Abstract
Associations between neurodegenerative and cerebrovascular neuropathologies and frailty are not well understood, especially in diverse populations. This study investigated these associations in an admixed Brazilian cohort. This cross-sectional study included participants aged 60 + from the Brazilian Biobank for Aging Studies (2004–2024). Neuropathology data covered nine markers and a neuropathological comorbidity score (NPC). Frailty was measured using a frailty index from 33 health deficits, and cognitive impairment was defined as a Clinical Dementia Rating ≥ 0.5, both based on post-mortem informant reports with the next of kin. Linear regression models, adjusted for age, sex, education, race, and cognitive impairment, assessed associations between neuropathology and frailty. Effect modification by cognitive status was also explored. Among 1,370 participants (mean age = 78.4 ± 9.3 years; 53% women), 45% were frail, and 38% had cognitive impairment. Neurofibrillary tangles (NFT) (β = 0.005, 95%CI = 0.000-0.008, p = 0.036), Lewy body disease pathology (β = 0.008, 95%CI = 0.003–0.012, p = 0.001), lacunar infarcts (β = 0.032, 95%CI = 0.012–0.052, p = 0.002), siderocalcinosis (β = 0.015, 95%CI = 0.002–0.030, p = 0.023), and hyaline arteriolosclerosis (β = 0.034, 95%CI = 0.021–0.048, p < 0.001) were associated with frailty, independent of cognitive impairment. Higher NPC scores were linked to higher frailty with stronger associations observed among cognitively impaired participants (β = 0.007, 95%CI = 0.003–0.011, p = 0.001), as indicated by a p-value for interaction = 0.007. There was a more pronounced association between neuropathology and frailty among cognitively impaired participants for NFT (β = 0.007, 95%CI = 0.001–0.015, p = 0.020) and hyaline arteriolosclerosis (β = 0.052, 95%CI = 0.031–0.073, p < 0.001). Unlike other neuropathologies, hyaline arteriolosclerosis was associated with higher frailty levels in participants without cognitive impairment (β = 0.020, 95%CI = 0.002–0.038, p = 0.023). Our findings suggest that neuropathology affects frailty independently of cognitive status, although its impact is compounded by cognitive impairment. Moreover, cerebrovascular pathology’s association with frailty in cognitively normal participants highlights the potential benefit of early cardiovascular risk management to reduce frailty risk, which is crucial in low- and middle-income countries considering the disproportionately high burden of cardiovascular and cerebrovascular conditions in these populations.
Supplementary Information
The online version contains supplementary material available at 10.1186/s40478-025-02003-1.
Keywords: Cognitive disorders, Frailty, Dementia, Neuropathology
Introduction
Frailty is a state of increased vulnerability due to the accumulation of age-related health deficits, leading to an impaired ability to maintain homeostasis in response to stressors [1]. Frailty is associated with falls, functional decline, hospitalization, and mortality. Frailty has also been linked to Alzheimer’s disease and other types of dementia [2–5]. Previous studies have shown that older adults classified as frail often exhibit a higher prevalence of biomarkers indicative of neurodegenerative and cerebrovascular diseases compared to those classified as robust [6–8].
With global aging, frailty prevalence is projected to rise significantly in the coming decades, particularly in low- and middle-income countries (LMIC). This sharper increase can be attributed to several factors, such as lower education levels, ethnic and socioeconomic disparities, a higher prevalence of cardiovascular risk factors, and unequal access to healthcare. These issues contribute to the earlier accumulation of health deficits over the lifespan, explaining the greater rise in frailty within lower-income regions [9–11]. Despite this scenario, most research on neuropathology, cognitive impairment, and frailty has disproportionately focused on high-income countries, leaving a significant gap in understanding how these conditions and their complex interactions manifest, particularly in more diverse and underserved populations [11, 12].
Earlier neuropathological studies from high-income countries, primarily involving individuals from retirement or religious communities, have suggested an association between neuropathology and frailty, independent of cognitive status [7, 13]. Notably, frail older adults may present a higher prevalence of dementia even with a relatively low burden of neuropathology [14]. Some researchers propose that neuropathology and frailty are interconnected through shared pathological mechanisms, such as neuroendocrine dysregulation, poor nutritional status, and chronic inflammation. Others suggest that neuropathology, particularly cerebrovascular pathology, may directly contribute to frailty [7, 15–17]. However, the precise mechanisms linking these two conditions remain unclear. Therefore, this study aimed to investigate the associations of frailty with neurodegenerative and cerebrovascular neuropathology using neuropathological data from an ethnically diverse sample with lower educational attainment from Brazil, a large LMIC experiencing rapid population aging and significant health disparities [18].
Methods
Participants
This observational cross-sectional study analyzed clinical and neuropathological data collected by the Biobank for Aging Studies (BAS) between 2004 and 2024. The BAS collects brains from deceased individuals aged ≥ 18 years undergoing autopsy at the São Paulo Autopsy Service. In Brazil, autopsy verification is mandatory to determine the cause of death in cases of non-traumatic deaths of unknown causes. Eligible participants have the brain donated by the next of kin after the signature of an informed consent. A post-mortem interview is performed with a next of kin who had at least weekly contact with the deceased for six months before death. Sociodemographic, medical history, clinical, functional, and neuropsychiatric information were collected through questionnaires applied by trained gerontologists [19].
For this study, we included only BAS participants who were 60 years or older, considering frailty is a well-established syndrome in the geriatric population. We excluded eligible participants who presented inconsistent clinical data or whose brain tissue was incompatible with neuropathological analysis (cerebrospinal fluid pH < 6.5 or major acute brain lesions including hemorrhages, infarctions, and trauma that required brain retention to determine the death cause).
