Artigos

Conteúdo (article):

Social inequalities in indicators of active aging: a population-based study

Autores:
Neuciani Ferreira da Silva Sousa1
1Departamento de Saúde Coletiva, Instituto de Saúde Coletiva, Universidade Federal de Mato Grosso, Cuiabá, Mato Grosso, Brasil.
E-mail: neuciani@yahoo.com.br
ORCID: https://orcid.org/0000-0002-7694-0811

Margareth Guimarães Lima2
Departamento de Saúde Coletiva, Faculdade de Ciências Médicas, Universidade Estadual de Campinas, Campinas, São Paulo, Brasil
E-mail: margarethglima@gmail.com
ORCID: https://orcid.org/0000-0001-6996-0745

Marilisa Berti de Azevedo Barros2
2 Departamento de Saúde Coletiva, Faculdade de Ciências Médicas, Universidade Estadual de Campinas, Campinas, São Paulo, Brasil
E-mail: marilisa@unicamp.br
ORCID: https://orcid.org/0000-0003-3974-195X


Resumo

O objetivo deste estudo foi analisar desigualdades em indicadores de envelhecimento ativo, segundo raça/cor, escolaridade, renda e posse de plano de saúde entre 986 idosos participantes do Inquérito de Saúde de Campinas, São Paulo - 2014/15. Estimaram-se as prevalências de participação dos idosos em 11 atividades e as razões de prevalência foram calculadas pela regressão de Poisson. Os resultados revelaram que brancos e negros participavam de forma semelhante em todas as atividades da dimensão social, porém na atividade física realizada no trabalho se observou predomínio de negros (14,1% versus 8,2%) e no uso da internet se constatou maior prevalência de brancos (RP = 2,11). Entre os idosos que possuíam maior escolaridade, maior renda e posse de plano de saúde foram observadas maiores prevalências de participação em atividades físicas de lazer, uso da internet, realização de cursos e atividades sociais, exceto cultos religiosos. Os resultados revelam que os idosos com maior acúmulo de recursos educacionais e financeiros tem maior acesso às atividades que são reconhecidamente associadas à saúde e bem-estar. O estudo também identificou importantes demandas para o SUS, pois a população que depende exclusivamente deste sistema apresentou menor participação em atividades benéficas à saúde.

Palavras-chave:
Desigualdade social; Disparidades nos Níveis de Saúde; Envelhecimento; Idoso



Abstract
The objective of this study was to analyze inequalities in active aging indicators according to race/skin color, level of education, income, and possession of health insurance among 986 older people who participated in the 2014/15 Campinas Health Survey. We estimated the prevalence of participation in 11 activity domains using Poisson regression. The findings reveal similar levels of participation among white and black people in all the domains of the social dimension. The prevalence of work-related physical activity was higher among black people (14.1% compared to 8.2% in white people) and the prevalence of internet use was higher among white people (PR = 2.11). The prevalence of participation in leisure time physical activity, internet use, courses, and in all domains of the social dimension except attendance at religious services was higher among respondents in the highest educational and income groups and among those with health insurance. The findings reveal that older people with a higher income and higher level of education are more likely to participate in activities associated with better health and well-being. The study also shows that older people place a significant demand on Brazil’s public health system since individuals who depend exclusively on public health services tend to participate less in activities that are shown to promote health benefits.
Key words: Social Inequality; Disparities in Health Status; Aging; Older People.




