- Email: [email protected]
Patterns of social engagement in the transition to later life
Maturitas 88 (2016) 90–95
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Patterns of social engagement in the transition to later life Katherine Burn a , Lorraine Dennerstein b , Colette Browning c , Cassandra Szoeke a,∗ a
Department of Medicine, University of Melbourne, Parkville, Victoria, Australia Department of Psychiatry, University of Melbourne, Parkville, Victoria, Australia c Royal District Nursing Service, St Kilda, Victoria, Australia b
a r t i c l e
i n f o
Article history: Received 23 February 2016 Received in revised form 17 March 2016 Accepted 22 March 2016 Keywords: Social engagement Ageing Women
a b s t r a c t As social roles in later life are shaped by significant life events and changes occurring in the late-life transition, examining social engagement in midlife may provide a context for interpreting the ageing social identity. This is particularly important for women, who are heavily dependent on social relationships and are more influenced by social losses and change. Objective: To examine major social changes occurring in the decade prior to late-life, starting from approximately 45–55 years of age. Study design: The study accessed data from the longitudinal prospective Women’s Healthy Ageing Project (WHAP). Participants were 493 women who had completed at least one assessment in the first 12 years of the study. Results: Living with a partner was common and stable amongst participants, while the proportion with children still at home decreased markedly (79–44%). Full-time employment also decreased as participants approached the average retirement age (40–13%). Volunteer work was popular throughout the study, increasing slightly with age, and minding grandchildren was common at the end of the study period (80%). Conclusions: Taken together, these findings suggest a notable deficit in participants’ social lives as they transition into later life, but with some evidence of compensation by increasing other social activities. © 2016 Elsevier Ireland Ltd. All rights reserved.
The transition from midlife to later life denotes a key shift in social roles. Many social relationships and roles are lost and social network sizes have been shown to reduce [1–3]. Key events in this stage may include retirement from the workforce, the death of a spouse or friends, and a decrease in participation in hobbies due to poor physical health [4–9]. The loss of social contacts is often self-perpetuating as the loss of one role can lead to the loss of others. Losing a companion may lead to lessened feelings of encouragement or security, leaving individuals unwilling to strike out on their own [10,11]. Furthermore, older adults may experience difficulty or feel a reluctance to form new social bonds so late in life, especially when they are mourning lost relationships [6]. These factors are especially poignant for older women, as women are more likely to suffer from losses such as widowhood and are more dependent on wider social contact for wellbeing [12–15]. The losses occurring in the transition to old age
∗ Corresponding author at: P.O. Box 2026, Parkville, Victoria 3050, Australia. E-mail address: [email protected] (C. Szoeke). http://dx.doi.org/10.1016/j.maturitas.2016.03.016 0378-5122/© 2016 Elsevier Ireland Ltd. All rights reserved.
can influence the shape of social activity in later life. In their discussion of widowhood, Van den Hoonaard et al. [12] highlighted the conceptualisation of loss as a process; moving through the events surrounding the loss, mourning, through to the reconstruction of the individual’s social identity. As later life social roles are shaped by significant life events and changes occurring in the late-life transition, examining social engagement in midlife may provide a context for interpreting the ageing social identity. Unfortunately, much of the work examining age-related changes to social roles is outdated and focuses on American or European populations. As social roles are heavily influenced by cultural factors, population-specific research is needed in order to inform research. The current study aimed to examine social roles in midlife in order to provide a context for social engagement in the mid-to-late life transition for Australians.
K. Burn et al. / Maturitas 88 (2016) 90–95
91
Table 1 Description of WHAP social measures. Social factor
Items
Employment
Currently employed Employment type (full-time, part-time, etc.) Non-employed status (retired, looking for work, etc.) Partner’s employment status (full-time, retired, etc.) Single, married/de facto, divorced, widowed, separated Outlines all coresidents Changes to household occupants since last interview Number of grandchildren Minding grandchildren (number of grandchildren, frequency, activities) Current volunteer status Types of volunteering (verbatim) Organisations that participant belongs to
Marital status Household composition Grandparenting Volunteerism Community organisations
Fig. 1. Summary of the Australian Bureau of Statistics’ [27] General Social Survey (GSS). The survey covers common social relationships, opportunities for social engagement or support, and factors that influence social participation. These items range from innate factors to factors that influence the general social conditions. The GSS items encompass the measures used in the World Health Organisation’s [28] survey of the social determinants of mental health. Because this study examined social engagement, only items that involved direct personal contact were included (Household and Extended Networks items). Age, education, and gender were also included as part of the WHAP protocol. Data regarding contact with friends, neighbours, and other relatives was not available. The final selection of items for the WHAP protocol are shown underlined.
1. Methods
1.2. Measures and procedure
1.1. Participants
Assessments were conducted according to the project protocol which has been previously published [16]. Participants were administered a core questionnaire at each timepoint. The questionnaires from each timepoint contained a selection from a pool of items assessing social activities, relationships, and environments, outlined in Table 1. These items are consistent with the Australian Bureau of Statistics’ General Social Survey (GSS); a national survey of the opportunities and factors for active social participation (see Fig. 1). Items of interest to the current study included marital status, household composition, employment status, partners’ employment status, participation in unpaid work, and grandparenting.
Participants were from the epidemiologically sampled longitudinal prospective Women’s Healthy Ageing Project (WHAP) which was initiated as the Melbourne Women’s Midlife Health Project. Briefly, a longitudinal cohort of Caucasian women within the Melbourne metropolitan area were identified by random telephone dialling in 1991 and re-interviewed annually over eight years then intermittently over five years. Details of the study protocol are outlined elsewhere [16].
92
K. Burn et al. / Maturitas 88 (2016) 90–95
Marital Status
Employment Status 70
Percentage of respondents
80 70 60
Married/de facto
50
Separated
40
Divorced Widowed
30
Single
20 10
Percentagen of respondents
90
0 1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
60 50 Full-time 40
Part-time Casual/Occasional
30
Not employed 20
Volunteer
10 0 1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
Fig. 2. Reported marital status between 1992 and 2004. Most participants were married, which remained constant across the study. Separation rates decreased sharply while divorce rates increased in 1994, which may represent progress towards relationship termination.
