Exercise Cornucopia 2010

COMMENTARY
Evidence Regarding the Benefits of Physical Exercise
J
UST OVER A QUARTER OF A ity as well as reduced risk of physi- herence to the resistance training
century ago, thought lead- cal disability and dependence, the program in this study had better
ers in the aging research most important health outcome, even Stroop test scores, a measure of
arena recognized that the more than death, for most older selective attention and conflict
promotion of increased people. In this issue of the Archives, resolution. Interestingly, although
physical activity was a 4 new articles move the scientific en- lower brain volume is typically as-
relatively unexplored but terprise in this area further along the sociated with poorer cognitive func-
potentially powerful avenue to pro- path toward the goal of understand- tion, the intervention that had a
moting healthy aging.1,2 Since that ing the full range of important aging- positive impact on executive func-
mandate to examine the role of ex- related outcomes for which exercise tion showed that the total brain vol-
ercise in preventive gerontology was has a clinically relevant impact. Two ume as measured by magnetic reso-
issued, and during the past decade of these studies4,5 evaluated the rela- nance imaging was actually lower
in particular, a growing body of evi- tionship between physical exercise after 12 months of exercise in the re-
dence has accumulated that has and decline in cognitive function, a sistance training group. The inter-
given legs to the hypothesis that the major contributor to loss of indepen- vention did not, however, have an
promotion of physical activity may dence and related institutionaliza- impact on other domains of cogni-
be the most effective prescription tion in elderly individuals. Opti- tion, such as working memory, nor
that physicians can dispense for the mism in this area, especially did the sample size, duration, and
purposes of promoting successful ag- cognition, is sorely needed owing to study population help in determin-
ing. Today it is recognized that vir- the failure of promising interven- ing the impact of the intervention on
tually all of the diseases and conditions aimed at preventing cognitive dementia incidence.
tions that lead to physical disability decline and dementia in the recent In the other 2 articles, the au-
in older adults have as part of their past.6,7 thors examined a composite mea-
etiology a component of personal In the study by Etgen et al,4 physi- sure of healthy aging8 and a mea-
lifestyle choices (eg, physical inac- cal activity was independently asso- sure of cost-effectiveness, 9 both
tivity) in addition to biological ag- ciated with reduced risk for cogni- increasingly important outcomes re-
ing and environmental exposure. tive function decline in a group of lated to debates on best practices for
Improving our understanding of the self-selected volunteers residing in maintaining the health and indepen-
relative contributions of these fac- a single community in the Bavarian dence of aging populations in de-
tors to aging-related loss of inde- region of Germany. Cognitive func- veloped countries. In the article by
pendence and the subsequent de- tion was assessed using a brief mea- Sun et al,8 data from the Nurses’
velopment and implementation of sure of global cognitive function, and Health Study demonstrate that mod-
prevention and treatment ap- the results essentially confirm simi- est levels of physical exercise dur-
proaches is the essence of the present lar epidemiologic findings by other ing middle age increased the prob-
research mission for scientists work- investigators while leaving open the ability of successful aging beyond 70
ing in the areas of geriatric medi- question of whether these results years, defined by low burden of co-
cine and gerontology. should be generalized beyond the morbidity, no mental health dis-
study population. These results es- ease, and preserved physical and
See also pages 170, 179, sentially add to the existing data sup- cognitive function. The study by
186, and 194 porting the need for well-designed Kemmler et al 9 again shows the
clinical trials testing the impact of health benefits of physical exercise
It is also now well established that physical activity on clinically mean- on risk factors for disability in older
higher quantities of physical activity ingful cognitive outcomes. In an- adults but does not establish defini-
have beneficial effects on numerous other article, Liu-Ambrose et al5 de- tive evidence for the cost-effective-
age-related conditions such as scribe the use of a broader battery ness of these interventions in this
osteoarthritis, falls and hip fracture, of neuropsychological tests to evalu- population. In summary, the find-
cardiovascular disease, respiratory dis- ate the impact of a 1-year weekly or ings of all 4 of these articles fit into
eases, cancer, diabetes mellitus, os- twice-weekly resistance training pro- the larger and optimistic view of ger-
teoporosis, low fitness and obesity, gram on cognitive decline in 155 Ca- ontologists regarding the power of
and decreased functional capacity, all nadian women aged 65 to 75 years. higher levels of physical activity to
conditions that greatly increase the The authors show that resistance aid in the prevention of late-life dis-
risk of reduced independence in late training reduced the risk for de- ability owing to either cognitive im-
life.3 Regular physical activity has also cline in measures of executive func- pairment or physical impairment,
been associated with greater longev- tion. Individuals who had high ad- separately or together.
(REPRINTED) ARCH INTERN MED/ VOL 170 (NO. 2), JAN 25, 2010 WWW.ARCHINTERNMED.COM
124
©2010 American Medical Association. All rights reserved.

The impact of unsuccessful ag- domized to a physical activity or to a Correspondence: Dr Pahor, Depart-
ing is readily apparent to both cli- health education program and fol- ment of Aging and Geriatric Re-
nicians and nonclinicians. Physical lowed for up to approximately 4 years search, Institute on Aging, Univer-
inactivity is one of the strongest pre- to assess the effects on major mobil- sity of Florida, PO Box 100107,
dictors of such unsuccessful aging ity disability, defined as inability to Gainesville, FL 32610-0107
for older adults and is perhaps the walk 400 m, cognition, injurious falls, (mpahor@aging.ufl.edu).
root cause of many unnecessary and and other health outcomes. The LIFE Financial Disclosure: None re-
premature admissions to long- study expands on the promising reported.
term care. Clinical trials targeting the sults of a pilot study10 that found the
1. Holloszy JO. Exercise, health, and aging: a need
diseases and conditions that popu- rate of onset of mobility disability was for more information. Med Sci Sports Exerc. 1983;
late the pathway to disability for lower among a group of older adults 15(1):1-5.
older people have demonstrated the who engaged in a structured exer- 2. Bortz WM II. Disuse and aging. JAMA. 1982;248
efficacy of physical activity to precise program for a year compared with (10):1203-1208.
a group of seniors who took part in a 3. van der Bij AK, Laurant MG, Wensing M. Effec-
vent or delay complications. How- tiveness of physical activity interventions for older
ever, while scientific value of these health education program for a year. adults: a review. Am J Prev Med. 2002;22(2):
studies cannot be disputed, whether The LIFE study will provide defini- 120-133.
the results can or should be trans- tive evidence regarding whether 4. Etgen T, Sander D, Huntgeburth U, Poppert H, Förstl
physical activity is effective and prac- H, Bickel H. Physical activity and incident cogni-
lated into recommendations for pre-
tive impairment in elderly persons: the INVADE
venting disability progression in the tical for preventing major mobility study. Arch Intern Med. 2010;170(2):186-193.
broader community of older adults disability. These results will have cru- 5. Liu-Ambrose T, Nagamatsu LS, Graf P, Beattie BL,
typically seeking care in the outpa- cial implications for public health pre- Ashe MC, Handy TC. Resistance training and ex-
tient setting is the burning ques- vention in a rapidly aging society and ecutive functions: a 12-month randomized con-
will fill an important gap in knowl- trolled trial. Arch Intern Med. 2010;170(2):170-
tion that remains. This is because 178.
most of these studies and other pub- edge for practicing evidence-based 6. DeKosky ST, Williamson JD, Fitzpatrick AL, et al;
lished studies have enrolled younger geriatric medicine. The study will also Ginkgo Evaluation of Memory (GEM) Study In-
and generally healthier older people. yield valuable information concern- vestigators. Ginkgo biloba for prevention of de-
ing the efficacy and effectiveness of mentia: a randomized controlled trial. JAMA. 2008;
A few major issues remain from 300(19):2253-2262.
the original mandate to explore the physical activity across a broad spec- 7. Shumaker SA, Legault C, Kuller L, et al; Wom-
efficacy of exercise in the 1980s. Of trum of important health outcomes. en’s Health Initiative Memory Study. Conjugated
these, the primary issue is the lack of The study will have an impact on both equine estrogens and incidence of probable de-
randomized controlled trials (RCTs) clinical practice and public health mentia and mild cognitive impairment in post-
policy and will, therefore, benefit in- menopausal women: Women’s Health Initiative
of adequate scope examining the fea- Memory study. JAMA. 2004;291(24):2947-2958.
sibility, safety, and impact of exer- dividuals and society. 8. Sun Q, Townsend MK, Okereke OI, Franco OH, Hu
cise over a long duration (⬎2 years) Along with the expected results FB, Grodstein F. Physical activity at midlife in re-
in persons who are at highest risk for of the LIFE study, the 4 new stud- lation to successful survival in women at age 70
ies in this issue of the Archives, and years or older. Arch Intern Med. 2010;170(2):
loss of independence on relevant geri- 194-201.
atric outcomes, such as onset of ma- the evidence from clinical studies 9. Kemmler W, von Stengel S, Engelke K, Häberle
jor disability. To address this impor- over the past 25 years, we have never L, Kalender WA. Exercise effects on bone min-
tant clinical and public health had greater reason to be hopeful re- eral density, falls, coronary risk factors, and health
question, the Lifestyle Interventions garding the potential for exercise to care costs in older women: the Randomized Con-
become a proven and generalizable trolled Senior Fitness and Prevention (SEFIP)
and Independence for Elders (LIFE) Study. Arch Intern Med. 2010;170(2):179-185.
study, a phase 3 multicenter RCT of strategy for promoting successful ag- 10. Pahor M, Blair SN, Espeland M, et al; LIFE Inves-
physical activity to prevent major mo- ing in the expanding population of tigators. Effects of a physical activity interven-
bility disability, will start recruiting older adults. tion on measures of physical performance: re-
sults of the Lifestyle Interventions and
early next year. A total of 1600 older Independence for Elders Pilot (LIFE-P) study.
sedentary persons who are at risk for Jeff Williamson, MD, MHS J Gerontol A Biol Sci Med Sci. 2006;61(11):
major mobility disability will be ran- Marco Pahor, MD 1157-1165.
125

ORIGINAL INVESTIGATION
Resistance Training and Executive Functions

A 12-Month Randomized Controlled Trial
Teresa Liu-Ambrose, PhD, PT; Lindsay S. Nagamatsu, MA; Peter Graf, PhD;
B. Lynn Beattie, MD; Maureen C. Ashe, PhD, PT; Todd C. Handy, PhD
Background: Cognitive decline among seniors is a press- Results: Both resistance training groups significantly im-
ing health care issue. Specific exercise training may com- proved their performance on the Stroop test compared
bat cognitive decline. We compared the effect of once- with those in the balance and tone group (Pⱕ.03). Task
weekly and twice-weekly resistance training with that of performance improved by 12.6% and 10.9% in the once-
twice-weekly balance and tone exercise training on the weekly and twice-weekly resistance training groups, re-
performance of executive cognitive functions in senior spectively; it deteriorated by 0.5% in the balance and tone
women. group. Enhanced selective attention and conflict reso-
lution was significantly associated with increased gait
Methods: In this single-blinded randomized trial, 155 speed. Both resistance training groups demonstrated re-
community-dwelling women aged 65 to 75 years living ductions in whole-brain volume compared with the bal-
in Vancouver were randomly allocated to once-weekly ance and tone group at the end of the study (P ⱕ.03).
(n = 54) or twice-weekly (n = 52) resistance training or
twice-weekly balance and tone training (control group)
Conclusion: Twelve months of once-weekly or twice-
(n = 49). The primary outcome measure was perfor-
weekly resistance training benefited the executive cog-
mance on the Stroop test, an executive cognitive test of
selective attention and conflict resolution. Secondary out- nitive function of selective attention and conflict reso-
comes of executive cognitive functions included set shift- lution among senior women.
ing as measured by the Trail Making Tests (parts A and
B) and working memory as assessed by verbal digit span Trial Registration: clinicaltrials.gov Identifier:
forward and backward tests. Gait speed, muscular func- NCT00426881
tion, and whole-brain volume were also secondary out-
come measures. Arch Intern Med. 2010;170(2):170-178
C
OGNITIVE DECLINE AMONG tion studies have shown that aerobic
persons 65 years or older exercise training enhances brain and
(hereinafter seniors) is a cognitive function.4 Whether resistance
pressing health care is- training has similar benefits on cognitive
sue. Effective pharmaco- function in seniors has received little
logic treatment of mild cognitive impair- investigation.5
Author Affiliations: Brain
ment and dementia remains a major We had 3 reasons to examine whether
Research Centre
(Drs Liu-Ambrose, Graf, medical challenge.1 Hence, effective pri- resistance training improves cognitive
Beattie, and Handy), Centre for mary prevention strategies for cognitive de- function in seniors. First, a meta-analysis6
Hip Health and Mobility cline would greatly benefit individuals and highlighted that the greatest benefit of
(Drs Liu-Ambrose and Ashe), society. aerobic exercise on cognition occurred
Vancouver Coastal Health when it was paired with resistance train-
Research Institute, Departments ing. There are plausible biological mecha-
of Physical Therapy See also pages 124, 179, nisms whereby resistance training might
(Dr Liu-Ambrose), Psychology 186, and 194 ameliorate cognitive function indepen-
(Drs Graf and Handy and dently of aerobic exercise. 5 Second, a
Ms Nagamatsu), Family
Observational studies suggest that 6-month trial7 indicated that resistance
Practice (Dr Ashe), and
Division of Geriatric Medicine, physical activity may limit age-associated training benefited memory performance
Faculty of Medicine cognitive decline. 2,3 However, those and verbal concept formation among
(Dr Beattie), University of studies did not distinguish between the seniors. This raised the possibility that a
British Columbia, 2 main types of physical activity— broader spectrum of cognitive functions
Vancouver, Canada. aerobic and resistance training. Interven- may also show improvement with resis-
170

tance training. Third, to our knowledge, no study to PRIMARY
date has examined the minimum frequency of resistance OUTCOME MEASURE
training (ie, once weekly or twice weekly) required for
cognitive benefits. However, the frequency of training This study focused on the following 3 executive cognitive func-
may influence long-term exercise adherence. If a rela- tions: selective attention and conflict resolution, set shifting,
tively standard resistance training program had cogni- and working memory. Our primary outcome measure was the
specific executive cognitive function of selective attention and
tive benefits and there was evidence of a minimally
conflict resolution, as measured by the Stroop test.13 We pre-
effective dose (frequency and duration), this would add viously demonstrated that this function responds to exercise
substantially to physicians’ options of exercise prescrip- training14 and used those observed changes in our sample size
tion for seniors. calculation.
We compared the effect of once-weekly and twice- For the Stroop test, we used 3 conditions. First, partici-
weekly resistance training with twice-weekly balance and pants were instructed to read out words printed in black ink
tone exercise training on the performance of executive (eg, blue). Second, they were instructed to read out the color
cognitive functions in senior women. We focused on ex- of colored x’s. Finally, they were shown a page with color
ecutive cognitive functions because they are highly as- words printed in incongruent colored inks (eg, the word
sociated with the ability to perform instrumental activi- blue printed in red ink). Participants were asked to name the
ink color in which the words are printed (while ignoring
ties of daily living8 and mobility.9
the word itself ). There were 80 trials for each condition, and
we recorded the time participants took to read each condi-
METHODS tion. The ability to selectively attend and control response
output was calculated as the time difference between the
STUDY DESIGN third condition and the second condition. Smaller time dif-
ferences indicate better selective attention and conflict
We conducted a randomized, controlled 52-week prospective resolution.
study of exercise from May 1, 2007, through April 30, 2008,
with 3 measurement periods (baseline, midpoint, and trial SECONDARY
completion). The assessors were blinded to the participants’ as- OUTCOME MEASURES
signments. However, the success of blinding was not formally
assessed throughout the trial. Secondary measures of executive cognitive functions were set
shifting and working memory. Also, to understand the wider
PARTICIPANTS range of effects resistance training may have on senior women,
we assessed gait speed, quadriceps muscular function, and
The sample consisted solely of women because cognitive whole-brain volume.
response to exercise differs between the sexes.6 From January
31 to April 30, 2007, we recruited participants with the use of Set Shifting
print advertisements and television features. Individuals
underwent screening by a standardized telephone interview. We used the Trail Making Tests parts A and B to assess set
Women who lived in Vancouver were eligible for study entry shifting.15 Part A assesses psychomotor speed and requires
if they (1) were aged 65 to 75 years; (2) were living indepen- the participant to draw lines that connect encircled numbers
dently in their own home; (3) scored 24 or more on the Mini- sequentially, such as drawing a line from 1 to 2, 2 to 3, and
Mental State Examination; and (4) had a visual acuity of at 3 to 4. Part B consists of encircled numbers and letters. Par-
least 20/40, with or without corrective lenses. We excluded ticipants were instructed to draw a line as quickly and as
those who (1) had a current medical condition for which accurately as possible from 1 to A, A to 2, 2 to B, B to 3, and
exercise is contraindicated, (2) had participated in resistance so on, until they completed the task. We recorded the
training in the past 6 months, (3) had a neurodegenerative amount of time (in seconds) it took to complete each task.
disease and/or a stroke, (4) had depression, (5) did not speak To index set shifting, we calculated the difference between
and understand English fluently, (6) were taking cholinester- part B and part A completion times. Smaller difference
ase inhibitors, (7) were receiving estrogen therapy, or (8) scores indicate better set shifting.
were receiving testosterone therapy.
The CONSORT (Consolidated Standards of Reporting Trials) Working Memory
flowchart in the Figure shows the number of participants in
the treatment arms at each stage of the study. Ethical approval We used the verbal digit span forward and backward tests to
was obtained from the Vancouver Coastal Health Research In- index the central executive component of working
stitute and the University of British Columbia’s Clinical Re- memory.16 Both tests consist of 7 pairs of random number
search Ethics Board. All participants provided written in- sequences that the assessor reads aloud at the rate of 1 per
formed consent. second. The sequence begins with 3 digits and increases by 1
at a time up to a length of 9 digits. The test includes 2
DESCRIPTIVE VARIABLES sequences of each length and testing ceases when the partici-
pant fails to recollect any 2 with the same length. The score
At baseline, participants underwent a physician assessment to recorded, ranging from 0 to 14, is the number of successful
confirm current health status and eligibility for the study. We sequences. For the verbal digit span forward test, the partici-
used the 15-item Geriatric Depression Scale10 to screen for de- pant’s task is to repeat each sequence exactly as it is given.
pression. Current level of physical activity was determined by For the verbal digit span backward test, the participant’s task
the Physical Activity Scale for the Elderly self-report question- is to repeat each sequence in reverse order. The difference
naire.11 General mobility was assessed by the timed Up and Go between the verbal digit span forward and backward test
test.12 scores was used as an index of the central executive compo-
171

262 Recruitment letters mailed Newspaper, television, flyer,
and e-mail advertisements sent
Telephone screen
346 Screened
41 Excluded
87 Eligible based on telephone
screen but no longer interested
Information session
218 Attended
36 Not interested
7 Excluded
1 Had terminal cancer
1 Taking antidepressants
5 Receiving estrogen, testosterone,
or growth hormone therapy
175 Consented and booked for

baseline assessment
15 Withdrew
1 Lost contact
14 Withdrew owing to time
commitment
Baseline assessment
160 Attended
1 Withdrew during assessment
1 Excluded because psychiatric
condition suspected by physician
158 Completed assessment
3 Withdrew
2 Owing to time commitment
1 Medical reason
155 Randomized
52 Received 2× RT 54 Received 1× RT 49 Received BAT 2×/wk
1 Withdrew owing to time 5 Withdrew 5 Withdrew

commitment 2 Owing to time commitment 4 For medical reasons
1 For medical reasons 1 For personal reasons
2 For personal reasons
51 Participated in midpoint assessment 49 Participated in midpoint assessment 44 Participated in midpoint assessment
5 Withdrew 2 Withdrew for medical reasons 2 Withdrew

1 Owing to time commitment 1 Owing to time commitment
2 For medical reasons 1 For medical reasons
1 For personal reasons
1 For unknown reasons
46 Participated in final assessment 47 Participated in final assessment 42 Participated in final assessment
135 Completed trial and included

in analysis
Figure. The CONSORT (Consolidated Standards of Reporting Trials) flowchart. BAT indicates twice-weekly balance and tone exercise training;
1⫻ RT, once-weekly resistance training; and 2⫻ RT, twice-weekly RT.
nent of working memory. Smaller difference scores indicate Muscular Function

better working memory.
In a subset of participants who were eligible (ie, those with no
Gait Speed significant preexisting knee, hip, or back condition), isotonic
Gait speed is a significant and independent predictor of falls quadriceps strength (single-repetition maximum lift [1-RM])
and fracture risk in older women.17 Participants were asked to and peak muscle power were assessed using an air-pressured
walk at their usual pace along a 4-m path. Gait speed (in me- digital resistance leg press machine (Keiser Sports Health Equip-
ters per second) was calculated from the mean of 2 trials. The ment, Fresno, California). The study physician screened all par-
test-retest reliability (intraclass correlation coefficient) of gait ticipants for eligibility. Two assessors completed all assess-
speed in our laboratory is 0.95.18 ments of 1-RM and peak muscle power; they attended two
172