Neuropathological assessment
Brain tissue was obtained within 24 h of death. The left hemisphere was fixed in 4% buffered paraformaldehyde, while selected brain areas from the contralateral hemisphere were frozen at − 80 °C. The following samples from the fixed hemisphere were embedded in paraffin: inferior and middle frontal gyrus, middle and superior temporal gyri, angular gyrus, superior frontal and anterior cingulate gyrus, visual cortex, hippocampal formation at the level of the lateral geniculate body, amygdala, basal ganglia at the level of the anterior commissure, thalamus, midbrain, pons, medulla oblongata, and cerebellum. Sections from all areas were stained with hematoxylin and eosin. Immunohistochemistry with antibodies against β-amyloid (Aβ) (4G8, 1:10.000; BioLegend #800,701), phosphorylated tau (AT8, 1:400; Invitrogen MN1020), TDP-43 (1:500, BioLegend #829,901), and α-synuclein (LB509, 1:500; Sigma-Aldrich MABN824) were performed in selected sections [20]. Detailed explanation of the use of antibodies and specified the brain regions where immunohistochemistry was performed can be found in Supplemental Table 1.
AD-type pathology was scored using the Braak and Braak staging system for neurofibrillary pathology and the Consortium to Establish a Registry for AD (CERAD) criteria for neuritic plaque [21, 22]. The neurofibrillary tangles (NFT) start at the transentorhinal regions (stage I), spreading to the cornu ammonis (CA1 subfield) of the hippocampus (stage II), then to the subiculum, basal ganglia, and amygdala (stage III), finally accumulating in the isocortex in the temporal region (stage IV), extending widely into the whole CA region, thalamus and hypothalamus (stage V) and then reaching the calcarine cortex and fascia dentata of the hippocampus (stage VI). The frequency of neuritic plaques (NP) was categorized according to the CERAD rating scheme of none (0), sparse (1–5 per ×100 field - I), moderate (6–15 per ×100 field - II), or frequent (> 15 per ×100 field - III).
Lewy-type pathology was classified according to the Braak et al. staging for Parkinson’s Disease (PD), the term Lewy body disease (LBD) pathology was used for diseases associated with Lewy bodies, thereby eliminating the distinction between PD, Parkinson’s Disease dementia, and dementia with Lewy Body [23]. The presence of LBD pathology was assessed considering six stages of progression, from the presence of Lewy bodies in the dorsal motor nucleus and anterior olfactory structure (stage 1), followed by lesions in caudal raphe nuclei (stage 2), lesions in midbrain particularly in the pars compacta of the substantia nigra (stage 3), presence of pathology on the transentorhinal regions and allocortex (stage 4), lesions on areas of the neocortex and prefrontal neocortex (stage 5), and finally lesions in sensory association areas of the neocortex and premotor areas (stage 6) [23].
The presence of cerebrovascular diseases was evaluated using hematoxylin and eosin-stained histological slides in all sampled areas [24]. Hyaline arteriolosclerosis was classified as present according to the presence of moderate and/or severe microvascular changes in three or more cortical areas. A graphic description of the grading process for small vessel arteriosclerosis can be found in Supplemental Fig. 1. Lacunar infarcts were registered by topography, stage, size, and number. The presence of lacunar infarcts was defined as one or more infarcts measuring ≤ 1.0 cm in any of the regions of the brain described above [24]. Siderocalcinosis, a vascular mineralization with an encrustation of calcium and iron in the middle layer, was evaluated in the basal ganglia and classified as present or absent [24]. Hippocampal sclerosis, defined by pyramidal cell loss and gliosis in CA1 and subiculum of the hippocampal formation, was noted and scored as present or absent [25]. Moreover, cerebral amyloid angiopathy (CAA) was considered present when moderate or severe lesions were observed widespread in the parenchyma in at least three different cortical areas [26].
TDP-43 pathology was assessed as stage 1, corresponding to TDP-43 pathology in the amygdala, stage 2 in the amygdala and hippocampus and stage 3, corresponding to pathology in the amygdala, hippocampus, and middle frontal gyrus [27].
All neuropathological assessments followed standardized protocols and were conducted by experienced neuropathologists. Inter-rater agreement was evaluated in a subset of cases, with an overall agreement of 90% and a kappa coefficient of 0.82, indicating substantial agreement.
A neuropathological comorbidity score (NPC) was constructed to account for the co-occurrence of neuropathological lesions in each subject. Previous studies have shown that such scores are associated with worsening neuropsychiatric symptoms, while also demonstrating an additive effect of multiple neuropathologies on cognitive impairment [19]. To develop the NPC, we performed a logistic regression using the presence of cognitive impairment measured by the Clinical Dementia Rating instrument (CDR ≥ 0.5) as the dependent variable, and all neuropathological variables as independent variables, adjusting for age at death, sex, education, and race. Neuropathological features associated at the 0.05 alpha level with cognitive impairment were included in the final model. Considering that each neuropathology was associated with different odds of cognitive impairment, points were assigned to each neuropathological variable by dividing their logistic regression coefficient by the coefficient with the lowest value, rounding to the nearest integer. The NPC for each participant was the sum of these points for each present neuropathological variable.
Frailty evaluation
The outcome was a frailty index calculated using information on 33 health-related items obtained from the post-mortem interview with the next of kin [28]. In this study, we adopted previously published standard procedures for developing the frailty index [29]. Deficits were health issues related to aging, present in 1–80% of the sample, not highly correlated with each other (r < 0.95), and not missing in over 5% of the sample. The items included sensory impairments, deglutition and dentition status, mental health indicators, the presence of chronic diseases, and functional status for performing basic and instrumental activities of daily living. The dichotomous variables were categorized as absent or present, with values of 0 or 1, respectively. Ordinal variables, on the other hand, were redefined on a scale from 0 to 1 with regular intervals, with higher values indicating more severe impairments. To avoid capturing deficits that were acutely present near the time of death, the next of kin was asked about the deceased’s characteristics from three months before death. A complete list of the 33 items composing the frailty index can be found in Supplemental Table 2. The frailty index was calculated as a proportion of deficits across the 33 evaluated health items, with scores ranging from 0 (no deficits) to 1 (total deficits). We adopted the cut point of ≥ 0.25 to define the frail group according to the frailty index scores, as suggested by previous studies, for descriptive purposes. However, for the statistical analysis, the frailty index was used as a continuous variable [28, 29]. Up to 10% of missing data were accepted for each subject, aiming to maximize the use of available data and minimize the need for imputation strategies [29].
