ORIGINAL
E n d o c r i n e
ARTICLE
R e s e a r c h
Association of Physical Activity with Trabecular
Microstructure and Cortical Bone at Distal Tibia
and Radius in Young Adult Men
M. Nilsson, C. Ohlsson, D. Sundh, D. Mellström, and M. Lorentzon
Center for Bone Research at the Sahlgrenska Academy, Institute of Medicine, University of Gothenburg,
416 45 Gothenburg, Sweden
Context: The relationship between physical activity, trabecular microstructure, and cortical geometry in weight-bearing and non-weight-bearing bone has not previously been studied in men.
Objective: The aim of this study was to investigate whether present (type and amount) and previous
duration of physical activity were associated with trabecular microstructure and cortical crosssectional area (CSA) in weight-bearing bone in young men.
Design and Setting: This was a cross-sectional, population-based study.
Participants: Participants included a cohort of 829 Swedish men between 22.8 and 25.7 yr old
(24.1 ⫾ 0.6 yr, mean ⫾ SD).
Main Outcome Measures: Several microstructural trabecular and cortical traits were assessed with
high-resolution three-dimensional peripheral quantitative computed tomography at distal tibia
and radius. A standardized questionnaire was used to collect information about physical activity
amount (hours per year), duration (years), and type (strain score 0 –3, based on ground reaction
forces).
Results: Men with the highest physical activity strain score had higher tibial trabecular bone volume
fraction (13.9⌬%) and trabecular number (12.7%) than men with the lowest strain score (P ⬍ 0.001).
Men in the group with the longest duration of physical activity had higher tibial cortical CSA
(16.1%) than the sedentary men (P ⬍ 0.001). Inclusion of all physical activity variables in a linear
regression model revealed that strain score independently predicted trabecular bone volume
fraction, and trabecular number (P ⬍ 0.001) and that duration of previous physical activity independently predicted cortical CSA (P ⬍ 0.001) of the tibia.
Conclusions: In this large cohort of young men, the degree of mechanical loading due to type of
physical activity was predominantly associated with trabecular microstructure, whereas duration
of previous physical activity was mainly related to parameters reflecting cortical bone size in
weight-bearing bone. (J Clin Endocrinol Metab 95: 2917–2926, 2010)
ven though the variance in bone mass is mostly genetically determined (1–3), it is well known that exercise
with loading of the bone is an important determinant of
bone mass (4 – 6). A recent study on older twin pairs suggested that the relative importance of physical activity
E
compared with genetic factors for structural bone strength
of the weight-bearing tibia was greater than the nonweight-bearing radius (7).
The mechanical strength of the bone and resistance
against fracture have been reported to be dependent on
ISSN Print 0021-972X ISSN Online 1945-7197
Printed in U.S.A.
Copyright © 2010 by The Endocrine Society
doi: 10.1210/jc.2009-2258 Received October 23, 2009. Accepted March 10, 2010.
First Published Online April 7, 2010
Abbreviations: BMD, Bone mineral density; BV/TV, trabecular bone volume fraction;
Cort.CSA, cortical cross-sectional area; Cort.Pm, cortical periosteal circumference;
Cort.Th, cortical thickness; CV, coefficient of variation; 3D, three-dimensional; D.Cort,
cortical bone density; D.Trab, trabecular bone density; GOOD, Gothenburg Osteoporosis
and Obesity Determinants; QCT, quantitative computed tomography; Tb.N, trabecular
number; Tb.Th, trabecular thickness; Tot.Area, total bone area; vBMD, volumetric BMD.
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
jcem.endojournals.org
2917
2918
Nilsson et al.
Physical Activity, Trabecular Microstructure, and Cortical Bone
bone size, volumetric density (8, 9), and trabecular bone
architecture (10). Thus, information on bone structure, as
provided by three-dimensional (3D) bone measurements,
could be a valuable addition to the commonly measured
areal bone mineral density (BMD) in estimating bone
strength (11). A previous study indicated that a given loss
of trabecular density affected the strength of the bone to a
higher extent if the loss was caused by reduced trabecular
number rather than by reduced trabecular thickness (12).
Ma et al. (13) recently reported, in a study on monozygotic
twin pairs, that physical activity during adulthood was
associated with higher volumetric trabecular density and
increased compressive strength in the distal tibia, but
whether the increased density was due to changes in trabecular number or trabecular thickness has not yet been
investigated.
Cohort studies demonstrated that physical training before and during puberty are associated with increased
bone acquisition in children and young adults (14 –16).
Several studies suggested that the type of physical activity
and the accompanying dynamic activity are of particular
importance (15, 17–20). The maximum effect is believed
to be achieved by weight-bearing physical activity including jumping actions, explosive actions like turning and
sprinting, and fairly few repetitions rather than endurance
or non-weight-bearing activities (21–25).
Although we previously reported, with data from the
Gothenburg Osteoporosis and Obesity Determinants
(GOOD) study, that physical activity is associated with
trabecular volumetric BMD (vBMD) in young men (15),
its association with trabecular bone microstructure has
not been investigated in men. High-load physical activity
has been associated with increased trabecular number (analyzed by magnetic resonance imaging) of weight-bearing
bone in a small cohort (n ⫽ 16) of female gymnasts (26),
but whether it is the type or the duration of physical activity that contribute the most to this improvement has not
yet been investigated. The aim of this study was to investigate whether present (type and amount) or previous duration of physical activity were associated with trabecular
bone microstructure and cortical bone size in weight-bearing and non-weight-bearing bone in young men.
Subjects and Methods
Subjects
The study subjects were initially enrolled in the population
based GOOD study with the aim to determine both environmental and genetic factors involved in the regulation of bone
mass (27). All study subjects in the original GOOD study were
contacted by letter and telephone and invited to participate in
this 5-yr follow-up study. Of the original 1068 subjects, 829
men, between 22.8 and 25.7 yr old (24.1 ⫾ 0.6 yr, mean ⫾ SD),
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
were included in the present study. A total of 133 subjects declined to participate and 106 men could not be reached with an
invitation. To be included in the original GOOD study, subjects
had to be between 18 and 20 yr of age and willing to participate
in the study (27). The original GOOD cohort was found representative of the general young male population in Gothenburg
(27). To determine whether the cohort of the present study also
was representative of the initial population, we compared the
age, height, weight, and amount of present physical activity (all
variables measured at the time of inclusion in the original GOOD
study) of the included subjects (n ⫽ 829) with the subjects that
were not included (n ⫽ 239) in the present study. There were no
significant differences between the included and not included
subjects in age (years) (included 18.9 ⫾ 0.6, not included
18.8 ⫾ 0.6, P ⫽ 0.114), height (centimeters) (181.6 ⫾ 6.7 vs.
180.9 ⫾ 6.9, P ⫽ 0.173), weight (kilograms) (73.9 ⫾ 11.8 vs.