30-minute training sessions before the baseline measurement using an ␣ level of less than .05, 52 participants per group
period. ensured a power of 0.80.
The initial load for quadriceps 1-RM assessment was the par-
ticipant’s own body mass. Participants pushed against the leg EXERCISE INTERVENTION
press over a 3-second count and then returned to the start po-
sition over a 3-second count. Load increased by 10% incre- The resistance training and balance and tone classes both be-
ments until participants were no longer able to lift the load gan 1 month after the baseline assessments were completed (ie,
through their available range of motion. The load (in new- May 2007). Classes were held at 2 locations: the local YMCA
tons) of the last successfully completed leg press was recorded and the Centre for Hip Health and Mobility research center.
and used for statistical analysis. All classes were led by certified fitness instructors who re-
After the completion of the quadriceps 1-RM testing, eli- ceived additional training and education from the study inves-
gible participants were given a 15-minute break. They then un- tigators. The classes were 60 minutes long, with a 10-minute
derwent quadriceps muscle power assessments during which warm-up, 40 minutes of core content, and a 10-minute
they completed leg press extensions at 6 relative loads of their cooldown. To ensure that programs were delivered faithfully
1-RM (ie, 40%, 50%, 60%, 70%, 80%, and 90%). Participants and consistently across sites, a research assistant who was not
performed the concentric portion of the leg press repetition as involved in delivering the study’s classes conducted quality as-
rapidly as possible and then slowly lowered the load during a sessments every month with the use of a standard form. At-
3-second count. Beginning at 40% of 1-RM, participants per- tendance was recorded daily by the assistants. Adherence, ex-
formed 3 repetitions at each relative 1-RM load. There was a pressed as the percentage of the total number of classes attended,
30-second rest between repetitions. The air-pressured digital was calculated from these attendance sheets.
resistance leg press machine recorded the power produced (in Specific strategies were implemented to promote partici-
watts). The peak quadriceps muscle power obtained by each pant engagement. These included (1) semimonthly newslet-
participant was used for statistical analysis. ters that featured personal accomplishments of the partici-
pants, healthy recipes contributed by the participants, and study
Whole-Brain Volume updates; (2) 3 social events (eg, the Winter Holiday Tea); (3)
personalized birthday cards; (4) follow-up for participants who
For those who met the inclusion criteria for magnetic reso- missed 2 consecutive classes without a reason; and (5) sup-
nance imaging and consented, whole-brain volume was mea- port and suggestions for overcoming barriers to participation.
sured via T1-weighted structural magnetic resonance images
obtained using a 3T scanner (Achieva; Philips Medical Sys- Resistance Training
tems UK Ltd, Surrey, England). Calculations of whole-brain vol-
ume and their percentage change across study time points were The resistance training program used a progressive, high-
made using the SIENA (Structural Image Evaluation, using Nor- intensity protocol. The air-pressured digital resistance leg press
malization, of Atrophy) method of automated analysis.19 SIENA machine and free weights were used to provide the training
is a longitudinal method that compares pairs of scans within stimulus. The leg press machine–based exercises consisted of
subjects and is available as part of the Functional Magnetic Reso- biceps curls, triceps extension, seated rowing, latissimus dorsi
nance Imaging of the Brain Software Library package.20 SIENA pull-down exercises, leg presses, hamstring curls, and calf raises.
has been shown to have an overall error rate of approximately The intensity of the training stimulus was at a work range of 6
0.2% of the absolute brain volume.19,21,22 It is designed to be to 8 repetitions (2 sets). The training stimulus was subse-
fully automatic, but careful evaluation of its intermediate out- quently increased using the 7-RM method, when 2 sets of 6 to
put is essential to ensuring accurate results. To minimize er- 8 repetitions were completed with proper form and without dis-
ror, we performed visual checks of intermediate output from comfort. Other key strength exercises included minisquats, mini-
3 critical processes: brain extraction, spatial alignment, and tis- lunges, and lunge walks. The number of sets completed and
sue segmentation. the load lifted for each exercise were recorded for each partici-
pant at every class.
RANDOMIZATION
Balance and Tone
The randomization sequence was generated by http://www
.randomization.com and was concealed until interventions were The balance and tone program consisted of stretching exer-
assigned. This sequence was held independently and remotely cises, range-of-motion exercises, basic core-strength exercises
by the research coordinator. Participants were enrolled and ran- including kegels (ie, exercises to strengthen the pelvic floor
domized by the research coordinator to one of the following 3 muscles), balance exercises, and relaxation techniques. Key bal-
groups: once-weekly resistance training (1⫻ RT), twice- ance exercises included tai chi–based forms (ie, the crane and
weekly resistance training (2⫻RT), or twice-weekly balance the tree pose), tandem stand, tandem walking, and single leg
and tone (BAT). stance (eyes opened and closed). Other than body weight, no
additional loading (eg, hand weights or resistance bands) was
SAMPLE SIZE applied to any of the exercises. There is no evidence that these
exercises improve cognitive function.4 This group served to con-
The required sample size for this study was calculated from trol for confounding variables such as physical training re-
predictions of 12-month changes in the Stroop test results. ceived by traveling to the training centers, social interaction,
Specifically, we predicted a 6% improvement for the 1⫻ RT and changes in lifestyle secondary to study participation.
and a 12% improvement for the 2⫻RT groups. We also esti-
mated a 10% deterioration in the BAT group (ie, the control ADVERSE EFFECTS
group). These estimates were based on our previous work,14
which demonstrated that a home-based program of strength Participants were questioned about the presence of any ad-
and balance retraining exercises significantly improved verse effects, such as musculoskeletal pain or discomfort, at each
Stroop test performance. Assuming a 20% attrition rate and exercise session. All instructors also monitored participants for
173

Table 1. Baseline Characteristics of the 155 Trial Participants a
Group
BAT 1ⴛ RT 2ⴛ RT All
Variable (n = 49) (n = 54) (n = 52) (N=155)
Age, y 70.0 (3.3) 69.5 (2.7) 69.4 (3.0) 69.6 (2.9)
Height, cm 161.0 (6.9) 160.9 (7.0) 162.8 (6.5) 161.6 (6.8)
Weight, kg 67.0 (11.5) 69.2 (16.2) 72.1 (16.8) 69.5 (15.2)
Education, No. (%)
No high school 1 (2.0) 1 (1.9) 1 (1.9) 3 (1.9)
Grades 9-12 without certificate or diploma 2 (4.1) 3 (5.6) 4 (7.7) 9 (5.8)
High school certificate or diploma 6 (12.2) 9 (16.7) 10 (19.2) 25 (16.1)
Trade or professional certificate or diploma 14 (28.6) 10 (18.5) 6 (11.5) 30 (19.4)
University certificate or diploma 7 (14.3) 12 (22.2) 9 (17.3) 28 (18.1)
University degree 19 (38.8) 19 (35.2) 22 (42.3) 60 (38.7)
MMSE score b 28.8 (1.2) 28.5 (1.3) 28.6 (1.5) 28.6 (1.3)
Falls in the last 12 months, No. (%) 16 (32.7) 13 (24.1) 20 (38.5) 49 (31.6)
Geriatric Depression Scale c 0.5 (1.8) 0.3 (1.1) 0.9 (2.3) 0.6 (1.8)
Functional Comorbidity Index d 2.2 (1.7) 1.8 (1.7) 2.3 (1.6) 2.1 (1.7)
Lawton and Brody Instrumental Activities of Daily Living Scale score e 8.0 (0) 8.0 (0.1) 7.9 (0.5) 8.0 (0.3)
PASE score 126.1 (51.0) 116.2 (61.4) 121.2 (60.4) 121.0 (57.7)
TUG, s 6.8 (1.4) 6.6 (1.4) 6.6 (1.4) 6.6 (1.4)
Abbreviations: BAT, balance and toning; MMSE, Mini-Mental State Examination; PASE, Physical Activity Scale for the Elderly; TUG, timed Up and Go test;
1⫻ RT, once-weekly resistance training; 2⫻ RT, twice-weekly RT.
a Unless otherwise indicated, data are expressed as mean (SD). Percentages have been rounded and may not total 100.
b Maximum was 30 points.
c Maximum was 15 points.
d Maximum was 18 points.
e Maximum was 8 points. The scale is described in Lawton and Brody.29
symptoms of angina and shortness of breath during the exer- beginning and the end of the intervention period were
cise classes. related to changes in gait speed.
STATISTICAL ANALYSIS RESULTS

All analyses were full analysis set,23 defined as the analysis set
that is as complete and as close as possible to the intention-to- DESCRIPTIVE VARIABLES, EXERCISE
treat ideal of including all randomized participants. Data were ADHERENCE, AND PHYSICAL ACTIVITY LEVELS
analyzed using SPSS statistical software (Windows version 17.0;
SPSS Inc, Chicago, Illinois). The mean (SD) age of the cohort was 69.6 (2.9) years,
Between-group differences in selective attention and con-
and the exercise adherence during the 1-year study pe-
flict resolution at the midpoint and at trial completion were com-
pared by multiple linear regression analysis. In the models, base- riod was 67.9%. Average adherence was 71.0% for the
line scores, experimental group, baseline Mini-Mental State 1⫻RT group, 70.3% for the 2⫻RT group, and 62.0% for
Examination score, baseline waist circumference,24,25 diagno- the BAT group. Baseline demographics and characteris-
sis of diabetes (yes/no),25-27 and visual edge contrast sensitiv- tics of the 155 participants who were randomized are
ity score28 were included as covariates. Two planned simple con- shown in Table 1. Physical activity levels (Physical Ac-
trasts were performed when there were significant main group tivity Scale for the Elderly scores) did not differ signifi-
effects. These contrasts were used to assess differences be- cantly between the groups at the midpoint (P =.98) or at
tween the 1⫻RT group and the BAT group and between the trial completion (P =.68).
2⫻RT group and the BAT group. In addition, difference con-
trasts were used within each RT group to assess when cogni-
tive benefits of resistance training were evident. The overall ␣ PRIMARY OUTCOME MEASURE
level was set at P⬍.05.
We analyzed our secondary outcome measures of execu- Table 2 shows the baseline, midpoint, and trial comple-
tive cognitive functions in the same manner as our primary out- tion results for the executive cognitive function tests. The
come measure with the exception that visual edge contrast sen- regression analyses revealed no significant between-
sitivity was not included as a covariate in the model for working group differences at the midpoint of the trial. However,
memory. at the end of the trial, there was a significant between-
For models of gait speed, quadriceps 1-RM, and peak
quadriceps muscle power, baseline scores and experimental
group difference in selective attention and conflict reso-
group were included as covariates. For models of percentage lution (P =.01). Planned simple contrasts indicated that
of change in whole-brain volume, presence of diabetes was both the 1⫻ RT and 2⫻RT groups had improved Stroop
included as a covariate. Finally, Pearson product moment test performance compared with the BAT group at trial
correlations were computed to determine whether changes completion (Pⱕ.03). Specifically, task performance im-
in selective attention and conflict resolution between the proved by 12.6% and 10.9% in the 1⫻ RT and 2⫻ RT
174

Table 2. Mean Values for Outcome Measures
Mean (SD)
Variable Baseline Midpoint Completion

2⫻ RT group n=52 n=51 n=46
Stroop test, CW − C, s 45.0 (15.8) 46.1 (17.2) 40.9 (14.9) a
Trail Making Test, part B − part A, s 49.5 (36.6) 41.7 (30.8) 38.7 (33.6)
Verbal digit span test, forward − backward 3.4 (2.4) 4.1 (2.9) 3.8 (2.1)
Gait speed, m/s 1.2 (0.2) 1.3 (0.2) 1.4 (0.2)
1-RM, N b 315.0 (66.9) 380.6 (84.0) 388.2 (82.4)
Peak muscle power, W c 624.9 (194.2) 707.2 (171.6) a 708.6 (161.2) a
Change in whole-brain volume from baseline, % d NA −0.02 (0.60) −0.43 (0.65) a
1⫻ RT group n=54 n=49 n= 47
Stroop test, CW − C, s 47.4 (26.2) 46.5 (25.8) 39.5 (14.1) a
Gait speed, m/s 1.2 (0.2) 1.4 (0.2) 1.4 (0.2)
1-RM, N b 323.9 (63.2) 371.0 (86.4) 386.5 (97.0)
Peak muscle power, W c 679.3 (184.2) 633.4 (192.3) 622.3 (204.4)
Change in whole-brain volume from baseline, % d NA −0.04 (0.48) −0.32 (0.54) a
BAT group n=49 n=44 n= 42
Stroop test, CW − C, s 44.0 (15.1) 49.2 (19.1) 43.8 (18.9)
Gait speed, m/s 1.2 (0.2) 1.3 (0.2) 1.4 (0.2)
1-RM, N b 338.2 (70.4) 363.9 (77.4) 356.3 (85.3)
Peak muscle power, W c 660.3 (229.7) 649.5 (208.8) 552.1 (194.0)
Change in whole-brain volume from baseline, % d NA 0.13 (0.67) 0.00 (0.63)
Abbreviations: BAT, balance and toning; C, Stroop test colored x ’s condition; CW, Stroop test color words condition; NA, not applicable; 1-RM, isotonic
quadriceps strength; 1⫻ RT, once-weekly resistance training; 2⫻ RT, twice-weekly RT.
a Significantly different from the BAT group at P ⬍ .05. For the Stroop test, the 95% confidence interval (CI) of the difference between the 1⫻ RT and BAT
groups was −13.8 to −2.5; between the 2⫻ RT and BAT groups, −12.2 to −0.8. For peak muscle power, 95% CI of the difference between the 2⫻ RT and BAT
groups at the midpoint was 22.2 to 151.3; between the 2⫻ RT and BAT groups at trial completion, 81.7 to 230.0. For the difference in whole-brain volume, the
95% CI of the difference between the 1⫻ RT and BAT groups was −0.76 to −0.04; between the 2⫻ RT and BAT groups, −0.89 to −0.12.
b For this analysis, 31 participants were included in the 2⫻ RT group at baseline, 26 at the midpoint, and 25 at trial completion; in the 1⫻ RT group, 30 at
baseline, 28 at the midpoint, and 27 at trial completion; and in the BAT group, 27 at baseline, 21 at the midpoint, and 24 at trial completion.
c For this analysis, 30 participants were included in the 2⫻ RT group at baseline, 23 at the midpoint, and 25 at trial completion; in the 1⫻ RT group, 29 at
baseline, 26 at the midpoint, and 27 at trial completion; and in the BAT group, 27 at baseline, 21 at the midpoint, and 24 at trial completion.
d For the 2⫻ RT group, 18 participants were included in the differences from baseline to the midpoint and from baseline to trial completion; for the 1⫻ RT
group, 28 in the differences from baseline to the midpoint and from baseline to trial completion; for the BAT group, 20 in the difference from baseline to the
midpoint and 18 from baseline to trial completion.
groups, respectively, whereas the BAT group demon- There were also between-group differences in the per-
strated a 0.5% deterioration. Within each RT group, dif- centage of change of whole-brain volume at trial comple-
ference contrasts demonstrated that Stroop test perfor- tion (P ⱕ.03) (Table 3). At the end of the study, the
mance was not significantly different from baseline to the 1⫻ RT and 2⫻ RT groups both demonstrated reduc-
midpoint (P=.79), but was significantly different from tions in whole-brain volume compared with the BAT
the midpoint to trial completion (P = .001). group (P ⱕ .03). Specifically, there was a 0.32% and a
0.43% reduction in whole-brain volume for the 1⫻RT
SECONDARY OUTCOME MEASURES and 2⫻RT groups, respectively. In contrast, there was
no change in whole-brain volume for the BAT group.
The regression analyses revealed no significant between- Improvement in selective attention and conflict reso-
group differences at the midpoint and at trial comple- lution during the 12-month intervention was signifi-
tion in set shifting and working memory (Table 2). cantly associated with improvement in gait speed (r=0.24;
There were no significant between-group differences P ⬍.01).
at the midpoint and at trial completion in gait speed and
quadriceps 1-RM. However, there were significant between- ADVERSE EVENTS
group differences in peak muscle power at the midpoint
(P⬍.01) and at trial completion (P⬍.001). Planned simple Results of the ␹2 test indicated significant group differ-
contrasts indicated that the 2⫻RT group increased peak ences (P=.02) in the proportion of participants report-
muscle power at the midpoint (P⬍.01) and at trial comple- ing adverse events. Specifically, musculoskeletal com-
tion (P⬍.001) compared with the BAT group. Specifi- plaints (eg, knee joint discomfort or bursa irritation in
cally, at trial completion, peak muscle power increased by the lateral hip) developed in 14 of 47 women (29.8%) in
13.4% in the 2⫻RT group but decreased by 8.4% and 16.3% the 1⫻RT group, in 5 of 46 (10.9%) in the 2⫻RT group,
for the 1⫻RT and BAT groups, respectively. and in 4 of 42 (9.5%) in the BAT group. All documented
175

that engaging in progressive resistance training as infre-
Table 3. Mean Change for Outcome Measures quently as once a week can significantly benefit execu-
tive cognitive function in community-dwelling senior
Mean (SD) Change a women.
Variable Midpoint Completion Cassilhas et al7 demonstrated that 6 months of thrice-
2⫻ RT group n=51 n=46
weekly moderate- or high-intensity resistance training im-
Stroop test, CW − C, s −0.96 (15.13) 5.01 (13.75) proved cognitive performance of memory and verbal con-
Trail Making Test, part B − part A, s 10.27 (40.25) 10.96 (36.92) cept formation among senior men. Our findings extend
Verbal digit span test, −0.67 (2.94) −0.47 (2.24) these results in several critical ways. Most notably, our
forward − backward results suggest that the effects of resistance training on
Gait speed, m/s 0.19 (0.17) 0.24 (0.16) executive cognitive functions appear to be selective; that
1-RM, N b 60.27 (51.18) 69.80 (74.75)
Peak muscle power, W c 74.68 (118.55) 72.42 (108.12)
is, resistance training enhanced selective attention and
Change in whole-brain volume, % d −0.02 (0.60) −0.43 (0.65) conflict resolution in older women, but cognitive abili-
1⫻ RT group n=49 n=47 ties associated with manipulating verbal information in
Stroop test, CW − C, s 0.28 (28.37) 6.22 (22.31) working memory and shifting between task sets or in-
Trail Making Test, part B − part A, s 4.91 (26.12) 7.3 (30.36) structions were not improved.
Verbal digit span test, −0.43 (2.63) 0.06 (2.54) Our study provides novel data relating the frequency
forward − backward
Gait speed, m/s 0.18 (0.19) 0.19 (0.19)
and duration of resistance training with cognitive ben-
1-RM, N b 42.15 (57.26) 44.22 (67.10) efits in women. We observed a cognitive benefit after 12
Peak muscle power, W c −27.54 (105.66) −78.61 (151.03) months of training but not at the 6-month trial mid-
Change in whole-brain volume, % d −0.04 (0.48) −0.32 (0.54) point. Cassilhas et al7 reported cognitive benefits after 6
BAT group n=44 n=42 months of resistance training in men. There were differ-
Stroop test, CW − C, s −4.27 (15.15) 0.26 (17.12) ences in the frequency of resistance training between the
Trail Making Test, part B − part A, s 2.17 (39.27) 8.64 (32.15)
2 studies (ie, once-weekly and twice-weekly training in
Verbal digit span test, −0.93 (3.42) −0.64 (2.70)
forward − backward our study vs thrice-weekly training in the Brazilian study);
Gait speed, m/s 0.17 (0.16) 0.22 (0.18) also, different cognitive functions may have different
1-RM, N b 24.73 (53.44) 18.15 (70.06) change trajectories with resistance training. Sex may also
Peak muscle power, W c −24.27 (132.56) −90.60 (144.58) be a moderating factor. Our study included women only,
Change in whole-brain volume, % d 0.13 (0.67) 0.00 (0.63) and the participants trained less frequently than did those
in the study by Cassilhas et al.7 Finally, differences in the
Abbreviations: See Table 2.
a Mean change for all cognitive measures is calculated as the difference control groups may have contributed to the lack of be-
between the baseline value and the midpoint or the completion value. tween-group differences in cognitive performance at 6
Positive change indicates improvement. Mean change for all performance months. The control group in the Brazilian study7 trained
measures is calculated as the difference between the midpoint and baseline
values or between the completion and baseline values. Positive change
only once weekly; our Canadian control group trained
indicates improvement. twice weekly.
b For this analysis, 31 participants were included in the 2⫻ RT group at We also demonstrated that enhanced selective atten-
baseline, 26 at the midpoint, and 25 at trial completion; in the 1⫻ RT group, tion and conflict resolution was associated with in-
30 at baseline, 28 at the midpoint, and 27 at trial completion; and in the BAT
group, 27 at baseline, 21 at the midpoint, and 24 at trial completion. creased gait speed. To our knowledge, this study is the
c For this analysis, 30 participants were included in the 2⫻ RT group at first to demonstrate this relationship. Our finding adds
baseline, 23 at the midpoint, and 25 at trial completion; in the 1⫻ RT group, weight to previous observations of a strong relationship
29 at baseline, 26 at the midpoint, and 27 at trial completion; and in the BAT
group, 27 at baseline, 21 at the midpoint, and 24 at trial completion.
between gait speed and cognitive function.30 The impli-
d For the 2⫻ RT group, 18 participants were included in the differences from cation for clinicians is that improved gait speed is a pre-
baseline to the midpoint and from baseline to trial completion; for the 1⫻ RT dictor of substantial reduction in mortality.31
group, 28 in the differences from baseline to the midpoint and from baseline to The design of our control group (ie, balance and ton-
trial completion; and for the BAT group, 20 in the difference from baseline to the
midpoint and 18 from baseline to trial completion. ing) may have also contributed to the lack of between-
group differences at 6 and 12 months in quadriceps 1-RM.
Our control group included balance training in their twice-
musculoskeletal symptoms resolved or diminished within
weekly program. Previous studies have demonstrated that
4 weeks of onset. There was also 1 fall in the BAT group;
balance training exercises can improve muscle strength.32,33
this fall did not result in injury.
In addition, in our own previous investigation of differ-
ent types of exercise training (ie, resistance training, agil-
COMMENT ity training, and stretching [ie, control] exercises) in se-
nior women with low bone mass, we did not find any
In community-dwelling women aged 65 to 75 years, 12 significant between-group differences in measures of quad-
months of progressive resistance training once or twice riceps strength and mobility.34
weekly improved selective attention and conflict resolu- We highlight that, although both resistance training
tion relative to twice-weekly balance and toning exer- groups enhanced selective attention and conflict resolu-
cises. We also found that resistance training twice tion by the end of the trial, there were more musculo-
weekly improved peak quadriceps muscle power, and skeletal adverse events in the 1⫻ RT group than in the
that resistance training once or twice weekly led to 2⫻RT and BAT groups. Hence, the possible increased
small but significant reductions in whole-brain volume. risk for musculoskeletal injury with once-weekly resis-
To our knowledge, this is the first study to demonstrate tance training must be weighed against its benefit of re-
176

