Personality and Other Lifelong Influences on Older-Age Health and Wellbeing: Preliminary Findings in Two Scottish Samples

Abstract

Recent observations that personality traits are related to later-life health and wellbeing have inspired considerable interest in exploring the mechanisms involved. Other factors, such as cognitive ability and education, also show longitudinal influences on health and wellbeing, but it is not yet clear how all these early-life factors together contribute to later-life health and wellbeing. In this preliminary study, we assessed hypothesised relations among these variables across the life course, using structural equation modelling in a sample assessed on dependability (a personality trait related to conscientiousness) in childhood, cognitive ability and social class in childhood and older age, education, and health and subjective wellbeing in older age. Our models indicated that both health and subjective wellbeing in older age were influenced by childhood IQ and social class, via education. Some older-age personality traits mediated the effects of early-life variables, on subjective wellbeing in particular, but childhood dependability did not show significant associations. Our results therefore did not provide evidence that childhood dependability promotes older-age health and wellbeing, but did highlight the importance of other early-life factors, particularly characteristics that contribute to educational attainment. Further, personality in later life may mediate the effects of early-life factors on health and subjective wellbeing. © 2016 The Authors. European Journal of Personality published by John Wiley & Sons Ltd on behalf of European Association of Personality Psychology.

Keywords: cognitive ability; education; health; longitudinal study; personality; subjective wellbeing.

Figure 1

Model of older‐age Health. Paths shown in grey were non‐significant at p < .05 and were therefore constrained to zero in the constrained model and removed from subsequent models. Variables in grey showed no significant association with any other variables and were therefore removed from the trimmed and final models. Variables shown in blue had no direct or indirect effect on the primary outcome, older‐age Health, and were therefore removed from the final model. Path labels represent standardised regression weights derived from the illustrated initial model, which, for retained paths, did not differ substantially in subsequent models.

Figure 2

Model of older‐age Wellbeing. Paths shown in grey were non‐significant at p < .05 and were therefore constrained to zero in the constrained model and removed from subsequent models. Variables in grey showed no significant association with any other variables and were therefore removed from the trimmed and final models. Variables shown in blue had no direct or indirect effect on the primary outcome, older‐age Health and were therefore removed from the final model. Path labels represent standardised regression weights derived from the illustrated initial model, which, for retained paths, did not differ substantially in subsequent models.

Figure 3

Model of older‐age Health via older‐age Wellbeing. Paths shown in grey were non‐significant at p < .05 and were therefore constrained to zero in the constrained model and removed from subsequent models. Variables in grey showed no significant association with any other variables and were therefore removed from the trimmed and final models. Variables shown in blue had no direct or indirect effect on the primary outcome, older‐age Health, and were therefore removed from the final model. Path labels represent standardised regression weights derived from the illustrated initial model, which, for retained paths, did not differ substantially in subsequent models.

Figure 4

Model of older‐age Wellbeing via older‐age Health. Paths shown in grey were non‐significant at p < .05 and were therefore constrained to zero in the constrained model and removed from subsequent models. Variables in grey showed no significant association with any other variables and were therefore removed from the trimmed and final models. Variables shown in blue had no direct or indirect effect on the primary outcome, older‐age Health, and were therefore removed from the final model. Path labels represent standardised regression weights derived from the illustrated initial model, which, for retained paths, did not differ substantially in subsequent models.