Abstract
Aims: Self-reported measures of heath status such as the Short Form-36 (SF-36) are widely used. This dissertation aims to measure how well directly assessed measures of health are reflected in the SF-36 general health and SF-36 mental health scales, as well as the single-item health question.
Methods: This cross-sectional study uses multiple linear regression to model the SF-36 general health, SF-36 mental health scales and the single-item health question as a function of a wide range of directly measured indicators of health status including physical health, mental health and personality measures in a cohort of nine hundred and seventy nine 26 year old New Zealanders.
Results: The SF-36 general health scale is associated as much with mental health measures and personality as it is with measured physical health markers. Physical health measures accounted for approximately 15% of the variability of the general health score. Mental health measures accounted for approximately 10% and personality measures accounted for 14% of the variability of the general health score. The SF-36 mental health scale is not significantly associated with a range of physical health measures, but is associated with psychiatric diagnosis and personality measures. Physical health measures accounted for only 2% of the variability, while mental health measures accounted for 21% and personality measures accounted for 27% of the variability of the SF-36 mental health score. The single-item health question reflected a range of directly assessed measures. Physical health measures accounted for 12%, mental health measures 8% and personality 9% of the variability of the single item health question.
Conclusions: A wide range of clinical factors influence the assessment of self-rated health in this cohort. Of particular note is the significant influence of personality on both general health and mental health measures which suggests an effect on the self rating process itself. The SF-36 general health scale is not primarily a measure of physical health, but is also associated with psychiatric diagnosis and personality. The SF-36 mental health score is primarily a measure of mental health and personality. A range of directly assessed measures are reflected in the single-item health question, but to a lesser degree than the SF-36 general health or mental health scores.