Aging Clin Exp Res DOI 10.1007/s40520-014-0253-8

ORIGINAL ARTICLE

Investigation of key factors affecting the balance function of older adults Fang Pu • Sichen Sun • Ling Wang • Yan Li • Hongmin Yu • Yang Yang • Yuanyuan Zhao • Shuyu Li

Received: 8 April 2014 / Accepted: 4 June 2014 Ó Springer International Publishing Switzerland 2014

Abstract Background and aims Previous studies have focused mainly on individual factors affecting the balance function of older adults. However, it is largely unknown whether the balance functions of older adults are affected by multiple factors occurring simultaneously, and what is predominant among these factors. Methods We adopted a cross-sectional study design and recruited 100 older adults from the community. Each participant was required to complete a questionnaire consisting of 20 questions related to four factors: sociodemographic, physical exercise, sleep condition and mental condition. We then evaluated all participants’ static and dynamic balance abilities using two balance tests performed using the Microsoft KinectTM system. We used MANOVA and FDR corrections to analyze each factor to determine which factors affected the balance parameters. Last, we identified the major factors related to balance by computing the percentage of primary factors with significant effects for each factor.

Electronic supplementary material The online version of this article (doi:10.1007/s40520-014-0253-8) contains supplementary material, which is available to authorized users. F. Pu
S. Sun
L. Wang
Y. Li
H. Yu
Y. Yang
Y. Zhao
S. Li (&) Key Laboratory of Rehabilitation Technical Aids of Ministry of Civil Affair, School of Biological Science and Medical Engineering, Beihang University, No.37 Xueyuan Road, Haidian District, Beijing 100191, China e-mail: [email protected] F. Pu State Key Laboratory of Virtual Reality Technology and Systems, Beihang University, No.37 Xueyuan Road, Haidian District, Beijing 100191, China

Results We found that static balance function was mainly affected by sociodemographic factors, sleep condition and mental condition. The dynamic balance function showed close relationships with physical exercise and sleep condition. Furthermore, sleep condition had a larger effect on static balance function than on dynamic balance function. We also observed an association between static balance function and mental condition. Conclusion A wide range of factors were associated with balance function in these older adults. The static and dynamic balance functions were related with different factors; this might provide useful information for older adults on maintaining good balance ability. Keywords Older people
Balance function
Kinect
Multiple factors

Introduction Balance function is crucial in our daily life; it maintains body posture and helps us to perform different activities and to respond appropriately to outside stimuli. With increasing age, the balance function of older people will decline because of age-related physiological changes, such as decreased nerve conduction velocity [1], increased central processing time [2], decreased muscle strength [3], and increased passive tissue stiffness [4]. The deteriorated balance function may largely increase the risk of falls in older adults [5]. Thus, it is important to determine the key factors affecting the balance function in older adults, to maintain and improve their balance function. Previous studies have found that some sociodemographic factors, such as gender and education level, affect balance functions in older adults. For example, Lord et al.

123

Aging Clin Exp Res

[6] reported that dynamic ability scores for a group of male older adults were significantly higher than those for females. Voos et al. [7] found that education level affected the cognitive function of older adults, as well as their performances on balancing tests. Importantly, balance functions in older people have been found to improve with physical exercise. Wong et al. [8] found that older people who practiced tai chi were more likely to be able to maintain a stable posture while doing challenging balance tasks. Furthermore, Seco et al. [9] showed that long-term physical exercise may improve the balancing ability of older people. In addition, several studies have demonstrated that sleeping conditions may also influence balance function. For instance, Ma et al. [10] found that subjects who experienced one night of sleep deprivation felt fatigued and demonstrated postural instability during balance function tests. As is evident from the studies discussed above, the majority of existing studies have focused on individual factors affecting the balance function in older adults. Such factors have included age, gender, sleep condition and physical exercise. It is largely unknown whether multiple factors influence the balance ability of older adults, as well as the contribution of each factor to balance function. Notably, there are few studies that have explored whether balance function in older people is affected by mental condition. The existing studies have reported that impaired mental condition could increase the risk of falls in a population of older adults. For instance, Quach et al. [11] showed that depression increased the risk of indoor and outdoor falls among community-dwelling older people. Martin et al. [12] found that reduced cognitive function was an important risk factor for falling in specific patient groups. The high risk of falls could be closely related to the deterioration of balance function in older people. However, it is unknown whether impaired mental condition might decrease balance function and further result in a high risk of falling. Thus, the aim of this study was to investigate how multiple factors, such as sociodemographic factors, physical exercise, sleep condition and mental condition, influence the balance ability of older adults, as well as to