Fig. 4. Paid and unpaid employment rates between 1992 and 2004. Note that participants were not asked about unpaid work in 2002 or 2004. Volunteerism was high throughout the study, with a slight increase in 1997 coinciding with the increase in those who were not employed. Casual and occasional work also increased over the course of the study, while full-time work decreased considerably.
Household Composition Participants' and their partners' employment status
90 90
70 60
Partner
50
Children
40
Parents/In-Laws Other adults
30
Lives alone
20 10 0 1992
1993
1994
1995
1996
1997
1998
1999
Fig. 3. Participants’ household composition 1992–1999. The proportion of participants living with a partner remained high throughout the study, while the proportion living with children dropped substantially. Note: household composition item was not a part of the questionnaire in 2002 and 2004.
Percentage of respondents
Percentage of respondents
80
80 70 60
Participants in full-time work
50
Partners in full-time work
40
Participants not employed
30 20
Partners not employed
10 0 1992 1993 1994 1995 1996 1997 1998 1999
Fig. 5. Participants’ employment rates compared to partners’ employment rates from 1992 to 1999. Item was not included in the questionnaire in 2002 and 2004. Participants were much less likely to be employed than their partners. However, both rates showed a negative trend across the study.
1.3. Data analysis Descriptive statistics such as response frequency were used to examine the prevalence of the different social connections across timepoints. Each role was then broken down into more detailed response categories and analysed using descriptive statistics. IBM SPSS was used to generate and graph all statistics. 1.4. Ethics The current study received approval from the University of Melbourne Human Research Ethics Council. Written informed consent was obtained from all participants prior to assessment at each timepoint. All data was de-identified. 2. Results The sample for the current study included 493 women who were interviewed at least once between 1992 and 2004. Sample sizes and age ranges for each timepoint are shown in Table 2. Marriage rates remained high and relatively stable across the study years, with a 5% decrease between 1992 and 2004 (Fig. 2). Similarly, the proportion of participants living with a partner remained high. The proportion living alone increased slightly from 1993 onwards, coinciding with the increase in sample divorce rates
(Fig. 3). Notably, the proportion of participants with children still at home dropped considerably from 78.8% in 1992 to 43.9% in 1999. Full-time employment decreased considerably between 1992 and 2004, dropping from 40% to only 13.1% of respondents. Although many of these would have retired, demonstrated by the increase in non-worker numbers, the percentage of participants in part-time employment increased slightly in 1998 and there was also an increase in casual and occasional work throughout the study (Fig. 4). The decrease in full-time working rates was mirrored in participants’ partners, which also decreased substantially (Fig. 5). While part-time working rates also decreased throughout the study, they did not experience as dramatic a change as full-time working rates, shown in Fig. 4. Participants not in paid employment increased markedly from 1997, when participants were aged approximately 50–61 years, coinciding with an increase in retirement rates. Of those who were not in paid employment, most indicated home duties as their main occupation (Fig. 6). Rates of volunteerism were noticeably high throughout the study, even when full-time employment rates were at their peak in 1991 (Fig. 4). Approximately 40% of participants were grandmothers, having an average of 3.99 (s.d. = 3.09) grandchildren. Of those who were grandmothers, nearly 80% stated that they were involved in minding their grandchildren, with more than half (53.5%) minding their grandchildren for at least a few hours a week.
K. Burn et al. / Maturitas 88 (2016) 90–95 Table 2 Sample sizes and age ranges by study year (1992–2004).
Status of Non-Workers 90
Percentage of respondents
80 Looking for full-time work Looking for part-time work Retired
70 60 50
Student
40 Non-worker
30 Home duties
20 10 0 1992
1993
1994
1995
1996
1997
1998
93
1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
N
Age Range
Mean (s.d) age
491 473 459 453 447 443 436 433 256 284
45.66–57.04 46.76–57.83 47.70–58.90 48.45–59.96 49.48–60. 78 50.50–61.78 51.56–62.75 52.77–63.71 55.93–67.16 57.97–68.70
49.74 (2.55) 50.81 (2.54) 51.83 (2.55) 52.76 (2.51) 53.66 (2.50) 54.73 (2.50) 55.73 (2.49) 56.75 (2.49) 59.83 (2.49) 61.76 (2.48)
1999
Fig. 6. Status of those not in paid employment between 1992 and 1999. A vast majority of non-workers reported that they were homemakers. While there was a decrease in participants seeking part-time work in 1996, there was a slight increase in retirement rates from 1997.
