Archives of Gerontology and Geriatrics 54 (2012) e315–e321

Contents lists available at SciVerse ScienceDirect

Archives of Gerontology and Geriatrics

journa l homepage: www.e lsev ier .com/ locate /archger

The relationship between accelerometer-determined physical activity (PA) andbody composition and bone mineral density (BMD) in postmenopausal women

Ales Gaba a,*, Ondrej Kapus a, Jana Pelclova b, Jarmila Riegerova a

a Department of Natural Sciences in Kinanthropology, Faculty of Physical Culture, Palacky University in Olomouc, Tr. Mıru 115, 771 11, Olomouc, Czech Republicb Center for Kinanthropology Research, Faculty of Physical Culture, Palacky University in Olomouc, Tr. Mıru 115, 771 11, Olomouc, Czech Republic

A R T I C L E I N F O

Article history:

Received 21 October 2011

Received in revised form 1 February 2012

Accepted 2 February 2012

Available online 8 March 2012

Keywords:

Proximal femur area

ActiGraph GT1M

Walking

Lean body mass

Body fat mass

A B S T R A C T

Studies of the relationships between BMD, PA and body composition have shown variable results.

Therefore, the aim of this cross-sectional study was to determine the relationships between

accelerometer-determined PA and selected body composition parameters to total and regional BMD

of the proximal femur in postmenopausal women. BMD and body composition were measured using

dual energy X-ray absorptiometry in 97 women with a mean age 63.63 � 5.23 years. PA was monitored

using an ActiGraph GT1M accelerometer. Correlation analysis did not show significant relationships between

PA variables and BMD, but increases in body composition variables were associated with increases in BMD.

Lean body mass was the strongest predictor of proximal femur BMD (r = 0.18–0.37), explaining 10% of the

variance for total femur, and 3–14% of the variance for regional femurs. Correlations increased when the

analysis was controlled for age (rp = 0.20–0.39). A significant relationship was also found between body fat

mass and BMD (r = 0.16–0.30; rp = 0.25–0.37). Analysis of differences between women with normal BMD and

osteopenic women showed statistically significant differences in age (p = 0.003; h2 = 0.09) and lean body

mass (p = 0.048; h2 = 0.04). In conclusion, body composition is a stronger predictor of proximal femur BMD

than PA variables. However, other studies are necessary to clarify the influence of long-term PA and exercise

type on proximal femur BMD.

� 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

A lack of regular PA and a sedentary lifestyle affect the quality oflife in elderly populations. Unfortunately, current population trendsshow a high prevalence of sedentary lifestyles in the EuropeanUnion, and the number of inactive individuals is rapidly increasing inolder age groups (Varo et al., 2003). PA is recommended as a primarymeans of preventing chronic diseases in adults and older adults.General recommendations for PA have focused on steps counts andintensity of PA. The most commonly cited healthy step goal is 10,000steps per day (Hatano, 1993) or detailed step-based recommenda-tion by Tudor-Locke and Bassett (2004). It has been suggested thatadults should perform at least 150 min of moderate PA, or 75 min ofvigorous PA per week (U.S. Department of Health and HumanServices, 2008). In addition, the American College of Sports Medicineand the American Heart Association (ACSM/AHA) recommendengaging in moderate PA for at least 30 min five times per week orvigorous PA for at least 20 min three times per week (Haskell et al.,2007). Combinations of moderate and vigorous PA can be alsoperformed to meet this recommendation.

* Corresponding author. Tel.: +420 777 945 875.

E-mail address: ales.gaba@upol.cz (A. Gaba).

0167-4943/$ – see front matter � 2012 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.archger.2012.02.001

Strong evidence has shown that regular PA significantly affectshuman health. Increasing the volume and intensity of PA isassociated with reduced risks of breast cancer (McTiernan et al.,2003), type 2 diabetes mellitus (Rana et al., 2007), cardiovasculardisease (Albright and Thompson, 2006) and obesity in postmeno-pausal women (Krumm et al., 2006; Kyle et al., 2004). The benefits ofregular PA are also reflected in bone health because it stimulatesosteoblast activity, decreases the rate of bone loss, and substantiallyreduces the risk of osteoporosis in old age (Warburton et al., 2006).The ACSM recommends regular weight-bearing endurance activitiesin combination with resistance exercise across the age spectrum tomaintain or maximize of bone mass (Kohrt et al., 2004).

PA that generates moderate and high intensity loading forcesmay also change body composition and muscle strength. Severalstudies have demonstrated associations of total body mass (Reid,2002), lean body mass (Beck et al., 2011), body fat mass (Dytfeldet al., 2011), and muscle strength (Winters and Snow, 2000) withtotal and regional BMD. Greater body weight has a positive effecton bone tissue due to its stimulating factor for osteogenic response.Individuals with higher body weight, and a respective higheramount of lean body mass, have a lower risk for developing healthdisorders (Winters and Snow, 2000; Reid, 2002). Lean body massgenerally leads to increased BMD through mechanical loading ofthe skeleton.

