Faculty of Sport, University of Ljubljana, ISSN 1318-2269  29 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
Povzetek
V raziskavo je bilo vključenih 40 poklicnih igralcev 
nogometnega kluba Železničar Niš in 40 študentov 
Fakultete za varnost pri delu, ki niso bili vrhunski 
športniki. Pri obeh skupinah smo določili indeks 
telesne mase (BMI) in stanje sestave kosti. Raziskava 
je vključevala merjenje mineralne gostote kosti (BMD) 
v ledvenem delu hrbtenice in kolku. Nihče izmed 
merjencev predhodno ni sodeloval v podobni raziskavi. 
Merilna naprava za določanje BMD je bil poseben 
DXP Dual Energy rentgenski Absorpciometry-DEXA 
merilnik gostote. V raziskavi smo ugotavljali tudi, ali 
je BMI lahko napovedovalec zmanjševanja mineralne 
gostote kosti na področji hrbtenice in kolkov.
Ključne besede: indeks telesne mase (BMI), mineralna 
gostota kosti (BMD), napovedovanje
Abstract
The research was conducted on a sample of 40 
players members of the Železničar Niš football 
club (professional athletes) and on the Faculty of 
Occupational Safety student population (40 non-
athletes) in order to determine the body mass index 
(BMI) and bone composition condition of the research 
subjects. The research included the measurement of 
bone mineral density (BMD) in the lumbar region of 
the spinal column and hip articulation of the subjects. 
None of subjects had previously participated in any such 
research. The measuring device used for the evaluation 
of BMD was a state-of-the-art Lunar DPX Dual Energy 
X-ray Absorpciometry-DEXA densitometer. This 
research also examined whether BMI can be a predictor 
of important decrease of bone mineral density values 
near the spinal column and near the hip articulation.
Key words: body mass index (BMI), bone mineral den-
sity (BMD), prediction.
¹University of Niš, Faculty of Sports and Physical 
Education, 18000 Niš, Serbia
²University of Novi Sad, Faculty of Sports and Physical 
Education, 21000 Novi Sad, Serbia
3
University of Ljubljana, Centre for University Sport, 1000 
Ljubljana, Slovenia
4
University of Ljubljana, Faculty of sport, 1000 Ljubljana, 
Slovenia
5
University of Niš, Faculty of Medicine, 18000 Niš, Serbia 
*Corresponding author:
Faculty of Sports and Physical Education,University of Niš
Čarnojevića 10a, 18000 Niš, Serbia
Tel. +381 18 510 900
E-mail: bubanjsasa@yahoo.co.uk
BODY MASS INDEX AS A PREDICTOR 
OF IMPORTANT DECREASE OF BONE 
MINERAL DENSITY
INDEKS TELESNE MASE KOT 
NAPOVEDOV ALEC ZMANJŠANJA 
MINERALNE GOSTOTE KOSTI
Saša Bubanj
¹
*
Ratko Stanković
¹
Borislav Obradović
2
Edvard Kolar
3
Jakob Bednarik
4
Radoslav Bubanj
¹
Emilija Petković
1
Aleksandar Dimić
5
Boris Đinđić
5

30 BMI as a predictor of bone mineral density Kinesiologia Slovenica, 15, 3, 29–39 (2009) 
INTRODUCTION
Regular physical activities exert direct and positive influence on the health and physical condition 
of people. Exercise reduces the risks of developing certain chronic diseases and contributes 
significantly to a longer life expectancy.
The skeletal system should be strong enough to support body weight as well as the efforts and 
loads to which people are exposed in life, as well as excess loads in certain situations including 
sports trainings and competitions. Various phenomena negatively influence the skeletal system; 
one of the most common is osteoporosis. 
Osteoporosis means “porous bones” and causes feebleness and brittleness of bones so that 
minimal physical activity could induce fracture. Generally, the skeletal system declines when a 
person possesses low levels of vitamin D, calcium, phosphor and other minerals in the bones. 
