Document 6BdgxND1JV9N4O7YO3jgY79K3

mm ESTABLISHED 1602 E. I. d u Po n t d e Ne mo u r s 51 Co mp a n y INCORPORATED Wil min g t o n , De l a w a r e 19898 EMPLOYEE RELATIONS DEPARTMENT HASKELL LABORATORY FOR TOXICOLOGY AND INDUSTRIAL MEDICINE Tfoypmtwy 7 1968 Sr. E. King, Director Department of Occupational Health University of Manchester Clinical Sciences Building York Place Manchester 13, England Dear Dr. King: Don Fowler has passed on your letter on the Zielhuis paper to me end I am writing this letter as a first reaction since time in short. It would not appear to me that your data support that of Zielhuis since your correlation coefficients are so low. These yield Ea or a coefficient of determination of no more than 4. hi, thus only 4. hi per cent of the variance is explained by the regression of Hb on PbU (or in the case of FbB on Hb of 7,29 per cent). These figures are insignificant and therefore no further manipulation is warranted. Dr. Zielhuis makes the same point in only accepting a correlation that has an R3 value of 65 per cent or greater. In a quick plot of data immediately available to me from the literature in which individual values are given and not ^ust group averages, I find no significant correlation as you can see from the enclosed graph. I am doubtful about the use of the bisector or '^principal components line" since Fb is presumably a cause of a lowered hemoglobin value, but the reverse is not true (over a reasonable range of values). Thus using both plots has little biological meaning and is not justified. Common sense would suggest that the relationship between hemoglobin and blood lead should be curvilinear with a "no effect" level occurring in the low blood lead range and an increasing effect as the blood lead values rise. We hope to have seme data available shortly on individuals that will enable ua to test this hypothesis. Yours sincerely, GJS/dyb Enc. ORIGINAL SIGNED BY G. J. STOPPS G. J. Stopps, M. D. ,B.S. Chief, Physiology Section K lr 0009023