BONE MINERAL DENSITY AND ACCELEROMETER-DETERMINED HABITUAL PHYSICAL ACTIVITY AND INACTIVITY IN POSTMENOPAUSAL WOMEN

It was well evidenced that the postmenopausal transition linked with estrogen changes is accompanied by weight gain and often leads to a decline in physical activity (Dubnov & Berry, 2005). However, particularly in postmenopausal women, physical activity appears to be important in preventing loss of bone mineral density (BMD) and osteoporosis (Warburton, Nicol, & Bredin, 2006) and thus contributes to the reduction of fracture risk (Nguyen, Sambrook, & Eisman, 1998). The relationship between physical activity and mineral bone density has been variously estimated in epidemiological studies (Nguyen, Sambrook, & Eisman, 1998), cross sectional studies (Coupland et al., 1999; Krumm et al., 2006), and intervention studies (Stengel et al., 2011; Yamazaki et al., 2004). Although the results of the studies were not always consistent, probably due to study design, weight-bearing endurance activities, mainly high intensity, as well as resistance exercise, seem to be eff ective in maintaining bone mass and hence, in reducing the relative risk for fracture (Kohrt et al., 2004). Postmenopausal women, as primarily aff ected by osteo penia/osteoporosis, should be addressed in order to convince them to participate in physical activity, because it helps to stimulate osteoblast activity, decrease the rate of bone loss, and thus substantially reduce the risk of osteoporosis in old age (Warburton et al., 2006). But yet, at the same time, the performed physical activity should be safe and age adjusted. As age increases, the portion of physical inactivity increases as well, and the habitual activities of daily living such as walking, housekeeping, stair climbing, and gardening create a big portion of total daily physical activity. This cross sectional study aimed to investigate the dif ferences between women with normal BMD and osteo penic women. In habitual physical activity, physical inactivity and meeting physical activity recommendations.


INTRODUCTION
It was well evidenced that the postmenopausal transition linked with estrogen changes is accompanied by weight gain and often leads to a decline in physical activity (Dubnov & Berry, 2005).However, particularly in postmenopausal women, physical activity appears to be important in preventing loss of bone mineral density (BMD) and osteoporosis (Warburton, Nicol, & Bredin, 2006) and thus contributes to the reduction of fracture risk (Nguyen, Sambrook, & Eisman, 1998).
The relationship between physical activity and mineral bone density has been variously estimated in epidemiological studies (Nguyen, Sambrook, & Eisman, 1998), cross sectional studies (Coupland et al., 1999;Krumm et al., 2006), and intervention studies (Stengel et al., 2011;Yamazaki et al., 2004).Although the results of the studies were not always consistent, probably due to study design, weight-bearing endurance activities, mainly high intensity, as well as resistance exercise, seem to be eff ective in maintaining bone mass and hence, in reducing the relative risk for fracture (Kohrt et al., 2004).
Postmenopausal women, as primarily aff ected by osteo penia/osteoporosis, should be addressed in order to convince them to participate in physical activity, because it helps to stimulate osteoblast activity, decrease the rate of bone loss, and thus substantially reduce the risk of osteoporosis in old age (Warburton et al., 2006).But yet, at the same time, the performed physical activity should be safe and age adjusted.As age increases, the portion of physical inactivity increases as well, and the habitual activities of daily living such as walking, housekeeping, stair climbing, and gardening create a big portion of total daily physical activity.
This cross sectional study aimed to investigate the dif ferences between women with normal BMD and osteo penic women.In habitual physical activity, physical inactivity and meeting physical activity recommendations.
par ticipate and gave their verbal and written consent.Data from fi ve participants had to be excluded due to incomplete physical activity records.Hence, the data of ninety-two women was analyzed in this study.The study was carried out in spring 2010 and 2011 according to the design and methods approved by the Faculty of Physical Culture Ethics Committee at Palacký University, Olomouc.Characteristics of the study subjects involving age, body composition variables, total body and femoral BMD and T-score, and smoking and sport issues are shown in TABLE 1.

MEASUREMENTS Bone mineral density and body composition
BMD and body composition were measured using a Dual X-ray Absorptiometry (DXA) Lunar Prodigy Primo TM (GE Healthcare, Great Britain).The region of interest was the proximal femur as it is one of the best predictors of hip and other fractures and is technically unaff ected by degenerative spinal disease.The DXA device was regularly calibrated before each measurement ses sion using a Lunar phantom.The DXA method shows high validity and reliability in the diagnosis of body composition (Heymsfi eld et al., 2005).Therefore, we also determined fat free mass (kg) and body fat mass (kg) by means of a total body DXA scan.The fat free mass index (FFMI; kg/m 2 ) and body fat mass index (BFMI; kg/m 2 ) were derived as fat free mass (kg) and body fat mass (kg), respectively, divided by height squared (m 2 ).The overall procedure (whole body and femoral scan) took approximately 30 minutes.

