INTRODUCTION

Population aging= and the needs of long-lived elderly people (80 years and over) have become a growing concern for managers, professionals, and academics. If, on the one hand, long-lived elderly individuals are an increasing number of successful survivors, on the other hand, the number of sick and dependent people who r= each advanced old age is also increasing⁽<= sup>1).

In this context,= the adaptation of health systems to meet the growing number of elderly people w= ho access health services becomes a challenge, which requires the design of lo= ng-term, specific, and complex health care⁽2). Data from the Brazilian public health system, called Sistema Único de Sa= úde (SUS, Unified Health System), indicate t= hat, in 2016, 24.9% of hospitalized people were 60 years of age or older and 14.= 2% were over 70 years of age⁽³⁾.

In the gerontolo= gical literature, hospitalization in old age is classified as a negative health outcome⁽⁴⁾, as it brings together a set of factors that lead to worsening health conditions, associated with predisposing conditions^(5=
), such as age and sex; sociodemographic or enabling conditions, such as educa= tional level, family income, and place of residence⁽⁶⁾; and health need= s, including chronic diseases, needing assistance on activities of daily livin= g, cognitive decline, and frailty^(4-8 .

Data from the E= studo Longitudinal da Saúde dos Idosos Brasileiros (ELSI-Brasil, Longitudinal Study of the Health of Elderly <= span class=3DGramE>Brazilians)⁽⁹⁾, conducted with 9= ,389 participants aged 50 years and over, observed that the prevalence of hospitalization in the last 12 months was 10.2%, being associated with loca= tion (rural area) and region of Brazil (Midwestern and Northern regions). Of the factors associated with health needs, hospitalization was more evident among adults and the elderly who had limitations in activities of daily living and chronic diseases, in order of relevance: stroke, cardiovascular disease, cancer, diabetes, depression, and hypertension.

In advanced old = age, some of these factors were also documented, with greater emphasis on the health profile and health needs of this population. The author= ⁽¹⁰⁾, when longitudinally investigating 861 elderly Germans aged between 85 and 1= 00 years, observed that there was a higher prevalence of hospitalization in the last six months among the elderly with a higher prevalence of depressive symptoms, functional dependence, chronic diseases, and greater social netwo= rk. In this study, social networks moderated the relationship between functional decline and hospitalization, indicating that the greater the functional decline, the greater the contact networks and the prevalence of hospitalization. As in the study by author⁽<= /span>¹¹⁾, no associations were observed between hospitalization and age, sex, income,= and educational level, which may be related to equ= ity in access in the investigated countries or to the effect of selectivity attributable to advanced old age. <= /span>
Brazilian studies on the subj= ect are scarce. It is the fastest-growing population in terms of proportion and= , at the same time, with the greatest number of health needs= ⁽¹⁾. Knowing the life, health, and hospitalization conditions of advanced old age can help prevention and management measures in the context of health care a= nd gerontological nursing. In this context, the guiding question of this study was: What are the sociodemographic and health conditions associated with the hospitalization of long-lived elderly Brazilians, in order to guide prevent= ion and case management actions? Thus, the aim of the present study was to investigate the factors associated with hospitalization of elderly people a= ged 80 years and over, residing in the Federal District, who reported having be= en hospitalized at least once in the year prior to the survey/assessment.

METHOD

Study location, desig= n, and period

= This is an observational, cross-sectional, descriptive, and quantitative study. = It was conducted in the city of Brasília/DF between the 2016 and 2018, in the Administrative Region of Águas Claras where the Universidade Católica de Brasília (UCB, Catholic University of Brasília) is located. Within t= he scope of the UCB, the research was conducted in three locations: the UCB University Hospital, for clinical, cognitive, and self-reported health assessment; the Physical Assessment and Training Laboratory for functional = and physical assessment; and the Laboratory for Clinical Analysis. These centers provide care to the population of Brasília, which is divided into 31 administrative regions, with a population of 2,570,160 inhabitants, 7.69% of whom are elderly⁽¹²⁾. The = Human Development Index (HDI) is 0.824, the highest in the country according to t= he Ranking of the Brazilian Institute of Geography and Sta= tistics⁽¹²⁾.

