The Genetics of Schizophrenic and Schizoid Disease1IIiiiiil..1IiiiII@

Leonard L. Heston

Science, New Series, Vol. 167, No. 3916. (Jan. 16, 1970), pp. 249-256.

Stable URL:http://links.jstor.org/sici?sici=0036-8075%2819700116%293%3AI67%3A3916%3C249%3ATGOSAS%3E2.0.CO%3B2-D

Science is currently published by American Association for the Advancement of Science.

Your use of the JSTOR archive indicates your acceptance of JSTOR' s Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless youhave obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, andyou may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.j stor.org/journals/aaas.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen orprinted page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive ofscholarly journals. For more information regarding JSTOR, please contact support@jstor.org.

http://www.j stor.org/Tue May 919:12:362006

The Genetics of Schizophrenicand Schizoid Disease

Summary

Photochenlical reactions in the tropo­sphere, stratosphere, ,and mesosphereare, to a large extent, reactions anl0ngluinor constituents of the atmosphere.The chemistry is markedly limited bythe minimum w'avelength of the solarradiation which penetrates to, a .givenatnl0spheric level. It is useful to differ­entiate between the chelnistry of citysmog and' that of the anlbient atmo­sphere, but the entire atmosphere isbeing polluted and the difference is oneof degree.

Both laboratory ,and :Held studies arecontributing to our knowledge of atmo­spheric photochemistry, and the mostconvincing conclusions have been ob­tain~d by combining the results of thetwo methods of investigation.

References and Notes

1. T. M. Donahue, Science 159, 489 (1968).2. R. D. Cadle, W. H. Fischer, E. R. Frank,

J. P. Lodge, Jr., J. Atmos. Sci. 25, 100(1968); R. W. Fenn, H. E. Gcrbcr, D. Wass­hausen, ibid. 20, 466 (1963).

3. C. E. Junge, Air Chemistry and Radiowactivity (Academic Press, New York, 1963).

4. R. D. Cadle and M. Ledford, Int. J. AirWater Pollute 10, 25 (1966); J. M. Hales,thesis, University of Michigan (1968).

5. G. Liuti, S. Dondes, P. Harteck, J. Amer.Chem. Soc. 88, 3212 (1966).

6. Many rate constants are temperature de­pendent. l-Iowever, for purposes of compari-

The contribution of genetic factorsto the etiology of schizophrenia hasbeen confirmed decisively. Because theinvestigations that have led to thisresult have uncovered questions cuttingacross several fields of inquiry, a freshlook at some central aspects of theschizophrenia probleln is warranted.These questions and the factual back­ground underlying, them are the mainconcerns of this article. Because em­phasis is placed on formulating testablehypotheses, the evidence is organizedin support of a particular genetictheory.

16 JANUARY 1970

son, only values for _ 300 0 K are given inthis article. The rate constants presentedwere in nlany cases selected from among sev­eral given" in the literature. Although consid­erable judgment was used in the selection,there was often little basis for choice amongseveral values. If we use the product of therate constant and reactant concentrations,the rate of disappearance of reactants orformation of products may be estimated asabove. The rates obtained in this way areused as a guide to the relative importanceof shnultaneous reactions occurring in theatmosphere.

7. R. D. Cadle, in Air Pollution, Handbook, P.L. Magill, F. R. Holden, C. Ackley, Eds.(McGraw-Hill, New York, 1956), pP. 3-1 to3-27.

8. E. R. Gerhard and H. F. Johnstone, Ind.Eng. Chem. 47, 972 (1955); T. C. Hall, Jr.,thesis, University of California, Los Angeles(1963); N. A. Renzetti and G. T. Doyle, l.Air Pollute Control Ass. 8" 293 (1959); Int.J. Air Pollute 2, 327 (1960).

