The relationship of soil-lead to house dust-lead in two inner city neighborhoods in Jackson, Mississippi (pathways of lead poisoning)

The objective of this study was to determine the relationship of soil-lead to dust-lead believed to be pathways to lead-poisoning and to test the null hypothesis that there is no relationship between soil-lead and dust-lead in and about dwellings located in Census Tracts 25 and 31. Thirty sites were targeted located in West Jackson, north (Tract 25) and south (Tract 31) of Jackson State University. The northern tract prior to desegregation was inhabited by all white families whereas the southern tract was from its inception all African-American residents.; At each site five soil samples taken from the street and along the drip-line of the house along with five dust wipe samples taken from the entrance, floor of the most frequented room, window sill and window trough. The samples were digested using microwave digestion pre-set for USEPA - SW846-3015 method. Analyses of dust and soil digestates were performed using Flame Furnace Atomic Absorption Spectrometry (FLAA) according to SW 846 method 7421.; The concentration of soil-lead taken at the street side in Census Tract 25 ranged form 17 to 168 ppm with a median concentration of 90 ppm. In Census Tract 31, the range was 37 to 222 ppm with a median of 125 ppm. The difference in the median values between the two groups was not statistically significant (p ≥ 0.071). In Census Tract 25 the drip-fine soil-lead concentration ranged from 18 to 21,080 ppm compared to 39 to 14,532 ppm in Census Tract 31. The median drip-fine lead concentration of 831 ppm in Census Tract 25. The range of dust-lead loading was 4 to 89,921 mug/ft2 in Census Tract 25 compared to 2 to 21,068 mug/ft2 in Census Tract 31. The median dust-lead loading was 33 mug/ft2 and 34 mug/ft 2, which is below EPA guidance level of 100 mug/ft2 hard floors.; The Mann-Whitney U Test, a non-parametric statistical method, revealed positive relations between soil-lead and dust-lead. The null hypothesis was rejected. These results compare favorably with Spearman rank order correlation. A correlation of 0.843 (p = 0.0436) exists for Census Tract 25 and 0.791 (p = 0.0305) for Census Tract 31.