Psychological Medicine

Age differences in depression and anxiety symptoms: a structural equation modelling analysis of data from a general population sample

a1 NHMRC Psychiatric Epidemiology Research Centre, The Australian National University, Canberra; and Department of Psychological Medicine, Monash University and Mental Health Research Institute of Victoria, Melbourne, Australia


Background. There is debate as to whether the elderly are really at lower risk for depressive disorders, or whether endorsement of symptoms is artefactually low. The present paper assesses the effects of age on anxiety and depression, and examines whether age has direct effects on self-report of individual symptoms independent of its effect on the underlying dimensions of anxiety and depression.

Methods. Structural equation modelling was used to assess the structure of the items and their associations with age and a number of demographic variables. The sample of 2622 participants aged between 18 and 79 years from Canberra (Australia) was drawn from the Electoral Roll. Two instruments were used: the anxiety and depression scales of Goldberg et al. (1988) and the Personal Disturbance Scale from the DSSI of Bedford et al. (1976).

Results. Both scales were found to fit satisfactorily to a two factor model. Age correlated negatively with depression. After controlling for the effects of gender, marital status, education and financial difficulty, direct effects of age were found on items from both instruments, indicating that certain depression items were associated with a differential probability of endorsement in older people, even when the level of depression was equal to that of younger people. Items with direct age effects reflected physical (feeling slowed down; waking early) and psychological (hopeless about the future) components of depression. Direct effects of age on items from both anxiety scales were also found.

Conclusions. The nature of the depression and anxiety experienced by younger and older people may differ qualitatively. Depression may be associated with an increase in somatic symptoms linked to physical changes and to an increase in endorsement of items which reflect the narrowing of opportunities in the long-term.

c1 Address for correspondence: Dr Helen Christensen, NHMRC Psychiatric Epidemiology Research Centre, the Australian National University, Canberra, ACT 0200, Australia.