Introduction
The relationship between aging and active living dates back to the 1950s in the United States and the activity theory1. However, it was the World Health Organization (WHO) that transformed the idea of “active aging” into a concept within global politics2 enshrined in the 2002 Madrid International Plan of Action on Ageing 3,4. As expected, the concept of active aging promoted by the WHO emphasizes the relationship between activity and health2,3 and – since it was developed in a European context – the participation and inclusion of older people as full citizens. It also focuses on a broader range of activities than those normally associated with active aging in the US context, such as productive activities5.
In the WHO approach, the term "active" refers to continuing participation in social, economic, cultural, physical, political, and civic affairs2. In other words, the definition considers economic participation and other forms of unpaid and non-productive participation, encompassing both formal and informal activities requiring physical or mental effort2,3. From the WHO perspective, active aging is not simply a choice, but rather a right bound to opportunities for health, participation, security, and lifelong learning6. This is because the possibility of autonomous choice is affected by inequalities, different life experiences, and oppressive social and economic conditions7. Thus, the risk of this strategy becoming coercive can be avoided if policy takes on an enabling role and responds to gender, race, class, cultural, and other differences accumulated over the life course.
In view of the above, research on active aging should pay special attention to understanding how social differences can hamper or differentiate active living8,9. However, studies in this area have devoted more effort to analyzing the association between participation in activities and health-related outcomes, such as subjective well-being, physical and emotional health, and survival10. Despite the importance of studies focusing on health, active aging approaches that disregard social differences run the risk of driving social exclusion and health inequalities, because not everybody is able to adhere to the model to an equal extent8. In this respect, studies show that there is a positive relation between socioeconomic status and participation in sociocultural activities11,12,13 and paid work14,15; however, the relation between socioeconomic status and physical activity, for example, remains unclear and may vary according to domain16,17. Moreover, the relationship between socioeconomic status and participation in learning or intellectual activities18,19 and the influence of race/skin color on active participation in society related to the direct effects of racial discrimination and the indirect effects of socioeconomic differences have been little explored. Finally, there is a lack of national and international studies on the role health services play in promoting active living across multiple domains.
Given that research in this area has paid little attention to social inequalities in active aging8,9 and that Brazilian studies exploring inequalities across multiple dimensions of active aging are limited to examining gender and age differences20,21, the aim of the present study was to analyze inequalities in indicators of active aging across social, physical, intellectual, and work dimensions according to race/skin color, level of education, income, and possession of private health insurance in the elderly population of a municipality in the southeast of Brazil.

