A first study of the statistics of pulmonary tuberculosis / by Karl Pearson.

  • Pearson, Karl, 1857-1936.
Date:
1907
Catalogue details

Licence: In copyright

Credit: A first study of the statistics of pulmonary tuberculosis / by Karl Pearson. Source: Wellcome Collection.

Provider: This material has been provided by The Royal College of Surgeons of England. The original may be consulted at The Royal College of Surgeons of England.

    16/30 (page 14)
    For the male pedigrees we find— •464,608 = r - 1 134,724 r2 + -228,679 v3 - 017,452 r\ whence the correlation coefficient = ‘33, to two decimal places. For the female pedigrees we have :— •667,718 = r + 1 047,991 r2 + '179,317 r3 + 038,549 r\ whence the correlation coefficient = 44. The average of these values = 385. This value is itself higher than has been found for the inheritance of cephalic index from mother to offspring, or in some shorter series for stature from parent to child. But it is somewhat smaller than the values obtained for long series of stature or pigmentation in man. The male pedigree results show the reduced correlation to be expected from less complete records. We see accordingly that parental inheritance lies between 4 and 6, approaching the upper limit, if we suppose the family histories when completed to give as great a tuberculous percentage as that observed by Dr. Thompson. Lastly, let us ask what effect it would have on the intensity of parental influ- ence if we supposed 13 per cent, and not 10 per cent, of the community to suffer from tuberculosis. Our tables now become :— Male Pedigrees. Parent. Female Pedigrees. Parent. T. N.T. Totals. T. N.T. Totals. T 79 361 440 T 157 509 666 N.T 78 2,867 2,945 N.T 157 4,300 4,457 Totals 157 3,228 3,385 Totals 314 4,809 5,123 From these tables we find— Equation from male pedigrees :— •842,527 = r + -946,738 r2 + -081,783 r3 + -023,793 r\ giving r = 55. Equation from female pedigrees :— ■885,075 = r + -869,594 r2 + 061,995 ? + '077,276 r\ giving r = -58 to nearest second place of decimals. It will thus be seen that an increase of 3 per cent, in the amount of tuberculosis in the general population from which the Crossley Sanatorium patients are drawn would lower the correlation coefficients by '04. An increase in the percentage of tuberculous offspring of tuberculous parents from the one in three of the incomplete family records to the one in two of Dr. Thompson’s complete family records * raises the correlation from about 4 to ’6. We may, I think, accordingly conclude with safety that the intensity of the inheritance factor in pulmonary tuberculosis is greater than 4 and less than *6- * This is the increase which is essential also on the Mendelian hypothesis, see p. 8.