Table 1 Sociodemographic data for all subjects

Materials and methods Subjects One hundred volunteers were recruited randomly from two communities in Beijing, China. The inclusion criteria were as follows: (1) more than 55 years of age; (2) the subject had not received previous special motor skills training. The following exclusion criteria were applied to all subjects: the existence of a neurological disorder, alcohol or drug abuse, or any physical disability or dyskinesia which could have affected the balance tests as assessed based on clinical evaluations and medical records. All subjects were screened using the Mini-Mental State Examination (MMSE). Thirteen subjects were then excluded because their MMSE scores were lower than 26 or they failed to complete the balance tests. The remaining 87 subjects consisted of 35 males and 52 females, ranging from 55 to 83 years of age, and an average MMSE score of 28.3 ± 1.7 [13]. The detailed demographic information is shown in Table 1. In our sample, 15 and 20 % of all the subjects indicated a low frequency (less than four times in a week) of smoking and drinking, respectively. Seventyseven percent of our sample had a chronic disease; this is comparable with the national percentage (78.93 %) of people with chronic disease, according to the ‘‘Report on chronic situation and development trend of elderly people in China’’ in 2005 [14]. These chronic diseases included diabetes, coronary heart disease, hyperlipidemia, arthritis, cervical spondylosis, lumbar disc herniation, osteoporosis

Age (years)

Gender

Height (cm)

Weight (kg)

BMI

Education (years)

64.82 ± 7.31 (55–83)

52 F/35 M

163.79 ± 6.46 (150–176)

66.75 ± 11.43 (45–115)

24.82 ± 3.56 (17.97–41.73)

10.92 ± 2.82 (0–16)

1: \60

1: male

1: \160

1: \60

1: \18.5

1: \6

2: 60–70

2: female

2: 160–170

2: 60–70

2: 18.5–24

2: 6–12

3: [170

3: [70

3: 25–28

3: 13–16

4: [28

4: [16

3: [70

123

identify the predominant factor affecting balance by analyzing the effect of each individual factor on balance function. Here, we adopted a cross-sectional study design. For the recruited elderly subjects, we investigated the factors that may be related to the balance functions of older people by questionnaire tests. And then we measured their static and dynamic balance abilities using two balance tests performed using the Microsoft KinectTM system. Statistical analysis was used to determine which factors affected the balance parameters.

Aging Clin Exp Res Fig. 1 Survey results for subjects’ drinking, smoking and chronic diseases. This figure shows that subjects in this study had different lifestyles and that they had some mild chronic disease, as shown along the Yaxis. The first four bars of the figure pertains to how many subjects smoke and drink, and the other bars pertain to how many subjects had certain diseases based on the information provided in their questionnaire. CHD coronary heart disease, CS cervical spondylosis, HIVD herniated intervertebral disc

and hypertension. More detailed information pertaining to these chronic diseases is shown in Fig. 1. However, none of the subjects used benzodiazepines or other drugs that could have affected the central and peripheral nervous system. Written informed consent was obtained before the experiment and the Ethics committee of the Beihang University approved the study. Questionnaire design The questionnaire included four primary factors and 20 questions. It included sociodemographic, physical exercise, sleep condition and mental condition factors. Each of these primary factors comprised a few individual factors (i.e., questions). We designed these questions based on the existing literature about factors affecting the balance function in older adults, as well as our hypotheses. The classification and labels of all factors are shown in Table 2. All subjects were asked to fill in the questionnaire accurately. For those subjects who were not able to complete the questionnaire due to visual problems or illiteracy, the questionnaire was completed by the experiment operator according to the oral choices made by these subjects. Instrumentation and data acquisition Balance testing was performed using Kinect (Microsoft, WA, USA) and an in-house developed software called ‘Motor Abilities Testing and Training System’ (MATT). Kinect is a motion-sensing device developed by Microsoft. It captures the positions of 20 bone markers by its internal human body recognition algorithm using the data acquired from its RGB camera and depth sensor. Going beyond its intended purpose of playing games, some researchers have