Osteoporosis can also be a consequence of the disturbances in the functioning of the endocrine 
system or the consequence of intensive administration of medicine, such as corticosteroids.
Most fractures caused by osteoporosis are in the area of spinal column, hip and wrist.
Bones are constantly changing by the process of remodelling in which the new bone tissue 
replaces the existing one (Bubanj, S. & Obradović, 2002). Bones develop from new binding tissue 
called mesenhim which grows two types of cells, new bone cells (osteoblasts) and large bone 
cells (osteoclasts). Osteoblasts generate bone tissue, and osteoclasts cause decay of bone tissue. 
Both types of cells act continuously and simultaneously, which is important for the normal 
development and formation of bone tissue (Bubanj, R., 1997 a; 1997 b). A complete cycle of bone 
remodelling lasts from two to three months. In a younger person, new bone cells are formed 
more rapidly and in larger quantity than the existent ones, which are reabsorbed so that the total 
bone mass increases. Bone mass reaches its peak in the middle thirties.
The main aims of this research were to determine possible existence of important decrease of bone 
mineral density (BMD) in subjects, and if body mass index (BMI) can be a predictor of important 
decrease of BMD in subjects, in the case of osteopenia or osteoporosis determination.
METHOD
Participants
The research was conducted on the sample of 80 male subjects. The sample was comprised of 40 
non-athletes chosen from the students population of the Faculty of Occupational Safety in Niš, 
who were not engaged in systematic sports training, aged between 19 and 35, and 40 football 
players of the Železničar football club in Niš, aged between 17 and 34, who have played this sport 
professionally for at least three years, and are competing in the First Serbian Football League. 
Measurements of body height, body weight and bone mineral density were performed in Radon 
Rehabilitation Centre in Niška Banja.
Va r iable s
The used variables included body mass index (BMI), reported in kg/m
2
; bone mineral density 
near the spinal column (BMDSPINE), reported in g/cm
2
; and bone mineral density near the hip 
articulation (BMDHIP), reported in g/cm
2
.

BMI as a predictor of bone mineral density 31 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
Instruments
Stadiometer was used for the measurement of body height, length of lower extremity, span 
of upper extremities and sitting height. It measured with accuracy of 0.1 cm. An abbreviated 
stadiometer was used for the measurement of body segments (length of lower extremity, upper 
extremity, feet, hand etc.). Medical scales which measure with an accuracy of 0.1 kg were used. 
Before measuring, the scales were placed on hard flat surface. The accuracy of measurement 
was verified by using weights 10 or 20 kg placed on the scales before measurement (Đurašković, 
2002).
The instrument for the evaluation of bone mineral density, when using the DEXA method was a 
Lunar DPX densitometer of current technology, which is based on dual photon absorptiometry 
of X-rays (Dual Energy X-ray Absorpciometry – DEXA apparatus). The accuracy and reproduc-
ibility of this instrument amounts to 1%. Anterior-posterior examination of lumbar vertebras 
L1-L4 was performed. The obtained data were processed by Lunar Software. Absolute values 
of bone mineral density (bone mineral density-BMD) were obtained, reported in g/cm² for each 
examined vertebra, mean values of vertebras density L1-L4 are reported in absolute values 
(g/cm
2
).Values which were +/- one standard deviation upon 100% of values, i.e. +/-12%, were 
conceived normally (Dimić, M, 2005).
Procedures
With a view of representing the distribution of analyzed variables, descriptive statistics were 
applied. A t-test for independent samples was used, as well as method of survival analyses in aim 
of diagnosis the important decrease of BMD based on BMI.
R ESULTS
Table 1: Descriptive statistics of BMD variables in subjects with confirmed osteopenia near the 
spinal column (N=6), and hip articulation (N=3) and non-affected subjects (absolute values of 
BMD in g/cm
2
).