Table 1Characteristics of study participants (n = 97).

Mean� SD Range

Age (yrs.) 63.63�5.23 50.00–77.00

YSM (yrs.) 12.98�6.03 1.00–27.00

Height (cm) 160.63�5.92 146.00–179.00

Weight (kg) 69.44�11.20 44.66–98.35

BMD (g/cm2)Total body 1.11� 0.09 0.92–1.27

Total femur 0.95� 0.11 0.70–1.22

Femoral neck 0.88� 0.10 0.66–1.19

Ward’s triangle 0.68� 0.11 0.48–0.95

Greater trochanter 0.79� 0.11 0.56–1.02

T-scoreTotal body �0.15�1.11 �2.60–1.80

Total femur �0.49� 0.88 �2.40–1.70

Femoral neck �1.14� 0.74 �2.70–1.10

Ward’s triangle �1.77� 0.83 �3.30–0.30

Greater trochanter �0.54� 0.94 �2.50–1.50

PALight PA (min/week) 531�270 105–1414

Moderate PA (min/week) 225�152 0–621

Vigorous PA (min/week) 9�24 0–195

Steps per day 9766�3644 2945–19.960

Physical inactivity (hours/day) 5.00�1.52 1.62–10.62

Body compositionLean body mass (kg) 40.65�4.42 30.30–51.87

Body fat mass (kg) 26.46�8.16 10.50–47.95

Body fat mass (%) 37.34�6.42 18.90–53.00

BMI (kg/m2) 26.92�4.22 18.45–39.84

FFMI (kg/m2) 16.65�1.52 12.74–21.26

BFMI (kg/m2) 10.27�3.20 3.98–19.54

Note: YSM – years since menopause.

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321e316

On the contrary, a decrease in muscle mass, a component of leanbody mass, and a related decline in muscle strength negativelyaffect physical function and general stability. This may lead to asignificant increase of falls, which is a common cause of seriousfractures among elderly adults. Women who suffer from osteopo-rosis have high risk of femoral fractures (neck and trochantericregions), which considerably affects mortality in old age (Centeret al., 1999). In addition, excluding the above-mentioned factors(i.e., PA, body composition and muscle strength) the risk ofosteoporosis increases with excessive use of alcohol (Maurel et al.,2011), tobacco use, eating disorders with excessive protein andcalcium deficiency and corticosteroid medications. Low BMD isalso influenced by heredity (Bartl and Frisch, 2009).

Maintaining healthy bones and avoiding osteoporotic fracturescannot be achieved by systematic monitoring of bone tissue alone,but must be evaluated in relation to objectively measured PA andbody composition. Unfortunately, the results have varied betweenstudies, and it remains unclear how these factors affect thevariability of bone tissue. Therefore, the main purpose of this cross-sectional study was to analyze the relationships betweenaccelerometer-determined PA and selected body compositionparameters to total and regional BMD of the proximal femur inpostmenopausal women.

2. Materials and methods

2.1. Subjects and design

This study was cross-sectional, descriptive, and non-randomizedanalysis of quantitative data. Evaluation of BMD, PA, and bodycomposition was performed in ninety-seven apparently healthypostmenopausal women, ranging in age from 50 to 77 years, whoattended education and PA programs at the University of the ThirdAge at Palacky University in Olomouc. All participants have not beenmenstruating for at least one year before examination. Women whowere currently taking or had taken hormone replacement therapy orany prescription medication for BMD improvements in the last twoyears were excluded from this research. Women who hadundergone major hip or knee surgery or bone densitometry in theprevious twelve months or who had physical handicap that mightinterfere with BMD, body composition and PA measurement (e.g.motor skills disorder, amputation, and paralysis) were alsoexcluded. The purpose and risks of the study were explained toeach participant, and written informed consent was obtained. Thestudy plan and design was approved by the Faculty of PhysicalCulture Ethics Committee at the Palacky University in Olomouc.Research was completed in April 2010 and 2011 at the bonedensitometry center in Olomouc (Czech Republic). Completeparticipant characteristics are shown in Table 1.

2.2. BMD and body composition assessment

We assessed BMD (g/cm2) of the total body and proximal femurusing a dual energy X-ray absorptiometer Lunar Prodigy PrimoTM

with enCORETM software version 12.20.023 (GE Healthcare, UK).Regions of interest at the proximal femur included the femoralneck, Ward’s triangle, and the greater trochanter. The DXA devicewas regularly calibrated before each measurement session using aLunar phantom with a precision error of �1%. Lean body mass (kg),body fat mass (kg), and percentage of body fat were measured via aDXA device that showed high validity and reliability in evaluatingbody composition (Heymsfield et al., 2005).

