Lifetime Trajectory of Physical Activity According
to Energy Expenditure and Joint Force
CHARLES R. RATZLAFF,1PAUL DOERFLING,2GAVIN STEININGER,2MIEKE KOEHOORN,3
JOLANDA CIBERE,1MATT LIANG,4DAVID R. WILSON,3JOHN ESDAILE,1AND JACEK KOPEC1
Objective. To develop and demonstrate the feasibility of a method for estimating lifetime hip and knee cumulative joint
force using survey data on physical activity, and to construct and describe lifetime trajectories of energy expenditure and
hip and knee joint force.
Methods. Exposure data on lifetime physical activity, including type (occupational, household, and recreation) and dose
(frequency, intensity, and duration), were collected from a Canada-wide population study of adults ages >45 years.
Subjects were ranked in 2 ways: in terms of physical activity–related energy expenditure and in terms of a cumulative
peak force index (CPFI) for the hip and knee, which is a measure of lifetime exposure and is a time/joint force product
involving years of force and subject bodyweight. A relative joint loading index was calculated as the ratio of joint force
(CPFI score) to energy expenditure.
Results. A total of 4,269 subjects completed the baseline measurements. Lifetime energy expenditure and hip and knee
CPFI scores were higher for occupational and household activity than sport. The mean lifetime energy expenditure from
total physical activity in the study sample was 119.1 metabolic equivalent-hours/week. Women had slightly higher total
lifetime energy expenditure and CPFI scores than men. The relative joint loading index was highest for male household
and sport activity and lowest for female occupational activity.
Conclusion. Lifetime cumulative hip/knee joint force trajectories were successfully constructed from survey data and
followed expected trends. Comparing energy expenditure with joint force revealed variation by age, sex, and activity type,
indicating these measures may help distinguish the numerous benefits of physical activity from possible risks.
Regular physical activity is associated with lowered inci-
dence of cardiovascular disease, diabetes mellitus, hyper-
tension, osteoporosis, and cancer (1,2), and is essential for
normal muscle and joint health (2,3). As such, physical
activity is an important measure in many health studies.
Many of the health benefits of physical activity accrue
from the metabolic changes associated with energy expen-
diture (e.g., cardiovascular health, increased glucose toler-
ance); however, energy expenditure is not the only com-
ponent of physical activity. For example, similar energy
expenditure may result from programs of swimming, walk-
ing, cycling, or jogging, whereas hip and knee joint force
varies considerably by sport (4–7). It is therefore important
to distinguish between the different components of phys-
ical activity and the health outcome investigated.
Although physical activity is vital for joint health and
may protect against osteoarthritis (OA), studies on some
high-intensity sports and strenuous occupations have
shown a relationship to OA (8,9). However, research on
the relationship between hip and knee health and moder-
ate and vigorous physical activities in the general popula-
Supported in part by the Canadian Institutes of Health
Research (grant 67008). Mr. Ratzlaff’s work was supported
by a Canadian Institutes of Health Research Fellowship and
a Michael Smith Foundation for Health Research award.
Dr. Koehoorn’s work was supported in part by a Michael
Smith Foundation for Health Research Senior Scholar
1Charles R. Ratzlaff, PT, PhD(Candidate), BSc, FCAMT,
Jolanda Cibere, MD, PhD, John Esdaile, MD, MPH, FRCPC,
FCAHS, Jacek Kopec, MD, PhD: University of British Colum-
bia and Arthritis Research Centre of Canada, Vancouver,
British Columbia, Canada;2Paul Doerfling, BA, MA, MBA,
Gavin Steininger, MASc(Mech), BSc(Stats): Arthritis Re-
search Centre of Canada, Vancouver, British Columbia,
University of British Columbia, Vancouver, British Colum-
Centre of Canada, Vancouver, British Columbia, Canada,
and Brigham and Women’s Hospital, Boston, Massachu-
Address correspondence to Charles R. Ratzlaff, PT, PhD-
(Candidate), BSc, FCAMT, Arthritis Research Centre of Can-
ada, 895 West 10th Avenue, Vancouver, British Columbia,
V5Z 1L7, Canada. E-mail: firstname.lastname@example.org.
Submitted for publication November 19, 2008; accepted in
revised form April 21, 2010.
3Mieke Koehoorn, PhD, David R. Wilson, DPhil:
4Matt Liang, MD, MPH: Arthritis Research
Arthritis Care & Research
Vol. 62, No. 10, October 2010, pp 1452–1459
© 2010, American College of Rheumatology
tion has not been consistent. At the hip, systematic re-
views by Lievense et al reported odds ratios for hip OA of
approximately 3 and 2 for heavy physical workload (10)
and sports (11), respectively; however, most of the studies
reviewed were cross-sectional. Three studies (12–14) have
shown no risk or a protective effect of physical activity.