Cognitive status and other potential confounders
Age at death, sex, education measured in years, and race and ethnicity (classified according to the Brazilian census criteria as White, Black, Brown, and Asian) were reported by the next of kin. The CDR was used to assess functional cognition [30]. This tool involves a semi-structured interview about six cognitive and functional domains: memory, orientation, judgment and problem-solving, community affairs, home and hobbies performance, and personal care. Given the nature of this study and as previously validated, the CDR was based solely on reports of the next of kin, with scores ≥ 0.5 defining cognitive impairment [31, 32].
Statistical analysis
We described sociodemographic, neuropathological features, and cognitive status for the entire sample and for each group according to frailty status. Means and standard deviation were used for continuous variables, and count and percentages for categorical variables. We compared these characteristics between the two frailty groups using the unpaired t-test for continuous variables and the chi-square test for categorical variables.
Linear regression models were used to investigate the associations between individual neuropathological lesions (main variables of interest) and the frailty index (outcome). We fitted three models: (1) unadjusted; (2) adjusted for sociodemographic factors (age at death, sex, race, and education); (3) fully adjusted, incorporating cognitive status alongside sociodemographic factors as a potential confounder. To evaluate whether cognitive status modified the associations between neuropathology and frailty, we added an interaction term between the neuropathologies and cognitive status in the multivariable models. In the interaction analysis, we included the main effect of neuropathology, the main effect of cognitive impairment, and the interaction term (neuropathology × cognitive impairment). We conducted stratified analyses when the p-value for interaction reached an alpha of 0.05. Finally, we investigated the association between NPC and the frailty index using the same linear regression models, also testing for an interaction between NPC and cognitive status on frailty scores. All statistical analyses were two-tailed, with an alpha level set at 0.05. Analyses were conducted using Stata software version 18.0 (StataCorp. LP, College Station, TX, USA).
Results
Of the 2,035 assessments conducted by the Biobank for Aging Studies between 2004 and 2024, 308 records were excluded for being from individuals younger than 60 years old. An additional 357 records were excluded due to missing data, including 47 for sociodemographic variables, 121 for the frailty index, and 189 for neuropathology, resulting in a final sample of 1,370 participants. The records excluded due to missing data on frailty and neuropathology did not differ in demographic characteristics from those of the included participants (Supplemental Table 3). Participants had a mean age at death of 78.4 ± 9.3 years, 52.7% were women, and 65.8% were White. Mean education was 4.4 ± 3.9 years, and 37.8% of the sample had cognitive impairment (Table 1).
Table 1.
Characteristics of the sample according to frailty status (n = 1,370)
| Total (n = 1,370) | Non-frail (n = 753) | Frail (n = 617) | P-value | |
|---|---|---|---|---|
| Age (years), mean (SD) | 78.37 (9.28) | 75.67 (8.94) | 81.67 (8.62) | <0.001* |
| Women, n (%) | 722 (52.70) | 345 (45.82) | 377 (61.10) | <0.001⊥ |
| Race, n (%) | 0.018⊥ | |||
| Asian | 30 (2.19) | 13 (1.73) | 17 (2.76) | |
| Black | 159 (11.61) | 83 (11.02) | 76 (12.32) | |
| Brown | 279 (20.36) | 135 (17.93) | 144 (23.34) | |
| White | 902 (65.84) | 522 (69.32) | 380 (61.58) | |
| Education (years), mean (SD) | 4.39 (3.91) | 4.81 (4.05) | 3.88 (3.68) | <0.001* |
| Cognitive impairment present (%) | 517 (37.74) | 86 (11.42) | 431 (69.85) | <0.001⊥ |
*Unpaired T test; ⊥ Chi-Square test
In this sample, 617 (45.0%) participants were classified as frail based on information from post-mortem interviews with the next of kin. The number of assessed deficits ranged from 30 to 33, with a mean number of 32.90 ± 0.36. The frailty index ranged from 0.01 to 0.73, with a mean of 0.25 ± 0.15. It was associated with age (β = 0.006, 95%CI = 0.005; 0.006, p < 0.001), and higher in women (0.27 ± 0.15) compared to men (0.22 ± 0.14, p < 0.001). The frailty index score at percentile 99 was 0.60, below the suggested threshold of 0.70 [29].
Frail individuals were more likely to be older, predominantly women, more frequently identified as Black, Brown, or Asian, to have less education, and to present cognitive impairment (Table 1). Moreover, as expected, frailty was associated with a higher frequency of cardiovascular, metabolic, and musculoskeletal diseases, as well as higher dependency on basic and instrumental activities of daily living (Supplemental Tables 4 and 5).
The frail participants also exhibited a higher frequency of neuropathological lesions. (Table 2) In the initial multivariate linear regression model adjusted for age at death, sex, education, and race, all neuropathological lesions were associated with increased frailty index scores. To determine whether this association was influenced by cognitive impairment, a second model was adjusted for cognitive status. In this model, higher Braak NFT stage and LBD pathology, along with the presence of lacunar infarcts, siderocalcinosis, and hyaline arteriolosclerosis, remained associated with elevated frailty index scores independent of cognitive status (Table 3).
Table 2.
