Click for next page ( 152


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 151
Chapter Xl ANALYSIS OF THE DATA CONCERNING DEATH DURING THE NINE-MONTH PERIOD FOLLOWING DELIVERY 11.1 The trait. Although from its very first usage the Genetics Short Form included a question concerning death following delivery, the collection of data with reference to death after the first week following delivery was not originally systematic. However, the instruction to the midwives regarding promptly notifying the ABCC in the event of an abnormal preg- nancy termination included a statement regard- ing death during the first days of life. As the autopsy program developed, the midwives often voluntarily informed the ABCC of death at any time during and after the first week postpartum. The routine follow-up visit was usually made between the seventh and fourteenth postpartum days sometimes later, usually in consequence of recurrent transportation crises in the "motor pool," seldom earlier, due largely to the system of collecting Genetics Short Forms from the midwives and also a natural and proper reluc- tance on the part of Japanese mothers, espe- cially during the winter months, to permit the examination of their children at an earlier date. At the time of this follow-up visit, any deaths which had occurred were noted. This system led to quite complete information regarding death during the first week of life. In the actual analysis, we have confined our attention to the occurrence of death during the first six days postpartum. In the strict sense, this is not the neonatal period, which is usually thought of as including the first month of life, but we shall apply the term "neonatal" to it as a matter of convenience. In addition to this information regarding neonatal death, there is available supplementary information from the "9-months program." As noted earlier, infants were selected for It follow-up studies at age 9 months on the basis of the terminal digit in their registration num- bers. Thus, depending on the personnel and facilities available, one month an attempt might be made to re-examine all infants whose regis- tration number ended in O or 5, the next month 0, 5, or 9, etc. At the time of the contacts necessary to arrange for this examination, it could be determined if the infant had died in the interval between our initial examination and the time of the second contact. There is thus available supplementary information for 21,788 infants concerning the occurrence of death be- tween birth and the ninth month following de- livery. The latter figure is somewhat approxi- mate because of variations in the "contact day," but no bias exists with reference to parental radiation history. 11.2 The genetic argument for radiation-in- d~ced changes in the neonatal death rote. The argument from which one might expect radiation-induced genetic changes in the neo- natal death rate is wholly analogous to the argu- ment for differences in the stillbirth rate with increasing parental exposure (see Sec. 9.2 ~ . Mutations detected in the first generation of offspring by a study of the neonatal death rate would consist largely of the dominant lethals (or possibly in a more strict sense the suble- thals) although some recessive lethals might also be recovered because of the fortuitous com- bination in some individuals of an induced lethal mutation with an allelic lethal mutation of spontaneous origin. Accordingly, we might expect as one manifestation of radiation-induced genetic damage an increase in the frequency of neonatal deaths with increasing parental ex posure.

