Combined Physical Activity and Sitting Time Association on Older Adults' Body Mass Index: Cross-Sectional Findings From the Brazilian Longitudinal Study of Aging.

This study examined isolated and combined associations of physical activity (PA) and sitting time (ST) with body mass index (BMI) among older Brazilian adults. We have analyzed baseline data of 8,177 participants aged 50 years and older from the Brazilian Longitudinal Study of Aging. Respondents were categorized into sufficient PA/low ST, sufficient PA/high ST, insufficient PA/low ST, or insufficient PA/high ST using the Short Version of International Physical Activity Questionnaire. Multinomial logistic regression analysis examined the associations of being underweight (BMI, <18 kg/m2), overweight (BMI, 25-29.9 kg/m2), and obese (BMI, ≥30 kg/m2) with PA and ST categories. High ST (≥165 min/day) was associated with overweight (odds ratio, 1.26; 95% confidence interval [1.11, 1.44]) and obesity (odds ratio, 1.43; 95% confidence interval [1.21, 1.70]). However, no association was observed between PA and BMI. Participants in the insufficient and sufficient PA/high ST categories were more likely to be obese after adjusting for sociodemographic and health condition. Our findings call for public health initiatives that consider reducing ST in older adults.

Admixture/fine-mapping in Brazilians reveals a West African associated potential regulatory variant (rs114066381) with a strong female-specific effect on body mass and fat mass indexes.

BACKGROUND/OBJECTIVES: Admixed populations are a resource to study the global genetic architecture of complex phenotypes, which is critical, considering that non-European populations are severely underrepresented in genomic studies. Here, we study the genetic architecture of BMI in children, young adults, and elderly individuals from the admixed population of Brazil. SUBJECTS/METHODS: Leveraging admixture in Brazilians, whose chromosomes are mosaics of fragments of Native American, European, and African origins, we used genome-wide data to perform admixture mapping/fine-mapping of body mass index (BMI) in three Brazilian population-based cohorts from Northeast (Salvador), Southeast (Bambuí), and South (Pelotas). RESULTS: We found significant associations with African-associated alleles in children from Salvador (PALD1 and ZMIZ1 genes), and in young adults from Pelotas (NOD2 and MTUS2 genes). More importantly, in Pelotas, rs114066381, mapped in a potential regulatory region, is significantly associated only in females (p = 2.76e-06). This variant is rare in Europeans but with frequencies of ~3% in West Africa and has a strong female-specific effect (95% CI: 2.32-5.65 kg/m2 per each A allele). We confirmed this sex-specific association and replicated its strong effect for an adjusted fat mass index in the same Pelotas cohort, and for BMI in another Brazilian cohort from São Paulo (Southeast Brazil). A meta-analysis confirmed the significant association. Remarkably, we observed that while the frequency of rs114066381-A allele ranges from 0.8 to 2.1% in the studied populations, it attains ~9% among women with morbid obesity from Pelotas, São Paulo, and Bambuí. The effect size of rs114066381 is at least five times higher than the FTO SNPs rs9939609 and rs1558902, already emblematic for their high effects. CONCLUSIONS: We identified six candidate SNPs associated with BMI. rs114066381 stands out for its high effect that was replicated and its high frequency in women with morbid obesity. We demonstrate how admixed populations are a source of new relevant phenotype-associated genetic variants.

Nationwide vitamin D status in older Brazilian adults and its determinants: The Brazilian Longitudinal Study of Aging (ELSI).

Little is known about vitamin D status in older adults in South America, where exposures to ultra-violet radiation are high. We examined the distribution of serum 25-hydroxyvitamin D (25OHD) concentration and its determinants in a nationally representative sample of Brazilians aged 50 years and older. Explanatory variables included environment and individuals' characteristics from the ELSI baseline survey (2015-16). Among the 2,264 participants (mean age = 62.6 years), the geometric mean of 25OHD concentration was 66.8 nmol/L. The prevalence of vitamin D deficiency (< 30 nmol/L) and insufficiency (< 50 nmol/L) were 1.7% (95% CI 1.0, 2.8) and 16% (95% CI 12, 20), respectively. Mean concentrations were lower in those geographical regions situated at lower latitudes. Those at the oldest age, women, self-classified as Black and Brown, living in urban areas and current smokers were more likely to have vitamin D insufficiency, independent of each other and other relevant factors. In contrast, individuals who eat fish regularly were considerably less likely to present lower concentration. Based on these findings it is possible to estimate that about 875,000 older Brazilians have vitamin D deficiency and 7.5 million its insufficiency.