Variables Valid N Mean Std. Dev. Std. Error Mean
bmdspine
.00 74 1.3404 .11480 .01334
1.00 6 1.0188 .04010 .01637
bmdhip
.00 77 1.3208 .13861 .01580
1.00 3 .8753 .07836 .04524
Table 1 shows descriptive statistics of BMD variables, upon examination near the spinal column 
and hip articulation. Osteopenia was confirmed in six subjects (one athlete and five non-athletes) 
near the spinal column, and in three subjects (non-athletes) near the hip articulation.
Table 2: Descriptive statistics of BMI variable (in kg/m
2
) of subjects with confirmed osteopenia 
near the spinal column (N=6) and non-affected subjects (N=74) 
Variable N Mean Std.Deviation Std. Error Mean
BMI .00 74 24.0621 2.71096 0.31514
BMI  1 6 21.4739 1.76204 0.71935

32 BMI as a predictor of bone mineral density Kinesiologia Slovenica, 15, 3, 29–39 (2009) 
Table 2 gives descriptive statistics of BMI variables, after examination in region of spinal column 
(L
1
-L
4
), of subjects in whom the important decrease of BMD values was confirmed (N=6), and 
non-affected subjects (N=74).
Table 3: T-test for independent samples and testing of the existence of differences in BMD values 
between subjects in whom the important decrease of BMD values was confirmed (N=6), and 
non-affected subjects (N=74), near the spinal column
Levene’s Test 
for Equality of 
Va r ia nces
t-test for Equality of Means
F Sig. t df
Sig. 
(2-tailed)
Mean 
Difference
Std. Error 
Difference
95% Confidence 
Interval of the 
Difference
Lower Upper
BMD
SPINE
Equal 
variances 
assumed
4.505 .037 15.224 13.445 .000 .32155 .02112 .27607 .36702
To test the existence of differences of BMD variable near the spinal column among subjects in 
whom the important decrease of BMD values was confirmed (N=6) and non-affected subjects 
(N=74), a t-test for independent samples was applied. The results are presented in Table 3. The 
significance of the t-test (p=0.000<0.05), indicates the existence of statistically important differ-
ences of BMD values between these two groups of subjects. 
Table 4: T-test for independent samples and testing of the existence of differences in BMD values 
between subjects in whom the important decrease of BMD values was confirmed (N=3), and 
non-affected subjects (N=77), near the hip articulation
Levene’s Test 
for Equality of 
Va r ia nces
t-test for Equality of Means
F Sig. t df
Sig. 
(2-tailed)
Mean 
Difference
Std. Error 
Difference
95% Confidence 
Interval of the 
Difference
Lower Upper
BMD
HIP
Equal 
variances 
assumed
1.044 .310 5.509 78 .000 .44545 .08085 .28448 .60642
To test the existence of differences of the BMD variable near the hip articulation, among subjects 
in whom the important decrease of BMD values was confirmed (N=3), and non-affected subjects 
(N=77), a t-test for independent samples was applied. The results are presented in Table 4. The 
significance of the t-test (p=0.000<0.05), indicates the existence of statistically important differ-
ences of BMD values between these two groups of subjects. 

BMI as a predictor of bone mineral density 33 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
Table 5: T-test for independent samples and testing of existence of differences in BMI values 
between subjects in whom the important decrease of BMD values was confirmed (N=6), and 
non-affected subjects (N=74), upon examination near the spinal column
Levene’s Test 
for Equality 
of Variances
t-test for Equality of Means
F Sig. t df
Sig. 
(2-tailed)
Mean 
Difference
Std. Error 
Difference
95% Confidence 
Interval of the 
Difference
Lower Upper
BMI
Equal 
variances 
assumed
0.301 0.585 2.292 78 0.025 2.58827 1.12924 0.34013 4.83641
To test the existence of differences of BMI variable near the spinal column, among subjects in 
whom the important decrease of BMD values was confirmed (N=6), and non-affected subjects 
(N=74), a t-test for independent samples was applied. The results are presented in Table 5. The 
significance of the t-test (p=0.025<0.05) indicates the existence of statistically important differ-
ences of BMI values between these two groups of subjects. The existence of differences indicates 
that variable BMI could be indicator of important decrease of bone mineral density values near 
the spinal column. In order to confirm this hypothesis, ROC analysis was performed (Diagram 
1 and Table 6).
