Prevalence of Frailty Indicators in Community-dwelling Older Adults From Northeastern Brazil, 2009-2015

BACKGROUND: Frailty is a clinical syndrome, and its development is multifactorial and 2 dynamic. The clinical indicators (physical measures and self-report) that characterize the 3 syndrome tend to vary across studies. To determine the contributions of the indicators in the 4 determination of frailty it is important to obtain data about the variations that occur among the 5 levels of frailty. The aim of this study was to a) survey the prevalence of the frailty syndrome 6 and of the indicators that compose the frailty phenotype in community-dwelling older adults, 7 and b) to evaluate the contribution of each indicator in the determination frailty. 8 METHODS: Prevalence study carried out with 163 older adults who participated in two 9 assessments; the first one was performed in 2009 and after 70 months the second assessment 10 was conducted. Assessment of physical measurements was performed to constitute the frailty 11 phenotype (gait speed and handgrip strength) alongside self-report (fatigue, unintentional 12 weight loss, and physical activity), as proposed by Fried. We used the McNemar ’s test and 13 Pearson ’ s chi-square to analyze the differences between means and Multinomial Logistic 14 Regression values. RESULTS: There was an increase in the number of pre-frail older adults (from 47.85% to 16 65.03%) and frail ones (from 7.98% to 9.82%). The frailty indicators increased significantly (+ 17 8.6% for walking; + 6.8% for fatigue; + 6.8% for grip strength; + 1.2% for physical activity), 18 except the “ loss ” item (-3%). The indicators highest of in 19 2009 (OR (OR 28.74; 20 items had the highest chance 21 for frailty slow gait (OR = 23.64; p<0.001) muscle 22 weakness (OR explanatory models of frailty changed in both assessments. The evolution of frailty signals the 26 necessity for interventions to be carried out with older adults to delay the progress of declining 27 faculties that threaten their health. 28

explanatory models of frailty changed in both assessments. The evolution of frailty signals the 26 necessity for interventions to be carried out with older adults to delay the progress of declining 27 faculties that threaten their health. The notion of frailty can be considered new and there is no consensus on its concept in 32 the scientific community; however, the most common understanding is that frail older adults 33 are more vulnerable and that they are at a higher risk of unfavorable morbidities (acute or 34 chronic diseases, falls and injuries, disabilities) and mortality 1.2 . Although frailty has long been 35 considered a synonym of disability and comorbidity, the development of studies on this issue 36 has contributed to raising important questions within the field of older adults' health 3 .

37
Characterized as a clinical syndrome, frailty has been studied for its multifactorial  This study takes into account the model proposed by Fried et al. 4 , responsible for the 44 operationalization of five frailty criteria, which are: unintentional weight loss corresponding to 45 5% of body weight, fatigue, low grip strength, slow gait, and low rate of weekly caloric 46 expenditure in physical exercises and domestic activities 4,5 . 47 The criteria are used to indicate frailty in older adults and comprise the following 48 classifications: non-frail, pre-frail (presence of one or two indicators), and frail (presence of 49 three or more indicators). Even though it is not a "gold standard" for assessing the syndrome, the exposed model has been widely used in multicentric studies, with its validity and predictive 51 value being discussed in different health outcomes in the older adults' group 6,7 . 52 The in-depth study of the syndrome, capable of allowing the investigation of its 53 indicators, enables early discovery of the pre-frail state; consequently, it facilitates the treatment 54 of older adults in a frank process of aging 8,9 . However, there are few studies in Brazil that follow 55 the evolution of the syndrome and variations in the indicators that make up the phenotype 7 , 56 especially in non-institutionalized older adults. 57 It is considered that studying and monitoring variations in the conditions of frailty and 58 its predictor variables, through prevalence studies and with a longitudinal cut, in addition to 59 identifying factors related to declining faculties and mortality in older adults, make it possible 60 to estimate the relationship between protective factors that contribute to the stability and delay 61 in the progress of the syndrome 3,10 . Thus, the present study aimed at surveying the prevalence 62 of frailty syndrome and the indicators that compose the frailty phenotype in community-  However, from the total number of participants, it was only possible to collect data from 83 163 older adults. The number of losses was due to deaths (n = 66), physical and cognitive 84 disabilities (n = 37), addresses that were lost, changes and/or not located (n = 108) and refusals 85 (n = 29). We worked with a paired database, so that only the older adults from the first FIBRA 86 study who also participated in the second moment of assessment remained in the database.  Table 1.  Sociodemographic data 111 We used a structured questionnaire on sociodemographic conditions (gender, age, 112 marital status, housing arrangement-living alone, education, retirement) of older adults. 115 Frailty was assessed according to the phenotype proposed by Fried et al. 4 . The presence 116 of three or more of the five criteria characterizes frailty and the fulfillment of one or two of the 117 criteria means pre-frailty.

163
For all analyzes, a significance level of 5% was used.

166
The two assessments pointed out statistically significant differences in the prevalence 167 for classified frailty, in which there was an increase of approximately 18% pre-frail older adults.

168
Among the frailty indicators, only the gait speed resulted in statistically significant differences 169 (p = 0.040). There was a small decrease in weight loss; but the remaining indicators increased 170 (  were more likely to become frail (Table 4).    We perceived an inversion of the items that best explained frailty in 2015 when  (Table 6).    (Table 7).

246
More than 70% of the studied sample of the second assessment period was identified as 247 frail or pre-frail. The prevalence found for frail (9.8%) and pre-frail (65%) older adults is similar

307
The Brazilian Association of Palliative Care 32 defines self-perceived fatigue as a 308 generalized tiredness noticed by the older adults after performing their daily activities, 309 manifesting itself physically and psychologically as mental fatigue and indisposition. We 310 believe that more attention must be given to the assessment of fatigue in the older adults, since 311 this condition has not been a concern for most healthcare professionals who care for the older

366
This study comprises some limitations that initially pertain to the excessive number of 367 losses of older adults who were expected to be part of the second assessment. The losses were 368 related to deaths and comorbidities that installed physical and cognitive disabilities in some of 369 them, there were missing addresses (older adults who moved homes, addresses that were not 370 located), and there were also some rejections. The period between the first and the second 371 assessment is considered long since it totaled more than five years (70 months).

372
Another limitation that should be considered in the present study was the limited number 373 of variables for a longitudinal study of the possible outcomes associated with the syndrome.

374
The older adults who were part of the first assessment who had not scored above the cutoff     The authors have no potential conflicts of interest to disclose.