Influence of the length of institutionalization on older adults' postural balance and risk of falls: a transversal study

OBJECTIVE: to ascertain the influence of the length of institutionalization on older adults' balance and risk of falls. METHOD: to evaluate the risk of falls, the Berg Balance Scale and the Timed Get Up and Go test were used; and for measuring postural balance, static stabilometry was used, with acquisition of the elliptical area of 95% and mean velocities on the x and y axes of center of pressure displacement. Parametric and nonparametric measures of association and comparison (α<0.05) were used. RESULTS: there was no significant correlation between the length of institutionalization and the tests for evaluation of risk of falling, neither was there difference between groups and within subgroups, stratified by length of institutionalization and age. In the stabilometric measurements, there was a negative correlation between the parameters analyzed and the length of institutionalization, and difference between groups and within subgroups. CONCLUSION: this study's results point to the difficulty of undertaking postural control tasks, showing a leveling below the clinical tests' reference scores. In the stabilometric behavior, one should note the reduction of the parameters as the length of institutionalization increases, contradicting the assumptions. This study's results offer support for the development of a multi-professional model for intervention with the postural control and balance of older adults living in homes for the aged.


Introduction
As well as the advance in age, aging is characterized by an inexorable functional decline of organs and systems, this being influenced by genetic factors, environmental determinants and lifestyle, which act at different levels of complexity (1) . This decline, although understood as a physiological process, has posed an important challenge to public health, in the search for maintenance of older adults' functional autonomy (2) and independence and the preservation of their quality of life (QL).
In this regard, abilities of the nervous system, and the sensory interactions with the motor responses, which are determinants for balance and postural control, present a reduction. As a result of this, for an older adult to fall becomes a probable risk (3) , reaching over 30% in non-institutionalized older adults; the majority of these older adults are recurrent fallers, which causes multiple harms to the health of this population (4)(5) .
The increase in longevity has raised various questions for the management of public policies, among which there is the increase in demand for Homes for the Aged (HA). Living in these institutions, however, can promote social isolation, reduction in mental and physical activities, and a worsening in the older adults' QL. Moving to HAs is strongly associated with a decline in the abilities to undertake activities of daily living (ADLs) and a progressive reduction in opportunities for mobility. Thus, for older adults, some activities which are apparently simple, such as walking, can become risky and difficult to undertake. Consequently, this contributes to the failure to undertake ADLs, inducing a hypokinetic routine for them, which becomes an intervening factor for falls (5) .
Based on the above, it is appropriate to acknowledge that length of institutionalization (LI) in HAs influences postural control and, as a consequence, the falls. Relating these variables, however, remains little explored in the literature. This study's objective was to ascertain the influence of LI on the postural balance and risk of falls of older adults resident in HAs, through functional tests of postural balance (6)(7) and through static stabilometry (8) .
Batista WO, Alves Junior ED, Porto F, Pereira FD, Santana RF, Gurgel JL.  Using the data obtained from consulting the medical records of the older adults included in the study, the sample was stratified and homogenized by: age ranges, considering the ages between 61 and 90 years old presented in this population; length of institutionalization (LI), whose periods in this sample varied between 07 and 231 months, there being no residents with LI between zero and six months or for exactly 64 months; and falls in the previous year, in which were considered the 12 months which preceded the study for the classification of the older adults as fallers and non-fallers. For these data, the respective medians were taken into consideration, the aim being that the same should have a symmetric division ( Figure 1).

Institutionalized Group Length 1 (LI1)
Length of institutionalization from 7 to 63 months In the test in question, the tasks are carried out in the following order: seated position to standing; to remain standing without support; to remain seated without support; from standing to seated position; transference from one chair to another; to remain standing up with the eyes shut; to remain standing up with the feet together; standing up, with the arms held forwards; standing up, to pick up an object from the ground; to turn around to look behind; to turn around 360°; position the feet alternatively on a step; to remain standing with one foot forward; and to remain standing on one leg. For this test, the cutoff point of 45 points was adopted, with intraclass correlation coefficients (ICC) for intraexaminer and interexaminer parameters of reliability of 0.99 and 0.98, respectively (6) .
The BBS was complemented by the Timed Get up and Go Test (TUG) (7) , which is also a test for tracking the older adult population's risk of falls; this has not been transculturally adapted, and is equivalent for measuring because it is a linear metric measurement, in meters (m) and time measurement in seconds (s). This test is widely used, and through it one can measure dynamic balance and walking speed with high intra-and interexaminer ICC (10) . It is carried out in the following way: measurement of the time (s) which an individual takes to get up from the chair, move 3m, go back and turn around to sit in the same chair. In this test, the cutoff point of 20 s was adopted, as recommended by the literature (3,7,10) .
For the laboratory evaluation of the balance, static stabilometry was used, which evaluates the postural balance through the quantification of the postural sway in the orthostatic position on a force platform.
Normally administered under different protocols, it has a broad application in areas such as rehabilitation, otolaryngology, orthopedics, pharmacology, geriatrics and sport training (8) . This test is considered the gold standard in the evaluation of balance and postural control.
For the test, a force platform was used (0.50m x 0.50m) constructed and calibrated in the Physical Education Institute of the Fluminense Federal University (11) . The variables for analysis of the  Significant differences were found in the comparisons undertaken between the older adults who fell and those who did not fall in all the tests suggested (Table 4).

