Ageing abstracts

Strand organiser: Gloria Chepnego-Langat, University of Southampton 

Ageing: Health and Wellbeing Monday 7 September 1.30pm 

Socio-economic correlates of self-rated health among persons aged 50 or higher in SHARE waves 2 and 3
Georgia Verropoulou, Maria Zakynthinou, University of Piraeus 

Background: Self-rated health (SRH), though a self-reported, general indicator of health, is a strong predictor of morbidity and mortality. As all health indicators, SRH exhibits a strong socio-economic gradient. Aims: The main aim of the analysis is to assess the importance of socio-economic factors related to present conditions in predicting SRH and contrast them to factors related to SES in childhood. Data and methods: To achieve this aim, data from waves 2 and 3 of the Survey of Health, Ageing and Retirement in Europe (SHARE) have been used. Wave 2 has collected information on socio-demographic and health indicators among persons aged 50 or higher in 13 countries, covering Europe from East to West and North to South; wave 3 includes retrospective information related to childhood SES and health conditions. Both waves were carried out in 2006/2007. Combining these two data sources, using information on the 20,289 persons who responded in both waves and logistic regression models an assessment of the relative importance of present versus past socio-economic conditions on SRH is carried out, controlling for country of residence and present/past health. Results: The findings indicate that both past and present SES are significant predictors of SRH, including current wealth, income, educational attainment, whether one ‘can make ends meet’ as well as occupation of main breadwinner at age 10, whether one had experience periods of hunger in childhood and ‘number of books read’ during childhood. The importance of these factors differentiates when males and female are considered separately or in broad age-groups. 


How has ageing changed over the last two decades? Health expectancies from the Cognitive Function and Ageing Studies
Carol Jagger1,2, Pia Wohland 1,2, Louise Robinson 1,2, Antony Arthur 3, Carol Brayne 4, Fiona E Matthews 1,5, on behalf of the Medical Research Council Cognitive Function and Ageing Collaboration,  1Institute of Health and Society, Faculty of Medicine, Newcastle University, 2 Newcastle University Institute for Ageing, 3 School of Health Sciences, University of East Anglia, 4 Department of Public Health and Primary Care, Cambridge Institute of Public Health, Cambridge University,  5MRC Biostatistics Unit, Cambridge Institute of Public Health

Background: Life expectancy continues to increase but whether these are increases in healthy years is of profound importance worldwide. We investigate how a range of health expectancies have changed in England across two decades using identical study design and methods. Methods: Baseline data from the Cognitive Function and Ageing Studies on populations aged 65 years and over in three geographically defined centres across England (Cambridgeshire, Newcastle, Nottingham) provided prevalence estimates for three health measures: self-perceived health (SPH: excellent or good/fair/ poor); cognitive impairment (CI: moderate or severe/mild/none assessed by Mini-Mental State Examination score 0-17/18-25/26-30); and disability in activities of daily living (ADL) (none/mild/moderate or severe). Health expectancies for the three regions combined were calculated by Sullivan’s method. Results: Between 1991 and 2011 gains in life expectancy at age 65 (men:4·5 years, women:3·6 years) were accompanied by equivalent gains in years free of CI (men:4·2 years, 95%CI 4·2-4·3; women:4·4 years, 95%CI 4·3-4·5) and declines in years with mild and moderate/severe cognitive impairment. Smaller gains occurred in years in excellent/good SPH (men:3·8 years, 95%CI 3·5-4·1; women:3·0 years, 95%CI 2·7-3·4) and in disability-free years (men:2.6 years, 95%CI 2.3-2.9; women:0.5 years, 95%CI 0.2-0.9) due to more years with mild disability. Similar results were observed at age 85. Interpretation: Though years with CI have reduced over the last two decades in England reflecting a compression of CI morbidity, disability shows more of a dynamic equilibrium (increases in years with mild disability). These results have wide-ranging implications for health services and for extending working life. 


Ethnic inequalities in health in later life
Evandrou, M., Falkingham, J., Feng, Z. and Vlachantoni, A. Centre for Research on Ageing and ESRC Centre for Population Change, Faculty of Social and Human Sciences, University of Southampton 