Methods
A cross-sectional study was conducted using data from the 2014/2015 Campinas Health Survey (ISACAMP, acronym in Portuguese). ISACAMP examines patterns, trends, and social disparities across multiple dimensions of health among people aged 10 years and over living in private households in urban areas in the municipality of Campinas in the State of São Paulo.
The ISACAMP used a multistage cluster sampling design. First, the population was divided into five strata corresponding to the city’s health districts: West, Northwest, North, Southeast, and South. Fourteen census tracts were then randomly selected from each stratum, resulting in 70 primary sampling units. All households within the sampling units were listed and then sampled. For the purposes of this study, the population was divided into three age groups: 10 to 19 years, 20 to 59 years, and 60 years and over. The minimum overall sample size was estimated to be 3,400 individuals, subdivided into 1,000 adolescents, 1,400 adults, and 1,000 older people. These numbers were defined based on a population proportion estimate of 0.50 with a margin of error of five percentage points, 95% confidence interval (95% CI), and design effect of 2. To obtain this sample size and considering an expected non-response rate of 27%, 22%, and 20%, respectively, for each age group based on previous surveys, the following numbers of households were randomly selected: 3,119 for adolescents, 1,029 for adults, and 3,157 for older people.
Within each household, all residents in the age group for which the household had been sampled were interviewed. This type of design is similar in terms of accuracy and more cost effective than designs that select only one person from each selected household22. The present study involved only people aged 60 years and over.
Each individual was given a final weight calculated by multiplying the design weight by the non-response weight and by the post-stratification weight using the age and sex distribution based on population projections performed by SEADE, the State of São Paulo’s data analysis system.
The 2014/2015 ISACAMP collected data using a pre-coded questionnaire containing mostly closed-ended questions in three thematic blocks. Data was collected by trained interviewers via face-to-face interviews optimized by the use of a tablet. The interviewers participated in a theoretical and practical training program that discussed expected behavior during interviews, the specific details of each question, how to enter data into the tablet, and the content addressed by the questionnaire. The interviewers were also provided with an instruction manual.
Active aging was defined in accordance with the WHO’s2 definition of the word “active”, which refers to participation in multi-dimensional social, physical, cultural, intellectual, economic, civic, and political activities. Questions concerning the following four dimensions of active aging were selected from the questionnaire:
Social dimension – respondents were asked about their current participation in four domains: a) participation in the family circle, using the question “Do you receive visits from or visit your friends and family?”; b) sociocultural activities, using the question “Do you participate in cultural or social activities (for example: cinema, theater, senior centers, bingo, clubs, older people\'s dance groups, parties, among others)?”; c) volunteering or participation in groups/associations, using the question “Do you do voluntary work or are you part of a sports, cultural, philanthropic, political, or religious group/association?”; d) religion, using the question “Do you attend a religious service at least once a week?”.
Physical activity dimension – this dimension was assessed using questions from the long version of the International Physical Activity Questionnaire (IPAQ)23. This instrument measures the weekly time spent doing moderate and/or vigorous physical activity in the following domains: work-related physical activity, transport-related physical activity, domestic activity, and leisure time physical activity. Level of physical activity in each of these domains is classified based on a physical activity score expressed in minutes per week. The score is computed by adding the minutes spent on moderate activities and the minutes spent on vigorous activities multiplied by two, thus taking into account the intensity of each activity as recommended by the WHO24. A score of over 150 minutes per week is the cut-off point used to classify individuals as active in terms of overall physical activity (regardless of domain) and in each specific domain. The differentiation of physical activity by domain is important for identifying and understanding which individual characteristics are associated with physical activity levels16.
Intellectual dimension – this dimension was assessed using two variables (internet use and participation in courses). Internet use was assessed based on “yes”/“no” answers to the following question: “Do you use the internet?”. Participation in courses was confirmed with a “yes” answer to either of the following questions: “Are you currently doing a course at a school or university?” and “Are you currently doing another type of course such as computing, languages, dance, arts etc.?”.
Work dimension – this dimension assesses participation in the following domains: a) paid work, assessed using the question “Do you currently carry out paid work or help a family member in his/her work?”; and b) work in retirement.
The following demographic and socioeconomic variables were used to characterize the study population: sex (female and male); age (60-69 years, 70-79 years, and 80 years and over); self-declared race/skin color (white, black/brown, other); level of education in years of schooling (0 to 3, 4 to 7, 8 or more); monthly family income in minimum salaries per capita (<1, 1 to 3, >3), and possession of private insurance.
The prevalence of each active aging domain was estimated according to self-declared race/skin color, level of education, income, and possession of private health insurance. Given the small number of people in the “other” race/skin color group (indigenous and yellow-skinned people) and group heterogeneity, only white and black/brown people were included in the analysis. Proportions were compared using Pearson\'s chi-squared test with Rao-Scott adjustment, adopting a significance level of p<0.05. The prevalence ratios for each active aging indicator and their respective 95% confidence intervals were calculated using Poisson regression and adjusted for sex and age to eliminate confounding. The analyses of race/skin color were also adjusted for level of education to determine whether potential associations with active aging indicators were explained by socioeconomic factors or by the direct effect of racial discrimination on participation in the activities.
The design effect was taken into account in all analyses using the Stata 14 survey module (Stata Corp., College Station, United States).
ISACAMP was approved by the Ethics Committee of the Faculty of Medical Sciences at the University of Campinas (approval number 409.714, 30 September 2013).