There have been more prospective studies at the knee, and
these have variously shown that exercise (including run-
ning) is not associated with OA (15,16), helps prevent OA
(17), or that certain occupations and work-related activi-
ties (18,19) and long-term participation in some sports
(including running) (20) have been associated with an
increased risk of OA. The inconsistent findings to date
may be due to individual variation in response to activity,
local factors (e.g., malalignment, previous injury), differ-
ent methodologies employed to measure physical activity
exposure, failure to capture accumulated physical activity
from all types of activity, or confounding, or may reflect
that physical activity is a construct with several compo-
nents that effect joint health differently. These challenges
have made it difficult to establish the true relationship
between physical activity and joint health.
Most studies on the relationship between physical ac-
tivity and joint health have focused on sport and/or occu-
pation and have not investigated the combined effect of
sport, occupation, and household activity (accounts for
most of the weekly physical activity in women), and have
not measured physical activity over the lifespan. Because
both idiopathic and secondary hip and knee OA have a
long induction and asymptomatic latency period, histori-
cal exposure is of considerable interest and potentially
allows for determination of etiologically important periods
of disease development. Studies that have measured long-
term physical activity have generally used exposure mea-
sures such as job title or several broad categories (e.g., low,
moderate, and high) without taking into account specific
activities. These methods of measuring physical activity
have not been designed to test biologic hypotheses about
the cumulative effect of different activities and the role of
joint forces. In summary, we do not know how dose and
type of joint force, nor combinations of activities, are as-
sociated with joint health.
The purpose of this study was 2-fold: to develop and
demonstrate the feasibility of a method for estimating cu-
mulative hip and knee joint force using survey data on
physical activity collected via a validated adaptive Inter-
net survey, and to construct and describe lifetime trajec-
tories of energy expenditure and hip and knee joint load by
sex and activity domain strata in a large Canadian sample.
MATERIALS AND METHODS
Subjects. The source population for the study cohort
included members of the Canadian Association of Retired
Persons (CARP), Canada’s largest age ?50 years advocacy
group with more than 400,000 members. An invitation to
participate in the physical activity and joint health study
was sent via e-mail to 28,000 CARP members who agreed
to receive such e-mail. Additionally, an advertisement was
included in a monthly online newsletter sent to 100,000
CARP members. A total of 6,052 people registered on the
study Web site, and 4,269 completed the survey (70% of
those registered). The Behavioral Research Ethics Board at
the University of British Columbia approved the study.
Data collection. Data were collected using the Lifetime
Physical Activity Questionnaire, a computer-based tool
that was adapted for this study based on existing validated
instruments (21,22). It was tested for reliability and valid-
ity in a subsample drawn from the current study sample
(23). The questionnaire incorporates cognitive recall tech-
niques that were adapted for self-administration.
All of the data were collected via the Internet using skip
logic technology that allowed the subjects to follow indi-
vidualized paths through the survey, moving forward
based on responses to previous questions. A Web site was
developed for this study that allowed subjects to log in
with a password, work on the survey, save responses, and
return later to continue. The questionnaire usually took
1–1.5 hours to complete.
In addition to physical activity, the questionnaire mea-
sured known hip and knee joint health risk factors and
anthropomorphic and sociodemographic variables, in-
cluding weight (current, maximum, and at age 20 years),
height (current and at age 20 years), general health, and
Physical activity measurement. Lifetime physical activ-
ity was measured across 3 domains: sport/recreation, oc-
cupation, and household. In the sport/recreation section,
the respondents were provided a list of 64 possible sports
and were permitted to add other sports. Respondents who
had performed an activity at least 100 times in their life-
time were prompted to provide detailed information. This
included the duration (years in the activity), frequency
(times per year), and average length of time per occasion.
In addition, for each type of sport, the respondents were
required to report time spent per hour (none, 1–5, 5–15,
15–30, 30–45, 45–60 minutes) in each of the following
bodily movements: sitting, standing, walking, running/
jogging, squatting or knee bending without lifting, and
squatting or knee bending with lifting or force.
The occupational section used an open format in which
the respondents indicated all of the jobs held over their
lifetime. For each occupation, details were collected on job
title or type, duration (years), average hours per week, and
whether the job was full time, part time, or seasonal. In
addition, for each job, the respondents were required to
report time spent in an 8-hour period (none, 0–1, 2–4, 5–7,
8 hours) in each of the following bodily movements: sit-
ting; standing; standing, holding, or moving objects greater
than 50 pounds (23 kg); walking; walking or carrying ob-
jects greater than 50 pounds (23 kg); moving or pushing
objects greater than 75 pounds (34 kg); using heavy tools;
squatting continuously; and kneeling continuously.
