Pre-frailty and frailty of elderly residents in a municipality with a low Human Development Index

OBJECTIVE: to identify the prevalence of the factors associated with pre-frailty and frailty of elderly residents in a municipality with a low Human Development Index METHOD: Cross-sectional study with a populational and household framework conducted with 316 elderly people. Frailty was determined from the presence of three or more of the following factors: (i) self-reported unintentional weight loss; (ii) lack of strength and energy; (iii) weakness; (iv) slowness; (v) low level of physical activity. The association between frailty and socio-demographic, behavioral and health factors was measured using the multinomial logistic regression technique. RESULTS: The prevalence of pre-frailty and frailty was 58.7% and 23.8%, respectively. The adjusted regression model showed that the state of pre-frailty was associated with gender, age group and BMI, and frailty was associated with gender, age group, hospitalization, functional capacity, and self-perceived health. CONCLUSION: The evidence presented in this study demonstrates more variables associated with the frailty condition, reinforcing the concept of a multifactorial clinical syndrome that may result in the loss of functionality.


Introduction
The aging process brings changes, which together with the increased prevalence of chronic diseases can cause the onset of geriatric syndromes, among which the frailty syndrome is highlighted. Frailty includes factors of different orders, being characterized as a syndrome resulting from the loss of physiological reserves and adaptation to stressors, where the energy deficit, sarcopenia, decreases the muscle strength and tolerance to effort leading to an exacerbated decline in multiple systems that places the individual in a condition of greater vulnerability (1) .
Frailty has increasingly emerged as an important concept, both in the clinical care of older people, as well as in studies on aging. As a clinical syndrome, it is usually associated with an increased risk of adverse situations such as falls, disabilities, institutionalization, and death (2) . There is no defined scientific consensus regarding the term frailty, its definition and its indicators, nor how it can be identified or even evaluated. However, the majority of studies that deal with frailty generally  (4) .
In Brazil, the majority of municipalities experience an increase in the number of elderly in the population each year. They present poor health indicators and this may contribute to the development of early frailty resulting in an unfavorable prognosis with dependence, hospitalization and severe complications in the subsequent years (4) . In addition, research into the identification of frailty in the elderly is at an early stage, therefore there are few population-based studies that present factors involved in the process of frailty in the Brazilian population.
When studying frailty it is essential to consider the context in which this process takes place. Since Brazil is a country of continental dimensions it is essential to understand how frailty develops in regions marked by pronounced social inequalities. Thus, the aim of this study was to identify the prevalence and the factors associated with pre-frailty and frailty of elderly residents in a community with a low human development index.
An ordinal variable with scores ranging from zero to five (0-5) was created from the sum of the points of all the components, with the following classification adopted: 0 points = not frail; 1-2 points = pre-frail; ≥3 points = frail (1) . All the individuals who responded to only 3 components and that were classified as frail were considered. Individuals who answered at least 4 components for the classification of frailty were considered eligible for the other classifications (9) . Thus, 286 subjects, classified according to frailty phenotype, were included in the analysis.  (11) and Instrumental Activities of Daily Living (IADL), using the Lawton scale (12) ).
Amount of medications used, Self-perceived health, and Cognitive status (assessed through the Mini-Mental State Examination questionnaire. using the validated a modified version (13) to verify the reliability of the answers -score >12 = not compromised and score ≤12 = compromised). When the cognitive status score was not reached, the informant was asked to responded to another instrument, which was the Functional Activities questionnaire of Pfeffer (14) , with

