American Journal of Epidemiology
Copyright © 1999 by The Johns Hopkins University School of Hygiene and Public Health
All rights reserved
Vol. 150, No. 12
Printed In U.&A.
Baseline and Previous Physical Activity in Relation to Mortality in
Elderly Men
The Zutphen Elderly Study
Fransje C. H. Bijnen,1'2 Edith J. M. Feskens,1 Carl J. Caspersen,3 Nico Nagelkerke,4 Willem L. Mosterd,2
and Daan Kromhout1
aged; cohort studies; exercise; men; mortality; physical fitness
(9). However, to our knowledge, changes in physical
activity have been investigated only in the Alameda
County Study (10) and, recently, the British Regional
Heart Study (11). In the Alameda County Study (10),
decreases in physical activity over the preceding 9
years were associated with increases in subsequent
mortality risk, and vice versa, compared with people
who did not change their activity level. In the British
Regional Heart Study (11), maintaining or taking up
light or moderate physical activity reduced mortality in
older men.
To support physical activity recommendations for the
elderly, it is useful to know which components of the
physical activity pattern may be especially important in
this respect Therefore, we examined the effect of baseline physical activity pattern on mortality risk in a
cohort of elderly Dutch men by taking into account their
physical activity pattern over the preceding 5 years.
In a cohort of predominantly middle-aged men,
beginning moderately vigorous sports activity was
associated with decreased mortality risk as compared
with not doing so (1). Because of the high mortality
rates among the elderly (2), it is of major public health
importance to know whether changes in physical activity also affect mortality risk in old age. If this is the
case, promotion of physical activity in old age could
be an important preventive measure for decreasing allcause mortality.
Inverse associations between baseline physical
activity and all-cause mortality have consistently been
reported for persons aged 60 years and older (3-8),
including participants in the Zutphen Elderly Study
Received for publication October 1, 1998, and accepted for publication March 17, 1999.
Abbreviations: Cl, confidence interval; RR, relative risk.
1
Department of Chronic Diseases, National Institute of Public
Hearth and the Environment, Bilthoven, The Netherlands.
2
Department of Medical Physiology and Sports Medicine, Faculty
of Medicine, Utrecht University, Utrecht, The Netherlands.
3
Division of Nutrition and Physical Activity, National Center for
Chronic Disease Prevention and Health Promotion, Atlanta, GA.
4
Management Team Computerization and Methodological
Consultancy, National Institute of Public Health and the Environment, Bilthoven, The Netherlands.
MATERIALS AND METHODS
Study population
The Zutphen Elderly Study is a longitudinal study of
lifestyle, chronic diseases, and health among elderly
1289
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
Understanding the effect of changes in physical activity on mortality risk may help researchers clarify
intervention strategies. This study investigated associations of physical activity at baseline and 5 years
previously with all-cause mortality risk in a cohort of 472 elderly Dutch men. Relative risks were estimated for
relations between mortality in 1990-1995 and physical activity levels in 1990 and 1985. Adjustments were made
for baseline age, chronic diseases, functional status, and lifestyle factors. In contrast to previous levels of
physical activity (adjusted p-trend = 0.39), baseline total time spent in physical activity was inversely associated
with mortality risk (p-trend = 0.004; for the most active fertile vs. the least active, relative risk = 0.44; 95%
confidence interval: 0.25, 0.80). No consistent associations were observed after fractionating total physical
activity into activities of differing intensity or into four different types of activity. Relative to maintaining a
physically active lifestyle (i.e., walking or bicycling for 20 minutes at least three times per week) in both survey
years, a gradient of increasing risk was observed from adopting an active lifestyle to becoming sedentary to
remaining sedentary (p-trend = 0.004). Recent levels of physical activity were more important for mortality risk
among elderly men than activity 5 years previously. Becoming or remaining sedentary was significantly
associated with increased mortality risk in comparison with remaining physically active. Afri J Epidemiol
1999:150:1289-96.
1290
Bijnenetal.
Examinations
Data were collected according to a standardized protocol. Clinical examinations were conducted by
trained medical staff at a local survey site. The participants completed questionnaires at home, and trained
research assistants checked the questionnaires for
missing values and inconsistencies.