Ethnic inequalities in health have been well known in the UK, however ethnic inequalities in health in later life are a relatively under-researched area. This paper explores ethnic inequalities in health among older people, controlling for social and economic disadvantage across ethnic groups in the UK. The first wave (2009-2011) of Understanding Society was analysed, which includes an Ethnic Minority Boost Sample. The two health outcomes used are the extent to which one’s health limits their typical activities, and self-rated health. Even after controlling for social and economic disadvantage across ethnic groups, elderly people from ethnic groups were still more likely than the White British to report that their health limits their typical activities and to report poor self-rated health. More specifically, Pakistani persons showed the poorest health outcomes. In the separate models for men and women, women from most ethnic groups were more likely than White British women to report that their health limits their typical activities; while, only Indian, Pakistani and Bangladeshi (South Asian) men showed higher odds than White British men in this respect. In term of self-rated health, Indian, Pakistani, Bangladeshi women and women from Other ethnic groups were more likely to report poor self-rated health than White British women, however, only Pakistani men and men from Other ethnic groups showed higher odds than White British men in this respect. Such results reflect the complexity of health inequalities among different ethnic groups in the UK, and the need to develop health policies which take into account differences in social and economic resources between different ethnic groups. 


The effect of Mindfulness Meditation on wellness in an ageing population.
Leigh Wilson1, Deborah Black2, 1Nan Tien Institute, NSW, Australia, 2University of Sydney 

Introduction Mindfulness meditation is used to reduce stress, alleviate anxiety and improve quality of life. The aim of this project was to investigate the effect of mindfulness meditation on self-reported health and wellbeing in a group of elderly residents living in self-care accommodation in Sydney, Australia. Method Twenty residents aged 65+ living in self-care retirement accommodation were screened at baseline for anxiety, stress and self-reported health using the SF12 and the Perceived Stress Scale (PSS). Participants were randomly allocated to either a 15 minute mindfulness meditation session once a week for 6 weeks and asked to practice 10 minutes of meditation daily or were asked to read a guide on how to improve wellness once each week for six weeks. Participants were assessed post intervention.

Results The mindfulness meditation group reported decreased stress and anxiety as measured on the SF12 and PSS, and reported better ability to sleep than those in the wellness guide group. Self-reported feelings of wellbeing increased in the mindfulness meditation group. Women in the meditation group were more likely than men to complete daily meditation as requested. All participants continued in the intervention for the full six weeks, although 3/10 in the wellness guide group did not read the relevant material every week.

Conclusions Mindfulness meditation is a useful tool to assist older people to reduce stress and anxiety and improve quality of life. Where possible, retirement service providers should encourage the provision of mindfulness meditation to assist in improving quality of life in older residents. 


Ageing: Ageing populations Tuesday 8 September 11.00 am 

The recent evolution of the crisis in Portugal – the effect of demographics on the labour market
Paula Albuquerque , SOCIUS/ CSG/ ISEG /University of Lisbon 

A labour market downturn affects different age groups in different degrees and is itself affected by demography. In this paper, we look at the relation between population ageing and labour market dynamics, focusing on the recent evolution of the Portuguese labour market. It is an interesting case study because it has been marked by a deep debt crisis and appertains to an aged society. Using quarterly data from the Portuguese Labour Force Survey, 2011-2014, we analyse the evolution of the unemployment rate - differentiated by age categories - looking simultaneously at the evolution of employment, unemployment and inactivity. We inspect transitions out of employment and out of unemployment to understand the driving forces of the changes in the labour market. We investigate the direct effect of population ageing on the evolution of the labour market outcomes, using a decomposition that measures the contribution to employment, to unemployment and to inactivity of observed changes in the age composition of the population, while holding the employment-to-population ratio, the unemployment-to-population ratio or the share of inactive population of each sex-age group constant. We reproduce the analysis with the numbers of hours worked instead of the numbers of persons in or out of work. We show that the crisis would have had an even greater negative impact on the unemployment rate if demographics were not exerting its pressure. 


The effect of parenthood on the pension protection of first and second-generation migrant women in Belgium
David De Wachter1 , Karel Neels1, Dimitri Mortelmans1, Tine Kil1, Lise Cordeel2, Nina Donvil2, Hans Peeters2, Nele Havermans2, 1Centre for Longitudinal & Life Course Studies, University of Antwerp, 2Centre for Sociological Research, University of Leuven

Pension protection is high on the agenda in western societies confronted with rapid population ageing. Especially the pension protection of migrant populations is a cause of concern since pension protection in Belgium is based directly or indirectly on (previous) labour force participation and migrant populations face less favourable prospects on the labour market. This situation becomes aggravated after migrant women have entered parenthood. While the impact of family formation on maternal employment has been well documented, no studies have addressed the question whether family formation has repercussions on the pension protection of migrant women. Using longitudinal micro-data from the Belgian Administrative Socio-Demographic panel this study investigates the effect of family formation on the pension protection of first and second-generation migrant women between 1999 and 2010. The results show that parenthood has a strong effect on pension protection by decreasing the likelihood of building up full-time pension rights through employment. Compared to women of Belgian origin, migrant women more rapidly build up pension rights through assimilated periods (e.g. unemployment, maternity/parental leave) or through derived periods based on a (previous) marriage to a working spouse. Whereas second-generation migrants are less likely than first-generation migrants to build up pension rights based on assimilated and derived periods, they are still less likely than women of Belgian origin to build up full-time pension rights through employment. Finally, differentials in pension protection after entry into parenthood are largely explained by differentials in pension protection before parenthood. 