Sample and inclusion = and exclusion criteria

A total of 208 elderly people aged 80 years or old= er agreed to participate in this study. The sample was obtained by convenience= and recruited at the Geriatrics outpatient clinic of the UCB University Hospita= l, which has an outpatient care program for the elderly. The choice of the convenien= ce sample was due to the possibility of selecting elderly people who could more easily access the study and because it is a quick method for data collectio= n. The inclusion criteria were: being 80 years of age or older, residing in the Federal District or nearby, having been hospitalized in the last year, consenting to participate in the study, and being available to conduct soci= al, physical, and laboratory assessments. In the case of elderly individuals wi= th cognitive decline, an additional criterion was used: having a companion or family caregiver to assist in the responses. The exclusion criteria were: b= eing bedridden, having severe hearing loss, uncontrolled psychiatric morbidities, and advanced-stage dementia syndrome.

Data collection<= /p>

Collection instruments

The assessment= instruments used were: questionnaires with sociodemographic information and access to health services; criteria for the frailty phenotype according to the authors⁽¹³⁾ adapted to Brazil(14); global cognitive assessment⁽¹⁵⁾; basic⁽¹⁶⁾ and instrumental activities of daily living⁽¹⁷⁾; self-reported chron= ic diseases; and polypharmacy (taking five or more medications).

Sociodemograph= ic information comprised: age (quantified in years); educational level (quantified in year= s of formal education); sex (male or female); marital status (single, widowed, married, divorced, separated); family income (quantified by the sum of the monthly earnings of the entire family per month); and household setup (livi= ng alone, with spouse and sons/daughters, with sons/daughters, grandchildren, = and great-grandchildren, and with other relatives and family members).

The global cognitive assessment was performed using the Mini Mental State Examination (MMSE), composed of 30 questions that assess orientation in time and space, episodic memory, immediate repetition, praxis, visuospatial functions, and = language⁽¹⁵⁾. To classify the el= derly with cognitive decline, a cutoff point was used in the MMSE: 17 points for = the illiterate elderly; 22 points for elderly individuals with 1 to 4 years of education; = 24 for those with 5 to 8 years of education; and 26 points for those with 9 ye= ars and over⁽¹⁴⁾. Regarding the variables related to functional performance, the following were used: absen= ce or presence of difficulties in Basic Activities of Daily Living (BADL) rela= ted to self-care (using the bathroom, eating, moving around the house, walking, bathing, doing personal hygiene)⁽16); and in the Instrumental Activities of Daily Living (IADL), which are: making food, taking transport, using the phone, shopping⁽¹⁷⁾. Elderly i= ndividuals with BADL or IADL limitations were classified as “with limitation” and those without any limitation as “independent.”

To assess frai= lty, a set of anthropometric and functional measures was used to compose the five criteria proposed by the authors⁽13): unintentional weight loss, fatigue, weakness, slowness, and low rate of ene= rgy expenditure. These criteria were assessed as follows: 1) Unintentional weig= ht loss was assessed by the question: “Have you lost weight unintentionally in= the last 12 months?”. Affirmative responses with a loss of more than 4.5 kg or = 5% of body weight met this study’s criteria. 2) Fatigue was assessed with the questions “I felt I had to make an effort to do everyday tasks” and “I coul= dn’t get my stuff done” extracted from the Center for Epidemiological Studies–Depression (CES-D) Scale⁽18), having filled in the fatigue criterion the elderly who answered always or m= ost of the time to any of the two questions presented. 3) Muscle strength was assessed using the Handgrip Strength test (HS), verified by the mean of thr= ee consecutive measurements with one-minute intervals to return the grip stren= gth of the dominant hand, measured in kilogram-force (kgf<= /span>) on a hydraulic dynamometer (Jamar, Model 5030J1, Lafayette Instruments Inc.= ). In this criterion, the cut-off points proposed by the a= uthors⁽¹³⁾ and adjusted for Brazil according to sex and Body Mass Index (BMI =3D weight/height²)⁽¹⁴⁾. 4) Gait speed was obtained by the mean time spent in three consecutive measurements on a 4.6m path. To fulfill the slow= ness criterion, the cutoff points proposed by the authors(¹³⁾ and adapted to Brazil, adjusted for sex and height, were adopted. 5) The participant’s level of physical activity was assessed using the Minnesota Leisure Time Activities Questionnaire, with weekly energy expenditure cut-o= ff points <270 kcal for women and < 383 kcal for men. Elderly people were considered frail if they met 4 to 5 of the aforementioned criteria, pre-fra= il if they met 1 to 3 criteria, and not frail any of the criteria.<= /span>