Diabetes, depressive symptoms, and mortality risk in old age: The role of inflammation.

BACKGROUND: Both diabetes and depression increase the mortality risk in the elderly. In this study, we evaluated mortality risk associated with the comorbidity between depression and diabetes. We also assessed the moderating role of inflammation in the mortality risk in this population. METHODS: We included a total of 1,183 community-dwelling older adults, divided into four groups: "neither diabetes nor depression"; "diabetes only"; "depression only," and "both diabetes and depression," and followed-up for a median of 13.5 years. We evaluated the inflammatory status by the high-sensitivity C-reactive protein (hs-CRP) levels. Date of death was computed by reviewing death certificates. We used Cox's proportional hazards models and additive interactions to evaluate the risk of mortality in the subject groups and the moderating effect of hs-CRP. RESULTS: Participants with both diabetes and depression had higher death risk (hazard ratio [HR]: 2.33; 95% confidence interval [CI]: 1.59-3.42) than those with each condition alone (HR diabetes: 2.08 95% CI: 1.56-2.76 HR depression: 1.26; 95% CI: 1.03-1.54). High level of hs-CRP, indicative of high inflammatory status, significantly moderated the risk of mortality in subjects with both diabetes and depression (Bonferroni-adjusted p = 0.0116). CONCLUSIONS: The coexistence of diabetes and depression symptoms is associated with the highest death risk in this population. This risk is moderated by inflammatory status.

Multiple inflammatory markers and 15-year incident ADL disability in admixed older adults: The Bambui-Epigen Study.

BACKGROUND: The ability of inflammatory markers to predict disability in later life has received growing attention. However, the current evidence came predominantly from Caucasians and the role of genomic ancestry has not been investigated. OBJECTIVE: We investigated the prognostic value of multiple citokynes and chemokines for incident disability in admixed older Brazilians and whether genomic African and Native American ancestry affects the association. DESIGN: Population-based longitudinal study. SETTING: The Bambui-Epigen (Brazil) Cohort Study of Aging. SUBJECTS: 1171 males and females aged ≥60 years over 15-year of follow-up. METHODS: Outcome examined was incident activity of daily living (ADL) disability assessed annually (10,039 measures were performed). Serum levels of citokynes (IL6, IL12, TNF, IL10, and IL1ß) and chemokines (CCL2, CCL5, CXCL8, CXCL9 and CXCL10) were measured at baseline. We used 370,539 Single Nucleotide Polymorphisms (SNPs) to estimate each individual genomic ancestry proportions. Potential confounding variables included a wide range of socio-demographic variables and health indicators. Statistical analyses were based on competing risk framework. RESULTS: The incidence rate of disability was 57.9 per 1000 person-years. IL6 level at the highest quartile showed an independent association with ADL disability (SRH=1.32; 95% CI: 1.03, 1.70). Other inflammatory markers showed no statistically significant associations with the outcome. Neither genomic African nor Native American ancestry had an effect modifier on the associations (P for interaction >0.05 for all). CONCLUSION: Among multi-inflammatory markers, only IL6 had the potential to identify people at increased risk of ADL disability, independently of ethno-racial background.

Socioeconomic inequalities in activities of daily living limitations and in the provision of informal and formal care for noninstitutionalized older Brazilians: National Health Survey, 2013.

BACKGROUND: This study assesses the association between socioeconomic factors and living arrangements with activity of daily living limitations (ADL) and the receipt of informal and formal care among non-institutionalized Brazilians aged ≥ 60 years. METHODS: Data come from a nationally representative survey (the Brazilian National Health Survey), conducted in 2013. Outcomes examined include the number of ADL tasks performed with limitations and number of tasks for which the individual received informal care (provided by unpaid relatives or friends), formal care, or no care. Key exposure variables were years of education and household assets. RESULTS: Functioning limitations were reported by 7,233 (30.1 %) of 23,815 survey participants. Of these, 5,978 reported needing help to perform at least one ADL task. There was a strong inverse gradient between physical functioning and both education and household assets that was independent of confounders. The provision of care showed an opposite trend, with the wealthiest being more likely to receive help for performing ADL tasks. The receipt of formal care was strongly correlated with highest education (Fully adjusted prevalence ratio [PR] = 1.64; 95 % CI 1.05, 2.58) and with the highest assets level (PR = 2.24; 95 % CI 1.38, 3.64). Living with someone else was associated with provision of care (formal or informal) for those at the lowest and intermediate educational and assets levels, but not for the wealthiest. CONCLUSION: Despite worse physical functioning, older Brazilians in worse socioeconomic conditions are much less likely to receive needed help in performing ADL tasks.

