding factors in the
model. Adjusted analyses were performed in the overall
group and stratified by age (⬍60 years and ⱖ60 years) and
gender. Similar analyses were not performed by race/ethnic
strata given the small number of subjects in some of our
Given that 97% of cases and 98% of controls reporting
any activity were in the light intensity activity category, an
analysis by activity intensity could not be performed. Instead, to test for a dose-response relationship, PA was
⬍ 0.0001
⬍ 0.0001
⬍ 0.0001
categorized by duration (none, ⬎0 h/week to ⬍5 h/week
and ⱖ5 h/week). Interactions between any level of PA and
LVM were assessed in the multivariate conditional logistic
regression model.
SAS version 7.0 (SAS Institute Inc., Cary, North Carolina) was used for all analyses.
Study population. From July 1, 1993 to December 31,
1998, 394 case subjects with first ischemic stroke and 413
stroke-free control subjects were enrolled. The baseline
clinical characteristics of the patients are shown in Table 1.
This was a predominantly elderly cohort (76% of the entire
cohort is age ⱖ60 years). Women constituted 54% of the
total. Overall, there were 17% non-Hispanic whites, 28%
non-Hispanic blacks and 55% Hispanics. There were no
differences in height, waist circumference, weight or body
surface area between cases and controls. Cases had a greater
prevalence of hypertension, coronary artery disease, atrial
fibrillation and diabetes mellitus.
Distribution of PA. The distribution of PA among our
study subjects is shown in Table 2. During the two weeks
before study enrollment, 47% of case subjects and 71% of
control subjects reported some PA (adjusted OR: 0.30 [95%
CI: 0.21 to 0.44]). The mean duration for any activity was
1.7 h/week ⫾ 3.1 for cases and 3.4 h/week ⫾ 4.5 for
controls. As expected in this predominantly elderly sample,
the predominant category of activity was of light intensity
(97% of cases and 98% of controls). Among subjects who
reported any PA, walking was the most frequently reported
form of exercise.
PA, LVM and the risk of stroke. In the entire group, the
risk of ischemic stroke increased as a function of LVM (Fig.
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
Table 2. Type and Duration of Physical Activity*
n (%)
Any activity
Light activity (⬃3–4.5 METS)
Moderate activity (⬃5–6 METS)
Heavy activity (⬎6.5 METS)
208 (53)
185 (47)
n (%)
120 (29)
293 (71)
167 (43)
3 (0.77)
23 (5.9)
1 (0.3)
1.5 ⫾ 2.9
0.04 ⫾ 0.76
0.08 ⫾ 0.47
0.009 ⫾ 0.16
267 (65)
6 (1.45)
87 (21)
3 (0.73)
3 (0.73)
2 (0.5)
2.9 ⫾ 4.25
0.03 ⫾ 0.39
0.3 ⫾ 0.81
0.03 ⫾ 0.39
0.02 ⫾ 0.35
0.008 ⫾ 0.12
6 (1.5)
0.03 ⫾ 0.32
10 (2.4)
11 (2.7)
0.05 ⫾ 0.44
0.04 ⫾ 0.33
4 (1.0)
7 (1.8)
0.009 ⫾ 0.10
0.04 ⫾ 0.32
5 (1.2)
13 (3.2)
0.015 ⫾ 0.15
0.04 ⫾ 0.24
*Some subjects reported participating in more than one activity.
METS ⫽ metabolic equivalents of the task.
1). Because of the substantially greater increase in risk
observed with the highest quartile of LVM, we evaluated
whether PA modified the risk of stroke associated with
increased LVM as defined by the highest quartile. Conditional logistic regression analysis demonstrated that the
effect of increased LVM on the risk of stroke was signifi-
Figure 1. Association between left ventricular (LV) mass and stroke. Left ventricular mass was categorized into quartiles. Adjusted and unadjusted odds
ratios (OR) and 95% confidence intervals (CI) for stroke risk are shown. The lowest quartile is the reference group.