Neuropathologic status of the sample according to frailty status (n = 1,370)
| Total (n = 1,370) | Non-frail (n = 753) | Frail (n = 617) | P-value | |
|---|---|---|---|---|
| Braak NFT Stage, n (%) | <0.001⊥ | |||
| 0 | 151 (11.02) | 114 (15.14) | 37 (6.00) | |
| I | 190 (13.87) | 135 (17.93) | 55 (8.91) | |
| II | 313 (22.85) | 196 (26.03) | 117 (18.96) | |
| III | 291 (21.24) | 153 (20.32) | 138 (22.37) | |
| IV | 212 (15.47) | 112 (14.87) | 100 (16.21) | |
| V | 121 (8.83) | 31 (4.12) | 90 (14.59) | |
| VI | 92 (6.72) | 12 (1.59) | 80 (12.97) | |
| CERAD NP score, n (%) | <0.001⊥ | |||
| None | 746 (54.45) | 472 (62.68) | 274 (44.41) | |
| Sparse | 212 (15.47) | 124 (16.47) | 88 (14.26) | |
| Moderate | 217 (15.84) | 105 (13.94) | 112 (18.15) | |
| Frequent | 195 (14.23) | 52 (6.91) | 143 (23.18) | |
| TDP-43, n (%) | <0.001⊥ | |||
| 0 | 824 (90.65) | 478 (93.91) | 346 (86.5) | |
| 1 | 39 (4.29) | 18 (3.54) | 21 (5.25) | |
| 2 | 38 (4.18) | 7 (1.38) | 31 (7.25) | |
| 3 | 8 (0.88) | 6 (1.18) | 2 (0.50) | |
| LBD pathology, n (%) | <0.001⊥ | |||
| 0 | 1,194 (87.15) | 688 (91.37) | 506 (82.01) | |
| I | 21 (1.53) | 12 (1.59) | 9 (1.46) | |
| II | 18 (1.31) | 11 (1.46) | 7 (1.13) | |
| III | 35 (2.55) | 13 (1.73) | 22 (3.57) | |
| IV | 48 (3.50) | 15 (1.99) | 33 (5.35) | |
| V | 38 (2.77) | 12 (1.59) | 26 (4.24) | |
| VI | 16 (1.17) | 2 (0.27) | 14 (2. 27) | |
| Cerebral amyloid angiopathy, n (%) | 95 (6.93) | 28 (3.72) | 67 (10.86) | <0.001⊥ |
| Hippocampal sclerosis, n (%) | 44 (3.21) | 11 (1.46) | 33 (5.35) | <0.001⊥ |
| Lacunar infarcts, n (%) | 140 (10.22) | 39 (5.18) | 101 (16.37) | <0.001⊥ |
| Siderocalcinosis, n (%) | 372 (27.15) | 182 (24.17) | 190 (30.79) | 0.006⊥ |
| Hyaline arteriolosclerosis, n (%) | 339 (29.12) | 155 (20.58) | 244 (39.55) | <0.001⊥ |
⊥ Chi-Square test; NFT: Neurofibrillary tangles; CERAD: Consortium to Establish a Registry for Alzheimer’s Disease; NP: Neuritic Plaques; LBD: Lewy body disease; TDP-43: Transactive DNA-binding protein 43
Table 3.
Association between neuropathology and the frailty index (n = 1,370)
| Unadjusted a | Model 1 b | Model 2 c | ||||
|---|---|---|---|---|---|---|
| β (95% CI) | P value | β (95% CI) | P value | β (95% CI) | P value | |
| Braak NFT Stage | 0.032 (0.028; 0.037) | < 0.001 | 0.021 (0.016; 0.026) | < 0.001 | 0.005 (0.000; 0.008) | 0.036 |
| CERAD NP score | 0.037 (0.030; 0.044) | < 0.001 | 0.023 (0.016; 0.030) | < 0.001 | 0.001 (-0.004; 0.007) | 0.697 |
| LBD pathology | 0.017 (0.011; 0.023) | < 0.001 | 0.013 (0.008; 0.019) | < 0.001 | 0.008 (0.003; 0.012) | 0.001 |
| TDP-43 | 0.045 (0.026; 0.065) | < 0.001 | 0.029 (0.011; 0.047) | 0.001 | 0.003 (-0.013; 0.018) | 0.738 |
| Cerebral Amyloid Angiopathy | 0.089 (0.057; 0.120) | < 0.001 | 0.057 (0.027; 0.087) | < 0.001 | 0.020 (-0.004; 0.044) | 0.102 |
| Hippocampal sclerosis | 0.107 (0.061; 0.113) | < 0.001 | 0.068 (0.025; 0.111) | 0.002 | 0.016 (-0.018; 0.050) | 0.369 |
| Lacunar infarcts | 0.086 (0.060; 0.113) | < 0.001 | 0.070 (0.046; 0.095) | < 0.001 | 0.032 (0.012; 0.052) | 0.002 |
| Siderocalcinosis | 0.035 (0.017; 0.053) | < 0.001 | 0.021 (0.004; 0.038) | 0.015 | 0.015 (0.002; 0.030) | 0.023 |
| Hyaline arteriolosclerosis | 0.077 (0.060; 0.095) | < 0.001 | 0.055 (0.038; 0.071) | < 0.001 | 0.034 (0.021; 0.048) | < 0.001 |
NFT: neurofibrillary tangles; CERAD: Consortium to Establish a Registry for Alzheimer’s Disease; NP: Neuritic Plaques; LBD: Lewy body disease; TDP-43: Transactive DNA-binding protein 43
a Unadjusted model: univariate linear regression model
b Model 1: linear regression adjusted for age at death, sex, race, and education
c Model 2: linear regression adjusted for age at death, sex, race, education, and cognitive impairment
In our sample, the neuropathological comorbidity score (NPC) ranged from 0 to 10 and included the following variables: NFT, NP, TDP-43, hippocampal sclerosis, and lacunar infarcts (Supplemental Table 6). The mean NPC was 1.68 ± 2.17. The NPC was higher in the frail group (mean 2.90 ± 2.64) in comparison to the non-frail group (mean 0.93 ± 1.37) (p = < 0.001) (Fig. 1a). Higher NPC was associated with a higher frailty index after adjustments for age at death, sex, race, and cognitive impairment (β = 0.005, IC 95%=0.002; 0.009, p = 0.004) (Fig. 1b).
Fig. 1.
(A) Boxplot of NPC scores according to frailty status. (B) Boxplot of the frailty index scores according to the neuropathological comorbidity score; NPC: Neuropathological comorbidity score
Cognitive status, determined through post-mortem next of kin interviews, significantly interacted with neuropathologies, modifying their association with frailty. Specifically, the relationships of Braak NFT stage, CERAD NP score, hyaline arteriolosclerosis, and NPC with frailty were more pronounced in the presence of cognitive impairment (Fig. 2a-d). In contrast, no significant interaction was observed for TDP-43, CAA, hippocampal sclerosis, lacunar infarcts, siderocalcinosis, or LBD pathology (Supplemental Fig. 2a-f). In stratified analyses by cognitive status, a stronger association was observed between higher Braak NFT stage, elevated NPC, and the presence of hyaline arteriolosclerosis with increased frailty index scores in participants with cognitive impairment. Interestingly, hyaline arteriolosclerosis remained significantly associated with higher frailty index scores even in participants without cognitive impairment, suggesting a persistent influence of this marker of small vessel disease on frailty. However, the CERAD NP score showed no significant association with frailty in either cognitive status group after stratification (Table 4).