1.0 0.8 0.6 0.4 0.2 0.0
1 - Specificity
1.0
0.8
0.6
0.4
0.2
0.0
ROC Curve
Sensitivity
Diagram 1: ROC curve (blue) refers to the area below curve

34 BMI as a predictor of bone mineral density Kinesiologia Slovenica, 15, 3, 29–39 (2009) 
Table 6: ROC analysis of BMI variable, upon BMD examination near the spinal column
Area Std. Error(a) Asymptotic Sig.(b)
Asymptotic 95% Confidence 
Interval
Lower Bound Upper Bound
0.813 0.082 0.011 0.652 0.975
a) Under the nonparametric assumption 
b) Null hypothesis: true area = 0.5
Based on the results of ROC analysis (Diagram 1 and Table 6), a conclusion can be drawn that the 
area below ROC curve (0.813) is significantly larger than 0.5 by reason of test significance, which 
is p=0.011<0.05. The test significance indicates that BMI can be an indicator of the important 
decrease of bone mineral density values near the spinal column.
Table 7: Values of BMI variable of all examined subjects are presented from minimum to maxi-
mum, upon BMD examination near the spinal column
Coordinates of the Curve
Test Result Variable(s): BMI
BMI Sensitivity 1 - Specificity BMI Sensitivity 1 - Specificity
17.115 41 1 1 23.78711 0.527027 0.166667
18.8603 0.986486 1 23.84862 0.513514 0.166667
19.61421 0.972973 1 23.90665 0.486486 0.166667
19.74588 0.959459 1 23.93217 0.486486 0
19.87619 0.959459 0.833333 23.96318 0.472973 0
20.19947 0.959459 0.666667 24.0219 0.459459 0
20.522 0.945946 0.666667 24.07495 0.432432 0
20.62231 0.932432 0.666667 24.1035 0.418919 0
20.73847 0.932432 0.5 24.13865 0.405405 0
20.91144 0.918919 0.5 24.19175 0.391892 0
21.13131 0.918919 0.333333 24.30564 0.378378 0
21.24864 0.905405 0.333333 24.40256 0.364865 0
21.39795 0.891892 0.333333 24.45193 0.351351 0
21.50127 0.878378 0.333333 24.5027 0.337838 0
21.5663 0.864865 0.333333 24.54673 0.324324 0
21.65423 0.851351 0.333333 24.63461 0.310811 0
21.69844 0.837838 0.333333 24.7635 0.297297 0
21.75876 0.824324 0.333333 24.87548 0.283784 0
21.98005 0.810811 0.333333 24.94618 0.27027 0
22.20879 0.783784 0.333333 25.01818 0.256757 0
22.26975 0.77027 0.333333 25.09357 0.243243 0
22.29597 0.756757 0.333333 25.18858 0.22973 0
22.35238 0.743243 0.333333 25.27813 0.216216 0
22.42171 0.72973 0.333333 25.30779 0.202703 0
22.4816 0.716216 0.333333 25.47444 0.189189 0

BMI as a predictor of bone mineral density 35 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
BMI Sensitivity 1 - Specificity BMI Sensitivity 1 - Specificity
22.55151 0.702703 0.333333 25.88135 0.175676 0
22.61761 0.689189 0.333333 26.17851 0.162162 0
22.82 0.675676 0.333333 26.24976 0.148649 0
23.1193 0.662162 0.333333 26.28076 0.135135 0
23.27354 0.648649 0.333333 26.40419 0.121622 0
23.31761 0.635135 0.333333 27.3184 0.108108 0
23.3614 0.621622 0.333333 28.24626 0.094595 0
23.3981 0.621622 0.166667 28.38129 0.081081 0
23.43215 0.608108 0.166667 28.40162 0.067568 0
23.4823 0.594595 0.166667 28.97252 0.054054 0
23.509 0.581081 0.166667 29.72688 0.040541 0
23.52904 0.567568 0.166667 30.18733 0.027027 0
23.60957 0.554054 0.166667 32.86126 0.013514 0
23.71834 0.540541 0.166667 36.26476 0 0
The value of the BMI variable that separates two groups of examined subjects (with osteopenia 
and without significant decrease of bone mineral density values), upon BMD examination near 
the spinal column is 21.13 kg/m² (Table 7).