73.8 ⫾ 12.1, P ⫽ 0.906), or amount of present physical activity
(hours per week) (4.3 ⫾ 5.1 vs. 4.6 ⫾ 6.0, P ⫽ 0.373) [using an
independent samples t test (mean ⫾ SD)]. The study was approved
by the regional ethical review board at the University of Gothenburg. Written and oral informed consent was obtained from
all study participants.
Assessment of present physical activity
A standardized self-administered questionnaire, based on a
validated physical activity questionnaire to measure the effect of
mechanical strain on bone mass (28) with amendments, was used
to collect information about patterns of present physical activity
in sports. Occupational or leisure manual labor was not considered. Information on the type of physical activity as well as time
(hours per week and weeks per year) and number of years spent
on all present physical activities in relation to sports was collected. A total of 529 subjects were physically active, whereas
300 had not participated in any physical activity during the past
12 months (sedentary). A total amount of physical activity
(hours per year) for each subject was calculated by multiplying
the time spent on physical activity (hours per week) with duration of physical activity (weeks per year) for each type of physical
activity during the past 12 months and then summarizing all the
products for all types of physical activity for each subject. Subjects were divided into sedentary (n ⫽ 300) and three equal
groups according to the amount of physical activity during the
past 12 months (hours per year). Tertiles divided the group as
follows: group 1 (10 –116 h/yr, n ⫽ 169), group 2 (117–239 h/yr,
n ⫽ 181), and group 3 (⬎239 h/yr, n ⫽ 179). Subjects were also
divided into sedentary (n ⫽ 300) and three equal groups according to the number of years they had been physically active with
their present sport activity. Sport activity with the longest duration, counted as number of active years from starting age until 12
months before inclusion, was used to divide the group as follows:
group 1 (⬍5 yr, n ⫽ 175), group 2 (5–12 yr, n ⫽ 177), and group
3 (⬎12 yr, n ⫽ 177).
Physical activity type (strain score) was categorized according
to peak strain score, based on ground reaction forces of physical
activity and classified according to a method previously described (29, 30). Activities involving jumping actions (e.g. gymnastics, handball, basketball) were given a strain score of 3, activities including explosive actions like turning and sprinting (e.g.
soccer, floor ball, badminton) were given a strain score of 2,
whereas other weight-bearing activities (e.g. jogging, martial
arts, strength training) were given a strain score of 1. Nonimpact
activities (e.g. swimming, bicycling, and sailing) were given a
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
jcem.endojournals.org
strain score of 0 (29, 30). The following types of physical activity,
with related strain score, were the most common among subjects
that were active in sports (each subject could have participated
in several types of sports): strength training (n ⫽ 241, strain
score ⫽ 1), soccer (n ⫽ 110, strain score ⫽ 2), running/jogging
(n ⫽ 104, strain score ⫽ 1), floor ball (n ⫽ 61, strain score ⫽ 2),
martial arts (n ⫽ 48, strain score ⫽ 1), bicycling/spinning (n ⫽
25, strain score ⫽ 0), badminton (n ⫽ 22, strain score ⫽ 2), tennis
(n ⫽ 18, strain score ⫽ 2), swimming/diving (n ⫽ 16, strain
score ⫽ 0), ice hockey (n ⫽ 15, strain score ⫽ 2), and handball
(n ⫽ 15, strain score ⫽ 3). Type of physical activity with the
highest strain score during the past 12 months was used to divide
the subjects into four groups according to the degree of mechanical loading on weight-bearing bone (strain score 0, n ⫽ 309;
strain score 1, n ⫽ 280; strain score 2, n ⫽ 220; strain score 3,
n ⫽ 20). Strain score (strain score 0 ⫺ strain score 3) was based
on ground reaction force for each type of physical activity. Strain
score 0 included both subjects with physical activity with strain
score ⫽ 0 (n ⫽ 9) and physically inactive subjects (sedentary).
2919
Bone microarchitectural measurement
A high-resolution 3D peripheral quantitative computed
tomography (QCT) device (XtremeCT; Scanco Medical AG,
Brüttisellen, Switzerland) was used to scan the ultradistal tibia
and the ultradistal radius of the nondominant leg and arm, respectively. The right arm and leg of right-handed men was defined as their dominant side, whereas the left arm and leg of
left-handed men was defined as their dominant side. Anatomically formed carbon fiber shells, especially designed for each type
of limb (Scanco Medical), were used to immobilize the subject’s
arm or leg during the scan. The measurements of the volume of
interest in the ultradistal tibia and radius, 1 cm in the proximal
direction and the whole cross-section in transversal direction,
were carried out according to a standardized protocol previously
described (31, 32). Briefly, a reference line was manually placed
at the center of the end plate of the distal tibia and distal radius.
The first computed tomography slice started 22.5 and 9.5 mm
proximal to the reference line for the tibia and radius, respectively. One hundred ten parallel computed tomography slices,
