Validity of the SF-36 five-item Mental Health Index for major depression in functionally impaired, community-dwelling elderly patients.J Am Geriatr Soc. 2005 Nov; 53(11):1978-85.JA
To examine criterion and construct validity of the five-item Mental Health Index (MHI-5) of the 36-item Short Form health survey (SF-36) in relation to the presence of major depression in functionally impaired, community-dwelling elderly patients and of eight subsamples defined by cognitive functioning, levels of functional impairment, and proxy report versus self-report.
Nineteen counties in western New York, West Virginia, and Ohio.
One thousand four hundred forty-four functionally impaired, community-dwelling Medicare beneficiaries aged 65 and older who participated in the Medicare Primary and Consumer-Directed Care Demonstration.
MHI-5, Mini-International Neuropsychiatric Interview Major Depressive Episode (MINI-MDE) module.
The MHI-5 demonstrated sufficient criterion validity (area under the receiver operating characteristic curve=0.837; sensitivity=78.7% and specificity=72.1% using a cutpoint of 59/60) with respect to the presence of depression for the entire sample. A significant correlation between MHI-5 scores and presence of major depression as identified using the MINI-MDE (Spearman correlation=-0.426, P<.001), a strong correlation between the MHI-5 and the SF-36 role emotional scale (Spearman correlation=0.522) and a weak correlation with the SF-36 physical functioning scale (Spearman correlation=0.133) provided evidence for construct validity. Additional evidence is provided by decline in mean MHI-5 score as level of formal education and number of close friends and relatives decreased. All eight subsamples demonstrated similar criterion and construct validity. A Cronbach alpha of 0.794 demonstrated internal consistency reliability.
This study provides evidence for adequate criterion and construct validity of the MHI-5 in relation to the presence of major depression among functionally impaired, community-dwelling elderly Medicare patients.