9. P. Drone, H. Lutsep, C. M. Noyes, J. F.Parcher, Environ. Sci. Technol. 2, 611 (1968).

10. H. F. Johnstone and D. R. CoughanoWf,Ind. Eng. Cheln. 50, 1169(1958).

11. R. D. Cadle and J. W. Powers, Tellus 18,176 (1966).

12. R. D. Cadle and R. C. Robbins, Discuss.Faraday Soc. 30, 155(1960).

13. E. L. Wong and A. E. Potter, Jr., J. Chem.Phys. 39, 2211 (1963).

14. R.D. Cadle, Discuss. Faraday Soc. 37, 66(1964). '

15. H. W. Ford, G. J. Doyle, N. Endow, J.Chem. Phys. 26, 1337 (1957).

16. J. P. Lodge, Jr., and J. B. Patc, Science 153,408 (1966).

17. }-L W. Ford and N. Endow, J. Chem. Phys.27, 1156 (1957).

18. F. Kaufman and J. R. Kelso, ibid. 40, 1162(1964).

19. J-L W. Ford, G. J. Doyle, N. Endow, ibid.26, 1336 (1957).

20. L. R. J. McHaney, thesis, University ofIllinois (1953).

21. L. A. Cavanagh, C. F. Schadt, E. Robin­son, Environ. Sci. Technol. 3, 251 (1969).

Leonard L. Heston

l~be Basic Evidence

During the first half of this century,systen1atic family studies demonstratedth'at the distribution of schizophrenia isthat of a genetic disease. Relatives ofschizophrenics were found to be af­fticted with the illness much morefrequently than members of the generalpopulation. The child of a schizophrenicparent, for example, was found to havea risk of schizophrenia about 15 timesthat of a member of the population atlarge. It was found that, among allc] asses of relatives, the closer the ge-

22. Of course, CO2 is involved ill plant photosYll­thesis.

23. P. A .. Leighton, Photochemistry of Air Pollu­tion (Academic Press, New York, 1961).

24. R. D. Cadle and H. S. Johnston, in Proceed­ings oj the Second National Ail' PollutionSymposium (Stanford Research Institute,Menlo Park, California, 1952), 1'1'. 28-34.

25. R. D. Cadle and J. W. Powers, J. Phys.Chent. 71, 1702' (1967).

26. W. E. Wilson, Jr., and J. T. O'Dollovan,J. Chenl. Phys. 47, 5455 (1967).

27. M. A. A. Clyne and B. A. Thrush, Proc.Roy. Soc. London Sere A Math. Phys. Sci.275, 559 (1963).

28. .J. N. Pitts, Jr., A. U. Khan, E. B. Smith,R. P. Wayne, Envil'oll. Sci. Technol. 3, 241(1969).

29. 1. M. Campbell and B. A. Thrush, Proc.Roy. Soc. London, Sere A Math. Phys. Sci.296, 222 (1967); F. Kaufman and J. R. Kelso,Discuss. Faraday. Soc. 37, 26 (1964); O. R.Lundell, R. D. Ketcheson, H. I. Schiff, paperpresented at the 12th International Symposiumon Combustion, Poitiers, France, 1968.

30. L. F. Phillips and H. 1. Schiff, J. Chem.Phys. 37, 1233 (1962).

31. F. Kauflllan, Ann. Geophys. 20, 106 (1964).32. R. l~. Cadle, R. Bleck, J~ P. Shedlovsky, I.

H. BUfford, J. Rosinski, A. L. Lazrus, J.Appl. Meteorol. 8, 348 (1969).

33. T. G. Scholz, L. E. I-Ieidt, E. A. Martell, D.II. Ehalt, Trans. Amer. Geophys. Union 50,176 (1969).

34. A. E. Bainbridge and L. E. Heidt, Tellus18, 221 (1966).

35. T. G. Kyle, D. G. Murcray, F. H. Murcray,W. J. Williams, J. Geophys. Res. 74, 3421(1969).

36. Based 011 U.S. Standard Atmosphere 1962(U.S. Government Printing Office, Washing­ton, D.C., 1962).

37. P. M. Furukawa, P. L. Haagenson, M. J.Scharberg, Nat. Center Atmos. Res. Tech.Note 26 (1967), pp. 0-55.

38. B. G. Hunt, J. Geophys. Res. 71, 1385 (1966).39. J. B. Pearce, ibid. 74, 853 (1969).40. The National Center for Atmospheric Re­

search is sponsored by the National ScienceFoundation.