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
Table 3. Effect of Left Ventricular Hypertrophy on Risk of Stroke Depending on Presence of Physical Activity
Unadjusted Odds Ratio (95% CI)
Total Cohort
40–60 yrs
⬎60 yrs
Adjusted Odds Ratio (95% CI)
Physical Activity
Absent (n ⴝ 328)
Physical Activity
Present (n ⴝ 472)
Physical Activity
Physical Activity
4.67 (2.66–8.19)
6.95 (2.73–17.74)
3.59 (1.77–7.26)
2.08 (1.35–3.22)
2.05 (1.11–3.78)
2.12 (1.15–3.92)
3.53 (1.94–6.42)
4.79 (1.78–12.92)
2.90 (1.37–6.14)
1.59 (0.99–2.57)
1.78 (0.90–3.49)
1.42 (0.72–2.80)
5.41 (2.03–14.44)
4.31 (2.24–8.31)
2.06 (0.81–5.25)
2.30 (1.38–3.83)
3.92 (1.39–11.07)
3.29 (1.64–6.59)
1.64 (0.60–4.52)
1.79 (1.03–3.13)
CI ⫽ confidence interval.
cantly attenuated by the presence of PA (Table 3). Increased
LVM had a 3.5-fold greater effect on the risk of stroke
among subjects who reported no PA. This increased risk of
stroke was more than halved among the group who reported
some level of PA.
In the unadjusted subgroup analysis (Table 3), the attenuation of the effect of increased LVM on stroke among
subjects who were physically active was observed among all
age and gender subgroups. In multivariate analysis, the
protective effect of PA persisted across gender and age but
was somewhat greater among men and subjects aged ⬍60.
Using the no-exercise group as reference, in adjusted
analysis we observed a dose-response relationship between
duration of PA and decrease in stroke risk. Increased LVM
was associated with the highest risk for stroke in the group
with no PA (OR: 3.52 [95% CI: 1.94 to 6.42]). The effect
of increased LVM on the risk of stroke progressively
decreased with increasing PA duration (⬎0 to ⬍5 h/week,
OR: 1.61 [95% CI: 0.94 to 2.75] and ⱖ5 h/week, OR: 1.43
[95% CI: 0.58 to 3.49]). There was a significant decrease in
mean LVM index with increasing duration of exercise
(Table 4).
To further clarify the relationship between the presence
of increased LVM and PA on the risk of stroke, we
stratified cases and controls by these two variables (Table 5).
Using subjects with neither increased LVM nor PA as the
reference group, we estimated the adjusted ORs for stroke
associated with some PA without increased LVM (OR:
0.40 [95% CI: 0.26 to 0.62]), increased LVM and some
reported PA (OR: 0.66 [95% CI: 0.39 to 1.12]) and
increased LVM without any reported PA (OR: 3.48 [95%
CI: 1.92 to 6.33]). The OR for stroke in the group with
increased LVM and PA was not significantly different from
that observed in subjects without increased LVM, and some
reported PA (p ⫽ 0.2).
The interaction between increased LVM and PA was
tested using multivariate conditional logistic regression. In
this model, after adjustment for other stroke risk factors,
there was a significant interaction between PA and increased LVM on the risk of stroke with a beta coefficient of
0.45 (95% CI: 0.22 to 0.94; p ⫽ 0.033). Therefore, PA
decreased the odds of stroke among patients with increased
LVM by 55% more than among those without increased
To assess for other confounding factors that may have
affected the level of PA in our study subjects, and especially
for baseline health status, we compared the baseline functional assessments among those subjects with increased
LVM and PA versus those with increased LVM and no PA.
These two groups showed no baseline differences in the
median prestroke or pre-enrollment Barthel ADL scores
and the total QWB scores. During the baseline functional
assessment, all subjects were asked whether their activity
was limited by a medical disorder or medication effect. No
significant difference in the response to this question was
found in subjects with increased LVM and PA versus those
with increased LVM and no PA.
Left ventricular mass is associated with coronary heart
disease, congestive heart failure, stroke or transient ischemic
attacks, all-cause mortality and sudden death (3–7). The
reasons for the increased risk are not immediately clear. It is
unknown if increased LVM has a direct role or is a marker
for other underlying conditions. However, given the significant cardiovascular risks associated with increased LVM,
methods to directly modify LVM have been proposed,
mostly involving pharmacologic therapy (28). Increased
LVM may be a marker of subclinical disease (29). Left
ventricular mass has been shown to correlate with carotid
wall thickness and luminal diameter (30,31) and to be
predictive of carotid atherosclerosis (32). Increased LVM
could represent a time-integrated indicator of the exposure
Table 4. Effect of PA on LVM by Duration of PA in the Entire Cohort
Mean LVM (SD)
Low PA
(n ⴝ 328)
Moderate PA
>0 to <5 h
(n ⴝ 327)
High PA
>5 h
(n ⴝ 145)
p Value
112.4 ⫾ 36.1
104.2 ⫾ 34.8
98.9 ⫾ 33.5
p ⫽ 0.0002
LVM ⫽ left ventricular mass; PA ⫽ physical activity.