The overall procedure (whole-body and femoral scan) tookapproximately 30 min and was performed with the participantlying in the supine position with their arms at their sides. All metalitems were removed from the participants to ensure the accuracy

of the measurement. The prevalence of osteopenia and osteoporo-sis was evaluated according to T-score (BMD normalized by theyoung adult reference BMD value) for the total femur region.According to the World Health Organization (World HealthOrganization, 2003), a normal BMD is defined as a T-score greaterthan �1.0, while a T-score between �1.0 and �2.5 indicates lowbone mass (osteopenia), and a T-score lower than �2.5 indicatesosteoporosis.

2.3. Anthropometric indices

Anthropometric data were collected by standard methods.Standing height was recorded to the nearest 0.5 cm using ananthropometer P-375 (Trystom, Czech Republic) before the DXAprocedure with subjects in light clothes without shoes. Body weightwas measured by the DXA device, as total body mass, to the nearest0.1 kg. Body mass index (BMI; kg/m2) was used as an indicator ofobesity. BMI is usually calculated as body weight (kg) divided byheight squared (m2). However, the significance of the BMI is notclear, because body mass is composed of two distinct components(fat-free mass and body fat mass). Therefore, we also calculated thefat-free mass index (FFMI; kg/m2) and the body fat mass index(BFMI; kg/m2). FFMI and BFMI were derived as fat-free mass (kg) andbody fat mass (kg), respectively, divided by height squared (m2). Forwomen within the normal BMI (18.5–24.9 kg/m2), derived BFMIvalues from 3.9 to 8.2 kg/m2 and FFMI values ranging from 14.6 to16.8 kg/m2 indicated normal body weight. The cut-off points forFFMI and BFMI corresponding to BMI values in healthy adults wereestablished by Kyle et al. (2004). In addition, we also used thepercentage of body fat as an indicator of obesity according topercentage body fat ranges reported by Gallagher et al. (2000).

2.4. PA assessment

The participants’ PA level was monitored for seven consecutivedays using an ActiGraph model GT1M (3.8 cm � 3.7 cm � 1.8 cm;

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321 e317

27 g) accelerometer (ActiGraph; LLC, Pensacola, FL, USA) thatregistered vertical acceleration in units of counts. Each acceler-ometer was calibrated according to the manufacturer’s recom-mendations before testing. The time sampling interval was set at15 s using the manufacturer’s software (ActiLife version 5.8) andstep mode was activated. Each accelerometer was attached to anelastic belt with a small pocket, was securely positioned near theright iliac crest, and was worn by participants beginningimmediately after the densitometry scan. Participants wereinstructed to wear the accelerometer while awake, and to removeit for water activities. Women who accumulated fewer than 10 h ofvalid PA recordings per day were excluded from the research.

Intensity levels (METs) were analyzed according the recommen-dation established by Freedson et al. (1998). Light PA (<3 METs) wasdefined as less than 1952 counts per minute, moderate PA wasdefined as between 1952 and 5724 counts per minute (3–6 METs),and vigorous PA was defined as greater than 5724 counts per minute(>6 METs). Women were considered sedentary if they accumulatedfewer than 150 min of moderate PA per week (U.S. Department ofHealth and Human Services, 2008) or 5000 steps per day (Tudor-Locke and Bassett, 2004).

2.5. Statistical analysis

Descriptive data are presented as mean and standard deviation(M � SD). Data were verified for normality of distribution (Shapiro–Wilk test). Pearson product moment correlations (r) were calculatedto determine the relationship between the dependent variables (BMDof proximal femur) and the independent variables (accelerometer-determined PA and body composition). Partial correlation coefficients(rp) were calculated between the above-mentioned dependent andindependent variables while controlling for age.

For additional analysis, women were divided into two groupsbased on the T-score of their proximal femur: normal BMD (n = 71)and osteopenic (n = 26). No women in our study sample had a T-score lower than �2.5. A one-way analysis of variance (ANOVA)and eta-squared (h2) were used to indicate the differences betweengroup of women with normal BMD and osteopenic women. Eta-squared was calculated via ANOVA [h2 = SSfactor/(SSfactor + SSerror)].The values of 0.01, 0.06, and 0.14 were interpreted as small,medium, and large effect size (Morse, 1999), respectively.[(Fig._1)TD$FIG]

step10,008000600040002000

100

200

300

400

500

600

700osteopenianormal

sedentary

Fig. 1. Adherence to the P

Statistical significance was set at p < 0.05. STATISTICA version9.0 software was used to complete all analyses.