netic relationship to a schizophrenicproband (or index case) is, the greateris the likelihood of schizophrenia inthe relative. Finally, and most tellingof all, 1110nozygotic twins were foundto be concordant with respect to schizo­phrenia about four tinles as often asdizygotic twins. Several authorities havecritically reviewed these basic data (1,2) . But, despite the supporting evidence,a genetic etiology for schizophreniawas not widely accepted, especially inthis country. It was pointed out thatthe investigators did not pay enoughattention to important procedural mat­ters, such as providing sanlpling safe­guards and insuring against bias 011 thepart of the investigator. But the par­anl0unt objection to a genetic interpre­tation of the evidence was the objectionthat the whole research strategy wasfaulty. The results of these studies, itwas held, were just as compatible withtransn1ission of schizophrenia throughthe social environment as with ttans­nlission through genes. T'he closer the

The author is assistant professor of psychiatryat the University of Iowa ~1edical School, IowaCity.

249

genetic relationship, the closer the so­cial relationship. Were genes or wasnoxious social learning responsible forthe familial clustering of schizophrenia?

Recently, several studies have beenaimed at closing those methodologicaland conceptual gaps. In these newerstudies diagnoses either were made byraters who did not know the geneticbackground of the subjects or weretaken unchanged from medical records.Care was taken to remove samplingbiases, and, most importantly, controlgroups were used. The strategy per­mitted separation of the effects of genesfrom the effects of social environmentthrough the use, as subjects, of childrenreared in adoptive or foster hOlnes.

The results of one such study areshown in Table 1 (3). The experi­mental subjects were individuals bornto schizophrenic mothers, and the con­trols were individuals born to parentswho had no record of psychiatric dis­turbance. The nlelnbers of both groupshad been permanently separated fromtheir biological mothers in the firstmonth of life and reared mainly infoster or adoptive homes. The subjects,as adults, were assessed through psy­chiatric interviews and review of everyavailable record-for example, school,police, Veterans Administration, andmedical-and then evaluated by ateam of clinicians. The significant ex­cess of schizophrenia found amongthose subjects whose biological motherswere schizophrenic seelns impossible toexplain except on a genetic basis. More­over, among those sanle experimentalsubjects, and thus also linked to schizo­phrenia by the evidence, was an evengreater excess of various apparentlynonschizophrenic disorders. The latterfinding, which is reflected in nearlyevery entry in Table 1, is a central con­cern throughout this article.

The preliminary results from a verysinlilar study which stressed exenlplaryinvestigative safeguards were much thesame. Rosenthal et al. (4) reported thatbiological children of schizophrenicsreared in adoptive honles exhibited"schizophrenic spectrum" disorders insignificant excess over sinli1arly rearedcontrols. The "schizophrenic spectrum"-an expression coined in a quite rea­sonable attempt to find a term thatwould encompass the various disorders'seen anl0ng biological relatives of schiz­ophrenics-included schizophrenia, pos~sible schizophrenia, borderline states,certain paranoid disorders, schizoid dis­orders, and the condition ,known asinadequate personality.

250

Karlsson (5), as one result of hisstudy of schizophrenia in Icel andicfamilies, found that 6 of 29 persons,sonle of them siblings, born to a schiz­ophrenic parent but reared in fosterhomes developed schizophrenia. Noneof their 28 foster sibs who were rearedin the same homes developed schizo­phrenia. This difference, too, is sig­nificant. Karlsson did not ascertain anydisorders other than typical schizo­phrenia alnong his subjects.