Physical activity
Seven days of physical activity was objectively assessed by using the ActiGraph GT1M accelerometer (Manufacturing Technology Inc., FL, USA).The data were intentionally collected during the moderate Springs of 2010 and 2011, avoiding warm summers and cold winters with icy and snowy surfaces.After the BMD measurement, participants were instructed on how to affi x the accelerometer snugly by means of an elastic belt cinched at the mid axillary line of the right hip.Moreover, they were asked to wear the instrument during all waking hours (at least 10 hours a day for a consecutive eight days), with the exception of times used to perform water activities.
Daily activity and inactivity logs were used for par ticipants to self record times (minutes) when the accelero meter was worn and activity (e.g.walking, conditioning activities, gardening, household activities, swimming, cycling) and inactivity patterns (watching TV, working at the computer, sitting while reading, tra veling, etc.) within each day of the week.Using manufacture supplied software, the time sampling interval of the accelerometers was set at 1 minute and the step mode was activated.Downloaded counts data were assessed and  et al. (2005).

Other data collection
All participants completed a questionnaire containing questions on potential risk factors for osteopenia/ osteoporosis including medication, smoking, physical activity in the past and menstrual history.

Data analysis
BMD at the proximal femur was used as the variable for sample grouping.The statistical package Statistica 9 (StatSoft, 2009) was used for data analysis with signifi cant levels set at p < 0.05.As the data was not found to be normally dis tributed for some variables examined (minutes), in these cases descriptive statistics were presented as medians (Me) and in an interquartile range (IQR).In cases of step based variables, mean (M) and standard deviation (SD) were used as the descriptive statistics and the t-test was used to test the diff erences in steps/day in days of the week between normal BMD and osteopenia groups.The Mann-Whitney U-test was used to test the diff erences in physical activity and physical inactivity between the normal BMD group and the osteopenia group.Pearsonʼs chi-square test (χ 2 ) was used to evaluate how women in the normal BMD group and in the osteopenia group diff er in their frequency of meeting PA recommendations (Haskell et al., 2007;Tudor-Locke & Bassett, 2004; U. S. Department of Health and Human Services, 2008).
In cases where we used the Mann-Whitney U-test, the Cohen's d (d = ) was calculated as an indicator of eff ect size (Cortina & Nouri, 2000).The values of 0.2, 0.5 and 0.8 were interpreted as having a small, medium and/or large eff ect (McCartney & Rosenthal, 2000;Thomas & Nelson, 2001).In case of Pearsonʼs chi-square test, Pearsonʼs ceoffi cient (r Ф = was used as an indicator of effect size (r = 0.1 small eff ect, r = 0.3 medium eff ect, r = 0.5 large eff ect) (McCartney & Rosenthal, 2000).

RESULTS
In comparison with European fi ndings (EU Sport Ministers, 2008), the women in this study were predominantly active with an average of 9,943 ± 3,623 steps/day, 8.8 ± 25 minutes of vigorous PA a week, 229 ± 154 minutes of moderate PA a week and 539 ± 271 minutes of light PA a week.
Fig. 1 depicts the diff erences in average steps/day in days of the week between the normal BMD group and the osteopenia group.Women with normal BMD had a slightly higher number of steps/day than women with osteopenia in six days of the week.However, these diff erences were not statistically signifi cant.  2 shows the diff erences in physical activity and physical inactivity between the group with normal BMD and the osteopenia group.Women with normal BMD spent signifi cantly more hours/day (p 0.05) being active than women with osteopenia.On the other hand, women with osteopenia perform significantly more household related PA (p 0.05).No signifi cant diff erences between the groups were found in relation to any other PA related variables.
When the data were analyzed according to physical inactivity, women with osteopenia spent signifi cantly more minutes/week while doing diff erent mental activities (reading, doing crossword puzzles, etc.) in a sitting position than women with normal BMD.Moreover, the osteopenia group was watching TV by 195 minutes a week more than the normal BMD group.However, this difference was not statistically significant.Both groups spent a similar amount of time while working at the computer or traveling.
No signifi cant diff erences were found between groups when the data were analyzed according to meeting different PA recommendations (TABLE 3).Nevertheless, the women with normal BMD were more successful in meeting all PA recommendations than women with osteopenia.The main diff erence was found in meeting the criterion of 10,000 steps/day.Approximately 22% more women with normal BMD as compared to women with osteopenia met this criterion.

