Correlation of body mass index with thoracic and abdominal panniculus

Obesity can play a significant role in chronic diseases, sudden unexpected death, and morbid obesity may be important as a cause of death for forensic pathologists. Our study attempted to determine if there is a correlation between panniculus measurements and body mass index (BMI) since BMI has been used in most studies to categorize obesity. Using data obtained from a review of 524 adult autopsies conducted at the University of Michigan from 1990 to 1992 we were able to show a correlation between both thoracic and abdominal panniculus and BMI (r2 = 0.335 and 0.296 respectively) which is statistically significant (p = 10(-47) and 10(-41) respectively). A prospective study confirmed the correlation (r2 = 0.552 for thoracic and 0.436 for abdominal panniculus) when the measurements were taken at the xyphoid process and 3 cm below the umbilicus. Using these data we calculated a panniculus index (PI) which is equal to the thoracic + abdominal panniculus in centimeters divided by the square of the height (in meters). The PI strongly correlated with BMI and was able to predict obesity. Using a BMI cutoff of 39 for morbid obesity, a PI value of 4.07 for females and 3.25 for males predicted morbid obesity with the probability of a false positive less than or equal to 2.5%. Mild and severe obesity could also be determined using the PI. Based on these data we've concluded that a concise mathematical relationship does exist between BMI and panniculus measurements. Therefore panniculus measurements can be used either as a surrogate measurement of morbid obesity or to support BMI calculations.