Age Differences in Unemployment Risk and Reemployment Outcomes in Late Working Life in Sweden.

Participation of older workers in the labor market depends, among other things, on older workers employment chances. This study examines age differences regarding risk of unemployment and reemployment outcomes in late working life in Sweden. Using Swedish registry data, we analyzed the probability of unemployment as well as work-related activity following unemployment (wage- or self-employment; exit; downward mobility) of all people born between 1954 and 1968 (aged 49-63) and registered in Sweden between the years 2012 and 2018. Results show that although risk of unemployment does not differ significantly across age groups, younger age groups are more likely to be reemployed as wage-employed while older age groups are more likely to be reemployed as self-employed or exit working life. After an unemployment period, older employees are more likely to have a lower wage than during their previous employment or become part-time unemployed. We conclude that different age groups have unequal chances in late working life in terms of reemployment, risk of exit, and risk of downward mobility following unemployment. Policies for extending working life and promoting inequality should include measures for increasing employability of older workers such as anti-discriminatory laws and dealing with skills mismatch.

Among older people, the risk of unemployment does not differ across age groups.Among older people, reemployment chances decrease by age.After unemployment, the probability of exit or self-employment increases by age among older people.Older employees have higher risk of downward mobility after unemployment.Policies for the demand side and employability of older workers are needed.

Patterns of long-term care utilization during the last five years of life among Swedish older adults with and without dementia.

AIMS: The aims of this study were to compare the patterns of long-term care (LTC) use (no care, homecare, residential care) among people with and without dementia aged 70+ in Sweden during their last five years of life and its association with sociodemographic factors (age, gender, education, cohabitation status) and time with a dementia diagnosis. METHODS: This retrospective cohort study included all people who died in November 2019 aged 70 years and older (n = 6294) derived from several national registers. A multinomial logistic regression was conducted to identify which sociodemographic factors predicted the patterns of LTC use. RESULTS: Results showed that the time with a dementia diagnosis and cohabitation status were important predictors that influence the patterns of LTC use during the last five years of life. Nearly three-quarters of people living with dementia (PlwD) used residential care during the last five years of life. PlwD were more likely to reside in residential care close to death. Women who lived alone, with or without dementia, used residential care to a higher degree compared to married or cohabiting women. CONCLUSIONS: Among people without a dementia diagnosis, as well as those who were newly diagnosed, it was common to have no LTC at all, or use LTC only for a brief period close to death. During the last five years of life, PlwD and those living alone more often entered LTC early and used residential care for a longer time compared to people without dementia and people living alone, respectively.

History of working conditions and the risk of old-age dependency: a nationwide Swedish register-based study.

AIMS: There is substantial evidence that previous working conditions influence post-retirement health, yet little is known about previous working conditions' association with old-age dependency. We examined job strain, hazardous and physical demands across working life, in relation to the risk of entering old-age dependency of care. METHODS: Individually linked nationwide Swedish registers were used to identify people aged 70+ who were not receiving long-term care (residential care or homecare) at baseline (January 2014). Register information on job titles between the years 1970 and 2010 was linked with a job exposure matrix of working conditions. Random effects growth curve models were used to calculate intra-individual trajectories of working conditions. Cox regression models with age as the timescale (adjusted for living situation, educational attainment, country of birth, and sex) were conducted to estimate hazard ratios for entering old-age dependency during the 24 months of follow-up (n = 931,819). RESULTS: Having initial adverse working conditions followed by an accumulation throughout working life encompassed the highest risk of entering old-age dependency across the categories (job strain: HR 1.23, 95% CI 1.19-1.27; physical demands: HR 1.36, 95% CI 1.31-1.40, and hazardous work: HR 1.35, 95% CI 1.30-1.40). Initially high physical demands or hazardous work followed by a stable trajectory, or initially low-level physical demand or hazardous work followed by an accumulation throughout working life also encompassed a higher risk of dependency. CONCLUSIONS: A history of adverse working conditions increased the risk of old-age dependency. Reducing the accumulation of adverse working conditions across the working life may contribute to postponing old-age dependency.

Length of the period with late life dependency: Does the age of onset make a difference?