attempted to use it for motion capture in rehabilitation testing and training. Furthermore, its accuracy and reliability have also been validated [15–17]. MATT was designed to test and train motor abilities, such as balance and coordination, using a series of tasks which received the data acquired by Kinect as input. Experiment process The subjects were required to complete two balance tests in the MATT system according to the instructions given by the experiment operator. The data were collected by Kinect. The tests included a single limb-standing test for evaluating static balance ability and a stepping test for dynamic balance ability. Experiment 1: stand on one leg test (SL) Subjects were asked to stand on one leg (their dominant leg), let their arms fall down naturally at the sides of their body, keep their heads upright, look straight ahead and remain focused. Meanwhile, to avoid the possible displacement of the mass center, subjects were asked to refrain from speaking or performing any body movement until they could not resist it any more. Kinect started to collect data 3 s after the experiment commenced. The system would automatically stop after data acquisition lasting 20 s. Five parameters including the time length of standing on one leg, root mean square of displacement (RMS), the envelope area of gravity center (EA), postural sway average speed (LNGV) and track length per unit area (LPA) were calculated and used for further statistical analysis.

123

Aging Clin Exp Res Table 2 List of all factors and results of the partial correlation Primary factor

Socio-demographic

Individual factor

1

Age

Static balance TIME

RMS

EA

LNGV

LPA

TIME

RMS

EA

LNGV

LPA

r2

0.323

0.03

–

–

–

–

–

–

–

–

P

0.000

0.114

2

Height

–

–

–

–

–

–

–

–

–

–

3

Weight

–

–

–

–

–

–

–

–

–

–

4

Gender

–

–

–

–

–

–

–

–

–

–

5

Education

–

0.001

–

0.005

–

–

–

–

–

–

–

0.732 –

–

0.499 –

–

–

–

–

–

–

–

–

–

–

–

–

–

0.001

–

–

–

–

–

–

r2 P

Physical exercise

Dynamic balance

6 7

Exercise frequency Exercise duration

r

2

P 8

Promenade

r2

0.966 –

–

–

–

–

–

–

P 9

Aerobics

10

Ball Games

11

Dreaminess

r2

0.825 –

–

–

–

–

–

–

0.015

–

–

–

–

–

–

–

–

–

–

–

0.012

–

0.034

0.026

–

–

–

–

–

0.093

0.149 –

–

–

–

–

P Sleep condition

r2

0.27

P 12

Difficulty falling asleep

r2

0.321 –

P

Mental condition

0.001

0.060

0.02

0.11

0.045

0.028

0.216

0.325

0.055

13

Egregorsis

–

–

–

–

–

–

–

–

–

–

14

Dyssomnia

–

–

–

–

–

–

–

–

–

–

15

Dyssomnia frequency

r2 P

–

–

–

0.018 0.231

0.012 0.329

–

–

–

–

0.009 0.4

16

Depression

r2

–

0.261

0.098

–

–

–

–

–

–

–

0.017

0.375

17

Memory deterioration

–

–

–

–

–

–

–

–

–

P r2

0.021

P

0.192

18

Lack of concentration

–

–

–

–

–

–

–

–

–

–

19

Emotional instability

–

–

–

–

–

–

–

–

–

–

2

0.001

0.016

0.016

–

–

–

–

–

–

–

P

0.886

0.260

0.257

20

MMSE score

r

Only the items which were significantly correlated in the MANOVA test are tested by partial correlation The significance of italicized values are p \ 0.1

