Temporal Association Between Nonfatal Self-Directed Violence and Tree and Grass Pollen Counts.

OBJECTIVES: Prior research suggests a possible association between pollen and suicide. No studies have examined the relationship between pollen and attempted suicide. This study examines the temporal association between airborne pollen counts and nonfatal suicidal and nonsuicidal self-directed violence (SDV) requiring an emergency department visit.

METHODS: Data on daily emergency department visits due to nonfatal SDV as identified by ICD-9 diagnosis criteria were extracted from emergency department medical records of Parkland Memorial Hospital in Dallas, Texas, between January 2000 and December 2003. Concurrent daily airborne tree, grass, and ragweed pollen data from the city of Dallas were extracted from the National Allergy Bureau online database. The data were analyzed using the time series method of generalized autoregressive conditional heteroskedasticity.

RESULTS: There were statistically significant and positive temporal associations between tree pollen counts and the number of nonfatal SDV events among women (P = .04) and between grass pollen counts and number of nonfatal SDV events among both men (P = .03) and women (P < .0001). There was no significant temporal association found between ragweed pollen counts and number of nonfatal SDV events.

CONCLUSIONS: The study findings suggest that an increase in nonfatal SDV is associated with changes in tree and grass pollen counts. This is the first study that has examined an association between seasonal variation in tree and grass pollen levels and nonfatal SDV event data. The study also used a narrowly defined geographic area and temporal window. The findings suggest that pollen count may be a factor influencing seasonal patterns in suicidal behavior.