In the health questionnaire, the number of self-reported chronic diseases was collected [= 0 to 1; 2 or + (heart disease, hypertension, stroke, cancer, rheumatoid arthriti= s, lung disease, depression, and osteoporosis)] and polypharmacy (less than fi= ve drugs and more than five drugs taken). Regarding the use and access to health services, the questions were analyzed: “About last year, did you need to be hospitalized?”, “Did you go to a medical appointment last year? If yes, how many?”, “What type of health service are you looking for?” and “Do you have= any difficulties in accessing the health service?” The answers to these questio= ns were categorized and the occurrence of hospitalization (yes or no), number = of medical consultations, type of service (SUS, health insurance or private medical appointment) and the prevalence of difficulties in accessing servic= es were obtained (answered with yes or no).

Data collection procedures

Data collectio= n was conducted in three stages: Initially, the elderly and their companions were invited to participate in the study and instructed on the objectives, procedures, guarantee of anonymity, and voluntary participation, as describ= ed and detailed in the Informed Consent Term (ICT). Those who wished to participate in the study signed the ICT and were referred to an individual interview to collect their sociodemographic data.

Subsequently, = the participants were referred to a geriatrician for clinical and health assess= ment purposes. At the end of the consultation, each elderly person received a voucher for blood collection and were sent to the university's physical assessment laboratory to conduct measurements of handgrip strength, gait, b= one densitometry, and other anthropometric data. After these steps, the participants were invited to see the doctor again for further assessments, = referrals, and feedback regarding their health status. All answers electronically coll= ected using Google Forms and sent to the database spreadsheet of the study.<= /o:p>

Analysis of Results

= Data were analyzed quantitatively, through comparisons of means between groups, performed by non-parametric tests (Mann Whitney or Kruskal-Wallis for continuous quantitative variables and Chi Square or Fischer for categorical quantitative variables). The variables with a p value below 0.10 in = the bivariate associations were ranked to compose a multiple logistic regression model. The final model was built using Wald's Forward Stepwise method, with adjustments for sociodemographic variables, functional dependence, and use = and access to health services, composed of variables with p<0.05. The dependent variable was categorized into 0 and 1, with 0 corresponding to elderly people who were not hospitalized in the last year and 1 to those who were hospitalized in the last year. For all analyses, a significance level = of 5% was used, or p<0.05.

Ethical aspect= s

The project was approved by the Research Ethics Committee of the UCB, under opinion nº 1.29= 0.368 and Certificado de Apresen= tação para Apreciação Ética (CAAE, Certificate of Presentation for Ethical Appreciation) No. 50075215.2.0000.0029, in accordance with the requirements of Resolution 466= /12 of the Brazilian National Council of Health, which provides for regulatory norms and guidelines for research involving human beings.=

RESULTS=

= Of the 208 elderly people assessed, 26.9% of the sample (n=3D44) reported being hospitalized in the previous year. In the bivariate analyses, the long-lived elderly who were hospitalized in = the last year were even longer lived, had lower educational levels, lived predominantly with their sons/daughters, grandchildren, and great-grandchildren; had a higher prevalence of cognitive decline, and two = of the frailty criteria (weight loss and decreased handgrip strength) (Table 1= ).