Table 8: Prediction effectiveness of significant decrease of bone mineral density values by us-
ing variable BMI of both groups of examined subjects, upon BMD examination near the spinal 
column
BMI test
To t a l
0 1
OSTSPINE  .00 68 6 74
1.00 2 4 6
To t a l 70 10 80
With the application of this prediction, the existence of significant decreases of bone mineral 
density values was not confirmed in 68 of the 74 subjects, without significant decrease of bone 
mineral density values near the spinal column. In four of the six subjects with osteopenia, sig-
nificant decreases of bone mineral density values near the spinal column were confirmed, which 
represents 88% of prediction effectiveness, by using BMI variable (Table 8).
Table 9: Descriptive statistics of BMI variable (in kg/m
2
) of subjects with confirmed osteopenia 
and non-affected subjects, upon BMD examination near the hip articulation
Variable N Mean Std. Deviation Std. Error Mean
BMI .00 77 23.8789 2.74862 .31323
BMI  1 3 23.5877 2.71910 1.56988
Table 9 presents descriptive statistics of BMI variables upon examination near the hip articula-
tion, in subjects with confirmed significant decreases of bone mineral density values (N=3) and 
non-affected subjects (N=77).

36 BMI as a predictor of bone mineral density Kinesiologia Slovenica, 15, 3, 29–39 (2009) 
Table 10: T-test for independent samples and testing of the existence of differences in BMI val-
ues between subjects with confirmed osteopenia (N=3) and non-affected subjects (N=77), upon 
examination near the hip articulation
Levene’s Test 
for Equality of 
Va r ia nces
t-test for Equality of Means
F Sig. t df
Sig. 
(2-tailed)
Mean 
Difference
Std. Error 
Difference
95% Confidence 
Interval of the 
Difference
Lower Upper
BMI
Equal 
variances 
assumed
.000 .992 .180 78 .858 .29128 1.61709 -2.92811 3.51066
To test the existence of differences of BMI variable near the hip articulation, between subjects 
with confirmed osteopenia (N=3) and non-affected subjects (N=77), a t-test for independent 
samples was applied, the results of which are presented in Table 10. 
The significance of t-test (p=0.858>0.05), does not indicate the existence of statistically important 
differences of BMI values between these two groups of subjects.
The non-existence of significant differences indicates that variable BMI could not be an indicator 
of important decrease of bone mineral density values near the hip articulation, which represents 
the result of ROC analysis (Diagram 2 and Table 11).
1.0 0.8 0.6 0.4 0.2 0.0
1 - Specificity
1.0
0.8
0.6
0.4
0.2
0.0
ROC Curve
Sensitivity
Diagram 2: ROC curve (blue) refers to the area below curve

BMI as a predictor of bone mineral density 37 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
Table 11: ROC analysis of BMI variable, upon BMD examination near the hip articulation
Area Std. Error(a) Asymptotic Sig.(b)
Asymptotic 95 % 
Confidence Interval
Lower Bound Upper Bound
.494 .183 .970 .136 .851
a) Under the nonparametric assumption 
b) Null hypothesis: true area = 0.5
Based on the results of ROC analysis (Diagram 2 and Table 11), that conclusion can be drawn that 
area below ROC curve (0.494) is not significantly different in relation to 0.5 (p=0.970>0.05).