3. Discussion This longitudinal prospective research catalogues the changes in fundamental social roles in Australian women during the midlife to later life transition. The length of this study, with multiple assessments spanning over 12 years, enable a detailed inspection of the major changes occurring during this period and provide insight into the social identities of women as they age. One of the most prominent changes occurring during this period was to household composition. The proportion of participants with children living at home decreased steadily over the course of the study. This is consistent with the literature suggesting that midlife women are prime candidates for empty nest; a post-parenting period where adult children are no longer living at home [17–19]. While this could conceivably have a negative impact on social activity, older adults often find ways to compensate for the empty nest by pursuing other social activities such as socialising with friends or community programs [20]. While there was a drop in full-time employment as would be expected in this age bracket, not all of these participants retired. There was a rise in casual and occasional employment rates that coincided with this decrease. This suggests that some participants are not ready to retire completely after leaving full-time work. Moving from full-time employment to retirement can be a difficult change for many, both emotionally and financially. Many older adults decide to ease into the transition by gradually reducing work hours. In Australia, just over one quarter of retirees worked parttime before they retired [21]. In addition, part-time employment rates did not decrease as noticeably as full-time employment rates. This may be due to the differing pressures between full-time and part-time jobs. Individuals who work only part-time have been shown to have higher job satisfaction and more favourable attitudes towards many aspects of their employment compared to their full-time colleagues [22]. Giannikis and Milhail [23] suggested that part-time employees have fewer opportunities for workplace disagreements and are less likely to be exposed to office politics, which may heighten their job satisfaction. Furthermore, individuals in part-time employment experience less dissonance in the work-life balance than their fulltime counterparts [24]. The proportion of participants in the labour force was much lower than the proportion of partners in employment, suggesting that men are still the household breadwinners in the older generations. This is supported by the high proportion of non-workers reporting home duties as their primary occupation, indicating a tendency towards more traditional marital roles. In addition, studies have shown that women are more likely to retire early than men
[21,25]. Just under half of retired women in Australia report their husband’s income as their primary fund for living costs, surpassing the number of those relying on a pension [21]. Retirement rates in the current cohort appeared to accelerate at around 50–60 years of age. Although this is earlier than the official retirement age of 65, this trend is consistent with population data demonstrating an average retirement age of 50 years in Australian women [21]. The number of participants who engaged in volunteer work was consistently high throughout the study, even when full-time employment rates were at their peak. This finding suggests that employment and unpaid work are not mutually exclusive activities. Carr and Kail [26] proposed that engaging in both volunteerism and paid employment may be more compatible than previously thought, as they reflect a desire to be actively and productively engaged in society. Their study found that older adults who volunteer while still in full-time employment are more likely to transition into partial retirement than full retirement and are therefore more likely to continue working. While volunteer rates remained high over 12 years, there was an increase in popularity after 1997, corresponding to the rise in non-workers. This increase in volunteerism suggests that older individuals may compensate for the loss of one major role by engaging in other social activities. Grandparenthood was also an important event in this cohort. The vast majority of grandmothers reported that they spent some time minding their grandchildren, most on a weekly basis. Information regarding grandparenting was only gathered in the last timepoint, and so trends cannot be observed through midlife. However, like volunteering, it is possible that grandmothers may have more time to invest in their grandchildren after they have ceased paid employment. Minding grandchildren would also provide a source of day to day activity and purpose that may help fill the deficit left by retirement. The findings of the current study offer insight into the significant life events and changes in some of the major social roles that occur in the midlife to later life transition. There may be a number of other roles that are involved or are influenced by this shift that were not included in this study. Individuals are more than a single role; in order to inform choices, more research is needed on the myriad of social roles that are common in ageing, the factors that enable engagement in these roles, and the impact of these roles on health in later life. Social engagement is well established in the literature as vital to health in ageing. Although much of this evidence was collected two to three decades ago, social determinants of health have experienced a return to focus in very recent years. Growing evidence suggests that disability and disease are not only determined by pathophysiology, but also by personal experiences and social contexts [27]. Maintaining high levels of social activity into later life has been found to reduce morbidity and mortality risk, while also improving cognitive and emotional health [28–32]. Although the loss of social roles in the transition from midlife would have negative impacts on health later on, the inclusion of new social roles in
94
K. Burn et al. / Maturitas 88 (2016) 90–95
this period may have protective effects. The changes demonstrated in these results highlight the need for longitudinal research from midlife into late life that can provide detailed information on the complex social profile. Further work is currently underway in this longitudinal cohort to examine a broader range of social activities occurring in later life and their relationship with health in ageing.
ten informed consent was obtained from all participants prior to assessment at each timepoint. All data was de-identified. Provenance and peer review This article has undergone peer review.