Although there were no significant differences regarding functional performance, polypharm= acy and chronic diseases, hospitalization in the previous year was associated w= ith a greater number of medical appointments (Table 2). In the final multiple logistic regression model, after using Wald's Forward Stepwise method, educ= ational level, frailty (weight loss), and number of medical appointments in the pre= vious year remained significant (Table 3).

Table 1 - Sociodemographic characterization of long-lived elderly people according to hospitalization in the previous year, Federal District - Brazil, 2016-2018.

Variables	Hospitalized n=3D44	Not hospitalized n=3D164	P value<= /span>
Age	M (SD)	85.02 (+ 4.59)	83.45 (+4.18)	0.003⁺
Family inc= ome		3.730.05 (+4.190.15)	4.545.74 (+4.872.14)	0.086⁺
= ; Sex	(%) Female <= span lang=3DEN-US style=3D'font-family:"Times New Roman",serif;mso-fareast-fon= t-family: "Times New Roman";color:black;mso-ansi-language:EN-US'>Male	61.4 38.6	65 35	0.653^*
Educational level	Illiterate	44.2	20.6	0.000^*
= ;	1 \|- 4 years	27.9	16.9
= ;	4 \|- 8 years	20.9	41.2
= ;	8 years or over	7.0	21.2
Household setup	Lives alone	29.8	24.9	0.496^*
	Lives with spouse and sons/daughters	23.4	20.7	0.691^*
	Lives with spouse, grandchildren, or great-grandchildren	4.3	1.8	0.318^*
	Lives with sons/daughters, grandchildren, or great-grandchildren	19.1	6.5	0.008^*
Marital st= atus	Married	9.1	8.0	0.516^*
	Single	34.1	38
	Separated	4.5	11
	Widowed	52.3	42.9

+ Mann Whitney test; *Chi-square test – p>0.05.
Source: The authors

Table 2 - Characterization of functional performance, hea= lth status, and use and access to health services by long-lived elderly people, according to hospitalization in the previous year, Federal District - Brazi= l, 2016-2018.

Variables		Hospitalized n=3D44 (%)	Not hospitalized n=3D164 (%)	P value
Chronic diseases	None One or two Three or over	2.5 50 47.5	3.6 46.4 50	0.837^*
Polypharmacy (% sim)		56.4	48.6	0.387^*
MMSE (% cognit= ive decline)		66.7	44.6	0.012^*
BADL (% difficulties)		34.1	28	0.429^*
IADL (%difficu= lties)		59.5	45.0	0.350^*
Frailty Phenotype (% frail)	Weight loss	40	16.4	0.002^*
= ;	Physical activity<= /p>	95.5	80.4	0.017^*
= ;	Fatigue	16.2	19.7	0.630^*
= ;	Handgrip strength<= /p>	16.7	13.5	0.654^*
= ;	Gait	30	15.6	0.069^*
Difficulty using health services (% of affirmative answers)		59.6	56.4	0.695^*
Number of medical appointments in the previous year (Mean and Standard Deviation)		6.0 (+5.17)	3.75 (+3.41)	0.003⁺
Type of service (%)	SUS Health insurance Private doctor	68.1 21.3 10.6	60.4 24.9 14.8	0.317^*

+ = Mann Whitney test; *Chi-square test – p>0.05.

Source: The authors

Table 3 – Multiple Logistic Regression – Final Model – Forward Stepwise Method

Final Model – Hospitalizati= on in long-lived elderly people*	B (SE)	OR	CI (95%)<= /p>	p-value
Educational le= vel (4 years \|- +)	-1.56 (0.44)	0.20	0.08-0.49=	0.000
Weight loss (a= ffirmative answers)	1.12 (0.45)	3.08	1.26-7.53=	0.014
Medical appoin= tments	0.11 (0.05)	1.12	1.01-1.24=	0.024
Constant (B0)<= o:p>	-1.52 (0.39)			0.000