In two ingeniously designed researchprojects, adopted individuals served asthe starting point. Wender et al. (6)studied the biological and adoptive par­ents of ten adopted schizophrenics andthe adoptive parents of ten normalpersons. The biological parents of theschizophrenics were found to exhibitsignificantly more psychopathology thaneither group of adoptive parents. Ina similar but wider-ranging study con­ducted by Kety et al. (7), psychopathol­ogy, again reported as "schizophrenicspectrunl" disorders, was found to beconcentrated in significant excess amongthe biological relatives of adopted schiz­ophrenics. The adoptive families ofschizophrenics were indistinguishablefrom the adoptive and biological fam­ilies of adopted controls. Since thepsychopathology found in these studieswas significantly greater alnong thegroup of biological relatives of theschizophrenic probands than anl0ng theadoptive relatives who actually livedwith thenl, this evidence too stronglyfavors genetic over social transmissionof schizophrenia.

The results of the studies of adoptedand foster children-results which arestrikingly consisterit from study tostudy, considering the vagaries of re­search in this area-present seeminglyinsurm"ountable difficulties for adherentsof environmental theories of schizo­phrenia. The evidence must surely conl­pel acknowledgnlent of a genetic con­tribution to schizophrenia, and prob­ably to related disorders as well. To gofurther, however, requires infornlationon other types of genetic relationshipsand larger nunlbers of subjects. Hap­pily, the older family studies can nowmeet these needs. For perhaps the mostimportant .contribution of the recentstudies of adopted and foster childrenis the fact that they have confirmed theresults of the older studies in all ma­terial respects. The familial clusteringof psychopathology that had been docu­mented in such detail has been linkedto one critical variable, a genetic rela­tionship to schizophrenia.

The Schizoid

The presence of so much psycho­pathology other than typical schizo­phrenia among relatives of schizo­phrenics was first noticed by physicianson visiting days in the earliest asylulns.Isaac Ray, writing in 1863,'gave a gooddescription (8). Because the relatives'disabilities resenlbled schizophrenia, in­vestigators associated with the Munichschool called these disabilities "schizoid"(schizophrenic-like) . Describing theschizoid individual, delimiting schizoidfrom psychiatric and general popula­tions, and placing the schizoid in relationto the schizophrenic were central con­cerns of the psychiatry of that day.After perhaps the longest detour inthe modern history of science, we havecome full circle in returning to thesame concerns. Meanwhile, problenls ofnomenclature have developed.

To me, "schizoid" and "schizophrenicspectrum" seem to denote precisely thesame disabilities, except that the latterterm also includes schizophrenia. Oneconsideration that may have led Kety(7), Rosenthal (4), Wender (6), andtheir co-workers to coin the new ternlis the obvious danger of confusing"schizoid" with "schizoid personality."The latter term, a diagnosis in theAmerican Psychiatric Association andWorld Health Organization nomencla­ture, although descended from descrip­tions of the abnornlal relatives of schiz­ophrenics, has evolved and changedin nleaning so that it is no 10ngeI: ap­plicable to most of those relatives. Forexample, it was not often applied torelatives of schizophrenics by the ratingclinicians in the studies of adopted andfoster children. But other diagnosescurrently considered applicable to suchindividuals also fit these relatives inl­perfectly, so no forlnal categorizationis now available. Because of a centraltrait of the schizoid-his. clinical re­semblance to the schizophrenic-andbecause of the desirability of nlaintain­ing continuity with older studies, Iuse the tern1 "schizoid" as a nanle forthe schizophrenic-like disabilities seenin relatives of schizophrenics, or forthe individual manifesting such dis­abilities.

Nearly all observers of the schizoidhave noted his clinical resemblance tothe schizophrenic, but clinical criteriaadequate to reliably distinguish theschizoid from members of a generalor a psychiatric population or evenfron1 other kinds of abnormal personswith a coincidental genealogical con-

SCIENCE, VOL. 167

Table 1. Results of a study of individuals born to schizophrenic mothers and reared inadoptive or foster homes, and of controls born to normal parents and similarly reared.

Table 2. Explicatives used by relatives of schizophrenics in describing their schizoid rela­tives. [After Slater (13)].

* The MHSRS is a global rating of psychopathology moving froln 0 to 100 with decreasing psycho­pathology. Total group mean, 72.8; S.D., 18.4. t One mental defective was also schizophrenic;another had antisocial personality. :1: Considerable duplication occurs in the entries under"neurotic personality disorder"; this designation includes subjects diagnosed as having varioustypes of personality disorder' and neurosis whose psychiatric <liability was judged to he a significanthandicap. § Group 1, highest social class; group 7, lowest.