3. Results

The BMD, PA, and body composition characteristics of thestudy participants are listed in Table 1. The mean age of thewomen was 63.63 � 5.23 years, the mean age of menopause was50.65 � 3.01 years, and the mean length of time since menopausewas 12.98 � 6.03 years. The mean total body BMD was 0.16 g/cm2

greater than BMD of proximal femur area. Ward’s triangle had thelowest BMD (0.68 � 0.11 g/cm2), whereas the femoral neck showedthe highest BMD (0.88 � 0.10 g/cm2) in the femoral head area. Theaverage T-score of the proximal femur (�0.49 � 0.88) was lowerthan the T-score of the whole body (�0.15 � 1.11). Using the T-scoreof proximal femur, we found that 73% of the participants had normalBMD and 27% were osteopenic. No instances of osteoporosis werefound. Thirty percent of the subjects reported being current smokers,and 5% reported having smoked previously.

Women preferred light PA (531 � 270 min/week), and vigorousPA was achieved for 9 � 24 min per week (Table 1). Although thewomen did not meet the suggested amount of vigorous PA, the meanamount of moderate PA was above the recommended threshold. Only38% of women performed less than 150 min of moderate PA per week.The average number of steps accumulated per day was 9766 � 3644,with 45% of participants accumulating more than 10,000 steps perday and 10% of participants accumulating less than 5000 steps per day(sedentary).

The analysis of adherence to the PA recommendations is shownin Fig. 1. Thirty-four percent of participants (21 with normal BMD,12 with osteopenia) accumulated fewer than 10,000 steps per dayand fewer than 150 min of moderate PA per week. In contrast, weregistered eighteen highly active women (13 with normal BMD, 5with osteopenia) who accumulated more than 12,500 steps per dayand 300 min of moderate PA per week.

The average BMI and percentage of body fat indicated theoverweight (BMI = 26.92 � 4.22 kg/m2; %BFM = 37.34 � 6.42%). Theprevalence of obesity related to body fat mass was 28%, and theprevalence of being overweight was 38% (Table 1). The analysis ofFFMI and BFMI is shown in Fig. 2. While the number of participantswith low FFMI or BFMI was low, the number of participants with high

s per day20,00018,00016,00014,00012,0000

A recommendations.

[(Fig._2)TD$FIG]

fat-

free

mas

s in

dex

(kg/

m2 )

13

14

15

16

17

18

19

20

21 normal

very

hig

hhi

ghno

rmal

low

osteopenia

very highhighnormal

20181614121086412

22

body fat mass index (kg/m2)

Fig. 2. Analysis of FFMI and BFMI.

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321e318

or very high FFMI and BFMI was high. We found very high FFMI andBFMI in eleven women (9 with normal BMD, 2 with osteopenia). Thelargest group (21 with normal BMD, 7 with osteopenia) was womenwith normal FFMI and high BFMI.

The analysis of the relationships between accelerometer-determined PA and body composition to proximal femur BMD ispresented in Tables 2 and 3. Pearson product moment correlationsrevealed no significant relationship between PA variables andproximal femur regions BMD (r values ranging from�0.11 to 0.09).Because changes in bone mass are related to age, we calculatedpartial correlation coefficients controlling for age. When partial

Table 2Correlations relationships between the BMD of the proximal femur and PA variables.

BMD total femur BMD femoral n

Correlation Partial

correlationyCorrelation

Light PA (min/week) �0.08 �0.13 �0.08

Moderate PA (min/week) 0.08 0.02 0.02

Vigorous PA (min/week) 0.03 �0.01 0.02

Steps per day �0.01 �0.09 �0.08

Physical inactivity (hours/day) 0.02 0.03 0.06

* p<0.05.y Controlling for age.

Table 3Correlations between the BMD of the proximal femur and body composition variables

BMD total femur BMD femoral neck

Correlation Partial

correlationyCorrelation Part

corr

Weight (kg) 0.36* 0.42* 0.34* 0.43

Lean body mass (kg) 0.32* 0.34* 0.32* 0.36

Body fat mass (kg) 0.30* 0.37* 0.26* 0.36

Body fat mass (%) 0.24* 0.31* 0.20* 0.30

BMI (kg/m2) 0.31* 0.39* 0.21* 0.32

FFMI (kg/m2) 0.28* 0.33* 0.17 0.24

BFMI (kg/m2) 0.27* 0.35* 0.20 0.31

* p<0.05.y Controlling for age.

correlations were used to account for the effect of age on therelationship between BMD of proximal femur regions and PA, onlythe relationship between steps per day and BMD of femoral neckwas significant (rp = �0.22).

