Skip to main content
. Author manuscript; available in PMC: 2015 Oct 6.
Published in final edited form as: Psychiatr Serv. 2011 Jul;62(7):720–726. doi: 10.1176/appi.ps.62.7.720

Table 3.

Predictors of antidepressant overuse among respondents to the 2005 Medical Expenditure Panel Survey (MEPS)a

Characteristic Adjusted ORb P
Age ≥65 (reference: <65)   .95 .03
Female (reference: male)   .79 .08
Race-ethnicity (reference: white)
 Hispanic 1.03 .88
 African American 1.60 .07
 Asian 2.00 .12
Education attainment (reference: postgraduate degree)
 Less than high school degree 1.14 .65
 High school graduate 1.13 .50
 College graduate   .26 .34
 Postgraduate degree   .40 .25
Income group (reference: high income)
 Poor or near poor   .88 .66
 Low income   .80 .39
 Middle income   .97 .90
 High income   .81 .35
Insurance coverage (reference: uninsured)
 Private insurance   .81 .22
 Medicare   .94 .80
 Medicaid   .89 .54
Fair or poor physical health (reference: average to good)c 1.18 .32
Fair or poor mental health (reference: average to good)d   .80 .02
a

Weighted N=23,026,608

b

Constructed with standard errors adjusted for clustering at the individual level and sample MEPS probability weight

c

Measured by the physical health index of the 12-Item Short Form Health Survey

d

Measured by the Kessler 6-Item Psychological Distress Scale