duced training time compared with twice-weekly resis- Funding/Support: This study was supported by operat-
tance training. ing grants from The Vancouver Foundation (BCM06-
An unexpected result was the reduced whole-brain vol- 0035, Dr Liu-Ambrose) and the Natural Sciences and En-
ume for both RT groups. Although reduced brain vol- gineering Research Council of Canada (Dr Handy), by
umes are commonly associated with impaired func- an establishment grant from the Michael Smith Foun-
tion,35 the groups who improved cognitive executive dation for Health Research (MSFHR) (CI-SCH-063[05-
function and muscular function in our study also had brain 0035], Dr Liu-Ambrose), and by a New Opportunities
volume reductions. There are precedents that parallel our Fund from the Canada Foundation for Innovation for es-
apparently paradoxical finding.36,37 In a ␤-amyloid im- sential infrastructure used in this study (Dr Liu-
munization trial among those with probable Alzheimer Ambrose). Drs Liu-Ambrose, Ashe, and Handy are MSFHR
disease, immunization led to significant clinical benefit, Scholars.
reduced ␤-amyloid load, and reduced brain volume.36 The Previous Presentations: This study was presented at the
investigators hypothesized that removal of ␤-amyloid 2009 Annual Meeting of the American College of Sports
and/or other protein constituents from brain tissue may Medicine; May 28, 2009; Seattle, Washington; and at the
have caused cerebral fluid shifts, resulting in brain vol- 19th International Association of Gerontology and Geri-
ume reductions on magnetic resonance imaging. How- atrics World Congress; July 7, 2009; Paris, France.
ever, we are very cautious in our interpretation of the Additional Contributions: The Vancouver South
whole-brain volume results and emphasize that this facet Slope YMCA management and members enthusiasti-
of the study, although not the first report of such a phe- cally supported the study by allowing access to par-
nomenon, needs further investigation. ticipants for the training intervention. Lindsay
Because our participant sample included women aged Katarynych, BSc, coordinated this study. We thank the
65 to 75 years only, the findings may not generalize to instructors for their commitment to the participants’
men or to women of other ages. Also, although we ob- health and safety.
served reduced whole-brain volumes, the study was not
designed to show which specific brain regions demon-
strated volumetric changes. REFERENCES
In conclusion, we provide novel randomized con-
1. Raschetti R, Albanese E, Vanacore N, Maggini M. Cholinesterase inhibitors in
trolled trial evidence that a pragmatic resistance train- mild cognitive impairment: a systematic review of randomised trials. PLoS Med.
ing program can enhance the executive cognitive func- 2007;4(11):e338.
tion of selective attention and conflict resolution and 2. Weuve J, Kang JH, Manson JE, Breteler MM, Ware JH, Grodstein F. Physical ac-
simultaneously improve muscular function in senior tivity, including walking, and cognitive function in older women. JAMA. 2004;
women. This has important clinical implications be- 292(12):1454-1461.
3. Abbott RD, White LR, Ross GW, Masaki KH, Curb JD, Petrovitch H. Walking and
cause cognitive impairment is a major health problem that dementia in physically capable elderly men. JAMA. 2004;292(12):1447-1453.
currently lacks a clearly effective pharmaceutical therapy 4. Colcombe SJ, Kramer AF, Erickson KI, et al. Cardiovascular fitness, cortical plas-
and because resistance training is not widely adopted by ticity, and aging. Proc Natl Acad Sci U S A. 2004;101(9):3316-3321.
seniors. The doses of resistance training we used in this 5. Liu-Ambrose T, Donaldson MG. Exercise and cognition in older adults: is there
a role for resistance training programmes? Br J Sports Med. 2009;43(1):
study fall within those recommended by the 2008 Physi- 25-27.
cal Activity Guidelines for seniors.38 6. Colcombe S, Kramer AF. Fitness effects on the cognitive function of older adults:
a meta-analytic study. Psychol Sci. 2003;14(2):125-130.
Accepted for Publication: October 7, 2009. 7. Cassilhas RC, Viana VA, Grassmann V, et al. The impact of resistance exercise
Correspondence: Teresa Liu-Ambrose, PhD, PT, Cen- on the cognitive function of the elderly. Med Sci Sports Exerc. 2007;39(8):
1401-1407.
tre for Hip Health and Mobility, Vancouver Coastal Re- 8. Royall DR, Palmer R, Chiodo LK, Polk MJ. Declining executive control in normal
search Institute, Department of Physical Therapy, Uni- aging predicts change in functional status: the Freedom House Study. J Am Geri-
versity of British Columbia. 357-2647 Willow St, atr Soc. 2004;52(3):346-352.
Vancouver, BC V5Z 3P1, Canada (tlambrose@exchange 9. van Iersel MB, Kessels RP, Bloem BR, Verbeek AL, Olde Rikkert MG. Executive
functions are associated with gait and balance in community-living elderly people.
.ubc.ca). J Gerontol A Biol Sci Med Sci. 2008;63(12):1344-1349.
Author Contributions: Dr Liu-Ambrose had full access 10. Yesavage JA. Geriatric Depression Scale. Psychopharmacol Bull. 1988;24(4):709-
to all the data in the study and takes responsibility for 711.
the integrity of the data and the accuracy of the data analy- 11. Washburn RA, McAuley E, Katula J, Mihalko SL, Boileau RA. The Physical Ac-
sis. Study concept and design: Liu-Ambrose, Nagamatsu, tivity Scale for the Elderly (PASE): evidence for validity. J Clin Epidemiol. 1999;
52(7):643-651.
Graf, Beattie, and Handy. Acquisition of data: Liu- 12. Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basic functional mo-
Ambrose, Nagamatsu, Ashe, and Handy. Analysis and in- bility for frail elderly persons. J Am Geriatr Soc. 1991;39(2):142-148.
terpretation of data: Liu-Ambrose, Nagamatsu, Graf, 13. Graf P, Uttl B, Tuokko H. Color- and picture-word Stroop tests: performance changes
Beattie, and Handy. Drafting of the manuscript: Liu- in old age. J Clin Exp Neuropsychol. 1995;17(3):390-415.
14. Liu-Ambrose T, Donaldson MG, Ahamed Y, et al. Otago home-based strength
Ambrose, Nagamatsu, Graf, and Handy. Critical revi- and balance retraining improves executive functioning in older fallers: a random-
sion of the manuscript for important intellectual content: Liu- ized controlled trial. J Am Geriatr Soc. 2008;56(10):1821-1830.
Ambrose, Nagamatsu, Beattie, Ashe, and Handy. Statistical 15. Spreen O, Strauss E. A Compendium of Neurological Tests. 2nd ed. New York,
analysis: Liu-Ambrose and Handy. Obtained funding: Liu- NY: Oxford University Press Inc; 1998.
Ambrose and Graf. Administrative, technical, and mate- 16. Wechsler D. Wechsler Adult Intelligence Scale—Revised. San Antonio, TX: Psy-
chological Corp, Harcourt Brace Jovanovich; 1981.
rial support: Liu-Ambrose, Nagamatsu, Beattie, Ashe, and 17. Dargent-Molina P, Favier F, Grandjean H, et al. Fall-related factors and risk of hip
Handy. Study supervision: Liu-Ambrose and Handy. fracture: the EPIDOS prospective study. Lancet. 1996;348(9021):145-149.
Financial Disclosure: None reported. 18. Liu-Ambrose T, Khan KM, Donaldson MG, Eng JJ, Lord SR, McKay HA. Falls-
177

related self-efficacy is independently associated with balance and mobility in older impairment and cognitive functioning in oldest-old subjects: the Leiden 85⫹ Study.
women with low bone mass. J Gerontol A Biol Sci Med Sci. 2006;61(8):832- Am J Geriatr Psychiatry. 2005;13(9):781-786.
838. 29. Lawton MP, Brody EM. Assessment of older people: self-maintaining and in-
19. Smith SM, Zhang Y, Jenkinson M, et al. Accurate, robust, and automated lon- strumental activities of daily living. Gerontologist. 1969;9(3):179-186.
gitudinal and cross-sectional brain change analysis. Neuroimage. 2002;17 30. Soumaré A, Tavernier B, Alperovitch A, Tzourio C, Elbaz A. A cross-sectional and
(1):479-489. longitudinal study of the relationship between walking speed and cognitive func-
20. Smith SM, Jenkinson M, Woolrich MW, et al. Advances in functional and struc- tion in community-dwelling elderly people. J Gerontol A Biol Sci Med Sci. 2009;
tural MR image analysis and implementation as FSL. Neuroimage. 2004;23 64(10):1058-1065.
(suppl 1):S208-S219. 31. Hardy SE, Perera S, Roumani YF, Chandler JM, Studenski SA. Improvement in
21. Smith SM, Rao A, De Stefano N, et al. Longitudinal and cross-sectional analysis usual gait speed predicts better survival in older adults. J Am Geriatr Soc. 2007;
of atrophy in Alzheimer’s disease: cross-validation of BSI, SIENA and SIENAX. 55(11):1727-1734.
Neuroimage. 2007;36(4):1200-1206. 32. Li JX, Xu DQ, Hong Y. Changes in muscle strength, endurance, and reaction of the
22. Camara O, Schnabel JA, Ridgway GR, et al. Accuracy assessment of global and lower extremities with tai chi intervention. J Biomech. 2009;42(8):967-971.
local atrophy measurement techniques with realistic simulated longitudinal Alz- 33. Xu DQ, Hong Y, Li JX. Tai chi exercise and muscle strength and endurance in
heimer’s disease images. Neuroimage. 2008;42(2):696-709. older people. Med Sport Sci. 2008;52:20-29.
23. International Conference on Harmonisation E9 Expert Working Group. ICH Har- 34. Liu-Ambrose T, Khan KM, Eng JJ, Janssen PA, Lord SR, McKay HA. Resistance
monised Tripartite Guideline: statistical principals in clinical trials. Stat Med. 1999; and agility training reduce fall risk in women aged 75 to 85 with low bone mass:
18(15):1905-1942. a 6-month randomized, controlled trial. J Am Geriatr Soc. 2004;52(5):657-
24. Wolf PA, Beiser A, Elias MF, Au R, Vasan RS, Seshadri S. Relation of obesity to 665.
cognitive function: importance of central obesity and synergistic influence of con- 35. Carlson NE, Moore MM, Dame A, et al. Trajectories of brain loss in aging and the
comitant hypertension: the Framingham Heart Study. Curr Alzheimer Res. 2007; development of cognitive impairment. Neurology. 2008;70(11):828-833.
4(2):111-116. 36. Fox NC, Black RS, Gilman S, et al; AN1792(QS-21)-201 Study. Effects of A␤ im-
25. Yaffe K, Kanaya A, Lindquist K, et al. The metabolic syndrome, inflammation, munization (AN1792) on MRI measures of cerebral volume in Alzheimer disease.
and risk of cognitive decline. JAMA. 2004;292(18):2237-2242. Neurology. 2005;64(9):1563-1572.
26. S Roriz-Filho J, Sá-Roriz TM, Rosset I, et al. (Pre)diabetes, brain aging, and 37. Sparks DL, Lemieux SK, Haut MW, et al. Hippocampal volume change in the Alz-
cognition. Biochim Biophys Acta. 2009;1792(5):432-443. heimer Disease Cholesterol-Lowering Treatment trial. Cleve Clin J Med. 2008;
27. Yaffe K, Weston AL, Blackwell T, Krueger KA. The metabolic syndrome and de- 75(suppl 2):S87-S93.
velopment of cognitive impairment among older women. Arch Neurol. 2009; 38. US Department of Health and Human Services. 2008 Physical activity guide-
66(3):324-328. lines for Americans. http://www.health.gov/paguidelines. Accessed July 15,
28. Gussekloo J, de Craen AJ, Oduber C, van Boxtel MP, Westendorp RG. Sensory 2009.
This Month in Archives of Internal Medicine Online @ www.archinternmed.com
Supplementary
Sign Up for Free Online Content Available for CME
• Table of Contents E-mail Alerts • Physical Activity and Incident Nonsurgical Weight Loss for
• Topic Collection E-mail Alerts Cognitive Impairment in Elderly Extreme Obesity in Primary Care
• RSS Feeds Persons: The INVADE Study Settings: Results of the Louisiana
Obese Subjects Study
See Also View Last Month’s Free Article

• Calendar of Events • Most Viewed Articles Physical Activity at Midlife in
• Physician Jobs • Most Sent Articles Relation to Successful Survival
• Backfiles of Articles Back to 1908 • Most Viewed Collections in Women at Age 70 Years
or Older
178

Physical Activity and Incident Cognitive Impairment

in Elderly Persons
The INVADE Study
Thorleif Etgen, MD; Dirk Sander, MD; Ulrich Huntgeburth, MD;
Holger Poppert, MD; Hans Förstl, MD; Horst Bickel, PhD
Background: Data regarding the relationship between out physical activity, fully adjusted multiple logistic re-
physical activity and cognitive impairment are limited and gression analysis showed a significantly reduced risk of
controversial. We examined whether physical activity is incident cognitive impairment after 2 years for partici-
associated with incident cognitive impairment during pants with moderate or high physical activity at base-
follow-up. line (odds ratio [OR], 0.57; 95% confidence interval [CI],
0.37-0.87 [P = .01]; and OR, 0.54; 95% CI, 0.35-0.83
Methods: As part of a community-based prospective co-
[P=.005]; respectively). Further subanalysis including
hort study in southern Bavaria, Germany, 3903 partici-
participants (n=2029) without functional impairment and
pants older than 55 years were enrolled between 2001
without prodromal phase of dementia resulted in an even
and 2003 and followed up for 2 years. Physical activity
higher reduction of risk of incident cognitive impair-
(classified as no activity, moderate activity [⬍3 times/
ment for participants with moderate or high physical ac-
wk], and high activity [ⱖ3 times/wk]), cognitive func-
tivity (OR, 0.44; 95% CI, 0.24-0.83 [P=.01]; and OR, 0.46;
tion (assessed by the 6-Item Cognitive Impairment Test),
95% CI, 0.25-0.85 [P=.01]; respectively) compared with
and potential confounders were evaluated. The main out-
no activity.
come measure was incident cognitive impairment after
2 years of follow-up. Conclusion: Moderate or high physical activity is asso-
Results: At baseline, 418 participants (10.7%) had cog-
ciated with a reduced incidence of cognitive impair-
nitive impairment. After a 2-year follow-up, 207 of 3485 ment after 2 years in a large population-based cohort of
initially unimpaired subjects (5.9%) developed incident elderly subjects.
cognitive impairment. Compared with participants with- Arch Intern Med. 2010;170(2):186-193
C
OGNITIVE IMPAIRMENT IN- demonstrate a benefit of physical exercise
cluding dementia is a in preserving cognitive function.11-14 One
growing worldwide pub- limitation of most of these studies is a pos-
lic health problem, and sible reverse causality, as a decline in ha-
the prevalence in elderly bitual exercise may be the result of a pro-
persons is between 10% and 22%.1-3 Ef- dromal phase of dementia.15 Furthermore,
fective prevention strategies would have the existing literature is limited by a re-
Author Affiliations: large public health implications by im-
Department of Psychiatry and stricted study population (either men7 or
proving quality of life and reducing eco- women8), telephone assessment of cognitive
Psychotherapy, Technische
nomic cost and social burden.
Universität München, Munich, function,8 small study population,4,9-14 or
Germany (Drs Etgen, Förstl, short interval of follow-up.9,12,13 Finally,
and Bickel); Department of
Neurology, Klinikum
See also pages 124, none of the studies was performed among
170, 179, and 194 European cohorts.
Traunstein, Traunstein,
Germany (Dr Etgen); Using data from the INVADE (Inter-
Department of Neurology, Physical activity has well-known ben- vention Project on Cerebrovascular Dis-
Technische Universität efits for many chronic diseases (eg, ische- eases and Dementia in the Community of
München, Munich, Germany mic heart disease, stroke, diabetes). How- Ebersberg, Bavaria) study, 16 we con-
(Drs Sander and Poppert); ducted the present prospective cohort
ever, the evidence for preventing or delaying
Department of Neurology, study to examine the association be-
Medical Park Hospital, cognitive decline is still controversial. The
Bischofswiesen, Germany results of recent longitudinal studies and tween physical activity and cognitive func-
(Dr Sander); and INVADE randomized trials suggest that physical ex- tion with special emphasis on early cog-
Study Group, Ebersberg, ercise enhances cognitive function in older nitive decline and with regard to several
Germany (Dr Huntgeburth). adults,4-10 whereas other studies could not potential confounders.
186

METHODS
Table 1. Instruments and Sources of Information Used
in the Structured Questionnaire
SUBJECTS
Method of Assessment Item
The INVADE study is a prospective and population-based co- Assessed Weight and height
hort study.16-18 All inhabitants of the district of Ebersberg, Ger- by general Blood pressure
many, born before 1946, older than 55 years in 2001, and en- practitioner 6CIT
Barthel Index
rolled in the largest statutory German health insurance fund
Rankin Scale
(Allgemeine Ortskrankenkasse [AOK]) were identified in the Medication
AOK database and then invited to participate (n=10 325). In Physical activity
the area of Ebersberg, more than 40% of all inhabitants older History of stroke
than 55 years were insured with the AOK. During the baseline History of ischemic heart disease
period of 2001 through 2003, 3908 subjects accepted the in- Smoking status
vitation, of which 3903 subjects could be included in the present Alcohol consumption
study. The remaining 5 subjects were excluded owing to miss- Participant’s self-report Geriatric Depression Scale
ing baseline 6-Item Cognitive Impairment Test (6CIT) score Living facility
(n=2) or missing details about physical activity (n=3). Laboratory test result Serum glucose
Lipids (total cholesterol, HDL-C, triglycerides)
Creatinine
EVALUATIONS
Abbreviations: 6CIT, 6-Item Cognitive Impairment Test;
HDL-C, high-density lipoprotein cholesterol.
The complete investigation at baseline and after 2 years of fol-
low-up was performed by general practitioners of the district
of Ebersberg (n = 65) and included a standardized question- PHYSICAL ACTIVITY
naire (Table 1), medical history, evaluation of several risk fac-
tors, a physical examination, a 12-lead electrocardiogram, and Physical activity at baseline was determined by asking partici-
an overnight fasting venous blood sample for analysis in a cen- pants the number of days per week they performed strenuous
tral laboratory. All data were entered in a central database af- activities (walking, hiking, bicycling, swimming, gardening, or
ter plausibility checks for further evaluation. After the initial other exercise). Similar to the classification used in other stud-
baseline investigation, the primary care physician performed ies, participants were allocated to the following 3 groups ac-
a physical examination of the participants every 3 months. The cording to their level of activity: no activity (no regular physi-
local institutional review board of the Technische Universität cal activity), moderate activity (physical activity ⬍3 times/
München, Munich, Germany, approved this study. All pa- wk), and high activity (physical activity ⱖ3 times/wk).6,25
tients provided written informed consent before entering the Impairment in activities of daily living was assessed by means
study. Details of the study design have been published.16 of the Barthel Index26 and the Modified Rankin Scale.27 Par-
ticipants with a Barthel Index score of 100 (includes the abil-
ity to climb stairs without help and to walk ⬎50 m) and a Modi-
CARDIOVASCULAR DISEASE STATUS fied Rankin Scale score of 0 were considered as being able to
AND RISK FACTORS perform physical exercise.
Information on current health status, medical history, cogni-
tive status, mood disorders, drug use, and former cardiovas- LABORATORY EXAMINATIONS
cular risk factors was obtained from the general practitioner
by a highly structured questionnaire (Table 1). Risk factors de- Overnight fasting blood samples were drawn from each sub-
termined included the following: body mass index (BMI) (cal- ject and were transferred on ice to a central laboratory that per-
culated as weight in kilograms divided by height in meters formed all analyses including measurements of fasting serum
squared), smoking status (never, former, or current), alcohol glucose, total cholesterol, high-density lipoprotein choles-
consumption (⬍7 drinks/wk or ⱖ7 drinks/wk), depression (15- terol (HDL-C), triglyceride, and serum creatinine levels.
item Geriatric Depression Scale [GDS] score ⱖ6),19 arterial hy-
pertension (treatment with antihypertensive medication or docu- COGNITIVE SCREENING
mented blood pressure ⱖ140 mm Hg systolic or ⱖ90 mm Hg
diastolic, measured in a standardized fashion),20 diabetes melli- Screening for cognitive function was performed by using the 6CIT.
tus (treatment with antidiabetic drugs or overnight fasting se- The 6CIT, also known as the “Short Blessed Test,” is a shortened
rum glucose levels ⱖ126 mg/dL [to convert to millimoles per form of The Blessed Information Memory Concentration Scale28
liter, multiply by 0.0555]), hyperlipidemia (treatment with lipid- and consists of 6 questions (asking for year, month, and time;
lowering medication or total cholesterol level ⱖ200 mg/dL [to counting backward from 20 to 1; saying the months of the year
convert to millimoles per liter, multiply by 0.0259] or triglyc- in reverse; and remembering an address with 5 components).29
eride level ⱖ150 mg/dL [to convert to millimoles per liter, mul- Scores between 0 and 7 points are considered normal and scores
tiply by 0.0113]),21 chronic kidney disease (estimated glomer- higher than 7 are consistent with cognitive impairment.29,30 The
ular filtration rate ⬍60 mL/min/1.73 m2),22 prevalent history 6CIT is a brief and simple test of cognition that correlates well
of ischemic heart disease (documented by previous myocar- with the Mini-Mental State Examination30 and has been used re-
dial infarction or angina pectoris, bypass surgery, or ⬎50% an- cently in large epidemiological studies.22,31 It performs better in
giographic stenosis of ⱖ1 major coronary artery), and preva- mild dementia than Mini-Mental State Examination and time
lent history of stroke (neurological deficit that persisted longer needed to perform the test is about 3 to 4 minutes, which makes
than 24 hours, evaluated by a neurologist). Myocardial infarc- the 6CIT a useful tool for cognitive screening in primary care.29
tion and stroke were diagnosed according to recent recom- The test itself was applied by the general practitioners, who were
mendations.23,24 all trained in the use of the 6CIT prior to the study.
187

Table 2. Baseline Characteristics of Participants by Physical Activity a
Activity
None Moderate High

