[Use of psychoactive substances and risky behavior].

The article by Bastos et al 1 (2008), published in the Re-vista de Saúde Pública, forms an important contribution towards studies on the use of psychoactive substances and their relationship with behavior leading to risks to sexual and reproductive health. As the authors emphasize in their introduction, there are many health problems relating to the use of alcohol and other substances, which makes it extremely relevant to study them on a population basis in Brazil. The aim of this letter is to highlight certain points relating to alcohol use that, in our opinion, were insuffi ciently discussed by these authors. The fi rst of these points relates to the notion of regular use. The authors established that the consumption of alcoholic drinks more than four times a week constituted regular use, through the question: " Have you, at any time of life, started to drink regularly (more than four times a week)? ". Although " frequency " is a signifi cant marker, recent studies have emphasized the importance of also assessing the quantity of alcohol consumed on each occasion, given the relationship between quantity and health problems. 6 In fact, frequent use is not necessarily risky use, and it may even be associated with protective factors. A study conducted among men whose lifestyle was considered healthy (nonsmokers with adequate diets, body mass index less than 25 and daily practice of physical activity) found decreased risk of myocardial infarct among those who consumed between 5.0 and 29.9g of alcohol per day, in comparison with those who abstained from all alcohol. 4 Although the defi nitions vary, risky use (" binge drinking ") is considered to consist of the intake of four or more drinks per occasion for women and fi ve or more for men, within a short space of time. 3 Laranjeira et al, 5 from a national sample, estimated that the prevalence of risky use in Brazil was 14% among men and 3% among women. In the study by Bastos et al, 1 among those who consumed alcohol frequently, there are subjects who consumed large quantities alongside others who consumed small quantities per occasion, which may have affected the analyses. Taking condom use as an example, it can be seen that the authors did not fi nd any association between " not using condoms " and making frequent use of alcohol. This probably occurred because subjects with very different …