Household activity covered 4 areas: 1) caring for chil-
dren, 2) caring for elderly individuals or individuals with
disabilities, 3) gardening, and 4) housework. For each
household area of activity, the participants were prompted
to provide detailed information on duration (years) and
Estimating Cumulative Hip/Knee Joint Force With Survey Data on PA 1453
average number of hours per week actively participating.
In addition, for each household area of activity, the sub-
jects were required to report time spent in an 8-hour pe-
riod for each of the bodily movements listed above for
occupational activity, with the exception of the use of
Primary exposure variables. Outcome variables were
constructed to describe lifetime physical activity in terms
of energy expenditure (metabolic equivalent [MET]–hours
per week) and cumulative hip and knee joint force (body-
weight hours). Participation in each specific activity (total
lifetime hours) was calculated by taking the product of
duration (years in activity), frequency (times per year), and
average length of participation per occasion. Physical ac-
tivity was then expressed as average weekly participation
(average hours per week), calculated by dividing the total
lifetime hours by respondent age and 52 (i.e., average
hours per week ? total lifetime hours/age/52). Calcula-
tions for both metrics were performed to obtain values for
each specific activity, and then were summed by activity
domain (sport, occupation, household), and finally for to-
Energy expenditure (MET-hours per week). To obtain a
measure of energy expenditure associated with physical
activity, average hours per week in physical activity in
each activity were multiplied by the specific MET associ-
ated with that activity, assigned using the Compendium of
Physical Activities (i.e., MET-hours per week ? average
hours per week ? MET, per each activity) (24). A MET is
defined as the ratio of the associated metabolic rate for a
specific activity as compared with the resting metabolic
Cumulative peak force index (bodyweight hours). To
obtain a measure of cumulative joint force at the hip and
knee, a cumulative peak force index (CPFI) score was
estimated for each joint (hip, tibiofemoral, patellofemoral)
and activity separately as the product of time spent in a
specific activity (total lifetime hours), bodyweight, and
typical peak joint force for that activity (% bodyweight;
i.e., CPFI score [bodyweight hours] ? total lifetime
hours ? bodyweight ? typical peak joint force, per each
To obtain a bodyweight estimate for each 5-year period,
a lifetime (age ?20 years) bodyweight trajectory was de-
rived using current weight, weight at age 20 years, and
maximum weight, and interpolated using a lowess (non-
parametric smooth) curve. This method is sufficiently flex-
ible to model the trajectory of weight over a respondent’s
lifetime, is more informative than a simple lifetime aver-
age, and allowed for the factoring in of changes in body-
weight over time. Bodyweight at age ?20 years is variable
and may be more difficult to accurately recall. Therefore,
joint force was calculated at age ?20 years. To assign a
typical peak joint force (in % bodyweight) for each activity
in the CPFI calculations, we carried out a comprehensive
review of the literature (4,6,25–33) and assigned a peak
force value to each activity. For example, typical peak hip
forces used in the calculations for walking, jogging, and
squatting were 3, 6, and 1.5 times bodyweight, respec-
Relative joint loading index. We also defined a relative
joint loading index for each joint as the ratio of CPFI score
to energy expenditure (i.e., joint loading index ? CPFI
[bodyweight hours]/energy expenditure [MET-hours]).
Statistical analysis. Descriptive statistics were per-
formed to characterize the study sample. For each type and
domain of activity, estimates were made for 5-year periods
in a person’s lifetime, starting at age 10 years for energy
expenditure and age 20 years for joint force. In order to
Figure 1. Lifetime trajectories for energy expenditure by activity type and sex. Mean values
for 5-year intervals over a person’s lifetime, averaged over all of the subjects, were calculated
and plotted. MET ? metabolic equivalent.
Table 1. Subject characteristics (n ? 4,269)
Subjects, no. (%)
Age, mean ? SD years
Current weight, mean ? SD pounds
Current body mass index, mean ? SD kg/m2
Some postsecondary education, %
61.5 ? 7.6
175 ? 41
27.3 ? 5.9
63.0 ? 7.8
193 ? 41
27.0 ? 5.3
60.6 ? 7.3
165 ? 38
27.5 ? 6.3
1454 Ratzlaff et al
calculate lifetime trajectories for both energy expenditure
and CPFI scores, mean values for 5-year intervals over a
person’s lifetime, averaged over all of the subjects, were
calculated and plotted. For each metric, trajectories over
the lifetime were described for total energy expenditure
and CPFI and across the demographic strata of the sample,
including activity domains (sport/recreation, occupa-
tional, household) and sex. Comparisons for the CPFI mea-
sure were made to expectations based on the literature and
known and/or rationalized relationships. All of the data
were analyzed using SPSS, version 15.0 (SPSS).