Fig. 2.
Linear predictions of the frailty index in participants with normal cognition (CDR = 0 in blue) and those with cognitive impairment (CDR ≥ 0.5 in red) according to the presence of (a) Braak NFT Stage; (b) CERAD NP score; (c) Arteriolosclerosis; (d) NPC; P-value represents the value for the interaction between each neuropathology and cognitive status in linear regression models adjusted for age at death, sex, education, and race; CDR: Clinical Dementia Rating; NFT: neurofibrillary tangles; CERAD: Consortium to Establish a Registry for Alzheimer’s Disease; NP: Neuritic Plaques; NPC: neuropathological comorbidity score
Table 4.
Associations between neuropathology and the frailty index stratified by cognitive status (n = 1,370)
| No cognitive impairment (n = 853) | Cognitive impairment (n = 517) | |||
|---|---|---|---|---|
| β (95% IC) | P value | β (95% IC) | P value | |
| Braak NFT Stage | 0.001 (-0.005; 0.007) | 0.678 | 0.007 (0.001; 0.015) | 0.020 |
| CERAD NP score | -0.006 (-0.014; 0.002) | 0.159 | 0.007 (-0.001; 0.016) | 0.115 |
| Hyaline arteriolosclerosis | 0.020 (0.002; 0.038) | 0.023 | 0.052 (0.031; 0.073) | < 0.001 |
| NPC | -0.001 (-0.007; 0.005) | 0.714 | 0.007 (0.003; 0.011) | 0.001 |
NFT: neurofibrillary tangles; CERAD: Consortium to Establish a Registry for Alzheimer’s Disease; NP: Neuritic Plaques; NPC: neuropathological comorbidity score
Linear regression model adjusted for age at death, sex, education, and race
Discussion
In this study, we found that individual neuropathologies and neuropathological comorbidity were significantly associated with frailty in an admixed population from a LMIC. This association was independent of cognitive status for NFT, LBD pathology, lacunar infarcts, siderocalcinosis, hyaline arteriolosclerosis, and NPC, supporting the concept that neuropathology is not solely associated with frailty due to cognitive impairment. This association might reflect either broader pathological processes common to both neuropathology and frailty or a direct effect of neuropathology on frailty. Interestingly, the association of hyaline arteriolosclerosis with frailty was observed even in cognitively normal individuals, suggesting an early and direct role of cerebrovascular health in frailty.
Although cognitive impairment appeared to strengthen the association of NFT, hyaline arteriolosclerosis, and NPC with frailty, we recognize that cognitive impairment and frailty are partially caused by overlapping neuropathological processes. Thus, individuals in the cognitively impaired group may have higher levels of these pathologies, which could explain the stronger association with frailty. In other words, rather than implying that neuropathology and cognitive impairment operate separately, our data underscore that a higher burden of neuropathology may drive both cognitive decline and frailty. Consequently, frailty could be exacerbated when pathology is severe enough to manifest as clinical cognitive impairment.
Frailty and cerebrovascular pathology share common pathophysiological mechanisms, as both are linked to cardiovascular conditions, oxidative stress, neuroendocrine dysregulation, and inflammation, all of which contribute to microvascular and endothelial dysfunction [33, 34]. These processes ultimately lead to the deposition of brain pathology and sarcopenia, a key finding in frailty, independently of cognitive function. It is also possible that cerebrovascular pathologies, such as small vessel disease and cerebral infarcts, may directly cause frailty as the accumulation of neuropathology in areas crucial for motor and executive function may contribute to impaired mobility, slower gait speed, and reduced lower body function [35–37]. The association between hyaline arteriolosclerosis and frailty was stronger in individuals with cognitive impairment. This may be due to a greater overall pathological accumulation or even that cognitive and executive dysfunction may exacerbate frailty. Interestingly, the relationship was also observed in those without cognitive impairment, which can indicate that non-cognitive pathways, such as those that impact mobility and muscle function, also play a critical role in linking cerebrovascular pathology and frailty [33, 36, 38]. This is particularly relevant for low- and middle-income countries, where individuals face an earlier onset of cardiovascular risk factors and greater exposure to adverse conditions—including food insecurity, air pollution, physical inactivity, infectious diseases, and limited healthcare access—leading to a higher prevalence of cerebrovascular pathology [19, 39]. Our results suggest that frailty may reflect underlying cardiovascular conditions, independent of cognitive status and that early interventions targeting cardiovascular health might help reduce both frailty and dementia.
Our findings demonstrated that the accumulation of neurodegenerative pathology was linked to higher frailty index scores. Previous reports have shown that Aβ deposition and LBD pathology are associated with impairments contributing to frailty, such as slower gait speed, reduced walking cadence, muscle weakness, mobility issues, and lower performance on physical tests like the Short Physical Performance Battery and Timed Up and Go, even in individuals without dementia or after controlling for cognitive status [40–42]. We found an interaction between cognitive impairment and intermediate to severe levels of NFT on frailty index scores. Braak NFT stages of III or more are associated with the emergence of cognitive impairment, which can contribute to sarcopenia, undernutrition, decreased resting metabolic rate, and slower walking speed— reflecting the accumulation of pathology severe enough to manifest clinical deficits and possibly impact on frailty status [15, 19]. In our sample, neuritic plaques (NP) were not associated with frailty after stratification for cognitive impairment. It has been shown that NP might not influence cognition directly and that its effect on cognition might be mediated through NFT formation, potentially diminishing NP’s role in cognitive and neuropsychiatric changes that can lead to weight loss, sarcopenia, and frailty status [43, 44].
While previous neuropathological studies suggested that the associations between Alzheimer’s disease (AD) pathology, macroinfarcts, Lewy body disease, and nigral neuronal loss with frailty were not modified by cognitive status, our findings indicate a stronger effect of NFT stages, cerebrovascular pathology, and the NPC score on frailty in those with cognitive impairment [6, 13]. Much of the earlier research was conducted within the Religious Orders Study (ROS) and the Rush Memory and Aging Project (MAP), involving predominantly older, highly educated, White participants from retirement or religious communities, many of whom shared specific living habits [45, 46]. In contrast, our population-based sample was younger, had less education, and presented a higher prevalence of cardiovascular and metabolic diseases. All these characteristics contribute to lower cognitive reserve, which might reduce the ability to manage neuropathological accumulation and the possible combined effects of neuropathology and cognitive impairment on frailty [47].