OCR for page 151
152 Genetic Effects of Atomic Bombs Chapter XI 11.3 Concomitant variables known to ;r~- e'~ce the occurrence of a neonatal death. The sources of extraneous variation affecting the frequency of neonatal deaths are not unlike those variables influencing the frequency of stillbirths (see Sec. 9.3~. Among the con- comitant variables common to both of the afore- mentioned indicators are maternal disease (principally syphilis I, nutrition, birth injury, maternal age, order of birth, and rate of repro 5.0- . PARITY 4.0- ~ 3.5- . 3.0 NEONATAL DEATHS (PER CENT) 2.5 2. 1.5- . 1 ~ . . 0.5- _ mother to infant with increasing maternal age. The age distribution among the exposure cells is such that this bias would tend to inflate the neonatal death rate among the non-exposed or less heavily exposed parents. It was felt that these effects could be ignored. 2. There is ample evidence attesting to the importance of nutrition and birth injury in the neonatal death rate, but these two variables will be ignored in our presentation because there - \ 1 4.5 . 7. ~ ~ 5 ~ 4 ... -- 3 2 ~ 1 ; I I I t I / t _~: ~/ ~ 15 20 25 30 35 40 4S MOTHER S AGE FIGURE 11.1 The distribution of the frequency of neonatal deaths by age of mother for specified parities. auction. In the case of neonatal death this list must be augmented by the addition of at least one more extraneous variable, namely, infantile infectious diseases. In the first six days of life, the principal infectious diseases, which could and frequently do lead to the death of an in- fant, are respiratory infections and infantile diarrhea. The observations and comments with regard to maternal disease, nutrition, and birth injury made in Chapter IX with regard to stillbirths are no less pertinent here. To recapitulate briefly, we noted that: 1. There existed a gradual but significant de- cline in the rate of transmission of syphilis from are no demonstrable or probable differences among exposure cells. With regard to the effect of maternal age and parity on the neonatal death rate the reader is referred to Figures 11.1 and 11.2 and Tables 11.1 and 11.2. The latter figure gives the distribution of the frequency of neonatal death by parity for specified maternal ages. The former figure gives the distribution of the frequency of neonatal death by maternal age for specified parities. From these two figures and tables we note that the effect of maternal age and parity on the frequency of neonatal death is strikingly similar to the effect of age and parity on the

OCR for page 151
Data Cor~cerr~ing Death Daring Period Following Delivery 153 stillbirth rate.t We find that the age-specific parity distributions or the parity-specific age distributions of the frequency of neonatal death tend to be U-shaped. However, unlike the still- birth data the minima of the age-specific parity distributions appear quite similar, whereas the minima of the parity-specific age distributions tend to occur at a higher level with increasing s.o . 