4. Conclusion References The results of this study provide context for social engagement as individuals progress into later life. Change in employment status is clearly a key event in ageing. The decline in full-time work and dramatic rise in the number of non-workers presents an opportunity for exploring new social roles. While there was a slight increase in volunteerism and casual or occasional work that coincided with decreasing employment rates, there was still a large proportion of the sample that was not accounted for. Coupled with the large number of adult children leaving the home, there is a notable deficit in participants’ social lives as they approach the transition into later life; however, there is some evidence of compensation in the high rates of volunteering and time minding grandchildren. Research is required to investigate the impact these changes have on social activity in later life. Contributors KB is the primary author of the paper. KB formulated the study design and wrote the paper. LD established the initial study design and oversaw the project. CB assisted with the structure of the study, and contributed to the writing and editing of the paper. CS contributed to the study concept, and the writing and editing of the paper. Conflict of interest CS has provided clinical consultancy and been on scientific advisory committees for the Australian CSIRO, Alzheimer’s Australia, University of Melbourne and other relationships which are subject to confidentiality clauses. She has been a named Chief Investigator on investigator driven collaborative research projects in partnership with Pfizer, Merck, Piramal, Bayer and GE. Her research program has received support from NHMRC, Alzheimer’s Association, Collier Trust, Scobie and Claire McKinnon Foundation, JO and JR Wicking Trust, Shepherd Foundation, Brain Foundation, Mason Foundation, Ramaciotti Foundation, Alzheimer’s Australia, and the Royal Australian College of Physicians. She may accrue revenues from patent in pharmacogenomics prediction of seizure recurrence. The other authors have no conflicts of interest to declare. Funding This research is funded by Alzheimer’s Australia, Australian Menopausal Society, Brain Foundation, Alzheimer’s Association, Shepherd Foundation, the Scobie and Claire McKinnon Foundation, the Collier Trust Fund, the Ramaciotti Foundation, Bayer Pharmaceuticals, the J.O. & J.R. Wicking Trust, Piramal Life Sciences, Mason Foundation and the National Health and Medical Research Council (NHMRC). Ethical approval The current study received approval from the University of Melbourne Human Research Ethics Council (HREC 931149). Writ-
[1] B. Cornwell, E.O. Laumann, L.P. Schumm, The social connectedness of older adults: a national profile, Am. Sociol. Rev. 73 (2008) 185–203. [2] M. McPherson, L. Smith-Lovin, M.E. Brashears, Social isolation in America: changes in core discussion networks over two decades, Am. Sociol. Rev. 71 (2006) 353–375. [3] K.J. Ajrouch, A.Y. Blandon, T.C. Antonucci, Social networks among men and women: the effects of age and socioeconomic status, J. Gerontol. B: Psychol. Sci. Soc. Sci. 60 (2005) S311–S317. [4] J.M. Smith, Loneliness in older adults: an embodied experience, J. Gerontol. Nurs. 38 (2012) 45–53. [5] S. Balandin, N. Berg, A. Waller, Assessing the loneliness of older people with cerebral palsy, Disabil. Rehabil. 28 (2006) 469–479. [6] G.J. McInnis, J.H. White, A phenomenological exploration of loneliness in the older adult, Arch. Psychiatr. Nurs. 15 (2001) 128–139. [7] B.A. Shaw, N. Krause, J. Liang, J. Bennett, Tracking changes in social relations throughout late life, J. Gerontol. B: Psychol. Sci. Soc. Sci. 62 (2007) S90–S99. [8] T. Van Tilburg, M.B. Van Groenou, Network and health changes among older Dutch adults, J. Soc. Issues 58 (2002) 697–713. [9] K. Glaser, M. Evandrou, C. Tomassini, Multiple role occupancy and social participation among midlife wives and husbands in the United Kingdom, Int. J. Aging Hum. Dev. 63 (2006) 27–47. [10] B. Havens, M. Hall, G. Sylvestre, T. Jivan, Social isolation and loneliness: differences between older rural and urban Manitobans, Can. J. Aging 23 (2004) 129–140. [11] M. Van Willigen, Differential benefits of volunteering across the life course, J. Gerontol. B: Psychol. Sci. Soc. Sci. 55 (2000) S308–S318. [12] D.K. van den Hoonaard, A. Martin-Matthews, K. Davidson, The legacy of Helena Znaniecka Lopata in widowhood research: enduring concepts, life lessons, new insights, J. Aging Stud. 27 (2013) 457–463. [13] D. McLaughlin, J. Leung, N. Pachana, L. Flicker, G. Hankey, A. Dobson, Social support and subsequent disability: it is not the size of your network that counts, Age Ageing 41 (2012) 674–677. [14] H. Chemaitelly, C. Kanaan, H. Beydoun, M. Chaaya, M. Kanaan, A.M. Sibai, The role of gender in the association of social capital, social support, and economic security with self-rated health among older adults in deprived communities in Beirut, Qual. Life Res. 22 (2013) 1371–1379. [15] V. Carayanni, C. Stylianopoulou, G. Koulierakis, F. Babatsikou, C. Koutis, Sex differences in depression among older adults: are older women more vulnerable than men in social risk factors? The case of open care centers for older people in Greece, Eur. J. Ageing 9 (2012) 177–186. [16] L. Dennerstein, E.C. Dudley, J.L. Hopper, J.R. Guthrie, H.G. Burger, A prospective population-based study of menopausal symptoms, Obstet. Gynecol. 96 (2000) 351–358. [17] E.B. Harkins, Effects of empty nest transition on self-report of psychological and physical well-being, J. Marriage Fam. (1978) 549–556. [18] P.K. Adelmann, T.C. Antonucci, S.E. Crohan, L.M. Coleman, Empty nest, cohort, and employment in the well-being of midlife women, Sex Roles 20 (1989) 173–189. [19] C.E. Barber, Transition to the empty nest, Aging Fam. (1989) 15–32. [20] D. Chen, X. Yang, S.D. Aagard, The empty nest syndrome: ways to enhance quality of life, Educ. Gerontol. 38 (2012) 520–529. [21] Australian Bureau of Statistics, Retirement and Retirement Intentions, Australian Bureau of Statistics, Canberra, 2013. [22] B.J. Eberhardt, A.B. Shani, The effects of full-time versus part-time employment status on attitudes toward specific organizational characteristics and overall job satisfaction, Acad. Manage. J. 27 (1984) 893–900. [23] S.K. Giannikis, D.M. Mihail, Modelling job satisfaction in low-level jobs: differences between full-time and part-time employees in the Greek retail sector, Eur. Manage. J. 29 (2011) 129–143. [24] A. Roeters, L. Craig, Part-time work, women’s work-life conflict, and job satisfaction: a cross-national comparison of Australia, the Netherlands, Germany, Sweden, and the United Kingdom, Int. J. Comp. Sociol. 55 (3) (2014) 185–203, 0020715214543541. [25] N.E. Rice, I.A. Lang, W. Henley, D. Melzer, Common health predictors of early retirement: findings from the English Longitudinal Study of Ageing, Age Ageing 40 (2011) 54–61. [26] D.C. Carr, B.L. Kail, The influence of unpaid work on the transition out of full-time paid work, Gerontologist 53 (1) (2012) 92–101. [27] D. O’Connor, A. Phinney, A. Smith, J. Small, B. Purves, J. Perry, et al., Personhood in dementia care: developing a research agenda for broadening the vision, Dementia 6 (2007) 121–142. [28] P.A. Thomas, Trajectories of social engagement and mortality in late life, J. Aging Health 24 (2012) 547–568.
K. Burn et al. / Maturitas 88 (2016) 90–95 [29] C.F.M. de Leon, K.B. Rajan, Psychosocial influences in onset and progression of late life disability, J. Gerontol. B: Psychol. Sci. Soc. Sci. 69 (2) (2014) 287–302, gbt130. [30] S. Ramlagan, K. Peltzer, N. Phaswana-Mafuya, Social capital and health among older adults in South Africa, BMC Geriatr. 13 (2013) 100. [31] K.R. Krueger, R.S. Wilson, J.M. Kamenetsky, L.L. Barnes, J.L. Bienias, D.A. Bennett, Social engagement and cognitive function in old age, Exp. Aging Res. 35 (2009) 45–60.
[32] J.S. Saczynski, L.A. Pfeifer, K. Masaki, E.S. Korf, D. Laurin, L. White, et al., The effect of social engagement on incident dementia: the Honolulu-Asia Aging Study, Am. J. Epidemiol. 163 (2006) 433–440.