Results
A total of 1,168 individuals aged 60 years and over were identified in the selected households; however, the non-response rate was 14% and losses for other reasons amounted to 1.5%, resulting in a final sample of 986 older people. Non-response rates varied between 6.3 and 22.6% across health districts, tending to be higher in areas with higher socioeconomic status: 22.6% and 18.6% in the higher status West and North districts, respectively; 13.7% and 13.1% in the lower status South and Southeast districts; and 6.3% in the Northwest district, in which residents have the lowest socioeconomic status. Post-stratification weights were used to reduce the effect of these differences.
Table 1 shows that the majority of the respondents were female (57.6%), in the 60 to 69 years age group (56.7%), white (71.2%), had less than 8 years of schooling (65.3%), had a per capita family income of 1 to 3 minimum salaries (55.3%), and did not posses private insurance (52.9%).
The analysis of the active aging profile by race/skin color (Table 2) showed similarities between blacks and whites in participation across all domains of the social dimension. After adjustment for sex and age, the prevalence of work-related physical activity was higher among black people (14.1% compared to 8.2% in white people) and the prevalence of internet use was higher among white people (PR = 2.11). These differences were not maintained after adjustment for level of education. However, this same adjustment showed that the prevalence of participation in a course and carrying out paid work was lower among white people.
Table 3 shows the prevalence of active aging indicators according to level of education. Differences between the educational groups were found across all domains of the social dimension except attendance at religious services. A comparison of the highest and lowest educational groups shows that the prevalence of participation in the family circle (PR = 1.10), sociocultural activities (PR = 2.30), and voluntary work/associations (PR = 1.89) was highest among respondents with at least 8 years of schooling. No statistically significant difference was found between the lowest and middle educational group in the physical activity dimension. However, a comparison of the highest and lowest educational groups shows that individuals with a higher level of education were more active in the leisure time physical activity domains and less active in the work-related physical activity domain (RP = 1.95 and RP = 0.62, respectively). The most striking differences between the lowest and highest educational group were found in the intellectual dimension; however, the estimates in this dimension were less precise. In the work dimension, the only statistically significant difference between the groups was that the prevalence of work in retirement was higher in the middle educational group than in the lowest educational group (PR = 2.51).
Table 4 shows that respondents with a higher income participated more in the family circle, sociocultural activities, and voluntary work/associations domains of the social dimension (PR = 1.13, PR = 2.20, and PR = 2.14, respectively) and in both domains of the intellectual dimension. In the work dimension, no statistically significant differences across income groups were found in the paid work domain; however, the prevalence of work in retirement was greater among those in the highest income group.
Table 5 shows that participation was greater among the group with private health insurance in all domains of the social dimension except attendance at religious services and in both domains of the intellectual dimension (internet use, PR = 3.75 and participation in courses, PR = 4.11). In the physical activity dimension, similarities were found between the two groups in the work-related physical activity, transport-related physical activity, and domestic activity domains; however, the prevalence of participation in leisure time physical activity was higher among respondents who possessed health insurance than those who depended exclusively on Brazil’s national health system, the Sistema Único de Saúde - SUS (PR = 1.89). No statistically significant differences were found across the groups in the work dimension.