4.5 4.0 3.5 3.0 2.5 NEONATAL DEATHS (PER CENT) 1.0 0.5 FIGURE 11.2 The distribution of the frequency of parity. These findings would suggest a signifi- cant effect of parity which varies by mother's age on the frequency of neonatal death (a rate of reproduction effect), but no significant inde- pendent effect of mother's age. This is consistent with Yerushalmy's (1938) observations in the United States. Now since the effect of parity is not independent of maternal age, adequate ad- justment to control variation in parity between ~ A full description of these data is to be published elsewhere. 2.0 \ ., \ / MATERNAL AGE I - 36-40 ~3~-as ..... 26-30 1 1 1 1 1 / / exposure cells would require also taking into account variation in maternal age. However, as in the case of stillbirths, a partitioning of the data as elegant as that required to effect the "best" control would seriously jeopardize the validity of the tests because of (a) the large number of empty cells, and (b) the very small expected values in the bulk of the occupied cells. 1 .,. PARI TY neonatal deaths by parity for specified maternal ages. Accordingly, the data were further partitioned only with regard to parity. We recognized five parity classes, namely, parity 1, parities 2-3, 4-5, ~7, and 8 and higher. The bias which might remain after this partitioning would be one difficult to specify adequately but probably would lead to an increase in neonatal mortality among the more heavily exposed cells because of increased maternal age within these latter cells. The possibility of a spurious irradiation effect

OCR for page 151
154 Genetic Ejects of Atomic Bombs Chapter XI TABLE 11.1 THE EFFECT OF MATERNAL AGE AT FIXED PARITY ON THE FREQUENCY OF NEONATAT DEATHS TotalNormal Neonatal % birthsbirths deaths neonatal AD Parity 1 < 21 1,8821,849 33 1.75 .00808 21-25 8,4798,356 123 1.45 .07422 26-30 3,1253,062 63 2.02 .04877 31-35 595587 8 1.34 .00793 36-40 171164 7 4.09 .01797 41 + 1816 2 11.11 .00733 Total 14,27014,034 236 - .16430 X25 = 21.293** IAD = .08215 PT = .26001 Parity 2 < 21 415 413 2 0.48 .01684 21-25 6,572 6,482 90 1.37 .OS416 26-30 6,294 6,225 69 1.10 .04273 31-35 1,399 1,383 16 1.14 .00585 36-40 360 46 - } 1.72 .01127 ........................ 15,086 14,902 184 .13086 X24 = 4.818 IAD = .06543 PT = .27488 Parity 3 < 21 .........................36 35 1 2.78 .00374 21-25 2,188 2,151 37 1.69 .05859 26-30 5,729 5,658 71 1.24 .02997 31-35 2,357 2,330 27 1.15 .02799 36-40 514 508 6 1.17 .00531 41 + 66 65 1 1.52 .00094 Total 10,890 10,747 143 - .12654 X25 = 3.864 IAD_ .06327 Parity 4 <: 21 . 21-25 26-30 31-35 36-40 41 + PT = 19843 474246012} 2.52 .05783 2,8282,78840 1.41 .00230 2,4872,45532 1.29 .03262 7367288 1.09 .02569 83821 1.20 .00183 Total 6,610 6,517 93 - .12028 X24 = 5.087 IAD = .06014 PT = .12044

OCR for page 151
Data Concerning Death Daring Period Following Delivery 155 TABLE 11.1- Continued TOta1 NOrma1 NeOnata1 % births births deaths neonatal AD ParitY 5 < 21 - - 21-25 7670 6 7.89 .08992 26-30 975 960 15 1.54 .00979 31-35 1,737 1,712 25 1.44 .01437 3~40 823 95 _} 0.98 .08534 TOta1 3,706 3,651 55 - .19942 X23 = 22.9994 ~IAD = .09971 ParitY 6 < 21 - - - - 21-25 ......................... 11 10 1 9.09 .03229 26-30 ......................... 259 254 5 1.93 .06458 31-35 ......................... 943 936 7 0.74 .18519 36~0 ......................... 785 775 10 1.27 .00237 41 + ......................... 135 131 4 2.96 .08594 . TOta1 2,133 2,106 27 - .37038 X24-11.484* IAD .18519 PT-.03887 ParitY 7 + < 21 ......................... - - 21-25 ......................... 3 3 ~ 1.72 .00052 2~30 ......................... 