There is a gap in knowledge about factors associated with the duration of late life dependency. In this study, we measured how the age at onset of late life dependency relates to the time spent with late life dependency. Using Swedish register data, we identified people 70 + who entered the period of late life dependency (measured by entering long-term care for help with PADLs) between June and December 2008. We followed this cohort (n = 17,515) for 7 years, or until death. We used Laplace regression models to estimate the median number of months with late life dependency by age group, gender, level of education and country of birth. We also calculated the crude percentiles (p10, p25, p50, p75 and p90) of month with late life dependency, by age group, gender and cohabitation status. Results show that the majority spent a rather long period with dependency, the median number of months were 40.0 (3.3 years) for women and 22.6 (1.9 year) for men. A higher age at entry was associated with a shorter duration of dependency, an association that was robust to adjustment for cohabiting at baseline, gender, education and country of birth. Our results suggest that older adults who postpone the start of dependency also compress the time with dependency, this lends support to the ambitions of public health initiatives and interventions targeting maintained independence in older adults.

Educational differences in long-term care use in Sweden during the last two years of life.

Background: In old age, many people experience a period of functional decline and require long-term care. Sweden has a universal largely tax-financed health and social care system that is used by all societal groups. However, few studies have investigated if educational groups use publicly paid long-term care equitably. The aim of this study was to explore educational differences in the use of long-term care, including both home care and institutional care, during the last two years of life in Sweden. Methods: We used linked register data on mortality and long-term care use, including all adults aged ⩾67 years who died in Sweden in November 2015 (N=6329). We used zero-inflated negative binomial regression models to analyse the number of months with long-term care by educational level, both crude and adjusted for age at death and cohabitation status. Men and women were analysed separately. Results: People with tertiary education died more commonly without using any long-term care compared to primary educated people (28.0% vs. 18.6%; p<0.001). In the adjusted model, educational differences in the estimated number of months with long-term care disappeared among men but remained significant among women (primary educated: odds ratio=17.3 (confidence interval 16.8-17.7); tertiary educated: odds ratio=15.8 (confidence interval 14.8-16.8)). Conclusions: Older adults spend considerable time in their last two years of life with long-term care. Only minor educational differences in long-term care use remained after adjustment for cohabitation status and age at death. This suggest that Sweden's publicly financed long-term system achieves relatively equitable use of long-term care at the end of life.

Long-term care use among people living with dementia: a retrospective register-based study from Sweden.

BACKGROUND: Although many people with dementia need progressive support during their last years of life little is known to what extent they use formal long-term care (LTC). This study investigates the use of LTC, including residential care and homecare, in the month preceding death, as well as the number of months spent in residential care, among Swedish older decedents with a dementia diagnosis, compared with those without a dementia diagnosis. METHODOLOGY: This retrospective cohort study identified all people who died in November 2019 in Sweden aged 70 years and older (n = 6294). Dementia diagnoses were collected from the National Patient Register (before death) and the National Cause of Death Register (death certificate). The use of LTC was based on the Social Services Register and sociodemographic factors were provided by Statistics Sweden. We performed regression models (multinomial and linear logistic regression models) to examine the association between the utilization of LTC and the independent variables. RESULTS: Not only dementia diagnosis but also time spent with the diagnosis was crucial for the use of LTC in the month preceding death, in particular residential care. Three out of four of the decedents with dementia and one fourth of those without dementia lived in a residential care facility in the month preceding death. People who were diagnosed more recently were more likely to use homecare (e.g., diagnosis for 1 year or less: home care 29%, residential care 56%), while the predicted proportion of using residential care increased substantially for those who had lived longer with a diagnosis (e.g., diagnosis for 7 + years: home care 11%, residential care 85%). On average, people with a dementia diagnosis stayed six months longer in residential care, compared with people without a diagnosis. CONCLUSIONS: People living with dementia use more LTC and spend longer time in residential care than those without dementia. The use of LTC is primarily influenced by the time with a dementia diagnosis. Our study suggests conducting more research to investigate differences between people living with different dementia diagnoses with co-morbidities.

People Living with a Dementia Diagnosis with No Eldercare at all: Who Are They?

