Underweight, overweight and obesity as risk factors for mortality and hospitalization

AIMS: The prevalence of overweight and obesity is increasing in many countries. We aimed to investigate differences in mortality and severe morbidity between underweight people (body mass index (BMI) or =30), and those with normal weights (BMI 18.5 to <25). METHODS: Random samples of the Swedish population aged 16-74 years in 1980-81 and 1988-89 were followed for 12 years with regard to all-cause mortality and mortality from circulatory diseases, all inpatient care, and inpatient care for circulatory and musculoskeletal diseases. Relative risks (RRs) for different levels of BMI were adjusted for age, longstanding illness, smoking, and educational level at baseline. In addition, analyses were made with delayed entry until the fourth-year after interview. RESULTS: Obesity and underweight, but not overweight, was associated with higher all-cause mortality. Among underweight men, the adjusted RR for all-cause mortality was 2.4 (95% confidence interval 1.6-3.6), and among underweight women it was 2.0 (1.5-2.7), but population attributable risks (PARs) were small, at 1.2% and 2.7%, respectively. Overweight was associated with increased risks for inpatient care for circulatory diseases, with PARs being 13.4% among men and 8.1% among women, and musculoskeletal diseases (PARs were 12.7% and 12.9%, respectively). Obese men and women had about 50% higher risks of all-cause mortality than normal-weight people, PARs being 3.2% and 3.8% respectively. CONCLUSIONS: This study supports the findings of other studies, in that overweight seems to be an exaggerated risk factor for all-cause mortality, but is related to other chronic disease. Underweight and obesity generally implies greater increases of RRs, but avoidance of overweight may have greater effect on the population level with regard to reduced cardiovascular and locomotor disease.