Are changes in self-rated health associated with memory decline in older adults? Bendayan, R., Piccinin, A. M., Hofer, S. M., & Muniz-Terrera, G


Keywords: Self-Rated Health; Memory; Cognitive Decline; Older Adults Introduction



Download 219,59 Kb.
Page3/8
Date01.10.2019
Size219,59 Kb.
1   2   3   4   5   6   7   8
Keywords: Self-Rated Health; Memory; Cognitive Decline; Older Adults

Introduction

Self-rated health (SRH) reflects the general perception of an individual’s health status and it is one of the most inclusive and informative indicators of health status (Jylhä, 2009). Moreover, it is known to be sensitive to health changes (Galenkamp et al., 2013) and has been shown to be a robust predictor of mortality and morbidity (Idler & Benyamini, 1997; Rutledge et al., 2010). Research to date has also found a positive association between SRH and cognitive performance in older adults (Anstey & Christensen, 2000; Earles, Connor, Smith & Park, 1997; Hultsch, Hertzog, Small & Dixon, 1999; Walker, Maxwell, Hogan & Ebly, 2004) and some studies also reported that poorer SRH predicts cognitive decline (Carmelli, Swan, LaRue & Eslinger, 1997; Sargent-Cox, Cherbuin, Sachdev & Anstey, 2011), cognitive impairment (Bond, Dickinson, Matthews, Jagger & Brayne, 2006) and all-cause dementia (John & Montgomery, 2013; Kato et al, 2013; Monthlaluc et al., 2011, Weisen, Frishman, Aronson & Wassertheil-Smoller, 1998; Yip, Brayne & Matthews, 2006).

In cognitive ageing research, most investigations have considered static baseline measures of SRH (Earles et al., 1997; John & Montgomorey, 2013; Kato et al, 2013; Monthlaluc et al., 2011; Sargent-Cox et al., 2011 Weisen et al., 1999; Yip et al., 2006) but recent publications highlight its dynamic nature (Leinonen, Heikkinen & Jylhä, 2001, 2002; Rohlfsen & Kronenfeld, 2014; Vogelsang, 2014; Wilson, Elliot, Eyles & Keller-Olaman, 2007). To date, only a few studies have examined the association between changes in SRH and cognition (Carmelli et al, 1997; Hultsch et al. 1999; Leinonen et al., 2001). For instance, Carmelli et al. (1997) considered changes in health rating and its association with cognitive change using two time points over 6 years. They identified three patterns of change in overall cognitive performance in older adults (i.e., decliners, non-changers and improvers) and found that decliners rated their health as poor at baseline and were also the only ones to show changes in SRH. Hultsch et al. (1999) examined the association between lifestyle and a number of cognitive tests (i.e., recall tasks, vocabulary, verbal fluency, working memory, reading comprehension, and comprehension and semantic speed) in 3 occasions over 6 years. They found that, although most of the cognitive variables (except fact recall, vocabulary and verbal fluency) and SRH were significantly associated at baseline, changes in the cognitive variables were not associated with baseline SRH or changes in SRH. Leinonen et al. (2001) identified different patterns of change in SRH between two time points with a 5 year interval between them and found that those whose psychomotor speed and perception decreased substantially also showed a decline in SRH. These inconsistencies across studies could be associated with the different cognitive domains considered, the use of only a limited number of measuring occasions to examine change or the characteristics of the samples considered. In addition, the above mentioned studies used analytical strategies (e.g., t-tests) that consider the average change of the group of individuals and do not account for between individual variability. However, each individual might have different scores at baseline and different rates of decline over time. Over the last decade, advances on longitudinal data analysis techniques have provided researchers the opportunity to explore how individuals change over time and how these changes vary between individuals (e.g., linear mixed models). Within this context, studies that examine the association between changes in SRH and changes in cognitive performance in older adults taking into account individual variability are needed.

The main aim of this paper is to examine the association between the different patterns of change in SRH and memory trajectories in older adults. In order to it, we analyse data from a nationally representative survey of US older adults, the Health Retirement Study (HRS), and replicate the statistical procedure in a nationally representative survey of English older adults, the English Longitudinal Study of Ageing (ELSA). Replication studies are essential to build scientific knowledge and recent publications have highlighted the need to promote systematic replication efforts (Koole & Lakens, 2012; Open Science, 2015), especially in longitudinal studies of aging (Hofer & Piccinin, 2009).


Methods

Sample

Data are drawn from the Health and Retirement Study (HRS) from the US, which is a biannual, longitudinal and nationally representative surveys that focus on adults aged 50 and over. Details can be found in their respective websites: http://hrsonline.isr.umich.edu/. For the analyses presented here, HRS data from 1998 (wave 4) to 2006 (wave 8) were included. Wave 4 of HRS was selected as “baseline” as 1998 is when HRS and AHEAD were fully integrated and when HRS assessment procedure matched other sister studies as the English Longitudinal Study of Ageing (ELSA). For the purpose of this study, the first wave considered will be labelled as baseline and the subsequent ones follow-ups.

For replication purposes, data from 2000 (wave 1) to 2008 (wave 5) of the sister study of HRS in England, ELSA, were also examined. Details can be found in http://www.elsa-project.ac.uk/. Both samples were restricted to respondents who were between 65 and 80 years old at the baseline and had at least three valid measurements for memory and self-rated health, following Singer and Willet (2003) recommendations. Therefore, the US sample consisted of a subsample of 6016 respondents (58.7% women) of the HRS and the English sample consisted of a subsample of 734 respondents (42.5% were women) of the ELSA. Relevant descriptive statistics are shown in Table 1 for HRS (Supplemental Table 1 for ELSA).

PLEASE, INSERT TABLE 1 AROUND HERE





Share with your friends:
1   2   3   4   5   6   7   8


The database is protected by copyright ©psysite.info 2019
send message

    Main page