TABLE 1. Characteristics and bone microstructure of the cohort according to the total amount of physical activity
during the past 12 months
Amount of present physical activity (h/yr)
Number of subjects
Age (yr)
Height (cm)
Weight (kg)
Calcium intake (mg/d)
Smoking (%)
Alcohol intake (cl/wk)
Total amount of present
physical activity (h/yr)
Duration of previous
physical activity (yr)
Tibia (n ⫽ 828)
D.Trab (mg/cm3)
BV/TV (%)
Tb.Th (␮m)
Cort.Th (mm)
Cort.Pm (mm)
D.Cort (mg/cm3)
Tot.Area (mm2)
Radius (n ⫽ 743)
D.Trab (mg/cm3)
BV/TV (%)⫺1
Tb.N (mm )
Tb.Th (␮m)
Cort.CSA (mm2)
Cort.Th (mm)
Cort.Pm (mm)
D.Cort (mg/cm3)
Tot.Area (mm2)
Sedentary
300
24.1 ⫾ 0.6
181.9 ⫾ 7.1
79.0 ⫾ 15.4
796 ⫾ 532
11.7
11.0 ⫾ 13.7
Group 1
(10 –116)
169
24.1 ⫾ 0.7
181.9 ⫾ 6.3
77.1 ⫾ 10.3
708 ⫾ 437
7.1
10.0 ⫾ 11.3
64 ⫾ 26
Group 2
(117–239)
181
24.0 ⫾ 0.6
182.8 ⫾ 6.2
78.4 ⫾ 11.9
801 ⫾ 483
4.4A
9.7 ⫾ 15.5
176 ⫾ 37
7.0 ⫾ 6.2
9.2 ⫾ 5.8B
215 ⫾ 32
17.9 ⫾ 2.7
87.6 ⫾ 11.6
1.26 ⫾ 0.27
116 ⫾ 10
873 ⫾ 33
878 ⫾ 151
214 ⫾ 34
17.8 ⫾ 2.8
87.0 ⫾ 10.7
1.28 ⫾ 0.28
a
118 ⫾ 9
872 ⫾ 32
913 ⫾ 135a
197 ⫾ 37
16.4 ⫾ 3.1
2.09 ⫾ 0.26
78.7 ⫾ 12.8
68 ⫾ 14
0.85 ⫾ 0.18
80 ⫾ 7
850 ⫾ 41
354 ⫾ 64
194 ⫾ 34
16.1 ⫾ 2.8
2.05 ⫾ 0.26
79.2 ⫾ 12.5
69 ⫾ 14
0.84 ⫾ 0.18
82 ⫾ 6A
843 ⫾ 42
370 ⫾ 52A
Group 3
(>239)
179
24.1 ⫾ 0.6
181.8 ⫾ 7.1
79.2 ⫾ 9.4
837 ⫾ 551
2.8A
11.4 ⫾ 20.3
425 ⫾ 194
ANOVA
P1
P2
0.478
0.414
0.394
0.109
0.658
11.6 ⫾ 6.1B,C
⬍0.001
221 ⫾ 30a,b
18.4 ⫾ 2.5a,b
89.2 ⫾ 11.2
1.31 ⫾ 0.27
118 ⫾ 9A
869 ⫾ 31
915 ⫾ 137A
229 ⫾ 31A,B,c
19.0 ⫾ 2.6A,B,c
89.3 ⫾ 10.7
1.40 ⫾ 0.34A,B,C
119 ⫾ 10A
872 ⫾ 33
920 ⫾ 148A
⬍0.001
⬍0.001
0.106
⬍0.001
0.002
0.510
0.004
⬍0.001
⬍0.001
0.057
⬍0.001
⬍0.001
0.855
⬍0.001
203 ⫾ 31bb
16.9 ⫾ 2.6
2.11 ⫾ 0.24b
80.7 ⫾ 11.5
71 ⫾ 12a
0.87 ⫾ 0.16
82 ⫾ 7a
844 ⫾ 35
365 ⫾ 53
212 ⫾ 36A,B,C
17.6 ⫾ 3.0A,B,C
2.17 ⫾ 0.26A,B,c
81.8 ⫾ 13.3
73 ⫾ 14A,B
0.89 ⫾ 0.19
83 ⫾ 8A
845 ⫾ 41
375 ⫾ 66A
⬍0.001
⬍0.001
0.001
0.061
0.004
0.071
0.002
0.211
0.002
⬍0.001
⬍0.001
0.002
0.111
0.020
0.109
⬍0.001
0.213
⬍0.001
Unadjusted values are given as mean ⫾ SD. Capital and lowercase letters represent P ⬍ 0.01 and P ⬍ 0.05, respectively. Capital bold type letters
represent P ⬍ 0.001.
a
P values for vs. sedentary; b vs. group 1; c vs. group 2. Differences between groups tested by ANOVA followed by least significant difference post
hoc test for continuous variables or by ␹2 for categorical variables.
1
ANOVA P values for unadjusted variables.
2
ANOVA P values for bone microstructure parameters adjusted for age, height, weight, calcium and alcohol intake, and smoking status.
2920
Nilsson et al.
Physical Activity, Trabecular Microstructure, and Cortical Bone
with a nominal isotropic resolution (voxel size) of 82 ␮m, were
obtained at each skeletal site, delivering a 3D representation of
an approximately 9-mm section of both the tibia and radius in
the proximal direction. At each skeletal site, the entire volume of
interest was automatically separated into a cortical and a trabecular region. From this separation and by previously described
methods to process the data (32), we obtained volumetric trabecular bone density (D.Trab; milligrams per cubic centimeter),
trabecular bone volume fraction (BV/TV, percent), trabecular
number (Tb.N; millimeters⫺1), trabecular thickness (Tb.Th; micrometers), cortical cross-sectional area (Cort.CSA; square millimeters), cortical thickness (Cort.Th; millimeters), cortical periosteal circumference (Cort.Pm; millimeters), volumetric cortical
bone density (D.Cort; milligrams per cubic centimeter), and total
bone area (Tot.Area; square millimeters).
The quality of the measurements on the tibia and radius were
assessed by a five-grade scale, recommended by the manufacturer (Scanco Medical), in which 1 had the highest quality, 2–3
acceptable quality (included in the analyses), and grade 4 –5 unacceptable quality (excluded from the analyses), due to artifacts
caused by inadequate limb fixation. A total of one measurement
of the leg and 86 measurements of the arm were considered to
have unacceptable quality (grade 4 and 5), leaving 828 subjects
for further analyses of the tibia and 743 for further analyses of
the radius. The coefficients of variation (CVs) for the used bone
measurements were obtained by three repeated measurements
according to the standardized protocol on two subjects. The CVs
ranged from 0.04 to 1.6% of the tibia (D.Trab, 0.2%; BV/TV,
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
0.3%; Tb.N, 1.6%; Tb.Th, 0.7%; Cort.CSA, 0.4%; Cort.Th,
0.3%; Cort.Pm, 0.1%; D.Cort, 0.1%; and Tot.Area, 0.04%)
and from 0.04 to 3.7% of the radius (D.Trab, 0.5%; BV/TV,
0.8%; Tb.N, 3.7%; Tb.Th, 3.7%; Cort.CSA, 0.9%; Cort.Th,
1.1%; Cort.Pm, 0.2%; D.Cort, 0.3%; and Tot.Area, 0.04%).
The same device, software, and operator were used throughout
the whole study.
Assessment of covariates
Height and weight were measured using standardized
equipment. The CV values were less than 1% for these measurements. A standardized self-administered questionnaire
was used to collect information about calcium and alcohol
intake and smoking (yes/no). Alcohol intake (centiliters of
pure ethanol per week) was calculated from questionnaire
data regarding weekly consumption of beer, wine, cider, and
liquors. Calcium intake (milligrams per day) was estimated
from dairy product intake.
