This study quantifies the risk of hospitalisation relating to overweight and obesity. Methods 241 949 45 and Up Study participants with linked hospital admissions and death data were followed from recruitment (February 2006 onwards) through to June 2009. Self-reported height and weight were used to classify patients into BMI categories, using cut-points of 15, 18.5, 20, 22.5, 25, 27.5, 30 and 35 kg/m 2 . Rates of incident hospitalisation by BMI were compared using Cox regression, adjusting for

more »

... range of confounders. Results Preliminary results on the first 103 040 participants show incident hospitalisation rates to be 299 (95% CI 294 to 304) per 1000 person years for males and 248 (95% CI 243 to 252) for females. Compared with those of healthy-weight (BMI 20e22.5 kg/m 2 ), rates in those with severe obesity (BMI 35e50 kg/m 2 ) were higher among males (HR: 1.36, 95% CI 1.21 to 1.54) and females (HR: 1.52; 95% CI 1.38 to 1.67). There were clear gradients as weight increased from healthy to higher BMI, more so among females than males, and in people aged 45e64 compared to those aged 65e84, with no evidence of increasing risk of hospitalisation with increasing BMI in those aged over 85. Conclusions Given the excess risk of hospitalisation among overweight and obese individuals, the burden to the health system attributable to overweight and obesity is likely to be substantial, particularly among middle-age adults. High BMI was not significantly related to hospitalisation in the elderly.