Depressive Symptom Severity and Correlates among Meals on Wheels Recipients

Background and Purpose: Despite the high vulnerability to depression among homebound older adults, routine depression assessment using a validated screening tool is rarely done by social service providers serving the homebound. Following a brief description of an adoption of the Patient Health Questionnaire -9 (PHQ-9) as its depression screening tool by a large urban Meals on Wheels (MOW) program, this study reports the assessment outcomes with respect to (1) the rates, severity, and correlates of depressive symptoms and differences by age group, gender, race/ethnicity, and cognitive health status; and (2) the relationship between depressive symptom severity and a diagnosis of depression and antidepressant medication use.

Methods: The sample consisted of 762 MOW clients for whom PHQ-9 assessments were done (Cronbach's alpha = .914) by the agency case managers at the time of comprehensive in-home assessments during the 8-month study period. Disabled adults younger than age 60 were included along with older adults aged 60 and older (71.4%) to examine age difference. Bivariate analysis, with chi-square statistics, was also done to examine difference in the rates and severity of depressive symptoms by gender, race/ethnicity, and cognitive status. Negative binomial regression analysis was used to determine the correlates of depression symptom severity.

Results: Of the sample, 29.4% scored 5+ on the PHQ-9, 17.3% had clinically significant depressive symptoms (PHQ-9 >= 10), and 8.4%, or a little less than half of those with clinically significant symptoms, had probable major depressive disorder. Significantly higher proportions of those who were under age 60, women, or cognitively impaired were found to have clinically significant depressive symptoms and probable major depressive disorder. No significant relationship between PHQ-9 scores and race/ethnicity was found in bivariate analysis. The results of the negative binomial regression analysis confirmed the significance of age, gender, and cognitive health status. In addition, the number of chronic medical conditions and the nutritional risk score were positively associated with the severity of depression symptoms; however, being African American, as opposed to non-Hispanic white, and the level of income were both negatively associated. Self-reported diagnosis of depression and antidepressant medication use were not significant correlates of the PHQ-9 score.

Conclusions and Implications: Routine depression screening of homebound individuals by their social service providers is likely to facilitate improved detection of depression and access to treatment. This study showed the feasibility and utility of using PHQ-9 as a tool to screen for depression of homebound adults by social service providers. The assessment outcomes revealed a high rate of depression and suggested that psychotherapy, along with antidepressant medication, may be necessary for some homebound adults for whom antidepressant medication alone may not work. An absolute majority of the subjects were low income (below 150% of the poverty line) and were likely to have multiple life stressors, such as financial difficulty, feelings of loneliness, and dependence on others for care, that make them feel depressed. In the face of these multiple life stressors, antidepressant medication alone may not be the most effective means of controlling depressive symptoms.