Discussion
The results demonstrate that the scores for the risk of falls (BBS and TUG) of the population studied (Tables   2, 3 (Table 4), even with the possibility of underreporting, which may mask an even worse scenario. These results corroborate evidence discovered in other Brazilian cities (5,13) , such as was evidenced in a retrospective study (14) which ascertained the prevalence of falls in residents in HAs in São Paulo. In the study, following analysis of 121 hospital records and 87 reports of falls in the period of 12 months, 114 events of falls suffered by 45 older adults were found; with a prevalence of 37.2% of older adults who fall, of whom 46.7% had suffered multiple falls (14) .
In the present study, when the results of the clinical tests of evaluation for the risk of falls were associated with the LI, in their respective HAs, weak correlations were found, without statistical significance (Table 1). In the same way, statistically-significant differences were not  (Tables 2 and 3).
These results show a leveling below the scores achieved by the older adults in the clinical tests, in which it is possible to make an inverse analogy to the principle of trainability (15) : the more the older adult's capacities are reduced, the more difficult it is to reduce them even further. Another similarity of these physical conditions is reported when one makes a parallel between sedentary aging and the physiological manifestation in the adaptation to the environment of microgravity (16) , both introduced by a situation of hypokinesia, leading to a condition of reduced mechanical overload.
The scarcity of correlated works in the literature, using residential time in homes for the aged and/or the chronological age as a manipulated variable, restricts a broader discussion of the results found here. However, when these results are dichotomized for those who fall and those who do not fall, one can note significant differences (Table 4). These findings are in accordance with similar experiments found in the literature (10,17) , which -through the BBS and TUG -differentiated the older adults as those who fall or do not fall. The results from the study showed that, in the population studied here, the LI and age are not different between those who fall and those who do not fall, showing a homogeneity of these variables, irrespective of the risk of falls (Table 4).
Although the TUG is used and recognized internationally as an instrument for tracking the risk of falls and its performance is associated with the history of these events, studies indicate that its capacity for predicting this phenomenon remains limited, as does its cutoff point which remains without variations between populations, thus making it a complementary measurement instrument which must be associated with Batista WO, Alves Junior ED, Porto F, Pereira FD, Santana RF, Gurgel JL.
other tests (10,18) . In addition to this, the standardization of the test's conditions combined with a control of confounding factors (age, sex and comorbidities) could provide better information regarding the predictive value for falls in older adults (10,18) .
In the same way, considering the clinical evaluation tests, the BBS, on its own, cannot definitively predict the risk of a fall, and no cutoff point was identified in in the present study, in which the above-mentioned test was complemented with the TUG (7) and the static stabilometry (8) .
When this study's subjects were subjected to Furthermore, older adults and young people differ regarding the geometrical limits of their base of support regarding displacement from the COP (20) . There is, however, evidence (21)(22)(23) corroborating this study's discoveries, in which the older adults reduce speed and area in order to maintain the efficiency of the task of postural control. This may be related to the fact that, simply, the older adults cannot sway more and with greater speed, so as to be able to maintain a greater margin within the limits of stability.
Another possible explanation for these results is that the cognitive functions may influence the motor task and the risk of falls (21) . Authors emphasize that there is an interdependence between postural and cognitive tasks.
They suggest that postural control and cognition need common resources and that the inconsistencies in the data and differences in the experimental designs make a broader understanding of the specific mechanisms of posture, cognition and double task difficult. In this scenario, one can include the fear of falling and depression, whose symptoms are a potential predictor of falls in institutionalized older adults (21)(22) .
The literature warns (23) that the onus of the cerebral abnormalities is correlated significantly with the decline in the control of balance. The signs of the aging of the brain, therefore, are concomitant with the degradation of the cognitive function and the reported age. This fact influences the processing speed of the postural control systems, which influence the integration of information from the sensory motor system in order to produce an action in the appropriate time and with the necessary precision for this function.
In the comparisons established within the groups and within the subgroups (Tables 2 and 3  showed there to be a significant correlation of the data with future falls, differentiating the older adults through stabilometry in the: older adults who fall, older adults who do not fall, and those who fall recurrently (24)(25) .
These institutionalized older adults demonstrate the use of different strategies for maintaining postural balance, reducing their sway from the COP in order to dimension their postural control with greater efficiency, suggesting that the motor aspects do not only influence the parameters of postural control, but also the cognitive functions (23) .

Conclusions
In