Source: Mercy JA, Kresnow M, O’Carrol PW,et al. Is suicide contagious? A study of the relation between exposure to the suicidal behavior of others and nearly lethal suicide attempts. Am J Epidemiol. 2001;154:120–127.The authors conducted a population-based case-control study in Houston to examine the association between adolescents’ nearly lethal suicide attempts and exposure to suicidal behavior of parents, relatives, friends, or acquaintances, and to accounts of suicide in the media. They interviewed 153 victims, 83 male, aged 13–34 who had been treated at 3 area hospital emergency departments between 1992 and 1995 for nearly lethal suicide attempts, and 513 control subjects identified by random digit dialing. Nearly lethal events were defined as those in which the attempter probably would have died without emergency medical or surgical intervention or those in which the attempter sustained any injury from a method with a high case fatality ratio (ie, gun or noose). Interviews were successfully completed with 63% of eligible cases and 60% of eligible controls. After controlling for possible confounding variables, nearly lethal suicide attempts were not significantly associated with suicidal behavior of a parent (OR=1.5; 95% CI, .6–3.6, P=.42) or a non-parent relative (OR=1.2; 95% CI, .7–2.0, P=.55). Suicidal behavior of a friend or acquaintance (OR=.6; 95% CI, .4–1.0, P=.05) and exposure to media accounts of suicide (OR=.2; 95% CI, .1–.3; P=.00) were both associated with a lower risk of nearly lethal suicide attempts. The authors conclude that exposure to suicide in the media and, to a lesser extent, exposure to suicide behavior of friends or acquaintances may protect against nearly lethal suicide attempts.Source: Takemura Y, Sakurai Y, Honjo S, et al. Relation between breastfeeding and the prevalence of asthma: The Tokorozawa Childhood Asthma and Pollinosis Study. Am J Epidemiol. 2001;154:115–119.These investigators at Japan’s National Defense Medical College studied the relationship between breastfeeding and the prevalence of asthma among children. They enrolled all public elementary and junior high school students in Tokorozawa, Japan (ages 6–15 years). Subjects included 2,315 students with asthma and 21,513 controls. Parents completed a standardized questionnaire. After adjustment for age, gender, parental smoking, and parental history of asthma, a higher prevalence of asthma was noted among children who had been breastfed (OR=1.198; 95% CI, 1.054–1.363; P<.01). The authors suggest that breastfeeding may be a risk factor for asthma in school-age children.Conventional wisdom tells us that suicide attempts by schoolmates and media exposure to suicide events are important risk factors for suicide in psychologically vulnerable adolescents and young adults.1 Mercy et al show the opposite: exposure to suicide behavior by acquaintances and media accounts of suicide appear to have a protective effect against nearly lethal suicide attempts. Many pediatricians also believe that breastfeeding protects against asthma.2,3 The study by Takemura et al suggests that, on the contrary, breastfeeding may be a risk factor for asthma. What gives?In the suicide study, the authors point out that conventional wisdom is derived from reports of clusters of suicide attempts in communities rather than isolated cases. They also suggest that recent media portrayals of suicide are more realistic and less glamorous than in the past. However, it is also important to note that this study examined near-lethal suicide attempts, which may differ in important ways from clusters of so-called “suicide gestures.” Also, the risk factors for completed suicides may differ from the factors influencing the attempts in this study. Clearly, the causes of the spectrum of suicidal behaviors in various contexts need further elucidation.Contrary to popular wisdom, studies of the relationship between breastfeeding and asthma have yielded disparate results. Similar numbers of studies have shown benefit, risk, or no effect at all. Takemura et al note that their study is limited by the potential for faulty recall of past events and by lack of information on parental smoking prior to the onset of asthma. They conclude only that their findings show the possibility of an undesirable effect. The etiology of asthma is almost certainly multifactorial. It is possible that what appears to be a protective factor in one study might appear as a contributory factor in another due to confounding. These 2 studies show the utility of epidemiologic investigations in both raising questions about our common beliefs and assisting us to formulate the next research questions.