Year of Publication

2016

College

Public Health

Date Available

6-8-2016

Degree Name

Master of Public Health (M.P.H.)

Committee Member

Glyn Caldwell, MD

Advisor

Steven Browning, PhD, MSPH

Co-Director of Graduate Studies

Lorie Chesnut, DrPH, MPH

Abstract

Depressive disorders are characterized as sharing affected mood, and somatic (e.g. sleep pattern, appetite, and unintentional changes in weight) and cognitive alterations from previous normal daily functioning that are clinically significant. Anxiety disorders are characterized by excessive fear, general or specific anxiety, and related behavioral disturbances that are also clinically significant. This project attempts to identify risk factors that may predict what groups are most likely to be affected by depression and/or anxiety.

Sponsored by the CDC, the National Health Interview Survey collects general health information for non-institutionalized individuals living in the US. Variables for use in logistic regression model construction were selected from the all-inclusive pool of data measurements taken during the 2014 NHIS person data module, using the literature as a guide. The sample was limited to adult respondents at least 18 years of age.

Odds ratios were calculated for each level of potential predictor variables comparing the two levels of the dependent variable: those with a self-reported depression, anxiety, or emotional problem (DAE) related functional limitation and those without. These were used as initial inclusion criteria for logistic regression modeling, which resulted in two final models. Each model contained eight variables accounting for age, marital status, education level, financial factors, and other functional limitations. The nine predictor model also included sex as a predictor (c=0.763), while the eight predictor did not (c=0.790). Different effects were observed in each model.

These models were designed for use as a tool for selecting groups for targeted public health intervention, not for use in a clinical setting, and could likely help distinguish groups that would be prime for investigating the prevalence and effects of DAE-related functional limitations. There were several limitations worth addressing in future research into this topic.

Included in

Public Health Commons

Share

COinS