Physical activity. In 1985 and 1990, physical activity was assessed with a questionnaire originally
designed for retired men (13). The questionnaire was
validated in a sample of 21 men from the 1990 survey
of the Zutphen Study (14). The 4-month test-retest correlation was 0.93 (p < 0.001), and the correlation with
total energy expenditure as measured by the doubly
labeled water technique was 0.61 (p < 0.01) (14). The
questionnaire asked about the frequency and duration
of walking and bicycling during the previous week; the
average amount of time spent weekly on hobbies and
gardening in both summer and winter; and the average
amount of time spent monthly on odd jobs and sports.
Time estimates were converted to minutes per week
for each type of activity, and minutes were summed to
yield "total physical activity" (15). Because walking,
bicycling, and gardening were the major components
of the total physical activity pattern (table 1), the other
activities were combined into one category ("other")
for further analyses. Time spent in total physical activity was also fractionated into time spent in physical
activities of "heavy" intensity, defined as activities
requiring an energy expenditure of at least 4 kcal/
kg'hour (e.g., brisk walking, bicycling at normal or
high speed, gardening), and activities of "nonheavy"
intensity (<4 kcal/kg»hour) (13). This cutoff point
takes into account the fact that elderly people generally
have a lower cardiorespiratory capacity than middleaged people. Finally, the men were classified as "phys-
ically active" or "sedentary" according to whether they
walked or cycled for 20 minutes at least three times per
week, a proxy for the assumption that regular physical
activity using large muscle groups is beneficial for
maintaining health (15, 16).
Other lifestyle characteristics. Information on cigarette smoking was collected by means of a standardized questionnaire, and participants were categorized
as "smokers" or "nonsmokers." During a home visit,
trained dietitians obtained information on diet during
the month preceding the interview using the crosscheck dietary history method (17). Dietary patterns
were classified as "healthy" or "other" on the basis of
cluster analyses (18). A healthy diet was characterized
by factors such as relatively low intakes of saturated
fatty acids and dietary cholesterol and a high intake of
fiber. Alcohol consumption was categorized as 0 g/day,
<20 g/day, and >20 g/day.
Health and mortality. In 1990, functional status
was assessed by means of a questionnaire on disabilities in activities of daily living (19). We categorized
functional status as "good" or "impaired." Good functional status was defined as needing no help in performing any of six items related to basic activities of
daily living (moving between rooms, eating, dressing,
washing and bathing, getting into and out of bed, and
using the lavatory) and four items related to mobility
(moving outdoors, using stairs, walking at least 400 m,
and carrying heavy objects a distance of 100 m).
Information on history of cardiovascular diseases
(myocardial infarction, angina pectoris, cerebrovascular accidents, and intermittent claudication), chronic
nonspecific lung diseases, cancer, and diabetes mellitus was obtained during medical examination by standardized questionnaires. Municipal registries provided
information on the vital status of the participants until
January 1995. Diagnoses were verified with hospital
discharge data and written information from the subject's general practitioner. All information was uniformly coded by three physicians. Two men were lost
to follow-up.
Statistical methods
SAS software (version 6.11) was used for all analyses (SAS Institute, Inc., Cary, North Carolina). Twosided significance levels of p = 0.05 were used. The
1990 survey was considered the baseline for the mortality follow-up. Cox's proportional hazards (survival)
analysis was used to investigate associations between
physical activity variables and all-cause mortality
between 1990 and 1995. Relative risks (with 95 percent confidence intervals and p values) for linear trend
in mortality are presented according to tertiles of physical activity (minutes/week), with the lowest tertile
Am J Epidemiol Vol. 150, No. 12, 1999
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
men living in Zutphen, The Netherlands (12). In spring
1985, a random sample of all men living in Zutphen
and born between 1900 and 1920 were invited to enter
the study; 939 (74 percent) of these men participated.
In spring 1990, the 718 surviving men from the 1985
sample were contacted for reexamination, and 560 (78
percent) participated. Between 1985 and 1990, the
majority of the participants (-95 percent) were still living independently.