Few studies have investigated the relationship between neuropathological comorbidity burden and frailty, yielding different results [6, 7, 48]. Although some reports have described the effects of multiple neuropathologies on frailty, showing that composite measures of neuropathological accumulation are associated with impaired gait functioning, worsened cognition, and higher frailty levels, another study found no significant association between neuropathological burden and worsening frailty trajectories after adjusting for confounders, such as cognitive impairment [7, 49]. In our study, the NPC was linked to higher frailty index scores independent of cognitive status. We also noted that this association was more pronounced in those with cognitive impairment. A previous study by our group demonstrated additive effects of multiple neuropathology on cognition and neuropsychiatric symptoms which could also influence frailty when pathology is sufficiently advanced [19].
Our study is among the first to investigate the relationship between neuropathology and frailty in a population-based sample from a low- and middle-income country. This diversity in ethnicity and educational background stands in contrast to prior investigations, which focused primarily on participants who were White and had higher levels of education [45, 46]. We examined a broad spectrum of neuropathologies for which data on their association with frailty were limited in the literature. To capture the cumulative effect of neuropathology on frailty, we developed a comprehensive neuropathological index.
However, our study also has limitations. The sampling methods may be biased towards deaths related to cardiovascular conditions, potentially underrepresenting other causes such as trauma and neoplasia. We acknowledge that frailty, as captured by the FI, and cognitive impairment may be influenced by shared underlying processes, including neuropathological accumulation. In this context, the stronger associations between neuropathology and frailty observed in individuals with cognitive impairment might reflect, at least in part, a greater overall burden of neuropathology, which could contribute to the observed interaction effect. Although we assessed clinical and functional variables based on next of kin reports from three months before death and any end-of-life declines are likely to have affected frailty index scores non-differentially, other factors related to informant characteristics or recall remain potential sources of bias. These include, for instance, the next of kin’s own educational level or understanding of the participant’s health status, which were not fully captured and thus persist as limitations in our analysis. However, previous work has validated the methods used in this study [31, 32]. While the informant-based CDR shows high sensitivity (86.6%) and specificity (84.4%) for classifying dementia, its sensitivity for mild cognitive impairment is notably lower (57.8%). Moreover, despite strong specificity in detecting normal cognition (93.7%), some individuals with borderline impairments could have been misclassified. Consequently, our binary approach to cognitive status provides a clear contrast between impaired and unimpaired groups, but may underestimate subtle levels of cognitive decline, potentially influencing the observed associations with frailty [31]. Finally, the cross-sectional nature of our study limits the ability to draw causal inferences from the associations we investigated.
Conclusion
In this large neuropathology study comprising 1,370 participants, we found that individual neuropathologies and neuropathological comorbidity burden were significantly associated with frailty. Notably, some of these associations, particularly those involving cerebrovascular pathologies, were independent of cognitive status. In addition, among the cognitive impaired, the association between certain neuropathologies, such as neurofibrillary tangles (NFT) and hyaline arteriolosclerosis, and frailty was stronger. A similar pattern was observed with neuropathological comorbidity (NPC). These findings underscore the complexity of the relationship between neuropathology and frailty, indicating that while neuropathology can affect frailty independently of cognitive status, its impact is compounded by cognitive impairment.
Future studies should explore whether controlling cardiovascular risk factors and, thereby reducing cerebrovascular pathology, could mitigate frailty. This is particularly crucial in low- and middle-income countries (LMICs), where the burden of cardiovascular and cerebrovascular conditions is disproportionately high [19, 39, 50]. Moreover, understanding the close relationship between neuropathology, frailty, and cognitive status may uncover novel strategies to improve cognitive function and reduce frailty, ultimately leading to better outcomes in aging populations.
Electronic supplementary material
Below is the link to the electronic supplementary material
Acknowledgements
Not applicable.
Author contributions
F.B.S.: study concept and design, analysis, interpretations of data, and manuscript drafting. M.J.R.A. and C.K.S: study concept and design, supervision of the project, data analysis and critical revision of the manuscript for intellectual content. A.F.O.J., V.R.P., R.E.P.L., and R.D.R.: data curation. L.T.G., C.A.P., and E.F. critical revision of the manuscript for intellectual content. All authors reviewed the manuscript and approved its final version.
Funding
This study was funded by the Capacity Building in International Dementia Research (CBIDR) Program (24CBIDR-1185483).
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
The procedures of the Biobank for Aging Studies have been fully described in previous publications and comply with both international and Brazilian regulations for research involving human subjects. The study was approved by local and federal ethics committees under the BAS [19, 51]. Voluntary informed consent, which outlined research procedures, objectives, risks, and benefits, as well as the option to withdraw at any time, was obtained by the interviewer and signed by the next of kin before any procedures were conducted. All materials used in the study were anonymized and coded. Data was accessed via the REDCap platform, restricted to the researchers involved in the study, with all access logged [52].