Age structure and the economy: Longer-term effects of immigration as an instrument to offset population ageing
Frank T. Denton, Byron G. Spencer, Department of Economics, McMaster University, Hamilton, Ontario

BACKGROUND Immigration is often thought of as an instrument for offsetting adverse effects of population aging on the economy but longer-run effects are frequently neglected. OBJECTIVE We ask what impact, in the longer run, alternative immigration strategies might have on the level of real income, especially per capita real income, when employed in the context of an aging population. We consider alternative choices of immigration rates and age distributions. We consider also, in conjunction with immigration, effects of productivity growth, increased labour participation of older people, longer life expectancy, and higher fertility rates. METHODS A simple generic demographic-economic model is used in simulation experiments to explore longer-run effects of immigration strategies. The “laboratory” is a fictional country Alpha, to which we assign realistic demographic characteristics typical of a country experiencing population aging. MOST SIGNIFICANT RESULTS In our experiments a very high and possibly unacceptable rate of immigration, with a heavy concentration of adult immigrants in younger working ages and a low proportion of children, is required to keep per capita real income from declining. A higher rate of productivity growth provides offset to population aging, with or without immigration, but to be realistic must be interpreted in relation to the rate in other countries. A higher rate of labour participation among older workers provides some offset too, but increases have to be large. Longer life expectancy, taken alone, lowers per capita real income, as do higher fertility rates. 


Population ageing and intergenerational relationships
Michela C. Pellicani, University of Bari 

Confronting population ageing, from the issues of elderly dependency to the increasing risks of unemployment and economic and social exclusion, intergenerational solidarity acquires (and will acquire even more in the future) a strategic importance. Population ageing, by definition, cannot be analysed in absolute terms but in relatively to the evolution of other age groups and generations. In demography the change can be interpreted as well as a succession of disparate generations. The more the change is rapid, the more intergenerational contrasts are pronounced. In generational terms and in an horizontal perspective, it is the amount of the different observed generations at the same fixed age that will vary; in a vertical perspective, it is the relationship between the different generations that will change.

To make the data talk, we first drew a reference picture underlining the past, the present and the future structural evolutions. Then thinking in intergenerational terms, we used two other indexes trying to answer to the following question: how heavy are the charges (in terms of support provided) on the average individual (in a statistical meaning) represented by their parents. The index that we adopted is based on the evaluation that the average individual makes of his situation giving more importance to the individual perspective (lived reality) instead of the population perspective.

We used the UN projections data (2013 revision and medium variant) and in particular the observed and projected age structures. The calculation units are the national populations. We chose to analyse the countries of the Mediterranean Basin because it represents, at the same time, a geopolitical region of very strong interconnections and of a space characterised (at the moment) by different stages of population ageing.

In such a context that brings into question the measures and instruments of social protection and welfare state, therefore, it is fundamental to carefully evaluate the structural evolutions to calibrate correctly the public policies and their equilibrium with the private (family) forms of solidarity.


Ageing: Health and Wellbeing Tuesday 8 September 1.45 pm 

Home Aaone: Determinants of living alone among older immigrants in Canada and the U.S.
S.M. Lee, B. Edmonston, University of Victoria, British Columbia

While living alone is not a new form of living arrangement, “going solo” has increased in many parts of the world. In this paper, we focus on “aging alone” or the increase in living alone among older people. Previous research on living arrangements of immigrants in Canada and the U.S. suggest that older immigrants may be less likely to live alone than older non-immigrants, because immigrants are more likely to co-reside with family members. We have two research questions. First, are older immigrants less likely than Canadian- or U.S.-born elderly to live alone? Second, what are the main determinants of living alone among older immigrants in each country? We examine data from the 2006 Canadian census and 2006 American Community Survey. The study samples are non-married older immigrants, aged 55 and older. Preliminary findings show that lower percentages of older immigrants in both countries live alone compared to native-born older people, but the large gaps are substantially reduced once various explanatory variables such as socioeconomic status and age are taken into account in logistic regression models. Comparisons of four gender/country groups of older immigrants (female/Canada, male/Canada, female/U.S., male/U.S.) confirm the positive effects of economic (income, education) and acculturation (duration of residence, language proficiency) factors on living alone among older immigrants. With few exceptions, determinants of living alone are similar for older immigrants in both countries. We discuss implications for studying living arrangements of older immigrants in other countries, and for housing, family and community social support, healthcare, and other services.