Discussion
The findings of this study show that white people and black people participate equally in all domains except work-related physical activity and internet use, where prevalence was higher among blacks and whites, respectively. In general, respondents with a higher level of education, higher income, and private health insurance show higher prevalence of participation in the social, intellectual, and physical activity dimensions. In the physical activity dimension, the difference in prevalence is particularly significant in the leisure time physical activity domain. Participation in religious services, transport-related and domestic physical activity, and paid work was similar across groups irrespective of race/color, level of education and income, and possession of private health insurance.
The findings regarding race/skin color and the social dimension are positive because they show that white and black people participate to a similar extent in the domains considered by the study, despite the historical accumulation of disadvantages by the black population in relation to the white population25,26,27 clearly associated with lower levels of participation in different dimensions of social life3. On the other hand, the differences between black and white people observed by this study show that the reduction of inequalities between the white and black population in other dimensions of life remains a major challenge in Brazil28. The first difference concerns work-related physical activity. This indicator is a marker of social inequalities29 and is not always associated with beneficial health effects30. A systematic review revealed that individuals with lower socioeconomic status and in occupations with low social prestige showed higher levels of work-related physical activity29. The higher level of work-related physical activity among black people therefore suggests that they occupy positions that require less professional qualifications and are more physically demanding. This reflects the reality of the Brazilian job market, where black people are the majority in the sectors with the worst working conditions – agriculture, the construction industry, and domestic work – and in precarious work, where labor rights are not protected25,28. The racial difference observed in internet use is disturbing since older people are already the population group in Brazil with lowest internet access18. This situation differs from that of the United States, for example, where digital inequality has decreased considerably over the last 18 months, with internet use by white and black people rising from 53% and 38%, respectively, in 2000, to 89% and 87%, respectively, in 201831.
A large part of the racial differences identified by the present study may be due to educational inequality between black and white people26,32. In this respect, after adjustment for level of education, differences in work-related physical activity and internet use disappeared and other differences were observed, showing that black people were more likely to be doing a course or carrying out paid work. However, this is not necessarily positive for black people and further in-depth study is needed to explore possible intrinsic differences in the activities in question.
With regard to socioeconomic status – measured according to level of education and income – the findings show that the prevalence of participation in the social dimension (participation in the family circle, sociocultural activities, and volunteering and work/associations) was higher in the groups with higher status. Other studies also showed that higher socioeconomic status was associated with a higher level of social participation11,12,13,14. One explanation for this association is that people with higher socioeconomic status have access to a range of resources (such as money, knowledge, prestige, and power) that contribute to healthy and active living33. However, social participation depends on both individual resources and the context of social inequality in which it is embedded. In countries with lower income inequality, for example, social services systems are more equitable, facilitating the participation of underprivileged groups34. Since deep inequalities persist in Brazil27, it is to be expected that more vulnerable segments of society face greater difficulty in participating in social activities.
With respect to physical activity, the findings show that level of education has a bidirectional effect, whereby the highest educational group shows a positive association with leisure time physical activity and an inverse association with work-related physical activity. The data also shows a direct association between income and overall physical activity, which is particularly significant in the leisure time physical activity domain. A systematic review observed considerable differences in the direction of inequalities in the physical activity domains, revealing an association between higher socioeconomic status and a higher level of leisure time physical activity, and between lower socioeconomic status and a higher level of work-related physical activity17. One explanation for this is that people with higher socioeconomic status are more likely to adhere to preventive programs and adopt healthy behaviors because they have more motivation and greater access to information and other resources such as sports clubs and gyms17,35. On the other hand, the inverse relation between level of education and work-related physical activity may be explained by the fact that people with a lower level of education are more likely to have jobs that have lower social prestige and are more physically demanding, leading to higher levels of energy expenditure29, as discussed above in relation of race/skin color.
With regard to the intellectual dimension, although the estimates are less precise, it is interesting to note that the groups that are more disadvantaged educationally and financially are significantly less likely to use the internet. There is a clear socioeconomic gradient in internet use in Brazil and the educational gradient is more pronounced than the income gradient, as shown in the present study18. These findings suggest that level of education is a key determinant of internet use since education potentiates the appropriation of rapidly changing technologies. This means that this indicator increasingly reflects social, economic, and cultural relations in the off-line world, including social inequalities36. Therefore, digital inclusion policies in Brazil should address not only investment in equipment, internet access in public places, and reduction in the cost of private internet access, but also continuous improvement in basic education so that all segments of society are able to explore, understand, and take ownership of the information available on the internet36,37.
The data also shows that the overall prevalence of participation in courses was low, revealing that the most educationally and financially privileged groups were more likely to engage in learning activities, reflecting the national reality32. An international study using data from 13 country studies and two cross-national analyses showed a relatively clear pattern across countries whereby those already better off in society are better able to access adult learning and obtain greater benefits in career progress. The study concludes that adult learning tends to reproduce and reinforce initial education resulting in educational selectivity, whereby people with a higher level of education are more likely to engage in other learning processes during the life course19. The findings of the present study therefore strengthen the argument for placing more emphasis on the complex relationship between social inequality and adult learning into old age19 to prevent education from increasingly becoming a commodity rather than a right.