The present analyses included the 472 men for
whom complete information was available with
respect to physical activity in 1985 and 1990 and other
selected lifestyle factors (e.g., cigarette smoking, alcohol consumption, and diet), as well as morbidity, functional status, and mortality follow-up between 1990
and 1995.
6"
s
TABLE 1.
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
I
Characteristics of 472 elderly men In 1990, by tertile of total time spent in leisure physical activity, Zutphen Elderty Study, 1985-1995
p
Characteristic
CO
CD
CD
Mean (SDt)
Physical activity (minutes/week)
Total activity
Tertile of total physical activity
Total
(N=472)
to
%
Lowest
("=158)
No.
Mean (SD)
Middle
(n»157)
No.
Mean (SD)
131 (89)
457(107)
No.
Highest
(n°157)
P
value*
No.
Mean (SD)
1,090(480)
559(492)
Type of activity
Walking
Bicycling
Gardening
Odd Jobs
Hobbies
Sports
138(163)
169(268)
112(191)
65(153)
59(190)
17(76)
61 (67)
33(51)
23(39)
7(21)
5(28)
2(12)
135(125)
160(134)
89(104)
38(57)
22(76)
13(42)
218(219)
315(394)
223 (278)
149(236)
149(301)
36(123)
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
Intensity of activity
"Heavy" physical activfty
Nonheavy physical activity
325(337)
233(341)
92(82)
39(59)
260(161)
197(157)
626(399)
465(479)
<0.01
<0.01
Age (years)
Alcohol Intake (g/day)
Sedentary
Healthy diet (yes)
Cigarette smoking (yes)
Impaired functional status
Diagnosed history of:
Cardiovascular diseases
Chronic nonspecific lung diseases
Cancer
Diabetes mellltus
75.1 (4.6)
10.4(13.5)
74.5 (4.3)
8.9(11.7)
76.5(5.1)
10.3(14.3)
11.5
12.7
23.6
3.2
18
20
37
5
<0.01
0.01
<0.01
0.46
0.65
<0.01
26.1
41
12
16
11
<0.01
0.02
0.43
0.11
74.3(4.1)
12.1 (14.2)
27.1
15.7
22.3
15.7
128
74
105
74
57.0
17.1
23.4
31.0
90
27
37
49
127
17.2
19.8
12.7
20
27
31
20
32.0
13.6
11.9
10.0
151
64
56
47
42.4
18.4
14.6
13.9
67
29
23
22
27.4
14.7
10.8
8.9
43
23
17
14
7.6
10.2
7.0
• Difference across physical activity tertiles by analysis of variance (for normally distributed variables), Kruskal-Wallls test (for skewed variables), or overall %' test (for categorical variables),
t SD, standard deviation.
f
o
m
CQ
CD
CD
a.
o
I
3'
m
3-
1292
Bijnenetal.
RESULTS
At baseline (1990), the study population (mean
age = 75.1 years, standard deviation 4.6) spent an
average of 80 minutes/day (standard deviation 70) on
total physical activity. Mean amounts of time spent in
physical activity were approximately 20 minutes/day,
1 hour/day, and 2.5 hours/day in the low, medium, and
high tertiles of total physical activity, respectively
(table 1). Compared with men in the lowest tertile of
activity, men in the highest tertile were younger, consumed more alcohol, and were less likely to have a history of diagnosed cardiovascular diseases or chronic
nonspecific lung diseases or impairments in functional
status. No significant differences were observed in the
proportions of men adhering to a healthy diet, currently smoking, or diagnosed with cancer. Because
cigarette smoking is a well established risk factor for
adverse health outcomes and because alcohol consumption differed significantly between tertiles,
adjustment for lifestyle factors was restricted to cigarette smoking and alcohol consumption.
Between 1985 and 1990, mean total time spent in
physical activity declined significantly from 95 minutes/day to 80 minutes/day. The Spearman correlation
coefficient for total times spent in physical activity in
1985 and 1990 was 0.55. Spearman coefficients for
correlations between various types of physical activity
in these two years were below 0.60, and those for correlations between time spent in activities of different
intensities were below 0.40. In 1985, men reporting any
physical activity were spending 45 percent of their total
time in activities requiring more than 4 kcal/kg»hour;
this figure dropped to 39 percent in 1990.