DISCUSSION
In this cross sectional, descriptive study, the associations between habitual physical activity and inactivity and femoral BMD were investigated.Out of the study sample, 72.8% women had normal femoral BMD and osteopenia was indicated in 27.2% of the women.However, no women with osteoporosis were found.These numbers support the fi ndings that osteopenia is more common than osteoporosis (Khosla & Melton, 2007) and, moreover, than the majority of fractures occur in the population of patients with osteopenia (Pasco et al., 2006).Within this information framework, an investigation of diff erences in physical activity and inactivity between women with normal BMD and osteopenia seems to be suitable.
The conclusion to be drawn from the meta-analyses of published studies is that high intensity resistance train ing as well as mixed loading exercise, combining jogging with other low impact loading activity (stair climb ing, walking), or a combination of high or low impact exercise with high magnitude exercise (resistance training) can be eff ective in preserving bone mass in post menopausal women (Martyn-St James & Carroll, 2006, 2011).In this study, only a few women participated in any resistance or high impact exercises and vi gorous PA, whereas a majority of the women reported household related physical activity.A similar portion of physical activities was reported by Wasburn et al. (1999) in the Physical Activity Scale for the Elderly (PASE), where approximately 60% of this questionnaireʼs points were due to household related activity, whereas only 10% of the total PASE scores were accounted for by leisure time exercise activity (walking, sports, muscular strength/endurance).We found out no signifi cant diff erences were found between normal BMD and the osteopenia group in resistance and high impact activities.As for household activity, the osteopenic women reported signifi cantly more minutes than women with normal BMD.However, based on accelerometer data, women with normal BMD had signifi cantly more total physical activity than osteopenic women.These results might suggest, consistent with the above mentioned analysis, that low intensity physical activity (e.g.household activities) has no impact on BMD.
Based on a physical inactivity analysis, the women with osteopenia spent signifi cantly more minutes (almost twice as many) while sitting than women with nor mal BMD.There is limited evidence about the association between physical inactivity and BMD (Nguyen, Sambrook, & Eisman, 1998).Nevertheless, the ma jority of studies concerning inactivity issues have examined the impact of physical inactivity on the risk of hip frac-ture (Gregg, Pereira, & Caspersen, 2000).There is considerable evidence from epidemiologic studies that physical inactivity is a risk factor for hip fracture (Gregg, Pereira, & Caspersen, 2000).This risk decreases when physical activity increases and conversely, the risk increases in individuals who became inactive (Feskanich, Willet, & Colditz, 2002).
Postmenopausal women prefer mainly moderate in tensity physical activity and walking (Gába et al., 2009).Several physical activity recommendations based on moderate intensity have been advocated as a target for a healthy lifestyle and, additionally, for more health benefi ts (Haskell et al., 2007;Tudor-Locke & Bassett, 2004; U. S. Department of Health and Human Services, 2008).However, the link between these targets and bone health is not clear.Hence, in this study the diff erence in meeting the above mentioned recommendations among women with diff erent levels of BMD was investigated.A higher, although not signifi cant, percentage in meeting all recommendations was observed in women with normal BMD comparing with women with osteopenia.The most apparent, but not signifi cant, diff erence was found in meeting the value of 10,000 steps/day.Boyer et al. (2011) suggest that the required step value for women with average body weight (65.1 kg) to maintain normal BMD (T-score of -1.0) is 4,892 steps/day.However, for lighter women even the value of 10,000 steps/day is insuffi cient, unless they increase their walking speed.Similarly, other studies reported walking speed as an important factor in femoral BMD (Coupland et al., 1999).However, in this study only volume (steps/day; minutes/day) and frequency were taken under consideration and walking speed was beyond the scope of this study.Nevertheless, based on accelerometer step data, women with normal BMD walked more and more often reached the target of 10,000 steps/day than women with osteopenia.Although the causal relationship between walking and femoral BMD was not examined in this stu dy, the results are consistent with the meta-analysis suggesting the signifi cant positive eff ect of walking on the preservation of BMD at the femoral neck (Martyn-St James & Carroll, 2008).
The study has several limitations.The small sample size within two unequal groups and the non random se lection of participants do not allow any generalization.Furthermore, this study is only descriptive and semi-longitudinal.Longitudinal studies would be more valuable in investigating the eff ect of habitual physical activity and inactivity on femoral BMD.While analyzing walking, walking speed was found to be an important factor (Boyer et al., 2011), however it is missing in this study.The above mentioned limitations might be taken under consideration while conducting further studies.

CONCLUSION
In conclusion, the results of this study show the differences in physical activity and physical inactivity in women with diff erent levels of femoral BMD.Regardless of the intensity of physical activity, women with nor mal BMD are signifi cantly more active than women with osteopenia.As for physical inactivity, women with osteopenia spend twice as much time sitting during different leisure time activities.
The T-score of the BMD of the pro ximal femur was calculated by a scanner software platform.According to the World Health Organization (2003), T-score ≥ -1 indicates normal BMD, T-score -1 and !-2.5 indicates osteopenia, and T-score ≤ -2.5 indicates osteoporosis.Within this classifi cation, 67 women had normal femoral BMD and osteo penia was indicated in 25 women.No women with osteoporosis were found within the study sample.

Fig. 1
Fig. 1Average number of steps/day in days of the week of the study subjects grouped by femoral bone density

TABLE 1
Characteristics of study subjects grouped by femoral bone density Legend: YSM -years since menopause, BMI -body mass index, BFMI -body fat mass index, FFMI -fat free mass index, BMD -bone mineral density cleaned according to procedures reported by Eslinger

TABLE 2
Physical activity and inactivity and percentage of meeting PA recommendations of the study subjects grouped by femoral T-score

TABLE 3
Percentage of meeting PA recommendations of the study subjects grouped by femoral T-score Legend: BMD -bone mineral density, MPA -moderate physical activity, VPA -vigorous physical activity, † -accelerometer based data