||All-population and area-based investigations of diet in central obesity are scarce. We used cross-sectional data from 50 339 individuals who responded to the HUNT3 survey of 2006-2008, which recruited from all county-residing adults 20 years and older, to investigate whether those with central obesity eat and drink differently than others. Answers to dietary questions were recoded and analyzed with multiple linear regression, using waist/hip ratio (WHR), age, and sex as explanatory variables. Frequencies of consumption or amounts of food, beverages, and meals were compared among WHR quartiles. Central obesity was present in the quartile with the highest WHR, WHR4 (WHR >/= 0.917 for women and 0.981 for men) but not in the quartile with the lowest WHR, WHR1 (WHR < 0.817 for women and 0.895 for men). Dietary variables differed markedly by age and sex. After adjustment for these factors and for multiple testing, we found significant differences between WHR4 and WHR1 for 19 of 30 dietary variables. Central obesity was associated with a lower intake of any bread, and of whole-grain bread in particular. Intake of fruits and berries, vegetables, and pasta and rice was less, and intake of sausages and hamburgers and boiled potatoes was more frequent. Intake of alcohol, tea, and fruit juice was lower in those with central obesity, whereas intake of sugar-free soft drinks and coffee was higher. The frequency of breakfast and lunch was lower and of nightly meals was higher in those with central obesity. In conclusion, in this large area-based population, central obesity was associated with differences in dietary habits, some of which (such as decreased consumption of whole-grain bread and increased intake of sugar-free drinks) are of possible clinical significance.