DISCUSSION
Concerning the values of body mass index (BMI), the confirmed important decrease of BMD, i.e. 
osteopenia in one athlete and five non-athletes near the spinal column, and in three non-athletes 
near the hip articulation, the discussion is given below.
In spite of high levels of professional sport physical activity, the determined important decrease 
of BMD near the spinal column (osteopenia) in one athlete is the most likely a consequence of 
genetics factors i.e., family case-history (mother has osteoporosis),
The determined important decrease of BMD near the spinal column i.e. osteopenia in five non-
athletes and determined important decrease of BMD near the hip articulation i.e. osteopenia in 
three non-athletes is most likely a consequence of several factors, such as lower level of physical 
activity, factors which have exerted influence before the commencement of their recreational 
sports activities and quotidian physical activities i.e. genetic factors, hormonal factors related 
to the phase of growth (puberty), presence of risk factors (smoking, alcohol, chronic diseases, 
administration of certain drugs, etc.), apparently influenced the BMD values.
The existence of significant differences in values of variable BMI (p=0.025), between examined 
subjects in whom it was confirmed important decrease of bone mineral density values (N=6) 
and examined subjects in whom it was not the case (N=74) near the spinal column indicates 
that variable BMI is an indicator of important decrease of bone mineral density (BMD) near 
the spinal column,
The non-existence of significant differences in values of variable BMI (p=0.858), between exam -
ined subjects in whom it was confirmed important decrease of bone mineral density values (N=3) 
and examined subjects in whom it was not the case (N=77) near the hip articulation, points out 
to the fact that variable BMI is not an indicator of important decrease of bone mineral density 
(BMD) near the hip articulation.
In research conducted by Galvan et al. (2009), the authors concluded that BMD is negatively 
associated with the adiponectin levels in men older than 60 years and this relationship is greater 
in those men with BMI >27, which suggests a plausible connection between bone and fat tissue. 
Bundred (2009) concluded that risk factors for premature bone loss and fracture include a low 
BMI, family history or personal history of fragility fracture after the age of 50, oral corticosteroid 
use more than six months and cigarette smoking. In research conducted by Garcia-Carrasco et al. 
(2009), authors indicated that chronic disease damage, low BMI, and cumulative corticosteroid 
dose are risks factors for low BMD in pre-menopausal patients with systemic lupus erythematosus 
(SLE).

38 BMI as a predictor of bone mineral density Kinesiologia Slovenica, 15, 3, 29–39 (2009) 
In research conducted by Nar-Demirer et al. (2009), the authors stated that patients with radial 
and/or lumbar and/or hip osteoporosis had a longer duration of diabetes, were older, and had 
a lower BMI, compared with non-osteoporotic subjects. In research conducted by Christo et 
al. (2008), the authors found that athletes with amenorrhea had lower bone-density z-scores at 
the spine and whole body, compared with athletes with eumenorrhea and control subjects, and 
lower hip z-scores, compared with athletes with eumenorrhea. Also, athletes with amenorrhea 
had lower BMI z-scores than athletes with eumenorrhea and lower insulin-like growth factor 
I-levels than the control subjects. In research conducted by Naka et al. (2005), the authors found 
significant positive correlations between BMD at every skeletal site and height, weight and grip 
strength in pre- and post-pubertal boys and girls. In research conducted by Misra et al. (2004), 
the authors found that all measures of bone mineral density (z-scores) were lower in girls with 
anorexia nervosa (AN) than in control subjects, with lean body mass, body mass index, and age 
at menarche emerging as the most important predictors of bone density. In research conducted 
by Barondess et al. (1997), the authors found that whole body bone mineral content (WBBMC), 
WBBMC/height, and whole body bone mineral density (WBBMD) were strongly and significantly 
correlated with weight, body mass index (BMI), and body composition.