Components of the metabolic syndrome and depressive symptoms in community-dwelling older people: the Bambuí Cohort Aging Study

Objective: To investigate the moderating effect of an increasing number of clustered metabolic syndrome (MetS) components on the association between MetS and depressive symptoms in a population-based cohort of older adults in Brazil. Methods: This analysis used data from the Bambuí Cohort Aging Study. Participants in this cross-sectional study comprised 1,469 community-dwelling older people aged ≥ 60 years. Analyses were performed to assess both the association between depressive symptoms and each individual MetS component and the association between depressive symptoms and clustering of an increasing number of MetS components. Results: High triglyceride level was the individual component that showed the strongest association with depressive symptoms (odds ratio [OR]: 1.47; 95% confidence intervals [95%CI] 1.19-1.81; p < 0.0001). Only the presence of three MetS components was associated with depressive symptoms (OR = 1.53; 95%CI 1.05-2.23; p = 0.025). No graded association was detected between increasing number of clustered MetS components and depressive symptoms. Conclusions: Increasing the number of MetS components did not impact the association with depressive symptoms. The association between high triglyceride level and depressive symptoms highlights the relevance of lipid metabolism abnormalities for the emergence of depressive symptoms in older adults.

Components of the metabolic syndrome and depressive symptoms in community-dwelling older people: the Bambuí Cohort Aging Study.

OBJECTIVE: To investigate the moderating effect of an increasing number of clustered metabolic syndrome (MetS) components on the association between MetS and depressive symptoms in a population-based cohort of older adults in Brazil. METHODS: This analysis used data from the Bambuí Cohort Aging Study. Participants in this cross-sectional study comprised 1,469 community-dwelling older people aged ≥ 60 years. Analyses were performed to assess both the association between depressive symptoms and each individual MetS component and the association between depressive symptoms and clustering of an increasing number of MetS components. RESULTS: High triglyceride level was the individual component that showed the strongest association with depressive symptoms (odds ratio [OR]: 1.47; 95% confidence intervals [95%CI] 1.19-1.81; p < 0.0001). Only the presence of three MetS components was associated with depressive symptoms (OR = 1.53; 95%CI 1.05-2.23; p = 0.025). No graded association was detected between increasing number of clustered MetS components and depressive symptoms. CONCLUSIONS: Increasing the number of MetS components did not impact the association with depressive symptoms. The association between high triglyceride level and depressive symptoms highlights the relevance of lipid metabolism abnormalities for the emergence of depressive symptoms in older adults.

Genomic Ancestry, Self-Rated Health and Its Association with Mortality in an Admixed Population: 10 Year Follow-Up of the Bambui-Epigen (Brazil) Cohort Study of Ageing.

BACKGROUND: Self-rated health (SRH) has strong predictive value for mortality in different contexts and cultures, but there is inconsistent evidence on ethnoracial disparities in SRH in Latin America, possibly due to the complexity surrounding ethnoracial self-classification. MATERIALS/METHODS: We used 370,539 Single Nucleotide Polymorphisms (SNPs) to examine the association between individual genomic proportions of African, European and Native American ancestry, and ethnoracial self-classification, with baseline and 10-year SRH trajectories in 1,311 community dwelling older Brazilians. We also examined whether genomic ancestry and ethnoracial self-classification affect the predictive value of SRH for subsequent mortality. RESULTS: European ancestry predominated among participants, followed by African and Native American (median = 84.0%, 9.6% and 5.3%, respectively); the prevalence of Non-White (Mixed and Black) was 39.8%. Persons at higher levels of African and Native American genomic ancestry, and those self-identified as Non-White, were more likely to report poor health than other groups, even after controlling for socioeconomic conditions and an array of self-reported and objective physical health measures. Increased risks for mortality associated with worse SRH trajectories were strong and remarkably similar (hazard ratio ~3) across all genomic ancestry and ethno-racial groups. CONCLUSIONS: Our results demonstrated for the first time that higher levels of African and Native American genomic ancestry--and the inverse for European ancestry--were strongly correlated with worse SRH in a Latin American admixed population. Both genomic ancestry and ethnoracial self-classification did not modify the strong association between baseline SRH or SRH trajectory, and subsequent mortality.