Characteristic (n = 584) (n = 1523) (n = 1796) P Value
Age, mean (SD), y 71.2 (9.1) 68.2 (7.7) 66.2 (6.9) ⬍.001
Male sex, No. (%) 191 (32.7) 617 (40.5) 790 (44.0) ⬍.001
Living in nursing home, No. (%) 30 (5.1) 16 (1.1) 10 (0.6) ⬍.001
History of stroke, No. (%) 39 (6.7) 50 (3.3) 46 (2.6) ⬍.001
History of ischemic heart disease, No. (%) 96 (16.4) 206 (13.5) 175 (9.9) ⬍.001
Diabetes mellitus, No. (%) 87 (14.8) 173 (11.4) 180 (10.1) .005
Hypertension, No. (%) 475 (81.3) 1184 (77.8) 1257 (70.0) ⬍.001
Hyperlipidemia, No. (%) 473 (81.0) 1229 (81.3) 1399 (78.4) .07
6CIT score ⬎7, No. (%) 125 (21.4) 160 (10.5) 132 (7.3) ⬍.001
6CIT, mean (SD), absolute score 4.5 (5.9) 2.5 (3.4) 2.3 (3.2) ⬍.001
Follow-up 6CIT, mean (SD), absolute score 4.7 (5.7) 2.9 (3.8) 2.3 (3.3) ⬍.001
Depression, GDS score ⱖ6, No. (%) 150 (25.7) 140 (9.2) 88 (4.9) ⬍.001
Current smoking, No. (%) 73 (12.5) 145 (9.5) 179 (10.0) .59
Alcohol, ⱖ7 drinks/wk, No. (%) 90 (15.4) 311 (20.4) 405 (22.6) .001
BMI, mean (SD) 28.6 (5.3) 27.8 (4.3) 27.4 (4.2) ⬍.001
Fasting glucose, mean (SD), mg/dL 100.3 (31.2) 96.4 (29.1) 94.6 (31.6) ⬍.001
Total cholesterol, mean (SD), mg/dL 217.1 (40.2) 219.1 (39.7) 219.0 (39.7) .55
HDL-C, mean (SD), mg/dL 55.8 (16.0) 58.1 (16.2) 58.7 (15.7) .001
Triglycerides, mean (SD), mg/dL 152.7 (85.8) 146.3 (83.4) 137.9 (81.5) ⬍.001
Systolic BP, mean (SD), mm Hg 139.4 (19.6) 140.5 (17.7) 139.1 (18.3) .10
Diastolic BP, mean (SD), mm Hg 81.7 (10.9) 82.6 (9.5) 82.2 (9.5) .14
Baseline eGFR, mean (SD), mL/min/1.73 m2 86.2 (41.9) 91.3 (37.7) 93.5 (34.6) ⬍.001
Abbreviations: 6CIT, 6-Item Cognitive Impairment Test; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared);
BP, blood pressure; eGFR, estimated glomerular filtration rate; GDS, Geriatric Depression Scale; HDL-C, high-density lipoprotein cholesterol.
SI conversion factors: To convert glucose to millimoles per liter, multiply by 0.0555; for cholesterol, by 0.0259; and for triglycerides, by 0.0113.
a Categorical variables are expressed as number (percentage) with P values calculated by the ␹2 test, and continuous variables are given as mean (SD) with
P values calculated by analysis of variance.
STATISTICAL ANALYSIS by the inclusion of participants with a prodromal phase de-

mentia, we performed a second subanalysis to assess only par-
All values are given as mean (SD) or 95% confidence intervals ticipants whose baseline 6CIT scores were lower than the 75th
(CIs) or as counts and percentages. Prevalence of cognitive im- percentile. The last subanalysis contained all participants with
pairment was calculated as the number of cognitively im- a baseline Barthel Index score of 100, a Modified Rankin Scale
paired subjects at baseline divided by the total number of par- of 0, and baseline 6CIT score lower than the 75th percentile.
ticipants. Incidence was calculated as the number of subjects For all statistical calculations, SPSS version 17.0.0 for Win-
with newly developed cognitive impairment divided by the total dows (SPSS Inc, Chicago, Illinois) was used. P⬍.05 was con-
number of initially unimpaired participants. We used ␹2 tests, sidered statistically significant.
independent sample t tests, and Mann-Whitney tests or uni-
variate analysis of variance for univariate analysis, as appro-
RESULTS
priate. Multiple logistic regression analysis was used to ana-
lyze the association of cognitive impairment as a function of
physical activity. Candidate covariates included age and BMI BASELINE CHARACTERISTICS
as continuous variables and sex, smoking (current smoking),
prevalent history of ischemic heart disease and/or stroke, hy-
pertension, diabetes, hyperlipidemia, alcohol consumption (ⱖ7
We included 3903 subjects in the analysis. Baseline char-
drinks/wk), chronic kidney disease, and depression at base- acteristics for participants by physical activity are summa-
line as categorical dichotomized variables. An unadjusted analy- rized in Table 2. Compared with participants with mod-
sis was used to examine the association between physical ac- erate or high activity, participants with no physical activity
tivity and all candidate covariates. Fully adjusted regression were older, more likely to be women, more likely to live in
models included all covariates found to be significant in un- a nursing home, and had a higher prevalence of most car-
adjusted analysis and those biological plausible covariates. The diovascularriskfactors.Nodifferenceswereobservedintotal
Hosmer-Lemeshow test results were not significant for all mul- cholesterol, blood pressure, and smoking. Of note, a lower
tiple regression analyses, thus indicating an adequate good- prevalence of alcohol consumption was observed in subjects
ness of fit. All regression analyses for incident cognitive im- with no activity. At baseline, 418 participants (10.7%) had
pairment at follow-up, which included participants with
corresponding complete data, also included adjustment for base-
cognitive impairment. The absolute 6CIT score was signifi-
line 6CIT score (used as a continuous variable). Several post cantly higher in the group with no physical activity com-
hoc subanalyses were calculated. The first included only par- paredwiththegroupswithmoderateorhighactivity(Table2).
ticipants who were able to perform physical exercise at base- The prevalence rates of cognitive impairment among par-
line based on a combined Barthel Index score of 100 and a Modi- ticipantswithno,moderate,andhighactivityatbaselinewere
fied Rankin Scale score of 0. To reduce a possible bias introduced 21.4%, 10.5%, and 7.3%, respectively.
188

Table 3. Cross-sectional Association of Physical Activity at Baseline and Cognitive Impairment at Baseline
Model No Activity Moderate Activity P Value High Activity P Value

Unadjusted
OR (95% CI) 1 [Reference] 0.43 (0.33-0.56) ⬍.001 0.29 (0.22-0.38) ⬍.001
No. cognitively impaired a/sample size, No. (%) 125/584 (21.4) 160/1523 (10.5) 132/1796 (7.3)
Adjusted for age and sex
OR (95% CI) 1 [Reference] 0.50 (0.38-0.65) ⬍.001 0.37 (0.28-0.49) ⬍.001
Fully adjusted b
OR (95% CI) 1 [Reference] 0.64 (0.48-0.87) .003 0.51 (0.37-0.70) ⬍.001
Abbreviatons: CI, confidence interval; OR, odds ratio.

a Number of participants with baseline cognitive impairment (6-Item Cognitive Impairment Test score ⬎7).
b Adjusted for age, sex, body mass index, depression, alcohol, diabetes, history of ischemic heart disease and/or stroke, hyperlipidemia, hypertension, chronic
kidney disease, and smoking.
CROSS-SECTIONAL ASSOCIATION BETWEEN line, there were 207 participants (5.9%) who developed in-
PHYSICAL ACTIVITY AND BASELINE cident cognitive impairment at the end of the follow-up pe-
COGNITIVE IMPAIRMENT riod. The incidence of new cognitive impairment among
participants with no, moderate, and high activity at base-
Findings from the univariate analysis showed a significant line was 13.9%, 6.7%, and 5.1%, respectively.
association between cognitive impairment at baseline and
the following covariates: age (P⬍.001), sex (P⬍.001), BMI LONGITUDINAL ASSOCIATION BETWEEN
(P⬍.001), diabetes (P=.005), hypertension (P⬍.001), his- PHYSICAL ACTIVITY AND COGNITIVE
toryofischemicheartdisease(P⬍.001)andstroke(P⬍.001), IMPAIRMENT AT FOLLOW-UP
chronic kidney disease (P⬍.001), alcohol consumption
(P=.001),anddepression(P⬍.001).Thefullyadjustedmodel Unadjusted analysis yielded a strong relationship be-
for age, sex, BMI, depression, alcohol, diabetes, history of tween physical activity at baseline and the development
ischemic heart disease and/or stroke, hyperlipidemia, hy- of incident cognitive impairment in participants with no
pertension, chronic kidney disease, and smoking showed activity compared with participants with moderate or high
asignificantlydecreasedriskofcognitiveimpairmentatbase- activity. After adjustment for age, sex, baseline cogni-
line for participants with moderate and high physical activ- tive function, BMI, depression, alcohol, diabetes, his-
ity compared with those without physical activity (Table 3). tory of ischemic heart disease and/or stroke, hyperlipid-
emia, hypertension, chronic kidney disease, and smoking,
FOLLOW-UP CHARACTERISTICS there remained a significant association with new cog-
nitive function impairment in participants with no physi-
The median time of follow-up was 778 days (range, 1035 cal activity compared with those with moderate or high
days; interquartile range, 80 days). Data on cognitive func- activity (Table 4).
tion could be obtained in 3369 subjects. Data were not
available for 534 participants (13.7%) because of death ANALYSIS OF SUBGROUPS WITHOUT
(n=106), change of health insurance company (n=25), FUNCTIONAL IMPAIRMENT AND WITHOUT
or incomplete data (n = 403). The 534 participants lost PRODROMAL PHASE OF DEMENTIA
to follow-up showed the following significant mean (SD)
differences in baseline characteristics vs the remaining A subanalysis of all those participants (baseline charac-
3369 participants: older (69.2 [9.3] vs 67.5 [7.5] years; teristics, eTable 1; http://www.archinternmed.com) who
P ⬍.001), higher diastolic blood pressure (83 [10] vs 82 were unimpaired in activities of daily living (Barthel In-
[10] mm Hg; P=.01), higher 6CIT score (3.5 [5.0] vs 2.6 dex score of 100 and Modified Rankin Scale score of 0)
[3.6]; P⬍.001), higher GDS score (3.1 [2.9] vs 2.3 [2.4]; showed similar results. In the fully adjusted model, there
P⬍.001), and lower estimated glomerular filtration rate was a significant association with incident cognitive func-
(87.2 [38.1] vs 92.3 [36.8] mL/min/1.73 m2; P=.004). tion impairment in participants with no physical activ-
The prevalence of stroke (6.0% vs 3.1%; P = .001) and is- ity at baseline compared with those with moderate or high
chemic heart disease (18.7% vs 11.6%; P⬍.001) were inactivity at baseline (eTable 2).
creased, but hyperlipidemia (76.8% vs 80.5%; P=.048) To reduce the problem of a possible reverse causality
and alcohol consumption (16.6% vs 21.3%; P=.01) were (see the “Comment” section), we performed a second sub-
less common in the participants lost to follow-up. analysis including only participants whose 6CIT scores
The group with high physical activity at baseline showed were lower than the 75th percentile, ie, with a 6CIT score
no change in absolute 6CIT score, whereas the groups with of 0 to 4 (baseline characteristics, eTable 3). Again, high
no activity and with moderate activity developed more el- and moderate physical activity compared with no activ-
evated 6CIT scores during follow-up (Table 2). After we ity was associated with a reduced risk of incident cog-
excluded participants with cognitive impairment at base- nitive impairment (eTable 4).
189

Table 4. Association of Incident Cognitive Impairment With Physical Activity at Baseline

Unadjusted
OR (95% CI) 1 [Reference] 0.45 (0.31-0.64) ⬍.001 0.33 (0.23-0.48) ⬍.001
No. cognitively impaired a/sample size, No. % 53/382 (13.9) 78/1166 (6.7) 75/1484 (5.0)
OR (95% CI) 1 [Reference] 0.51 (0.35-0.75) .001 0.43 (0.29-0.63) ⬍.001
Fully adjusted b
OR (95% CI) 1 [Reference] 0.57 (0.37-0.87) .01 0.54 (0.35-0.83) .005
Abbreviations: CI, confidence interval; OR, odds ratio.

a Number of participants with baseline cognitive impairment (6-Item Cognitive Impairment Test [6CIT] score ⬎7).
b Adjusted for age, sex, body mass index, baseline 6CIT score, depression, alcohol, diabetes, history of ischemic heart disease and/or stroke, hyperlipidemia,
hypertension, chronic kidney disease, and smoking.
Table 5. Baseline Characteristics of Participants Without Functional Impairment a and With a 6CIT Score Lower
Than the 75th Percentile b by Physical Activity c
Activity
Characteristic None (n = 241) Moderate (n = 971) High (n = 1342) P Value

Age, mean (SD), y 68.2 (7.7) 66.7 (7.1) 65.4 (6.3) ⬍.001
Male sex, No. (%) 77 (32.0) 397 (40.9) 580 (43.2) .004
Living in nursing home, No. (%) 2 (0.8) 2 (0.2) 2 (0.1) ⬍.001
History of stroke, No. (%) 4 (1.7) 23 (2.4) 23 (1.7) .50
History of ischemic heart disease, No. (%) 27 (11.2) 110 (11.3) 115 (8.6) .06
Diabetes mellitus, No. (%) 32 (13.3) 99 (10.2) 126 (9.4) .20
Hypertension, No. (%) 193 (80.1) 732 (75.4) 919 (68.5) ⬍.001
Hyperlipidemia, No. (%) 195 (80.9) 789 (81.3) 1036 (77.2) .06
6CIT, mean (SD), absolute score 1.3 (1.5) 1.1 (1.4) 1.1 (1.5) .04
Follow-up 6CIT, mean (SD), absolute score 2.3 (2.9) 1.8 (2.5) 1.6 (2.4) .001
Depression, GDS score ⱖ6, No. (%) 35 (14.5) 78 (8.0) 50 (3.7) ⬍.001
Current smoking, No. (%) 37 (15.4) 94 (9.7) 126 (9.4) .02
Alcohol, ⱖ7 drinks/wk 37 (15.4) 195 (20.1) 300 (22.4) .04
BMI, mean (SD) 29.2 (5.4) 27.7 (4.4) 27.3 (4.1) ⬍.001
Fasting glucose, mean (SD), mg/dL 99.5 (26.3) 95.0 (23.9) 94.6 (32.6) .05
Total cholesterol, mean (SD), mg/dL 217.2 (39.7) 220.2 (39.7) 219.0 (39.5) .56
HDL-C, mean (SD), mg/dL 55.8 (16.4) 58.2 (15.7) 59.2 (15.9) .008
Triglycerides, mean (SD), mg/dL 149.5 (80.4) 144.9 (82.3) 135.8 (82.2) .006
Systolic BP, mean (SD), mm Hg 140.0 (17.7) 139.5 (17.3) 138.2 (17.4) .10
Diastolic BP, mean (SD), mm Hg 82.8 (9.5) 82.9 (9.3) 82.3 (9.0) .27
Baseline eGFR, mean (SD), mL/min/1.73 m2 96.0 (40.2) 94.7 (38.4) 94.2 (33.6) .77
Abbreviations: 6CIT, 6-Item Cognitive Impairment Test; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared);
BP, blood pressure; eGFR, estimated glomerular filtration rate; GDS, Geriatric Depression Scale; HDL-C, high-density lipoprotein cholesterol.
SI conversion factors: To convert glucose to millimoles per liter, multiply by 0.0555; cholesterol, by 0.0259; and triglycerides, by 0.0113.
a “Without functional impairment” was defined by a Barthel Index score of 100 and Modified Rankin Scale score of 0.
b 6CIT score of 0 to 4.
c Categorical variables are expressed as number (percentage) with P values calculated by the ␹2 test, and continuous variables are given as mean (SD) with
P values calculated by analysis of variance.
The last subanalysis containing all participants tivity yielded a significant association with incident cog-
(n = 2029) with a baseline Barthel Index score of 100, nitive impairment after 2 years. This result remained
Modified Rankin Scale score of 0, and baseline 6CIT score statistically significant even after adjustment of impor-
lower than the 75th percentile (baseline characteristics, tant potential confounders including age, sex, baseline
Table 5) resulted in an even more reduced risk of in- cognitive status, depression, chronic kidney disease, and
cident cognitive impairment for participants with mod- cardiovascular risk factors. In addition, no clear dose-
erate or high activity compared with no activity (Table 6). response relationship between physical activity and in-
cident cognitive impairment was found, since there was
COMMENT no benefit of high physical activity over moderate physi-
cal activity.
This population-based prospective study of a large co- Some prospective cohort studies about the relation-
hort of elderly subjects found that lack of physical ac- ship between physical exercise and cognitive decline
190

Table 6. Association of Incident Cognitive Impairment With Physical Activity at Baseline for Participants Without Baseline Functional
Impairment a and With a Baseline 6CIT Score Lower Than the 75th Percentile b

Unadjusted
OR (95% CI) 1 [Reference] 0.49 (0.28-0.87) .01 0.40 (0.23-0.70) .001
No. cognitively impaired c/sample size, No. (%) 19/214 (8.9) 38/835 (4.6) 45/1197 (3.8)
OR (95% CI) 1 [Reference] 0.51 (0.28-0.91) .02 0.45 (0.25-0.80) .006
Fully adjusted d
OR (95% CI) 1 [Reference] 0.44 (0.24-0.83) .01 0.46 (0.25-0.85) .01
Abbreviations: 6CIT, 6-Item Cognitive Impairment Test; CI, confidence interval; OR, odds ratio.
a “Without functional impairment” was defined by a Barthel Index score of 100 and Modified Rankin Scale score of 0.
b 6CIT score of 0 to 4.
c Number of participants with incident cognitive impairment (6CIT score ⬎7).
d Adjusted for age, sex, body mass index, baseline 6CIT score, depression, alcohol, diabetes, history of ischemic heart disease and/or stroke, hyperlipidemia,
hypertension, chronic kidney disease, and smoking.
have found a protective association, whereas others phase of dementia. First, in the INVADE study, we as-
have failed to find this association. Several differences sessed cognitive impairment instead of dementia, whereas
among study design and study parameters probably most other studies examined the association between
yielded these inconsistent results. For example, the physical activity and dementia. Therefore, cognitive func-
sample size of some prospective cohort studies was tion was assessed by the 6CIT, which screens for cogni-
small (⬍500 participants),4,11,14 which limits multivari- tive impairment and performs better in mild dementia
ate analyses, and there were study population discrep- or prodromal phase of dementia.29,30 Second, during the
ancies including restriction by sex7,8 and diverse age follow-up period, all participants with cognitive impair-
groups ranging from participants older than 55 years4 to ment at baseline (and potentially low activity linked with
older than 75 years.11,14 There was also a great variabil- the prodromal phase of dementia) were excluded, which
ity among studies concerning the adjustment for pos- should reduce the potential for this classification error.
sible confounders, ranging from adjustment for age and Third, to reduce a possible bias introduced by the inclu-
sex only6 to adjustment for several confounders such as sion of participants with “borderline cognitive impair-
age, sex, education, medical history, cholesterol, apoli- ment” who might therefore present already with a re-
poprotein E4, or smoking.5,7 Predominantly, diverse duced physical activity, an additional analysis was
cognitive tests assessing manifest dementia rather than performed that included only participants with a cogni-
cognitive impairment were applied. Some studies used tive function assessment lower than the 75th percentile
the Modified Mini-Mental State Examination or a simi- in the 6CIT.
lar test,4,5,8,25 but other tests like the Cognitive Abilities This study cannot fully explain the mechanism be-
Screening Instrument were also applied.6,7 Occasionally, hind the observed association between incident cogni-
cognitive function was assessed by telephone inter- tive impairment and physical activity. Several factors may
view.8 The majority of the cohort studies rated physical contribute to a possible protective effect of physical ac-
activity according to self-reporting,5,6,14,25 a more objec- tivity. For example, physical exercise leads to a reduced
tive method (treadmill and oxygen uptake) was rarely risk of cardiovascular diseases (eg, hypertension, diabe-
used.4 Only some studies undertook efforts to account tes, stroke), which are associated with cognitive de-
for reverse causality.6,7,10 Limitations of the few con- cline.34 Moreover, physical activity may even directly im-
trolled intervention trials comprised a small sample size prove cerebral perfusion35 and induce angiogenesis in the
(⬍150 participants) and a short time of follow-up (ⱕ1 cerebral cortex.36 In experimental studies, physical ac-
year).9,10,12,13 Nevertheless, a recent meta-analysis of 16 tivity resulted in an increase of neurogenesis37 and neu-
prospective studies with 163 797 participants without rotrophic factors,38 especially in the hippocampus, which
dementia at baseline and 3219 cases of dementia at has an important role in the pathogenesis of cognitive
follow-up found a decreased relative risk of dementia in impairment and dementia. In summary, these and other
the highest physical activity category compared with factors underline a possible protective effect of physical
the lowest (0.72; 95% CI, 0.60-0.86).32 activity against cognitive decline and link pathophysi-
One limitation of these studies is a possible reverse ological evidence with results from prospective cohort
causality.15 It is now widely accepted that manifesta- studies.
tions of behavior changes (including decline in habitual The strengths of our study are the large number of pa-
exercise) related to dementia with insidious onset can oc- tients, the complete nature of the data set, the longitu-
cur years before a person crosses a threshold that allows dinal assessment of cognitive performance, and the regu-
a definitive diagnosis of dementia to be made.33 Our study lar examination by general practitioners. Another
design included several steps to reduce the potential effect advantage includes the ability to adjust for multiple vas-
of changes in physical exercise related to a prodromal cular risk factors that may affect cognitive function such
191

