Objectives: to assess longitudinal relationships between habitual levels of
physical activity and indices of psychological wellbeing in older people.
Design: baseline assessment with 4- and 8-year follow-ups.
Subjects: 1042 people originally aged 65 and over randomly sampled from gen
eral practitioner Lists in Nottingham, UK.
Methods: logistic regression analysis of selected T1 (1985) and T2 (1989) v
ariables, with depression at T2 as dependent; multiple regression analyses
of selected T1, T2 and T3 (1993) variables, with life Satisfaction at T2 (m
odel 1) or T3 (model 2) as dependent.
Main outcome measures: questionnaire-assessed levels of physical activity;
14-item Symptoms of Anxiety and Depression scale; 13-item Life Satisfaction
Index; health, demographic and social activity variables.
Results: in the logistic regression model, depression at T2 was most strong
ly associated with depression [odds ratio (OR) = 7.13; 95% confidence inter
val (CI) = 3.25 -15.64; P < 0.001] and lower physical health status (OR = 1
.26 per unit change in score; 95% CI = 1.17 - 1.42; P < 0.001) at T1. Lower
levels of outdoor/leisure activities at T1 were also associated with some
increased risk of depression 4 years later (OR = 0.92 per hour of activity;
95% CI = 0.85 - 0.99; P < 0.05). Similar predictive patterns emerged from
the multiple regression analyses where, in both models, earlier levels of l
ife satisfaction, social engagement and health accounted for most of the ex
plained variance in life satisfaction (R-2 = 0.42 for model I; R-2 = 0.35 f
or model 2). Again, however, earlier levels of physical activity (as walkin
g and housework) did contribute significantly, although modestly: to longit
udinal changes in morale.
Conclusions: while the results provide some support for the conclusion that
physical activity contributes independently to the promotion and maintenan
ce of psychological wellbeing in later life, this contribution is, at best,
extremely modest.