95
Contents lists available at ScienceDirect
Maturitas journal homepage: www.elsevier.com/locate/maturitas
Patterns of social engagement in the transition to later life Katherine Burn a , Lorraine Dennerstein b , Colette Browning c , Cassandra Szoeke a,∗ a
Department of Medicine, University of Melbourne, Parkville, Victoria, Australia Department of Psychiatry, University of Melbourne, Parkville, Victoria, Australia c Royal District Nursing Service, St Kilda, Victoria, Australia b
a r t i c l e
i n f o
Article history: Received 23 February 2016 Received in revised form 17 March 2016 Accepted 22 March 2016 Keywords: Social engagement Ageing Women
a b s t r a c t As social roles in later life are shaped by significant life events and changes occurring in the late-life transition, examining social engagement in midlife may provide a context for interpreting the ageing social identity. This is particularly important for women, who are heavily dependent on social relationships and are more influenced by social losses and change. Objective: To examine major social changes occurring in the decade prior to late-life, starting from approximately 45–55 years of age. Study design: The study accessed data from the longitudinal prospective Women’s Healthy Ageing Project (WHAP). Participants were 493 women who had completed at least one assessment in the first 12 years of the study. Results: Living with a partner was common and stable amongst participants, while the proportion with children still at home decreased markedly (79–44%). Full-time employment also decreased as participants approached the average retirement age (40–13%). Volunteer work was popular throughout the study, increasing slightly with age, and minding grandchildren was common at the end of the study period (80%). Conclusions: Taken together, these findings suggest a notable deficit in participants’ social lives as they transition into later life, but with some evidence of compensation by increasing other social activities. © 2016 Elsevier Ireland Ltd. All rights reserved.
The transition from midlife to later life denotes a key shift in social roles. Many social relationships and roles are lost and social network sizes have been shown to reduce [1–3]. Key events in this stage may include retirement from the workforce, the death of a spouse or friends, and a decrease in participation in hobbies due to poor physical health [4–9]. The loss of social contacts is often self-perpetuating as the loss of one role can lead to the loss of others. Losing a companion may lead to lessened feelings of encouragement or security, leaving individuals unwilling to strike out on their own [10,11]. Furthermore, older adults may experience difficulty or feel a reluctance to form new social bonds so late in life, especially when they are mourning lost relationships [6]. These factors are especially poignant for older women, as women are more likely to suffer from losses such as widowhood and are more dependent on wider social contact for wellbeing [12–15]. The losses occurring in the transition to old age
∗ Corresponding author at: P.O. Box 2026, Parkville, Victoria 3050, Australia. E-mail address: [email protected] (C. Szoeke). http://dx.doi.org/10.1016/j.maturitas.2016.03.016 0378-5122/© 2016 Elsevier Ireland Ltd. All rights reserved.
can influence the shape of social activity in later life. In their discussion of widowhood, Van den Hoonaard et al. [12] highlighted the conceptualisation of loss as a process; moving through the events surrounding the loss, mourning, through to the reconstruction of the individual’s social identity. As later life social roles are shaped by significant life events and changes occurring in the late-life transition, examining social engagement in midlife may provide a context for interpreting the ageing social identity. Unfortunately, much of the work examining age-related changes to social roles is outdated and focuses on American or European populations. As social roles are heavily influenced by cultural factors, population-specific research is needed in order to inform research. The current study aimed to examine social roles in midlife in order to provide a context for social engagement in the mid-to-late life transition for Australians.
K. Burn et al. / Maturitas 88 (2016) 90–95
91
Table 1 Description of WHAP social measures. Social factor
Items
Employment
Currently employed Employment type (full-time, part-time, etc.) Non-employed status (retired, looking for work, etc.) Partner’s employment status (full-time, retired, etc.) Single, married/de facto, divorced, widowed, separated Outlines all coresidents Changes to household occupants since last interview Number of grandchildren Minding grandchildren (number of grandchildren, frequency, activities) Current volunteer status Types of volunteering (verbatim) Organisations that participant belongs to
Marital status Household composition Grandparenting Volunteerism Community organisations
Fig. 1. Summary of the Australian Bureau of Statistics’ [27] General Social Survey (GSS). The survey covers common social relationships, opportunities for social engagement or support, and factors that influence social participation. These items range from innate factors to factors that influence the general social conditions. The GSS items encompass the measures used in the World Health Organisation’s [28] survey of the social determinants of mental health. Because this study examined social engagement, only items that involved direct personal contact were included (Household and Extended Networks items). Age, education, and gender were also included as part of the WHAP protocol. Data regarding contact with friends, neighbours, and other relatives was not available. The final selection of items for the WHAP protocol are shown underlined.
1. Methods
1.2. Measures and procedure
1.1. Participants
Assessments were conducted according to the project protocol which has been previously published [16]. Participants were administered a core questionnaire at each timepoint. The questionnaires from each timepoint contained a selection from a pool of items assessing social activities, relationships, and environments, outlined in Table 1. These items are consistent with the Australian Bureau of Statistics’ General Social Survey (GSS); a national survey of the opportunities and factors for active social participation (see Fig. 1). Items of interest to the current study included marital status, household composition, employment status, partners’ employment status, participation in unpaid work, and grandparenting.
Participants were from the epidemiologically sampled longitudinal prospective Women’s Healthy Ageing Project (WHAP) which was initiated as the Melbourne Women’s Midlife Health Project. Briefly, a longitudinal cohort of Caucasian women within the Melbourne metropolitan area were identified by random telephone dialling in 1991 and re-interviewed annually over eight years then intermittently over five years. Details of the study protocol are outlined elsewhere [16].
92
K. Burn et al. / Maturitas 88 (2016) 90–95
Marital Status
Employment Status 70
Percentage of respondents
80 70 60
Married/de facto
50
Separated
40
Divorced Widowed
30
Single
20 10
Percentagen of respondents
90
0 1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
60 50 Full-time 40
Part-time Casual/Occasional
30
Not employed 20
Volunteer
10 0 1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
Fig. 2. Reported marital status between 1992 and 2004. Most participants were married, which remained constant across the study. Separation rates decreased sharply while divorce rates increased in 1994, which may represent progress towards relationship termination.