Experiment 2: step on the same place with eyes closed test (SE)

Statistical analyses Reliability test of questionnaire

First, each subject was required to stand quietly, and a range cycle was defined by setting the middle point as between the subject’s feet and 20 cm as the radius. Then the subject stepped following the rhythm of the audio (frequency of 120 steps/min) while keeping his/her eyes closed. Based on the spatial coordinates captured by Kinect, the program could automatically calculate the COM (center of mass) of each subject during the test. The test stopped once the Kinect identified that the feet of the subject exceeded the range cycle. Finally, the same five parameters as in the above experiment were calculated by Kinect.

123

To ensure the reliability of the questionnaire, we analyzed the answers of the questionnaires, shown in Table 3, using a scale reliability analysis (i.e., correlation and factor analysis). We calculated the correlation of the score for each item and the total scores for the questionnaire. A higher coefficient meant a higher internal consistency. This indicated that the question item in this questionnaire reliably measured the same latent variables as the whole questionnaire. In addition, we also performed a homogeneity test using principle component analysis during the

Aging Clin Exp Res

extraction of factor analysis. Gulliksen [18, 19] originally used the item homogeneity test as an estimate of scale reliability, which indicated the item redundancy and narrowness of a scale. Items should be selected which are loaded maximally by the factor representing that scale. Normality and homogeneity of variances tests The general assumptions for MANOVA are multivariate normality of data and homogeneity of variances; thus, we used a Kolmogorov–Smirnov test to check the normality of our data, and the Levene’s test to check for homogeneity of variances. MANOVA analysis We used SPSS 17.0 (SPSS Inc., Chicago, IL, USA) to perform a MANOVA analysis to examine the impact of each individual factor on each measure of balance function within each primary factor. Thus, we tested Table 3 Analysis of questionnaire items Items

Exercise frequency Exercise duration Promenade

Items and total correlation

0.741** 0.617** -0.094

Results Homogeneity test Communalities 0.235

Factor loading 0.485

0.454

0.673

0.544

-0.737

Aerobics

0.274*

0.268

0.518

Ball games

0.363**

0.234

0.483

Swimming

0.367**

0.005

-0.067

Bicycling

0.085

0.022

-0.148

Dreaminess

0.609**

0.428

0.654

Difficulty falling asleep

0.432**

0.156

0.395

Egregorsis

0.355**

0.123

0.351

Dyssomnia

0.970**

0.958

0.979

Dyssomnia frequency

0.944**

0.86

0.928

Depression

0.407**

0.356

0.596

Memory deterioration

0.419**

0.367

0.606

Lack of concentration Emotional instability

0.583**

0.593

0.770

0.389**

0.296

0.544

MMSE score Standard

0.759** [0.35

N 9 M associations in this study, where N is the number of all individual factors (N = 20) and M is the number of measures by two balance tests (M = 10). P values \0.05 were considered statistically significant. We also employed multiple testing corrections to reduce the false positives using the false discovery rate (FDR) method. After FDR correction, the results with significant associations are shown in Fig. 2. After MANOVA analysis, we further performed discriminant function analysis for those statistically significant factors to describe the contribution of each dependent variable to a function describing differences among factor levels. To further investigate the effect of each primary factor on balance function, we calculated the percentage of individual factors with high significance (P \ 0.05) for each primary factor in all four primary factors. In addition, we performed the partial correlation analysis among the items that showed significant correlations in MANOVA tests, which measured the mutual relationship between two variables when other variables were held constant with respect to the two variables.

