Body mass index, change in body mass index, and survival in old and very old persons.

OBJECTIVES: To examine how body mass index (BMI) and change in BMI are associated with mortality in old (70-79) and very old (≥80) individuals. DESIGN: Pooled data from three multidisciplinary prospective population-based studies: OCTO-twin, Gender, and NONA. SETTING: Sweden. PARTICIPANTS: Eight hundred eighty-two individuals aged 70 to 95. MEASUREMENTS: BMI was calculated from measured height and weight as kg/m(2) . Information about survival status and time of death was obtained from the Swedish Civil Registration System. RESULTS: Mortality hazard was 20% lower for the overweight group than the normal-underweight group (relative risk (RR) = 0.80, P = .011), and the mortality hazard for the obese group did not differ significantly from that of the normal-underweight group (RR = 0.93, P = .603), independent of age, education, and multimorbidity. Furthermore, mortality hazard was 65% higher for the BMI loss group than for the BMI stable group (RR = 1.65, P < .001) and 53% higher for the BMI gain group than for the BMI stable group (RR = 1.53, P = .001). Age moderated the BMI change differences. That is, the higher mortality risks associated with BMI loss and gain were less severe in very old age. CONCLUSION: Old persons who were overweight had a lower mortality risk than old persons who were of normal weight, even after controlling for weight change and multimorbidity. Persons who increased or decreased in BMI had a greater mortality risk than those who had a stable BMI, particularly those aged 70 to 79. This study lends further support to the belief that the World Health Organization guidelines for BMI are overly restrictive in old age.