Fig. 4. Paid and unpaid employment rates between 1992 and 2004. Note that participants were not asked about unpaid work in 2002 or 2004. Volunteerism was high throughout the study, with a slight increase in 1997 coinciding with the increase in those who were not employed. Casual and occasional work also increased over the course of the study, while full-time work decreased considerably.
Household Composition Participants' and their partners' employment status
90 90
70 60
Partner
50
Children
40
Parents/In-Laws Other adults
30
Lives alone
20 10 0 1992
1993
1994
1995
1996
1997
1998
1999
Fig. 3. Participants’ household composition 1992–1999. The proportion of participants living with a partner remained high throughout the study, while the proportion living with children dropped substantially. Note: household composition item was not a part of the questionnaire in 2002 and 2004.
Percentage of respondents
Percentage of respondents
80
80 70 60
Participants in full-time work
50
Partners in full-time work
40
Participants not employed
30 20
Partners not employed
10 0 1992 1993 1994 1995 1996 1997 1998 1999
Fig. 5. Participants’ employment rates compared to partners’ employment rates from 1992 to 1999. Item was not included in the questionnaire in 2002 and 2004. Participants were much less likely to be employed than their partners. However, both rates showed a negative trend across the study.
1.3. Data analysis Descriptive statistics such as response frequency were used to examine the prevalence of the different social connections across timepoints. Each role was then broken down into more detailed response categories and analysed using descriptive statistics. IBM SPSS was used to generate and graph all statistics. 1.4. Ethics The current study received approval from the University of Melbourne Human Research Ethics Council. Written informed consent was obtained from all participants prior to assessment at each timepoint. All data was de-identified. 2. Results The sample for the current study included 493 women who were interviewed at least once between 1992 and 2004. Sample sizes and age ranges for each timepoint are shown in Table 2. Marriage rates remained high and relatively stable across the study years, with a 5% decrease between 1992 and 2004 (Fig. 2). Similarly, the proportion of participants living with a partner remained high. The proportion living alone increased slightly from 1993 onwards, coinciding with the increase in sample divorce rates
(Fig. 3). Notably, the proportion of participants with children still at home dropped considerably from 78.8% in 1992 to 43.9% in 1999. Full-time employment decreased considerably between 1992 and 2004, dropping from 40% to only 13.1% of respondents. Although many of these would have retired, demonstrated by the increase in non-worker numbers, the percentage of participants in part-time employment increased slightly in 1998 and there was also an increase in casual and occasional work throughout the study (Fig. 4). The decrease in full-time working rates was mirrored in participants’ partners, which also decreased substantially (Fig. 5). While part-time working rates also decreased throughout the study, they did not experience as dramatic a change as full-time working rates, shown in Fig. 4. Participants not in paid employment increased markedly from 1997, when participants were aged approximately 50–61 years, coinciding with an increase in retirement rates. Of those who were not in paid employment, most indicated home duties as their main occupation (Fig. 6). Rates of volunteerism were noticeably high throughout the study, even when full-time employment rates were at their peak in 1991 (Fig. 4). Approximately 40% of participants were grandmothers, having an average of 3.99 (s.d. = 3.09) grandchildren. Of those who were grandmothers, nearly 80% stated that they were involved in minding their grandchildren, with more than half (53.5%) minding their grandchildren for at least a few hours a week.
K. Burn et al. / Maturitas 88 (2016) 90–95 Table 2 Sample sizes and age ranges by study year (1992–2004).
Status of Non-Workers 90
Percentage of respondents
80 Looking for full-time work Looking for part-time work Retired
70 60 50
Student
40 Non-worker
30 Home duties
20 10 0 1992
1993
1994
1995
1996
1997
1998
93
1992 1993 1994 1995 1996 1997 1998 1999 2002 2004
N
Age Range
Mean (s.d) age
491 473 459 453 447 443 436 433 256 284
45.66–57.04 46.76–57.83 47.70–58.90 48.45–59.96 49.48–60. 78 50.50–61.78 51.56–62.75 52.77–63.71 55.93–67.16 57.97–68.70
49.74 (2.55) 50.81 (2.54) 51.83 (2.55) 52.76 (2.51) 53.66 (2.50) 54.73 (2.50) 55.73 (2.49) 56.75 (2.49) 59.83 (2.49) 61.76 (2.48)
1999
Fig. 6. Status of those not in paid employment between 1992 and 1999. A vast majority of non-workers reported that they were homemakers. While there was a decrease in participants seeking part-time work in 1996, there was a slight increase in retirement rates from 1997.