55 54 1) 31-35 ......................... 479 471 8 1.67 .00285 36-40 ......................... 1,167 1,146 21 1.80 .034S4 41 + 483 476 7 1.45 .03221 XO3 = .254 IAD = .03506 2,187 2,150 37 .07012 PT = .03985 All Parities <21 2,3372,301 36 1.54 .00392 21-25 17,80117,532 269 1.51 .02307 26-30 19,26519,001 264 1.37 .01053 31-35 9,9979,874 123 1.23 .02378 36-40 4,5564,488 68 1.49 .00480 41 + 926911 15 1.62 .00252 TOta1 54,882 X25 = 4.641 Heterogeneity X2 65.158** r 54,107 775 I'AD (for age) =.07447 DF=23

OCR for page 151
56 Genetic Ejects of Atomic Bombs Chapter XI TABLE 11.2 THE EFFECT OF PARITY AT FIXED MATERNAL AGE ON THE FREQUENCY OF NEONATAL DEATHS Mother's age: < 21 Total Normal Neonatal % births births deaths neonatal AD 1 1,8821,849 33 1.75 .11310 2 415413 2 0.48 .12393 4 + 3435 _ } 2.50 .01083 . . . . . . . . . . . . . . . . . . . . . . . . . . . Total 2,3372,301 36 .24786 X2O = 3.872 IAD = .12393 PT = .04258 Mother's age: 21-25 1 8,479 8,356 123 1.45 .01937 2 6,572 6,482 90 1.37 .03515 3 2,188 2,151 37 1.69 .01486 4 472 460 12 2.54 .01837 5 76 70 6 7.89 .01831 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 11101} 7.14 .00298 7 -- 33 Total 17,80117,532269 .10904 X25 = 28.736~* IAD = .05452 PT = 32435 Mother's age: 26-30 1 3,125 3,062 63 2.02 .07749 2 6,294 6,225 69 1.10 .06625 3 5,729 5,658 71 1.24 .02883 4 2,828 2,788 40 1.41 .00479 5 975 960 15 1.54 .00629 6 259 254 5 1.93 .00557 ' ' 55 54 1 1 81 .00095 I t ------- - Total 19,265 19,001 264 - .19018 X 6 = 14.791* IAD = .09509 PT = .35103 Mother's age: 31-35 1 595 587 8 1.34 .00559 2 1,399 1,383 16 1.14 .00998 3 2 357 2,330 27 1.15 .01646 . . . . . . . . . . 4 2,487 2,455 32 1.29 .01153 5 1,737 1,712 25 1.44 .02987 6 943 936 7 0.74 .03788 7 + 479 471 8 1.67 .01734 Total 9,997 9,874 123 .12866 X26 = 3.589 IAD = .06433 PT = .18215 Mother's age: 36- O 1 171 164 7 4.09 .06640 2 360 353 7 1.94 .02429 3 514 508 6 1.17 .02496 4 736 728 8 1.09 .04456 5 823 814 9 1.09 .04902 6 785 775 10 1.27 .02562 7 + 1,167 1,146 21 1.80 .05348 _ Total 4,556 4,488 68 .28834 X26 = 11.456 IAD = .14417 PT = .083O1

OCR for page 151
Data Concerning Death Daring Period Following Delivery 157 TABLE 11.2- Continued Total Normal Neonatal % births births deaths neonatal AD Mother s age: 41 + 2 46 46 2: 3.13 .06528 3 66 65 1 1.52 .00468 4 83 82 _ } 0.56 .12763 5 6 135 131 4 2.96 .12287 7 + 483 476 7 1.45 .05584 Total 926 911 15 .37630 X-4 = 3.781 IAD = .18815 PT = 01687 All ages: 1 14,270 14,034 236 1.65 .04514 2 15,086 14,902 184 1.22 .03800 3 10,890 10,747 143 1.31 .01411 4 6,610 6,517 93 1.41 .00045 5 3,706 3,651 55 1.48 .00349 6 2,133 2,106 27 1.27 .00408 7 + 2,187 2,150 37 1.69 .00801 . Total 54,882 54,107 775 - - X26 = 12.462 I"AD (for parity) = .08320 Heterogeneity %2 = 53.763** DF = 23 on the neonatal death rate due to a non-random distribution (by parental exposure) of contacts with the agents responsible for the more serious infectious infantile diseases such as the respira- tory infections and infantile diarrhea is difficult to evaluate. We are not aware of any widespread outbreaks of respiratory infections or diarrhea among the newborn infants of Hiroshima and Nagasaki during the interval 1948-1953. 