Cohort Change in Living Conditions and Lifestyle Among Middle-Aged Swedes: The Effects on Mortality and Late-Life Disability

Abstract

The results of gerontological research today are based on cohorts that were born and lived through historically unique circumstances. People now entering old age have had different socioeconomic living conditions and different lifestyles. In Sweden today, people aged 55–64 are much better educated and more likely to participate in political and cultural activities, compared to previous cohorts. They are more likely to have a good cash margin and are less likely to report having economic hardship in childhood. They are more active and have better teeth. On the other hand, they report more physical symptoms and women in this age group are more likely to smoke while men less so. This chapter will use nationally representative longitudinal data stretching from 1968 to 2000 to compare birth cohorts at mid-life in regards to living conditions, activity participation, and health behavior. We will then follow-up the older birth cohorts to examine associations between socioeconomic and lifestyle factors with mortality and disability in late life. Can we assume that the nature of these associations will be similar for all cohorts? And if so, to what extent can we expect improvements in future cohorts of old people?

Appendix for Cohort change in living conditions and lifestyle among middle aged Swedes:The effects on mortality and late-life disability

Material and Methods

Data material

Data come from the Level of Living Survey (LNU), a nationally representative study of the Swedish adult population aged 18–75. LNU was first carried out in 1968, and subsequently in 1974, 1981, 1991, 2000 and 2010. Younger individuals and immigrants are added to the sample to keep it nationally representative. The LNU surveys were followed up by the Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD). The SWEOLD sample consists of persons who were included in the LNU sample but passed the upper age limit of 75 years. Data collections were carried out in 1992, 2002, 2004 and 2010.

In both surveys, professional interviewers conducted structured interviews with participants in their homes. The interviews addressed questions about work life, family situation, health behaviours, economic conditions, living conditions and health status. The SWEOLD interviews excluded questions about working conditions and instead had more questions about health, function and the ability to manage everyday life.

Together, the surveys have both a cross sectional and a longitudinal character. While each survey wave represents a cross section of the Swedish population at the time, there is also a panel of individuals that is followed over time. This study uses data from middle-aged adults (ages 55–64) from the LNU surveys conducted in 1968 (n = 935), 1981 (n = 812), 1991 (n = 698) and 2000 (n = 843). Table 16.1 uses this repeated cross sectional design, while results in Table 16.4 are based on a longitudinal design with follow-ups of the birth cohorts from 1968 (followed up in SWEOLD 1992) and 1981 (followed up in SWEOLD 2004).

Table 16.1 Descriptive characteristics of the study sample, divided by gender and cohort. (Values in percentage)

Outcome variables

Mortality

The follow-up time with regard to mortality was 15 years from the time of interview. Consequently, the sample interviewed in 2000 was not included in these analyses. Date of death was obtained from the Swedish National Cause of Deaths Register.

Late-life disability

This measure includes both instrumental activities of daily living(IADL,including food purchase, food preparation, and cleaning of the house) and activities of daily living(ADL,including the ability to eat by yourself, to visit the toilet, to dress and undress, to get in and out of bed, and to bathe or shower). Each IADL disability given 1 point, inability to bathe or shower 2 points, and each ADL disability 3 points. The final score ranged from 0–15, grouped into the following categories: 0 points (no disability), 1–2 points, 3–6 points, 7–15 points (disabled). Late-life disability/function was based on the older persons’ actual situation at the time of follow-up in 1992 (for those interviewed in 1968) or 2004 (for those interviewed in 1981). Data came from the SWEOLD study.

Independent variables

All independent variables were measured at mid-life (average age 59) in 1968, 1981, 1991 and 2000.

Current smoking was divided into no current smoking, moderate smoking (less than 10 cigarettes or equivalent per day) and heavy smoking (10 cigarettes or more). In the analyses investigating associations with mortality and late-life disability, a dichotomous variable of current smoking (no/yes) was used.

Three activities were included in the index of socio-cultural activities: going to the movies, going to the theatre, museums, exhibitions, and attending study circles. Participation frequency for each activity was rated as never (0), sometimes (1), and often (2). The index was dichotomized, with persons scoring 1 or higher on the index were coded as active.

Political activities included questions about ever having contacted an authority in order to influence a decision, delivered a speech at a meeting, written an article or a letter to the editor, or participated in a demonstration. Persons answering yes to any of these questions were coded politically active.

Low social network activitiesconsisted of two questions regarding visiting friends and visiting relatives. Responses were never (0), sometimes (1), and often (2). Persons scoring 0 or 1 were categorized as having low social network activities.

Civil status was self-reported and coded as unmarried (never-married), divorced, widow/widower and married/cohabiting.

Education was measured as years of schooling. Since the level of education has increased remarkably since the 1960s, we dichotomized education based on quartiles in the analyses of mortality and late-life disability. Persons in the highest quartile were categorized as highly educated. In the analyses of late-life disability a separate education variable was used that separated the highest education quartile among survivors only.

Occupation was measured as own occupation, categorized into unskilled blue-collar worker, skilled blue-collar worker, lower white-collar worker, and higher white-collar worker. Housewives were coded according to their husband’s occupation.

Employment status was categorized as “employed” (working for pay), “housewife” and “other”. The “other” category consisted of persons who were retired (mostly disability pension), on sick-leave, unemployed, studying etc.

Having no cash margin indicates not being able to withdraw 12,000 SEK (year 2000) from own account within a week. In 1968, 1981 and 1991 equivalent amounts were used.

Economic hardships in childhood was measured retrospectively, possible answers were yes and no.

The index of circulatory problems consisted of myocardial infarction, chest pain, heart weakness, high blood pressure and dizziness. These items were subjectively rated as no problems (0), moderate problems (1) and severe problems (3). Myocardial infarction gave 3 points independent of severity. The summed score was divided into no problems (0 points), moderate problems (1–2 points), and severe problems (3+ points). In the analyses investigating associations with mortality and late-life disability a dichotomous variable of circulatory problems (no/yes) was used.

The index of psychological distress consisted of self-reported depressions, nervous problems (e.g., anxiety and nervousness), general fatigue and sleep problems. The variable construction was the same as that of circulatory problems. In the analyses investigating associations with mortality and late-life disability a dichotomous variable of psychological distress (no/yes) was used.

Impaired mobility is a composite measurement of the respondents’ self-assessed ability to walk 100 m fairly briskly without difficulties and going up and down stairs without difficulties. Individuals who are able to perform these two tasks are coded 0; the others are considered to have impaired mobility and are coded 1.

Dental status is a self-reported item where the respondents are asked to judge their own dental status. People reporting that they had most of their own teeth, either in good shape or with many fillings and crowns, were considered to have good dental status. People reporting that their teeth were in bad shape or that they had no or few own teeth were considered to have poor dental status.

Table 16.2 Outcome variables: 15-year mortality and late-life disability (23–24 years of follow-up)

Table 16.3 Associations between lifestyle factors, socioeconomic factors and health status with 15-year mortality, divided by cohort. Hazard ratios

Table 16.4 Associations between lifestyle factors, socioeconomic factors and health status in midlife with late-life disability (index of ADL and IADL), divided by cohort. Odds ratios from ordered logistic regressions