as hypertension, diabetes mellitus, or history of stroke. tent: Etgen, Sander, Huntgeburth, Poppert, Förstl, and
Finally, and in contrast to most other studies, we also Bickel. Statistical analysis: Etgen and Bickel. Obtained fund-
included chronic kidney disease in the list of possible con- ing: Huntgeburth, Förstl, and Bickel. Administrative, tech-
founders because several recent studies found impaired nical, and material support: Sander, Huntgeburth, Pop-
renal function to be an independent predictor of cogni- pert, Förstl, and Bickel. Study supervision: Sander,
tive impairment.22,39,40 Despite the comprehensive na- Huntgeburth, Förstl, and Bickel.
ture of the data set, this study also has several limita- Financial Disclosure: None reported.
tions. First, the definition of physical activity was based Funding/Support: The Bavarian health insurance com-
on a questionnaire rather than a more precise and ob- pany AOK funded this work.
jective method, such as motion sensors or doubly la- Role of the Sponsor: The sponsor had no role in the de-
beled water measurement. Second, the follow-up pe- sign and conduct of the study; in the collection, man-
riod for cognitive decline of 2 years is relatively short. agement, analysis, and interpretation of the data; or in
Third, the assessment of cognitive function was based only the preparation, review, or approval of the manuscript.
on the use of the 6CIT. However, it has been shown that Online-Only Material: eTables 1 through 4 are avail-
the 6CIT is equivalent to the Mini-Mental State Exami- able at http://www.archinternmed.com.
nation in identifying dementia.41 Current data even sug-
gest a better performance of 6CIT compared with other
REFERENCES
tests in patients with mild dementia, favoring its use as
a screening tool for cognitive function in primary care.29 1. Graham JE, Rockwood K, Beattie BL, et al. Prevalence and severity of cognitive
Fourth, although the AOK represents the insurance plan impairment with and without dementia in an elderly population. Lancet. 1997;
with the largest market share, a bias due to the partici- 349(9068):1793-1796.
pation rate based on voluntary participation or member- 2. Luck T, Riedel-Heller SG, Kaduszkiewicz H, et al; AgeCoDe Group. Mild cogni-
ship in the AOK cannot be ruled out. Last, we cannot fully tive impairment in general practice: age-specific prevalence and correlate re-
sults from the German study on ageing, cognition and dementia in primary care
exclude the bias by the observational nature of this study patients (AgeCoDe). Dement Geriatr Cogn Disord. 2007;24(4):307-316.
design, as, for example, participants with cognitive im- 3. Plassman BL, Langa KM, Fisher GG, et al. Prevalence of cognitive impairment
pairment at the beginning could be less likely to get in- without dementia in the United States. Ann Intern Med. 2008;148(6):427-434.
volved or those participants who were lost to follow-up 4. Barnes DE, Yaffe K, Satariano WA, Tager IB. A longitudinal study of cardiores-
piratory fitness and cognitive function in healthy older adults. J Am Geriatr Soc.
had a higher risk of death. 2003;51(4):459-465.
In conclusion, the present study found that in a gen- 5. Laurin D, Verreault R, Lindsay J, MacPherson K, Rockwood K. Physical activity
eral elderly population, moderate or high physical activ- and risk of cognitive impairment and dementia in elderly persons. Arch Neurol.
ity, compared with no physical activity, are indepen- 2001;58(3):498-504.
dently associated with a lower risk of developing incident 6. Larson EB, Wang L, Bowen JD, et al. Exercise is associated with reduced risk for
incident dementia among persons 65 years of age and older. Ann Intern Med.
cognitive impairment after 2 years of follow-up. Future 2006;144(2):73-81.
large randomized controlled intervention trials assess- 7. Abbott RD, White LR, Ross GW, Masaki KH, Curb JD, Petrovitch H. Walking and
ing the quantity (eg, no activity vs moderate vs high ac- dementia in physically capable elderly men. JAMA. 2004;292(12):1447-1453.
tivity) and quality (aerobic exercise or any type of physi- 8. Weuve J, Kang JH, Manson JE, Breteler MM, Ware JH, Grodstein F. Physical ac-
tivity, including walking, and cognitive function in older women. JAMA. 2004;
cal exercise, like balance and strength training or even 292(12):1454-1461.
integrated physical activities like dancing or games) of 9. Williamson JD, Espeland M, Kritchevsky SB, et al; LIFE Study Investigators.
physical activity that is required to prevent or delay a de- Changes in cognitive function in a randomized trial of physical activity: results
cline in cognitive function are recommended.42 These trials of the lifestyle interventions and independence for elders pilot study. J Gerontol
should also focus on selected cognitive tests that are clini- A Biol Sci Med Sci. 2009;64(6):688-694.
10. Lautenschlager NT, Cox KL, Flicker L, et al. Effect of physical activity on cogni-
cally relevant and transparent and heighten the repro- tive function in older adults at risk for Alzheimer disease: a randomized trial. JAMA.
ducibility of results for future research.42 Randomized con- 2008;300(9):1027-1037.
trolled trials could also account for the problem of reverse 11. Broe GA, Creasey H, Jorm AF, et al. Health habits and risk of cognitive impair-
causality, since this design would ensure a homoge- ment and dementia in old age: a prospective study on the effects of exercise,
smoking and alcohol consumption. Aust N Z J Public Health. 1998;22(5):621-
neous distribution of participants without functional im- 623.
pairment and without baseline cognitive impairment 12. Hill RD, Storandt M, Malley M. The impact of long-term exercise training on psy-
among the different intervention groups. chological function in older adults. J Gerontol. 1993;48(1):P12-P17.
13. Madden DJ, Blumenthal JA, Allen PA, Emery CF. Improving aerobic capacity in
Accepted for Publication: November 16, 2009. healthy older adults does not necessarily lead to improved cognitive performance.
Psychol Aging. 1989;4(3):307-320.
Correspondence: Thorleif Etgen, MD, Department of 14. Verghese J, Lipton RB, Katz MJ, et al. Leisure activities and the risk of dementia
Neurology, Klinikum Traunstein, 83278 Traunstein, in the elderly. N Engl J Med. 2003;348(25):2508-2516.
Cuno-Niggl-Strasse 3, Germany (thorleif.etgen 15. Solfrizzi V, Capurso C, D’Introno A, et al. Lifestyle-related factors in predemen-
@klinikum-traunstein.de). tia and dementia syndromes. Expert Rev Neurother. 2008;8(1):133-158.
16. Sander D, Schulze-Horn C, Bickel H, Gnahn H, Bartels E, Conrad B. Combined
Author Contributions: Dr Etgen had full access to all the effects of hemoglobin A1c and C-reactive protein on the progression of subclini-
data in the study and takes responsibility for the integ- cal carotid atherosclerosis: the INVADE study. Stroke. 2006;37(2):351-357.
rity of the data and the accuracy of the data analysis. Study 17. Chonchol M, Gnahn H, Sander D. Impact of subclinical carotid atherosclerosis
concept and design: Etgen, Sander, Förstl, and Bickel. Ac- on incident chronic kidney disease in the elderly. Nephrol Dial Transplant. 2008;
quisition of data: Huntgeburth, Poppert, and Bickel. Analy- 23(8):2593-2598.
18. Desbien AM, Chonchol M, Gnahn H, Sander D. Kidney function and progression
sis and interpretation of data: Etgen, Sander, and Bickel. of carotid intima-media thickness in a community study. Am J Kidney Dis. 2008;
Drafting of the manuscript: Etgen, Sander, and Bickel. Criti- 51(4):584-593.
cal revision of the manuscript for important intellectual con- 19. Weyerer S, Eifflaender-Gorfer S, Kohler L, et al; German AgeCoDe Study group
192

(German Study on Ageing, Cognition and Dementia in Primary Care Patients). short Orientation-Memory-Concentration Test of cognitive impairment. Am J
Prevalence and risk factors for depression in non-demented primary care at- Psychiatry. 1983;140(6):734-739.
tenders aged 75 years and older. J Affect Disord. 2008;111(2-3):153-163. 31. Yaffe K, Weston AL, Blackwell T, Krueger KA. The metabolic syndrome and de-
20. Chobanian AV, Bakris GL, Black HR, et al; Joint National Committee on Preven- velopment of cognitive impairment among older women. Arch Neurol. 2009;
tion, Detection, Evaluation, and Treatment of High Blood Pressure. National Heart, 66(3):324-328.
Lung, and Blood Institute; National High Blood Pressure Education Program Co- 32. Hamer M, Chida Y. Physical activity and risk of neurodegenerative disease: a sys-
ordinating Committee. Seventh report of the Joint National Committee on Pre- tematic review of prospective evidence. Psychol Med. 2009;39(1):3-11.
vention, Detection, Evaluation, and Treatment of High Blood Pressure. Hypertension. 33. Linn RT, Wolf PA, Bachman DL, et al. The “preclinical phase” of probable Alz-
2003;42(6):1206-1252. heimer’s disease: a 13-year prospective study of the Framingham cohort. Arch
21. De Backer G, Ambrosioni E, Borch-Johnsen K, et al; Third Joint Task Force of Neurol. 1995;52(5):485-490.
European and Other Societies on Cardiovascular Disease Prevention in Clinical 34. Etgen T, Bronner M, Sander D, Bickel H, Sander K, Forstl H. Somatic factors in cog-
Practice. European guidelines on cardiovascular disease prevention in clinical nitive impairment [in German]. Fortschr Neurol Psychiatr. 2009;77(2):72-82.
practice. Eur Heart J. 2003;24(17):1601-1610.
35. Rogers RL, Meyer JS, Mortel KF. After reaching retirement age physical activ-
22. Etgen T, Sander D, Chonchol M, et al. Chronic kidney disease is associated with
ity sustains cerebral perfusion and cognition. J Am Geriatr Soc. 1990;38(2):
incident cognitive impairment in the elderly: the INVADE study. Nephrol Dial
123-128.
Transplant. 2009;24(10):3144-3150.
36. Black JE, Isaacs KR, Anderson BJ, Alcantara AA, Greenough WT. Learning causes
23. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction rede-
synaptogenesis, whereas motor activity causes angiogenesis, in cerebellar cor-
fined—a consensus document of The Joint European Society of Cardiology/
tex of adult rats. Proc Natl Acad Sci U S A. 1990;87(14):5568-5572.
American College of Cardiology Committee for the redefinition of myocardial
37. van Praag H, Shubert T, Zhao C, Gage FH. Exercise enhances learning and
infarction. J Am Coll Cardiol. 2000;36(3):959-969.
24. Adams H, Adams R, Del Zoppo G, Goldstein LB; Stroke Council of the American hippocampal neurogenesis in aged mice. J Neurosci. 2005;25(38):8680-
Heart Association; American Stroke Association. Guidelines for the early man- 8685.
agement of patients with ischemic stroke: 2005 guidelines update a scientific 38. Berchtold NC, Kesslak JP, Pike CJ, Adlard PA, Cotman CW. Estrogen and exer-
statement from the Stroke Council of the American Heart Association/American cise interact to regulate brain-derived neurotrophic factor mRNA and protein ex-
Stroke Association. Stroke. 2005;36(4):916-923. pression in the hippocampus. Eur J Neurosci. 2001;14(12):1992-2002.
25. Middleton L, Kirkland S, Rockwood K. Prevention of CIND by physical activity: 39. Hailpern SM, Melamed ML, Cohen HW, Hostetter TH. Moderate chronic kidney
different impact on VCI-ND compared with MCI. J Neurol Sci. 2008;269(1-2): disease and cognitive function in adults 20 to 59 years of age: Third National
80-84. Health and Nutrition Examination Survey (NHANES III). J Am Soc Nephrol. 2007;
26. Wade DT, Collin C. The Barthel ADL Index: a standard measure of physical disability? 18(7):2205-2213.
Int Disabil Stud. 1988;10(2):64-67. 40. Kurella M, Chertow GM, Fried LF, et al. Chronic kidney disease and cognitive im-
27. de Haan R, Limburg M, Bossuyt P, van der Meulen J, Aaronson N. The clinical mean- pairment in the elderly: the health, aging, and body composition study. J Am Soc
ing of Rankin “handicap” grades after stroke. Stroke. 1995;26(11):2027-2030. Nephrol. 2005;16(7):2127-2133.
28. Blessed G, Tomlinson BE, Roth M. The association between quantitative mea- 41. Davous P, Lamour Y, Debrand E, Rondot P. A comparative evaluation of the short
sures of dementia and of senile change in the cerebral grey matter of elderly orientation memory concentration test of cognitive impairment. J Neurol Neu-
subjects. Br J Psychiatry. 1968;114(512):797-811. rosurg Psychiatry. 1987;50(10):1312-1317.
29. Brooke P, Bullock R. Validation of a 6 item cognitive impairment test with a view 42. Angevaren M, Aufdemkampe G, Verhaar HJ, Aleman A, Vanhees L. Physical activity
to primary care usage. Int J Geriatr Psychiatry. 1999;14(11):936-940. and enhanced fitness to improve cognitive function in older people without known
30. Katzman R, Brown T, Fuld P, Peck A, Schechter R, Schimmel H. Validation of a cognitive impairment. Cochrane Database Syst Rev. 2008;(3):CD005381.
193

Exercise Effects on Bone Mineral Density,

Falls, Coronary Risk Factors, and
Health Care Costs in Older Women
The Randomized Controlled Senior Fitness and Prevention (SEFIP) Study
Wolfgang Kemmler, PhD; Simon von Stengel, PhD; Klaus Engelke, PhD;
Lothar Häberle, PhD; Willi A. Kalender, PhD, MD
Background: Physical exercise affects many risk fac- low-up] for the exercise group: 1.77% [1.26% to 2.28%]
tors and diseases and therefore can play a vital role in gen- vs controls: 0.33% [−0.24% to 0.91%]; P⬍.001), femoral
eral disease prevention and treatment of elderly individu- neck (exercise group: 1.01% [0.37% to 1.65%] vs con-
als and may reduce costs. We sought to determine whether trols: −1.05% [−1.70% to −0.40%]; P⬍.001), and fall rate
a single exercise program affects fracture risk (bone min- per person during 18 months (exercise group: 1.00 [0.76
eral density [BMD] and falls), coronary heart disease (CHD) to 1.24] vs controls: 1.66 [1.33 to 1.99]; P=.002). The 10-
risk factors, and health care costs in community- year CHD risk was significantly affected in both sub-
dwelling elderly women. groups (absolute change for the exercise group: −1.96%
[95% CI, −2.69% to −1.23%] vs controls: −1.15% [−1.69%
Methods: We conducted a randomized, single-blinded,
to −0.62%]; P = .22), with no significant difference be-
controlled trial from May 1, 2005, through July 31, 2008, tween the groups. The direct health care costs per partici-
recruiting women 65 years or older who were living in-
pant during the 18-month intervention showed nonsig-
dependently in the area of Erlangen-Nuremberg, Ger-
nificant differences between the groups (exercise group:
many. In all, 246 women were randomly assigned to an
18-month exercise program (exercise group) or a well- €2255 [95% CI, €1791-€2718] vs controls: €2780 [€2187-
ness program (control group). The exercise group (n=123) €3372]; P=.20).
performed a multipurpose exercise program with special
emphasis on exercise intensity; the controls (n=123) fo- Conclusion: Compared with a general wellness pro-
cused on well-being with a low-intensity, low-frequency gram, our 18-month exercise program significantly im-
program. The main outcome measures were BMD, the proved BMD and fall risk, but not predicted CHD risk,
number of falls, the Framingham-based 10-year CHD risk, in elderly women. This benefit occurred at no increase
and direct health care costs. in direct costs.
Results: For the 227 women who completed the 18- Trial Registration: clinicaltrials.gov Identifier:
month study, significant exercise effects were observed for NCT00267839
BMD of the lumbar spine (mean [95% confidence inter-
val (CI)] percentage of change in BMD [baseline to fol- Arch Intern Med. 2010;170(2):179-185
A
DVANCED AGE IS ASSOCI - diseases, physical exercise affects a vari-
ated with a variety of age- ety of risk factors and diseases3 and there-
related diseases that in- fore plays a vital role in general disease pre-
creasingly stress health vention and treatment of the elderly. So
care systems. Approxi- far, positive exercise effects have been de-
mately 25% of the German population 65 termined for insulin resistance/glucose in-
tolerance,4,5 blood pressure,6,7 blood lipid
Author Affiliations: Institute of levels,8,9 body composition,10,11 abdomi-
Medical Physics (Drs Kemmler, See also pages 124, 170, nal adiposity, 12 bone mineral density
von Stengel, Engelke, and
Kalender) and Institute of
186, and 194 (BMD),13,14 osteoarthritis, low back pain15
Biometry and Epidemiology and risk of falls.16,17
(Dr Häberle), Friedrich- years or older have more than 5 chronic Numerous exercise regimens empha-
Alexander University of diseases, and approximately 50% have 2 size speed, strength, power, endurance,
Erlangen-Nuremberg, to 4.1,2 Contrary to pharmaceutical agents, or coordination.18,19 However, because
Erlangen, Germany. which are typically dedicated to specific specific stimuli cause specific adapta-
179

METHODS
Table 1. Baseline Characteristics
A semi-blinded, randomized 18-month controlled trial was con-
Mean (SD) ducted to compare the effects of an intense exercise program
Exercise Control with a low-intensity, low-frequency control program on mul-
Group Group P tiple risk factors and HCCs of elderly community-dwelling
Characteristic (n=123) (n=123) Value women. The study was blinded for the outcome assessors and
Age, y 68.9 (3.9) 69.2 (4.1) .45
participants but not for the investigator (W.K.). Unlike other
Median (95% CI) 68 (65-78) 68 (65-78)
exercise studies, participants were blinded to the underlying
Height, cm 161.8 (6.1) 160.5 (5.8) .08 hypothesis by postulating a different study aim (wellness) for
Weight, kg 68.1 (10.9) 69.5 (12.0) .33 the control group. In this study, the primary goal of the well-
Total body fat, % 36.3 (5.9) 37.4 (5.6) .09 ness protocol was to positively affect well-being and to briefly
Age at menopause, y a 48.8 (5.9) 49.8 (5.3) .16 introduce topics related to a healthy lifestyle and physical ac-
Energy intake, kcal/d b 1574 (411) 1535 (392) .51 tivity. Both groups were trained separately to prevent contact be-
Calcium intake, mg/d b 828 (414) 816 (356) .89 tween the cohorts. The effectiveness of the blinding in the con-
Baseline physical activity index c 4.2 (1.1) 4.3 (1.2) .44 trol group was proven in structured interviews conducted by the
Exercise volume, min/wk a 85 (102) 67 (111) .45 primary investigators at the end of the 18 months.
V̇O2 peak, mL/kg/min d 24.1 (4.1) 22.9 (4.2) .06 The study protocol was approved by the ethics committee
Smokers, No. b 4 4 ⬎.99 of the Friedrich-Alexander University of Erlangen-Nuremberg
Diabetes mellitus, No. a 10 11 .76 and the Bundesamt für Strahlenschutz. The study was de-
Hypertension, No. e 51 60 .25 signed and executed by the Institute of Medical Physics, and
High cholesterol, No. f 64 52 .80 randomization and statistical procedures were performed by
Metabolic syndrome, No. g 45 50 .43 the Institute of Biometry and Epidemiology, both at Friedrich-
Osteoporosis, No. h 27 23 .53 Alexander University of Erlangen-Nuremberg. All study par-
Osteoarthritis, No. a 9 9 ⬎.99
ticipants gave written informed consent. The study was fully
Multiple morbidity, No. i 67 75 .30
registered under http://www.clinicaltrials.gov.
In the present study, we tested the hypothesis that our mul-
Abbreviations: CI, confidence interval; V̇O2 peak, peak oxygen
consumption.
tipurpose exercise protocol significantly affects BMD, fall fre-
SI conversion factor: To convert energy intake from kilocalories to quency, CHD risk, and HCCs compared with a low-volume,
kilojoules, multiply by 4.186. low-intensity “wellness” program.
a As assessed by baseline questionnaire.
b Four-day dietary protocol.
c Self-rated physical activity score (1 indicates very low; 7, very high).
PARTICIPANTS
d Stepwise treadmill test to voluntary maximum.
e Defined as diastolic blood pressure of greater than 90 mm Hg and/or
Six hundred fifty-nine women, all members of Siemens Health
systolic blood pressure of greater than 140 mm Hg.
Insurance, who were 65 years or older and were living inde-
f Defined as total cholesterol of greater than 250 mg/dL (to convert to pendently in the area of Erlangen-Nuremberg replied to our
millimoles per liter, multiply by 0.0259). individualized information letter (n=7500). Participants were
g According to International Diabetes Federation guidelines.22 recruited by mail between May 1, 2005, and January 31, 2006
h Defined as having a bone mineral density T-score of less than −2.5 at the
(±1 week). Two hundred eighty-three women were not admit-
lumbar spine or total hip. ted to the study because of diseases (eg, alcoholism and hy-
i Defined as 2 or more diseases. Besides the factors given in the table, all
other diseases (including diabetes and osteoarthritis) were assessed by
percortisolism) or medication usage (eg, bisphosphonates, hor-
baseline questionnaire. mone therapy, glucocorticoids, and laxatives) affecting bone
metabolism or fall risk during the past 2 years (n=254), a his-
tory of cardiovascular disease (eg, stroke or cardiac events)
(n=13), acute or chronic inflammatory diseases (n=2), known
tions, exercise protocols that focus on coronary heart secondary osteoporosis (n=2), participation in exercise stud-
disease (CHD) risk factors usually differ from regi- ies during the preceding 2 years (n=10), or very low physical
mens that target muscle or bone strength or those that capacity (⬍50 W at cycle ergometry) or athletic history dur-
focus on reduction in fall risk. Thus, for the preven- ing the past decade (n=2). Three hundred seventy-six women
tion of CHD, fracture risk, and frailty, women should were invited to meetings to receive detailed study informa-
tion. Primarily because of the randomization procedure and the
perform endurance, resistance, and balance exercise inability to choose the preferred study arm, 80 individuals re-
programs. Because of reduced mobility or time con- fused to participate. Thus, 296 women were assigned by com-
straints, it is unlikely that elderly individuals are will- puter-generated block randomization (first sequence: 150 sub-
ing to participate simultaneously in several different jects and a block size of 3⫻2; last sequence: 146 subjects and
exercise programs.20 Therefore, it was the aim of the a block size of 2⫻2) stratified for age to 3 intervention arms:
Senior Fitness and Prevention (SEFIP) study to an exercise program (exercise group; 123 women), a control
develop a multipurpose exercise program and to show group (123 women), or an exercise and whole-body vibration
its overall effectiveness in reducing fracture risk (BMD group (50 women). The whole-body vibration group consti-
and falls) and 10-year CHD risk along with its effect tuted a substudy with extended eligibility criteria (ie, exclu-
on health care costs (HCCs) (all equally weighted sion in case of knee or hip implant).21 Owing to the different
eligibility criteria, divergent study aims, and additional vibra-
study end points). Furthermore, to ensure transfer- tion protocol, the whole-body vibration group was not in-
ability of our program, the exercise program was con- cluded in the present analysis.
ducted under training conditions that are applicable to All participants remained in the same group throughout the
ambulatory exercise groups (eg, low group session fre- study.22 Table 1 gives the baseline characteristics of the ex-
quency and with low demand for training materials, ercise and control groups, and the Figure shows the partici-
rooms, and instructors). pant flow through the study.
180

659 Assessed for eligibility
363 Excluded
15 For diseases affecting bone metabolism or fall risk
239 For medication affecting bone metabolism or fall risk
13 For history of profound coronary heart diseases (stroke, cardiac events)
2 For acute or chronic inflammatory diseases
2 For secondary osteoporosis
10 Participated in exercise studies during past 2 y
2 Had very low physical capacity (<50 W during ergometry)
80 Refused to participate after detailed introduction to study protocol
296 Randomized (stratified for age)
SEFIP study SEFIP study 50 Participated in whole-body

vibration substudy
123 Allocated to exercise group 123 Allocated to control group
123 Received intervention 123 Received intervention
8 Lost to follow-up 11 Lost to follow-up

