O besity and abdominal adiposity have been shown in prospective studies to be risk factors for cardiovascular disease and particularly for diabetes (1-8). In cross-sectional studies, both are related with risk factors for these diseases (9 -12), but there are few publications on the effects of changes in abdominal adiposity (13). We characterized men and women who gained and lost abdominal adiposity over 9 years and describe the incidence and the improvement in cardiometabolic risk factors according to changes in waist circumference.
RESEARCH DESIGN AND METHODS -From the Data from anEpidemiological Study on the Insulin Resistance Syndrome (DESIR) cohort (9,14,15), 1,868 men and 1,939 women aged 30 -64 years at baseline were followed over 9 years. The 73% of the baseline cohort that we studied were older, less frequently abdominally obese, hypertriglyceridemic, hyperinsulinemic, smokers, and fewer had metabolic syndrome. Cardiometabolic abnormalities and the metabolic syndrome were defined according to the National Cholesterol Education Program (NCEP) criteria (16), except high blood pressure, which included antihypertensive treatment (Table 1); hyperinsulinemia was defined by upper quartiles of fasting insulin Ն57.3 pmol/l for men and Ն52.8 pmol/l. The incidence and improvement of cardiometabolic risk factors were studied by age-adjusted logistic regression, according to waist change: Յ Ϫ3.0 cm, Ϫ2.9 to ϩ2.9 cm, 3.0 -6.9 cm, and Ն7.0 cm. Statistical significance was defined as P Ͻ 0.05.RESULTS -The median increase in waist circumference was 3 cm in men and 4 cm in women; 25 and 34% of men and women, respectively, increased their waist by Ն7 cm, 14% decreased their waist by Ն3 cm, and 29% remained stable (Ϯ2.9 cm).Men whose waist decreased were older and had a larger waist circumference and BMI at baseline. Age was not significantly related to waist change in women; however, women who became slimmer had a larger baseline waist but similar BMI. Men who decreased alcohol intake reduced their waist circumference. Stopping smoking was associated with an increase in waist circumference, but smoking at baseline was associated with a gain in waist circumference in only men. Baseline physical activity did not influence waist change, but an increase was associated with a decreasing waist circumference.The incidence of abdominal obesity was 10% in men and 15% in women (Table 1). Of all risk factors, high blood pressure had the highest incidence (48 and 30%, respectively), and the incidences of the metabolic syndrome were 8% for men and 7% for women. The metabolic syndrome and all cardiometabolic factors showed significant trends that became worse with an increasing waist (with one exception, LDL cholesterol in women). The odds ratios (95% CI) for an incident metabolic syndrome were 7.9 (4.4 -13.9) in men and 4.7 (2.7-8.0) in women who increased their waist by Ն7 cm, compared with a stable waist circumference. Results were not changed after adjusting for baseline waist circumference or BMI. Adjusting for 9-year BMI change, ...