Waist circumference and mortality: impact of associated risk factors

This study was designed to evaluate the risks of all-cause and cardiovascular mortality in subjects with large waist circumferences, with or without associated risk factors, and to determine whether or not waist circumference might identify high-risk subjects.

The population included 55,800 men (aged 52.1 ± 8.2 years) and 28,937 women (aged 54.2 ± 9.1 years) who had undergone a health checkup at the Preventive and Clinical Investigations Centre between January 1999 and December 2004 with a mean follow-up of 4.7 ± 1.7 years. An increased waist circumference was defined as those in the last quintile of distribution. Mortality risk for each waist-circumference quintile, with or without associated risk factors (hypertension, diabetes, elevated LDL cholesterol), was evaluated using Cox's regression models, including age, gender, tobacco and alcohol consumption, and physical activity.

The percentage of subjects with hypertension, diabetes and raised LDL cholesterol levels increased from the first waist-circumference quintile to the last. After adjusting for variables, all-cause mortality risk did not increase significantly with large waist circumference only (HR: 1.19 [0.84-1.68]), but was significantly higher when an increased waist circumference was associated with at least one risk factor (HR=1.58 [1.26-1.98]; 3.70 [2.05-6.68] for three risk factors). Similar results were observed for cardiovascular mortality (HR: 0.85 [0.19-3.68] with only large waist circumference and 3.56 [2.05-6.57] when waist circumference was associated with at least one risk factor).

In a population with low-to-moderate mortality risk, waist circumference alone did not identify high-risk subjects, thus suggesting that a more global approach is necessary.