Obesity produces a high morbidity and mortality. The prevalence of obesity in America increased by 50 percent in the last 15 years, and similar dramatic increases were noted in most countries of the world. The etiology of the worldwide increase in obesity is not clear. Five different viruses have been reported to cause obesity in animal models, but a role for viral infections in human obesity has been considered only recently. Human adenovirus-36 (AD-36) produces a syndrome of increased visceral and total body fat deposition, with paradoxically decreased serum cholesterol (CHOL) and triglycerides (TG), in chickens, mice, and monkeys. Screening for Ad-36 antibodies in serum in 3 US cities showed 28 percent of obese humans, but only 5 percent of lean humans, were antibody positive (AB+). Serum CHOL and TG of AB+ obese humans were significantly lower than in antibody negative (AB-) people. Twin pairs (n=26) discordant for Ad-36 antibodies were identified. AB- twins had a higher body mass index (BMI) and percent body fat. This application will evaluate twin registries in 3 sites: New York, NY; Richmond, VA, and Brisbane, Australia.
The specific aims are: 1) To obtain anthropometric data and to perform assays for Ad-36 antibodies on stored serum samples of these twins, 2) To identify twin pairs discordant for Ad-36 antibodies and determine if they differ in anthropometric and other variables, 3) To recall discordant twin pairs for blood samples, history, and physical exam to determine if the variables change over time, 4) To determine the """"""""specificity"""""""" of the Ad-36 effects by assessing effects of infection with """"""""control"""""""" adenoviruses, Ad-2 and Ad-9. Hypotheses: 1) Ad-36 antibody positive (AB+) co-twins will have a higher initial BMI, percent body fat, and waist circumference, and serum CHOL and TG, compared to their antibody negative (AB-) co-twins, 2) These differences will increase over time, documented at the repeat evaluation, 3) Co-twins discordant for antibodies to Ad-2 and Ad-9 will not differ in the above variables. Data and stored serum from 2000 twin pairs will be analyzed initially. Pilot data suggest about 20 percent of twin pairs will be discordant for Ad-36 antibodies, so about 400 twin pairs will be evaluated for changes over time. Humans cannot be deliberately infected with Ad-36, so the study design of evaluating body weight and serum lipids in twins and discordant twins over time will provide the strongest available evidence that Ad-36 produces obesity in humans. Demonstration that a virus produces obesity and alters serum lipids in humans has worldwide implications for the diagnosis, prevention, and treatment of obesity and lipid disorders. The potential impact on worldwide public health may be enormous.