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Rockwood K, Howlett SE (2018) Fifteen years of progress in understanding frailty and health in aging. BMC Med 16(1):1–4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Boyle PA, Buchman AS, Wilson RS, Leurgans SE, Bennett DA (2010) Physical frailty is associated with incident mild cognitive impairment in community-based older persons. J Am Geriatr Soc 58(2):248–255 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Buchman AS, Boyle PA, Wilson RS, Tang Y, Bennett DA (2007) Frailty is associated with incident Alzheimer’s disease and cognitive decline in the elderly. Psychosom Med 69(5):483–489 [DOI] [PubMed] [Google Scholar]
- 4.Wang C, Ji X, Wu X, Tang Z, Zhang X, Guan S et al (2017) Frailty in relation to the risk of Alzheimer’s disease, dementia, and death in older Chinese adults: a seven-year prospective study. J Nutr Health Aging 21(6):648–654 [DOI] [PubMed] [Google Scholar]
- 5.Trebbastoni A, Canevelli M, D’Antonio F, Imbriano L, Podda L, Rendace L et al (2017) The impact of frailty on the risk of conversion from mild cognitive impairment to Alzheimer’s disease: evidences from a 5-year observational study. Front Med 4(OCT) [DOI] [PMC free article] [PubMed]
- 6.Buchman AS, Yu L, Wilson RS, Schneider JA, Bennett DA (2013) Association of brain pathology with the progression of frailty in older adults. Neurology 80(22):2055–2061 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Buchman AS, Yu L, Wilson RS, Boyle PA, Schneider JA, Bennett DA (2014) Brain pathology contributes to simultaneous change in physical frailty and cognition in old age. J Gerontol - Ser Biol Sci Med Sci 69(12):1536–1544 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Maltais M, De Souto Barreto P, Hooper C, Payoux P, Rolland Y, Vellas B (2019) Association between brain β-amyloid and frailty in older adults. J Gerontol - Ser Biol Sci Med Sci 74(11):1747–1752 [DOI] [PubMed] [Google Scholar]
- 9.Siriwardhana DD, Hardoon S, Rait G, Weerasinghe MC, Walters KR (2018) Prevalence of frailty and prefrailty among community-dwelling older adults in low-income and middle-income countries: a systematic review and meta-analysis. BMJ Open 8(3) [DOI] [PMC free article] [PubMed]
- 10.Wu AH, Setiawan VW, Stram DO, Crimmins EM, Tseng CC, Lim U et al (2023) Racial, ethnic, and socioeconomic differences in a deficit accumulation frailty index in the multiethnic cohort study. J Gerontol - Ser Biol Sci Med Sci 78(7):1246–1257 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hoogendijk EO, Rijnhart JJM, Kowal P, Pérez-Zepeda MU, Cesari M, Abizanda P et al (2018) Socioeconomic inequalities in frailty among older adults in six low- and middle-income countries: results from the WHO study on global ageing and adult health (SAGE). Maturitas 115:56–63 [DOI] [PubMed] [Google Scholar]
- 12.Livingston G, Huntley J, Liu KY, Costafreda SG, Selbæk G, Alladi S et al (2024) Dementia prevention, intervention, and care: 2024 report of the lancet standing commission. Lancet 404(10452):572–628 [DOI] [PubMed] [Google Scholar]
- 13.Buchman AS, Schneider JA, Leurgans S, Bennett DA (2008) Physical frailty in older persons is associated with alzheimer disease pathology. Neurology 71(7):499–504 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wallace LMK, Theou O, Godin J, Andrew MK, Bennett DA, Rockwood K (2019) Investigation of frailty as a moderator of the relationship between neuropathology and dementia in Alzheimer’s disease: a cross-sectional analysis of data from the rush memory and aging project. Lancet Neurol 18(2):177–184 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Robertson DA, Savva GM, Kenny RA (2013) Frailty and cognitive impairment-a review of the evidence and causal mechanisms. Ageing Res Rev 12(4):840–851 [DOI] [PubMed] [Google Scholar]
- 16.Halil M, Cemal Kizilarslanoglu M, Emin Kuyumcu M, Yesil Y, Cruz Jentoft AJ (2015) Cognitive aspects of frailty: mechanisms behind the link between frailty and cognitive impairment. J Nutr Health Aging 19(3):276–283 [DOI] [PubMed] [Google Scholar]
- 17.Walston J, Hadley EC, Ferrucci L, Guralnik JM, Newman AB, Studenski SA et al (2006) Research agenda for frailty in older adults: toward a better understanding of physiology and etiology: summary from the American Geriatrics Society/National Institute on Aging research conference on frailty in older adults. J Am Geriatr Soc 54(6):991–1001 [DOI] [PubMed] [Google Scholar]
- 18.Andrade JM, Duarte YA, de Alves O, Andrade LC, Junior FCD, de Lima-Costa PRB (2018) Frailty profile in Brazilian older adults: ELSI-Brazil. Rev Saude Publica 52:1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Suemoto CK, Ferretti-Rebustini REL, Rodriguez RD, Leite REP, Soterio L, Brucki SMD et al (2017) Neuropathological diagnoses and clinical correlates in older adults in Brazil: a cross-sectional study. PLoS Med 14(3):1–22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Paradela RS, Justo AFO, Paes VR, Leite REP, Pasqualucci CA, Grinberg LT et al (2023) Association between APOE-ε4 allele and cognitive function is mediated by Alzheimer’s disease pathology: a population-based autopsy study in an admixed sample. Acta Neuropathol Commun 11(1):205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Braak H, Braak E (1991) Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol (Berl) 82(4):239–259 [DOI] [PubMed] [Google Scholar]
- 22.Mirra SS, Heyman A, McKeel D, Sumi SM, Crain BJ, Brownlee LM et al (1991) The consortium to establish a registry for Alzheimer’s disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology 41(4):479–486 [DOI] [PubMed] [Google Scholar]
- 23.Braak H, Del Tredici K, Rüb U, De Vos RAI, Jansen Steur ENH, Braak E (2003) Staging of brain pathology related to sporadic Parkinson’s disease. Neurobiol Aging 24(2):197–211 [DOI] [PubMed] [Google Scholar]
- 24.Grinberg LT, Thal DR (2010) Vascular pathology in the aged human brain. Acta Neuropathol (Berl) 119(3):277–290 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Rauramaa T, Pikkarainen M, Englund E, Ince PG, Jellinger K, Paetau A et al (2013) Consensus recommendations on pathologic changes in the hippocampus: a postmortem multicenter inter-rater study. J Neuropathol Exp Neurol 72(6):452–461 [DOI] [PubMed] [Google Scholar]
