Determinants of oral cancer at the national level: just a question of smoking and alcohol drinking prevalence?

In addition to individual-based prevention strategies, the burden of oral cancer could be decreased by controlling its national level determinants. Population-based studies have found smoking, drinking, and wealth to be associated with oral cancer incidence and mortality rates. However, these studies merely reported trends, or did not account for confounders or for intercorrelation between predictor variables. This ecologic study sought to investigate oral cancer determinants at the country level. The male, age-standardized mortality rate was the dependent variable. The explanatory variables, obtained from reliable international agencies, were life expectancy, frequency of physicians, gross national product (GNP), expenditure on health, literacy rate, human immunodeficiency virus (HIV) prevalence, smoking prevalence, alcohol drinking prevalence, drinking modality, average daily calorie consumption, and average calorie intake from fruit and vegetables. Common factor analysis was used to generate a new dimension that incorporated all of the strongly intercorrelated variables. These were life expectancy, physician frequency, GNP, expenditure on health, literacy rate, calorie consumption, smoking prevalence, and drinking modality. According to this dimension, arbitrarily called the country development level (CDL), countries were split into quartiles. The ecologic risk for high mortality from oral cancer, estimated using logistic regression analysis, was three to five times higher among the second, third, and fourth CDL quartiles than among the first CDL quartile, which included the highest-income countries. HIV, drinking prevalence, and fruit and vegetable intake did not affect significantly mortality. These results suggest that it might be possible to improve oral cancer mortality by modifying country-based determinants related to aberrant lifestyles (not only smoking and drinking prevalence) and improving healthcare system efficiency, approximately estimated by CDL, as well as general socioeconomic and cultural conditions.