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
Table 5. Interaction of LVH and Physical Activity on the Risk
of Stroke
3.48 (1.92–6.33)
0.40 (0.26–0.62)
0.66 (0.39–1.12)
Adjusted odds ratios and 95% confidence intervals for stroke risk are shown.
LVH ⫽ left ventricular hypertrophy; PA ⫽ physical activity.
to established risk factors (8,33) such as hypertension and
diabetes, whose effect over time might result in significant
cerebrovascular pathology, thus affecting the risk of stroke
and cardiovascular disease.
Relationship of LVM, PA and stroke risk. Consistent
with prior observations from the Framingham Heart study
(4), our study confirms that LVM is a risk factor for stroke
in a multiethnic population after adjustment for other stroke
risk factors. This study, however, is the first ever to
investigate the relationship of LVM and PA on the risk of
stroke. In recent years, there has been accumulating evidence supporting the protective effects of light-to-moderate
PA on stroke risk, independent of hypertension, lipid levels
and other stroke risk factors among both men and women
(13–15). Our findings suggest that the effect of increased
LVM on the risk of stroke is modified by the presence of
PA. A significant interaction was observed between increased LVM and PA. Increased LVM was associated with
an increased risk of stroke, especially among sedentary
patients, while PA decreased the risk of stroke among those
with increased LVM to a level comparable to that of
patients without increased LVM. Susceptibility to the effect
of increased LVM on the risk of stroke was found to
increase with decreasing duration of PA, and, conversely,
additional protection from the effects of increased LVM was
observed with increasing duration of exercise.
Potential mechanisms underlying the associations. The
mechanism by which PA attenuates the risk of stroke
associated with increased LVM is unclear. A proposed
mechanism may involve LVM regression. Although we
could not assess for LVM regression over time in our
cross-sectional study, we did observe significantly lower
LVM in subjects with PA versus those without it. However,
because the beneficial effect of PA was greatest among
subjects with documented increased LVM than among
those without increased LVM, mass regression alone is
likely insufficient to explain the results of this study. Also, it
is not established if LVM regression is an independent
predictor of improved prognosis. It is possible that PA
decreases the stroke risk through its known beneficial effect
on endothelial function (34,35). Physical activity has also
been shown to lower blood pressure levels (20), increase
insulin sensitivity (36) and favorably influence lipid profiles
(37). It is also possible that these effects of PA on other
established stroke risk factors may affect their interaction on
the risk of stroke associated with increased LVM. The
somewhat greater reduction in the stroke risk from increased LVM with the presence of PA observed among men
and subjects aged ⬍60 years old may be related to greater
duration of PA performed by subjects in these subgroups.
Study limitations. Our design was an observational casecontrol study and not a prospective randomized trial. Therefore, causal inferences cannot be made. Our PA assessments
were not designed to evaluate lifelong exercise practices,
exercise patterns at younger ages, physical conditioning or
direct quantitative estimations of energy expenditure. Physical activity was assessed during the preceding two weeks,
which may or may not be a reliable indicator of more
sustained PA patterns. Physical activity could be a marker of
general good health. However, we performed several analyses to assess for differences in several indexes of quality of
life, which may have affected the level of individual PA.
Among subjects with increased LVM, who were the ones
who benefited most from PA, the frequency of activity
limitations because of a medical condition or medication
effect, the pre-enrollment or prestroke ADL score and the
total QWB scores were all similar among those who were
physically active and those who were not (p ⬎ 0.05). This
observation reduces the likelihood that a better basal health
status could be the underlying factor responsible for the
decreased stroke risk observed in subjects with increased
LVM and PA.
Physical activity was not classified according to intensity,
but rather by duration, because almost all of the subjects
were in the light intensity activity category, and the predominant activity in this category was walking. As an
estimate of intensity, walking may expend between 2.5 to
12 METS (38). Given our predominantly elderly cohort,
it is likely that walking was a leisure time PA with an
estimated energy expenditure of 3 to 4.5 METS. However,
this estimate of intensity may not be sensitive enough to
establish if a relationship exists between intensity of exercise
and a protective effect on the risk of stroke associated with
increased LVM.
