Epidemiology and Genetics of Obesity in three Sardinian Genetic Isolates

Obesity is a serious health problem that is becoming ever more common. In the US, the NHANES studies have noted a doubling in the prevalence of obesity over the last 20 years (from 15% to 30%), and European data depicts a population becoming ever more similar to the United States: an estimated 135 million Europeans are obese. In Italy, the most recent data reveal that 33.4% of adults are overweight and 9.1% are obese, with higher prevalence in the south (11.4%) than in the north (7.5%). The causes of obesity consist of genetic and environmental components. Previous studies in twins and families have shown that obesity has a large familiarity, even if inside a family it is difficult to ascertain if genetic components influence environment or vice versa. Some of the difficulties in identifying the specific factors that contribute to obesity may be reduced if the study is carried out in a genetic isolate. Ogliastra, a region of eastern Sardinia consisting of 23 small mountain villages with a mean number of inhabitants of about 2000 and which have been isolated for centuries is one such example. The aim of the study was to determine obesity prevalence in the three villages of Talana, Perdasdefogu and Urzulei in the Ogliastra region and to explore its relationship to socio-environmental factors and, with our extensive genealogical information, to assess the genetic contribution to BMI. The study design was a cross-sectional population-based survey. Participants had a blood sample taken, blood pressure, weight and height measured, and completed a standardized questionnaire about living habits, clinical and family history and socio-demographic data. The research conformed to the Declaration of Helsinki, and verbal and written informed consent were obtained from all participants at the examination site. Fieldwork took place over a two year period starting in October 2001. Weight categories of BMI were as defined by the World Health Organization: underweight is BMI<18.5, normal weight is 18.5<=BMI< 25, overweight is 25<= BMI<30, obese is BMI>=30. Descriptive statistics and logistic regression analyses were performed by using STATA 8. Estimates of prevalence were standardized by the direct method to the age and sex distribution of the 2002 adult population of Sardinia. Heritability and genome-wide linkage analyses on BMI were performed by using the SOLAR package. Our sample consisted of 637 people in Talana, 1310 in Perdasdefogu and 866 in Urzulei (age>=18 years). Talana and Perdasdefogu had an average BMI in the range of overweight while Urzulei had an average BMI that fell in the range of normal weight. We estimated an obesity prevalence of 18% in Talana, 15.2% in Perdasdefogu and 11.6% in Urzulei. Overall, we found greater prevalence of obesity in men in Talana and Urzulei, and in women in Perdasdefogu. As with gender, the distribution of obesity differs with age: for men, it increases as age increases to 75 years old in Talana and Perdasdefogu, but only to age 64 in Urzulei; while for women it increases as age increases to 64 years old in all the villages. One of the most important features of obesity is that it is an important risk factor for many other pathologies. This prompted us to examine the comorbidity of the pathologies most frequently associated with obesity and to see if the same global trends were also observed in Ogliastra (Tab.1). To further examine obesity in Ogliastra, we looked at some social and serological parameters that have been found, in some circumstances, to associate with obesity (Tab.2). Multiple logistic regressions of various obesity comorbidities on weight categories adjusting for age, marital status, smoking and education were performed. For all comorbidities, odds of disease is higher in overweight and obese subjects, especially for metabolic syndrome, diabetes and hypertension. As for tumours, odds of disease is higher in underweight subjects, although the difference is not statistically significant. To assess the genetic contribution to variation in BMI, we then estimated the heritability of BMI in the three villages and performed a genome-wide linkage analysis for this trait on Talana extended pedigrees. We estimated a BMI heritability of 36% in Talana, 29% in Perdasdefogu and 30% in Urzulei (p<0.000001). Preliminary results of the genome-wide linkage analysis on BMI in Talana allowed the identification of suggestive loci.