The fact that there were no differences between the socioeconomic groups in participation in the labor market is surprising, given that other studies have shown a direct gradual association between socioeconomic status and work and a particularly pronounced association with level of education14,15. The relationship between financial resources and work is less clear since low-income individuals may have to work to maintain a minimum standard of living, while those with high incomes are more likely to have a higher level of education and, given that education is positively associated with employment, are more likely to remain employed15,38. However, this does not exclude the possibility that those with sufficient income to maintain a dignified and satisfactory standard of living opt to stop working as soon as possible in order to enjoy their old age and participate in other types of activities. It is important to note, however, that the findings of this study in relation to socioeconomic status and participation in the labor market should be interpreted with caution since the indicator used does not capture other inequalities between social groups in other aspects of the labor market14,15,38.
The data presented show that individuals who possessed private health insurance were more active in the social and intellectual dimensions and in the leisure time physical activity domain. This finding is particularly relevant because it illustrates that older people place a significant demand on the SUS. In this respect, service provision providers should focus not only on meeting the physical needs of older people, but also on targeting services that connect older people with their community39. Social participation is highly valued by older people and should be encouraged within the SUS in view of its potential for promoting physical and mental health and generating social benefits by increasing this group’s community contributions12. Encouraging volunteering and creating community groups that develop cultural and education activities are examples of measures that should be promoted within the SUS39, as provided by the National Older People’s Health Policy40. These actions are vital for promoting social and digital inclusion, optimizing social connections, developing new skills, and preserving cognitive functioning3. However, promoting active aging goes beyond the provision of basic health services and it is necessary to address the barriers that limit social participation, such as lack of leisure time due to the burden of family commitments, difficulties in getting around in urban areas, and lack of guidance on leisure activities41.
The findings also reveal inequalities in the physical activity dimension, particularly in participation in leisure time physical activity between individuals with and without private health insurance. This reinforces the importance of strategies such as the Health and Fitness Gym program and family health support units for improving access to body practices and physical activity42,43. In this respect, these spaces should broaden the scope of activities provided in order to promote the participation of subgroups that differ in terms of gender, age, health status, time available, and individual preferences44,42.
The present study has some limitations, such as the absence of information on frequency of participation in sociocultural activities and volunteering and groups/associations and on the reasons for and frequency of internet use. These details could provide a deeper insight into inequalities, because activities can have diverging effects on health and well-being depending on the context in which they are carried out. Another limitation is survival bias, given that individuals with lower socioeconomic status have a greater risk of premature death and are therefore less likely to be included in the study. This type of bias tends to reduce the strength of association45. Study strengths include: the use of a population-based sampling method and the size of the study sample, enabling the assessment of the majority of active aging indicators with a good level of precision; the quality of the data collected; and the use of indicators that have been little explored with the older population, thus adopting a multidimensional approach to active aging.
Final Considerations
In addition to discussing the challenges of promoting active aging, this study reveals the magnitude of social inequalities across multiple active aging indicators, thus providing a deeper insight into this question. The findings show that older people with a higher income and higher level of education are more likely to participate in activities associated with better health and well-being, particularly in the social and intellectual dimensions and leisure time physical activity domain. They also show that educational inequalities need to be addressed. As mentioned above, education is a key determinant of active aging, not only because it potentiates participation in activities, but also because it enhances quality of life as people age.
More specifically, the findings corroborate the positive relationship between socioeconomic status and participation in sociocultural activities, but contradict the positive association between socioeconomic status and paid work, suggesting the need for further research in this area. In addition, the study confirms that the relationship between socioeconomic status and physical activity varies according to domain, being positive for leisure time physical activity and negative for work-related physical activity. The findings also show differences in work-related physical activity and internet use between race/skin color groups, which may be attributed to the indirect effect of socioeconomic differences between these groups, rather than the direct effect of racial discrimination. Finally, the study shows that older people place a significant demand on the SUS since individuals who depend exclusively on public health services tend to participate less in activities that are shown to promote health benefits (intellectual and cultural activities, volunteering and participation in groups/associations, and leisure time physical activity). Promoting active aging should therefore go beyond the provision of basic health services to include strategies that foster social participation as a way of enhancing the health and well-being of older people.
The findings show that the promotion of active aging presupposes tackling social inequalities throughout the life course with a view to providing fairer solutions that are sensitive to differences across all segments of society and reduce the causes and extent of health inequalities.
Acknowledgements
We are grateful to: the State of São Paulo Research Foundation (FAPESP, acronym in Portuguese), for funding the ISACAMP; the Ministry of Health and Department of Health of Campinas City Council, for providing complementary funding for the ISACAMP; the Coordination of Improvement of Higher Education Personnel (CAPES, acronym in Portuguese), for the PhD scholarship provided to N.F.S.S.; and the National Council for Scientific and Technological Development (CNPq, acronym in Portuguese), for the research grant provided to M.B.A.B..