* Final model: Chi-square of the model with 29.93, degrees of freedom: 3, p.0.000. R2=3D0.249. The reference condition for “Educational level” was the category “0 |- 4 years of education”. For “Weight loss,” the category “affi= rmative answers.” The “Medical Appointments” variable was analyzed as continuous. B= =3D beta. SE =3D Standard error. OR =3D Odds ratio. CI =3D Confidence interval.=

Source: The authors

DISCUSSION

In the present study, there were bivariate associations between hospitalization in the previous year and age, educational level, household setup, cognitive decline, frailty criteria (weight loss and handgrip strength), an= d a greater number of medical appointments. However, after adjustments according to the multiple logistic regression model, only educational level, frailty (weight loss), and number of medical appointments in the previous year remained associated with hospitalization. These findings indicate that both enabling conditions such as educational level, = as well as health needs and frailty (weight loss criterion), are associated wi= th hospitalization in long-lived elderly Brazilians^(<=
/sup>⁶⁾.

Regarding the association between educational leve= l and hospitalization, the findings confirm previous studies that indicate a high= er prevalence of hospitalization among the elderly with lower educational levels⁽^6-8). This is probably be= cause elderly individuals with lower educational levels tend to have less access = to goods and services, lower income, inhabit regions with greater social vulnerability⁽¹⁹⁾ and have diffe= rences in lifestyle-related variables⁽²⁾. However, international studie= s conducted with long-lived elderly people did not find associations between sociodemographic conditions and hospitalization, which indicates that conditions of equity in access to health can moderate these associations⁽^10-11). In Brazil, with a population or generation of long-lived elderly people exp= osed to social inequalities, it remains to be seen how these variables interact = with each other throughout their lives.

Regarding the association between hospitalization = and frailty, multiple studies highlight that hospitalization is an outcome of <= span class=3DGramE>frailty⁽^20-23), associated with = worse health status and that can worsen depending on the clinical status and after hospital discharge⁽²⁰⁾. Associations between the BMI, frailty syndrome, and functional worsening after hospital discharge were observed i= n a Brazilian study with hospitalized elderly people^(<=
/sup>²⁰⁾. In the literature, the criterion of frailty for weight loss is associated w= ith depression, loss of muscle and bone mass, decreased inflammatory response, = and general health status⁽²¹⁾.= When frailty and weight loss coexist, they are associated with increased mortality⁽^21-22). For this reaso= n, measures to prevent nutritional risk in healthy and long-lived elderly peop= le are necessary, given that weight loss represents one of the main causes of pathophysiology of frailty⁽^21).

It is possible that the group of individuals who w= ere hospitalized in the previous year, despite having a similar prevalence in the number of chronic diseases= in relation to the non-hospitalized long-lived elderly, is clinically more complex. For this reason, a possible consequence would be a hyper-attendanc= e of elderly people to hospital care and complex health care services, which cou= ld be minimized if there was a long-term care network compatible with their ne= eds, including rehabilitation, prioritization of chronic health conditions, and = care management⁽^23-24).

In this study, although it was not possible to det= ect the causes of hospitalization, the findings highlight the need to implement health policies aimed at this population, focusing on the qualification of health professionals, the design of educational interventions for the family and community, and the use of assessment protocols to the establishment of clinical guidelines for the care of hospitalized elderly people⁽²⁴⁾.

Despite the similar prevalence in the functional performance of BADL and IADL in this study, other studies emphasize that hospitalization is associated with a greater functional dependence as it increases the demands for care⁽^2=
5). It is possible that, due to the high prevalence of functional limitations, = it was not possible to observe a significant relationship between functional performance and hospitalization. In this sense, further studies with more detailed measures of functionality in long-lived elderly people are justifi= ed.

According to the data presented, a considerable nu= mber of these elderly people referred to themselves as “SUS-dependent.” In this context, the challenge of assistance to advanced old-age people becomes even greater. Articulating the service network with assistance and care demands = has been the target of health care policies for the elderly, through the implementation of lines of care and qualification of health teams. More investments are needed and more programs must be designed in order to integ= rate a network of social assistance and comprehensive care for the elderly.