0.050 [0.2

0.224 [0.35

* The significance level is 0.05, items are significantly correlated ** The significance level is 0.01, items are significantly correlated The italic values mean that they don’t reach the standard thresholds

The questionnaire results For each individual factor, we divided the values into different levels for MANOVA. The division range for each sociodemographic factor is shown in Table 1. For the individual factors, each subject chose one level based on his/her experience. In this study, we performed a reliability test for the questionnaire; two items (swimming and bicycling) were rejected for item homogeneity, because they had very low commonalities and factor loadings. The effects of individual factors on balance function The results of the two balance function tests are shown in Table 4. We performed MANOVA analyses for each individual factor to investigate its effect on balance function. We then performed an FDR correction to correct for multiple comparisons by adjusting the significance level. The results of these statistical analyses are demonstrated in Fig. 2. The X-axis represents 20 individual factors (the corresponding relationships between label and name are shown in Table 2) as well as the four primary factors. The Y-axis represents the measurements of balance ability. The color of each block represents the P value (corrected), where a red hue from light to deep color indicates an increase in statistical significance. The effect of each primary factor on balance function, computed using the percentage of statistically significant (P \ 0.05) individual factors, is shown in Table 5. The amount of variability for

123

Aging Clin Exp Res

Fig. 2 Statistical results of the FDR corrections. The results of the FDR corrections are shown in Fig. 1. The X-axis represents 20 individual factors (the corresponding relationship between label and name is shown in Table 2) as well as the four primary factors. The Y-

axis represents the measurements of static balance and dynamic balance ability. The color of each block represents the P value, where a red hue from a light to deep color indicates an increase in statistical significance

Table 4 Results from the static and dynamic balance tests

further found that there were linear dependence between sleep condition, mental condition and static balance function. These results are shown in Table 2.

Static balance

Dynamic balance

TIME (s)

13.14 ± 7.16

14.85 ± 8.39

RMS (mm)

18.11 ± 16.71

61.86 ± 13.27

EA (mm2)

1,272.77 ± 1,541.24

12,636.27 ± 6,362.08

LNGV (mm/s)

32.27 ± 29.12

87.69 ± 51.89

LPA (mm-1)

0.38 ± 0.27

0.1 ± 0.05

Table 5 Percentage of each primary factor in all four primary factors during balance tests Sociodemographic (%)

Physical exercise (%)

Sleep condition (%)

Mental condition (%)

Static balance

12

0

36

24

Dynamic balance

0

12

4

0

the dependent variable explained by the main factor is listed in Table 2. We found that there were different effects of each primary factor on the static and dynamic balance functions. The main factors affecting the static balance function included sociodemographic factors, sleep condition and mental condition. The dynamic balance function was primarily affected by physical exercise and sleep condition. Among these factors, sleep conditions showed more important effects on balance function compared with other factors. Through partial correlation analysis, we

123

Discussion In the present study, we enrolled 87 older subjects and used a questionnaire to survey several factors, including physical exercise, mental condition, sleep condition and sociodemographic factors. The subjects’ balance functions were evaluated using two balance tests. We found that the main factors affecting the balance function of older people involved sociodemographic factors, physical exercise, sleep condition and mental condition. Our results may provide useful information for older adults for maintaining good balance ability during normal aging. In this study, we found that some sociodemographic factors, such as age and education level had significant effects on static balance function. Many studies have reported a reduction of balance function with increasing age; this may be due to increased reaction time, decreased muscle strength and bone loss. In addition, our results showed that one of the parameters measured in the balance function test was significantly different among subject groups with different education levels. This finding is supported by a previous study, which showed a positive correlation between level of education and balance function (as measured by the Berg Balance Scale) [7]. A possible explanation for this may be that individuals with limited education are less likely to have had enough

