ORIGINAL ARTICLE doi: 10.1111/j.1463-1326.2008.00895.x PROCEED: Prospective Obesity Cohort of Economic Evaluation and Determinants: baseline health and healthcare utilization of the US sample* A. M. Wolf, 1 N. Finer, 2 A. A. Allshouse, 3 K. B. Pendergast, 3 B. H. Sherrill, 3 I. Caterson, 4 J. O. Hill, 5 L. J. Aronne, 6 H. Hauner, 7 C. Radigue, 8 C. Amand 9 and J.-P. Despres 10 1 University of Virginia School of Medicine, Virginia, USA 2 University of Cambridge, Cambridge, UK 3 RTI Health Solutions, North Carolina, USA 4 University of Sydney, Sydney, Australia 5 University of Colorado School of Medicine, Colorado, USA 6 Columbia University College of Physicians and Surgeons, New York, USA 7 Technical University of Munich, Munchen, Germany 8 Sanofi-Aventis Recherche et De ´veloppement, Paris, France 9 Keyrus Biopharma, Cedex, France 10 Hospital Laval, Quebec, Canada Aim: To summarize baseline characteristics, health conditions, resource utilization and resource cost for the US population for the 90-day period preceding enrolment, stratified by body mass index (BMI) and the presence of abdominal obesity (AO). Methods: PROCEED (Prospective Obesity Cohort of Economic Evaluation and Determinants) is a multinational, prospective cohort of control (BMI 20–24.0 kg/m 2 ), overweight (BMI 25–29.9 kg/m 2 ) and obese (BMI  30 kg/m 2 ) subjects with AO and without AO [non-abdominal obesity (NAO)], defined by waist circumference (WC) >102 and 88 cm for males and females, respectively. Subjects were recruited from an Internet consumer panel. Outcomes were self-reported online. Self-reported anthropometric data were validated. Prevalence of conditions and utiliza- tion is presented by BMI class and AO within BMI class. Differences in prevalence and means were evaluated. Results: A total of 1067 overweight [n ¼ 474 (NAO: n ¼ 254 and AO: n ¼ 220)] and obese [n ¼ 493 (NAO: n ¼ 39 and AO: n ¼ 454)] subjects and 100 controls were recruited. Self-reported weight (r ¼ 0.92) and WC (r ¼ 0.87) were correlated with measured assessments. Prevalence of symptoms was significantly higher in groups with higher BMI, as were hypertension (p < 0.0001), diabetes (p < 0.0001) and sleep apnoea (p < 0.0001). Metabolic risk factors increased with the BMI class. Among the overweight class, subjects with AO had significantly more reported respiratory, heart, nervous, skin and reproductive system symptoms. Overweight subjects with AO reported a significantly higher prevalence of diabetes (13%) compared with overweight subjects with NAO (7%, p ¼ 0.04). Mean healthcare cost was significantly higher in the higher BMI classes [control ($456  937) vs. overweight ($1084  3531) and obese ($1186  2808) (p < 0.0001)]. Conclusion: An increasing gradient of symptoms, medical conditions, metabolic risk factors and healthcare utilization among those with a greater degree of obesity was observed. The independent effect of AO on health and healthcare utilization deserves further study with a larger sample size. Keywords: abdominal obesity, body mass index, cost analysis, economics, population studies Received 11 March 2008; accepted 18 March 2008 Correspondence: Anne Wolf, MS, RD, 5030 Rutherford Road, Charlottesville, VA 22901, USA. E-mail: amw6n@virginia.edu *This study was conducted by RTI Health Solutions and Sanofi-Aventis Recherche ´ et De ´veloppement and funded by Sanofi- Aventis Recherche et De ´veloppement. 1248 j Diabetes, Obesity and Metabolism, 10, 2008, 1248–1260 # 2008 The Authors Journal Compilation # 2008 Blackwell Publishing Ltd