Physical activity and all-cause mortality among older Brazilian adults: 11-year follow-up of the Bambuí Health and Aging Study.

OBJECTIVE: To investigate the association between physical activity (eg, energy expenditure) and survival over 11 years of follow-up in a large representative community sample of older Brazilian adults with a low level of education. Furthermore, we assessed sex as a potential effect modifier of this association. MATERIALS AND METHODS: A population-based prospective cohort study was conducted on all the ≥60-year-old residents in Bambuí city (Brazil). A total of 1,606 subjects (92.2% of the population) enrolled, and 1,378 (85.8%) were included in this study. Type, frequency, and duration of physical activity were assessed in the baseline survey questionnaire, and the metabolic equivalent task tertiles were estimated. The follow-up time was 11 years (1997-2007), and the end point was mortality. Deaths were reported by next of kin during the annual follow-up interview and ascertained through the Brazilian System of Information on Mortality, Brazilian Ministry of Health. Hazard ratios (95% confidence intervals [CIs]) were estimated by Cox proportional-hazard models, and potential confounders were considered. RESULTS: A statistically significant interaction (P<0.03) was found between sex and energy expenditure. Among older men, increases in levels of physical activity were associated with reduced mortality risk. The hazard ratios were 0.59 (95% CI 0.43-0.81) and 0.47 (95% CI 0.34-0.66) for the second and third tertiles, respectively. Among older women, there was no significant association between physical activity and mortality. CONCLUSION: It was possible to observe the effect of physical activity in reducing mortality risk, and there was a significant interaction between sex and energy expenditure, which should be considered in the analysis of this association in different populations.

The association between nutritional status and cognitive impairment in brazilian community-dwelling older adults assessed using a range of anthropometric measures - the Bambui study / Associação entre o estado nutricional avaliado por várias medidas antropométricas e o comprometimento cognitivo em idosos brasileiros vivendo em comunidade - projeto Bambuí

In most studies, body mass index (BMI) has been used as the main measurement of nutritional status. However, BMI does not differentiate between body fat and muscle mass. OBJECTIVE: To investigate the association between nutritional status and cognitive impairment in a population of Brazilian elderly. METHODS: Participants (n=1,496) from the Bambuí Cohort Study of Aging were selected based on the results for the two variables nutritional status and cognitive impairment (MMSE score). Gender, age, education, lifestyle, ApoE, chronic diseases, depressive symptoms, current use of hypnotic or sedative medication and functional disability were used as confounding factors for adjusting the logistic regression. RESULTS: Cognitive impairment was associated with lower BMI (OR: 0.91; CI: 0.86-0.95), waist circumference (OR: 0.97; CI: 0.95-0.99), triceps skinfold thickness (OR: 0.92; CI: 0.89-0.96) among the younger participants (60-69 years), while lower arm muscle circumference (OR: 0.88; CI: 0.80-0.98) and corrected arm muscle area (OR: 0.96; CI: 0.93-0.99) were associated with cognitive impairment among the older participants (70 years and over). CONCLUSION: There was a difference of association between anthropometric measures and cognitive impairment after stratifying by age group. In the group aged between 60 and 69, cognitive impairment was associated with measures related to fat mass, while in the group aged over 70, cognitive impairment was associated with measures related to muscle mass. This finding suggests that investigation of nutritional status in the elderly using anthropometric measures should not be restricted only to the use of BMI, and should also, differ according to age.

