Words of advice from the Wild, Wild Weststay on guard when "lead" is rising in the airmay now ring true in a more literal sense. Paul Stretesky of Colorado State University in Fort Collins and Michael Lynch of the University of South Florida, Tampa, found that elevated levels of airborne lead correlate with higher murder rates. Their results are published online this week in the Journal of the American Medical Association (JAMA).

The researchers collected data from the Environmental Protection Agency (EPA) on the concentration of airborne lead in all 3,111 U.S. counties in 1990 and matched them with homicide records for the same year from the Centers for Disease Control (CDC) National Center for Health Statistics. In their analysis, they ruled out the effects of other pollutants and of indicators potentially affecting crime rates, including poverty levels, race and education. The results showed that higher murder rates were statistically related to high levels of airborne lead. Indeed, murders were up to 4 times more common in counties with the most heavy metal in the air.

The effects of lead poisoningalso called saturnismtypically include seizures, mental retardation and, in children, a potentially fatal encephalopathy. Adults chronically exposed to the metal are also known to suffer from a range of symptoms, among them anorexia, abdominal disorders and personality changes. Recent studies on young children and students, however, suggest that chronic exposure to low levels of airborne lead could also cause a tendency toward aggressive and violent behavior.

"The results of this study contribute to the emerging and controversial issue concerning lead's role in ... violent behaviors," the authors of the JAMA paper wrote. "Our findings indicate that persons who commit homicide tend to be exposed to higher levels of lead in the environment." They cautioned, however, that further studies based on individual statisticsnot geographical dataare needed. They admitted, for example, that the homicide rates used in the study represent the victim's rather than the murderer's county of residence, a variable that could introduce errors.