Volume 54 | Number 3 | June 2019

Abstract List

Vanessa Panaite PhD, Nicholas W. Bowersox PhD, ABPP, Kara Zivin, Dara Ganoczy M.P.H., Hyungjin Myra Kim Sc.D., Paul N. Pfeiffer


Assess whether neighborhood characteristics predict patient‐reported outcomes for depression.

Data Sources

VA electronic medical record data and U.S. census data.

Study Design

Retrospective longitudinal cohort.

Data Extraction Methods

Neighborhood and individual characteristics of patients (N = 4,269) with a unipolar depressive disorder diagnosis and an initial Patient Health Questionnaire (PHQ‐9) score ≥10 were used to predict 50 percent improvement in 4‐8‐month PHQ‐9 scores.

Principal Findings

The proportion of a patient's neighborhood living in poverty (OR = 0.98; 95% CI: 0.97‐.1.00; 0.03) was associated with lower likelihood of depression symptom improvement in addition to whether the patient was black (OR = 0.76; 95% CI:0.61‐0.96; =0.02) had PTSD (OR = 0.59; 95% CI:0.50‐0.69; <0.001) or had any service‐connected disability (OR = 0.73; 95% CI:0.61‐0.87; <0.001).


Neighborhood poverty should be considered along with patient characteristics when determining likelihood of depression improvement.