A total of 4,269 subjects completed the baseline measure-
ments (Table 1). The sample was 63% female and 93%
white. Sixty-six percent had postsecondary education.
Lifetime trajectories for energy expenditure (Figure 1) in
sport/recreational activity were highest prior to age 20
years and peaked by age 25 years. For both energy expen-
diture and joint force (Figures 2 and 3), occupational and
household measures were much higher than sport/recre-
ation over the lifetime, increasing rapidly from ages 20–35
years (particularly male occupation and female house-
hold) before peaking and then slowly declining.
The mean energy expenditure from total physical activ-
ity in the study sample was 119.1 MET-hours/week. This
level of activity is equivalent to 17 MET-hours per day or
12 hours of 1.42 METS of activity per day. Overall, when
accounting for total activity, including household, women
reported expending on average more energy than men over
the lifetime (126 versus 107 MET-hours per week per
year). Occupation energy expenditure for the overall sam-
ple was slightly higher than household (53.9 versus 49.8
MET-hours/week), whereas sport/recreational MET-hours
per week was approximately 3 times less than either
household or occupational activity (16.8 MET-hours/
Women reported approximately 1.74 times higher mean
energy expenditure in household activity as compared
with occupational activity (70.5 versus 40.5 MET-hours/
week). Compared with sport/recreational activity, women
spent approximately 7 times more energy on average in
household activity (70.5 versus 7.3 MET-hours/week).
Men reported spending approximately 2.2 times more en-
ergy in occupational as compared with household activity
(63.9 versus 28.6 MET-hours/week), and approximately 2
times more energy in household as compared with sport/
recreational activity (28.6 versus 16.9 MET-hours/week).
Women’s energy expenditure in household activity was
approximately 2.5 times that of men, whereas men’s en-
ergy expenditure in occupational activity was approxi-
mately 1.5 times that of women. For sport/recreation, men
expended on average over the lifetime approximately 2
times more energy than women (16.9 versus 7.3 MET-
Hip and tibiofemoral joint CPFI scores were substan-
tially higher than patellofemoral joint scores. Trajectories
for the hip and knee CPFI scores by sex and activity do-
main are shown in Figures 2 and 3. We defined a relative
joint loading index as the ratio of cumulative joint load to
MET to examine activities that produce high physical ex-
(CPFI) score (in kg hours per year) by activity type and sex. Mean values for 5-year intervals
over a person’s lifetime, averaged over all of the subjects, were calculated and plotted. BW ?
Lifetime trajectories for the knee (tibiofemoral) cumulative peak force index
Figure 3. Lifetime trajectories for the hip cumulative peak force index (CPFI) score (in kg
hours per year) by activity type and sex. Mean values for 5-year intervals over a person’s
lifetime, averaged over all of the subjects, were calculated and plotted. BW ? bodyweight.
Estimating Cumulative Hip/Knee Joint Force With Survey Data on PA1455
ertion without excessive joint force (low ratio) and vice
versa (high ratio). For example, swimming would have a
low ratio, whereas sustained squatting would have a high
ratio. The lifetime trajectory for the knee (tibiofemoral)
loading index is shown in Figure 4. Male household and
sport activities had the highest relative joint load ratio.
Occupational activity for women had the lowest ratio, at
approximately 50% of male household activity. House-
hold ratios for both sexes were higher than occupational.
Hip and patellofemoral relative joint loading ratios
showed similar trends to tibiofemoral by sex and activity
strata (data not shown).
To our knowledge, this is the first attempt to quantify
cumulative physical activity–related lifetime hip and knee
joint force, an exposure that may prove to be an important
measure to further understand the benefits and potential
risks of different doses and types of physical activity at the
population level. In particular, it may help to distinguish
between energy expenditure and the biomechanical effects
of different levels, types, and combinations of physical
activity that may be determinants of hip and knee joint
health. Although additional work needs to be done in
determining the reliability and validity of the lifetime joint
force measure, the current findings are encouraging. We
were able to construct long-term trajectories for cumula-
tive hip and knee joint force using data from an extensive
adaptive online survey.
The reliability and validity of the lifetime survey on
which the joint force measure is based were good (23). The
joint force measure is new, however, and validation was
limited here in comparing strata of the sample with ex-
pected or known differences from other physical activity
measures. The lifetime energy expenditure (MET-hours
per week) estimates were similar to several other studies
reporting this metric (34–36). Sport/recreation and occu-
pational activity accounted for more energy expenditure
and joint force for men than women, whereas household
energy expenditure and joint force were higher for women.