Overall, 118 participants (25 percent) died between
1990 and 1995. In the data for 1990, 61 men from the
lowest physical activity tertile died and 23 men from
the highest tertile died. The unadjusted all-cause mortality rates in the three tertiles ranged from 101.2 per
1,000 person-years to 32.7 per 1,000 person-years,
respectively. Following adjustment for age, chronic
diseases, functional status, and the lifestyle factors cigarette smoking and alcohol consumption (but not for
levels of physical activity in 1990 and 1985), mortality
risks were 0.61 (95 percent confidence interval (CI):
0.39, 0.95) for men in the middle tertile and 0.50 (95
percent CI: 0.30, 0.84) for men in the highest tertile, as
compared with the lowest.
After the data were adjusted for the above variables
and simultaneously stratified by levels of physical
activity in 1990 and 1985 (table 2), the 5-year mortality risks showed a clear inverse association with total
physical activity in 1990 (p-trend = 0.004). No association with 1985 data was noted. After time spent in total
physical activity was fractionated into four distinct
types of physical activity in the regression model (table
2), only bicycling in 1990 showed a clear inverse association. When total physical activity was fractionated
into time spent in physical activities of different intensities (table 2), no significant association was observed
with either more intense activity or less intense activity.
Because tertiles are measures that are relative to the
performance of the study population, we also applied a
distinct cutoff point between a sedentary lifestyle and a
physically active lifestyle for both survey years. The
proportion of men who walked or cycled for 20 minutes
at least three times per week increased nonsignificantly
between 1985 and 1990, from 71 percent to 73 percent.
The majority of the men (56 percent) maintained a
physically active lifestyle between 1985 and 1990,
while approximately equal proportions of the men
changed their lifestyle in either direction or remained
sedentary during both years (table 3). The baseline
characteristics for each of the groups are shown in table
3. The men who maintained a physically active lifestyle
or adopted an active lifestyle were not significantly
younger than the men who did not do so, but a smaller
proportion of these men showed impaired functional
status or a diagnosed history of cardiovascular diseases
or chronic nonspecific lung diseases.
Compared with men who had a physically active
lifestyle in both years, mortality gradually increased (ptrend = 0.004) from men who adopted an active
lifestyle (relative risk (RR) = 1.36) to men who became
sedentary (RR = 1.72) and men who remained sedentary during both years (RR = 2.01) (table 4). The
increased risk associated with a sedentary lifestyle,
either adopted or maintained, differed significantly from
the risk of men who maintained a physically active
Lifestyle during both years. Twenty-three percent of allAm J Epidemiol
Vol. 150, No. 12, 1999
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
serving as the reference category. Dummy variables
for tertiles of time spent in physical activity assessed in
1985 and 1990 were entered simultaneously into the
regression models as explanatory variables.
Because persons with clinical disease may decrease
their physical activity, increased levels of physical
activity may be associated with fewer disease outcomes in the absence of a causal relation. To preserve
the power of our analyses, we adjusted for this potential confounding primarily by including history of cardiovascular diseases, cancer, or chronic nonspecific
lung diseases, diabetes mellitus, and functional status
in 1990 as covariables in the multivariate model rather
than excluding men with diseases or impairments.
Using the adjusted population attributable risk (20),
we calculated the extent to which observed mortality
could be attributed to not maintaining a physically
active lifestyle.
Physical Activity Change and Mortality in Elderly Men
I--. S
I
cause mortality could be attributed to not maintaining a
physically active lifestyle in both survey years.
Sp
00
C§ CNJ
d d d
d
d
d
DISCUSSION
09
a
T- in
CO O
o> co en CM
T-: d T-: T^
"5
o co u) ^
o o
d d
!»• CNJ i n •tf
2 ™
d d d d
•a "5
•a
c
ca
5
8S
CM
T-' 6
T^
in
en
en en
C\i T-'
en p" co
p in
I!