Na maioria dos estudos o índice de massa corporal (IMC) é usado como a principal medida de avaliação do estudo nutricional. Entretanto, o IMC não apresenta capacidade de diferenciar a gordura corporal da massa muscular. OBJETIVO: Investigar a associação do estado nutricional e o comprometimento cognitivo na população idosa de Bambuí. MÉTODOS: Participaram do estudo 1496 idosos que responderam simultaneamente as variáveis do estado nutricional e o comprometimento cognitivo (avaliado através do escore do MMSE). As seguintes variáveis: sexo, idade, educação, estilo de vida, ApoE, doenças crônicas, sintomas depressivos, uso de medicamentos hipnóticos e sedativos e incapacidade funcional foram utilizadas como fatores de confusão na regressão logística multivariada. RESULTADOS: O comprometimento cognitivo foi associado com os baixos valores de: IMC (OR: 0.91; CI: 0.86-0.95), circunferência da cintura (OR: 0.97; CI: 0.95-0.99), dobra cutânea triciptal (OR: 0.92; CI: 0.89-0.96) entre os idosos mais jovens (60-69 anos). Enquanto que baixos valores da circunferência (OR: 0.88; CI: 0.80-0.98) e da area muscular do braço corrigida (OR: 0.96; CI: 0.93-0.99) foram associados com o comprometimento cognitivo entre os idosos mais velhos (70 anos ou mais).CONCLUSÃO: Existe uma diferença entre a associação das medidas antropométricas e o comprometimento cognitivo após a estratificação por idade. Nos participantes entre 60 a 69 anos, o comprometimento cognitivo foi associado a medidas relacionadas com o tecido adiposo enquanto que no grupo com 70 anos ou mais, o comprometimento cognitivo foi associado a medidas relacionadas com a massa muscular. Esses achados sugerem que a investigação do estado nutricional dos idosos não se deve restringir somente ao IMC, sendo necessário variações devido a idade.

The association between nutritional status and cognitive impairment in brazilian community-dwelling older adults assessed using a range of anthropometric measures - the Bambui study.

In most studies, body mass index (BMI) has been used as the main measurement of nutritional status. However, BMI does not differentiate between body fat and muscle mass. OBJECTIVE: To investigate the association between nutritional status and cognitive impairment in a population of Brazilian elderly. METHODS: Participants (n=1,496) from the Bambuí Cohort Study of Aging were selected based on the results for the two variables nutritional status and cognitive impairment (MMSE score). Gender, age, education, lifestyle, ApoE, chronic diseases, depressive symptoms, current use of hypnotic or sedative medication and functional disability were used as confounding factors for adjusting the logistic regression. RESULTS: Cognitive impairment was associated with lower BMI (OR: 0.91; CI: 0.86-0.95), waist circumference (OR: 0.97; CI: 0.95-0.99), triceps skinfold thickness (OR: 0.92; CI: 0.89-0.96) among the younger participants (60-69 years), while lower arm muscle circumference (OR: 0.88; CI: 0.80-0.98) and corrected arm muscle area (OR: 0.96; CI: 0.93-0.99) were associated with cognitive impairment among the older participants (70 years and over). CONCLUSION: There was a difference of association between anthropometric measures and cognitive impairment after stratifying by age group. In the group aged between 60 and 69, cognitive impairment was associated with measures related to fat mass, while in the group aged over 70, cognitive impairment was associated with measures related to muscle mass. This finding suggests that investigation of nutritional status in the elderly using anthropometric measures should not be restricted only to the use of BMI, and should also, differ according to age.

Na maioria dos estudos o índice de massa corporal (IMC) é usado como a principal medida de avaliação do estudo nutricional. Entretanto, o IMC não apresenta capacidade de diferenciar a gordura corporal da massa muscular. OBJETIVO: Investigar a associação do estado nutricional e o comprometimento cognitivo na população idosa de Bambuí. MÉTODOS: Participaram do estudo 1496 idosos que responderam simultaneamente as variáveis do estado nutricional e o comprometimento cognitivo (avaliado através do escore do MMSE). As seguintes variáveis: sexo, idade, educação, estilo de vida, ApoE, doenças crônicas, sintomas depressivos, uso de medicamentos hipnóticos e sedativos e incapacidade funcional foram utilizadas como fatores de confusão na regressão logística multivariada. RESULTADOS: O comprometimento cognitivo foi associado com os baixos valores de: IMC (OR: 0.91; CI: 0.86-0.95), circunferência da cintura (OR: 0.97; CI: 0.95-0.99), dobra cutânea triciptal (OR: 0.92; CI: 0.89-0.96) entre os idosos mais jovens (60-69 anos). Enquanto que baixos valores da circunferência (OR: 0.88; CI: 0.80-0.98) e da area muscular do braço corrigida (OR: 0.96; CI: 0.93-0.99) foram associados com o comprometimento cognitivo entre os idosos mais velhos (70 anos ou mais). CONCLUSÃO: Existe uma diferença entre a associação das medidas antropométricas e o comprometimento cognitivo após a estratificação por idade. Nos participantes entre 60 a 69 anos, o comprometimento cognitivo foi associado a medidas relacionadas com o tecido adiposo enquanto que no grupo com 70 anos ou mais, o comprometimento cognitivo foi associado a medidas relacionadas com a massa muscular. Esses achados sugerem que a investigação do estado nutricional dos idosos não se deve restringir somente ao IMC, sendo necessário variações devido a idade.

