To estimate the strength of a suspected causal association between fearfulness and carrying a lethal weapon among urban middle-school students, while holding constant other suspected risk factors.
A prospective study of an epidemiological sample assessed at baseline in 1992 and 1 year later, with relative risk estimates derived from the conditional form of the multiple logistic regression model used to hold constant alternative explanatory variables.
An urban environment in the mid-Atlantic region of the United States.
An epidemiological sample of 1131 youths enrolled in public middle schools.
Main Outcome Measures:
Carrying a lethal weapon for protection or defense during a 1-year observation interval after the baseline assessment of fears and other suspected risk factors.
Of the 1131 youths, 194 (17%) reported carrying a lethal weapon for protection or defense during the 1-year interval of follow-up observation after baseline; 937 youths (83%) reported that they had not carried a lethal weapon for any reason. Self-reported fears, deviant peer affiliation, and worrying were associated with risk of starting to carry a weapon. For youths with the lowest worrying scores, the lowest neighborhood danger scores, and the least affiliation with deviant peers, self-reported fears were associated with risk of starting to carry a lethal weapon (relative risk estimate, 1.68; 95% confidence interval, 1.11-2.52; P=.01), even after holding constant age, sex, and conduct problems. However, the fear of crowded or closed-in places and the fear of leaving home alone were more salient risk factors than the fear of using public transportation or the fear of open spaces.
In this study, youths with fears were at greater risk of carrying a lethal weapon for protection or defense, even when alternative explanatory variables were taken into account. Pending confirmation by other investigators, this new finding could point out a useful target for public health interventions to reduce the carrying of weapons and associated violence in urban America.Arch Pediatr Adolesc Med. 1997;151:555-560