In comparing the ratio of cumulative joint force with en-
ergy expenditure (joint loading index), male household
and sport ratios were highest, whereas occupational ratios
for both sexes were much lower. These differences, as well
as changes over the lifetime (by 5-year age period), were
expected in this sample and provide face validity to the
cumulative joint force and ratio metrics.
Of importance in introducing these new measures is that
cumulative joint force has a biologically plausible relation-
ship with hip and knee joint health. Moderate mechanical
loading is necessary for homeostasis and to maintain
healthy articular cartilage (3). Under normal physiologic
conditions, articular cartilage provides a nearly friction-
less surface for the transmission and distribution of joint
loads, exhibiting little or no wear over decades of use (37).
However, physical activity and joint load can elicit either
anabolic or catabolic processes, depending on the inten-
sity and duration, age at onset, and amount and type of
activity (3). Insufficient load through a joint compromises
bone and cartilage health (38), whereas overloading the
knee and hip joints could lead to cartilage breakdown and
failure of ligamentous and other structural support (39). It
is not known what level or type(s) of physical activity is
optimal for joint protection and what may increase the risk
of joint deterioration.
Some of the uncertainty may be the result of study
designs that were not designed to test biologic hypotheses
about the cumulative effect or critical levels and combina-
tions of different activities and the role of joint forces.
Biologically, it seems likely that different activities have a
cumulative effect, and that this effect depends on the
mechanical forces transmitted through the joint (40). How-
ever, this theory has never been tested empirically. It is
therefore important to attempt to quantify exposure to
long-term joint force.
There were two unexpected findings. First, the overall
lifetime energy expenditure and joint force estimates were
slightly higher for women than men. Over the past several
decades, women have increased both occupational and
sport/recreational activity, while largely maintaining high
levels of household activity. Women often spend ?40
hours a week at a full-time job and anywhere from 25–45
hours a week working in the home. We found that on
average over the lifetime, women spend 4.23 hours per day
in household activity. The only previous study to estimate
lifelong joint load found that women differ in the quality
and quantity of physical load and that a reduction of high
Figure 4. Lifetime trajectories for the knee joint loading index (the ratio of weekly cumu-
lative peak force index score to weekly energy expenditure) by activity type and sex. Mean
values for 5-year intervals over a person’s lifetime, averaged over all of the subjects, were
calculated and plotted. MET ? metabolic equivalent.
1456Ratzlaff et al
physical load at home could probably lower the risk of
knee OA later in life (19). It is possible that the high overall
and household levels of joint force in women may contrib-
ute to the higher incidence of knee and hip OA in older
women, and are worthy of investigation in future studies.
A second unexpected finding was the relatively lower
levels of sport-related energy expenditure and joint force.
Although early epidemiologic studies based physical ac-
tivity estimates primarily on occupation, physical activity
has declined significantly in most occupations (41). There-
fore, since the 1990s, there was a greater emphasis on
investigating sports and leisure-time activity over the life
course. However, we found that the occupation and house-
hold domains remain the most important in terms of ac-
cumulation of joint force and energy expenditure. This
may in part be due to the older average age of our cohort
reflecting historical work patterns prior to the 1990s, and
an age-related decline in sport-related activity. Further-
more, we were able to measure sport-related activity rela-
tive to occupation and household activity, something few
studies have compared. Previous studies investigating
sport activity alone may not have concluded that sport
levels were substantially lower than occupation and
household; studies that considered sport activities alone
may have been confounded by occupational and/or house-
hold physical activity–related joint force.
The energy expenditure estimates we found are consis-
tent with the previous literature (22,34–36). In a study of
the relationship between breast cancer and lifetime phys-
ical activity in 2,470 women (average age 56 years, body
mass index 27.5 kg/m2) (35), Friedenreich et al reported
that women expended 127.8 MET-hours/week over the
lifetime, whereas our study reported 126.0 MET-hours/
week. Comparisons by type of activity were also similar by
household (66.1 versus 70.5), occupational (48.1 versus
40.5), and sport/recreation (13.6 versus 7.3). The differ-
ences may in part relate to the older age of women (average
age 60.6 years versus 56 years) in our study, and are
relatively small, given that lifetime activity was measured.
A similar study in men investigating lifetime physical
activity and prostate cancer provided estimates that were
also close to ours (36). In a criterion validity study assess-
ing past-year physical activity (using a questionnaire sim-
ilar to the lifetime version on which our survey was
based), self-report data were correlated with data from
accelerometers and activity logs (34). An acceptable level
of reliability and validity was reported and the point esti-
mates for energy expenditure for total and activity domain
were similar to our estimates.