[A population-based study on factors associated with functional disability among older adults in the Great Metropolitan Belo Horizonte, Minas Gerais State, Brazil]. / Estudo de base populacional dos fatores associados à incapacidade funcional entre idosos na Região Metropolitana de Belo Horizonte, Minas Gerais, Brasil.

This study assessed factors associated with functional disability in old age in Greater Metropolitan Belo Horizonte, Minas Gerais State, Brazil. The theoretical model was based on predisposing (socio-demographic), extra-individual (social support, use of health services), and intra-individual factors (health conditions). The study was conducted in a probabilistic sample of 1,786 subjects aged > 60 years. The dependent variable was disability defined as mild (some difficulty) or severe (total dependence) in performing at least one basic activity of daily living. Prevalence of disability was 16% (8% mild; 8% severe). Age and worse self-rated health were independently and positively associated with mild and severe disability. Self-reported hypertension and arthritis were associated with mild disability, while diabetes and stroke were associated with severe disability. Severe disability was independently and negatively associated with number of visits by friends in the previous 30 days. According to the results, chronic conditions associated with disability in the study population are preventable, and severely disabled elderly had less extra-family social support.

Estudo de base populacional dos fatores associados à incapacidade funcional entre idosos na Região Metropolitana de Belo Horizonte, Minas Gerais, Brasil / A population-based study on factors associated with functional disability among older adults in the Great Metropolitan Belo Horizonte, Minas Gerais State, Brazil

O objetivo deste trabalho foi determinar os fatores associados à incapacidade funcional entre idosos na Região Metropolitana de Belo Horizonte, Minas Gerais, Brasil. O modelo teórico adotado foi baseado em características predisponentes (sócio-demográficas), fatores extra-individuais (apoio social, uso de serviços de saúde) e intra-individuais (condições de saúde). Participaram do estudo 1.786 idosos (> 60 anos) selecionados por meio de amostra probabilística. A variável dependente foi a incapacidade funcional, definida como incapacidade leve ou moderada (alguma dificuldade) e grave (total dependência) para realizar atividades da vida diária. A prevalência da incapacidade foi de 16 por cento (8 por cento leve e 8 por cento grave). Idade e pior auto-avaliação da saúde apresentaram associações positivas e independentes com ambos os níveis de incapacidade. Hipertensão e artrite apresentaram associações com incapacidade leve ou moderada, enquanto diabetes e acidente vascular cerebral apresentaram associações com incapacidade grave. Associação negativa com incapacidade grave foi observada para visita de amigos nos últimos trinta dias. Esses resultados mostram que as condições crônicas associadas à incapacidade na população estudada são passíveis de prevenção e que o apoio social externo à família é menor em idosos com incapacidade grave.

This study assessed factors associated with functional disability in old age in Greater Metropolitan Belo Horizonte, Minas Gerais State, Brazil. The theoretical model was based on predisposing (socio-demographic), extra-individual (social support, use of health services), and intra-individual factors (health conditions). The study was conducted in a probabilistic sample of 1,786 subjects aged > 60 years. The dependent variable was disability defined as mild (some difficulty) or severe (total dependence) in performing at least one basic activity of daily living. Prevalence of disability was 16 percent (8 percent mild; 8 percent severe). Age and worse self-rated health were independently and positively associated with mild and severe disability. Self-reported hypertension and arthritis were associated with mild disability, while diabetes and stroke were associated with severe disability. Severe disability was independently and negatively associated with number of visits by friends in the previous 30 days. According to the results, chronic conditions associated with disability in the study population are preventable, and severely disabled elderly had less extra-family social support.