11.4 The data. In Tables 11.3 and 11.4 are given the distribution of neonatal deaths by sex of infant, parental exposure and city without and with allowance for parity. The corresponding analytical results are given in Tables 11.5 and 11.6. Let us consider first the evidence for an irradiation effect from the data prior to partitioning by parity. From inspection of Table 11.3 and from the results given in Table 11.5 we note a significant mother-father exposure interaction. The occurrence of this interaction considerably complicates further analysis of these data since, as in the non- orthogonal chase of the analysis of variance, the disproportion among the numbers of observa- tions in the various ways of classification leads to distortion of main effects tests based upon the marginal totals, due to the unequal contributions to the marginal class totals. Under the circum stances it would seem that the most appropriate tests for those ways of classification other than those involved in the interactions, here mother's exposure and father's exposure, would be the sum of the chi-squares obtained by testing a given way of classification at each level of the ways of classification involved in the interac- tion. Thus, in this instance, the test of the city effect would be obtained by summing the six- teen chi-squares and their degrees of freedom obtained by testing the effect of city on the vari- ate at each of the sixteen mother-father levels. The disadvantage of this procedure is that all sixteen tests contribute equally to the sum, whereas the power of the individual tests will vary considerably. Ideally, perhaps, one would weight the individual contributions of these tests to the sum by their power functions. This has not been done, and hence the sum tests will probably be somewhat biased. When the data are analyzed as just outlined, for only one of the eight exposure tests when the exposure of one parent is held constant does a significant effect of mother's or father's ex- posure emerge. The weight ~ this one positive finding should be given in the face of seven non-significant, similar tests is debatable. There are significant differences between the sexes but

OCR for page 151
158 ~Genetic Efects of Atomic Bombs Chapter XI TABLE 1 1.3 THE FREQUENCY OF NEONATAL DEATHS BY PARENTAL EXPOSURE, CITY, AND SEX OF INFANT (Unrelated parents) Male infants Hiroshima o Fathers l{d En 2] d En 33 d lip En 4-54 d UP 1 2 751 7 .0093 909 11 .0121 1,081 16 .0148 551 10 .0181 Total 10,066 155 .0154 3,975 58 .0146 1,637 20 .0122 56766 114 .0179.0183 34-5 304206 7 .0340 .0296 190 219 258 2 .0078 108 51 3 - .0588 79 2 .0253 116 2 .0172 En13,197 1,98780345716,444 Totals d203 181412247 up.0154 .0091.0174.0263.0150 Male infants Nagasaki cow : - c, Fathers , 123 4-5Total rn 7,417 1,087 124 72 8,700 1]d 130 19 3 152 up .0175 .0175 .0242 .0175 rn4,7322,055137 1037,027 2]d80312 113 up.0169.0151.0146 .0161 rn34913150 13543 34d101 -11 up.0287.0076 .0203 rn2745518 15362 4-5gd221 5 up.0073.0364.0556 -.0138 rn12,7723,328329 20316,632 Totals d222536 281 up.0174.0159.0182 -.0169

OCR for page 151
Data Cor~cerr~ir~g Death During Period Following Delivery TABLE 11.3 Continued Female infants Hiroshima Fathers 1 2 3 4-5 Total rEn8,0867112711779,245 1g d871152105 up.0108.0155.0185.0113.0114 rn2,4668861821183,652 2g d32121 45 up.0130.0135.0055 .0123 En1,039199246751,559 3] d81 9 up.0077.0050 .0058 rn533905558736 4-5gd71 -~ up.0131.0111--.0105 rn12,1241,88675442815,192 Totals d1342562167 p.0111.0133.0080.0047.0110 Female infants Nagasaki , ~ o Fathers . 1234-5Total rn6,8431,023112648,042 lid82122197 up.0120.0117.0179.0156.0121 rn4,3731,967131706,541 2/ d60262 88 up.0137.01320153 .0135 En3731394321576 3/ d611 8 ~ p.0161.0072.0233-.0139 rn272581613359 t5g d51 6 up.0184.0172 .0167 En11,8613,18730216815,518 Totals d1534051199 up.0129.0126.0166.0060.0128 159