Dementia is a non-curable disease that progressively affects people's ability to handle their everyday life. Still, previous Swedish research found that many people living with dementia (PlwD) do not use any eldercare. In this study, we investigated the association between not using eldercare services and key social background factors: years with dementia, hospital care, and country of birth for PlwD. We identified all people aged 65+ diagnosed with dementia between January 2006 and March 2015 (n = 43,372) using secondary data analysis of information taken from multiple Swedish nationwide registers. Results showed that not using eldercare was more common among younger age groups and men. The likelihood of having no eldercare was three times higher for those cohabiting; notably, almost a fifth of the PlwD who had the diagnosis for four years or more did not have eldercare at all. Finally, people born outside Europe used eldercare less than persons born in Sweden. Since the public dementia policy in Sweden is basically geared toward PlwD in a later stage and primarily in residential care, there has been almost no policy development around the needs and rights of PlwD in their homes. This study provides a ground for reconsidering dementia policy regardless of the ethnic and cultural backgrounds of PlwD.

Trends over two decades in life expectancy with complex health problems among older Swedes: implications for the provision of integrated health care and social care.

BACKGROUND: Due to population aging, it is essential to examine to what extent rises in life expectancy (LE) consist of healthy or unhealthy years. Most health expectancy studies have been based on single health measures and have shown divergent trends. We used a multi-domain indicator, complex health problems (CHP), indicative of the need for integrated medical and social care, to investigate how LE with and without CHP developed in Sweden between 1992 and 2011. We also addressed whether individuals with CHP more commonly lived in the community in 2011 compared to earlier years. METHODS: CHP were defined as having severe problems in at least two of three health domains related to the need for medical and/or social care: symptoms/diseases, cognition/communication, and mobility. The Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD), a nationally representative survey of the Swedish population aged ≥ 77 years with waves in 1992, 2002 and 2011 (n≈2000), was used to estimate the prevalence of CHP. Age- and gender-specific death rates were obtained from the Human Mortality Database. The Sullivan method was deployed to calculate the remaining life expectancy with and without CHP. The estimates were decomposed to calculate the contribution of changes from morbidity and mortality to the overall trends in LE without CHP. RESULTS: Between 1992 and 2011, both total LE (+ 1.69 years [95% CI 1.56;1.83] and LE without CHP (+ 0.84 years [-0,87;2.55]) at age 77 increased for men, whereas LE at age 77 increased for women (+ 1.33 [1.21;1.47]) but not LE without CHP (-0.06 years [-1.39;1.26]). When decomposing the trend, we found that the increase in LE with CHP was mainly driven by an increase in the prevalence of CHP. Among individuals with CHP the proportion residing in care homes was lower in 2011 (37%) compared to 2002 (58%) and 1992 (53%). CONCLUSIONS: The findings, that an increasing number of older people are expected to live more years with CHP, and increasingly live in the community, point towards a challenge for individuals and families, as well as for society in financing and organizing coordinated and coherent medical and social services.

Changes in Secondary Healthcare Use Over Retirement Transition: Examining Social Differences With Swedish Register Data.

Background: Despite its relevance for healthcare expenditures and public health, few studies have examined how secondary healthcare use changes during the retirement transition. We therefore use Swedish register data to examine whether retirement is associated with intensified secondary healthcare use overall and for specific subgroups based on gender and education. Methods: The sample was all individuals registered in Sweden who retired from paid work in 2010. We used Generalised Estimating Equations models to analyse changes in two indicators of secondary healthcare use, namely specialist visits and hospitalisation, from 3 years prior to 5 years after retirement. Results: Retirement is not associated with changes in specialist visits or hospitalisation per se. Three years before retirement, women were more likely to visit a specialist but less likely to be hospitalised than men; these gender differences disappeared approximately 1 year before retirement. Women with high education were more likely to visit a specialist than women with low education across the entire retirement transition, particularly post-retirement. Significant differences with regard to specialist visits between male educational groups only emerged 12 months after retirement. There were no educational differences with regard to hospitalisation. Conclusions: We conclude that secondary healthcare use in Sweden does not generally change with retirement. However, over the course of retirement gender differences in secondary healthcare use tend to decrease and within-gender educational differences tend to increase. We interpret the results as reflecting the role of labour market institutions in contributing to gender differences but repressing educational differences in secondary healthcare use.