In contrast, correlations between body composition variablesand total BMD of the proximal femur indicated that increases in allbody composition variables were associated with increases inproximal femur BMD. Lean body mass was the strongest predictorof BMD of proximal femur (r values ranging from 0.18 to 0.37),explaining 10% of the variance at total femur, and 3–14% of thevariance at the regional femur. Correlations increased when the

eck BMD Ward’s triangle BMD greater trochanter

Partial

correlationyCorrelation Partial

correlationyCorrelation Partial

correlationy

�0.16 �0.03 �0.09 �0.11 �0.14

�0.09 0.00 �0.09 0.12 0.09

�0.04 �0.01 �0.06 0.09 0.07

�0.22* �0.07 �0.19 0.00 �0.04

0.08 �0.04 �0.03 �0.01 �0.01

.

BMD Ward’s triangle BMD greater trochanter

ial

elationyCorrelation Partial

correlationyCorrelation Partial

correlationy

* 0.21* 0.28* 0.36* 0.39*

* 0.18 0.20* 0.37* 0.39*

* 0.16 0.25* 0.26* 0.31*

* 0.14 0.23* 0.19 0.23*

* 0.13 0.22* 0.27* 0.32*

* 0.10 0.15 0.29* 0.32*

* 0.12 0.22* 0.22* 0.27*

Table 4PA characteristics and body composition among groups according to total body BMD categories.

Normal BMD (n = 71) Osteopenia (n = 26) F h2

Mean� SD Mean� SD

Age (yrs.) 62.69�4.82 66.19�5.56 9.26* 0.09yy

YSM (yrs.) 12.26�5.75 14.92�6.47 3.79 0.04y

Height (cm) 161.32�6.00 158.74�5.33 3.73 0.04y

Weight (kg) 70.78�9.61 65.78�14.26 3.91 0.04y

PALight (min/week) 542�275 502�259 0.41 0.00

Moderate (min/week) 240�151 187�150 2.29 0.02y

Vigorous (min/week) 9�26 7�19 0.20 0.00

Steps per day 9925�3426 9331�4226 0.50 0.01y

Physical inactivity (hours/day) 5.05�1.47 4.90�1.71 0.20 0.00

Body compositionLean body mass (kg) 41.19�3.95 39.19�5.32 4.02* 0.04y

Body fat mass (kg) 27.16�7.94 24.52�9.88 2.01 0.02y

Body fat mass (%) 37.82�6.03 36.04�7.36 1.46 0.02y

BMI (kg/m2) 27.26�3.91 25.99�4.92 1.73 0.02y

FFMI (kg/m2) 16.77�1.48 16.33�1.64 1.64 0.02y

BFMI (kg/m2) 10.49�3.00 9.66�3.68 1.26 0.01y

*p<0.05; yh2>0.01, yyh2>0.06.

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321 e319

analysis was controlling for age (rp values ranging from 0.20 to0.39). Additionally, body fat mass explained 9% of the variance attotal femur, and 3–7% of the variance at regional femur. Nosignificant relationship was found between Ward’s triangle BMDand body composition when analysis was not controlled for age.

The comparisons of PA and body composition between womenwith normal proximal femur BMD and osteopenic women areshown in Table 4. Statistically significant differences were found inage (p = 0.003; h2 = 0.09) and lean body mass (p = 0.048; h2 = 0.04).Furthermore, the small effect size was observed for all variablesexcept light and vigorous PA and physical inactivity. Both groupsperformed more than 150 min of moderate PA per week andaccumulated less than 10,000 steps per day. The average BMI andpercentage body fat indicated the overweight in both groups.

4. Discussion

In this study, we evaluated the relationship of objectivelymeasured PA and body composition with total and regional BMD ofthe proximal femur area. BMD is commonly used for diagnosingosteoporosis. Our results confirm recent findings (Wu et al., 2011)showing that Ward’s triangle has the lowest BMD and T-score inthe femoral neck. In our study, the mean T-score of total andregional femur indicated healthy bones, with the exception ofWard’s triangle (T-score lower than�1.0). In addition, although theprevalence of osteoporosis related to the femoral neck in 50 to 84-year-old postmenopausal women is approximately 21% (Kaniset al., 2008), we did not find any women with osteoporosis.

Because low BMD contributes to the incidence of fractures inpostmenopausal women, strategies that positively stimulate bonetissue may reduce mortality related to osteoporotic fractures.Although BMD can be increased through pharmacologic therapy,PA is the only intervention that increases bone mass. For example,Douchi et al. (2000) found that exercising women have higher BMDin the dominant arm and lumbar spine than sedentary women. Across-sectional cohort study including data from 6032 postmeno-pausal women between the ages of 50 and 79 years (Beck et al.,2011) demonstrated that PA positively affects bone health, withbones becoming stronger and more resistant to fracture when theirgeometric response to load is improved. Nevertheless, we observedno significant relationship between PA variables and proximalfemur regions BMD, except for the relationship between steps perday and femoral neck BMD. Furthermore, we did not find

significant differences in PA variables between women withnormal BMD and osteopenic women.