4 Lost interest 8 Lost interest
2 Moved 1 Died
2 Absent during follow-up testing 2 Absent during follow-up testing
Included in final statistical analysis
115 In exercise group 112 In control group
Figure. Flowchart for the 246 participants in the Senior Fitness and Prevention (SEFIP) study.
STUDY INTERVENTION 3. Functional gymnastics, isometric strength training, and

stretching sequences with 1 to 3 sets of isometric floor exer-
A high-intensity exercise program (primary intervention) was cises for trunk flexors and extensors, hip flexors and exten-
compared with an intermittent low-frequency wellness pro- sors, and leg abductors and adductors. Typically, 10 to 15 ex-
gram that focused on well-being. The study intervention took ercises were performed with 6 to 10 seconds of maximum
place from January 1, 2006, to September 30, 2007 (±1 week). intensity and 20 to 30 seconds of active rest. During the rest
Participants of both groups were supervised by certified train- periods, continuous stretching for the corresponding muscle
ers who were instructed weekly by the principal investigator group was performed. Exercises were replaced every 6 to 18
(W.K.). Individual training logs were analyzed every 10 weeks weeks by more strenuous ones.
to determine participant adherence to the exercise or wellness 4. Two to 3 sets of upper body exercises (low and high belt
program and attendance. For ethical reasons and because rowing and belt shoulder raises23) with 10-15 repetitions and
BMD was a primary study end point, calcium (1500 mg/d) a 2-second concentric–1-second static–2-second eccentric time
and cholecalciferol (500 IU/d) supplements (Opfermann Arz- under tension intermittent with 30 to 40 seconds of continu-
neimittel, Wiehl, Germany) were provided for all partici- ous stretching were executed using elastic belts (Thera-Band;
pants in both groups. However, adherence to the calcium Ludwig Artzt GmbH, Hadamar, Germany). Participants were
and cholecalciferol intake regimen was not determined in encouraged to perform at their maximum exertion level mi-
this study. Apart from the exercise intervention, subjects nus 2 repetitions. To ensure a progressive development of
were requested to maintain their usual lifestyle and exercise strength, different belts (3.5 kg, 4.5 kg, and 6 kg per 100% ex-
habits. tension) were used. Furthermore, 2 sets with 3 unilateral dy-
namic weight-bearing leg exercises were performed in a cir-
cuit mode with 1 minute of exercise and 1 minute of active rest,
EXERCISE PROGRAM with stretching exercises for the corresponding muscle groups.
Participants were asked to execute 8 repetitions per leg
The weekly exercise program consisted of two 60-minute su- (2-second–0-second–2-second time under tension mode) on
pervised group classes and two 20-minute home training ses- a maximum exertion level minus 2 repetitions. Intensity was
sions. Group classes were structured into 4 sequences: progressively increased by enlarging the amplitude of the move-
1. A warm-up/aerobic dance sequence with progressively ment, changing the velocity of the concentric execution, and
higher-impact elements performed for 20 minutes at 70% to introducing more strenuous exercises.
85% maximum heart rate, as assessed during a treadmill test
to a voluntary maximum. The home training session emphasized strength and flex-
2. Five minutes of static and dynamic balance training. ibility exercises, with 1 to 2 sets of 6 to 8 isometric exercises, 2
Briefly, exercises were performed in a standing position under to 3 belt exercises with 2 sets of 10 to 15 repetitions, and in-
conditions of progressively increasing postural instability. termittent stretching exercises. Every 12 weeks, a new home
181

training routine with more intense exercises replaced the ex- Baseline questionnaires were used to determine the partici-
isting protocol. Participants were regularly encouraged by the pants’ social status and living conditions; health status and medi-
instructors to consistently practice the home training session. cal conditions; osteoporotic and general health risk factors, in-
cluding falls and fall history; and prestudy physical activity and
exercise levels. To control changes caused by confounding fac-
WELLNESS PROGRAM tors (eg, medication, physical activity, and exercise level) and
adverse effects (defined as a harmful or abnormal result), fol-
To blind the participants, the control group performed a pro- low-up questionnaires and structured interviews were per-
gram that focused on well-being and was designed not to cause formed by the same physician.
physical adaptations. These participants executed a low-fre- Nutritional intake was assessed using 4-day dietary proto-
quency, low-intensity protocol for 60 minutes once a week for cols and analyzed with the use of Prodi-4.5/03 Expert software
10 weeks followed by 10 weeks of rest. This training-rest cycle (Wissenschaftlicher Verlag, Freiburg, Germany). Based on the re-
was repeated throughout the study. Each training session started sults, a maximum of 1500 mg of calcium and 500 IU of chole-
with 5 to 10 minutes of walking at 50% to 60% maximum heart calciferol were given to supplement the diets of all participants.
rate and finished with 10 minutes of a muscular relaxation se-
quence. The main topic changed from week to week. Within each
of the four 10-week blocks, the following activities were per- STATISTICAL ANALYSES
formed: relaxation, games/interaction, general coordination, en-
durance, balance, dances, body sensitivity, muscle strength, The estimated sample size was based on fall frequency. To de-
breathing, and flexibility. tect a 33% difference between groups in the fall rate during the
Endurance and strength-training sessions were executed with 18-month intervention(effect size=0.33), 115 participants per
low to moderate intensity without progressive increments of group were required for a 5% error probability with 80% sta-
exercise intensity or duration. tistical power (1-tailed). To adjust for subjects who were lost
to follow-up, 123 participants were included in each group. We
performed an intention-to-treat analysis that included all sub-
MEASUREMENTS jects with 18-month follow-up data.
The allocation sequence and group assignment were per-
Our primary outcome measures consisted of BMD at the lum- formed by the Institute of Biometry and Epidemiology. Par-
bar spine (LS) and proximal femur, fall frequency (fall rate), ticipants were enrolled by the Institute of Medical Physics.
projected 10-year CHD risk, and HCCs. Bone mineral density was log transformed to obtain the nor-
Our secondary outcome measures included the number of mally distributed data required for the analysis of variance with
fallers per group, number of fallers with injurious falls, and over- repeated measurements. This was performed as a mixed linear
all number of fractures per group. In addition, the following model, with the treatment group and time as fixed factors and
variables constituted the 10-year CHD risk score and were ana- with subjects as the random factor. By using a mixed linear model,
lyzed: low- and high-density lipoprotein cholesterol levels, blood the baseline values of participants who were lost to follow-up
pressure, smoking, and the presence of diabetes mellitus. could be included in the analysis. We report herein the P val-
Baseline and 18-month measurements were obtained by the ues of post hoc tests according to the Tukey-Kramer method.
same researcher and at the same time of day (±1 hour). All as- The number of falls (fall rate with 95% confidence inter-
sessments were determined in a blinded fashion. vals [CIs]) was compared between groups by means of nega-
Height was determined with a stadiometer, and weight was tive binominal regression.
measured on a digital scale while the participant wore minimal The 10-year CHD risk and health cost data could not be trans-
clothing. Body fat and BMD at the LS and proximal femur were formed to normally distributed data and were therefore ana-
assessed with dual-energy x-ray absorptiometry (QDR 4500, lyzed nonparametrically with the Wilcoxon rank sum test in
Discovery upgrade; Hologic, Bedford, Massachusetts) using stan- the 2-sample cases and with the Wilcoxon signed rank test in
dard protocols specified by the manufacturer. The coefficient the 1-sample cases. Contrary to the analysis of variance with
of variance for the dual-energy x-ray absorptiometry scan of repeated measurements, only complete data (ie, data from base-
the LS was 0.9%, and the corresponding value for the femoral line and follow-up), are evaluated with the nonparametric tests.
neck was 1.0%. Effect sizes were based on the absolute difference (SD) be-
Falls were defined according to the PROFANE (Prevention tween the baseline and 18-month follow-up values in the ex-
of Falls Network Europe) group.24 Injurious falls and overall ercise and control groups and were calculated using Cohen d.
fractures were monitored daily with the use of fall calendars All tests were 2-sided, with P⬍.05 considered statistically sig-
compiled by the participants. Outcome assessors contacted sub- nificant. Raw P values without further adjusting are pre-
jects who fell and nonresponders monthly by telephone. The sented. All statistical evaluations were performed with SAS soft-
6-month prestudy fall rate was recorded at the study start in ware (version 9.1; SAS Institute Inc, Cary, North Carolina).
response to a questionnaire.
The 10-year CHD risk was calculated according to the Framing-
ham Risk Calculator by Wilson et al.25 Briefly, points were added RESULTS
to each risk factor for sex, age, low- and high-density lipoprotein
cholesterol level, systolic and diastolic blood pressure, diabetes The Figure shows the participant flow during the SEFIP
mellitus, and smoking. Based on a score sheet, the 10-year CHD study. All participants are included in the final analysis;
risk for each subject was given as a percentage value. 115 women in the exercise group (93.5%) and 112 par-
In close collaboration with Siemens Health Insurance ticipants in the controls (91.1%) provided follow-up out-
(Siemens Betriebs Krankenkasse), total HCCs (excluding den-
tal costs) were determined for the 6 months before the study
comes data. Twelve participants lost interest; 8 of them
and for the 18-month intervention. Final data for HCCs were cited study-related reasons with respect to the exercise
delivered in June 2008. Dental costs for the 18-month inter- protocol (4 women in the exercise group and 3 con-
vention were comparable between the groups (€527 for the ex- trols) or because of the calcium and cholecalciferol supple-
ercise group vs €511 for the controls). (At the time of the study, mentation (1 woman in the control group). Two partici-
€1⬇$1.50.) pants moved to a different city and 1 woman died. Four
182

Table 2. Changes in the Exercise and Control Group
Mean (SD)
Exercise Control Absolute Difference, Effect Size,

Variable (n=115) (n = 112) Mean (95% CI) P Value a Cohen d
BMD at LS, g/cm2
Baseline 0.919 (0.157) 0.927 (0.149)
18 mo 0.936 (0.167) 0.930 (0.152)
Difference 0.017 (0.027) 0.003 (0.028) 0.014 (0.006 to 0.021) ⬍.001 0.49
BMD at femoral neck, g/cm2
Baseline 0.706 (61.1) 0.703 (0.112)
18 mo 0.713 (54.6) 0.696 (0.110)
Difference 0.007 (0.024) −0.008 (0.024) 0.015 (0.008 to 0.021) ⬍.001 0.60
Overall fall rate
6 mo before study 0.37 (0.68) 0.41 (0.74) 0.04 (−0.22 to 0.14) .66 ⬍0.10
18-mo total 1.00 (1.37) 1.66 (1.79) 0.66 (0.25 to 1.07) .002 0.41
10-y CHD risk, %
Baseline 10.5 (4.2) 11.2 (5.0)
18 mo 8.5 (3.4) 10.1 (4.7)
Difference −1.96 (3.80) −1.15 (2.84) 0.81 (−0.08 to 1.70) .22 0.24
Health care costs, € b
6 mo before study 833 (1397) 869 (1589) −36 (−413 to 340) .85 ⬍0.10
18-mo total 2255 (2596) 2780 (3318) 525 (−1273 to 380) .20 0.18
Abbreviations: BMD, bone mineral density; CHD, coronary heart disease; LS, lumbar spine.
a Significance levels are given for between-group differences only. Further information is listed in the “Results” section.
b €1 ⬇ $1.50.
participants were unable to attend the final assessment. riod. Furthermore, relative risk was calculated as 0.54
The mean (SD) overall attendance rate in the exercise (95% CI, 0.35-0.84; P=.01) for subjects who fell and 0.33
group was 76.3% (8.1%) for the group training session (0.15-0.74; P =.01) for those who experienced injurious
(for the controls, 72.0% [8.7%]) and 42.2% (5.3%) for falls in the exercise group. Overall fractures due to falls
the home training session. Furthermore, as determined were twice as high in the controls (n=12) than in the ex-
by training logs kept by both groups, adherence to the ercise group (n=6). There were no fall-induced spine or
exercise (exercise execution, strain thresholds, modali- proximal femur fractures, and the group difference was
ties, and progression) and wellness protocols was excel- not significant (P =.90).
lent for completion of the sessions performed. The 10-year CHD risk significantly (P ⬍ .001) im-
Other than the intervention, no changes in physical proved in both groups (exercise group: −1.96% [95% CI,
activity or exercise level were determined for either group. −2.69% to −1.23%] vs controls: −1.15% [−1.69% to
No musculoskeletal injuries, falls, or CHD events oc- −0.62%]; P=.22). Changes in 10-year CHD risk in the
curred during the training sessions. In addition, no ad- exercise group were primarily based on significant changes
verse effects (ie, pain or quality-of-life reductions) of our in high-density lipoprotein cholesterol level (exercise
exercise or wellness protocol were observed. group: 6.5% [95% CI, 4.3% to 8.7%] vs controls: 1.8%
Blinding of the control group was successful; 81.3% of [3.1% to 6.0%]; P=.002), and low-density lipoprotein cho-
the participants believed they were part of the primary in- lesterol level (exercise group: −1.9% [−4.5% to 0.7%] vs
tervention. In the exercise group, all of the participants exercise group: 3.1% [−0.1% to 6.3%]; P=.02). Both groups
believed they were part of the primary intervention. further experienced significant changes in systolic (ex-
Table 2 shows between-group differences and the nu- ercise group: −3.5% [95% CI, −5.8% to −1.3%] vs con-
merical changes of the end points. At baseline, none of trols: −4.8% [−7.1% to −2.5%]; P =.43) and diastolic (ex-
the differences were significant. ercise group: −8.7% [−10.9% to −6.6%] vs controls: −7.6%
Mean (95% CI) BMD at the LS (1.77% [1.26% to 2.28%]) [−9.9% to −5.3%]; P=.48) blood pressure that were com-
and femoral neck (1.01% [0.37% to 1.65%]) significantly parable between the groups.
increased in the exercise group (P⬍.001 for both), whereas, Other variables (diabetes and smoking) contributing to
among the controls, BMD did not significantly change at the 10-year CHD risk score were not significantly affected.
the LS (0.33% [−0.24% to 0.91%]; P = .25) and signifi- Prestudy HCCs did not differ between the groups
cantly decreased at the femoral neck (−1.05% [−1.70% to (Table 2). At 18 months, HCCs per participant were higher
−0.40%]; P⬍.001). Significant group differences (P⬍.001) in the control group (exercise group: €2255 [95% CI,
were determined for both measures (Table 2). €1791-€2718] vs controls: €2780 [€2187-€3372]; P=.20
Prestudy fall rates (falls per person per group for a [not included were dental costs, which were comparable
6-month period) did not differ between groups (Table 2). between groups]), but between-group differences were not
However, significant differences between the exercise significant. These costs included the intervention costs per
group (1.00 [95% CI, 0.76-1.24]) and controls (1.66 [1.33- participant of €383 (€328-€437) in the exercise group and
1.99]) were observed for the 18-month intervention pe- €175 (€134-€216) in the controls.
183

Significant changes in the confounding variables (life- cies used in the control group apparently still had an effect
style and nutritional intake) were not observed during on health factors with low strain thresholds (eg, blood
the intervention period. However, 8 women in the ex- pressure)31 and may have prevented significant group dif-
ercise group and 10 in the controls modified their blood ferences for 10-year CHD risk and HCCs. As with previ-
pressure, cholesterol level, and dosage of hypothyroid- ous exercise studies, it is difficult to assess the role of cross-
ism or hyperthyroidism medication. over and inadequate blinding in mediating or masking
effects. From a statistical point of view, it may be prefer-
able to focus on just one primary end point. However, un-
COMMENT
like pharmaceutical studies with restricted applicability,
the central aim of this study was to determine the effect
The results of our exercise program clearly demonstrated of an exercise program on the multiple morbidity situa-
positive effects on the most relevant risk factors (fracture tion of an elderly cohort.
and CHD risk factors) for elderly women, although the CHD With respect to BMD,13,15,32 falls,16,17,33 and CHD risk,9,34,35
risk factors were not significantly better compared with the impact of the observed positive changes was compa-
those of the control group. The 50%-lower fracture inci- rable to more specific interventions. Besides these clinical
dence in the exercise group was probably a result of the end points, this study also demonstrated a trend toward
decreased fall rate in conjunction with increased BMD. lower HCCs in the exercise group (exercise group: €2255
However, this surprising result was not significant, and vs controls: €2780), although the difference did not reach
fracture rate was not a main study end point. Even the con- significance. Two main reasons may prevent a higher sta-
trol group, who participated in the wellness intervention, tistical power of our study in this area: First, the unin-
significantly reduced their 10-year CHD risk, mainly tended positive effects on the primary study end points in
through a favorable effect on blood pressure. Thus, con- the control group may affect HCCs in this group. Second,
trary to the high-impact, high-intensity strategy required the total HCCs for our cohort were far below the €6100
to affect BMD,14 the rather smooth effort exerted in the calculated by the Statistical Federal Bureau as the average
wellness program was apparently able to significantly affect for German women who are 65 to 80 years old.36
blood pressure in this cohort. In summary, this contribution extends the existing data
We performed an exercise program with fewer group in that a single multipurpose exercise program that is based
sessions than typically recommended26,27 but with high ex- on a low-volume, high-intensity philosophy and is de-
ercise intensity during the aerobic and strength section. signed for the elderly improves overall fitness, maintains
Although this strategy was successful for our clinical end bone health, and reduces fall risk. Because this training
points, one may worry whether this approach increases regimen can be easily adopted by other institutions and
joint or low back pain in elderly subjects15 and may there- health care providers, a broad implementation of this pro-
fore be inadequate in this population. However, similar gram is feasible.
to our recent high-intensity exercise program for early post-
menopausal women,28 any negative changes in pain indi- Accepted for Publication: November 18, 2009.
cators or quality of life were detected in the present study. Correspondence: Wolfgang Kemmler, PhD, Institute of
This result was indirectly supported by a low dropout rate Medical Physics, Friedrich-Alexander University of Er-
and a moderate to high attendance rate.29,30 langen-Nuremberg, Henkestrasse 91, 91052 Erlangen,
Our study possesses several strengths. Our cohort was Germany (wolfgang.kemmler@imp.uni-erlangen.de).
a homogeneous community-dwelling group of women Author Contributions: Dr Kemmler had full access to
65 years or older. The blinding strategy was successful. all the data in the study and takes responsibility for the
The study duration was sufficiently long to detect rel- integrity of the data and the accuracy of the data analy-
evant changes of physiological variables. Potential ef- sis. Study concept and design: Kemmler, von Stengel, En-
fects of lifestyle changes, diseases, medication, and nu- gelke, and Kalender. Acquisition of data: Kemmler, von
trition were strictly controlled. The exercise regimen was Stengel, and Engelke. Analysis and interpretation of data:
progressively augmented during the intervention pe- Kemmler and Häberle. Drafting of the manuscript:
riod, and the group sessions were strictly supervised by Kemmler, von Stengel, Engelke, and Kalender. Critical
certified trainers. The favorable attendance at the group revision of the manuscript for important intellectual con-
sessions and the favorable dropout rates30 indicate the tent: Kemmler, von Stengel, Engelke, and Häberle. Sta-
attractiveness of the exercise program. However, the low tistical analysis: Häberle. Obtained funding: Kemmler, von
adherence to the home training reported by the exercise Stengel, Engelke, and Kalender. Administrative, techni-
group participants again demonstrates the reluctance of cal, and material support: Kemmler, von Stengel, En-
most elderly individuals to exercise on their own. Health gelke, and Kalender. Study supervision: Kemmler and von
care costs were directly assessed rather than estimated Stengel.
or interpolated. Finally, the demands and relative costs Financial Disclosure: None reported.
for training materials were low; thus, the program can Funding/Support: This study was supported by the Siemens
be easily adopted by others. Betriebs Krankenkasse, Behinderten- und Rehabilitations-
There are potential limitations to our study. Although Sportverband Bayern, Netzwerk Knochengesundheit e.V.,
not reflected by physical fitness measures (strength and Opfermann Arzneimittel GmbH, Thera-Band, Institute of
aerobic capacity did not favorably change in the control Sport Science, and Institute of Medical Physics.
group) or by the responses to the physical activity ques- Role of the Sponsors: The funding organizations played
tionnaire, the rather low exercise intensities and frequen- no role in the design or conduct of study. Siemens Be-
184

triebs Krankenkasse was responsible for the collection in older people living in the community. Cochrane Database Syst Rev. 2009;
(2):CD007146.
of the total HCCs; however, none of the funding part-
18. Baechle TR. Essentials of Strength Training and Conditioning. Champaign, IL:
ners influenced the analyses, the interpretation of the Human Kinetics; 1994.
data, or the preparation, review, or approval of the 19. Bompa TO. Periodization: Theorie and Methodology of Training. Champaign, IL:
manuscript. Human Kinetics; 1999.
Additional Contributions: Jerry Mayhew, PhD, of Tru- 20. Marcus R. Exercise: moving in the right direction. J Bone Miner Res. 1998;13(12):
man State University, Kirksville, Missouri, provided help- 1793-1796.
21. von Stengel S, Kemmler W, Engelke K, Bebenek M, Mayhew JL, Kalender WA.
ful critical remarks.
Effects of whole body vibration training on the fracture risk of postmenopausal wom-
en: preliminary results of the Erlangen Longitudinal Vibration Study (ELVIS) [abstract].
REFERENCES Med Sci Sports Exerc. 2008;40(5)(suppl):S313. doi:10.1249/01.mss.0000323210
.26037.f0.
22. Alberti KG, Zimmet P, Shaw J. Metabolic syndrome: a new world-wide definition:
1. Tesch-Römer C. Zukunftsfähige Altersvorsorge: deutsche und internationale
Perspektiven. Paper presented at: Lebenslagen älterer Menschen in Deutsch- a consensus statement from the International Diabetes Federation. Diabet Med.
land; November 30, 2004; Berlin, Germany. 2006;23(5):469-480.
2. Wiesner G, Bittner E. Zur Inzidenz und Prävalenz von Mehrfachkrankheiten in 23. Pearl B. Keys to the Inner Universe: Encyclopedia on Weight Training. Vol 1. Phoe-
Deutschland. Arbeitsmed Sozialmed Umweltmed. 2005;40(9):490-498. nix, AZ: Bill Pearl Enterprises; 1982.
3. Pedersen BK, Saltin B. Evidence for prescribing exercise as a therapy in chronic 24. Lamb SE, Jorstad-Stein EC, Hauer K, Becker C; Prevention of Falls Network Eu-
disease. Scand J Med Sci Sports. 2006;16(suppl 1):3-63. rope and Outcomes Consensus Group. Development of a common outcome data
4. Boulé NG, Haddad E, Kenny GP, Wells GA, Sigal RJ. Effects of exercise on gly- set for fall injury prevention trials: the Prevention of Falls Network Europe
cemic control and body mass in type 2 diabetes mellitus: a meta-analysis of con- consensus. J Am Geriatr Soc. 2005;53(9):1618-1622.
trolled clinical trials. JAMA. 2001;286(10):1218-1227. 25. Wilson PW, D’Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB.
5. Sigal RJ, Kenny GP, Wasserman DH, Castaneda-Sceppa C. Physical activity/ Prediction of coronary heart disease using risk factor categories. Circulation. 1998;
exercise and type 2 diabetes. Diabetes Care. 2004;27(10):2518-2539. 97(18):1837-1847.
6. Cornelissen VA, Fagard RH. Effects of endurance training on blood pressure, blood 26. Erlichman J, Kerbey AL, James WP. Physical activity and its impact on health out-
pressure–regulating mechanisms, and cardiovascular risk factors. Hypertension. comes: paper 1: the impact of physical activity on cardiovascular disease and all-
2005;46(4):667-675. cause mortality: an historical perspective. Obes Rev. 2002;3(4):257-271.
7. Kelley GA, Kelley KS. Progressive resistance exercise and resting blood pres- 27. Kohrt WM, Bloomfield SA, Little KD, Nelson ME, Yingling VR; American College
sure: a meta-analysis of randomized controlled trials. Hypertension. 2000; of Sports Medicine. American College of Sports Medicine Position Stand: physi-
35(3):838-843. cal activity and bone health. Med Sci Sports Exerc. 2004;36(11):1985-1996.
8. Houde SC, Melillo KD. Cardiovascular health and physical activity in older adults: 28. Kemmler W, Lauber D, Weineck J, Hensen J, Kalender W, Engelke K. Benefits of 2
an integrative review of research methodology and results. J Adv Nurs. 2002; years of intense exercise on bone density, physical fitness, and blood lipids in early
38(3):219-234. postmenopausal osteopenic women: results of the Erlangen Fitness Osteoporosis
9. Kelley GA, Kelley KS, Tran ZV. Aerobic exercise and lipids and lipoproteins in wom- Prevention Study (EFOPS). Arch Intern Med. 2004;164(10):1084-1091.
en: a meta-analysis of randomized controlled trials. J Womens Health (Larchmt). 29. Asikainen TM, Kukkonen-Harjula K, Miilunpalo S. Exercise for health for early
2004;13(10):1148-1164. postmenopausal women: a systematic review of randomised controlled trials.
10. Stiegler P, Cunliffe A. The role of diet and exercise for the maintenance of fat-free
Sports Med. 2004;34(11):753-778.
mass and resting metabolic rate during weight loss. Sports Med. 2006;36(3):
30. Wallace BA, Cumming RG. Systematic review of randomized trials of the effect of
239-262.
exercise on bone mass in pre- and postmenopausal women. Calcif Tissue Int. 2000;
11. Toth MJ, Beckett T, Poehlman ET. Physical activity and the progressive change
67(1):10-18.
in body composition with aging: current evidence and research issues. Med Sci
31. Asikainen TM, Miilunpalo S, Oja P, et al. Randomised, controlled walking trials in
Sports Exerc. 1999;31(11)(suppl):S590-S596.
postmenopausal women: the minimum dose to improve aerobic fitness? Br J Sports
12. Kay SJ, Fiatarone Singh MA. The influence of physical activity on abdominal fat:
Med. 2002;36(3):189-194.
a systematic review of the literature. Obes Rev. 2006;7(2):183-200.
13. Bonaiuti D, Shea B, Iovine R, et al. Exercise for preventing and treating osteoporo- 32. Kelley GA. Aerobic exercise and bone density at the hip in postmenopausal wom-
sis in postmenopausal women. Cochrane Database Syst Rev. 2002;(3):CD000333. en: a meta-analysis. Prev Med. 1998;27(6):798-807.
14. Martyn-St James M, Carroll S. High-intensity resistance training and postmeno- 33. Carter ND, Kannus P, Khan KM. Exercise in the prevention of falls in older people:
pausal bone loss: a meta-analysis. Osteoporos Int. 2006;17(8):1225-1240. a systematic literature review examining the rationale and the evidence. Sports
15. Vuori IM. Dose-response of physical activity and low back pain, osteoarthritis, Med. 2001;31(6):427-438.
and osteoporosis. Med Sci Sports Exerc. 2001;33(6)(suppl):S551-S586, dis- 34. Kelley GA, Kelley KA, Tran ZV. Aerobic exercise and resting blood pressure: a
cussion 609-610. meta-analytic review of randomized, controlled trials. Prev Cardiol. 2001;4
16. Gillespie LD, Gillespie WJ, Robertson MC, Lamb SE, Cumming RG, Rowe BH. (2):73-80.
Interventions for preventing falls in elderly people [update of: Cochrane Data- 35. Stewart KJ, Bacher AC, Turner K, et al. Exercise and risk factors associated with
base Syst Rev. 2001;(3):CD000340]. Cochrane Database Syst Rev. 2003;(4): metabolic syndrome in older adults. Am J Prev Med. 2005;28(1):9-18.
CD000340. 36. Statistisches Bundesamt. Gesundheit in Deutschland. Berlin, Germany: Statis-
17. Gillespie LD, Robertson MC, Gillespie WJ, et al. Interventions for preventing falls tisches Bundesamt; 2006.
185

