Self-rated health and life prognosis

Arch Med Res. 2003 Jul-Aug;34(4):343-7. doi: 10.1016/S0188-4409(03)00052-3.

Abstract

Background: Effects of self-rated health on survival using prospective surveys were evaluated during the past 20 years mainly in Europe and the US. An overview on this field of research was necessary to know the range of risk ratio (RR) of poor self-rated health.

Methods: The MEDLINE database from 1982 until the end of 2001 was assessed. Self-rated health, mortality, and associated terms were used as key words and information retrieval was executed. After reference papers were broadly collected, 30 papers that included relative risk or odds ratio (OR) to express risk of poor self-rated health on survival were precisely reviewed. RR or OR of poor self-rated health against excellent self-rated health-controlling factors also recognized as relating to survival was calculated in these papers using multiple logistic regression or Cox regression analysis.

Results: Multiple logistic regression analysis was frequently used in the first 10 years and Cox regression analysis was subsequently adopted in the following 10 years. Nearly one half of the study subjects were followed up for 5 years or less, and two thirds of the studies had a target population <5,000 persons. Mortality was largely dependent on age and ranged from 4.6 to 33%. When 54 data with 95% confidence interval (95% CI) were checked, statistical significance was observed in 41 data (75.9%).

Conclusions: Self-rated health is an independent predictor of survival that controls for other related health indicators or covariates. These types of research should be conducted not only in Western countries, but also in other areas including Japan.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Attitude to Health
  • Cohort Studies
  • Female
  • Health Status
  • Health Surveys
  • Health*
  • Humans
  • MEDLINE
  • Male
  • Middle Aged
  • Odds Ratio
  • Prognosis
  • Proportional Hazards Models
  • Prospective Studies
  • Regression Analysis
  • Risk
  • Sex Factors
  • Time Factors