Development of healthcare use across contemporary retirement pathways: results from a register based cohort study.

AIM: We aimed to understand the interplay between retirement pathways and healthcare use in the postponed and structurally changing context of retirement. METHODS: Based on Swedish register data on income and healthcare use, we applied combined sequence and cluster analysis to identify typical pathways into retirement and analysed their relation to healthcare use developments. RESULTS: We detected five distinct pathways into retirement. Level of healthcare use was significantly higher for the pathway via disability pensions. We saw an overall increase in healthcare use across the retirement process that was related to age rather than to the different pathways. CONCLUSIONS: Level of healthcare use at the beginning of the retirement process may be related to selection into different pathways of retirement. We did not find clear evidence across several healthcare measures that different pathways lead to different developments in healthcare use.

Eldercare services for people with and without a dementia diagnosis: an analysis of Swedish registry data.

BACKGROUND: The growing number of people living with dementia (PlwD) implies an increase in the demand for eldercare services in Sweden like in many other countries. Few studies have analyzed the use of eldercare services for PlwD. The aim of the present study is to investigate the association between demographic factors (age, sex, cohabiting status) and the use of municipal eldercare services (including both home care and residential care) for older adults with dementia compared to older adults without dementia in Sweden. METHODS: This study used several nationwide Swedish registers targeting all individuals aged 65 and above living in Sweden in 2014 and still alive 31st of March 2015 (n = 2,004,409). The primary outcomes variables were different types of eldercare service, and all participants were clustered based on age, sex, cohabiting status, and dementia diagnosis. In addition to descriptive statistics, we performed multivariate logistic regression models for binary outcomes and linear regression models for continuous outcomes. RESULTS: Results showed that (1) older age is a significantly strong predictor for the use of eldercare services, although PlwD start using eldercare at an earlier age compared with people without dementia; (2) women tend to receive more eldercare services than men, especially in older age, although men with dementia who live alone are more likely than women living alone to receive eldercare; (3) having a dementia diagnosis is a strong predictor for receiving eldercare. However, it was also found that a substantial proportion of men and women with dementia did not receive any eldercare services. CONCLUSIONS: We found that people with a dementia diagnosis use more as well as start to use eldercare services at an earlier age than people without dementia. However, further research is needed to investigate why a substantial part of people with a dementia diagnosis does not have any eldercare at all and what the policy implications of this might be.

Going web or staying paper? The use of web-surveys among older people.

BACKGROUND: Web-surveys are increasingly used in population studies. Yet, web-surveys targeting older individuals are still uncommon for various reasons. However, with younger cohorts approaching older age, the potentials for web-surveys among older people might be improved. In this study, we investigated response patterns in a web-survey targeting older adults and the potential importance of offering a paper-questionnaire as an alternative to the web-questionnaire. METHODS: We analyzed data from three waves of a retirement study, in which a web-push methodology was used and a paper questionnaire was offered as an alternative to the web questionnaire in the last reminder. We mapped the response patterns, compared web- and paper respondents and compared different key outcomes resulting from the sample with and without the paper respondents, both at baseline and after two follow-ups. RESULTS: Paper-respondents, that is, those that did not answer until they got a paper questionnaire with the last reminder, were more likely to be female, retired, single, and to report a lower level of education, higher levels of depression and lower self-reported health, compared to web-respondents. The association between retirement status and depression was only present among web-respondents. The differences between web and paper respondents were stronger in the longitudinal sample (after two follow-ups) than at baseline. CONCLUSIONS: We conclude that a web-survey might be a feasible and good alternative in surveys targeting people in the retirement age range. However, without offering a paper-questionnaire, a small but important group will likely be missing with potential biased estimates as the result.

Prediction of (Non)Participation of Older People in Digital Health Research: Exergame Intervention Study.