Statistical analysis
All data were analyzed using SPSS software, version 15.0 for
Windows (SPSS Inc., Chicago, IL). Differences in characteristics
and bone microstructure between subjects divided according to
total amount of physical activity during the past 12 months (sedentary and amount groups 1–3), the degree of mechanical loading (strain score 0 –3), or duration of previous physical activity
(sedentary and duration groups 1–3) were calculated using
TABLE 2. Characteristics and bone microstructure of the cohort according to the degree of mechanical loading in
weight-bearing bone due to present physical activity
Degree of mechanical loading (strain score) in weight-bearing bone
Number of subjects
Age (yr)
Height (cm)
Weight (kg)
Calcium intake (mg/d)
Smoking (%)
Alcohol intake (cl/wk)
Total amount of present
physical activity(h/yr)
Duration of previous physical
activity (yr)
Tibia (n ⫽ 828)
D.Trab (mg/cm3)
BV/TV (%)
Tb.Th (␮m)
Cort.Th (mm)
Cort.Pm (mm)
D.Cort (mg/cm3)
Tot.Area (mm2)
Strain
score 0
309
24.1 ⫾ 0.6
181.9 ⫾ 7.0
79.0 ⫾ 15.3
791 ⫾ 529
11.7
11.1 ⫾ 13.5
100 ⫾ 82#
Strain
score 1
280
24.1 ⫾ 0.6
181.8 ⫾ 6.8
77.6 ⫾ 10.2
792 ⫾ 515
5.0A
9.2 ⫾ 16.5
216 ⫾ 200
2.9 ⫾ 4.2#
6.5 ⫾ 4.8a
215 ⫾ 32
17.9 ⫾ 2.7
87.5 ⫾ 11.6
1.26 ⫾ 0.27
116 ⫾ 10
873 ⫾ 33
879 ⫾ 150
215 ⫾ 30
17.9 ⫾ 2.5
88.1 ⫾ 11.5
1.27 ⫾ 0.30
118 ⫾ 9a
869 ⫾ 33
906 ⫾ 139a
Strain
score 2
220
24.0 ⫾ 0.6
182.5 ⫾ 6.1
78.4 ⫾ 11.0
773 ⫾ 458
3.6A
11.5 ⫾ 15.4
228 ⫾ 163a
ANOVA
Strain
score 3
20
24.0 ⫾ 0.6
184.6 ⫾ 7.5
84.9 ⫾ 8.6
851 ⫾ 645
10.0
14.6 ⫾ 21.2
359 ⫾ 275A–C
0.504
0.254
0.071
0.912
12.7 ⫾ 6.3A,B
13.6 ⫾ 4.9A,B
⬍0.001
228 ⫾ 33A,B
19.0 ⫾ 2.7A,B
89.2 ⫾ 10.3
1.39 ⫾ 0.29A,B
119 ⫾ 9A
874 ⫾ 30
923 ⫾ 138A
245 ⫾ 37A,B,c
20.4 ⫾ 3.1A,B,c
88.3 ⫾ 9.5
1.43 ⫾ 0.35A,b
123 ⫾ 10A,b
871 ⫾ 31
982 ⫾ 157A,b
⬍0.001
⬍0.001
0.396
⬍0.001
⬍0.001
0.321
⬍0.001
1
P
P2
0.187
0.002
⬍0.001
⬍0.001
0.272
⬍0.001
⬍0.001
0.054
⬍0.001
Unadjusted values are given as mean ⫾ SD. Capital and lowercase letters represent P ⬍ 0.01 and P ⬍ 0.05, respectively. Capital bold type letters
represent P ⬍ 0.001.
a
P values for vs. strain score 0; b P values for vs. strain score 1; c P values for vs. strain score 2. Differences between groups tested by ANOVA
followed by least significant difference post hoc test for continuous variables or by ␹2 for categorical variables.
1
ANOVA P values for unadjusted variables.
2
ANOVA P values for bone microstructure parameters adjusted for age, height, weight, calcium and alcohol intake, and smoking status.
#
n ⫽ 9 (physically inactive subjects were excluded in the analysis for total amount of physical activity).
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
jcem.endojournals.org
ANOVA followed by least significant difference post hoc test for
continuous variables. A ␹2 test was used for categorical variables
(Tables 1–3). The independent predictors of various bone parameters of the tibia were tested using multiple linear regression
analysis, including covariates age, height, weight, calcium and
alcohol intake, and smoking together with strain score, total
amount of physical activity during the past 12 months, and
duration of previous activity (Table 4) in the same analysis.
This analysis, excluding strain score as a predictor variable,
was also performed for the radius. Amount of present physical
activity and duration of previous physical activity were entered as continuous variables, whereas strain score was entered as a categorized variable (strain 0 –3). Due to the high
correlation observed between the physical activity variables,
a physical activity indicator variable (physical activity yes/no)
was introduced to reduce problems of multicollinearity between the other three investigated physical activity parameters (Table 4). The percentage of the variation (R2) of each
bone parameter, explained by the whole model, all combined
physical activity variables or by each individual physical activity variable, was calculated using the linear regression
2921
model. R2 for each variable was calculated as the R2 change
of the entire model when adding each variable, until all variables were included in the regression model (Table 4). Weight
was not normally distributed and was logarithmically transformed before entered into the regression model.
Results
Association between physical activity, cortical
bone structure, and trabecular microstructure
Characteristics of the cohort divided according to the total
amount of present physical activity, the degree of mechanical
loading due to physical activity, or duration of previous physical activity are shown in Tables 1–3, respectively.
Total amount of physical activity (hours per year) and
degree of mechanical loading in weight-bearing bone
(tibia) due to physical activity (strain score 0 –3) during the
past 12 months as well as duration of previous physical
TABLE 3. Characteristics and bone microstructure of the cohort according to duration of previous physical activity
Duration of previous physical activity
Number of subjects
Age (yr)
Height (cm)
Weight (kg)
Calcium intake (mg/d)
Smoking (%)
Alcohol intake (cl/wk)
Duration of previous
physical activity (yr)
Total amount of present
physical activity (h/yr)
Tibia (n ⫽ 828)
D.Trab (mg/cm3)
BV/TV (%)
Tb.Th (␮m)
Cort.Th (mm)
Cort.Pm (mm)
D.Cort (mg/cm3)
Tot.Area (mm2)