Physical Activity at Midlife in Relation to Successful

Survival in Women at Age 70 Years or Older
Qi Sun, MD, ScD; Mary K. Townsend, ScD; Olivia I. Okereke, MD; Oscar H. Franco, MD, ScD, PhD;
Frank B. Hu, MD, PhD; Francine Grodstein, ScD
Background: Physical activity is associated with re- increases in successful survival were observed begin-
duced risks of chronic diseases and premature death. ning at the third quintile of activity: odds ratios (ORs)
Whether physical activity is also associated with im- (95% confidence intervals [CIs]) in the lowest to high-
proved overall health among those who survive to older est quintiles were 1 [Reference], 0.98 (0.80-1.20), 1.37
ages is unclear. (1.13-1.65), 1.34 (1.11-1.61), and 1.99 (1.66-2.38)
(P⬍.001 for trend). Increasing energy expenditure from
Methods: A total of 13 535 Nurses’ Health Study par- walking was associated with a similar elevation in odds
ticipants who were free of major chronic diseases at base- of successful survival: the ORs (95% CIs) of successful
line in 1986 and had survived to age 70 years or older as survival across quintiles of walking were 1 [Reference],
of the 1995-2001 period made up the study population. 0.99 (0.80-1.21), 1.19 (0.97-1.45), 1.50 (1.24-1.82), and
We defined successful survival as no history of 10 major 1.47 (1.22-1.79) (P ⬍.001 for trend).
chronic diseases or coronary artery bypass graft surgery
and no cognitive impairment, physical impairment, or Conclusion: These data provide evidence that higher lev-
mental health limitations. els of midlife physical activity are associated with excep-
tional health status among women who survive to older
Results: After multivariate adjustment for covariates, ages and corroborate the potential role of physical activ-
higher physical activity levels at midlife, as measured by ity in improving overall health.
metabolic-equivalent tasks, were significantly associ-
ated with better odds of successful survival. Significant Arch Intern Med. 2010;170(2):194-201
T
HE PAST CENTURY HAS WIT- nesses.3-6 Indeed, limited epidemiologic
nessed a dramatic increase studies conducted primarily among older
in life expectancy in the male populations have identified several
United States, from 47.3 modifiable midlife risk factors, such as
years in 1900 to 75.2 years smoking and obesity, associated with the
for men and 80.4 years for women in 2005.1 probability of exceptional health among
Together with a decreased birth rate and those who survive to older ages.3-5,7
the aging of baby boomers, it is projected Physical activity is a well-established ap-
that by 2030, 1 in every 5 Americans will proach to reducing risks of many chronic
be 65 years or older.2 Older adults are dis- diseases,8-12 and potentially other aspects
Author Affiliations: proportionately affected by chronic dis- of health.13-19 However, findings from lim-
Departments of Nutrition eases and functional disabilities, and the at- ited existing studies of the relation be-
(Drs Sun and Hu) and tendant medical and social costs are tween midlife physical activity and over-
Epidemiology (Drs Townsend, all health and well-being at older ages have
Hu, and Grodstein), Harvard been inconsistent. For example, several
School of Public Health, Boston, See also pages 124, 170, studies among men and women found that
Massachusetts; Channing 179, and 186 physical activity increased healthy ag-
Laboratory, Department of ing,3,7 disability-free survival,20 or self-
Medicine, Brigham and tremendous.2 However, development of reported physical and overall health,21
Women’s Hospital and Harvard
chronic diseases and disabilities is not in- while a more recent study of Japanese
Medical School, Boston
(Drs Okereke, Hu, and evitable among aged populations. Several American men reported a null associa-
Grodstein); and Health Sciences studies have demonstrated that as many as tion with successful survival.5 Moreover,
Research Institute, University of 10% to 50% of those 65 years or older can evidence specifically among women is
Warwick, Coventry, England maintain physical and cognitive integrity lacking, despite the fact that women live
(Dr Franco). and remain free of major chronic ill- longer than men; thus, identifying risk fac-
194

tors for successful survival is particularly important among The physical activity questionnaire has been validated in a
women. Finally, limited research has addressed the dose- similar population (the NHS II).25 In a representative sample
response relationship and intensity of activities in rela- of 147 nurses, the physical activity scores based on this ques-
tion to successful survival. tionnaire administered 2 years apart were reasonably corre-
lated, given some true changes in activity across 2 years; the
Herein, we use data from the Nurses’ Health Study
test-retest correlation coefficient (r) was 0.59. The question-
(NHS)22 to further explore the relation between midlife naire estimate of physical activity levels was highly correlated
physical activity, including walking, and successful ag- with those reported in 1-week recalls (r=0.79) and those logged
ing as measured by a full spectrum of health outcomes, in diaries during the year (r=0.62).
including incidence of chronic diseases, cognitive and
physical functioning, and mental status.23 ASCERTAINMENT OF CHRONIC DISEASES
METHODS A wide variety of major chronic diseases (ie, cancer, diabetes,

coronary heart disease, stroke, Parkinson disease, and mul-
tiple sclerosis) were reported by participants in 1976 and in
STUDY POPULATION biennial follow-up questionnaires. The self-reports were con-
firmed by study physicians through a variety of methods, such
The NHS is an ongoing prospective cohort study initially com- as medical record review, pathology report review, telephone
prising 121 700 female registered nurses, aged 30 to 55 years, interview, or supplementary questionnaire inquiries. The self-
who responded to a baseline questionnaire in 1976. Fol- report of incidence of chronic diseases among these nurses has
low-up questionnaires have been administered to the partici- been previously demonstrated to be highly valid.26-29
pants every 2 years since 1976 to collect and update the infor-
mation on incidence of diseases and demographic and lifestyle
risk factors. In 1986, we started collecting detailed informa- ASSESSMENT OF PHYSICAL FUNCTION
tion on physical activity. Through 2000, the close of fol- AND MENTAL HEALTH
low-up for most participants in the present analyses, the fol-
low-up rate was greater than 95%. In 1992, 1996, and 2000, we added the Medical Outcomes Sur-
vey Short-Form Health Survey (SF-36) to the follow-up ques-
tionnaires to assess the physical and mental status of the par-
ASSESSMENT OF PHYSICAL ACTIVITY ticipants. The SF-36 is a 36-item questionnaire that measures
eight health concepts, including limitations of physical activi-
In 1986, we inquired about the average time per week in the past ties, usual role activities, social activities, mental health, bodily
year participants spent on leisure-time physical activities, includ- pain, vitality, and general health perceptions. The validity and
ing walking or hiking outdoors; jogging (ⱖ10 min/mile); run- reproducibility of the SF-36 have been extensively examined
ning (⬍10 min/mile); bicycling; lap swimming; playing tennis; and reported elsewhere.30
doing calisthenics, aerobics, aerobic dance, and/or rowing ma-
chine exercise; and playing squash or racquet ball. For each ques-
tion, there were 10 possible response categories (range, 0 to ⱖ11 ASSESSMENT OF COGNITIVE FUNCTION
h/wk). Furthermore, we inquired about flights of stairs climbed
each day, and, for walkers, the usual walking pace: easy or ca- From 1995 to 2001, we invited all nurses 70 years or older who
sual (⬍2.0 mph), normal (2.0-2.9 mph), brisk (3.0-3.9 mph), and were free of stroke to participate in a cognitive function study.
very brisk (ⱖ4.0 mph). Based on this information, we calculated Of 21 202 invited nurses, 19 415 (92%) agreed to participate and
energy expenditure in metabolic-equivalent tasks (METs) mea- were administered the Telephone Interview for Cognitive Sta-
sured in hours per week.24 Each MET-hour is the caloric need tus (TICS),31 which is modeled on the Mini-Mental State Ex-
per kilogram of body weight per hour of activity divided by the amination.32 Scores on the TICS have a range of 0 (worst) to 41
caloric need per kilogram of weight per hour at rest. According (perfect), with a score lower than 31 indicating cognitive im-
to this standard, we assigned a MET value of 12.0 to running; 8.0 pairment.31 The high test-retest reliability and validity of TICS
to stair-climbing; 7.0 to jogging, bicycling, lap swimming, and compared with in-person cognitive testing have been demon-
playing tennis and other racquet sports; 6.0 to aerobics and cal- strated previously.32 Trained study nurses who were unaware of
isthenics; and 2.5 to 4.5 to walking, depending on the pace. In the study hypothesis and exposure status of the participants ad-
other words, for example, running for an hour would generate ministered the TICS with high inter-interviewer reliability.31 Ow-
12 METs’ energy expenditure; climbing stairs for an hour would ing to the availability of cognitive data from this group, the present
generate 8 METs’ energy expenditure, and so on. The same amount analysis was conducted among these participants.
of energy expenditure can be achieved by various physical ac-
tivities. For example, to achieve 30 METs/wk, a woman can run DEFINITION AND ASCERTAINMENT
for 2.5 h/wk or swim for 4.3 h/wk. In analyses of activity inten- OF SUCCESSFUL AGING
sity, we defined activity with a MET value larger than 6 as vig-
orous; walking was defined as a moderate-intensity activity ow- To evaluate the overall health status of the study participants,
ing to the lower MET value. we used the concept of successful aging first outlined by Rowe
For the current analysis, we used 1986, when detailed and Kahn,23 which takes into account both comorbidities and
physical activity information was first obtained, as the study disabilities. The working definition of successful aging has been
baseline. Moreover, in all analyses, we only considered introduced in detail elsewhere.33 Briefly, our definition of suc-
physical activity reported in 1986 because we wanted to cessful aging addressed 4 domains: (1) no history of cancer (ex-
minimize the possibility of reverse causation with aging (ie, cept nonmelanoma skin cancer), diabetes, myocardial infarc-
if poor underlying health status caused decreased physical tion, coronary artery bypass graft surgery (CABG), congestive
activity rather than the opposite). At baseline in 1986, the heart failure, stroke, kidney failure, chronic obstructive pulmo-
mean age was 60 years for our study participants, and there- nary disease, Parkinson disease, multiple sclerosis, or amyotro-
fore, midlife was defined as age 60 years for the purposes of phic lateral sclerosis; (2) no impairment in cognitive function
this report. (TICS score ⱖ31); (3) no physical disabilities (no limitations on
195

moderate activities and no more than moderate limitations on still possible that long-term physical disabilities at baseline might
more demanding physical performance measures); and (4) no have biased our analysis. To address this issue, rather than com-
mental health limitations (mental health score ⬎84, which is the pare women with the least and the most activity, we repeated
median score in our study population). Any participant who sur- the analysis only within participants who reported having at
vived to at least age 70 years and met all these criteria was de- least a minimum level of activity, which we defined as walk-
fined as a successful survivor; the remaining participants who sur- ing at least 1 hour per week or performing any vigorous activ-
vived to at least aged 70 years but had a chronic disease history, ity at least 20 minutes per week at baseline. Second, to best ad-
CABG, cognitive impairment, physical or mental health limita- dress the independent effects of walking as exercise, we estimated
tions were defined as usual survivors. Since the cognitive func- the ORs associated with walking METs after excluding women
tion of most study participants was assessed in the 1999-2000 who both walked and participated in vigorous activity. Fi-
period (87.5%), we used the year 2000 to define chronic dis- nally, to examine the robustness of our definition of success-
ease status. Similarly, physical and mental health domains were ful aging, we repeated the analysis using an alternative defini-
primarily derived from the SF-36 administered in 2000. tion that included the same criteria for chronic disease status,
We excluded nurses who had a history of any of the rel- but used median score to define the cut points for cognitive,
evant chronic diseases or CABG at baseline (n=2361) or who physical, and mental health domains.33 We conducted this analy-
had missing physical activity data at baseline (n=2724). We sis because, while the domains we used for considering suc-
further excluded those who skipped more than 2 items on the cessful survival are widely accepted, the specific criteria for de-
mental health scale at 70 years or older or more than 5 items fining “successful” within each domain is less established.
on the physical function scale in the SF-36 (n=795). After these
participants were excluded, data from 13 535 nurses were avail-
RESULTS
able for analysis. All participants gave informed consent. The
study protocol was approved by the institutional review board
of the Brigham and Women’s Hospital. PRIMARY ANALYSIS
STATISTICAL ANALYSIS Of the total of 13 535 participants, 1456 (10.8%) met the
criteria for successful survivor. Table 1 summarizes the
We grouped the study participants into quintiles of total METs. baseline characteristics of the participants in 1986. As ex-
We used logistic regression to assess the odds ratios (ORs) of pected, successful survivors were more active than usual
successful survival vs usual survival associated with each quin- survivors. The successful survivors were also leaner and
tile, defining the lowest quintile as the reference level. In mul- less likely to smoke than usual survivors and had a slightly
tivariate logistic regression models, we adjusted for variables lower prevalence of hypertension or high cholesterol levels.
defined in 1986, including age at baseline (in years); educa- Table 2 summarizes the age- and multivariate-
tion (registered nurse, bachelor’s degree, master’s degree, or doc-
adjusted ORs of successful survival associated with quin-
torate); marital status (unmarried, married, widow, separated,
or divorced); if married, husband’s education (less than high tiles of total physical activity METs and walking METs.
school, some high school, high school graduate, college gradu- After adjustment for multiple covariates, the ORs for suc-
ate, or graduate school); postmenopausal hormone use (never, cessful survival across quintiles were 1 [reference], 0.98,
past, or current use); smoking status (never, past, current 1-14 1.37, 1.34, and 1.99 for total METs (P ⬍ .001 for trend).
cigarettes/d, current 15-24 cigarettes/d, or current ⱖ25 cigarettes/ We also found associations of similar strength between
d); family history of heart disease, diabetes, or cancer (yes or walking METs and the odds of successful aging. After mul-
no); dietary polyunsaturated to saturated fat ratio (in quin- tivariate adjustment of covariates, ORs for successful sur-
tiles); intakes of trans fat, alcohol, and cereal fiber (all in quin- vival across walking METs quintiles were 1 [reference],
tiles); and intakes of fruits and vegetables and red meat (in ter- 0.99, 1.19, 1.50, and 1.47 (P⬍.001 for trend). Further
tiles). Since moderate-intensity physical activity such as walking
adjustment for possible intermediate variables, such as
was associated with lower risk of chronic diseases in previous
studies by our research group,8,10 we further examined walk- body mass index (BMI), history of hypertension, and his-
ing METs and pace in relation to successful survival in the tory of hypercholesterolemia, did not change these as-
present study. When examining the associations for walking sociations materially.
MET quintiles, we further adjusted for vigorous activity METs Independent of the total physical activity levels, in-
to minimize potential confounding by vigorous physical activ- creasing walking pace was also strongly associated with
ity. Similarly, when we examined the associations for walking a significant increase in odds of successful aging
pace, we further controlled for total METs. (Table 3). Compared with women whose walking pace
Tests of linear trend across increasing MET quintiles were was easy, women with a moderate walking pace had a
conducted by treating the quintiles as a continuous variable and 90% increase in the odds of successful aging; women
assigning the median score for each quintile as its value. All P
whose walking pace was brisk or very brisk had 2.68-
values were 2 sided. Ninety-five percent confidence intervals
(95% CIs) were calculated for ORs. Data were analyzed with fold increased odds. To help disentangle the effects of the
the Statistical Analysis Systems software package, version 9.1 amount walked on the association with walking pace, we
(SAS Institute Inc, Cary, North Carolina). stratified the analysis by lower or higher levels of walk-
ing METs. Walking pace was similarly associated with
increased odds of successful aging for both groups.
SENSITIVITY ANALYSES
Acknowledging the interrelationship between BMI and
We performed 3 secondary sensitivity analyses to examine the physical activity, we also examined the joint associations
robustness of observed associations. First, although we ex- of BMI in 1986 and total physical activity with successful
cluded anyone with major chronic diseases at baseline, and im- survival (Figure). The positive associations between physi-
posed an average 14-year lag period between the assessment cal activity and successful aging persisted within each BMI
of activity levels and the assessment of successful survival, it is category (calculated as weight in kilograms divided by height
196

Table 1. Baseline Characteristics of Successful Survivors and Usual Survivors in the Nurses’ Health Study22
Successful Survivors Usual Survivors

Characteristic (n = 1456) a (n = 12 079) a P Value b
Age at baseline, y 60.1 (2.5) 60.6 (2.5) ⬍.001
Age at cognitive function assessment, y 73.7 (2.1) 74.2 (2.3) ⬍.001
Physical activity, METs (h/wk) 19.1 (22.0) 14.1 (19.7) ⬍.001
Walking activity, METs (h/wk) 9.5 (11.5) 7.2 (9.7) ⬍.001
BMI 23.8 (3.3) 25.5 (4.4) ⬍.001
Waist circumference, cm 76.2 (8.2) 80.6 (10.3) ⬍.001
Waist to hip ratio 0.78 (0.08) 0.79 (0.07) ⬍.001
Saturated fat intake, g/d 12.5 (2.5) 12.9 (2.4) ⬍.001
Polyunsaturated fat intake, g/d 5.8 (1.3) 5.9 (1.3) .003
Polyunsaturated to saturated fat ratio 0.50 (0.15) 0.50 (0.14) .07
trans Fat intake, g/d 1.8 (0.6) 1.9 (0.5) ⬍.001
Alcohol intake, g/d 7.1 (9.9) 6.7 (9.9) .15
Cereal fiber intake, g/d 4.2 (2.1) 4.0 (2.0) .004
Red meat intake, serving/d 1.0 (0.5) 1.1 (0.5) ⬍.001
Fruits and vegetables, serving/d 5.4 (2.0) 5.2 (2.0) ⬍.001
Smoking status c ⬍.001
Never smoked 53.5 46.4
Past smoker 34.8 36.2
Current smoker, 1-14 cigarettes/d 5.7 6.3
Current smoker, 15-24 cigarettes/d 4.3 6.9
Current smoker, ⱖ25 cigarettes/d 1.7 4.2
Education ⬍.001
Registered nurse 74.0 79.0
Bachelor’s degree 17.3 14.9
Master’s degree 8.0 5.8
Doctorate 0.7 0.4
Husband’s education c .01
Less than high school 2.2 2.6
Some high school 5.4 6.4
High school graduate 39.4 42.9
College graduate 28.7 27.6
Graduate school 24.2 20.5
Marital status .47
Married 63.2 62.2
Widowed 33.2 34.6
Separated, divorced, or never married 3.6 3.2
PMH status c .01
Never used 34.1 31.0
Current user 36.9 36.7
Past user 29.0 32.4
Family history
Heart disease 15.0 17.7 .01
Diabetes 26.7 29.7 .01
Cancer 16.7 18.3 .14
History of hypertension 21.5 33.5 ⬍.001
History of high cholesterol 12.2 17.8 ⬍.001
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); METs, metabolic-equivalent tasks (measured in
hours per week; each MET-hour is the caloric need per kilogram of body weight per hour of activity divided by the caloric need per kilogram of weight per hour at
rest); PMH, postmenopausal hormone use.
a Data are reported as means (SDs) for continuous variables or percentages for categorical variables.
b P values are based on the t test for continuous variables or ␹2 test for categorical variables.
c Proportions are based on nonmissing values.
in meters squared). Nonetheless, women who were both ful aging. The multivariate ORs (95% CIs) comparing any
lean (BMI, 18.5-22.9) and active (highest tertile of total vs none were 1.66 (1.30-2.14) for jogging, 1.87 (1.33-
METs) had the highest odds of successful survival in com- 2.61) for running, 1.34 (1.03-1.74) for playing tennis, and
parison with women who were overweight (BMI, ⱖ25) and 1.23 (1.09-1.39) for doing aerobics or calisthenics.
sedentary (bottom tertile of total METs): the OR was 3.44
(95% CI, 2.74-4.31). SECONDARY ANALYSIS
We also considered specific types of vigorous activi-
ties. After controlling for moderate-intensity activity METs, We observed similar associations for total METs when
we found that several individual vigorous activities were we restricted our analysis to women who were capable
each associated with significantly elevated odds of success- of performing at least low- to moderate-intensity activi-
197

Table 2. Odds of Successful Survival Among Women 70 Years or Older in the Nurses’ Health Study22
by Physical Activity Level at Midlife
Total Physical Activity Quintile a

