Search

Unpaid caregivers provide most of the support for individuals with social care needs in England and maintaining caregiver health and wellbeing is important to sustaining their role. Sleep disturbance has been linked to poorer health and lower quality of life, and, among caregivers, may be a reason for ceasing care. However, our understanding of caregiver sleep from longitudinal and population-based studies is limited. This thesis investigated relationships between caregiving and sleep disturbance among women and men aged fifty and over in England. It was guided by the stress process framework and insights from the sociology of sleep. The conceptual framework for the study focused on characteristics of caregiving and periods of transition and stability likely to be stressful, as well as the influence of gendered social roles, and changes in partner health among couples. The research aimed to identify, first, how care hours and relationship to the care recipient were associated with sleep disturbance; second, differences in patterns of sleep disturbance across a care episode and by care duration, and third, associations with co-resident care and partner health in the context of cohabiting couples. Fixed effect models were the main statistical approach used to analyse change within individuals, on data drawn primarily from Waves 4, 6 and 8 of the English Longitudinal Study of Ageing (ELSA). Wave 1 of the ELSA COVID-19 sub-study was also analysed, during a period likely to exacerbate stressors for some caregivers. Results indicated limited associations between some aspects of unpaid care, partner health and sleep disturbance. Spouse care and care of 20 or more hours per week were linked to small increases in the probability of sleep disturbance among men. Among women, sleep disturbance was related to care durations of two or more waves, and unexpectedly, ceasing care, while depressive symptoms accounted for associations in some analyses. Among couples, no association was found with co-resident care, but several partner health conditions were linked to sleep disturbance among women. As expected, changes in individual health were strongly associated with sleep disturbance. Limited evidence supported hypothesised sleep disturbance linked to caregiving, however, sleep disturbance may be more transient than could be measured by the data available. Future research would benefit from availability of more frequently measured sleep data enabling closer modelling of changing circumstances and sleep.

Background & Objectives. In light of population ageing, policies aimed at extending working lives have been implemented in several higher income countries, including the UK. However, given older adults' substantial contributions through unremunerated forms of work, such as informal care and volunteering, longer working lives may have repercussions for engagement in unpaid activities, which are important for their economic value and their influence on older people's health and wellbeing. Using the British Household Panel Survey and its continuation, the UK Household Longitudinal Study, this thesis examined the effect of employment status on engagement in unpaid forms of work in mid to later life (55-70). Two different approaches were used. Firstly, pathways of engagement in multiple activities are examined using two-stage latent class analysis (LCA), to provide a nuanced picture of how individuals combine engagement in paid and unpaid activities over time. Secondly, the longitudinal relationship between employment status and engagement in informal caregiving and volunteering is examined using within-between random effects (WB-RE) models that disentangle direct influences from selection effects. Results. Findings using two-stage LCA identified three distinct engagement pathways: paid workers (43%), mixed activities (housework) (46%), and mixed activities (volunteers) (11%). Paid workers were more likely to work full-time and less likely to engage in unpaid work. Individuals classified in the other two pathways were more likely to combine paid and unpaid activities. Gender differences were conspicuous, as more women conformed to the second and third pathways. Being classified in one of the three engagement paths was associated to baseline sociodemographic and health characteristics. WB-RE models showed that full-time employment was consistently associated with lower odds of caregiving (any and at least 20 hours per week, hpw), through both within and between person effects. Time-lagged models showed that women who work part-time had lower odds of caregiving 20 or more hpw, while men in full-time self-employment had decreased odds of caregiving at all intensities (between person effects). Among caregivers, full-time employment was associated with lower odds of more intensive caregiving (20 hpw) at follow-up, according to both within and between person effects; additionally, part-time paid work and self-employment were negatively associated with providing more intensive care according to between person effects. Full-time employment was consistently negatively associated with monthly and weekly volunteering, through within and between person effects. Time-lagged models showed that part-time employment and full-time self-employment led to lower odds of monthly volunteering (within person effects) among men. Part-time employment also led to decreased odds of weekly volunteering. Men who were in part-time self-employment had higher odds of monthly and weekly volunteering (between person effects). Among female volunteers, full-time employment led to lower odds of more frequent volunteering at follow-up (within person effects). Among male and female volunteers, those who were in full-time self-employed had lower odds of more frequent volunteering (between person effects). Conclusion. Full-time employment led to a lower likelihood of engagement in caregiving and volunteering, with both direct and selection effects driving this relationship. The effect of part-time work and self-employment was mainly mediated through selection effects. Individuals' engagement in unpaid work should be taken into account in the context of longer working lives, as participation in full-time employment limits engagement in informal caregiving and volunteering in mid to later life. Findings have also important implications for active ageing related policies, as they emphasise the importance of gender, socioeconomic and health inequalities as determinants of engagement in paid and unpaid work in later life.