In this study, we measured only the intensity and volume of PA.However, Hamilton et al. (2010) have shown that bone tissuequality is primarily dependent on exercise type. To benefit bonetissue, PA should include weight-bearing exercises, special trainingto strengthen the muscles and exercise with high impact loading(Kohrt et al., 2004). Previous research has highlighted the positiveeffect of whole-body vibration training in reducing the risk forosteoporosis by increasing lumbar BMD (Stengel et al., 2011). Incontrast, lower impact exercises that reduce the effect of gravity,such as swimming, do not significantly improve BMD (Bartl andFrisch, 2009).

As our data showed, postmenopausal women preferred light ormoderate PA. However, we did not observe a significant relation-ship between these PA variables and proximal femur regions BMD.Although our study did not show a significant relationshipbetween moderate PA and bone mass, Feskanich et al. (2002)have demonstrated that moderate PA by postmenopausal womensubstantially lowers the risk of hip fracture. Specifically, the risk ofhip fractures decreases by 6% for every increase of 3 MET-hours perweek of PA, which is roughly equivalent to 1 h of walking per week.In contrast, Chubak et al. (2006) have reported a randomizedcontrolled study showing that yearlong moderate-intensityaerobic exercise intervention does not affect total body BMD inoverweight or obese postmenopausal women.

Walking, an alternative of moderate PA, is the most common typeof activity among elderly adults. It has been extensively promoted asa means of preventing osteoporosis. However, studies have shownvariable effects of walking on bone health. Martyn-St James andCarroll (2008) have shown that although regular walking had nosignificant effect on spinal BMD preservation in postmenopausalwomen, a significant positive effect was evident at the femoral neck.Although these findings do not confirm that regular walkingpreserves bone mass, several studies have demonstrated thepositive consequence of walking on proximal femur BMD. Forexample, Boyer et al. (2011) have reported that a minimum of 4892steps (walking at 1.00 m/s) per day is required for postmenopausalwomen with an average body weight (65.10 kg) to maintain aproximal femur T-score of�1.0. Substantially more steps per day arerequired for lighter individuals (18,568 steps per day) and less stepsper day are required for heavier individuals (1638 steps per day).These results correspond with our study, in which participants(weight = 69.44 kg) who accumulated 9766 steps per day had a

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321e320

mean proximal femur T-score of�0.48. As stated above, the numberof steps accumulated per day was the only PA variable thatsignificantly correlated with proximal femur regional BMD.

There has been discussion regarding the influence of lean bodymass and body fat mass on bone tissue. Several studies have shownthat lean body mass is a better predictor of BMD than body fat massin young (Miller et al., 2004), premenopausal (Winters and Snow,2000) and postmenopausal women (Gjesdal et al., 2008).Postmenopausal women with higher lean body mass havesignificantly stronger femur geometry than their counterpartswith less lean body mass (Beck et al., 2011). Winters and Snow(2000) have shown that lean body mass explained 24% of thevariance for total femur BMD and 17% of the variance for femoralneck BMD in premenopausal women. Although our findingsconfirm these results, we found that lean body mass explained only10% of the variance for total femur and 3–14% for the variance atregional femur. Based on additional analysis, we also observed thatlean body mass was statistically higher in women with normalBMD. In contrast, Dytfeld et al. (2011) reported a strongerassociation between body fat mass and femoral neck BMD thanbetween lean body mass and femoral neck BMD in postmeno-pausal women with osteoporosis. Our study sample includedparticipants who had a relatively high absolute and relativeamount of body fat mass. Thus, this may explain why our studysample did not find women with osteoporosis. Although the higherproportion of body fat mass is associated with increasing of totaland regional BMD, adiposity is also associated with increased riskof many chronic diseases such as cardiovascular disease (Lavieet al., 2009) or type 2 diabetes mellitus (Rana et al., 2007).

This study has several limitations. The small size of studysample could limit the statistical power. Study subjects were notchosen randomly and do not represent the general population ofpostmenopausal women, as recent research (Gaba et al., 2009) hasshown that female students of the University of the Third Age areoften active. Furthermore, the study sample did not includesubjects with T-scores lower than �2.5. Therefore, follow-upresearch should focus on an elderly population with insufficient PAand persons with osteoporosis. The objectivity of the presentedresults may also have been affected by the study design. Becausethis study was cross-sectional and BMD is dependent on the age, itmay not provide the same results as a longitudinal or semi-longitudinal study. However, other researchers have observedsimilar results, which support our findings. Because of the intensityand volume of PA observed in this study, follow-up research shouldmonitor the relationship between the type of PA and bone health.Our results also show that lean body mass strongly predicts totaland regional proximal femur BMD. Therefore, future researchcould investigate the muscle strength of lower extremity viaisokinetic dynamometry that depends on lean body mass.