Radius (n ⫽ 743)
D.Trab (mg/cm3)
BV/TV (%)
Tb.N (mm⫺1)
Tb.Th (␮m)
Cort.CSA (mm2)
Cort.Th (mm)
Cort.Pm (mm)
D.Cort (mg/cm3)
Tot.Area (mm2)
ANOVA
Group 1
(<5 yr)
175
24.1 ⫾ 0.7
181.8 ⫾ 6.6
76.7 ⫾ 10.6
766 ⫾ 508
7.4
9.9 ⫾ 14.7
2.4 ⫾ 1.3
Group 2
(5–12 yr)
177
24.1 ⫾ 0.6
182.1 ⫾ 6.5
78.9 ⫾ 12.0
805 ⫾ 505
2.3A,b
9.6 ⫾ 19.0
8.4 ⫾ 2.1
Group 3
(>12 yr)
177
24.0 ⫾ 0.6
182.7 ⫾ 6.6
79.1 ⫾ 8.9
779 ⫾ 475
4.5A
11.7 ⫾ 14.4
17.1 ⫾ 2.1
0.383
0.537
0.205
0.886
161 ⫾ 155
240 ⫾ 201B
272 ⫾ 193B
⬍0.001
215 ⫾ 32
17.9 ⫾ 2.7
87.6 ⫾ 11.6
1.26 ⫾ 0.27
116 ⫾ 10
873 ⫾ 33
878 ⫾ 151
214 ⫾ 31
17.9 ⫾ 2.6
88.7 ⫾ 11.4
1.26 ⫾ 0.29
117 ⫾ 9
869 ⫾ 34
892 ⫾ 131
219 ⫾ 29
18.2 ⫾ 2.4
87.8 ⫾ 11.0
1.32 ⫾ 0.31a,b
119 ⫾ 9A
871 ⫾ 31
918 ⫾ 143A
231 ⫾ 34A–C
19.2 ⫾ 2.8A–C
89.1 ⫾ 10.3
1.41 ⫾ 0.30A,B,C
120 ⫾ 9A,B
874 ⫾ 31
938 ⫾ 142A,B
⬍0.001
⬍0.001
0.455
⬍0.001
⬍0.001
0.396
⬍0.001
⬍0.001
⬍0.001
0.365
⬍0.001
⬍0.001
0.069
⬍0.001
197 ⫾ 37
16.4 ⫾ 3.1
2.09 ⫾ 0.26
78.7 ⫾ 12.8
68 ⫾ 14
0.85 ⫾ 0.18
80 ⫾ 7
850 ⫾ 41
354 ⫾ 64
195 ⫾ 34
16.3 ⫾ 2.8
2.06 ⫾ 0.26
79.3 ⫾ 12.9
67 ⫾ 14
0.84 ⫾ 0.18
81 ⫾ 7
840 ⫾ 40
359 ⫾ 55
205 ⫾ 33a,B
17.1 ⫾ 2.7a,B
2.12 ⫾ 0.24b
81.0 ⫾ 12.1
72 ⫾ 14A,B
0.88 ⫾ 0.19
83 ⫾ 7A,b
845 ⫾ 41
375 ⫾ 56A,b
208 ⫾ 34A,B
17.3 ⫾ 2.9A,B
2.14 ⫾ 0.25B
81.4 ⫾ 12.3
73 ⫾ 13A,B
0.88 ⫾ 0.16
83 ⫾ 7A,b
847 ⫾ 36
376 ⫾ 59A,b
0.001
0.001
0.039
0.095
⬍0.001
0.093
⬍0.001
0.062
⬍0.001
0.002
0.002
0.075
0.106
0.001
0.068
⬍0.001
0.030
⬍0.001
Sedentary
300
24.1 ⫾ 0.6
181.9 ⫾ 7.1
78.9 ⫾ 15.4
796 ⫾ 532
11.7
11.0 ⫾ 13.7
1
P
P2
0.538
Unadjusted values are given as mean ⫾ SD. Capital and lowercase letters represent P ⬍ 0.01 and P ⬍ 0.05, respectively. Capital bold type letters
represent P ⬍ 0.001.
a
P values for vs. sedentary, b vs. group 1, c vs. group 2. Differences between groups tested by ANOVA followed by least significant difference post
hoc test for continuous variables or by ␹2 for categorical variables.
1
ANOVA P values for unadjusted variables; 2 ANOVA P values for bone microstructure parameters adjusted for age, height, weight, calcium and
alcohol intake, and smoking status.
2922
Nilsson et al.
Physical Activity, Trabecular Microstructure, and Cortical Bone
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
TABLE 4. Linear regression analysis of the association between physical activity parameters, trabecular bone
microstructure, and cortical bone size in young men
Total amount of
present physical
activity (h/yr)
R2 (%)
Tibia (n ⫽ 828)
D.Trab (mg/cm3)
BV/TV (%)
Tb.N (mm⫺1)
Cort.CSA (mm2)
Cort.Th (mm)
Cort.Pm (mm)
Tot.Area (mm2)
Radius (n ⫽ 743)
D.Trab (mg/cm3)
BV/TV (%)
Tb.N (mm⫺1)
Cort.CSA (mm2)
Cort.Th (mm)
Cort.Pm (mm)
Tot.Area (mm2)
P
0.5
0.5
0.3
—
—
—
—
0.032
0.032
0.048
0.116
0.190
0.652
0.859
2.2
2.2
1.9
0.5
0.5
—
—
⬍0.001
⬍0.001
⬍0.001
0.040
0.041
0.987
0.563
Degree of
mechanical
loading (strain
score) in weightbearing bone
Duration of
previous physical
activity (yr)
Physical activity
R2 (%)
P
R2 (%)
P
(yes/no) R2 (%)
Alla R2 (%)
1.3
1.3
0.6
1.3
1.0
—
—
⬍0.001
⬍0.001
0.009
⬍0.001
0.001
0.950
0.784
0.5
0.5
0.4
1.7
1.0
0.3
0.3
0.041
0.041
0.040
⬍0.001
0.001
0.036
0.030
1.5
1.5
0.8
1.0
1.0
0.1
0.1
5.2
5.2
3.2
9.9
5.9
1.8
1.7
—
—
—
0.7
—
—
—
0.080
0.078
0.176
0.022
0.118
0.079
0.092
0.4
0.4
0.6
0.1
0.3
0.3
0.2
3.8
3.8
2.7
2.3
1.2
2.0
2.0
Linear regression including all of the following variables simultaneously: age, height, weight, smoking, calcium and alcohol intake, present physical
activity (yes/no), total amount of present physical activity, strain score (tibia analyses only), and duration of previous physical activity. P values and
R2 are presented.
a
R2 for all four physical activity parameters.
activity (years) were associated with tibial D.Trab, BV/
TV, Tb.N, Cort.CSA, Cort.Th, Cort.Pm, and Tot.Area
(Tables 1–3 and Figs. 1 and 2). Subjects in the group with
the highest amount of present physical activity (group 3)
had higher tibial BV/TV (6.3⌬, 6.7⌬, and 3.3⌬%), Tb.N
(4.3, 4.0, and 3.0%), and Cort.CSA (14.5, 10.4, and 7.7%)
than subjects in all other subgroups with a lower amount of
present physical activity (sedentary, amount groups 1 and 2),
respectively (Table 1 and Figs. 1A and 2A). Men in the groups
with the highest physical activity strain score (strain scores 3
and 2) had higher tibial BV/TV (13.9⌬ and 6.0⌬%, respectively), Tb.N (12.7 and 4.6%), and Cort.CSA (17.9 and
10.6%) than men with the lowest weight-bearing physical
activity strain score (strain score 1) (Table 2 and Figs. 1B and
2B). Similarly, subjects in the group with the longest duration
of physical activity (duration group 3) had higher tibial
BV/TV (7.5⌬, 7.7⌬, and 5.5⌬%), Tb.N (5.4, 6.8, and 3.6%),
and Cort.CSA (16.1, 15.2, and 7.9%) than both sedentary
subjects and subjects with a shorter duration of previous
physical activity (duration groups 1 and 2), respectively (Table 3 and Figs. 1C and 2C).
Because the strain score assessment pertains solely to
loading the weight-bearing skeleton experiences, the analyses of microstructure at the radius were only performed
concerning total amount and duration of physical activity
and not according to strain score. Both the total amount
of present physical activity (hours per year) and duration
of previous physical activity before the past 12 months
were associated with D.Trab, BV/TV, Tb.N, Cort.CSA,
Cort.Pm, and Tot.Area at the radius (Tables 1 and 3).
