Depression was studied in a community sample of 962 males and 1555 females aged 55 years and over living in Kentucky in 1981. The sample was representative of the population in Kentucky in that age group and quite similar to that US population. The Center for Epidemiologic Studies Depression Scale was used as a measure of depression, and 13.7% of the males and 18.2% of the females were at or above a previously established cutpoint of 20 for adults over age 55 years. Significant relationships to depression were found in both sexes for age, education, income, housing quality, marital status, and health. For females, the age-depression relationship was not linear. By far the strongest relationship was with self-reported physical health. Significant proportions of those with self-reported kidney or bladder disease, heart trouble, lung trouble, hardening of the arteries, and stroke were above the depression cutpoint. For those conditions, physicians could expect high levels of concomitant depression in about one fourth of males and at least one third of females. These levels of depression were not found for those with high blood pressure, stomach ulcers, cancer, or diabetes. Over half of the sample reported taking prescribed medication and over half had needed a physician's care in the previous six months. Only 3.9% of the males and 3.2% of the females admitted to needing help for mental health problems. Thus, older adults with depression would probably be more likely to seek help from physicians than from services or professionals with explicit mental health labels.

Five scales were assessed as mental health measures for older persons: Affect Balance, The Center for Epidemiological Studies Depression Scale, General Well-Being, LSI-Z Life Satisfaction, and Trait-Anxiety Inventory. These scales were administered to a community sample of 279 older persons and a clinical sample of 109 older persons who were in psychiatric inpatient units. In both samples, the internal consistency reliabilities for the anxiety, depression, and well-being scales were moderately high to high, for the life satisfaction scale they were acceptable, but the reliabilities for the affect balance scale suggest some caution in its use. For validity, multivariate analyses of variance found that all scales significantly discriminated between the two samples. The well-being and then depression scales were the strongest discriminators while the life satisfaction scale had the weakest validity. Cutting points for the well-being and depression scales are suggested for estimating the proportions of older persons who would be probable at-risk for disorder that requires intervention.

As participants in a panel study, 234 older adults were interviewed before, as well as after, serious flooding occurred in southeastern Kentucky. Floods are not uncommon in this area, but these were more widespread than most, and resulted in both previously exposed and newly exposed subsamples of disaster victims. Flood impact was measured at both personal and community levels. With preflood symptoms controlled, there were modest flood effects on both trait anxiety and weather-specific distress in older adults without prior flood experience, but no flood effects in older adults who had been in floods before. Thus, the study provides support for the "inoculation hypothesis" and other conceptualizations that emphasize the advantage of being familiar or experienced with a stressor that is at hand. An implication is that "experienced" victims could be a valuable resource in prevention efforts.

We examined the additive and interactive roles of six sociodemographic factors, three resources and three categories of life events, in the development of depressive symptoms. A probability sample of 1,233 persons, 55 years of age and older, were interviewed twice in their homes, with a 6-month interval between Times 1 and 2. All of the subjects were below a Center for Epidemiologic Studies Depression Scale (CES-D) cut point of 16 at the first interview. An onset group (n = 66) was identified that had increases in depression to above a CES-D cut point of 20 at Time 2. A hierarchical discriminant analysis found significant effects for the following factors after initial symptom levels were statistically controlled: physical health, social support, the Social Support X Health interaction, loss events, and the interaction of social support and loss events. Thus, health and social support played both additive and interactive roles, life events had weak effects, and sociodemographic factors did not contribute to depressive onset. Depression appears to be the most prevalent functional psychiatric disorder of older adults. Recent community studies have reported that between 10% and 18% of the elderly have clinically important levels of depressive symptoms (Ban, 1978;

No abstract

Because they were participating in a concurrent panel study, 222 older adults had been asked how much help they would expect to receive in a hypothetical emergency before experiencing two separate floods. For the subsample suffering losses or injuries during the floods, Study 1 examined the accuracy of their expectations, as well as possible changes in them, as a result of help actually received. Generally, victims received much less help than they had expected to receive prior to the floods. Preflood expectation of support predicted help from kin, whereas loss and education predicted help from nonkin sources. However, subsequent expectations did not change as a result of the level of help received. Using the total sample, Study 2 examined the broader issue of whether the disaster itself affected subsequent perceptions of support. Flood exposure, as measured at both individual and community levels, was associated with declines in perceptions of support and social participation.

As part of a larger panel study interviews were obtained from 3 samples of older adults: 45 persons who had recently lost a spouse, 40 who had lost a parent or child, and 45 who were not bereaved. Assessments were conducted before and after the deaths. In the widowed sample, health remained quite stable, but depression increased sharply, then remained elevated. Changes were minimal in the sample who had lost a parent or child and in the nonbereaved sample. Multiple regression procedures were used to identify factors that contribute to depression and health 9 months after the spouse's death. Postbereavement depression was associated with higher prebereavement depression, higher financial pressures, higher global stress, fewer new interests, and lower social support. Health was a function of prebereavement health, new interests, financial pressures, and global stress. In general, life events and resources had stronger effects in the widowed sample than in the comparison samples.

Education appears to confer a lifelong advantage for healthy aging. Part of this advantage is accounted for by the relationship between education and trait negative affect. Higher educational attainment is related to lower levels of trait negative affect; lower negative affect results in better health and life satisfaction.