There are some limitations to this study. The interpre-
tation of data using self-report physical activity measures
requires caution and is limited to characterizing large
groups of people rather than individuals because of large
within-person variability and problems with recall (42–
44). Despite these limitations, it has been repeatedly
shown that physical activity questionnaires are both prac-
tical and valid when used appropriately for large-scale
epidemiologic studies (44–47). Falkner et al studied the
reliability of recall of physical activity in the last 30 years
and found acceptable levels of recall (48). Validity studies
of physical activity questionnaires using accelerometers
and/or physical activity logs have shown acceptable valid-
ity (46,49). In a review focused on the limitations of phys-
ical activity questionnaires, Shephard concluded that al-
though detailed interpretation and attempts to estimate
precise dose are inadvisable, the use of data to monitor
change in population activity and provide categorical es-
timates is of value (44). If the questionnaire is adequately
designed for a particular population and has acceptable
reliability and validity, the instrument should be able to
rank order adults by category of activity levels and by
sociodemographic groups, providing a relative distribu-
tion of historical physical activity (43). Moreover, when
used for etiologic analysis, a precise measure of an irrele-
vant variable (e.g., recent physical activity exposure) may
be of little value, whereas an imprecise measure of a rele-
vant variable may be useful.
We used a lifetime questionnaire that was based on
existing validated instruments (21,22) developed carefully
and pilot tested and validated in a separate study (23).
Test–retest reliability with an 8-month interval was mod-
erate to very good (intraclass correlation coefficient range
0.58–0.82). Construct validity was established in a series
of substudies using the current study sample with Spear-
man’s correlation coefficients ranging from 0.34–0.71.
The calculation of joint force was based in part on as-
signing typical peak forces for given activities or positions.
The typical peak joint force is not the same for every
individual and may not precisely represent the forces in
our study subjects. Furthermore, the biomechanical stud-
ies we reviewed to obtain typical force values presented
peak joint force for that activity. The use of peak (as op-
posed to mean) force in the calculation may result in
overestimation of absolute lifetime joint load. However,
because we are interested in relative and not absolute
values, the biases introduced in the estimates would not
significantly influence our findings.
The study had a number of strengths. A population-
based study design and large sample size were employed.
A novel feature of the questionnaire was its Internet-based
administration, which permitted the use of skip logic to
maximize efficiency, minimize respondent burden, elimi-
nate missing data, and allow the subjects control of time
management. This may have increased compliance, com-
pletion rates, and accuracy, and allowed capture of all of
the relevant aspects of physical activity. We were also able
to draw distinctions between time spent in different bodily
positions, which allowed for complete classification of the
total volume of physical activity and loading aspects of
joint force. The inclusion of household activities was crit-
ical because they account for most of the weekly energy
expenditure in women. Lastly, we collected comprehen-
sive data on physical activity that will allow us to test
biologic hypotheses about the cumulative effect of differ-
ent activities and the role of joint forces in hip and knee
In conclusion, lifetime joint force trajectories for the hip
and knee were constructed from data using a validated
survey, and followed expected trends by age, sex, and
physical activity domain, indicating the measure has face
validity. The creation of a joint loading index may be an
important exposure that separates energy expenditure
Estimating Cumulative Hip/Knee Joint Force With Survey Data on PA 1457
from joint force, and may reveal activity patterns that
maximize health benefits while minimizing the risk of
joint overload. Future research should examine the rela-
tionship between joint force and OA and investigate the
interaction of long-term joint force and local factors (e.g.,
alignment, injury, neuromuscular control). If the amount
and type of physical activity associated with protective
and more damaging loads can be identified, it could lead to
relatively inexpensive and practical recommendations.
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors ap-
proved the final version to be submitted for publication. Mr.
Ratzlaff had full access to all of the data in the study and takes
responsibility for the integrity of the data and the accuracy of the
Study conception and design. Ratzlaff, Liang, Wilson, Kopec.
Acquisition of data. Ratzlaff, Doerfling, Liang, Kopec.
Steininger, Koehoorn, Cibere, Esdaile, Kopec.
1. Blair SN. Physical activity, clinical medicine, and public
health. Curr Sports Med Rep 2007;6:71–2.
2. Macera CA, Hootman JM, Sniezek JE. Major public health
benefits of physical activity. Arthritis Rheum 2003;49:122–8.
3. Griffin TM, Guilak F. The role of mechanical loading in the
onset and progression of osteoarthritis. Exerc Sport Sci Rev
4. Neptune RR, Kautz SA. Knee joint loading in forward versus
backward pedaling: implications for rehabilitation strategies.