There were no significant differences in age, height,
weight, or daily calcium and alcohol intake between the
four subgroups divided by total amount, strain score, or
duration (Tables 1–3). However, subjects in the group
with the highest strain score (strain score 3) had been more
physically active, during the past 12 months, than subjects
in all of the other subgroups, and subjects in the strain
score 2 group had been more physically active than subjects in the group with the lowest strain score (Table 2).
Total amount and duration of physical activity as well as
strain score of physical activity in weight-bearing bone
were highly correlated (n ⫽ 829; r ⫽ 0.78 – 0.87, P ⬍
0.001) in the whole cohort but to a lesser extent in the
training men (r ⫽ 0.15– 0.48). Furthermore, subjects in
both the amount and duration groups 2 and 3 were less
frequently smokers than subjects in the sedentary group
(Tables 1 and 3), whereas subjects in strain score 1 and
strain score 2 groups were less frequently smokers than
subjects in strain score 0 group (Table 2).
To determine whether present total amount of physical
activity, degree of mechanical loading (strain score), or
duration of previous physical activity were independent
predictors of bone parameters in weight-bearing bone
(tibia), a linear regression analysis was used. This analysis
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
jcem.endojournals.org
2923
the radius (Table 4). In contrast, both duration of previous physical activity and total amount of present physical activity were
independently associated with Cort.CSA.
Neither amount of present physical activity
nor previous duration of physical activity
independently predicted Cort.Pm or Tot.Area at the radius (Table 4).
Discussion
In this population-based study, we investigated the relationship between trabecular microstructure and cortical
bone size in weight-bearing bone, using
high-resolution 3D peripheral QCT
and habits of physical activity in young
men. A previous randomized controlled
trial of the effects of Tai Chi Chun
FIG. 1. Total amount of physical activity (A) (hours per year) and degree of mechanical
exercise on BMD in postmenopausal
loading in weight-bearing bone (tibia) due to physical activity (B) (strain score 0 –3) during the
women reported that the intervention
past 12 months as well as duration of previous physical activity (C) (years) were associated
resulted in reduced trabecular bone loss
with tibial Tb.N (millimeters⫺1). Sedentary (n ⫽ 299); amount group 1 (10 –116 h/yr, n ⫽
169), amount group 2 (117–239 h/yr, n ⫽ 181), and amount group 3 (⬎239 h/yr, n ⫽ 179);
of the distal tibia (33). To our knowlduration group 1 (⬍5 yr, n ⫽ 175), duration group 2 (5–12 yr, n ⫽ 177), and duration group
edge, this is the first study examining
3 (⬎12 yr, n ⫽ 177). Strain score 0 (n ⫽ 308), strain score 1 (n ⫽ 280), strain score 2 (n ⫽
the association between exercise and
220), and strain score 3 (n ⫽ 20). Strain score 0 included both subjects with physical activity
with strain score ⫽ 0 (n ⫽ 9) and physically inactive subjects (sedentary). Unadjusted values
trabecular bone microstructure in men.
are given as mean ⫾ SD. Capital and lowercase letters represent P ⬍ 0.01 and P ⬍ 0.05,
We demonstrated a positive association
respectively. Capital bold type letters represent P ⬍ 0.001. P values are for vs. sedentary or
between present physical activity and
strain score 0 (a), vs. amount group 1, strain score 1 or duration group 1 (b), or vs. amount
augmented trabecular bone volume
group 2, strain score 2 or duration group 2 (c), respectively. Differences between groups
tested by ANOVA followed by least significant difference post hoc test (n ⫽ 828).
fraction due to increased trabecular
number but not altered trabecular
(including age, height, weight, calcium and alcohol intake, thickness. Furthermore, we found that parameters reflectand smoking as covariates) revealed that total amount and ing trabecular microstructure in weight-bearing bone was
strain score of present physical activity independently premainly associated with degree of current mechanical loaddicted D.Trab, BV/TV, and Tb.N in weight-bearing bone
ing due to type of present physical activity, indicating the
(Table 4). Strain score also predicted Cort.CSA and
importance of mechanical loading on trabecular microCort.Th, whereas duration of physical activity predicted
structure. In contrast, our results suggest that number of
all investigated bone parameters of the tibia (Table 4).
years spent on physical activity could be of importance in
Strain score was predominantly associated with trabecaugmenting the cortical cross sectional area and thickness
ular bone parameters and could alone explain 1.3% of
the variation in BV/TV (Table 4). Duration of previous at the tibia. Thus, our findings support that present
physical activity mainly predicted cortical Cort.CSA (within the last year) in addition to past (previous years)
and could explain 1.7% of the total variation in this physical activity could be important for the quickly rebone trait. All physical activity parameters could ex- modeled trabecular bone. In addition, our results further
plain 5.2 and 9.9% of the variation in BV/TV and imply that previous physical activity, during growth, is of
importance for the development of cortical bone size.
Cort.CSA, respectively (Table 4).
The results of the present study are consistent with anThe above-used linear regression analysis (without
other
study and our previous findings in that an associaphysical activity strain score) was used to investigate the
association between physical activity and bone parame- tion between trabecular vBMD and mechanical loading
ters of the radius. In this model, total amount of present via amount of physical activity was found (15, 34). Conphysical activity, but not duration of previous physical sistent with our previous results from the GOOD cohort,
activity,wasindependentlyassociatedwithBV/TVandTb.Nat the group with the lowest amount of physical activity
2924
Nilsson et al.
Physical Activity, Trabecular Microstructure, and Cortical Bone
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
as much as 12.7% in subjects in the
group with the highest physical activity
strain score compared with subjects
with the lowest weight-bearing physical
activity strain score. Both bone size and
volumetric density is of great importance concerning mechanical strength
of the bone and resistance against fracture (8, 9). In accordance with previous
findings in the same cohort (15), we also
demonstrated a positive association between physical activity duration and
augmented cortical bone size. Because
our present study reveals a positive association between high degree of mechanical loading, due to physical activity,
and trabecular bone microstructure
and between number of years spent on
physical activity and cortical bone parameters, we suggest that both highload physical activity and a start of exFIG. 2. Total amount of physical activity (A) (hours per year) and degree of mechanical
ercise at a young age could be important
loading in weight-bearing bone (tibia) due to physical activity (B) (strain score 0 –3) during the
past 12 months as well as duration of previous physical activity (C) (years) were associated
contributors to increased bone strength
with tibial Cort.CSA (square millimeters). Sedentary (n ⫽ 299); amount group 1 (10 –116 h/yr,
and resistance against fracture.
