Nursing Research July/August 2004 Vol 53, No 4 251 Depression Among Southern Rural African American Women A Factor Analysis of the Beck Depression Inventory-II Faye A. Gary ▼ Hossein N.Yarandi  Background: This is the first reported study involving a factor analysis of the Beck Depression Inventory-II, which was administered to a sample of southern rural African American women.  Objective: To determine the factor structure of the Beck Depression Inventory-II using data collected from southern rural African American women.  Methods: Using a correlational, descriptive design, 206 south- ern rural African American women were invited to participate in a face-to-face interview that occurred in a variety of com- munity-based settings.  Results: The factor analysis of the Beck Depression Inventory- II resulted in a two-factor solution. Symptoms such as pes- simism and worthlessness loaded high on the first factor (cognitive). The second factor explained somatic-affective symptoms of depression, with factor loadings high on tired- ness and fatigue and loss of energy.  Conclusions: The application of the Beck Depression Inven- tory-II among African American people would generate needed information about how depressive symptoms may be expressed among them. Knowledge gained from this study promises to be useful for developing appropriate research studies and population-specific treatment approaches for this group of women.  Key Words: African American women  Beck Depression Inventory-II be 10% to 25% for women and 5% to 12% for men, with a point prevalence of approximately 5% to 10% for women and 3% to 6% for men. The estimated economic burden of depression from treatment costs, mortality, and lost productivity exceeds $43 billion a year (Pinus, 2001; Scanlon, 2002). It is important to distinguish between depressive symp- toms and depression itself. Depressive symptoms can be detected by psychometric instruments such as the Beck Depression Inventory-II (BDI-II) (Beck, Steer, & Brown, 1996). Clinical depression, however, implies a formal psy- chiatric diagnosis based on a systematic comparison of the individual’s mental health history and his or her signs and symptoms with predetermined criteria (APA, 1996, 2000). African American women reportedly manifest more depressive symptoms than their White counterparts, but these differences dissipate when socioeconomic status is considered (Beck et al., 1996; Kessler et al., 1994; Kessler et al., 2003). However, African American women are at higher risk for depression, primarily because of their over- representation in the nation’s lower socioeconomic group (Kessler et al., 2003; Tomes, Brown, Semenya, & Simpson, 1990), their chronic medical problems (Jackson-Triche et al., 2003), their limited access to health and mental health services (Gamble, 1997; Gary, Yarandi, & Rivers, 2001; Oppenheimer & Shultz, 1999; Snowden, 2001), and the stigma of mental illness embedded in some African Ameri- can communities (Schreiber, Stern, & Wilson, 1998, 2000; Snowden, 2001). Depression interferes with an individual’s essential functional abilities, thwarts productivity, and can increase the risk for other severe health problems (LaVeist, Bowie, & Cooley-Quille, 2000; Marmer, 2003; Olfson et Faye A. Gary, EdD, MS, RN, FAAN, is Medical Mutual of Ohio Professor of Nursing for Vulnerable and At-Risk Persons, Frances Payne Bolton School of Nursing, Case Western Reserve Univer- sity, Cleveland, Ohio. Hossein N. Yarandi, PhD, is Associate Professor, College of Nurs- ing and Biostatistics Unit, University of Florida, Gainesville. epression manifests as a disturbance in mood, with common symptoms such as persistent sadness or despair, insomnia, decreased appetite, anhedonia, hope- lessness, irritability, low self-esteem, and suicidal ideation (American Psychiatric Association [APA], 1996, 2000). It is the most pervasive psychiatric problem observed in pri- mary care settings throughout the world (Endler, Macrodimitris, & Kocovski, 2000). In the United States, the lifetime prevalence of major depression is estimated to D