3. Discussion This longitudinal prospective research catalogues the changes in fundamental social roles in Australian women during the midlife to later life transition. The length of this study, with multiple assessments spanning over 12 years, enable a detailed inspection of the major changes occurring during this period and provide insight into the social identities of women as they age. One of the most prominent changes occurring during this period was to household composition. The proportion of participants with children living at home decreased steadily over the course of the study. This is consistent with the literature suggesting that midlife women are prime candidates for empty nest; a post-parenting period where adult children are no longer living at home [17–19]. While this could conceivably have a negative impact on social activity, older adults often find ways to compensate for the empty nest by pursuing other social activities such as socialising with friends or community programs [20]. While there was a drop in full-time employment as would be expected in this age bracket, not all of these participants retired. There was a rise in casual and occasional employment rates that coincided with this decrease. This suggests that some participants are not ready to retire completely after leaving full-time work. Moving from full-time employment to retirement can be a difficult change for many, both emotionally and financially. Many older adults decide to ease into the transition by gradually reducing work hours. In Australia, just over one quarter of retirees worked parttime before they retired [21]. In addition, part-time employment rates did not decrease as noticeably as full-time employment rates. This may be due to the differing pressures between full-time and part-time jobs. Individuals who work only part-time have been shown to have higher job satisfaction and more favourable attitudes towards many aspects of their employment compared to their full-time colleagues [22]. Giannikis and Milhail [23] suggested that part-time employees have fewer opportunities for workplace disagreements and are less likely to be exposed to office politics, which may heighten their job satisfaction. Furthermore, individuals in part-time employment experience less dissonance in the work-life balance than their fulltime counterparts [24]. The proportion of participants in the labour force was much lower than the proportion of partners in employment, suggesting that men are still the household breadwinners in the older generations. This is supported by the high proportion of non-workers reporting home duties as their primary occupation, indicating a tendency towards more traditional marital roles. In addition, studies have shown that women are more likely to retire early than men
[21,25]. Just under half of retired women in Australia report their husband’s income as their primary fund for living costs, surpassing the number of those relying on a pension [21]. Retirement rates in the current cohort appeared to accelerate at around 50–60 years of age. Although this is earlier than the official retirement age of 65, this trend is consistent with population data demonstrating an average retirement age of 50 years in Australian women [21]. The number of participants who engaged in volunteer work was consistently high throughout the study, even when full-time employment rates were at their peak. This finding suggests that employment and unpaid work are not mutually exclusive activities. Carr and Kail [26] proposed that engaging in both volunteerism and paid employment may be more compatible than previously thought, as they reflect a desire to be actively and productively engaged in society. Their study found that older adults who volunteer while still in full-time employment are more likely to transition into partial retirement than full retirement and are therefore more likely to continue working. While volunteer rates remained high over 12 years, there was an increase in popularity after 1997, corresponding to the rise in non-workers. This increase in volunteerism suggests that older individuals may compensate for the loss of one major role by engaging in other social activities. Grandparenthood was also an important event in this cohort. The vast majority of grandmothers reported that they spent some time minding their grandchildren, most on a weekly basis. Information regarding grandparenting was only gathered in the last timepoint, and so trends cannot be observed through midlife. However, like volunteering, it is possible that grandmothers may have more time to invest in their grandchildren after they have ceased paid employment. Minding grandchildren would also provide a source of day to day activity and purpose that may help fill the deficit left by retirement. The findings of the current study offer insight into the significant life events and changes in some of the major social roles that occur in the midlife to later life transition. There may be a number of other roles that are involved or are influenced by this shift that were not included in this study. Individuals are more than a single role; in order to inform choices, more research is needed on the myriad of social roles that are common in ageing, the factors that enable engagement in these roles, and the impact of these roles on health in later life. Social engagement is well established in the literature as vital to health in ageing. Although much of this evidence was collected two to three decades ago, social determinants of health have experienced a return to focus in very recent years. Growing evidence suggests that disability and disease are not only determined by pathophysiology, but also by personal experiences and social contexts [27]. Maintaining high levels of social activity into later life has been found to reduce morbidity and mortality risk, while also improving cognitive and emotional health [28–32]. Although the loss of social roles in the transition from midlife would have negative impacts on health later on, the inclusion of new social roles in
94
K. Burn et al. / Maturitas 88 (2016) 90–95
this period may have protective effects. The changes demonstrated in these results highlight the need for longitudinal research from midlife into late life that can provide detailed information on the complex social profile. Further work is currently underway in this longitudinal cohort to examine a broader range of social activities occurring in later life and their relationship with health in ageing.
ten informed consent was obtained from all participants prior to assessment at each timepoint. All data was de-identified. Provenance and peer review This article has undergone peer review.