BACKGROUND: The use of digital technologies is increasing in health care. However, studies evaluating digital health technologies can be characterized by selective nonparticipation of older people, although older people represent one of the main user groups of health care. OBJECTIVE: We examined whether and how participation in an exergame intervention study was associated with age, gender, and heart failure (HF) symptom severity. METHODS: A subset of data from the HF-Wii study was used. The data came from patients with HF in institutional settings in Germany, Italy, the Netherlands, and Sweden. Selective nonparticipation was examined as resulting from two processes: (non)recruitment and self-selection. Baseline information on age, gender, and New York Heart Association Functional Classification of 1632 patients with HF were the predictor variables. These patients were screened for HF-Wii study participation. Reasons for nonparticipation were evaluated. RESULTS: Of the 1632 screened patients, 71% did not participate. The nonrecruitment rate was 21%, and based on the eligible sample, the refusal rate was 61%. Higher age was associated with lower probability of participation; it increased both the probabilities of not being recruited and declining to participate. More severe symptoms increased the likelihood of nonrecruitment. Gender had no effect. The most common reasons for nonrecruitment and self-selection were related to physical limitations and lack of time, respectively. CONCLUSIONS: Results indicate that selective nonparticipation takes place in digital health research and that it is associated with age and symptom severity. Gender effects cannot be proven. Such systematic selection can lead to biased research results that inappropriately inform research, policy, and practice. TRIAL REGISTRATION: ClinicalTrial.gov NCT01785121, https://clinicaltrials.gov/ct2/show/NCT01785121.

Long-Term Care Use During the Last 2 Years of Life in Sweden: Implications for Policy to Address Increased Population Aging.

OBJECTIVES: To map out the total use of long-term care (LTC; ie, home care or institutional care) during the last 2 years of life and to investigate to what extent gender differences in LTC use were explained by cohabitation status and age at death. DESIGN: The National Cause of Death Register was used to identify decedents. Use of LTC was based on the Social Services Register (SSR) and sociodemographic factors were provided by Statistics Sweden. SETTING AND PARTICIPANTS: All persons living in Sweden who died in November 2015 aged ≥67 years (n = 5948). METHODS: Zero inflated negative binomial regression was used to estimate the relative impact of age, gender, and cohabitation status on the use of LTC. RESULTS: Women used LTC to a larger extent [odds ratio (OR) 2.17, 95% confidence interval (CI) 1.92-2.50] and for a longer period [risk ratio (RR) 1.14, 95% CI 1.11-1.18] than men. When controlling for age at death and cohabitation status, gender differences in LTC attenuated (OR 1.47, 95% CI 1.28-1.72) and vanished in regard to the duration. In the controlled model, women used LTC for 15.6 months (95% CI 15.2-16.0) and men for 14.1 months (95% CI 13.7-14.5) out of 24 months. The length of stay in institutional care was 7.2 (95% CI 6.8-7.5) and 6.2 months (95% CI 5.8-6.6), respectively. CONCLUSIONS AND IMPLICATIONS: A substantial part of women's greater use of LTC was due to their higher age at death and because they more often lived alone. Given that survival continues to increase, the association between older age at death and LTC use suggests that policy makers will have to deal with an increased pressure on the LTC sector. Yet, increased survival among men could imply that more women will have access to spousal caregivers, although very old couples may have limited capacity for extensive caregiving at the end of life.

A research tool for measuring non-participation of older people in research on digital health.

BACKGROUND: Healthcare services are being increasingly digitalised in European countries. However, in studies evaluating digital health technology, some people are less likely to participate than others, e.g. those who are older, those with a lower level of education and those with poorer digital skills. Such non-participation in research - deriving from the processes of non-recruitment of targeted individuals and self-selection - can be a driver of old-age exclusion from new digital health technologies. We aim to introduce, discuss and test an instrument to measure non-participation in digital health studies, in particular, the process of self-selection. METHODS: Based on a review of the relevant literature, we designed an instrument - the NPART survey questionnaire - for the analysis of self-selection, covering five thematic areas: socioeconomic factors, self-rated health and subjective overall quality of life, social participation, time resources, and digital skills and use of technology. The instrument was piloted on 70 older study persons in Sweden, approached during the recruitment process for a trial study. RESULTS: Results indicated that participants, as compared to decliners, were on average slightly younger and more educated, and reported better memory, higher social participation, and higher familiarity with and greater use of digital technologies. Overall, the survey questionnaire was able to discriminate between participants and decliners on the key aspects investigated, along the lines of the relevant literature. CONCLUSIONS: The NPART survey questionnaire can be applied to characterise non-participation in digital health research, in particular, the process of self-selection. It helps to identify underrepresented groups and their needs. Data generated from such an investigation, combined with hospital registry data on non-recruitment, allows for the implementation of improved sampling strategies, e.g. focused recruitment of underrepresented groups, and for the post hoc adjustment of results generated from biased samples, e.g. weighting procedures.

