Journal of General Internal Medicine

, Volume 32, Issue 4, pp 416–422

The Association Between Neighborhood Environment and Mortality: Results from a National Study of Veterans

  • Karin Nelson
  • Greg Schwartz
  • Susan Hernandez
  • Joseph Simonetti
  • Idamay Curtis
  • Stephan D. Fihn
Original Research

Abstract

Background

As the largest integrated US health system, the Veterans Health Administration (VHA) provides unique national data to expand knowledge about the association between neighborhood socioeconomic status (NSES) and health. Although living in areas of lower NSES has been associated with higher mortality, previous studies have been limited to higher-income, less diverse populations than those who receive VHA care.

Objective

To describe the association between NSES and all-cause mortality in a national sample of veterans enrolled in VHA primary care.

Design

One-year observational cohort of veterans who were alive on December 31, 2011. Data on individual veterans (vital status, and clinical and demographic characteristics) were abstracted from the VHA Corporate Data Warehouse. Census tract information was obtained from the US Census Bureau American Community Survey. Logistic regression was used to model the association between NSES deciles and all-cause mortality during 2012, adjusting for individual-level income and demographics, and accounting for spatial autocorrelation.

Participants

Veterans who had vital status, demographic, and NSES data, and who were both assigned a primary care physician and alive on December 31, 2011 (n = 4,814,631).

Main Measures

Census tracts were used as proxies for neighborhoods. A summary score based on census tract data characterized NSES. Veteran addresses were geocoded and linked to census tract NSES scores. Census tracts were divided into NSES deciles.

Key Results

In adjusted analysis, veterans living in the lowest-decile NSES tract were 10 % (OR 1.10, 95 % CI 1.07, 1.14) more likely to die than those living in the highest-decile NSES tract.

Conclusions

Lower neighborhood SES is associated with all-cause mortality among veterans after adjusting for individual-level socioeconomic characteristics. NSES should be considered in risk adjustment models for veteran mortality, and may need to be incorporated into strategies aimed at improving veteran health.

KEY WORDS

risk adjustment veteran socioeconomic factors public health clinical epidemiology 

Supplementary material

11606_2016_3905_MOESM1_ESM.docx (19 kb)
ESM 1(DOCX 19 kb)
11606_2016_3905_MOESM2_ESM.docx (21 kb)
ESM 2(DOCX 20 kb)
11606_2016_3905_MOESM3_ESM.docx (20 kb)
ESM 3(DOCX 19 kb)
11606_2016_3905_MOESM4_ESM.docx (21 kb)
ESM 4(DOCX 20 kb)
11606_2016_3905_MOESM5_ESM.docx (17 kb)
ESM 5(DOCX 17 kb)

Copyright information

© Society of General Internal Medicine 2016

Authors and Affiliations

  • Karin Nelson
    • 1
    • 2
    • 3
    • 4
  • Greg Schwartz
    • 5
  • Susan Hernandez
    • 4
  • Joseph Simonetti
    • 1
    • 3
  • Idamay Curtis
    • 5
  • Stephan D. Fihn
    • 2
    • 3
    • 4
    • 5
  1. 1.VA Puget Sound Healthcare System, Health Services Research and Development Seattle-Denver COINSeattleUSA
  2. 2.VA Puget Sound Healthcare System, General Internal Medicine ServiceSeattleUSA
  3. 3.School of Medicine, Department of MedicineUniversity of WashingtonSeattleUSA
  4. 4.School of Public Health, Department of Health ServicesUniversity of WashingtonSeattleUSA
  5. 5.VHA Office of Analytics and Business IntelligenceSeattleUSA

Personalised recommendations