I
d
CO" CD
I s * CO
C\J
d d
co
S
s
Si
o> r~ co
en
•<t < D a>
d d r^ d
T-
T<D
T^
d
a
£3
CD
c
co cp
c\i en
38
d
d d d d
o> is.
S op
oq co p
CD
i- 1 i- 1 f-1
S
S" ID
T-"
CO
c
o
d
C\J
CD
O)
CO
co
cvi
CM" Tf"
^f
£
N
d d d d
to
6'
co
:
I
£
S p K
2
C\J
i
d T^ d T^
B
8
CO
N
S
f
OO
II
m m T- i^
o
in r^
d d d d
d
i^ i^ to co
co en
S
I!
m
* en en to
in co c5
OC
CC
d
o
2 t
0) 0) ID
| s . CO
3
d d d i-1
5
d d
e-g
o 111
1
CD
f|
$•
$
CO i
StDOO
T-
Am J Epidemiol Vol. 150, No. 12, 1999
'S
S.i- -Z.
c
li
1s
To our knowledge, associations between withinsubject changes in physical activity and total mortality
risk were specifically investigated only in the Alameda
County Study (10, 21), the Harvard Alumni Study (1),
the Copenhagen male study (22), the Gothenburg study
of women (23), and the British Regional Heart Study of
older men (11). These studies differed with respect to
methodology and study population. A major strength of
our study is that the same validated physical activity
questionnaire was used for both survey years (14). We
observed that recent physical activity levels were more
important for mortality risk than past levels. This is in
line with reports on middle-aged Harvard alumni (3)
and British civil servants (24) and the older men in the
British Regional Heart Study (11), but it contrasts with
the findings of the Copenhagen study of men aged
53-75 years (22). However, the physical activity questionnaires in the Copenhagen study were not identical
for each survey year and appear not to have been validated. Furthermore, no adjustments were made for
morbidity, except that men with a history of cardiovascular diseases were excluded.
The particular importance of recent physical activity
compared with prior physical activity as observed for
tertiles of time spent in physical activity was also seen
when physical activity was dichotomized into sedentary and physically active lifestyles. Because the categorization between a physically active lifestyle and a
sedentary lifestyle was based on a proxy for generally
recommended guidelines on maintaining health among
adults (15, 16), our results indicate that these guidelines may also apply to elderly people.
Part of the deterioration associated with old age may
result from physical inactivity (25). Our data confirm
that it is very appropriate to prevent a decrease in
physical activity among elderly people. The lowest
mortality risk was observed among men who remained
physically active. This is consistent with results for the
older men in the British Regional Heart Study (11), the
Harvard Alumni Study of men aged 45-84 years (1),
and the Copenhagen study (22). In the Alameda
County Study (10) and the Gothenburg study (23), no
distinction was made between subjects who remained
physically active and subjects who remained inactive
in the "no change in physical activity" group.
However, these studies confirmed the hypothesis that
decreased physical activity might be associated with
increased mortality risk. In the Gothenburg study, a
twofold increased risk was observed among 38- to 60year-old women who decreased their physical activity
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
m CM a in
cq oq r~ tp
T-~
ID
1293
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
TABLE 3.
ro
2.
01
Characteristics of 472 elderly men In 1990 according to change in physical activity between 1985 and 1990, Zutphen Elderly Study, 1985-1995
Physically active (walking or bicycling for 20 minutes 23 times per week) in 1985 and 1990
Characteristic
In
Yes-Yes
(n == 266 (56%))
No-Yes
(n = 78(17%))
Yes-No
(n:= 69(15%))
1990
Mean (SDt)
Physical activity (minutes/week)
Total activity
%
No.
Mean (SD)
%
No.
Mean(SD)
%
No.
No-No
(n = 59(13%))
Mean (SD)
%
p
value*
No.