4. Conclusion References The results of this study provide context for social engagement as individuals progress into later life. Change in employment status is clearly a key event in ageing. The decline in full-time work and dramatic rise in the number of non-workers presents an opportunity for exploring new social roles. While there was a slight increase in volunteerism and casual or occasional work that coincided with decreasing employment rates, there was still a large proportion of the sample that was not accounted for. Coupled with the large number of adult children leaving the home, there is a notable deficit in participants’ social lives as they approach the transition into later life; however, there is some evidence of compensation in the high rates of volunteering and time minding grandchildren. Research is required to investigate the impact these changes have on social activity in later life. Contributors KB is the primary author of the paper. KB formulated the study design and wrote the paper. LD established the initial study design and oversaw the project. CB assisted with the structure of the study, and contributed to the writing and editing of the paper. CS contributed to the study concept, and the writing and editing of the paper. Conflict of interest CS has provided clinical consultancy and been on scientific advisory committees for the Australian CSIRO, Alzheimer’s Australia, University of Melbourne and other relationships which are subject to confidentiality clauses. She has been a named Chief Investigator on investigator driven collaborative research projects in partnership with Pfizer, Merck, Piramal, Bayer and GE. Her research program has received support from NHMRC, Alzheimer’s Association, Collier Trust, Scobie and Claire McKinnon Foundation, JO and JR Wicking Trust, Shepherd Foundation, Brain Foundation, Mason Foundation, Ramaciotti Foundation, Alzheimer’s Australia, and the Royal Australian College of Physicians. She may accrue revenues from patent in pharmacogenomics prediction of seizure recurrence. The other authors have no conflicts of interest to declare. Funding This research is funded by Alzheimer’s Australia, Australian Menopausal Society, Brain Foundation, Alzheimer’s Association, Shepherd Foundation, the Scobie and Claire McKinnon Foundation, the Collier Trust Fund, the Ramaciotti Foundation, Bayer Pharmaceuticals, the J.O. & J.R. Wicking Trust, Piramal Life Sciences, Mason Foundation and the National Health and Medical Research Council (NHMRC). Ethical approval The current study received approval from the University of Melbourne Human Research Ethics Council (HREC 931149). Writ-
[1] B. Cornwell, E.O. Laumann, L.P. Schumm, The social connectedness of older adults: a national profile, Am. Sociol. Rev. 73 (2008) 185–203. [2] M. McPherson, L. Smith-Lovin, M.E. Brashears, Social isolation in America: changes in core discussion networks over two decades, Am. Sociol. Rev. 71 (2006) 353–375. [3] K.J. Ajrouch, A.Y. Blandon, T.C. Antonucci, Social networks among men and women: the effects of age and socioeconomic status, J. Gerontol. B: Psychol. Sci. Soc. Sci. 60 (2005) S311–S317. [4] J.M. Smith, Loneliness in older adults: an embodied experience, J. Gerontol. Nurs. 38 (2012) 45–53. [5] S. Balandin, N. Berg, A. Waller, Assessing the loneliness of older people with cerebral palsy, Disabil. Rehabil. 28 (2006) 469–479. [6] G.J. McInnis, J.H. White, A phenomenological exploration of loneliness in the older adult, Arch. Psychiatr. Nurs. 15 (2001) 128–139. [7] B.A. Shaw, N. Krause, J. Liang, J. Bennett, Tracking changes in social relations throughout late life, J. Gerontol. B: Psychol. Sci. Soc. Sci. 62 (2007) S90–S99. [8] T. Van Tilburg, M.B. Van Groenou, Network and health changes among older Dutch adults, J. Soc. Issues 58 (2002) 697–713. [9] K. Glaser, M. Evandrou, C. Tomassini, Multiple role occupancy and social participation among midlife wives and husbands in the United Kingdom, Int. J. Aging Hum. Dev. 63 (2006) 27–47. [10] B. Havens, M. Hall, G. Sylvestre, T. Jivan, Social isolation and loneliness: differences between older rural and urban Manitobans, Can. J. Aging 23 (2004) 129–140. [11] M. Van Willigen, Differential benefits of volunteering across the life course, J. Gerontol. B: Psychol. Sci. Soc. Sci. 55 (2000) S308–S318. [12] D.K. van den Hoonaard, A. Martin-Matthews, K. Davidson, The legacy of Helena Znaniecka Lopata in widowhood research: enduring concepts, life lessons, new insights, J. Aging Stud. 27 (2013) 457–463. [13] D. McLaughlin, J. Leung, N. Pachana, L. Flicker, G. Hankey, A. Dobson, Social support and subsequent disability: it is not the size of your network that counts, Age Ageing 41 (2012) 674–677. [14] H. Chemaitelly, C. Kanaan, H. Beydoun, M. Chaaya, M. Kanaan, A.M. Sibai, The role of gender in the association of social capital, social support, and economic security with self-rated health among older adults in deprived communities in Beirut, Qual. Life Res. 22 (2013) 1371–1379. [15] V. Carayanni, C. Stylianopoulou, G. Koulierakis, F. Babatsikou, C. Koutis, Sex differences in depression among older adults: are older women more vulnerable than men in social risk factors? The case of open care centers for older people in Greece, Eur. J. Ageing 9 (2012) 177–186. [16] L. Dennerstein, E.C. Dudley, J.L. Hopper, J.R. Guthrie, H.G. Burger, A prospective population-based study of menopausal symptoms, Obstet. Gynecol. 96 (2000) 351–358. [17] E.B. Harkins, Effects of empty nest transition on self-report of psychological and physical well-being, J. Marriage Fam. (1978) 549–556. [18] P.K. Adelmann, T.C. Antonucci, S.E. Crohan, L.M. Coleman, Empty nest, cohort, and employment in the well-being of midlife women, Sex Roles 20 (1989) 173–189. [19] C.E. Barber, Transition to the empty nest, Aging Fam. (1989) 15–32. [20] D. Chen, X. Yang, S.D. Aagard, The empty nest syndrome: ways to enhance quality of life, Educ. Gerontol. 38 (2012) 520–529. [21] Australian Bureau of Statistics, Retirement and Retirement Intentions, Australian Bureau of Statistics, Canberra, 2013. [22] B.J. Eberhardt, A.B. Shani, The effects of full-time versus part-time employment status on attitudes toward specific organizational characteristics and overall job satisfaction, Acad. Manage. J. 27 (1984) 893–900. [23] S.K. Giannikis, D.M. Mihail, Modelling job satisfaction in low-level jobs: differences between full-time and part-time employees in the Greek retail sector, Eur. Manage. J. 29 (2011) 129–143. [24] A. Roeters, L. Craig, Part-time work, women’s work-life conflict, and job satisfaction: a cross-national comparison of Australia, the Netherlands, Germany, Sweden, and the United Kingdom, Int. J. Comp. Sociol. 55 (3) (2014) 185–203, 0020715214543541. [25] N.E. Rice, I.A. Lang, W. Henley, D. Melzer, Common health predictors of early retirement: findings from the English Longitudinal Study of Ageing, Age Ageing 40 (2011) 54–61. [26] D.C. Carr, B.L. Kail, The influence of unpaid work on the transition out of full-time paid work, Gerontologist 53 (1) (2012) 92–101. [27] D. O’Connor, A. Phinney, A. Smith, J. Small, B. Purves, J. Perry, et al., Personhood in dementia care: developing a research agenda for broadening the vision, Dementia 6 (2007) 121–142. [28] P.A. Thomas, Trajectories of social engagement and mortality in late life, J. Aging Health 24 (2012) 547–568.
K. Burn et al. / Maturitas 88 (2016) 90–95 [29] C.F.M. de Leon, K.B. Rajan, Psychosocial influences in onset and progression of late life disability, J. Gerontol. B: Psychol. Sci. Soc. Sci. 69 (2) (2014) 287–302, gbt130. [30] S. Ramlagan, K. Peltzer, N. Phaswana-Mafuya, Social capital and health among older adults in South Africa, BMC Geriatr. 13 (2013) 100. [31] K.R. Krueger, R.S. Wilson, J.M. Kamenetsky, L.L. Barnes, J.L. Bienias, D.A. Bennett, Social engagement and cognitive function in old age, Exp. Aging Res. 35 (2009) 45–60.
[32] J.S. Saczynski, L.A. Pfeifer, K. Masaki, E.S. Korf, D. Laurin, L. White, et al., The effect of social engagement on incident dementia: the Honolulu-Asia Aging Study, Am. J. Epidemiol. 163 (2006) 433–440.
95