n ⫽ 169), amount group 2 (117–239 h/yr, n ⫽ 181), and amount group 3 (⬎239 h/yr, n ⫽
To distinguish between the relative
179); duration group 1 (⬍5 yr, n ⫽ 175), duration group 2 (5–12 yr, n ⫽ 177), and duration
group 3 (⬎12 yr, n ⫽ 177). Strain score 0 (n ⫽ 308), strain score 1 (n ⫽ 280), strain score 2
role of degree of mechanical loading
(n ⫽ 220), and strain score 3 (n ⫽ 20). Strain score 0 included both subjects with physical
and present amount or previous duraactivity with strain score ⫽ 0 (n ⫽ 9) and physically inactive subjects (sedentary). Unadjusted
tion of physical activity and because it is
values are given as mean ⫾ SD. Capital and lowercase letters represent P ⬍ 0.01 and P ⬍
well known that physical activity types
0.05, respectively. Capital bold type letters represent P ⬍ 0.001. P values are for vs. sedentary
or strain score 0 (a), vs. amount group 1, strain score 1 or duration group 1 (b), vs. amount
with high strain score give the most fagroup 2, strain score 2 or duration group 2 (c), respectively. Differences between groups
vorable bone acquisition (21–23), we
tested by ANOVA followed by least significant difference post hoc test (n ⫽ 828).
categorized the reported physical activity types according to peak strain score
had lower trabecular vBMD, indicating that only high
in weight-bearing bone. This categorization was based on
amounts of exercise can increase amount of trabecular
ground reaction forces of physical activity and classified
bone (15, 34). By using high-resolution peripheral QCT,
according to a method previously described (29, 30). To
we could in the present study further discriminate between
our knowledge, a similar classification does not exist for
which of the trabecular bone microarchitectural parameloading of non-weight-bearing bone, i.e. upper extremiters that most highly contribute to this increase in trabecties. The analyses according to strain score were therefore
ular vBMD, associated with physical activity.
It has previously been reported, in computer-generated not applied on measurements of the radius. Both present
models using 3D reconstructions of trabecular bone ar- amount and previous duration of physical activity were
chitecture, that loss of trabeculae is of greater importance associated with several microarchitectural parameters at
for the bone strength rather than reduced trabecular thick- the radius, but only the present amount was independently
ness (12). In the present study, all physical activity pa- associated with trabecular parameters. It is plausible that
rameters in general, and physical activity strain score in factors other than mechanical loading could have contribparticular, were found to be associated with increased uted to this improvement of bone parameters of the radius.
trabecular number, indicating physical activity with a One explanation could be that systemic factors, e.g. GH,
high-load physical activity results in increased bone being released after endurance exercise could be of imstrength due to improved trabecular bone microstructure. portance in determining microarchitectural parameters
Although type of physical activity was found to be most (35, 36). However, because some of the most common
strongly associated with trabecular bone volume fraction physical activities reported in the present study involves
(13.9% difference), trabecular number was increased by the upper extremities, e.g. strength training, soccer, floor
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
ball, and martial arts, the mechanical loading could also,
at least partly, be the cause of the augmentation of bone
parameters of the radius.
There are some limitations associated with the present
study. Due to the loss of men from the original cohort at
follow-up, we cannot be certain that the presently investigated cohort is representative of the original cohort. Furthermore, the results from the present study derive from
investigations on 23- to 25-yr-old men and may not therefore be applicable to other age groups. Present and former
physical activity habits were assessed using a retrospective
self-reporting questionnaire, which could have limited the
ability of the subjects to recall their history of physical
activity and cause bias and misclassification. However, by
using a standardized self-administered questionnaire,
based on a validated physical activity questionnaire, with
amended questions concerning physical activity habits
over the whole year to measure the effect of mechanical
loading on bone mass (28), we believe that we have increased our possibilities to collect accurate information
about physical activity habits. Furthermore, some studies
have reported that people can recall activity patterns of up
to 10 yr ago with high reliability and that recall of more
vigorous activity, like sports and exercise, is more accurate
than recall of less intensive activities (37–39). We also
assessed calcium intake using a questionnaire, but our
questions were focused on calcium intake from dairy
products, which have been shown to be the source of about
65% of the daily calcium intake (40). Thus, the calculated
calcium intake in our population is most likely underestimated. In addition, we have not been able to control for
supplements of either calcium or other minerals or vitamins,
which could have confounded the presented analyses. The
cross-sectional design does not allow for direct cause-effect
relationships to be established. Using a regression model (Table 4) with all physical activity variables, which were highly
correlated, could give rise to problems of multicollinearity.
By using an indicator variable (physical activity yes/no), the
ability to discern the independent contribution of the physical activity variables was increased.
In conclusion, we demonstrate, in a large cohort of
young men, that the degree of mechanical loading due to
type of present physical activity was predominantly associated with trabecular microstructure, whereas duration
of previous physical activity was mainly associated parameters of cortical bone size in weight-bearing bone.
Acknowledgments
Address all correspondence and requests for reprints to: Mattias
Lorentzon, Associate professor, M.D., Division of Endocrinology, Department of Internal Medicine, Gröna Stråket 8, Sahlg-
jcem.endojournals.org
2925
renska University Hospital, 413 45 Gothenburg, Sweden.
E-mail: [email protected].
This work was supported by the Swedish Research Council,
the Swedish Foundation for Strategic Research, European Commission, the Lundberg Foundation, the Torsten and Ragnar
Söderberg’s Foundation, Petrus and Augusta Hedlund’s Foundation, the Avtal för Läkarutbildning och Forskning (ALF/LUA)
grant from the Sahlgrenska University Hospital, the Novo Nordisk Foundation, and The European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis (ESCEO)-Amgen Fellowship Award (to M.L.).
Disclosure Summary: All authors have nothing to disclose.
References
1. Giguère Y, Rousseau F 2000 The genetics of osteoporosis: ‘complexities and difficulties.’ Clin Genet 57:161–169
2. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PN,
Eberl S 1987 Genetic determinants of bone mass in adults. A twin
study. J Clin Invest 80:706 –710
3. Rizzoli R, Bonjour JP, Ferrari SL 2001 Osteoporosis, genetics and
hormones. J Mol Endocrinol 26:79 –94
4. Lanyon LE 1996 Using functional loading to influence bone mass and
architecture: objectives, mechanisms, and relationship with estrogen of
the mechanically adaptive process in bone. Bone 18:37S– 43S
5. Kannus P, Haapasalo H, Sankelo M, Sievänen H, Pasanen M,
Heinonen A, Oja P, Vuori I 1995 Effect of starting age of physical
activity on bone mass in the dominant arm of tennis and squash
players. Ann Intern Med 123:27–31
6. Welten DC, Kemper HC, Post GB, Van Mechelen W, Twisk J, Lips
P, Teule GJ 1994 Weight-bearing activity during youth is a more
important factor for peak bone mass than calcium intake. J Bone
Miner Res 9:1089 –1096
7. Mikkola TM, Sipilä S, Rantanen T, Sievänen H, Suominen H,
Kaprio J, Koskenvuo M, Kauppinen M, Heinonen A 2008 Genetic
and environmental influence on structural strength of weight-bearing and non-weight-bearing bone: a twin study. J Bone Miner Res
23:492– 498
8. Silva MJ 2007 Biomechanics of osteoporotic fractures. Injury
38(Suppl 3):S69 –S76
9. Orwoll ES 2003 Toward an expanded understanding of the role of
the periosteum in skeletal health. J Bone Miner Res 18:949 –954
10. Griffith JF, Genant HK 2008 Bone mass and architecture determination: state of the art. Best Pract Res Clin Endocrinol Metab 22:
737–764
11. Sievänen H, Kannus P, Järvinen TL 2007 Bone quality: an empty
term. PLoS Med 4:e27
12. van der Linden JC, Homminga J, Verhaar JA, Weinans H 2001
Mechanical consequences of bone loss in cancellous bone. J Bone
Miner Res 16:457– 465
13. Ma H, Leskinen T, Alen M, Cheng S, Sipilä S, Heinonen A, Kaprio
J, Suominen H, Kujala UM 2009 Long-term leisure time physical
activity and properties of bone: a twin study. J Bone Miner Res
24:1427–1433
14. Bass S, Pearce G, Bradney M, Hendrich E, Delmas PD, Harding A,
Seeman E 1998 Exercise before puberty may confer residual benefits
in bone density in adulthood: studies in active prepubertal and retired female gymnasts. J Bone Miner Res 13:500 –507
15. Lorentzon M, Mellström D, Ohlsson C 2005 Association of amount
of physical activity with cortical bone size and trabecular volumetric
BMD in young adult men: the GOOD study. J Bone Miner Res
20:1936 –1943
16. Tobias JH, Steer CD, Mattocks CG, Riddoch C, Ness AR 2007
Habitual levels of physical activity influence bone mass in 11-year-
2926
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
Nilsson et al.