Clin Biomech (Bristol, Avon) 2000;15:528–35.
5. Bergmann G, Deuretzbacher G, Heller M, Graichen F, Rohl-
mann A, Strauss J, et al. Hip contact forces and gait patterns
from routine activities. J Biomech 2001;34:859–71.
6. Taylor WR, Heller MO, Bergmann G, Duda GN. Tibio-femoral
loading during human gait and stair climbing. J Orthop Res
7. Bergmann G, Graichen F, Rohlmann A. Hip joint loading
during walking and running, measured in two patients. J Bio-
8. Rossignol M, Leclerc A, Allaert FA, Rozenberg S, Valat JP,
Avouac B, et al. Primary osteoarthritis of hip, knee, and hand
in relation to occupational exposure. Occup Environ Med
9. Kujala UM, Kaprio J, Sarna S. Osteoarthritis of weight bearing
joints of lower limbs in former elite male athletes. BMJ 1994;
10. Lievense A, Bierma-Zeinstra S, Verhagen A, Verhaar J, Koes
B. Influence of work on the development of osteoarthritis of
the hip: a systematic review. J Rheumatol 2001;28:2520–8.
11. Lievense AM, Bierma-Zeinstra SM, Verhagen AP, Bernsen
RM, Verhaar JA, Koes BW. Influence of sporting activities on
the development of osteoarthritis of the hip: a systematic
review. Arthritis Rheum 2003;49:228–36.
12. Rogers LQ, Macera CA, Hootman JM, Ainsworth BE, Blairi
SN. The association between joint stress from physical activ-
ity and self-reported osteoarthritis: an analysis of the Cooper
Clinic data. Osteoarthritis Cartilage 2002;10:617–22.
13. Hootman JM, Macera CA, Helmick CG, Blair SN. Influence of
physical activity-related joint stress on the risk of self-re-
ported hip/knee osteoarthritis: a new method to quantify
physical activity. Prev Med 2003;36:636–44.
14. Karlson EW, Mandl LA, Aweh GN, Sangha O, Liang MH,
Grodstein F. Total hip replacement due to osteoarthritis: the
importance of age, obesity, and other modifiable risk factors.
Am J Med 2003;114:93–8.
15. Lane NE, Oehlert JW, Bloch DA, Fries JF. The relationship of
running to osteoarthritis of the knee and hip and bone mineral
density of the lumbar spine: a 9-year longitudinal study.
J Rheumatol 1998;25:334–41.
16. Felson DT, Niu J, Clancy M, Sack B, Aliabadi P, Zhang Y.
Effect of recreational physical activities on the development
of knee osteoarthritis in older adults of different weights: the
Framingham study. Arthritis Rheum 2007;57:6–12.
17. Foley S, Ding C, Cicuttini F, Jones G. Physical activity and
knee structural change: a longitudinal study using MRI. Med
Sci Sports Exerc 2007;39:426–34.
18. McAlindon TE, Wilson PW, Aliabadi P, Weissman B, Felson
DT. Level of physical activity and the risk of radiographic and
symptomatic knee osteoarthritis in the elderly: the Framing-
ham study. Am J Med 1999;106:151–7.
19. Sandmark H, Hogstedt C, Vingard E. Primary osteoarthrosis of
the knee in men and women as a result of lifelong physical
load from work. Scand J Work Environ Health 2000;26:20–5.
20. Cheng Y, Macera CA, Davis DR, Ainsworth BE, Troped PJ,
Blair SN. Physical activity and self-reported, physician-diag-
nosed osteoarthritis: is physical activity a risk factor? J Clin
21. Vuillemin A, Guillemin F, Denis G, Huot J, Jeandel C. A
computer-assisted assessment of lifetime physical activity:
reliability and validity of the QUANTAP software. Rev Epi-
demiol Sante Publique 2000;48:157–67.
22. Friedenreich CM, Courneya KS, Bryant HE. The lifetime total
physical activity questionnaire: development and reliability.
Med Sci Sports Exerc 1998;30:266–74.
23. De Vera M, Ratzlaff C, Doerfling P, Kopec J. Reliability and
validity of an Internet-based questionnaire measuring lifetime
physical activity. Am J Epidemiol 2010. In press.
24. Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM,
Strath SJ, et al. Compendium of physical activities: an update
of activity codes and MET intensities. Med Sci Sports Exerc
25. Mow VC, Hayes WC. Basic orthopaedic biomechanics. New
York: Raven Press; 1991.