P Value
Characteristic 1 (Lowest) 2 3 4 5 (Highest) for Trend b
Activity level, METs (h/wk), median (range) 0.9 (0.2-2.3) 3.6 (2.4-5.1) 7.9 (5.2-11.4) 16.2 (11.5-22.8) 37.1 (ⱖ22.9) NA
Usual/successful survivors, No./No. 2603/213 2349/195 2466/307 2382/303 2279/438 NA
Age-adjusted model 1 [Reference] 1.01 (0.83-1.24) 1.53 (1.28-1.84) 1.57 (1.31-1.89) 2.39 (2.01-2.85) ⬍.001
Multivariate model 1 c 1 [Reference] 0.98 (0.80-1.20) 1.37 (1.13-1.65) 1.34 (1.11-1.61) 1.99 (1.66-2.38) ⬍.001
Multivariate model 2 d 1 [Reference] 0.96 (0.78-1.18) 1.30 (1.08-1.57) 1.25 (1.03-1.51) 1.76 (1.47-2.12) ⬍.001
Walking Quintile a
Activity level, METs (h/wk), median (range) 0 (0-0.5) 2.0 (0.6-2.5) 3.0 (2.7-4.5) 7.5 (5.0-11.2) 20.0 (ⱖ12.5) NA
Usual/successful survivors, No./No. 2231/195 2536/230 2553/295 2423/379 2336/357 NA
Age-adjusted model 1 [Reference] 1.04 (0.86-1.28) 1.32 (1.09-1.60) 1.82 (1.52-2.18) 1.80 (1.50-2.17) ⬍.001
Multivariate model 1 c 1 [Reference] 0.99 (0.80-1.21) 1.19 (0.97-1.45) 1.50 (1.24-1.82) 1.47 (1.22-1.79) ⬍.001
Multivariate model 2 d 1 [Reference] 0.99 (0.80-1.22) 1.15 (0.94-1.40) 1.42 (1.17-1.72) 1.37 (1.10-1.67) ⬍.001
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CI, confidence interval;
METs, metabolic-equivalent tasks (measured in hours per week; each MET-hour is the caloric need per kilogram of body weight per hour of activity divided by the
caloric need per kilogram of weight per hour at rest); NA, not applicable.
a Unless otherwise noted, data are reported as odds ratios (95% confidence intervals).
b Estimates of P value for linear trend are based on linear scores derived from the medians of each physical activity category.
c Multivariate model was adjusted for age at baseline (in years); education (registered nurse, bachelor’s degree, master’s degree, or doctorate); marital status
(unmarried, married, widow, separated, or divorced); if married, husband’s education (less than high school, some high school, high school graduate, college
graduate, or graduate school); postmenopausal hormone use (never, past, or current use); smoking status (never, past, current 1-14 cigarettes/d or 15-24
cigarettes/d or ⱖ25 cigarettes/d); family history of heart disease, diabetes, or cancer (yes or no); dietary polyunsaturated to saturated fat ratio (in quintiles);
intakes of trans fat, alcohol, and cereal fiber (all in quintiles); and intakes of fruits and vegetables and red meat (in tertiles). For walking METs, vigorous physical
activity METs were further adjusted.
d Further adjusted for BMI category (⬍18.5, 18.5-22.9, 23.0-24.9, or ⱖ25.0), history of hypertension (yes or no), and history of hypercholesterolemia (yes or
no).
ties at baseline: the ORs (95% CIs) across total METs quin- examined the association of physical activity with over-
tiles were 1 [reference], 1.53 (1.20-1.95), 1.38 (1.08- all health status as evaluated by multiple domains
1.77), 1.83 (1.44-2.32), and 2.04 (1.61-2.58) (P⬍ .001 among those who have survived to older ages. In addi-
for trend). Likewise, associations for walking METs were tion, existing data are primarily for men,3,5,7 despite the
largely unchanged when we repeated the analysis among fact that women live, on average, longer than men.
women who did not engage in any vigorous activity: the Among Cardiovascular Health Study3 participants and
ORs (95% CIs) were 1.32 (1.03-1.69) for women in the male Harvard college alumni,7 midlife physical activity
middle tertile and 1.64 (1.32-2.04) for women in the high- was associated with an improved overall health status at
est tertile of walking METs. older ages. In contrast, among male Japanese Ameri-
Finally, of 13 535 participants, 1252 (9.3%) met the cans, midlife physical activity was not associated with
criteria of the alternate successful survival definition. In the probability of exceptional overall health at older
analyses of physical activity and this alternate defini- ages.5 In the study of Japanese Americans, adjustment
tion, we found similar associations. For example, the ORs of risk factors that can mediate the effects of physical
(95% CIs) for total activity METs quintiles were 1 [ref- activity on human health, such as plasma glucose and
erence], 1.27 (1.02-1.57), 1.49 (1.22-1.83), 1.63 (1.33- triacylglycerol levels, hypertension, and BMI, is likely
2.00), and 1.93 (1.58-2.36), indicating that our results one explanation for the null association. Despite this, it
were robust to different definitions of successful sur- is difficult to directly compare our findings with those
vival (P⬍.001 for trend). of these studies because our cohort included only
women, for whom physical activity patterns tend be dif-
COMMENT ferent from those of men. Nonetheless, similar to the
Cardiovascular Health Study and the Harvard alumni
In this large study of women, we documented a strong, study, we observed a strong, positive association
positive association between midlife leisure-time physi- between physical activity and exceptional survival at
cal activity and the odds of successful survival or excep- age 70 years or older in women. Our observations are
tional overall health in later life. This included a posi- also compatible with previous studies that used
tive relation between moderate-intensity activity, such disability-free survival or self-rated overall health as a
as walking, and odds of maintaining overall health sta- surrogate measure of successful survival.20,21
tus among aging women. In previous studies of successful survival, walking was
There is persuasive evidence supporting an inverse not distinguished from more vigorous activities. While
association between physical activity and many indi- approximately 85% of Americans do not participate in
vidual aspects of health, including multiple chronic dis- any regular vigorous physical activities, 44% walk for ex-
eases, cognitive function, physical function, and mental ercise.35 Consistent with the literature on walking in re-
health.9-19,34 However, fewer epidemiologic studies have lation to chronic diseases and other specific, adverse health
198

Table 3. Odds of Successful Survival Among Women 70 Years or Older in the Nurses’ Health Study22 by Walking Pace at Midlife
Walking Pace b
Easy Moderate Brisk/Very Brisk P Value

Characteristic a (⬍2.0 mph) (2.0-2.9 mph) (ⱖ3.0 mph) for Trend
Overall
Usual/successful survivors, No./No. 1972/98 6244/672 3738/678 NA
Age-adjusted model 1 [Reference] 2.16 (1.74-2.68) 3.60 (2.89-4.48) ⬍.001
Multivariate model 1 1 [Reference] 1.90 (1.52-2.38) 2.68 (2.13-3.37) ⬍.001
Walking METs ⬍3.1
Walking METs ⱖ3.1
Abbreviations: CI, confidence interval; METs, metabolic-equivalent tasks (measured in hours per week; each MET-hour is the caloric need per kilogram of body
weight per hour of activity divided by the caloric need per kilogram of weight per hour at rest); NA, not applicable.
a Regression models were adjusted for the same sets of covariates as detailed in the footnotes of Table 2 plus total physical activity (METs, hours per week; in
quintiles) in multivariate models.
b Unless otherwise noted, data are reported as odds ratios (95% confidence intervals).
outcomes,36-38 our results suggest that energy expendi-

ture from walking at a moderate to brisk pace could also 3.5 BMI
increase the likelihood of exceptional survival. Given that 18.5-22.9
3.0 23.0-24.9
walking is a sustainable exercise that can often be easily
Odds Ratio of Successful Aging
≥25.0
incorporated into people’s daily schedules, our observa- 2.5
tions provide initial support for the consideration of walk-
2.0
ing in broad public health recommendations.
Importantly, in the present study, being physically ac- 1.5
tive was associated with increased odds of successful sur-
1.0
vival for both lean and overweight women. This obser-
vation was consistent with previous findings by our 0.5
research group that physical activity was related to a sub-
0.0
stantial reduction in risk of chronic diseases and prema- >15.0 4.1-15.0 ≤4.0
ture death among participants with various body Total Physical Activity, METs (h/wk)
weights.10,34,39 Together, our data strongly support the no-
tion that, regardless of body weight, engaging in physi-
cal activity may increase the probability of preserving op- Figure. Participants’ body mass index (BMI) (calculated as weight in
kilograms divided by height in meters squared) and physical activity at
timal health. Meanwhile, our study also demonstrated that baseline in relation with the odds of successful survival in the Nurses’ Health
maintaining a healthy body weight and high physical ac- Study.22 The odds ratios were adjusted for the model 1 covariates detailed in
tivity levels simultaneously at midlife likely convey the footnote c of Table 2. MET indicates metabolic-equivalent tasks, measured in
hours per week. Each MET-hour is the caloric need per kilogram of body
highest odds of successful survival. weight per hour of activity divided by the caloric need per kilogram of weight
The strengths of the current study include a compre- per hour at rest.
hensive measurement of overall health of aging women,
large sample size, high follow-up rate, accurate self-
reported incidence of chronic diseases, and validated average 14-year lag period between exposure and out-
methods to quantify physical and mental disabilities and come assessments—to both address reverse causation as
cognitive function. Further unique aspects of our study well as the biologic likelihood that health and chronic
are the focus on women (who live longer than men on conditions at older ages are influenced by lifestyle fac-
average and thus merit particular attention in consider- tors adopted at younger ages.
ing risk factors for successful survival) and the exami- Our study also has several limitations. First, the gen-
nation of walking (one of the more common types of ac- eralizability of the current study may be limited to women
tivity among women). An additional strength derives from who were primarily of European ancestry and largely
the multiple analyses conducted to consider possible re- healthy at midlife. Further research should be con-
verse causation. For example, we excluded anyone with ducted in minority populations and populations with vari-
existing chronic diseases at baseline and also imposed an ous specific health issues in earlier life. In addition, we
199

considered successful survival as of age 70 years. Whether the manuscript for important intellectual content: Sun,
the observed associations can be generalized to popula- Townsend, Okereke, Hu, and Grodstein. Statistical analy-
tions at much older ages is unknown. sis: Sun, Townsend, and Okereke. Obtained funding:
Second, although our questionnaire to measure physi- Franco, Hu, and Grodstein. Administrative, technical, and
cal activity has been validated in a similar population and material support: Sun, Okereke, Franco, and Hu. Study
has shown reasonable accuracy, the self-reported physi- supervision: Hu and Grodstein.
cal activity levels were inevitably subject to measure- Financial Disclosure: None reported.
ment error. However, since these data were collected be- Funding/Support: This research was supported by re-
fore any of the study outcomes occurred, the measurement search grants AG13482, AG15424, and CA40356 from
errors would most likely be nondifferential and bias true the National Institutes of Health and grant DK46200 from
associations to the null. the Pilot and Feasibility program sponsored by the Bos-
Third, as in any observational study, residual con- ton Obesity Nutrition Research Center. Dr Sun was sup-
founding is also an alternative explanation of our obser- ported by a postdoctoral fellowship from Unilever Cor-
vations. However, the strength and the dose-response gra- porate Research. Dr Townsend was supported by the Yerby
dient of the multivariate associations support a causal Postdoctoral Fellowship Program. Dr Hu is a recipient
relationship between physical activity and successful ag- of American Heart Association Established Investigator
ing. In addition, the homogeneity of our study popula- Award.
tion with respect to demographic characteristics and ac- Role of the Sponsors: The funding sources had no role in
cess to health care further reduce possibilities for the collection, analysis, and interpretation of the data or
confounding. in the decision to submit the manuscript for publication.
Fourth, we did not assess physical and mental health Additional Contributions: Frans van der Ouderaa, PhD,
status at baseline. Therefore, long-term physical impair- provided insightful comments.
ment or mental limitations might have biased our ob-
servations. However, when we restricted our analysis to
women with sufficient function to engage in at least low REFERENCES
to moderate physical activity levels at baseline, we ob-
served similar associations. 1. National Center for Health Statistics. 2007 Chartbook on Trends in the Health of
Americans. Hyattsville, MD: National Center for Health Statistics; 2007.
Finally, approximately 16% of eligible women were 2. He W, Sengupta M, Velkoff VA, DeBarros KA. 65⫹ in the United States: 2005.
excluded from the present analysis because of missing Washington, DC: US Census Bureau; 2005.
physical activity data at baseline. These participants had 3. Newman AB, Arnold AM, Naydeck BL, et al; Cardiovascular Health Study Re-
slightly higher BMIs; worse physical, cognitive, and men- search Group. “Successful aging”: effect of subclinical cardiovascular disease.
tal status at older ages; and were less likely to be active Arch Intern Med. 2003;163(19):2315-2322.
4. Reed DM, Foley DJ, White LR, Heimovitz H, Burchfiel CM, Masaki K. Predictors
at baseline than women who provided data on their physi- of healthy aging in men with high life expectancies. Am J Public Health. 1998;
cal activity. This combination could lead to bias toward 88(10):1463-1468.
the null. 5. Willcox BJ, He Q, Chen R, et al. Midlife risk factors and healthy survival in men.
In summary, the present study provides new evi- JAMA. 2006;296(19):2343-2350.
6. von Faber M, Bootsma-van der Wiel A, van Exel E, et al. Successful aging in the
dence that midlife physical activity, including walking, oldest old: who can be characterized as successfully aged? Arch Intern Med. 2001;
is associated with increased odds of exceptional health 161(22):2694-2700.
among women who are initially healthy at midlife and 7. Vaillant GE, Mukamal K. Successful aging. Am J Psychiatry. 2001;158(6):839-847.
survive to older ages. Since the American population is 8. Hu FB, Sigal RJ, Rich-Edwards JW, et al. Walking compared with vigorous physi-
aging rapidly2 and nearly a quarter of Americans do not cal activity and risk of type 2 diabetes in women: a prospective study. JAMA.
1999;282(15):1433-1439.
engage in any leisure-time activity,40 our findings ap- 9. Hu FB, Stampfer MJ, Colditz GA, et al. Physical activity and risk of stroke in women.
pear to support federal guidelines regarding physical ac- JAMA. 2000;283(22):2961-2967.
tivity to promote health among older people and fur- 10. Manson JE, Hu FB, Rich-Edwards JW, et al. A prospective study of walking as
ther emphasize the potential of activity to enhance overall compared with vigorous exercise in the prevention of coronary heart disease in
women. N Engl J Med. 1999;341(9):650-658.
health and well-being with aging. The notion that physi- 11. Paffenbarger RS Jr, Hyde RT, Wing AL, Lee IM, Jung DL, Kampert JB. The as-
cal activity can promote successful survival rather than sociation of changes in physical-activity level and other lifestyle characteristics
simply extend the lifespan may provide particularly strong with mortality among men. N Engl J Med. 1993;328(8):538-545.
motivation for initiating activity. 12. Thune I, Furberg AS. Physical activity and cancer risk: dose-response and can-
cer, all sites and site-specific. Med Sci Sports Exerc. 2001;33(6)(suppl):S530-
S550, S609-S610.
Accepted for Publication: October 12, 2009. 13. Almeida OP, Norman P, Hankey G, Jamrozik K, Flicker L. Successful mental health
Correspondence: Qi Sun, MD, ScD, Department of Nu- aging: results from a longitudinal study of older Australian men. Am J Geriatr
trition, Harvard School of Public Health, 655 Hunting- Psychiatry. 2006;14(1):27-35.
ton Ave, Boston, MA 02115 (qisun@hsph.harvard.edu). 14. Fiatarone MA, O’Neill EF, Ryan ND, et al. Exercise training and nutritional supple-
mentation for physical frailty in very elderly people. N Engl J Med. 1994;330
Author Contributions: Dr Grodstein had full access to (25):1769-1775.
all of the data in the study and takes responsibility for 15. Laurin D, Verreault R, Lindsay J, MacPherson K, Rockwood K. Physical activity
the integrity of the data and the accuracy of the data analy- and risk of cognitive impairment and dementia in elderly persons. Arch Neurol.
sis. Study concept and design: Sun, Townsend, Okereke, 2001;58(3):498-504.
Franco, Hu, and Grodstein. Acquisition of data: Hu and 16. Strawbridge WJ, Deleger S, Roberts RE, Kaplan GA. Physical activity reduces
the risk of subsequent depression for older adults. Am J Epidemiol. 2002;156
Grodstein. Analysis and interpretation of data: Sun, (4):328-334.
Townsend, Okereke, Franco, Hu, and Grodstein. Draft- 17. Vita AJ, Terry RB, Hubert HB, Fries JF. Aging, health risks, and cumulative disability.
ing of the manuscript: Sun and Franco. Critical revision of N Engl J Med. 1998;338(15):1035-1041.
200

18. Yaffe K, Barnes D, Nevitt M, Lui LY, Covinsky K. A prospective study of physical 30. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey
activity and cognitive decline in elderly women: women who walk. Arch Intern (SF-36), I: conceptual framework and item selection. Med Care. 1992;30(6):
Med. 2001;161(14):1703-1708. 473-483.
19. Yates LB, Djousse L, Kurth T, Buring JE, Gaziano JM. Exceptional longevity in 31. Stampfer MJ, Kang JH, Chen J, Cherry R, Grodstein F. Effects of moderate al-
men: modifiable factors associated with survival and function to age 90 years. cohol consumption on cognitive function in women. N Engl J Med. 2005;352
Arch Intern Med. 2008;168(3):284-290. (3):245-253.
20. Leveille SG, Guralnik JM, Ferrucci L, Langlois JA. Aging successfully until death 32. Brandt J, Spencer M, Folstein M. The telephone interview for cognitive status.
in old age: opportunities for increasing active life expectancy. Am J Epidemiol. Neuropsychiatry Neuropsychol Behav Neurol. 1988;1:111-117.
1999;149(7):654-664. 33. Sun Q, Townsend MK, Okereke OI, Franco OH, Hu FB, Grodstein F. Adiposity and
21. He XZ, Baker DW. Body mass index, physical activity, and the risk of decline in weight change in mid-life in relation to healthy survival after age 70 in women:
overall health and physical functioning in late middle age. Am J Public Health. prospective cohort study. BMJ. 2009;339:b3796.
2004;94(9):1567-1573.
34. Li TY, Rana JS, Manson JE, et al. Obesity as compared with physical activity in
22. Colditz GA, Manson JE, Hankinson SE. The Nurses’ Health Study: 20-year con-
predicting risk of coronary heart disease in women. Circulation. 2006;113(4):
tribution to the understanding of health among women. J Womens Health. 1997;
499-506.
6(1):49-62.
35. Department of Health and Human Services. Physical Activity and Health: A Re-
23. Rowe JW, Kahn RL. Successful aging. Gerontologist. 1997;37(4):433-440.
port of the Surgeon General. Atlanta, GA: National Center for Chronic Disease
24. Ainsworth BE, Haskell WL, Leon AS, et al. Compendium of physical activities:
Prevention and Health Promotion; 1996.
classification of energy costs of human physical activities. Med Sci Sports Exerc.
36. Hu FB, Manson JE. Walking: the best medicine for diabetes? Arch Intern Med.
1993;25(1):71-80.
25. Wolf AM, Hunter DJ, Colditz GA, et al. Reproducibility and validity of a self-administered 2003;163(12):1397-1398.
physical activity questionnaire. Int J Epidemiol. 1994;23(5):991-999. 37. Weuve J, Kang JH, Manson JE, Breteler MM, Ware JH, Grodstein F. Physical ac-
26. Colditz GA, Willett WC, Rotnitzky A, Manson JE. Weight gain as a risk factor for tivity, including walking, and cognitive function in older women. JAMA. 2004;
clinical diabetes mellitus in women. Ann Intern Med. 1995;122(7):481-486. 292(12):1454-1461.
27. Alonso A, Hernan MA, Ascherio A. Allergy, family history of autoimmune diseases, 38. Hakim AA, Curb JD, Petrovitch H, et al. Effects of walking on coronary heart dis-
and the risk of multiple sclerosis. Acta Neurol Scand. 2008;117(1):15-20. ease in elderly men: the Honolulu Heart Program. Circulation. 1999;100(1):
28. Barr RG, Herbstman J, Speizer FE, Camargo CA Jr. Validation of self-reported 9-13.
chronic obstructive pulmonary disease in a cohort study of nurses. Am J Epidemiol. 39. Hu FB, Willett WC, Li T, Stampfer MJ, Colditz GA, Manson JE. Adiposity as com-
2002;155(10):965-971. pared with physical activity in predicting mortality among women. N Engl J Med.
29. Sun Q, Ma J, Campos H, et al. A prospective study of trans fatty acids in eryth- 2004;351(26):2694-2703.
rocytes and risk of coronary heart disease. Circulation. 2007;115(14):1858- 40. Mokdad AH, Ford ES, Bowman BA, et al. Prevalence of obesity, diabetes, and
1865. obesity-related health risk factors, 2001. JAMA. 2003;289(1):76-79.
201

Exercise Cornucopia 2010

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Exercise Cornucopia 2010

Enviado por

Direitos autorais:

Formatos disponíveis

COMMENTARY

Evidence Regarding the Benefits of Physical Exercise

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

Resistance Training and Executive Functions

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

175 Consented and booked for

158 Completed assessment

52 Received 2× RT 54 Received 1× RT 49 Received BAT 2×/wk

1 Withdrew owing to time 5 Withdrew 5 Withdrew

51 Participated in midpoint assessment 49 Participated in midpoint assessment 44 Participated in midpoint assessment

5 Withdrew 2 Withdrew for medical reasons 2 Withdrew

46 Participated in final assessment 47 Participated in final assessment 42 Participated in final assessment

135 Completed trial and included

nent of working memory. Smaller difference scores indicate Muscular Function

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

STATISTICAL ANALYSIS RESULTS

©2010 American Medical Association. All rights reserved.

Variable Baseline Midpoint Completion

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

This Month in Archives of Internal Medicine Online @ www.archinternmed.com

See Also View Last Month’s Free Article

©2010 American Medical Association. All rights reserved.

Physical Activity and Incident Cognitive Impairment

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

None Moderate High

STATISTICAL ANALYSIS by the inclusion of participants with a prodromal phase de-

©2010 American Medical Association. All rights reserved.

Model No Activity Moderate Activity P Value High Activity P Value

Abbreviatons: CI, confidence interval; OR, odds ratio.

©2010 American Medical Association. All rights reserved.

Model No Activity Moderate Activity P Value High Activity P Value

Abbreviations: CI, confidence interval; OR, odds ratio.

Characteristic None (n = 241) Moderate (n = 971) High (n = 1342) P Value

©2010 American Medical Association. All rights reserved.

Model No Activity Moderate Activity P Value High Activity P Value

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

Exercise Effects on Bone Mineral Density,

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

296 Randomized (stratified for age)

SEFIP study SEFIP study 50 Participated in whole-body

8 Lost to follow-up 11 Lost to follow-up

Included in final statistical analysis

115 In exercise group 112 In control group

STUDY INTERVENTION 3. Functional gymnastics, isometric strength training, and

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

Exercise Control Absolute Difference, Effect Size,

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

©2010 American Medical Association. All rights reserved.

Physical Activity at Midlife in Relation to Successful

©2010 American Medical Association. All rights reserved.

METHODS A wide variety of major chronic diseases (ie, cancer, diabetes,

©2010 American Medical Association. All rights reserved.