Proactive healthcare for frail elderly persons: study protocol for a prospective controlled primary care intervention in Sweden.

INTRODUCTION: The provision of healthcare services is not dedicated to promoting maintenance of function and does not target frail older persons at high risk of the main causes of morbidity and mortality. The aim of this study is to evaluate the effects of a proactive medical and social intervention in comparison with conventional care on a group of persons aged 75 and older selected by statistical prediction. METHODS AND ANALYSIS: In a pragmatic multicentre primary care setting (n=1600), a prediction model to find elderly (75+) persons at high risk of complex medical care or hospitalisation is used, followed by proactive medical and social care, in comparison with usual care. The study started in April 2017 with a run-in period until December 2017, followed by a 2-year continued intervention phase that will continue until the end of December 2019. The intervention includes several tools (multiprofessional team for rehabilitation, social support, medical care home visits and telephone support). Primary outcome measures are healthcare cost, number of hospital care episodes, hospital care days and mortality. Secondary outcome measures are number of outpatient visits, cost of social care and informal care, number of prescribed drugs, health-related quality of life, cost-effectiveness, sense of security, functional status and ability. We also study the care of elderly persons in a broader sense, by covering the perspectives of the patients, the professional staff and the management, and on a political level, by using semistructured interviews, qualitative methods and a questionnaire. ETHICS AND DISSEMINATION: Approved by the regional ethical review board in Linköping (Dnr 2016/347-31). The results will be presented in scientific journals and scientific meetings during 2019-2022 and are planned to be used for the development of future care models. TRIAL REGISTRATION NUMBER: NCT03180606.

Underestimated Health Inequalities Among Older People-A Consequence of Excluding the Most Disabled and Disadvantaged.

OBJECTIVES: The present study analyzed whether estimates of health inequalities in old age are sensitive to the exclusion of people living in institutions and nonuse of proxy interviews. METHOD: Pooled data from the 2004 wave (n = 1,180, aged 69-100, primarily interviewed over the phone) and the 2011 wave (n = 931, aged 76-101, primarily interviewed face-to-face) of the Swedish Panel Study of Living Conditions of the Oldest Old were used to compare absolute and relative differences in disability and mortality between people with compulsory education and people with more than compulsory education. RESULTS: Both absolute and relative health inequalities would have been underestimated in a survey that excluded institutionalized people and proxy-interviewed community dwellers. The same patterns were found in men and women and regardless of the mode of data collection (telephone or face-to-face interview). The degree of underestimation was lower in those 85 years and older than in those 69 to 84 years. DISCUSSION: A survey that only includes people who live in the community and can participate without the help of a proxy might give the impression that those with low levels of education have less extensive health disadvantages than they actually have.

Is the level of education associated with transitions between care settings in older adults near the end of life? A nationwide, retrospective cohort study.