In conclusion, no significant relationship was found betweenaccelerometer-determined PA and proximal femur BMD. Althoughthe correlations increased when the analysis controlled for age,only the relationship between steps per day and femoral neck BMDwas significant. Our results support the hypothesis that proximalfemur BMD is more dependent on body composition than PA. Leanbody mass was a stronger predictor of proximal femur BMD thanbody fat mass. Ward’s triangle was the only area of the proximalfemur that showed no significant relationship with PA, even withbody composition when analysis did not control for age. Analysis ofdifferences between women with normal BMD and osteopenicwomen showed statistically significant differences in age and leanbody mass.

Conflicts of interest

None.

Acknowledgments

The study was supported by a research grant from the Ministryof Education, Youth and Sports of the Czech Republic (No. MSM6198959221) ‘‘Physical activity and inactivity of the inhabitants ofthe Czech Republic in the context of behavioral changes’’ and aresearch grant from Palacky University in Olomouc (ID:FTK_2011_014) ‘‘Evaluation of bone tissue of the proximal femurin women with different levels of physical activity’’.

References

Albright, C., Thompson, D., 2006. The effectiveness of walking in preventingcardiovascular disease in women: a review of the current literature. J. Women’sHealth 15, 271–280.

Bartl, R., Frisch, B., 2009. Osteoporosis: Diagnosis, Prevention, Therapy, second ed.Springer, Berlin.

Beck, T.J., Kohlmeier, L.A., Petit, M.A., Wu, G.L., Leboff, M.S., Cauley, J.A., Nicholas, S.,Chen, Z., 2011. Confounders in the association between exercise and femur bonein postmenopausal women. Med. Sci. Sports Exerc. 43, 80–89.

Boyer, K., Kiratli, B., Andriacchi, T., Beaupre, G., 2011. Maintaining femoral bonedensity in adults: how many steps per day are enough? Osteoporos. Int. 1–8.

Center, J.R., Nguyen, T.V., Schneider, D., Sambrook, P.N., Eisman, J.A., 1999. Mortalityafter all major types of osteoporotic fracture in men and women: an observa-tional study. Lancet 353, 878–882.

Chubak, J., Ulrich, C.M., Tworoger, S.S., Sorensen, B., Yasui, Y., Irwin, M.L., Stanczyk, F.Z.,Potter, J.D., McTiernan, A., 2006. Effect of exercise on bone mineral density andlean mass in postmenopausal women. Med. Sci. Sports Exerc. 38, 1236–1244.

Douchi, T., Yamamoto, S., Oki, T., Maruta, K., Kuwahata, R., Yamasaki, H., Nagata, Y.,2000. The effects of physical exercise on body fat distribution and bone mineraldensity in postmenopausal women. Maturitas 35, 25–30.

Dytfeld, J., Ignaszak-Szczepaniak, M., Gowin, E., Michalak, M., Horst-Sikorska, W.,2011. Influence of lean and fat mass on bone mineral density (BMD) in postmen-opausal women with osteoporosis. Arch. Gerontol. Geriatr. 53, 237–242.

Feskanich, D., Willett, W., Colditz, G., 2002. Walking and leisure-time activity andrisk of hip fracture in postmenopausal women. J. Am. Med. Assoc. 288,2300–2306.

Freedson, P.S., Melanson, E., Sirard, J., 1998. Calibration of the Computer Science andApplications, Inc. accelerometer. Med. Sci. Sports Exerc. 30, 777–781.

Gaba, A., Pelclova, J., Pridalova, M., Riegerova, J., Dostalova, I., Engelova, L., 2009. Theevaluation of body composition in relation to physical activity in 56–73 y. oldwomen. A pilot study. Acta Univ. Palacki. Olomuc., Gymn. 39, 21–30.

Gallagher, D., Heymsfield, S.B., Heo, M., Jebb, S.A., Murgatroyd, P.R., Sakamoto, Y.,2000. Healthy percentage body fat ranges: an approach for developing guide-lines based on body mass index. Am. J. Clin. Nutr. 72, 694–701.

Gjesdal, C.G., Halse, J.I., Eide, G.E., Brun, J.G., Tell, G.S., 2008. Impact of lean mass andfat mass on bone mineral density. The Hordaland Health Study. Maturitas 59,191–200.

Hamilton, C.J., Swan, V.J., Jamal, S.A., 2010. The effects of exercise and physicalactivity participation on bone mass and geometry in postmenopausal women: asystematic review of pQCT studies. Osteoporos. Int. 21, 11–23.

Haskell, W.L., Lee, I.M., Pate, R.R., Powell, K.E., Blair, S.N., Franklin, B.A., Macera, C.A.,Heath, G.W., Thompson, P.D., Bauman, A., 2007. Physical activity and publichealth: updated recommendation for adults from the American College ofSports Medicine and the American Heart Association. Circulation 116,1081–1093.

