Late-life depression: the differences between early- and late-onset illness in a community-based sample
Janssen J, Beekman ATF, Comijs HC, Deeg DJH and Heeren TJ;
Commented by , 20 Feb 2006
Background
In older people with depression, several studies based in secondary care have found differences in aetiology and symptoms between early onset depression (EOD) and late onset depression (LOD), where LOD is defined as depression presenting de novo after age 60. EOD is in general associated with increased family history of depression/genetic susceptibility; LOD with increased cerebrovascular pathology. The aim of this study was to see if this trend is replicated in a community sample of older depressives.
Methods
Participants were drawn from the wider Longitudinal Ageing Study Amsterdam project. This has followed up a random sample of 3,107 55-85 year olds in the Netherlands since baseline assessment in 1992. After initial screening for depression, diagnosis of depression was made using the Diagnostic Interview Schedule, leading to DSM-III diagnoses of Major Depressive Disorder (MDD) or double depression (MDD superimposed on dysthymia).
The baseline diagnostic assessment was repeated at 3 and 6 years. A large amount of demographic information on participants was available for analysis. Analysis was by logistic regression and odds ratios of the associations between independent variables (cognition, health, life events, depressive symptoms) compared with age of onset of depression.
Results
129 subjects (of the original sample of 3,107) had MDD during the 6 year span.
With respect to aetiology, LOD was weakly associated with being older, being widowed, and having impaired cognition. EOD was associated with double depression, and experiencing World War II as a traumatic event, but not family history.
With respect to depressive symptoms, the only depressive symptom associated with LOD was weight loss. EOD was associated with co-morbid anxiety, feeling worthless, and suicidal thoughts.
Dr Seymour's comments
This study examines the question: is age of onset of depression in older people relevant to how their depression is managed? Previous literature on EOD vs LOD can be summed up in Brodaty’s phrase: "different aetiology, same phenomenology".
The main finding of this study is with respect to aetiology: in the community sample, there were some differences in aetiology between EOD and LOD, but the previous finding in secondary care of increased genetic susceptibility in EOD, but increased vascular pathology in LOD was not confirmed. This is important because it implies that all older people with depression should be treated with equal vigour, ie., it should not be assumed that older people with LOD have vascular depression and hence a worse prognosis.
There are several limitations to this study which the authors acknowledge. The prevalence and incidence of MDD detected was lower than in previously reported epidemiological studies, and at all stages of the study the older and frailer subjects were at a greater risk of dropping out, potentially skewing the results. Age of onset of depression relied on self report, which may be inaccurate, particularly in people with double depression.
Why do the findings of this community-based study differ from previous studies in secondary care? Only a minority of older patients with MDD are referred to secondary care, potentially biasing secondary care results: EOD cases are more likely to be known to psychiatric services, and LOD cases are more likely to be referred after "trigger" events such as transient ischaemic attacks. Therefore, this community-based study is an important contribution to the literature.
In summary, EOD and LOD may not be as aetiologically distinct as previously thought, so therapeutic nihilism in LOD should be avoided.