BACKGROUND: End-of-life transitions between care settings can be burdensome for older adults and their relatives. AIM: To analyze the association between the level of education of older adults and their likelihood to experience care transitions during the final months before death. DESIGN: Nationwide, retrospective cohort study using register data. SETTING/PARTICIPANTS: Older adults (⩾65 years) who died in Sweden in 2013 ( n = 75,722). Place of death was the primary outcome. Institutionalization and multiple hospital admissions during the final months of life were defined as secondary outcomes. The decedents' level of education (primary, secondary, or tertiary education) was considered as the main exposure. Multivariable analyses were stratified by living arrangement and adjusted for sex, age at time of death, illness trajectory, and number of chronic diseases. RESULTS: Among community-dwellers, older adults with tertiary education were more likely to die in hospitals than those with primary education (55.6% vs 49.9%; odds ratio (OR) = 1.21, 95% confidence interval (CI) = 1.14-1.28), but less likely to be institutionalized during the final month before death (OR = 0.83, 95% CI = 0.76-0.91). Decedents with higher education had greater odds of remaining hospitalized continuously during their final 2 weeks of life (OR = 1.12, 95% CI = 1.02-1.22). Among older adults living in nursing homes, we found no association between the decedents' level of education and their likelihood to be hospitalized or to die in hospitals. CONCLUSION: Compared with those who completed only primary education, individuals with higher educational attainment were more likely to live at home until the end of life, but also more likely to be hospitalized and die in hospitals.

Alcohol consumption in very old age and its association with survival: A matter of health and physical function.

BACKGROUND: Alcohol consumption in very old age is increasing; yet, little is known about the personal and health-related characteristics associated with different levels of alcohol consumption and the association between alcohol consumption and survival among the oldest old. METHODS: Nationally representative data from the Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD, ages 76-101; n=863) collected in 2010/2011 were used. Mortality was analyzed until 2014. Alcohol consumption was measured with questions about frequency and amount. Drinks per month were calculated and categorized as abstainer, light-to-moderate drinker (0.5-30 drinks/month) and heavy drinker (>30 drinks/month). Multinomial logistic regressions and Laplace regressions were performed. RESULTS: Compared to light-to-moderate drinkers, abstainers had lower levels of education and more functional health problems, while heavy drinkers were more often men, had higher levels of education, and no serious health or functional problems. In models adjusted only for age and sex, abstainers died earlier than drinkers. Among light-to-moderate drinkers, each additional drink/month was associated with longer survival, while among heavy drinkers, each additional drink/month was associated with shorter survival. However, after adjusting for personal and health-related factors, estimates were lower and no longer statistically significant. CONCLUSIONS: The association between alcohol consumption and survival in very old age seems to have an inverse J-shape; abstention and heavy use is associated with shorter survival compared to light-to-moderate drinking. To a large extent, differences in survival are due to differences in baseline health and physical function.

Bias in estimates of alcohol use among older people: selection effects due to design, health, and cohort replacement.

BACKGROUND: There is a growing awareness of the need to include the oldest age groups in the epidemiological monitoring of alcohol consumption. This poses a number of challenges and this study sets out to examine the possible selection effects due to survey design, health status, and cohort replacement on estimates of alcohol use among the oldest old. METHODS: Analyses were based on three repeated cross-sectional interview surveys from 1992, 2002 and 2011, with relatively high response rates (86 %). The samples were nationally representative of the Swedish population aged 77+ (total n = 2022). Current alcohol use was assessed by the question "How often do you drink alcoholic beverages, such as wine, beer or spirits?" Alcohol use was examined in relation to survey design (response rate, use of proxy interviews and telephone interviews), health (institutional living, limitations with Activities of Daily Living and mobility problems) and birth cohort (in relation to age and period). Two outcomes were studied using binary and ordered logistic regression; use of alcohol and frequency of use among alcohol users. RESULTS: Higher estimates of alcohol use, as well as more frequent use, were associated with lower response rates, not using proxy interviews and exclusion of institutionalized respondents. When adjusted for health, none of these factors related to the survey design were significant. Moreover, the increase in alcohol use during the period was fully explained by cohort replacement. This cohort effect was also at least partially confounded by survey design and health effects. Results were similar for both outcomes. CONCLUSIONS: Survey non-participation in old age is likely to be associated with poor health and low alcohol consumption. Failure to include institutionalized respondents or those who are difficult to recruit is likely to lead to an overestimation of alcohol consumption, whereas basing prevalence on older data, at least in Sweden, is likely to underestimate the alcohol use of the oldest old. Trends in alcohol consumption in old age are highly sensitive for cohort effects. When analysing age-period-cohort effects, it is important to be aware of these health and design issues as they may lead to incorrect conclusions.