Aging Clin Exp Res

cognitive resources available in terms of visual perception, memory, divided attention, coordination, manual dexterity, and motor sequencing [20, 21], which might result in lessefficient postural control and balance ability. Our results suggest that those older people with a lower education level should pay more attention to maintaining and improving their balance functions. An interesting finding of our study was that sleep condition had an effect on both static and dynamic balance function. According to our knowledge, very few studies have investigated the effect of sleep condition on balance function. The existing studies have mainly focused on the effects of acute sleep deprivation on balance function. For example, Ma et al. [10] reported that subjects who experienced one night of sleep deprivation felt fatigued and demonstrated postural instability during balance tests. After 24 h of sleep deprivation, postural control in healthy subjects was significantly affected both in anteroposterior and lateral directions [22]. Sleep deprivation may lead to reduced adaptation ability and lapses in attention and thereby affect postural stability [23]. Thus, we speculated that those older adults who experienced poor sleep conditions may have had worse cognitive function and felt fatigued, which resulted in poor balance function. Our results suggest that good sleep quality is beneficial for maintaining balance function in older people. Another interesting finding was that mental condition had a large effect on static balance function in older people. The aging population is likely to experience some mental problems such as depression, inability to concentrate, irritability, and emotional behavior. However, it is largely unknown whether the balance function of older people is affected by mental condition. To date, there are only a few studies reporting the association between the risk of falls and mental condition. For instance, Quach et al. [11] showed that depression increased the risk of indoor and outdoor falls among community-dwelling older people. Furthermore, Kvelde et al. [24] found that a higher level of depressive symptoms at baseline resulted in a greater likelihood of falling during follow-up. In our study, depression was shown to have the most significant effect on balance ability among mental factors. Therefore, we speculated that the increase of falling risk in older people was partially explained by the reduced balance ability caused by depression. In addition, we also evaluated the effects of cognitive function on balance function using scores from the Mini-Mental State Examination. We found that MMSE score had significant effects on static balance function. A previous study reported that reduced cognitive function was identified as an important risk factor of falling in specific patient groups [12]. Additionally, Montero et al. [25] found that early disturbances in cognitive processes such as attention, executive function, and working memory

were associated with the prediction of future mobility loss and falls. Thus, we inferred that the reduction of cognitive function would cause balance instability, which further increased the risk of falling in older adults. The impact of physical exercise on balance function has been widely reported [8, 26]. Our results showed that the frequency of exercise had an impact on balance function, suggesting that older people should be more involved in physical exercise. Our findings concur with those from previous studies. Barnett et al. [26] found that an exercise group performed significantly better than the control group on 3 of 6 balance measures: postural sway on the floor with eyes open and eyes closed and coordinated stability. Furthermore, older adults who had practiced tai chi had better balance function, and were more likely to be able to maintain a stable posture while doing challenging tasks compared with controls [8]. Notably, we also found that almost all types of exercise only affected the dynamic balance function. A previous study using a similar exercise intervention reported improvements in muscle strength, reaction time and walking speed [26]. In addition, it was widely reported that the exercise could improve the blood circulation and the spirometric variables [27, 28]. Those improvements caused by physical exercise might be able to explain why dynamic function rather than static balance function could be improved by daily exercise. Several issues need to be addressed further. First, it should be with caution that we had not given any causal relationships among these factors and balance functions of elderly people because of the cross-sectional and observational design in our study. In future, the follow-up study would be helpful to clarify this issue. Second, the subjects were randomly enrolled from two communities in Beijing, China. By comparing the average height and weight between our sample and the general Chinese population (See Table 3S), we found that the weights of our sample matched well with the whole nation, but the heights of our sample in many age ranges were higher than those of the entire country, which may be explained by economic factors [29]. Thus, it should be with caution that our conclusions are extended to other regions or countries due to different characteristics of subjects. Third, our subjects had high BMIs with an average 24.8, which was beyond the upper limit of the normal range for China. According to the report on national physical quality released by General Administration of Sport of China in 2010, people with a BMI higher than 24 are considered to be obese and need to pay more attention to losing weight [30]. This might induce some bias in our experimental results. One study has demonstrated that people with a lower BMI have better resistance, flexibility, balance and strength [31]. In future, it would be important to evaluate this factor in older