References

1. Havighrst RJ. Flexibility and the Social Roles of the Retired. Am J Sociol. 1954; 59:309-311.
2. World Health Organization. Active Ageing: a Policy Framework. Geneva: World Health Organization; 2002.
3. Centro Internacional de Longevidade Brasil. Envelhecimento Ativo: Um marco político em resposta à revolução da longevidade. Rio de Janeiro: Centro Internacional de Longevidade Brasil; 2015.
4. Walker A. A strategy for active ageing. Int Soc Secur Rev. 2002; 55:121-39.
5. Walker A. Active ageing: Realising its potential. Australas J Ageing. 2015; 34:2-8.
6. United Nations. General Assembly. A/RES/46/91 (Implementation of the International Plan of Action on Ageing and related activities), 16 December 1991.
7. Holstein MB, Minkler M. Self, Society, and the ‘‘New Gerontology’’. Gerontologist. 2003; 43: 787–96.
8. Timonen V. Beyond Successful and Active Ageing: A Theory of Model Ageing. Bristol, UK: Policy Press; 2016.
9. Barslund M, von Werder M, Zaid A. Inequality in active ageing: evidence from a new individual-level index for European countries. Ageing Soc 2017;1-27.
10. Adams KB, Leibbrandt S, Moon H. A critical review of the literature on social and leisure activity and wellbeing in later life. Ageing Soc. 2011; 31: 683–712.
11. van Groenou MB, Deeg DJH. Formal and informal social participation of the ‘young-old’ in The Netherlands in 1992 and 2002. Ageing Soc 2010; 30:445–65.
12. Galenkamp H, Gagliardi C, Principi A, Golinowska S, Moreira A, Schmidt AE, et al. Predictors of social leisure activities in older Europeans with and without multimorbidity. Eur J Ageing 2016; 13:129–43.
13. Vézina M, Crompton S. Volunteering in Canada. Canadian Social Trends 2012; 37- 55.
14. Ilinca S, Rodrigues R, Schmidt A, Zolyomi E. Gender and Social Class Inequalities in Active Ageing: Policy meets Theory. European Centre for Social Welfare Policy and Research: Vienna; 2016.
15. Uppal S. Labour market activity among seniors. Perspectives on Labour and Income. 2010; 11:5-18.
16. Notthoff N, Reisch P, Gerstorf D. Individual Characteristics and Physical Activity in Older Adults: A Systematic Review. Gerontology 2017; 63: 443-59.
17. Beenackers MA, Kamphuis CBM, Giskes K, Brug J, Kunst AE, Burdorf A, et al. Socioeconomic inequalities in occupational, leisure-time, and transport related physical activity among European adults: A systematic review. Int J Behav Nutr Phys Act 2012; 9:116.
18. Instituto Brasileiro de Geografia e Estatística. Pesquisa nacional por amostra de domicílios: acesso à internet e à televisão e posse de telefone móvel celular para uso pessoal 2015. Rio de Janeiro: IBGE, 2016.
19. Kilpi-Jakonen E, Vilhena DV, HP Blossfeld. Adult learning and social inequalities: Processes of equalisation or cumulative disadvantage? Int Rev Educ 2015; 61:529-46.
20. Ribeiro PCC, Neri AL, Cupertino APFB, Yassuda MS. Variabilidade no envelhecimento ativo segundo gênero, idade e saúde. Psicol Estud 2009; 14:501-9.
21. Sousa NFS, Lima MG, Cesar CLG, Barros MBA. Envelhecimento ativo: prevalência e diferenças de gênero e idade em estudo de base populacional. Cad Saúde Pública. 2018; 34:e00173317.
22. Alves MCGP, Escuder MML, Claro RM, Silva NN. Sorteio intradomiciliar em inquéritos de saúde. Rev Saúde Pública 2014; 48:86-93.
23. Matsudo S, Araújo T, Matsudo V, Andrade D, Andrade E, Oliveira LC et al. Questionário internacional de atividade física (IPAQ): Estudo de validade e reprodutibilidade no Brasil. Rev Bras Ativ Fís Saúde 2001; 6:5-18.