Pathways to death by socioeconomic status
Cecilia Potente, University of Oxford, Nuffield College

Socioeconomic status plays an ambiguous role in later life health: scholars are not unanimous about the existence of a convergence of health outcomes or a divergent pattern among socioeconomic groups as a result of the accumulation of risks during the life-course. The perspective through which socioeconomic status and health are examined in this work is innovative. Age loses centrality in this analysis of health outcomes in later life, since health trajectories among different SES are compared with respect to a fixed end point: death. The central idea is that socioeconomic status might have a critical role in shaping health in proximity to death partially irrespective of age. Moreover, the emphasis is on considering the multidimensionality both of health outcomes (ADL, IADL, physical functioning, and depression) and socioeconomic indicators (education, income, wealth, and occupation). Data are from the English Longitudinal Study of Aging (Wave 1 - 5) including individuals aged 65 and over. Through latent class analysis the most common patterns to death are selected and analyzed. The goal is to see how different SES indicators predicts the belonging to a certain trajectory of health outcomes. Accumulation of risk theory (or cumulative advantage) would predict a more abrupt decline in health for more advantaged groups. On the contrary, in case of convergence among health outcomes in later life considerable differences in the trajectories for different SES individuals would not be observed. This analysis contributes towards a better understanding of the social gradient in health during old age.


Ageing in Russia: Comparative analysis of Russia's rank in the international indexes of active ageing
Oxana Sinyavskaya, Maria Varlamova, Institute for Social Development Studies, Higher School of Economics, Maastricht University

This paper is aimed at applying and analyzing international active ageing indices in Russia, including the Active Ageing Index (AAI), developed by European Centre Vienna, and Global AgeWatch Index by HelpAge International, to provide the base for cross-national comparison and development of a comprehensive national policy on active ageing. Our research was motivated by the following questions (1) to what extent can the international approaches to measure active ageing be applied to the Russian context and data? (2) to what extent a country’s position in the ranking is sensitive to the index methodology and data used? (3) whether and under what conditions Russia can improve its positions in the active ageing indices? To answer these questions, we estimated the AAI for Russia based on eight data sources and recalculated some of the AgeWatch Index results based on reliable data. The methodology of both indices and the quality and adequacy of the data used are discussed in detail in the paper. The results show that ranking of Russia according to these indices varies considerably from the 65th place out of 96 countries by the Global AgeWatch Index to the 18th place among 29 countries (28 EU countries plus Russia) by the AAI. Nevertheless, both indices draw rather similar pictures of active ageing potential in Russia. We provide some recommendations on how the indicators can be modified to capture some peculiarities of the ageing context in Russia and other countries with similar demographic, economic and social context.


Measuring Frailty: A comparison of Fried’s Frailty Phenotype and Rockwood’s Frailty Index using the English Longitudinal Study of Ageing (ELSA)
Marshall, A.1, Buckly, J.2, Mekli, K.2, Nazroo, J.2 , 1University of St Andrews, 2University of Manchester

This paper addresses the uncertainty around how frailty should be measured by comparing the Fried and Rockwood frailty assessment measures using data from the English Longitudinal study of Ageing (ELSA). The frailty index developed by Rockwood is based on the proportion of the ‘deficits’ that an individual has experienced. The Fried approach categorises individuals into states of ‘robust’, ‘pre-frail’ and ‘frail’ based on the presence of three or more frailty indicators: unintentional weight loss, slow walking speed, subjective exhaustion, low grip strength and low levels of physical activity. We evaluate the success of each frailty measure in predicting death (and other adverse outcomes). The results of our survival analysis suggest a comparable performance of the Fried and Rockwood measures of frailty in terms of predicting mortality. As might be expected we observed significantly increased risk of mortality with increasing frailty according to each frailty measure. Harrells C statistic an indicator of model fit was comparable for the continuous frailty index (Harrells C=0.79), the discrete version of the frailty index (Harrells C=0.78) and the fried frailty measures (Harrells C=0.79). Interestingly we observed a similar model fit using measures of wealth and self-reported health as opposed to measures of frailty. In conclusion, prediction of mortality over an 8 year period for older people (60+) living in the community is comparable whether the frailty phenotype or index is used. The choice of measure might then reflect the particular setting. The frailty phenotype is advantageous in clinical setting, whilst the Frailty index is useful in community environments where the collection of clinical information within the frailty phenotype is challenging.