680(503)
614 (455)
298(345)
243(371)
<0.01
Type of activity
Walking
Bicycling
Gardening
Oddiobs
Hobbles
Sports
184(168)
238(312)
115(188)
75 (168)
57(191)
12(65)
142(182)
166(220)
146(219)
65(117)
57(191)
38(101)
38(82)
34(60)
94(157)
50(145)
68(212)
14(88)
42(72)
21 (63)
72(201)
36(129)
57 (166)
14(68)
<0.01
<0.01
0.01
<0.01
0.95
<0.01
Intensity of activity
"Heavy" physical activity
Nonheavy physical activity
286 (378)
395(353)
275 (347)
339(303)
116(170)
182(281)
79(202)
164(254)
<0.01
<0.01
Age (years)
Alcohol Intake (g/day)
Healthy diet (yes)
Cigarette smoking (yes)
Impaired functional status
Diagnosed history of:
Cardiovascular diseases
Chronic nonspecific lung diseases
Cancer
Diabetes mellltus
74.9 (4.4)
9.6(12.5)
76.3 (5.4)
8 1 (10.4)
74.7 (4.6)
14.5 (15.8)
75.2 (4.5)
11.7(16.6)
15.0
20.7
8.7
40
55
23
18.0
19.2
12.8
14
15
10
17.4
24.6
27.5
12
17
19
27.1
7.9
12.4
9.8
72
21
33
26
26.9
15.4
10.3
3.9
21
12
8
3
44.9
21.7
13.0
14.5
31
15
9
10
13.6
30.5
37.3
8
18
22
45.8
27.1
10.2
13.6
27
16
6
8
* Difference across physical activity tertiles by analysis of variance (for normally distributed variables), Kruskal-Wallls test (for skewed variables), or overall x* test (for categorical variables),
t SD, standard deviation.
en
o
Z
o
CO
CO
CO
0.10
0.06
0.86
0.34
<0.01
<0.01
<0 01
0.91
0.13
Physical Activity Change and Mortality in Elderly Men
TABLE 4. Relation of change In physical activity between
1985 and 1990 to all-cause mortality (1990-1995) among 472
elderly men, Zutphen Elderly Study, 1985-1995
Physically active
(walking or
bicycling fof
20 minutes
£3 times per week)
1985
1990
Yes
No
Yes
No
Yes
Yes
No
No
No.
of
deaths
Rate
per
1,000
personyears
48
18
27
25
40.8
53.8
100.6
113.4
Adjusted
relative
risk'.t
95%
confidence
Interval
1.00$
1.36
1.72
2.01
0 .78, 2.36
1.04, 2.85
1.19,3.39
as compared with women showing no change in physical activity. In our study, men who became inactive
had a 72 percent increased risk of mortality compared
with men who remained active, but their risk did not
differ significantly from that of men who remained
inactive. In addition, in the British Regional Heart
Study (11) and the Harvard Alumni Study (1), no significant difference in risk was observed among men
who became inactive when compared with men who
remained inactive.
Some concerns have been raised about the small,
nonsignificant excess risks associated with adopting a
physically more active lifestyle in the Copenhagen
(22) and Gothenburg (23) studies. However, the
British observations (11) and our observations of
elderly men, who may be particularly prone to complications, do not substantiate these concerns.
Instead of an expected increase in the prevalence of
a sedentary lifestyle between 1985 and 1990 due to
aging of the population, prevalence remained stable at
approximately 28 percent. It seems unlikely that overreporting of physical activity especially played a role in
1990, because physical activity questions have regularly been asked since 1960 as part of a large set of
questionnaires and other examinations. However, the
prevalence of a sedentary lifestyle in 1990 may have
been underestimated because of a secular trend due to
unexpectedly pleasant weather (26). The observed stability of the prevalence of a sedentary lifestyle might
also partly illustrate that our study population was a
select group of men, despite being a randomly selected
cohort, because they had to survive until at least 1990
to be eligible for mortality follow-up. This survival
selection may have resulted in declined strength of
observed associations between physical activity and
mortality in comparison with middle-aged populations.
Am J Epidemiol
Vol. 150, No. 12, 1999
One advantage of the Zutphen Elderly Study was
that it enabled us to take important health-related
behaviors such as cigarette smoking, alcohol consumption, detailed dietary information, and functional
status into consideration. Because the proportion of
people adhering to a healthy diet did not differ
between physical activity categories in 1990, we
decided not to adjust for diet in our final analyses.