24. World Health Organization. Global recommendations on physical activity for health. Geneva: World Health Organization; 2010.
25. Instituto de Pesquisa Econômica Aplicada [et al.]. Retrato das desigualdades de gênero e raça. 4ª ed. Brasília: Ipea, 2011.
26. Instituto de Pesquisa Econômica Aplicada. Retrato das desigualdades de gênero e raça - 1995 a 2015. Brasília: Ipea, 2017.
27. Oxford Committee for Famine Relief Brasil. A distância que nos une: um retrato das desigualdades brasileiras. São Paulo: OXFAM Brasil, 2017.
28. Charão C. O longo combate às desigualdades raciais. Desenvolvimento 2011; (70): 22-31.
29. Smith L, McCourt O, Sawyer A, Ucci M, Marmot A, Wardle J, et al. A review of occupational physical activity and sedentary behaviour correlates. Occup Med (Lond) 2016; 66:185–92.
30. Abu-Omar K, Rütten A. Relation of leisure time, occupational, domestic, and commuting physical activity to health indicators in Europe. Prev Med 2008; 47:319–23.
31. Pew Research Center. Internet use by race/ethnicity. 2017.
32. Instituto Brasileiro de Geografia e Estatística. Educação 2016. Rio de Janeiro: IBGE, 2017.
33. Phelan JC, Link BG, Tehranifar P. Social Conditions as Fundamental Causes of Health Inequalities: Theory, Evidence, and Policy Implications. J Health Soc Behav 2010; 51(S): S28–S40.
34. Lancee B, van de Werfhorst HG. Income inequality and participation: A comparison of 24 European countries. Soc Sci Res 2012; 41:1166–78.
35. Del Duca GF, Rombaldi AJ, Knuth AG, Azevedo MR, Nahas MV, Hallal PC. Associação entre nível econômico e inatividade física em diferentes domínios. Rev Bras Ativ Fís Saúde 2009; 14:123-31.
36. van Deursen AJAM, van Dijk JAGM. The digital divide shifts to differences in usage. New Media Soc 2014; 16:507-26.
37. Mattos FAM, Chagas GJN. Desafios para a inclusão digital no Brasil. Perspect Ciênc Inf 2008; 13:67-94.
38. Wajnman S, Oliveira AM, Oliveira EL. Os idosos no mercado de trabalho: tendências e consequências. In: Camarano AA, organizadora. Os novos idosos brasileiros: muito além dos 60? Rio de Janeiro: Instituto de Pesquisa Econômica Aplicada; 2004. p. 453-79.
39. Douglas H, Georgiou A, Westbrook J. Social participation as an indicator of successful aging: an overview of concepts and their associations with health. Aust Health Rev 2017; 41: 455–62
40. Brasil. Portaria nº 2.528/MS de 19 de outubro de 2006. Aprova a Política Nacional de Saúde da pessoa idosa. Diário Oficial da União. Brasília: DF,20 de outubro de 2006.
41. Levasseur M, Généreux M, Bruneau JF, Vanasse A, Chabot E, Beaulac C, et al. Importance of proximity to resources, social support, transportation and neighborhood security for mobility and social participation in older adults: results from a scoping study. BMC Public Health 2015; 15:503.
42. Carvalho FFB, Nogueira JAD. Práticas corporais e atividades físicas na perspectiva da Promoção da Saúde na Atenção Básica. Ciênc Saúde Colet 2016; 21:1829-38.
43. Fernandes AP, Andrade ACZ, Costa DAS, Dias MAS, Malta DC, Caiaffa WT. Programa Academias da Saúde e a promoção da atividade física na cidade: a experiência de Belo Horizonte, MG, Brasil. Ciênc Saúde Colet 2017; 22:3903-14.
44. Costa TB, Ribeiro LHM, Neri AL. Prevalence of and factors associated with leisure-time physical activity in older adults from seven Brazilian cities: data from the FIBRA study. Rev Bras Ativ Fís Saúde 2015; 20:174-83.
45. Lima-Costa MF, Barreto SM. Tipos de estudos epidemiológicos: conceitos básicos e aplicações na área do envelhecimento. Epidemiol Serv Saude 2003; 12: 189-201.