Th= is study has the following limitations: a convenience sample obtained in an outpatient setting and restricted to the context of the Administrative Regi= on of Ág= uas Claras (B= razil) where the UCB is located, with its own cultural diversity as it is a metropolitan region with many migrants from different locations in Brazil; = in some cases, the family member, caregiver, or respondent may have exacerbated the situations of dependency of the elderly, which can configure an importa= nt response bias in terms of data related to functional dependency and to the = use and access to health services; a large sample loss while conducting assessm= ents was related to the frailty criterion, due to the significant number of wheelchair users and the high prevalence of disabilities in the investigated sample, individuals who were unable to perform gait tests; and, finally, the research protocol did not include questions about the quality of the interaction between the members of the household.<= /span>

Re= garding the contributions to nursing, the present study contributes as a reflection= on the implementation of prevention and case management measures, raising the = need to track health needs and sociodemographic conditions associated with a hig= her prevalence of hospitalization in long-lived elderly people. The assessment = of functional performance and frailty criteria (especially nutritional risk), associated with greater articulation and effectiveness in primary and secon= dary health care, can contribute to the development of a more resolute health sy= stem. In the context of gerontological nursing, it is necessary to train and sens= itize the nursing team for the proper as= sessment, screening, and management of cases, with a view to devel= oping a health system that is integrated to the biopsychosocial needs of long-liv= ed elderly patients.

CLOSING REMARKS

By analyzing the factors associated with the hospitalization of long-lived elderly people living in the Federal District= and hospitalized in the previous year, a sample with lower educational levels, higher nutritional risk, and a higher number of medical consultations in the previous year is identified. Taken together, these findings support the need to adopt case management measures, as well as prevention a= nd management of frailty in the long-lived elderly, to guide proper care and prevent hospitalization.

ACKNOWLEDGMENTS

The authors would like to thank the Coordenação de Aperfei= çoamento de Pessoal de Nível Superior (CAPES, Coordination for the Improvement of Higher Education Personnel) for funding the research project entitled: “Patterns of physical, cognitive, and psychosocial aging in elderly people aged 80 years and over living in different contexts” providing an opportunity to create an Network of Academic Cooperation in the field of Gerontology. The Postgraduate Programs in Gerontology of the Universidade Estadual<= /span> de Campinas (UNICAMP, State University of Campinas), the UCB and the Universidade de Passo<= /span> Fundo (UPF, University of Passo Fundo) participated in this inter-institutional cooperation.

REFERENCES

1.&n= bsp; Neri AL, Borim FSA, Assumpção D. Octogenários em Campinas: Dados do Fibra 80+. Campinas: Editora Alínea, 2019.

2.&n= bsp; Gutierrez BAO, Salmazo-Silva H, Guimar= ães C, Campino AC. Impacto econômico da doença de Alzheimer no Brasil: é possív= el melhorar a assistência e reduzir custos? Ciênc. saúde coletiva [Internet]. 2014 [cited&nbs= p;2019 Dec 30]; 19(11): 4479-86. A= vailable from: http://dx.doi.org/10.1590/1413-81232= 0141911.03562013 .

3.&n= bsp; DATASUS [Internet]. Brasília, DF: Ministério da Saúde. 2008- . Morbidade Hospitalar do Sistema Único de Saúde - Brasil 2016.

4.&n= bsp; Heldmann P, Werner C, Belala N. et= al. Early inpatient rehabilitation for acutely hospitalized older patients: a systematic review= of outcome measures. BM= C Geriatr [Internet]. 2019 [cited 2019 Dec 30]; 19, 189. Available from: http://dx.do= i.org/10.1186/s12877-019-1201-4

5.&n= bsp; Andersen R, Newman JF. Societal and individual determinants of medic= al care utilization in the United States. Milbank Mem Fund Q Health So= c. [Internet]. 1973 [cited 2019 Dec 30];51(1):95-124. Available from: http:= //dx.doi.org/10.1111/j.1468-0009.2005.00428= .x <= /span>