Paranoid eccentricities: suspicious, sensitive, sullen, touchy, grouchy, morose, resentful,unforgiving, difficult, quarrelsome, self-conscious, jealous, litigious, critical, and others.

Eccentricities: giggly, opinionated, pedantic, narrow-minded, meticulous, obstinate, humor­less, rigid, little-minded, spiritualists, and many others.

Lack oj feeling: passive, cruel, calculating, placid, hard and stingy, unsympathetic, cold,withdrawn, little-feeling, and others.

Reserve: shy, serious,haughty, snobbish, studious, unforthcoming, taciturn, unsociable,seeks solitude, and so on.

Anergic: dependent, tired, slack, unreliable, subservient, and so on.

.054

.006

.021

0.0006.024.052.017.052

Exact probability(Fisher's test)

84.55.4

94.011.671

69

473035.82265.2

549

13

11112

721

Experi­luental

COJnpared with 3 percent of the generalpopulation. l:'he expected reciprocalrelationship, an excess of schizophrenicsal1long Inental defectives or their rela­tives, was found by Penrose (15) andBook (16) alllong nlental defectives butnot by Reed and Reed (17) in theirInonumental survey of the relatives of11lental defectives. Also, Kallmann founda nluch higher rate of Inental deficiency(10.8 percent) anlong relatives of sinl­pIe schizophrenics, where there is aclinical conlmonality of sorts, thanalTIOng relatives of other subtypes inthe Kraepelinian classification. The evi­dence for or against an association be­t.ween schizophrenia and mental de­ficiency is inconclusive, land nloredata are needed before the nlatter canbe decided.

Obviously there is 111uch yet to belearned before we can describe anddelinl it schizoidia. H:owever, the samething can be said of schizophrenia it­self, and in this regard study of theschizoid Blay lighten sonle dark corners.

215

2.17

503336.31980.loo27

14.24.7

103.712.484

74

Control

llladc, his concept of the schizoid is ofcritical inlportance (11). From his de­scription (11, p. 102) it is clear thathe relied heavily on the schizoid's clin­ical reselnblance to the schizophrenic.Kalhl1ann regarded the distinguishingfeatures of the schizoid to be the "fun~

dall1ental sylnptonls of schizophrenia inthe nlilder fornl of characterologicalabnonnalities . . . dOlninating the per­sonality of the individual in question."Kallnlann also looked analytically attraits other than those obviously asso­ciated with schizophrenia or schizoidiathat seenled to occur in excess anlongrelatives of schizophrenics, with theainl of including or excluding then1fron1 the group of schizoid traits. Onvarious grounds he excluded all thetraits that he considered.

One of the traits which Kallnlannconsidered and rejected, nlental de­ficiency, perhaps deserves another look.About 6 to 10 percent of schizophrenics(see 14) and their first-degree relatives(see 3, 11) are ll1cntally subnormal, as

N ulnber of subjectsN unlbcr of malesAge, nlcan (years)N'umber adoptedMHSRS, means*N'umber with schizophreniaNUlllber with mental deficiency (LQ. < 70) tN'unlber with antisocial personalitiesNUlnber with neurotic personality disorder~(

Persons spending more than 1 year inpenal or psychiatric institutionNUlnber'rotal years incarcerated

NUlubcr of felonsN Ulllber serving in armed forcesNU111ber discharged from armed forces on

psychiatric or behavioral groundsSocial group, first home, nlean ~

Social group, present, mean ~

I.Q., 111eanYears in school, 111eanN'uluber of children, totalN'umber of divorces, totalNU111ber never married, > 30 years of age

nection to a schizophrenic are nlostinlperfect (9). Though unsatisfactory,the only means of identifying many­perhaps most-schizoids remains gen­ealogical, and a clinical understandingof the schizoid can best be gained byreading descriptions of abnornlal rela­tives of schizophrenics (see 10-13 forgood examples). The circularity thusintroduced is regrettable but inescap­able. The schizoid exists, and he sonlC­tinlCS shows as much itnpairnlcnt psy­chiatrically as a typical schizophrenic.

