aging, mental health, and well being 81

satisfaction with life, on average, than young
and middle aged adults, although the differ
ences are substantively modest (Campbell
et al. 1976). These age differences have been
consistent for more than 30 years and do not
result from older adults being more advan
taged than young and middle aged adults in
objective life conditions (Horley & Lavery
1995). Aspiration theory explains age differ
ences in life satisfaction. According to this
theory, individuals are satisfied with life when
there is little discrepancy between their aspira
tions and their achievements and, conversely,
are dissatisfied when there is a large discre
pancy. Older adults higher levels of subjective
well being result from their lower aspirations,
on average, than those of young and middle
aged persons. It remains unclear whether
these age differences result from cohort differ
ences or the dynamics of aging.
One cannot understand age distributions of
mental illness without taking into account the
difference between organic and non organic
psychiatric disorders. Organic disorders involve
structural changes in the anatomy of the
brain and include dementia. These disorders
are typically and appropriately not included in
sociological investigations. Non organic diag
noses include depressive disorders, anxiety dis
orders, psychotic disorders, and substance use
disorders (alcohol, illegal drugs, abuse of pre
scribed medications). Most sociological studies
focus on depression; thus, social epidemiology
is primarily the study of the distribution of
depression.
Depression in later life exhibits an epide
miologic paradox. Rates of depressive disorder
(i.e., disorder meeting diagnostic standards)
are lowest among older adults, highest among
young adults, and intermediate among middle
aged adults. But a different pattern is observed
for depressive symptoms, where the oldest old
report higher levels of symptoms than adults
of other ages (Blazer et al. 1991; Mirowsky &
Ross 1989). Definitive evidence about the cause
of this paradox of low diagnoses and high
symptoms is lacking, but most observers believe
that criteria other than the pure number of
symptoms (e.g., persistence over time) exclude
some older adults from qualifying for a diagno
sis of depression.
SOCIAL ANTECEDENTS OF
MENTAL HEALTH AND
SUBJECTIVE WELL BEING

A common, if not consensual, theory of the
social precursors of depression in later life is
emerging in the research literature (George
2004). Loosely speaking, it is a model of stra
tification or social disadvantage and stress.
The general premise is that social disadvan
tage puts individuals on pathways that expose
them to more proximate determinants of
depression and distress. Although applications
of the basic model utilize both cross sectional
and longitudinal data, it is a stage model of
increasingly proximate predictors of psychia
tric disorders in general and depression or
distress in particular. There are five stages in
the model.
The first, most distal stage includes basic
demographic variables (e.g., age, sex, race, or
ethnicity) that represent fundamental aspects
of social location and are in fact bases of
stratification in society. The second stage
includes measures of early events and achieve
ments, most commonly educational attainment
and childhood traumas (e.g., child abuse, sex
ual abuse, parental divorce). The third stage
includes indicators of later achievements, pri
marily SES (occupation, income) and family
characteristics (marital status, fertility history).
As a group, the first three stages provide fairly
extensive information about social status. The
general hypothesis is that disadvantaged status
increases the risk of depression and distress.
The fourth stage of the model includes
indicators of social integration. The most
commonly used indicators measure personal
attachments to social structure, such as orga
nizational and religious participation. More
recently, investigators have examined the
effects of characteristics of the residential
environment, such as measures of disorganiza
tion and transience, poverty levels, and rates
of criminal victimization. Personal attachments
to social structure are expected to decrease the
risk of depression and distress; residence in
disorganized, poor, and unsafe neighborhoods
is expected to increase risk. The fifth stage
of the model includes the most proximate
antecedents of depression and distress and