123

Aging Clin Exp Res

subjects with different BMIs. Fourth, in the current study, our data did not follow the normal distribution. We also excluded the possibility of non-normality of our data. However, the F test is robust to non-normality, if the nonnormality is caused by skewness rather than by outliers [32]. Using the homogeneity of variance, we found that the parameters acquired by the two balance function tests were homoscedastic except for a few parameters (see Table 4S). Thus, we performed MANOVA analyses for each individual factor within each primary factor. In future, the normality criteria may be met by including more samples in the experiment. Furthermore, we chose the ‘standing on one foot’ test and ‘closing the eyes when stepping’ test for the evaluation of balance ability; only time and gravity movement parameters were calculated and used in our analysis. It would be valuable to use more balance tests as well as more measures to evaluate the balance function of older people. Finally, in this study, our questionnaire had been designed to involve four primary factors. It would be interesting to further investigate the effects of more factors such as executive function, flexibility and travel modes on the balance function in older people. Acknowledgments The authors wish to thank all subjects who participated in the research and acknowledge the support from the National Natural Science Foundation of China (Grant No. 61101008, 11072022) and Microsoft Research Asia (FY12-RES-THEME-096). Conflict of interest

None.

References 1. Valerio BC, Nobrega JA, Tilbery CP (2004) Neural conduction in hand nerves and the physiological factor of age. Arq Neuropsiquiatr 62(1):114–118 2. Finkel D, Reynolds CA, McArdle JJ, Pedersen NL (2007) Age changes in processing speed as a leading indicator of cognitive aging. Psychol Aging 22(3):558–568. doi:10.1037/0882-7974.22. 3.558 3. Perry MC, Carville SF, Smith IC, Rutherford OM, Newham DJ (2007) Strength, power output and symmetry of leg muscles: effect of age and history of falling. Eur J Appl Physiol 100(5):553–561. doi:10.1007/s00421-006-0247-0 4. DeVita P, Hortobagyi T (2000) Age increases the skeletal versus muscular component of lower extremity stiffness during stepping down. J Gerontol Ser A Biol Sci Med Sci 55(12):B593–B600 5. Melzer I, Benjuya N, Kaplanski J (2004) Postural stability in the elderly: a comparison between fallers and non-fallers. Age Ageing 33(6):602–607. doi:10.1093/ageing/afh218 6. Lord SR, Sambrook PN, Gilbert C, Kelly PJ, Nguyen T, Webster IW et al (1994) Postural stability, falls and fractures in the elderly: results from the Dubbo Osteoporosis Epidemiology Study. Med J Aust 160(11):684–5, 8–91 7. Voos MC, Custodio EB, Malaquias J Jr (2011) Relationship of executive function and educational status with functional balance in older adults. J Geriatr Phys Ther 34(1):11–18. doi:10.1097/ JPT.0b013e3181ff2452