6.&n= bsp; Almeida APSC, Nunes BP, Duro SMS, Facchini LA. Socioeconomic determinants of access to health services among older adults: a systematic review. Rev Saude Publica [Intern= et] 2017 [cited 2019 Dec 30];15;51:50. Available from: http://doi.org/10.1590/S1518-8787.2017051006661

7.&n= bsp; Silva AMM, Mambrini JVM, Peixoto SV, Malta DC, Lima-Costa MF. Use of health service= s by Brazilian older adults with and without functional limitation. Rev= Saúde Publica. [Internet]. 2017 [cited 2019 Dec 20];51 = Supl 1:5s. Available from: https://doi.= org/10.1590/s1518-8787.2017051000243

8.&n= bsp; Phelan EA, Borson S, Grothaus L, Balch S, Larson EB. Association of incident dementia with hospitalizations. JAMA [Int= ernet]. 2012 [cited 2019 Nov 30];= 307(2): 165-72. Available from:= http://dx.doi.org/10.1001/jama.2011.1964

9.&n= bsp; Silva AMM, Mambrini JVM, Souza Junior PRB, Andrade FB, Lima-Cost= a MF. Hospitalizations among older adults: results from ELSI-Brazil. [Inte= rnet]. 2018 [cited 2019 Dec30]; 52(Suppl 2): 3s. <= span class=3DSpellE>from: http://dx.doi.org/10.11606/s1518-8787.2018052000639

10.&= nbsp; Ha= jek A, Brettschneider C, Eisele M, Kaduszkiewickz H= , Mamone S. et al. Correlates of hospitalization among = the oldest old: results of the AgeCoDeQualiDe prosp= ective cohort study. Aging Clin Exp Res [Internet]. 2019 [cited Dec 30]; Epub ahead of print. = Available= from: http://dx.doi.org/10.1007/s40520-019-01315-2

11.&= nbsp; Walter‐Ginzburg A, Chetrit A, Medina C, Blumstein T, Gindin J, Modan B. Physic= ian Visits, Emergency Room Utilization, and Overnight Hospitalization in the Old‐Old in Israel: The Cross‐Sectional and Longitudinal Aging S= tudy (CALAS). JAGS [Internet]. 2001 [cited 2019 Nov 28]; 49: 549-56. Available f= rom: http://dx.doi.org/10.1046/j.1532-5415.2001.49112.x.

12.&= nbsp; Instituto Brasileiro de Geografia e Estatística (IBGE). Sinopse do Censo Demográfico 2010, Rio de Janeiro: 2010= .

13.<= span style=3D'font:7.0pt "Times New Roman"'> Fried L, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sc= i. [Internet]. 2001 [cited 2019 Nov 25]; 56:M146-56. Available from: https://doi.org/10.1093/gerona/56.3.M146

14.<= span style=3D'font:7.0pt "Times New Roman"'> = Neri AL, Yassuda MS, Araújo LF, Eulálio MC, Cabral B= E, Araújo LF, Siqueira ME, Santos GA. Metodologia e perfil sociodemográfico, cognitivo e de fragilidade de idosos comunitários de sete cidades brasileir= as: Estudo FIBRA. CSP [Interne= t] 2013 [cited 2019 Nov 25]; 29(= 4): 778-92. Available from: http://dx.doi.org/10.1590/S0102-311X2013000400015<= /span>

15.<= span style=3D'font:7.0pt "Times New Roman"'> Brucki SMD, Nitrini R, Caramelli P, Bertolucci PHF, Okamoto IH. Sugestões para o uso do Miniexame do Estado Mental no Brasil. = Arq. Neuro-Psiquiatr.&nbs= p;[Internet] 2003 [cited 2019 Nov 20]; 61(3B): 777-81. Available from: http://dx.doi.org/10.1590/S0004-282X2003000500014<= /span>