Hatano, Y., 1993. Use of the pedometer for promoting daily walking exercise. Int.Coun. Health Phys. Educ. Recreat. 29, 4–8.

Heymsfield, S., Lohman, T., Wang, Z., Going, S., 2005. Human Body Composition.Human Kinetics, Champaign, IL.

Kanis, J.A., Burlet, N., Cooper, C., Delmas, P.D., Reginster, J.Y., Borgstrom, F., Rizzoli,R., 2008. European guidance for the diagnosis and management of osteoporosisin postmenopausal women. Osteoporos. Int. 19, 399–428.

Kohrt, W.M., Bloomfield, S.A., Little, K.D., Nelson, M.E., Yingling, V.R., 2004. Ameri-can College of Sports Medicine Position Stand: physical activity and bonehealth. Med. Sci. Sports Exerc. 36, 1985–1996.

Krumm, E.M., Dessieux, O.L., Andrews, P., Thompson, D.L., 2006. The relationshipbetween daily steps and body composition in postmenopausal women. J.Women’s Health 15, 202–210.

Kyle, U.G., Morabia, A., Schutz, Y., Pichard, C., 2004. Sedentarism affects body fatmass index and fat-free mass index in adults aged 18 to 98 years. Nutrition 20,255–260.

Lavie, C.J., Milani, R.V., Ventura, H.O., 2009. Obesity and cardiovascular disease: riskfactor, paradox, and impact of weight loss. J. Am. Coll. Cardiol. 53, 1925–1932.

Martyn-St James, M., Carroll, S., 2008. Meta-analysis of walking for preservation ofbone mineral density in postmenopausal women. Bone 43, 521–531.

Maurel, D.B., Boisseau, N., Benhamou, C.L., Jaffre, C., 2011. Alcohol and bone: reviewof dose effects and mechanisms. Osteoporos. Int. [ahead of print].

McTiernan, A., Kooperberg, C., White, E., Wilcox, S., Coates, R., Adams-Campbell, L.L.,Woods, N., Ockene, J., 2003. Recreational physical activity and the risk of breastcancer in postmenopausal women: the Women’s Health Initiative Cohort Study.J. Am. Med. Assoc. 290, 1331–1336.

A. Gaba et al. / Archives of Gerontology and Geriatrics 54 (2012) e315–e321 e321

Miller, L.E., Nickols-Richardson, S.M., Wootten, D.F., Ramp, W.K., Herbert, W.G.,2004. Relationships among bone mineral density, body composition, and iso-kinetic strength in young women. Calcif. Tissue Int. 74, 229–235.

Morse, D.T., 1999. MINSIZE2: a computer program for determining effect size andminimum sample size for statistical significance for univariate, multivariate,and nonparametric tests. Educ. Psychol. Meas. 59, 518–531.

Rana, J., Li, T., Manson, J., Hu, F., 2007. Adiposity compared with physical inactivityand risk of type 2 diabetes in women. Diabetes Care 30, 53.

Reid, I.R., 2002. Relationships among body mass, its components, and bone. Bone 31,547–555.

Stengel, S.V., Kemmler, W., Bebenek, M., Engelke, K., Kalender, W.A., 2011. Effects ofwhole body vibration training on different devices on bone mineral density.Med. Sci. Sports Exerc. 43, 1071–1079.

Tudor-Locke, C., Bassett, D., 2004. How many steps/day are enough?: Preliminarypedometer indices for public health. Sports Med. 34, 1–8.

U.S. Department of Health and Human Services, 2008. 2008 Physical ActivityGuidelines for Americans be Active, Healthy, and Happy! U.S. Dept. of Healthand Human Services, Washington, D.C..

Varo, J.J., Martinez-Gonzalez, M.A., De Irala-Estevez, J., Kearney, J., Gibney, M.,Martinez, J.A., 2003. Distribution and determinants of sedentary lifestyles inthe European Union. Int. J. Epidemiol. 32, 138–146.

Warburton, D.E., Nicol, C.W., Bredin, S.S., 2006. Health benefits of physical activity:the evidence. Can. Med. Assoc. J. 174, 801–809.

Winters, K.M., Snow, C.M., 2000. Body composition predicts bone mineral densityand balance in premenopausal women. J. Women’s Health Gender-Based Med.9, 865–872.

World Health Organization, 2003. Prevention and Management of Osteoporosis:Report of a WHO Scientific Group. World Health Organization, Geneva.

Wu, Q., Lefante, J.J., Rice, J.C., Magnus, J.H., 2011. Age, race, weight, and genderimpact normative values of bone mineral density. Gender Med. 8, 189–201.