Results
The prevalence of pre-frailty and frailty of the elderly people living in the urban area of the municipality of Lafaiete Coutinho-BA was 57.8 and 23.8%, respectively. Table 1 shows the crude analysis of the independent variables that composed the socio-demographic, behavioral, and health condition factors and their association with frailty in elderly people.
Reis Júnior WM, Carneiro JAO, Coqueiro RS, Santos KT, Fernandes MH.   In the first step the variables of the sociodemographic block were analyzed, and the variables that remained associated with the frail stage were: female gender (p=0.031) and age group ≥80 years (p<0.001) and, with the pre-frail stage, also female gender (p=0.037) and age group ≥80 years (p=0.001).
In the second step the variables of the health conditions block were included and those that continued to be associated with the frail classification were: hospitalization in the previous year, one or more times (p=0.033), functional capacity for both dependent in the IADL (p<0.001), as well as in the BADL and IADL (p=0.007), and also self-perceived negative health (p=0.007). Only the underweight BMI variable (p=0.015) remained associated with the pre-frail classification.
The variables related to behavioral aspects, i.e., consumption of alcoholic beverages and tobacco use were not included in the adjusted model, as they did not present at least 20% statistical significance (p≤0.20).
frailty prevalence values were higher when compared to international studies (16) . The explanation for the wide variation in the prevalence of frailty in the elderly is probably related to socioeconomic differences among the elderly people studied, as well as the methodological differences in the use of some instruments that differ from those that compose the items used (16) .
In the present study women had a higher prevalence in both the pre-frail and frail categories, corroborating other studies that indicated that elderly women are frailer than elderly men, and presented a frailty prevalence of over 13% for the men and 17% for the women. Over three years, the women had twice the pre-frailty incidence rates compared to the men (17) .
In a study conducted in Taiwan, which estimated the prevalence of frailty and identified factors associated with frailty, from the data of the Survey of Health and Living Status of the Elderly, it was found that the prevalence of frailty increased with age and was higher in women (18) .
Longevity is another aspect that should be taken into consideration. A national study, which is part of the multicentric project Frailty in Elderly Brazilians (Fibra), analyzed prevalence characteristics and associated factors related to frailty and, of the sociodemographic variables included in the model, only age was associated, even when adjusted for the other variables, thus demonstrating the influence of the aging process on the emergence of frailty (19) .
Weight loss has been reported to be important information related to the prediction of frailty.
However, its participation in the phenomenon is still unclear, considering that weight loss does not precisely represent muscle loss, and a simultaneous increase in fat mass could mask the real estimate (20) .
The studies that have investigated the association between the BMI and frailty are still inconclusive.
Some have shown that pre-frail and frail elderly men presented lower BMI values and that the prevalence of frailty is slightly higher in normal and underweight men (20). Others, however, have reported an association of increased BMI values with the frail situation, mainly observed in frail elderly women (21) .
The findings of the present study were consistent with the results of the study in which five Latin American and Caribbean cities were investigated, including São Paulo, Brazil, which verified that higher BMI values for women were related to a greater probability of frailty (9) .
The observation of the association between dependence in performing instrumental and basic activities and frailty is presented in national and international studies, demonstrating the degree of injury that this impairment generates in the elderly, by directly limiting them in their autonomy, resulting in a decrease in the quality of life (1) . The data obtained are consistent with the literature in that, after adjustment, an association between the BADL, IADL and frailty was also found (18) .
However, it is critical to know how the process of functional disability takes place and its relationship to frailty, therefore important adverse factors are interrelated to the two events, such as fatigue, low levels of physical activity and decreased muscle strength, which have been suggested as predictors for functional disability (22) .
In this study, self-perceived health was associated with being frail, corroborating other findings (18) . There are suggestions in the literature that this issue is influenced by the life trajectory and the experiences of the elderly people and the way they deal with adverse situations (1) , and overcoming these situations has been sustained by the theory that proposes the relationship between resilience in the elderly and frailty (23) .
Some limitations of the study are that this research is characterized as a cross-sectional type study, in which there is no possibility of establishing cause and effect, and that some of the instruments used required subjective or self-reported information, which may lead to memory bias. Longitudinal studies and the use of more objective instruments are needed to make the inferences about the predictive indicators of frailty more robust.

Conclusion
The prevalence of pre-frailty and frailty was 58.7% and 23.8%, respectively. The evidence presented in this study showed that the state of pre-frailty was associated with gender, age group and BMI. While the state of frailty was associated with gender, age group, hospitalization, functional capacity and self-perceived health. There were more variables associated with the frailty condition, which reinforces the concept of a multifactorial clinical syndrome that may result in the loss of functionality. Furthermore, it is worth mentioning once again that the high prevalence of elderly people in the pre-frail condition constitutes Rev. Latino-Am. Enfermagem 2014 July-Aug.;22(4):654-61.
important information, considering the possibility of deterioration and becoming frail over a short period, which will result in more care and spending on the health of these elderly people.
Therefore, the results published on the subject should serve to guide and improve the healthcare policies for the elderly, as well as assist the professionals involved in their care, whose purpose should be to prevent deterioration and its consequences, and the development of the preliminary stages of frailty into an advanced situation.