OCR for page 151
z a o A. Led z It Us t~s td ~ ~ o an ~ z~ o o z ~7 Em d. - - ~ fir ~ No ~ GN ~ ~ ~ ret - (-I Cot DO ~ O ret fir ~ _I ~ ~ _. ~ ~ ~ r<` ~ ~ ~ ~ [-4 ON ~ r, I _~ O ~4 - = - ~= ~ tax mm ~ x 1 ~ ~ 00 - ~ ~ ~ x ~ ~ ~1 0~= ~ ~ 0- Oom O O ~ O ~ =~ - - ~ - SCAN - ~ - =~= \0~ - Am O ~ ~ ret vet ~ ~c; ~ ;N ~en, ~ ret td ~ Oo ~ a ~ O ~ O O ~ Oo _ ~ err 0 c~=- oo-^ =-o GN~= v~ -- r`+v ~0 '_~ - +~ ~= - ~v~ ~ o ~om E ~c~ ~ O ~ O - ^ O ~ O O ~00 ~ -4 . ~ . ~ . . . '_ . - mo ~OX m~` V~ - ^ - X 00^ ~x~ -~r~ r~ -~1- ~ ~ ~\om ~ =^ O m~0 ~ O ~ O ~ O +~- -> Ct L, o ~ X o 'd. O G ~l l _. | ~| X _d 1^ Cd L Iv- - 1 - 0 1 o 1 1 ~1 1 - ~ - v ~ ~1 ~1 1 ~1 1 =~^ o =1 1 ~ 1 1 - ~= X 1 1 ~ 1 1 - - x ~ ~ ~ X o 1 1 1~ o o -- ~ o l I ~o-3 ~ 1 1 ~ 1 1 1 -~~o co C~ ~~) o ~ ~ ~ v~ 1 1~ 1 1 r~ C~xv~ x ~ ~ ~ ~ 1 11 1 x 0 o, 0 . . . . - ~oOm w~O~D ~000 - ~m ~ 1 1 '_ooO ~d ~ ~ X-= v ~ ~r ~U~ 1 1 v~ o ~ ~ ~ ~ ~0 ~0 ~ ~ 0 0 0 0 - ~~^~~ ~1 1 ~1 1 r ~1 1 =+c~ v ~V ~ C ~1 1 -4 1 1 1 1 oco 0 0 ~0 . . 1 1 1 --- G`_ ^ oO I I CN 1 1 ~O O _ o ~1 1 1 o ~o ~0 0 0 ~ C~ V} 1 -~- ~ V~ - U~ o. (~4 ~ ~ ~ 1 1 ~ 1 1 ~ 1 1~ 1 1 ~0 ~ ~ - ~= x 1 1 ~ o1 1 _ X XU~ ~0 00 _ v~ e~ ~ ^~.0 00~-_ ~- 1-.r~= ~ (~, ~ ~ ~ ~ ~ X rrs ~C~ ~ '' - - ~ N t_ _~ ~- o ~ OoO O C ~. - -'o~ oo 1 1 ~ 1 1 _1 1 1 - . 1 1 ~ 1 1 ~ 1 1 ~ 1 1 _ ~A ~ ~ o r~- ~oo~ ~~- v~D V~~l_ V-+ ( ~C~ O X O ~ 1- G~ ~ 1- ~ax~ ~ - 0 ~ O -~ O O _. O _~-~ -1-~ ^~ ^~1- 0 1 1 ~oom CN ~ ~- O ~ X V~ ~ ~ 1 1 -~ ~ -I ~ O ~ ~ O O O O _ O. +00 \0_- _~- -1_0N 1_+- _~_+ cd 00- -~~ -=1- ~_4C;~ O ~r^- ~ ~O ~ ~ ~ ~ ~ X - X ~ ~ ~- ~< ~ ~oo O ~ O ,~` O O. O `^ O _ e4 - ~~0 ~r^= ooV~oO - mv~ ~ 1 1 ~r`~_ O O O O O ~ ~ -~- ~_v~= ~v~^ -~= ~ 1 1 o~rm ~ ~ ~ ~ - 0 ~ ~ ~ ~ I | ~ N ~ ~ O 0^ 0 ~ 0 0 ~ O . _. . C~= x - ~ r~oo~ ~= ~m ~r~r~ v-~O x - + - ~= ~mr` - _ O ~ O o O O ~ O 1-m~ X C;NO ~X~ u~oo~ oo - = ~ C\O =1_ - ~C~I 00= ~ ~ - 1 ~ m ~ ~ ~ ~ ~ _ v~ ~ ~ ~0O ~ _ v ~O 00 0 ~O 0 0 \0 0 r~ - 160 ~-- -0 00 CsCN_ 00 ~le~ CN_4 _d O O -o ,~ cN ~ ~_ x ~ = ~- o ~- V- ~ ~ ~ _~ _4 I_ _~ ~ ~ G~ ~_ E - ~ ,~r ~C, (~, (~, -~+ o 1 1 ~- I ~ _' r~ o, 0 ^~~ G~ - O oo - - O ~ O =~0 ~r oo c~ 0 ~ 0 _ ~ 0 ~ 0 ~ ~ ~ 0 0 o, 0 ~ I_ ~ ~ v~ ~ ~ ~ ~ ~ <) ~ X 1 a:) _ oo ~r~o ='-~ =~xr oo-~ 0 ~ 1 ~o~ C~ ~ 0 ~ a, ~ _ ~ v~ ~ 1 v~ ~ ~ \C, 0 c~ 0 _. 0 0 1 c<` 0 GN ~0 ~ 1-~ - = O - ~ - - - 0 1 ~ GN C;~ x ~ 0 ~ - ~r ~ ~-~ ~ ~ 1 =~ - ~ ~ ~ ~,., ~4c~ o 1 ~-~ ~ cr,0 I_ 0 c~ o ~ = = o - = ~ x ~ - ~ O Cx C ==o =~ - - ~ - - _ - ~ w 1 =_= x _ c4 ~ - o _ ~ ~ ~ 0 1 - 1 ~- c' {<, O \C, O ,~, O O 1 'd' O V) C! C) ~ ~ C =) G C ~ ~ C ~ CL. C ~ ~ C ~ C C ~ ~ OCR for page 151
Data Concerning Death Daring Period Following Delivery 161 TABLE 11.5 CH! SQUARE ANALYSIS OF THE FRE QUENCY OF NEONATAL DEATHS BY PARENTAL EXPOSURE, CITY, AND SEX (Unrelated ParentS) (a) 4 X 4 SOUrCe DF %2 p TABLE 11.5 Continued TOta1 ..................... 63 83.974 .05-.10 InteraCtiOnS, first Order CS ..................... CM ..................... CF ..................... SM ..................... SF ..................... MF ..................... Sex a Main effects MOtherS: FatherS 1......... FatherS 2...... FatherS 3...... FatherS 4,5.... FatherS: MOtherS 1.... MOtherS 2........ MOtherS 3........ MOtherS 4,5...... M1F' ......... M1F2 ........ M1F3,4,6 ....... M2F1 ......... M2F? ........ M2F3,J,6 ...... M3,4,5F1 ....... M3.4,SF2 ...... M394,5F3,4,6 .... Sum .................. M1F1 1 M2F1 1 M3,4,5F1 1 M1F2 1 M2F2 1 M9,4,N,F2 1 M1F3,4,6 1 M2F3,.,6 1 M3,A,5F3,4~,5 ........ 1 1 0.070 3 3.192 3 5.336 3 2.202 3 5.224 9 22.986 3 1.223 3 4.365 3 12.480 3 4.378 3 6.258 3 5.280 3 7.132 3 2.479 1 13.296 1 0.051 1 1.840 1 2.504 1 0.077 1 0.070 1 2.080 1 0.204 1 6.266 9 26.388 2.048 0.091 1.837 0.361 0.168 3.265 0.780 1.093 0.036 .................. 9 9.679 .70-.80 .30-.50 Total ... .10-.20 .50-.70 .10-.20 001-.01 (b) 3 X 3 Source Interactions, first order CS ..................... CM ..................... CF ..................... SM ..................... SF ..................... MF ..................... .70-.80 .20-.30 001-.01 .20-.30 Main effects .10-.20 Sex (S) a .10-.20 .05-.10 .30-.50 < .001 .80-.90 .10-.20 .10-.20 .70-.80 .80-.90 .10-.20 .50-.70 .01-.02 Mother's exposure 2... Mother's exposure 3... Mother's exposure 4,5.. City ~ C ~ Mother's exposure-2. . . Mother's exposure 3... Mother's exposure 4,5.. Mother (M) ............. .001-.01 Father (F) a Mother's exposure 2... Mother's exposure-3. . . Mother's exposure-4,5. . .10-.20 .70-.80 .10-.20 50- 70 Sum .50-.70 .05-.10 .30-.50 .20-.30 .80-.90 .30-.50 aTo increase the numbers of observations, cate- gories 3, 4, and 5 have been pooled in these tests. a adjusted for mothers. DF X2 P .50-.60 35 39.587 0.161 .50-.70 0.425 .80-.90 4.354 .10-.20 2.258 .30-.50 5.007 .05-.10 5.441 .20-.30 1 0.106 1 0.432 1 2.604 . 3 3.142 1 1.005 1 0.454 1 0.980 . 3 2.439 2 6.216 3.708 1.108 2.421 6 7.237 .70-.80 .50-.70 .10-.20 .30-.50 .30-.50 .50-.70 .30-.50 .30-.50 .02-.05 .10-.20 .50-.70 .20-.30 .20-.30

OCR for page 151
162 Genetic Efects of Atomic Bombs Chapter XI TABLE 1 1.6 CHI-SQUARE ANALYSIS OF THE FRE QUENCY OF NEONATAL DEATHS BY PARENTAL EXPOSURE, CITY, AND PARrrY (Unrelated parents) Total Interactions, first order CM 3 3.192 .30-.50 CF 3 4.855 .10-.20 CP 4 14.401 .001-.01 ME 9 22.986 .001-.01 MP 12 21.436 .02-.05 FP 12 13.334 .30-.50 Main effects (only CP interaction taken into account) Parity (P) Hiroshima ~ H ~ Nagasaki (N) ... City (C) Parity Class 1 Class- 2 Class 3 Class 4 Class 5 Mother (M) H: Parity Class 1 Class 2 Class 3 Class 4 Class 5 N: Parity Class 1 Class 2 Class 3 Class 4 Class 5 Source DF %2 p 189.601 .05-.10 .. 159 Sum ........... Father (F) H: Parity Class 1 Class 2 Class 3 Class 4 Class 5 N: Parity Class 1 Class 2 Class 3 Class 4 Class 5 Sum 3 3 3 - 30 4 5.872 4 32.045 l 3 11.753 3 3.424 3 0.809 3 2.882 3 0.775 5.235 1.977 2.552 2.409 4.662 3 3 3 3 3 30 36.478 3 4.260 3 2.387 3 10.236 3 2.559 3 1.976 18.585 <.001 0.458 .30-.50 0.420 .50-.70 1.720 .10-.20 0.200 .50-.70 .001-.01 .30-.50 .80-.90 .30-.50 .80-.90 From Tables 11.4 and 11.6, we note that when neonatal deaths are further partitioned by parity rather extensive heterogeneity in the data is revealed. Both the city-parity and mother-father interactions are significant. To adjust the main effects tests as indicated in Chapter VI leads, in this instance, to such ex tensive partitioning as to seriously jeopardize the validity of the tests. However, one could adjust for one of the interactions at a time. Accordingly, two sets of "main effects" tests could be generated. If one ignores the mother father interaction, the main effects