16.<= span style=3D'font:7.0pt "Times New Roman"'> Katz S<= span lang=3DEN-US style=3D'font-size:12.0pt;font-family:"Times New Roman",serif; color:black;background:white;mso-ansi-language:EN-US'>, Ford AB, Moskowitz RW, Jackson BA, Jaffe MW= , Cleveland MA. Studies of i= llness in the aged. The index of ADL: A standardized measure of biological and psychosocial function. JAMA [Internet]. 1963 [cited 2019 Nov 10]; 185: 914–19. Available from: = http://dx.doi.org/10.1001/jama.1963.03060120024016=

17.&= nbsp; Lawton MP, Brody P. Assessment of older people: Self maintaining and instrumental activities of daily living. The Gerontologist [Internet] 1969 [cited 2019 Nov 05]; 9:179–86. Available from: http://www.eurohex.eu/bibliography/pdf/Lawton_Gerontol_1969-1502121986/Law= ton_Gerontol_1969.pdf

18.&= nbsp; Radlof, LS. The CES-D scale: a self-report depression scale for research in the general population. Applied psychological measurement [Internet].1977 [cited 2019 Nov 10]; 1:385-401. Available from: http://dx.do= i.org/10.1177/014662167700100306

19.&= nbsp; Pe= reira Nunes B, Uhlmann Soares M, Soares Wachs L, Moraes Volz P, de Oliveira Saes M, Manjourany Silva Duro S, Thumé E, Facchini LA. Hospitalização em idosos: assoc= iação com multimorbidade, atenção básica e plano de s= aúde. Revista de Saúde Pública [Internet]. 2017 [cited 2019 Nov 05];51:1-10. Available from: http://www.redalyc.org/articulo.oa?id=3D67249591042

20.&= nbsp; Carvalho TC, Valle AP, Jacinto AF, Mayoral VFS, Boas PJFV. Impacto da hospitalização na funcionalidade de idosos: estudo de coorte. Rev. bras. geriatr. gerontol. [Internet]. 2018 Apr [cited 2019 Dec 30];21(2):134-142. Available from: http://dx.doi.org/10.1590/1981-22562018021.170143<= /span>. =

21.&= nbsp; Fougère B, Morley JE. Weight loss is a major cause of frailty. J Nutr Health Aging [Internet]. 2017 [cited 2019 Nov 01]; 21: 933. Ava= ilable from: https://doi.or= g/10.1007/s12603-017-0971-7

22.&= nbsp; Bo= rges MK, Canevelli M, Cesari M, Aprahamian I. Frailty as a predictor of cognitive disorders: A systematic review and meta-analysis. Front Med (Lausanne) [Int= ernet]. 2019 [cited 2019 Nov 3]; 6:26. Available from: http://dx.doi.org/10.3389/fmed.2019.00026

23.&= nbsp; M<= span style=3D'background:white'>arcoux V, Chouinard M-C, Di= adiou F, Dufour I, Hudon C. Screening tools to identi= fy patients with complex health needs at risk of high use of health care servi= ces: A scoping review. PLoS ONE [Internet]. 2017 [ci= ted 2019 Nov 01]; 12(11): e0188663. Available from: http://dx.doi.org/10.1371/journal.pone.0188663.eCo= llection 2017.

24.&= nbsp; Organização Mundial de Saúde (OMS). Resumo. Relatório Mundial de Envelhecimento e Saúde. [Internet] 2015 [cited 2019 Nov 01]. Available fro= m: https://sbgg.org.br//wp-content/uploads/2015/10/OMS-ENVELHECIMEN= TO-2015-port.pdf

25.&= nbsp; Giacomin= KC, Duarte YAO, Camarano AA, Nunes DP, Fernandes D. Cuidados e limitações funcionais em atividades cotidianas – ELSI-Brasil. Rev Saude Publica. [In= ternet]. 2018 [cited 2019 Nov 3]; 52 (Supl 2):9s. Availa= ble from: https://doi.org/10.11606/S1518-8787.2018052000650

Submission= := 2021-12-28=

Approvall= := 2022-02-08