The adjusted research considered a young, male population, with relatively good social status. 
Although body mass index (BMI) is related to bone mineral density (BMD), the potential 
mechanism(s) of this relation have yet to be defined. The results obtained by the research indicate 
the role of oestrogen and other hormonal factors in mediating fat mass effects on the skeleton.
REFERENCES
Barondess, D. A., Nelson, D. A., Schlaen, S. E. (1997). Whole body bone, fat, and lean mass in black and 
white men. Journal of Bone and Mineral Research,12(6): 96 –71.
Bubanj, R. (1997a). Basics of applied biomechanics in sports. Niš: SIA.
Bubanj, R. (1997b). Basics of applied biomechanics in kinesiology. Niš: SIA.
Bubanj, S., Dimić, A., Stanković, R., Obradović, B., Bubanj, R., Bubanj, M. & Bojanić, V. (2007). Influence 
of isometric and ballistic muscular potential on differences in bone mineral density in region of spinal 
column and hip articulation at sportsmen and non-sportsmen. Balneoclimatologia, 31(4), 169–181.
Bubanj, S. & Obradović, B. (2002). Mechanical force and bone density. Facta Universitatis series Physical 
Education and Sport, 1(9): 37–50.
Bundred, N. (2009). Aromatase inhibitors and bone health. Current Opinion in Obstetrics and Gynecology, 
21(1): 60–67.
Christo, K., Prabhakaran, R., Lamparello, B., Cord, J., Miller, K., Goldstein, M., Gupta, N., Herzog, D., 
Klibanski, A. & Misra, M. (2008). Bone Metabolism in Adolescent Athletes with Amenorrhea, Athletes 
with Eumenorrhea, and Control Subjects. Pediatrics, 121(6): 1127–1136.
Dimić, M. (2005). The influence of anticonvulsive drugs on bone metabolism of children affected by epilepsy. 
Unpublished doctoral dissertation, Niš: Faculty of Medicine.
Đurašković, R., (2002). Sports medicine. Niš: SVEN.
Galvan, R., Cordova, N., Saucedo, R., Vargas, C., Campos, S., Halley, E., Avelar, F. & Zarate, A. (2009). 
Adiponectin is associated with low bone mineral density in elderly men. European Journal of Endocrinol-
ogy, 160(2): 289–293.
Garcia-Carrasco, M., Sandoval-Cruz, H., Escarcega., O., Jimenez-Hernandez, M., Etchegaray-Morales, I., 
Soto-Vega, E., Munoz-Guarneros, M., Lopez-Colombo, A., Deleze-Hinojosa, M., Cervera, R. (2009). Risks 

BMI as a predictor of bone mineral density 39 Kinesiologia Slovenica, 15, 3, 29–39 (2009)
factors for low bone mineral density in pre-menopausal Mexican women with systemic lupus erythema-
tosus. Clinical Rheumatology, 28(1): 65–70.
Misra, M., Aggarwal, A., Miller, K. K., Almazan, C., Worley, M., Soyka, L. A., Herzog, D. B. & Klibanski, 
A. (2004). Effects of anorexia nervosa on clinical, hematologic, biochemical, and bone density parameters 
in community-dwelling adolescent girls. Pediatrics, 114(6): 1574–1583.
Nar-Demirer, A., Bascil-Tutuncu, N. & Ertorer, E. (2009). Prevalence of osteoporosis and factors affecting 
bone mineral density among postmenopausal Turkish women with type 2 diabetes. Journal Of Diabetes And 
Its Complications, 23: 12– 7.
Naka, H., Iki, M., Morita, A. & Ikeda, Y. (2005). Effects of pubertal development, height, weight, and grip 
strength on the bone mineral density of the lumbar spine and hip in peripubertal Japanese children: Kyoto 
kids increase density in the skeleton study (Kyoto KIDS study). J. Bone Miner. Metab., 23(6): 463–9.