- 26.Rudolf Thal D, Ghebremedhin E, Yamaguchi H, Del Tredici K, Braak H (2002) Two types of sporadic cerebral amyloid angiopathy. J Neuropathol Exp Neurol 61. Available from: https://academic.oup.com/jnen/article/61/3/282/2610063 [DOI] [PubMed]
- 27.Nelson PT, Lee EB, Cykowski MD, Alafuzoff I, Arfanakis K, Attems J et al (2023) LATE-NC staging in routine neuropathologic diagnosis: an update. Acta Neuropathol (Berl) 145(2):159–173 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rockwood K, Mitnitski A (2011) Frailty defined by deficit accumulation and geriatric medicine defined by frailty. Clin Geriatr Med 27(1):17–26 [DOI] [PubMed] [Google Scholar]
- 29.Theou O, Haviva C, Wallace L, Searle SD, Rockwood K (2023) How to construct a frailty index from an existing dataset in 10 steps. Age Ageing 52(12) [DOI] [PMC free article] [PubMed]
- 30.Morris JC (1993) The clinical dementia rating (cdr): current version and scoring rules. Neurology 43(11):2412–2414 [DOI] [PubMed] [Google Scholar]
- 31.de Ferretti REL, Damin AE, Brucki SMD, Morillo LS, Perroco TR, Campora F et al (2010) Post-mortem diagnosis of dementia by informant interview. Dement E Neuropsychol 4(2):138–144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Waite L, Grayson D, Jorm AF, Creasey H, Cullen J, Bennett H et al (1999) Informant-based staging of dementia using the clinical dementia rating. Alzheimer Dis Assoc Disord 13(1):34–37 [DOI] [PubMed] [Google Scholar]
- 33.Ducca EL, Gomez GT, Palta P, Sullivan KJ, Jack CR, Knopman DS et al (2022) Physical frailty and brain white matter abnormalities: the atherosclerosis risk in communities study. J Gerontol Biol Sci Med Sci 78(2):357–364 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ijaz N, Jamil Y, Brown CH, Krishnaswami A, Orkaby A, Stimmel MB et al (2024) Role of cognitive frailty in older adults with cardiovascular disease. J Am Heart Assoc 13(4):e033594 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Siejka TP, Srikanth VK, Hubbard RE, Moran C, Beare R, Wood A et al (2018) Frailty and cerebral small vessel disease: a cross-sectional analysis of the Tasmanian study of cognition and gait (TASCOG). J Gerontol - Ser Biol Sci Med Sci 73(2):255–260 [DOI] [PubMed] [Google Scholar]
- 36.Kant I, de Bresser J, Montfort SJT, Verlaan JJ, Zacharias N, Winterer G et al (2018) The association between brain volume, cortical brain infarcts, and physical frailty. Neurobiol Aging 70:247–253 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Maltais M, de Souto Barreto P, Moon SY, Rolland Y, Vellas B (2019) Prospective association of white matter hyperintensity volume and frailty in older adults. Exp Gerontol 118:51–54 [DOI] [PubMed] [Google Scholar]
- 38.Avila-Funes JA, Pelletier A, Meillon C, Catheline G, Periot O, Treviño-Frenk I et al (2017) Vascular cerebral damage in frail older adults: the AMImage study. J Gerontol - Ser Biol Sci Med Sci 72(7):971–977 [DOI] [PubMed] [Google Scholar]
- 39.Miranda JJ, Barrientos-Gutiérrez T, Corvalan C, Hyder AA, Lazo-Porras M, Oni T et al (2019) Understanding the rise of cardiometabolic diseases in low- and middle-income countries. Nat Med 25(11):1667–1679 [DOI] [PubMed] [Google Scholar]
- 40.Sathyan S, Ayers E, Gao T, Milman S, Barzilai N, Rockwood K et al (2019) Frailty and risk of incident motoric cognitive risk syndrome. J Alzheimers Dis 71(s1):S85–93 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Wennberg AMV, Savica R, Hagen CE, Roberts RO, Knopman DS, Hollman JH et al (2017) Cerebral amyloid deposition is associated with gait parameters in the Mayo clinic study of aging. J Am Geriatr Soc 65(4):792–799 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Yoon DH, Lee JY, Shin SA, Kim YK, Song W (2018) Physical frailty and Amyloid-β deposits in the brains of older adults with cognitive frailty. J Clin Med 7(7) [DOI] [PMC free article] [PubMed]
- 43.Nelson PT, Jicha GA, Schmitt FA, Liu H, Davis DG, Mendiondo MS et al (2007) Clinicopathologic correlations in a large alzheimer disease center autopsy cohort: neuritic plaques and neurofibrillary tangles do count when staging disease severity. J Neuropathol Exp Neurol 66(12):1136–1146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Bennett DA, Schneider JA, Wilson RS, Bienias JL, Arnold SE (2004) Neurofibrillary tangles mediate the association of amyloid load with clinical alzheimer disease and level of cognitive function. Arch Neurol 61(3):378–384 [DOI] [PubMed] [Google Scholar]
- 45.Bennet D, Schneider J, Arvanitakis Z, Wilson R. (2012) Overview and findings from the religious orders study. Curr Alzheimer Res 9(6):628–645 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Bennett DA, Schneider JA, Buchman AS, De Leon CM, Bienias JL, Wilson RS (2005) The rush memory and aging project: study design and baseline characteristics of the study cohort. Neuroepidemiology 25(4):163–175 [DOI] [PubMed] [Google Scholar]
- 47.Stern Y (2013) Cognitive reserve in ageing. Lancet Neurol 11(11):1006–1012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Wallace LMK, Theou O, Darvesh S, Bennett DA, Buchman AS, Andrew MK et al (2020) Neuropathologic burden and the degree of frailty in relation to global cognition and dementia. Neurology 95(24):e3269–e3279 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Wallace LMK, Theou O, Godin J, Ward DD, Andrew MK, Bennett DA et al (2021) 10-year frailty trajectory is associated with Alzheimer’s dementia after considering neuropathological burden. Aging Med 4(4):250–256 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Arokiasamy P, Uttamacharya, Kowal P, Capistrant BD, Gildner TE, Thiele E et al (2017) Chronic noncommunicable diseases in 6 Low- and Middle-Income countries: findings from wave 1 of the world health organization’s study on global ageing and adult health (SAGE). Am J Epidemiol 185(6):414–428 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Grinberg LT, Nitrini R, Suemoto CK, de Lucena Ferretti-Rebustini RE, Leite REP, Farfel JM et al (2013) Prevalence of dementia subtypes in a developing country: a clinicopathological study. Clinics 68(8):1140–1145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG (2009) Research electronic data capture (REDCap) - A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inf 42(2):377–381 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.