As with affective disorders in the early and mid-life ages, late-life mood disorders are quite heterogeneous. Age of onset appears to be an important clinical variable in the genetic study of affective illness. Therefore, two groups of affectively ill elderly patients are usually differentiated: those who had an early onset of illness with continued symptoms or episodes in their later life, and those who had a late onset of illness (usually defined as a first onset after the age of 60 years)⁹. When differentiated by age of onset, late-onset depressive disorders have qualities suggesting that they may be genetically different from early onset forms, including clinical differences in expression and non-conformity to expected genetic models¹⁰.

For example, basic genetic models propose that diseases with strong genetic influences are most often found in earlier onset and/or more severe forms of an illness. However, late-onset affective syndromes are typically among the most severe and most refractory to treatment^11,12. Further, many studies have consistently found that patients with late-onset depression and bipolar disorders have less familial aggregation of mood disorders than those with the early onset affective disorders^13-18. Also, late-onset mood disorders are also less likely to display the gender disparity present in early onset mood disorders¹³. These considerations challenge our models of genetic influence for late-life affective disorders.

Two explanations have been proposed for the age-related variation in genetic influence for affective illnesses: multifactorial inheritance and increased phenocopy expression with age. Multifactorial inheritance (also called polygenic inheritance) suggests that inheritance patterns are a result of a combination between multiple interacting genes and environmental factors. Examples of traits that have multi- factorial inheritance include stature, intelligence or blood pressure¹⁰. According to the multifactorial theory, the onset of an affective disorder results from the additive effects of several genetic and environment factors. One could also assume that early onset illnesses probably result from greater genetic loading, while late-onset illnesses may have less genetic loading but increased occurrence of environmental events with age.

The second explanation, increased phenocopies, may explain the limited specificity of the affective disorders diagnosis. Both bipolar disorder and major depressive disorder are syndrome diagnoses; that is, the diagnosis is based on the presence of a collection of observed symptoms (phenotypes) rather than an etiological cause. In this theory, late-onset affective illness is a collection of pheno- copies. Phenocopies are illnesses with a characteristic phenotype but different underlying causes.

It has long been known that depressive syndromes are very common in neurological disorders such as Parkinson’s disease, Alzheimer’s disease and stroke; more recently added to this list would be the proposed subtype of depression called ‘vascular depression’¹⁹‘²⁰. Taking these into account, the age-related variation in late-life affective disorders may be explained by increasing phenocopies in older adults who have late-onset affective disorders. Patients with early onset affective disorders (which would include elderly patients with previous diagnosis of affective disorders) may be, therefore, more likely to have a genetic cause. As research sharpens our understanding of these diseases, the ‘heterogeneous’ group of affective disorders will be differentiated into various subtypes, each with a more specifically defined course, prognosis and treatment. Each new subtype will also suggest new ways that genes and the environment may contribute to the expression of the illness.

There is one further consideration in studying genetic contributions to late-life affective disorders. This is the fact that not all subjects with higher or lower genetic susceptibility may reach the developmental age of expression. Competing causes of death must be taken into account (i.e. individuals at high risk of developing a disorder may not express the disease symptoms during their lifetime). This in turn means that the actual risk of developing a disease is decreased when other causes of mortality are in play²¹.

First-degree relatives (parents, siblings, offspring) of depressed adult patients are twice as likely to develop depression as those in the general population^{22 – 26}. In two twin studies of adult populations, the relative risk for monozygotic twins to develop depression was 1.9, while the risk for dizygotic twins was 1.2^27,28. Adoption studies have suggested that certain environmental influences, especially parental loss and parental alcoholism, may be predictors of depression²⁹. Heritabil- ity, therefore, is estimated at around 40% in the adult population³⁰; however, it has also been estimated to be lower in the elderly^31,32. Evidence of only limited heritability (16%) was found in a sample of elderly reared-apart and reared-together Swedish twins studied for characteristics of depression⁴. In contrast, a sample of Danish twins 75 years of age and older revealed that depression symptomatology is moderately heritable in late life (approximately 35%)⁵. Moreover, both studies consistently implicated environmental factors as the major source of variance in depression symptoms among the elderly.