123

8. Wong AM, Lin YC, Chou SW, Tang FT, Wong PY (2001) Coordination exercise and postural stability in elderly people: effect of Tai Chi Chuan. Arch Phys Med Rehabil 82(5):608–612. doi:10.1053/apmr.2001.22615 9. Seco J, Abecia LC, Echevarria E, Barbero I, Torres-Unda J, Rodriguez V et al (2013) A long-term physical activity training program increases strength and flexibility, and improves balance in older adults. Rehabil Nurs 38(1):37–47. doi:10.1002/rnj.64 10. Ma J, Yao YJ, Ma RM, Li JQ, Wang T, Li XJ et al (2009) Effects of sleep deprivation on human postural control, subjective fatigue assessment and psychomotor performance. J Int Med Res 37(5):1311–1320 11. Quach L, Yang FM, Berry SD, Newton E, Jones RN, Burr JA et al (2013) Depression, antidepressants, and falls among community-dwelling elderly people: the MOBILIZE Boston Study. J Gerontol Ser A Biol Sci Med Sci. doi:10.1093/gerona/glt084 12. Martin KL, Blizzard L, Srikanth VK, Wood A, Thomson R, Sanders LM et al (2013) Cognitive function modifies the effect of physiological function on the risk of multiple falls––a populationbased study. J Gerontol Ser A Biol Sci Med Sci 68(9):1091–1097. doi:10.1093/gerona/glt010 13. Folstein MF, Folstein SE, McHugh PR (1975) ‘‘Mini-mental state’’. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12(3):189–198 14. Report on chronic situation and development trend of elderly people in China. Available at: http://www.lovepama.com/love pama2.pdf. Accessed 2005 15. Clark RA, Pua YH, Fortin K, Ritchie C, Webster KE, Denehy L et al (2012) Validity of the Microsoft Kinect for assessment of postural control. Gait Posture 36(3):372–377. doi:10.1016/j.gait post.2012.03.033 16. Obdrzalek S, Kurillo G, Ofli F, Bajcsy R, Seto E, Jimison H et al (2012) Accuracy and robustness of Kinect pose estimation in the context of coaching of elderly population. Conf Proc 2012:1188–1193. doi:10.1109/EMBC.2012.6346149 17. Yang Y, Pu F, Li Y, Li S, Fan Y, Li D (2014) Reliability and validity of Kinect RGB-D sensor for assessing standing balance. Sens J IEEE 14(5):1633–1635. doi:10.1109/JSEN.2013.2296509 18. Gulliksen H (1950) Theory of mental tests. Wiley, New York 19. Boyle GJ (1985) Self-report measures of depression: some psychometric considerations. Br J Clin Psychol 24(1):45–59 20. Snowdon DA, Ostwald SK, Kane RL, Keenan NL (1989) Years of life with good and poor mental and physical function in the elderly. J Clin Epidemiol 42(11):1055–1066 21. Nitrini R, Caramelli P, Herrera E Jr, Charchat-Fichman H, Porto CS (2005) Performance in Luria’s fist-edge-palm test according to educational level. Cogn Behav Neurol 18(4):211–214 22. Patel M, Gomez S, Berg S, Almbladh P, Lindblad J, Petersen H et al (2008) Effects of 24-h and 36-h sleep deprivation on human postural control and adaptation. Exp Brain Res (Experimentelle Hirnforschung Experimentation cerebrale) 185(2):165–173. doi:10.1007/s00221-007-1143-5 23. Gomez S, Patel M, Berg S, Magnusson M, Johansson R, Fransson PA (2008) Effects of proprioceptive vibratory stimulation on body movement at 24 and 36 h of sleep deprivation. Clin Neurophysiol 119(3):617–625. doi:10.1016/j.clinph.2007.10.058 24. Kvelde T, McVeigh C, Toson B, Greenaway M, Lord SR, Delbaere K et al (2013) Depressive symptomatology as a risk factor for falls in older people: systematic review and meta-analysis. J Am Geriatr Soc 61(5):694–706. doi:10.1111/jgs.12209 25. Montero-Odasso M, Verghese J, Beauchet O, Hausdorff JM (2012) Gait and cognition: a complementary approach to understanding brain function and the risk of falling. J Am Geriatr Soc 60(11):2127–2136. doi:10.1111/j.1532-5415.2012.04209.x 26. Barnett A, Smith B, Lord SR, Williams M, Baumand A (2003) Community-based group exercise improves balance and reduces

Aging Clin Exp Res falls in at-risk older people: a randomised controlled trial. Age Ageing 32(4):407–414 27. Fatima SS, Rehman R, Saifullah, Khan Y (2013) Physical activity and its effect on forced expiratory volume. JPMA 63(3):310–312 28. Hamdorf PA, Withers RT, Penhall RK, Haslam MV (1992) Physical training effects on the fitness and habitual activity patterns of elderly women. Arch Phys Med Rehabil 73(7):603–608 29. Boix C, Rosenbluth F (2006) Bones of contention: the political economy of height inequality. Am Polit Sci Rev 108(1):1–22. doi:10.1017/S00030554130005551

30. Report on national physical quality in 2010. Available at: http://www. gov.cn/test/2012-04/19/content_2117320.htm. Accessed 2011.9.2 31. Vaquero-Cristobal R, Martinez Gonzalez-Moro I, Alacid Carceles F, Ros Simon E (2013) Strength, flexibility, balance, resistance and flexibility assessment according to body mass index in active older women. Revista espanola de geriatria y gerontologia 48(4):171–176. doi:10.1016/j.regg.2012.11.003 32. Mardia KV (1971) The effect of non-normality on some multivariate tests and robustness to non-normality in the linear model. Biometrika 58(1):105–112

123