Instead of excluding men, we adjusted not only for
history of major chronic diseases but also for impaired
functional status in 1990. Therefore, we do not believe
that subclinical diseases played a major role in our conclusions. The strength of the associations remained
similar after additional adjustment for subjective health
status. The association was mostly weakened for men
who became sedentary (RR = 1.65 (95 percent CI:
1.00, 2.74) instead of RR = 1.72). This was more pronounced after exclusion of men who died within 1 year
of follow-up (RR = 1.38; 95 percent CI: 0.78, 2.45).
Although we cannot exclude the possibility that residual confounding by subclinical disease was present,
these results may just as well underline the importance
of recent levels of activity. Furthermore, both types of
adjustment (i.e., adjustment for subjective health status
and exclusion of men who died within 1 year of followup) had only a minor effect among men who remained
sedentary (RR = 1.88 (95 percent CI: 1.11, 3.19) and
RR = 1.97 (95 percent CI: 1.13, 3.44), respectively,
instead of RR = 2.01).
In conclusion, more recent levels of physical activity were particularly important with respect to mortality risk in our population of elderly men, being
inversely associated. Remaining sedentary or adopting
a sedentary lifestyle were both associated with
increased mortality risk as compared with remaining
physically active.
ACKNOWLEDGMENTS
This study was supported by grants from the Netherlands
Heart Foundation (grant 92.356), the Netherlands
"Praeventiefonds," and the Netherlands Organization for
Scientific Research (NWO).
The authors are indebted to many people involved in this
study, especially Drs. E. B. Bosschieter and B. P. M.
Bloemberg for coordinating activities in Zutphen; die fieldwork teams and Drs. S. O. Keli, S. Kalmijn, and I. Miedema
for coding the morbidity and mortality data; Dr. J. N. Morris
(London School of Hygiene and Tropical Medicine) for providing the physical activity questionnaire; and Drs. A. J.
Schuit (National Institute of Public Health and the
Environment) and E. Bol (Utrecht University) for critical
review of the manuscript.
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
* Adjusted for age, prevalence of disease (cardiovascular diseases, cancer, chronic nonspecific lung diseases, diabetes mellitus), functional status, cigarette smoking, and alcohol intake in
1990.
t p trend for change from "yes-yes" to "no-no" = 0.0O4.
% Referent.
1295
1296
Bijnenetal.
REFERENCES
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
Am J Epidemiol Vol. 150, No. 12, 1999
Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on February 26, 2014
1. Paffenbarger RS Jr, Hyde RT, Wing AL, ct al. The association
of changes in physical-activity level and other lifestyle characteristics with mortality among men. N Engl J Med 1993;
328:538-45.
2. WHO Study Group on Epidemiology and Prevention of
Cardiovascular Diseases in the Elderly. Epidemiology and
prevention of cardiovascular diseases in elderly people: report
of a WHO study group. (WHO Technical Report Series, no.
853). Geneva: World Health Organization, 1995.
3. Paffenbarger RS Jr, Hyde RT, Wing AL, et al. Physical activity, all-cause mortality, and longevity of college alumni. N
Engl J Med 1986;314:605-13.
4. Kaplan GA, Seeman TE, Cohen RD, et al. Mortality among the
elderly in the Alameda County Study: behavioral and demographic risk factors. Am J Public Health 1987;77:307-12.
5. Grand A, Grosclaude P, Bocquet H, et al. Disability, psychosocial factors and mortality among the elderly in a rural
French population. J Clin Epidemiol 1990;43:773-82.
6. Rakowski W, Mor V. The association of physical activity with
mortality among older adults in the Longitudinal Study of
Aging (1984-1988). J Gerontol 1992;47:M 122-9.
7. Simonsick EM, Lafferty ME, Phillips CL, et al. Risk due to
inactivity in physically capable older adults. Am J Public
Health 1993;83:1443-50.