tests of city, parity, and exposure would be as given in Table .20-.30 11.6. If the city-parity interaction were ignored, < .00i the tests of exposure and city would be as given in Table 11.5a, and there would, in addition, be a test of parity. One must obviously view the results of these two sets of tests as approximate. However, in neither instance does evidence emerge indicating a significant, consistent effect of parental exposure. The results from the "at-birth" examinations are corroborated by the Endings at 9 months of age, presented in Table 11.7. The reader should note the following departures from the usual classifications of the 9-months data: (1) sex of infant is ignored, (2) exposure categories ~ 4 and 5 have been pooled, and (3) deaths from birth to 9 months are recorded. The purpose of these latter procedures was to bring to bear on the problem of infant mortality fol lowing dissimilar exposure experiences on the part of the parents, the largest possible number of observations. No attempt has been made to control concomitant variation which would pre sumably be no less of a problem here than else where in these data, and would, in the main, tend to inflate the death rate among infants born to the more heavily exposed parents. The analy sis of these data is presented in Tame 11.8. We note no effect of city, maternal or paternal exposure, or evidence of heterogeneity between the cities or exposure cells. 11.5 Summary. No consistent, significant effect of parental exposure on neonatal mortality emerges from the data obtained in Hiroshima and Nagasaki on deaths occurring in the first six days postpartum. Analysis of neonatal and infantile mortality obtained at the time of the 9-months examination fails to disclose an effect of parental exposure on infant survival. No consistent exposure trend is exhibited by these data. .10 - .20 .50-.70 .30-.50 .30-.50 .10-.20 .10-.20 .20-.30 .30-.50 .01-.02 .30-.50 .50-.70 1.085 .70-.80 4.322 .20-.30 3.156 .30-.50 2.322 .50-.70 0.846 .80-.90 33.149 .30-.50 not the cities. Omission of the category 1 parents from the analysis results in the disappearance of the M-F interaction (Table 1l.5b). As be- fore, there is no significant effect of father's exposure but now a barely significant effect of mother's exposure appears.

OCR for page 151
Data Concerning Death Daring Period Following Delivery TABLE 1 1.7 THE FREQUENCY OF DEATHS BE TWEEN BIRTH AND NINE MONTHS OF AGE BY PARENTAL EXPOSURE AND CITY (Unrelated parents) . , . H~rosn~ma i_ v, con 163 Mothers 12 3-4-5Total rn 5,6781,803 1,0648,545 1g d 2548548 387 up .0447.0471.0451 .0453 En4825281881,198 23d2427859 up.0498.0511.0426.0492 En319186264769 3-~54d1681438 up.0502.0430.0530.0494 - rn6,4792,5171,51610,512 Total] d29412070484 .up.0454.0477.0462.0460 Nagasaki , ~ Mothers , ~ - ~ 1 2 3~-5Total En 4,225 2,599 3967,220 14 d 209 125 25359 p .0495 .0481 .0631.0497 En5831,0511171,751 2\ d2061283 up.0343.0580.0171.0474 En12014966335 3-4-5gd47516 up.0333.0470.0758.0478 En4,9283,7995799,306 Total] d23319332458 up.0473.0508.0553.0492 TABLE 11.8 ANALYSIS OF THE FREQUENCY OF DEATHS BETWEEN BIRTH AND NINE MONTHS OF AGE, BY PARENTAL EXPOSURE AND CITY (Unrelated parents) Source DF %2 Total ..... 13.146 p .70-.80 Interactions, first order CM 2 0.382 .80-.90 CF 2 0.658 .70-.80 MF 4 4.647 .30-.50 Main effects City 1 1.097 .20-.30 Mothers 2 1.081 .50-.70 Fathers 2 0.087 .95-.98