Physical Activity, Trabecular Microstructure, and Cortical Bone
old children from the United Kingdom: findings from a large population-based cohort. J Bone Miner Res 22:101–109
Turner CH 1998 Three rules for bone adaptation to mechanical
stimuli. Bone 23:399 – 407
Kontulainen S, Sievänen H, Kannus P, Pasanen M, Vuori I 2002 Effect
of long-term impact-loading on mass, size, and estimated strength of
humerus and radius of female racquet-sports players: a peripheral
quantitative computed tomography study between young and old starters and controls. J Bone Miner Res 17:2281–2289
Nilsson M, Ohlsson C, Mellström D, Lorentzon M 2009 Previous
sport activity during childhood and adolescence is associated with
increased cortical bone size in young adult men. J Bone Miner Res
24:125–133
Nikander R, Sievänen H, Uusi-Rasi K, Heinonen A, Kannus P 2006
Loading modalities and bone structures at nonweight-bearing upper
extremity and weight-bearing lower extremity: a pQCT study of
adult female athletes. Bone 39:886 – 894
Fehling PC, Alekel L, Clasey J, Rector A, Stillman RJ 1995 A comparison of bone mineral densities among female athletes in impact
loading and active loading sports. Bone 17:205–210
Heinonen A, Oja P, Kannus P, Sievänen H, Haapasalo H, Mänttäri
A, Vuori I 1995 Bone mineral density in female athletes representing
sports with different loading characteristics of the skeleton. Bone
17:197–203
Taaffe DR, Robinson TL, Snow CM, Marcus R 1997 High-impact
exercise promotes bone gain in well-trained female athletes. J Bone
Miner Res 12:255–260
Nikander R, Sievänen H, Heinonen A, Kannus P 2005 Femoral neck
structure in adult female athletes subjected to different loading modalities. J Bone Miner Res 20:520 –528
Nikander R, Kannus P, Rantalainen T, Uusi-Rasi K, Heinonen A,
Sievanen H 17 November 2009 Cross-sectional geometry of weightbearing tibia in female athletes subjected to different exercise loadings. Osteoporos Int doi: 10.1007/s00198-009-1101-0
Modlesky CM, Majumdar S, Dudley GA 2008 Trabecular bone
microarchitecture in female collegiate gymnasts. Osteoporos Int 19:
1011–1018
Lorentzon M, Mellström D, Ohlsson C 2005 Age of attainment of
peak bone mass is site specific in Swedish men—the GOOD Study.
J Bone Miner Res 20:1223–1227
J Clin Endocrinol Metab, June 2010, 95(6):2917–2926
28. Kemper HC, Bakker I, Twisk JW, van Mechelen W 2002 Validation
of a physical activity questionnaire to measure the effect of mechanical strain on bone mass. Bone 30:799 – 804
29. Neville CE, Murray LJ, Boreham CA, Gallagher AM, Twisk J, Robson
PJ, Savage JM, Kemper HC, Ralston SH, Davey Smith G 2002 Relationship between physical activity and bone mineral status in young
adults: the Northern Ireland Young Hearts Project. Bone 30:792–798
30. Groothausen J, Siemer H, Kemper HCG, Twisk J, Welten DC 1997
Influence of peak strain on lumbar bone mineral density: an analysis
of 15 year physical activity in young males and females. Pediatr
Exerc Sci 9:159 –173
31. MacNeil JA, Boyd SK 2007 Load distribution and the predictive
power of morphological indices in the distal radius and tibia by high
resolution peripheral quantitative computed tomography. Bone 41:
129 –137
32. Laib A, Häuselmann HJ, Rüegsegger P 1998 In vivo high resolution
3D-QCT of the human forearm. Technol Health Care 6:329 –337
33. Chan K, Qin L, Lau M, Woo J, Au S, Choy W, Lee K, Lee S 2004 A
randomized, prospective study of the effects of Tai Chi Chun exercise on bone mineral density in postmenopausal women. Arch Phys
Med Rehabil 85:717–722
34. Nara-Ashizawa N, Liu LJ, Higuchi T, Tokuyama K, Hayashi K,
Shirasaki Y, Amagai H, Saitoh S 2002 Paradoxical adaptation of
mature radius to unilateral use in tennis playing. Bone 30:619 – 623
35. Kraemer WJ, Ratamess NA 2005 Hormonal responses and adaptations to resistance exercise and training. Sports Med 35:339 –361
36. Ohlsson C, Mohan S, Sjögren K, Tivesten A, Isgaard J, Isaksson O,
Jansson JO, Svensson J 2009 The role of liver-derived insulin-like
growth factor-I. Endocr Rev 30:494 –535
37. Blair SN, Dowda M, Pate RR, Kronenfeld J, Howe Jr HG, Parker G,
Blair A, Fridinger F 1991 Reliability of long-term recall of participation in physical activity by middle-aged men and women. Am J
Epidemiol 133:266 –275
38. Falkner KL, Trevisan M, McCann SE 1999 Reliability of recall of
physical activity in the distant past. Am J Epidemiol 150:195–205
39. Slattery ML, Jacobs Jr DR 1995 Assessment of ability to recall physical activity of several years ago. Ann Epidemiol 5:292–296
40. Wikberger C, Eidstedt M 2009 Consumption of food and nutritive
values, data up to 2007. Report 2009:5. Jönköping, Sweden: Swedish Board of Agriculture