8. Ruigomez A, Alonso J, Anto JM. Relationship of health
behaviours to five-year mortality in an elderly cohort. Age
Ageing 1995;24:113-19.
9. Bijnen FC, Caspersen CJ, Feskens EJ, et al. Physical activity
and 10-year mortality from cardiovascular diseases and all
causes: The Zutphen Elderly Study. Arch Intern Med 1998;
158:1499-505.
10. Kaplan GA, Haan MN. Is there a role for prevention among
the elderly? Epidemiological evidence from the Alameda
County Study. In: Ory MG, Bond K, eds. Aging and health
care: social science and policy perspective. London, United
Kingdom: Routledge and Kegan Paul Ltd, 1989:27-51.
11. Wannamethee SG, Shaper AG, Walker M. Changes in physical activity, mortality, and incidence of coronary heart disease
in older men. Lancet 1998;351:1603-8.
12. Feskens EJ, Bloemberg BP, Pijls LT, et al. A longitudinal study
on elderly men: The Zutphen Study. In: Schroots JJ, ed.
Aging, health and competence: the next generation of longitudinal research. Amsterdam, The Netherlands: Elsevier Science
Ltd, 1993:327-33.
13. Caspersen CJ, Bloemberg BP, Saris WH, et al. The prevalence
of selected physical activities and their relation with coronary
heart disease risk factors in elderly men: the Zutphen Study,
1985. Am J Epidemiol 1991;133:1078-92.
Westerterp KR, Saris WH, Bloemberg BP, et al. Validation of
die Zutphen Physical Activity Questionnaire for the Elderly
with doubly labeled water. (Abstract). Med Sci Sports Exerc
1992;24(suppl):S68.
Bijnen FC, Feskens EJ, Caspersen CJ, et al. Physical activity
and cardiovascular risk factors among elderly men in Finland,
Italy, and the Netherlands. Am J Epidemiol 1996; 143:553- 61.
American College of Sports Medicine Position Stand. The
recommended quantity and quality of exercise for developing
and maintaining cardiorespiratory and muscular fitness in
healthy adults. Med Sci Sports Exerc 1990;22:265-74.
Bloemberg BP, Kromhout D, Obermann-de Boer GL, et al.
The reproducibility of dietary intake data assessed with the
cross-check dietary history method. Am J Epidemiol 1989;
130:1047-56.
Huijbregts PP, Feskens EJ, Kromhout D. Dietary patterns and
cardiovascular risk factors in elderly men: the Zutphen
Elderly Study. Int J Epidemiol 1995;24:313-20.
Hoeymans N, Feskens EJ, van den Bos GA, et al. Measuring
functional status: cross-sectional and longitudinal associations
between performance and self-report (Zutphen Elderly Study
1990-1993). J Clin Epidemiol 1996;49:1103-10.
Rothman KJ. Modern epidemiology. Boston, MA: Little,
Brown and Company, 1986.
Kaplan GA, Strawbridge WJ, Cohen RD, et al. Natural history
of leisure-time physical activity and its correlates: associations with mortality from all causes and cardiovascular disease over 28 years. Am J Epidemiol 1996; 144:793-7.
Hein HO, Suadicani P, Sorensen H, et al. Changes in physical
activity level and risk of ischaemic heart disease: a six-year
follow-up in the Copenhagen male study. Scand J Med Sci
Sports 1994;4:57-64.
Lissner L, Bengtsson C, Bjorkelund C, et al. Physical activity
levels and changes in relation to longevity: a prospective
study of Swedish women. Am J Epidemiol 1996;143:54-62.
Morris JN, Clayton DG, Everitt MG, et al. Exercise in leisure
time: coronary attack and death rates. Br Heart J 1990;63:
325-34.
Pescatello LS, DiPietro L. Physical activity in older adults: an
overview of health benefits. Sports Med 1993;15:353-64.
Bijnen FC, Feskens EJ, Caspersen CJ, et al. Age, period, and
cohort effects on physical activity among elderly men during
10 years of follow-up: the Zutphen Elderly Study. J Gerontol
A Biol Sci Med Sci 1998;53:M235-41.