Conclusions. We demonstrated that an interaction exists
between PA and increased LVM on the risk of ischemic
stroke. Having increased LVM without PA conferred the
greatest risk of ischemic stroke, whereas the presence of PA
brought the stroke risk among those in the highest quartile
of LVM down to the level of patients in the lowest quartile.
This benefit was apparent for light intensity activities, such
as walking, and a modest incremental benefit may be gained
from increasing the duration of activity. Although the exact
mechanism underlying the role of increased LVM in the
pathogenesis of ischemic stroke is not clear, our study
suggests that the stroke risk associated with increased LVM
may be potentially modifiable by nonpharmacologic means
such as PA.
It is recommended that every adult accumulate at least
30 min of PA on most, preferably all, days of the week (39).
While PA can be considered a means of reducing the risk of
stroke regardless of the presence of increased LVM, our
data indicate a stronger influence on patients with increased
Rodriguez et al.
Physical Activity, LV Mass and the Risk of Stroke
The authors gratefully acknowledge the support of J. P.
Mohr, MD, Director of Cerebrovascular Research. They
also wish to thank Lynette M. De Los Santos, BS, and Ms.
Inna Titova, BS, for their assistance in the preparation of
the manuscript.
Reprint requests and correspondence: Dr. Marco R. Di Tullio,
Department of Medicine/Division of Cardiology, Columbia University, 622 West 168th Street, PH 3-342, New York, New York
10032. E-mail: [email protected]
1. American Heart Association. 2000 Heart and Stroke Statistical Update. Dallas, TX: American Heart Association, 1999.
2. Brodrick J, Brott T, Kothari R, et al. The Greater Cincinnati/
Northern Kentucky Stroke study: preliminary first ever rates of stroke
among blacks. Stroke 1998;29:415–21.
3. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart study. N Engl J Med 1990;322:
4. Bikkina M, Levy D, Evans JC, et al. Left ventricular mass and risk of
stroke in an elderly cohort. JAMA 1994;272:33–6.
5. Liao Y, Copper RS, McGee DL, Mensah GA, Ghali JK. The relative
effects of left ventricular hypertrophy, coronary artery disease and
ventricular dysfunction on survival among black adults. JAMA 1995;
6. Ghali JY, Liao Y, Simmons B, Castaner A, Cao G, Cooper RS. The
prognostic role of left ventricular hypertrophy in patients with or
without coronary artery disease. Ann Intern Med 1992;117:826 –31.
7. Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH.
Relationship of left ventricular mass and geometry to morbidity and
mortality in men and women with essential hypertension. Ann Intern
Med 1991;114:345–52.
8. Benjamin EJ, Levy D. Why is left ventricular hypertrophy so predictive of morbidity and mortality? Am J Med Sci 1999;317:168 –75.
9. Christmas C, Andersen RA. Exercise and older patients: guidelines for
the clinician. J Am Geriatric Soc 2000;48:318 –24.
10. Froelicher VF. Exercise, fitness, and coronary heart disease. Bouchard
C, Shephard RJ, Stephens T, et al., eds. Exercise, Fitness, and Health:
A Consensus of Current Knowledge. Champaign, IL: Human Kinetics Publishers, 1990:429 –50.
11. Dunn AL, Marcus BH, Kainpert JB, Garcia ME, Kohl HW, III, Blair
SN. Comparison of lifestyle and structured interventions to increase
physical activity and cardiorespiratory fitness. JAMA 1999;281:327–
12. Hu FB, Stampfer MJ, Colditz GA, et al. Physical activity and risk of
stroke in women. JAMA 2000;284:1784 –5.
13. Kiely DK, Wolf PA, Cupples LA, Beiser AS, Kannel WB. Physical
activity and stroke risk: the Framingham study. Am J Epidemiol
14. Abbot RD, Rodriguez BL, Burchfiel CM, Curb JD. Physical activity
in older middle aged men and reduced risk of stroke: the Honolulu
heart program. Am J Epidemiol 1994;139:581–93.
15. Sacco RL, Gan R, Albala B, et al. Leisure-time physical activity and
ischemic stroke risk. Stroke 1998;29:380 –7.
16. Rost R. The athlete’s heart. Cardiol Clin 1997;15:493–512.
17. Maron BJ. Structural features of the athlete heart as defined by
echocardiography. J Am Coll Cardiol 1986;10:1214 –21.