26. Van den Bogert AJ, Read L, Nigg BM. An analysis of hip joint
loading during walking, running, and skiing. Med Sci Sports
27. Bergmann G, Graichen F, Rohlmann A, Linke H. Hip joint
forces during load carrying. Clin Orthop Relat Res 1997;335:
28. Flynn TW, Soutas-Little RW. Patellofemoral joint compres-
sive forces in forward and backward running. J Orthop Sports
Phys Ther 1995;21:277–82.
29. McNitt-Gray JL. Kinetics of the lower extremities during drop
landings from three heights. J Biomech 1993;26:1037–46.
30. Ericson MO, Bratt A, Nisell R, Nemeth G, Ekholm J. Load
moments about the hip and knee joints during ergometer
cycling. Scand J Rehabil Med 1986;18:165–72.
31. Rohrle H, Scholten R, Sigolotto C, Sollbach W, Kellner H.
Joint forces in the human pelvis-leg skeleton during walking.
J Biomech 1984;17:409–24.
32. Nemeth G, Ekholm J, Arborelius UP. Hip load moments and
muscular activity during lifting. Scand J Rehabil Med 1984;
33. Taylor SJ, Walker PS. Forces and moments telemetered from
two distal femoral replacements during various activities.
J Biomech 2001;34:839–48.
34. Friedenreich CM, Courneya KS, Neilson HK, Matthews CE,
Willis G, Irwin M, et al. Reliability and validity of the Past
Year Total Physical Activity Questionnaire. Am J Epidemiol
35. Friedenreich CM, Bryant HE, Courneya KS. Case-control
study of lifetime physical activity and breast cancer risk. Am J
36. Friedenreich CM, McGregor SE, Courneya KS, Angyalfi SJ,
Elliott FG. Case-control study of lifetime total physical activ-
ity and prostate cancer risk. Am J Epidemiol 2004;159:740–9.
37. Guilak F, Setton LA, Kraus VB. Structure and function of
articular cartilage. In: Garrett WE, Speer KP, Kirkendall DT,
editors. Principles and practice of orthopaedic sports medi-
1458Ratzlaff et al
cine. Philadelphia: Lippincott Williams & Wilkins; 2000. p.
38. Arokoski JP, Jurvelin JS, Vaatainen U, Helminen HJ. Normal
and pathological adaptations of articular cartilage to joint
loading. Scand J Med Sci Sports 2000;10:186–98.
39. Felson DT, Lawrence RC, Dieppe PA, Hirsch R, Helmick CG,
Jordan JM, et al. Osteoarthritis: new insights. Part 1: the dis-
ease and its risk factors. Ann Intern Med 2000;133:635–46.
40. Chen CT, Burton-Wurster N, Lust G, Bank RA, Tekoppele JM.
Compositional and metabolic changes in damaged cartilage
are peak-stress, stress-rate, and loading-duration dependent.
J Orthop Res 1999;17:870–9.
41. Kriska AM, Knowler WC, LaPorte RE, Drash AL, Wing RR,
Blair SN, et al. Development of questionnaire to examine
relationship of physical activity and diabetes in Pima Indians.
Diabetes Care 1990;13:401–11.
42. Janz KF. Physical activity in epidemiology: moving from
questionnaire to objective measurement. Br J Sports Med
43. Kriska AM, Sandler RB, Cauley JA, LaPorte RE, Hom DL,
Pambianco G. The assessment of historical physical activity
and its relation to adult bone parameters. Am J Epidemiol
44. Shephard RJ. Limits to the measurement of habitual phys-
ical activity by questionnaires. Br J Sports Med 2003;37:
45. Cust AE, Armstrong BK, Friedenreich CM, Slimani N, Bau-
man A. Physical activity and endometrial cancer risk: a re-
view of the current evidence, biologic mechanisms and the
quality of physical activity assessment methods. Cancer
Causes Control 2007;18:243–58.
46. Chasan-Taber L, Erickson JB, Nasca PC, Chasan-Taber S,
Freedson PS. Validity and reproducibility of a physical activ-
ity questionnaire in women. Med Sci Sports Exerc 2002;34:
47. Godin G, Jobin J, Bouillon J. Assessment of leisure time exer-
cise behavior by self-report: a concurrent validity study. Can
J Public Health 1986;77:359–62.
48. Falkner KL, Trevisan M, McCann SE. Reliability of recall of
physical activity in the distant past. Am J Epidemiol 1999;
49. Matthews CE, Shu XO, Yang G, Jin F, Ainsworth BE, Liu D, et
al. Reproducibility and validity of the Shanghai Women’s
Health Study Physical Activity Questionnaire. Am J Epide-
Estimating Cumulative Hip/Knee Joint Force With Survey Data on PA1459