18. Rapaport E. Pathophysiological basis of ventricular hypertrophy. Eur
Heart J 1982;3 Suppl A:29 –33.
JACC Vol. 39, No. 9, 2002
May 1, 2002:1482–8
19. Pelliccia A, Culasso F, Di Paolo FM, Maron BJ. Physiologic left
ventricular cavity dilatation in elite athletes. Ann Intern Med 1999;
20. Hagberg JM, Brown MD. Does exercise training play a role in the
treatment of essential hypertension? J Cardiovasc Risk 1995;2:296 –
21. Kokkinos PF, Narayan P, Colleran JA, et al. Effects of regular exercise
on blood pressure and left ventricular hypertrophy in African American men with severe hypertension. N Engl J Med 1995;333:1462–7.
22. Sacco RL, Elkind M, Albala B, et al. The protective effect of alcohol
consumption on ischemic stroke. JAMA 1999;281:53–60.
23. Sacco RL, Roberts K, Albala B, et al. Race-ethnicity and determinants
of carotid atherosclerosis in a multiethnic population: the Northern
Manhattan Stroke study. Stroke 1997;28:929 –35.
24. Gentry EM, Kalsbeek WD, Hogelin GC, et al. The Behavioral Risk
Factor Surveys, II: design, methods, and estimates from combined
state data. Am J Prev Med 1985;1:9 –14.
25. Moss AJ, Parsons VL. Current estimates from the National Health
Interview Survey, United States, 1985. Washington, DC: National
Center for Health Statistics. US Department of Health and Human
Services. Publication PHS 86-1588, Vital Health Stat 10 (160); 1986.
26. McPhillips JB, Pellettera KM, Barrett-Connor E, Wingard DL,
Criqui MH. Exercise patterns in a population of older adults. Am J
Prev Med 1989;5:65–72.
27. Schiller NB, Shah PM, Crawford M, et al. for the American Society
of Echocardiography, Committee on Standards, Subcommittee of
Quantitation of Two Dimensional Echocardiograms. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr 1989;2:358 –67.
28. Devereux RB. Therapeutic options in minimizing left ventricular
hypertrophy. Am Heart J 2000;139:S9 –S14.
29. Devereaux RB, Alderman MH. Role of preclinical cardiovascular
disease in the evolution from risk factor exposure to development of
morbid events. Circulation 1993;88:1444 –55.
30. Roman MJ, Saba PS, Pini R, et al. Parallel cardiac and vascular
adaptation in hypertension. Circulation 1992;86:1909 –18.
31. Kronmal RA, Smith VE, O’Leary DH, Polak JF, Gardin JM, Manolio
TA. Carotid artery measures are strongly associated with left ventricular mass in older adults (a report from the Cardiovascular Heart
study). Am J Cardiol 1996;77:628 –33.
32. Okin PM, Devereaux RB, Kligfield P. Association of carotid atherosclerosis with electrocardiographic myocardial ischemia and left ventricular hypertrophy. Hypertension 1996;28:3–7.
33. Post WS, Larson MG, Levy D. Impact of left ventricular structure on
the incidence of hypertension: the Framingham Heart study. Circulation 1994;90:179 –85.
34. DeSouza CA, Shapiro LF, Clevenger CM, et al. Regular aerobic
exercise prevents and restores age-related declines in endothelium
dependent vasodilation in healthy men. Circulation 2000;102:1351–7.
35. Hirofumi T, Dinemnno FA, Monahan KD, Clevenger CM, DeSouza
CA, Seals DR. Aging, habitual exercise, and dynamic arterial compliance. Circulation 2000;102:1270 –5.
36. Mayer-Davis EJ, D’Agostino R, Jr., Karter AJ, et al. Intensity and
amount of physical activity in relation to insulin sensitivity: the Insulin
Resistance Atherosclerosis study. JAMA 1998;279:666 –74.
37. Stefanick ML, Mackey S, Sheehan M, Ellsworth N, Haskell WL,
Wood PD. Effects of diet and exercise in men and postmenopausal
women with low levels of HDL cholesterol and high levels of LDL
cholesterol. N Engl J Med 1998;339:12–20.
38. Ainsworth BE, Haskell WL, Leon AS. Compendium of physical
activities: classification of energy costs of human physical activities.
Med Sci Sports Exerc 1993;25:71–80.
39. NIH Consensus Development Panel on Physical Activity and Cardiovascular Health. Physical activity and cardiovascular health. JAMA