Several problenlatical behaviors havebeen associated with the schizoid.Aillong nlales, antisocial behavior hasbeen found cOlnlnonly enough to war­rant the older subdesignation "schi'zoidpsychopath." Entries in the· police rec­ords of the schizoid psychopaths in D1Y

study reflectcd impulsive, seelningly il­logical crime such as arson, unreasoningassault, and poorly planned theft (3).Social isolation, heavy intake of alcohol,and sexual deviance have been notedfrequently. Other schizoids, both nlaleand female, have been described as ec­centric, suspicion-ridden recluses. T'he111ai n disability of still other schizoids,lTIOStly feln,ales, has been found to heincapacitating attacks of panic or un­reasoning fear in response to ordinarysocial challenges.

On a more technical level the reseln­blance to schizophrenia is, more ap­parent. Rigidity of thinking, bluntingof affect, anhedonia, exquisite sensi­tivity, suspiciousness, and a relativepoverty of ideas-in variable conlbina­tions and intensities-characterize boththe schizoid and the schizophrenic,through such characteristics ,are lesspro111inent in the former. Though schiz­oids do not show a well-Inarked thoughtdisorder, delusions, and hallucinations~

descriptions of sonle of the behaviorallapses of schizoids, especially the schiz­oid psychopath, are bizarre enough tosuggest micropsychotic episodes.

SI ater took a different approach. He1isted a series of explicatives, partiallyreproduced in Table 2, used by rela­tives of schizophrenics when describingtheir abnormal but nonschizophrenicrelatives (13). Slater went on to say( 13, p. 83) that "the same or similarwords or phrases occur in descriptionsof abnormal personalities from otherkinds of families, but much less fre­quently, not in such concentrated fornl,and they are usually submerged bydescriptions of a very different tone."

Because K'alhnann's investigations ofthe families of schizophrenics were byfar the most extensive that have been

16 JANUARY 1970 251

Table 3. Data on monozygotic twins of schizophrenics.

::: Investigators' diagnoses: ? schizophrenia, schizophreniform, transient schizophrenia, reactive psychosis,borderline state, schizoid, suicide, psychopathic, neurosis, and variations of these diagnoses. -;- FromShields, Gotteslnan, and Slater (44).

Other Nonnal,

Investigator Pairs Schizophrenia significant or mild(No.) (No.) abnonnality* abnornlality

(No.) (No.)

Essen-Moller (19) 9 0 8 1Slater (13) 37 18 11 8Tienari (41) 16 1 12 3Kringlen (12) 45 14 17 14Inouye (42) 53 20 29 4GottesnUln and Shields (23) 24 10 8 6Kalhnann -;- (43) 174 103 62 9

Totals 358 166 (46.4% ) 147 (41.1%) 45 (12.6% )

Schizophrenia is defined operationally,not etiologically. It is the clinician whodeternlines whether schizophrenia ispresent. But of course the limits ofthe clinical entity lnay not correspondto those of the ,etiological entity. Infact the linking of schizoidia to schiz­ophrenia by genetic evidence raisesserious questions about the etiologicalreality of the clinical definition of schiz­ophrenia. There has always been a fuz­zy border about schizophrenia alongwhich several nanled entities, includingabortive, ambulatory, borderline, latent,pseudoneurotic, pseudopsychopathic,and reactive schizophrenia and the"schizotype" of Meehl (18) haveseemed to lie. These terms nlay bestbe viewed as attelnpts to cope withan operationally defined border be-

. tween schizoidia and schizophrenia thatis clinically imprecise because it is bio­logically unreal.

Quantitative Aspects

Given a schizophrenic who has amonozygotic twin, the enlpirical prob­ability that his twin will also be schiz­ophrenic has been found to be about0.46 (Table 3). Most of the remaining54 percent of monozygotic twins ofschizophrenics have also been foundto be abnornlal. From clinical descrip­tions included in five studies (12, 13,19-21 ) it appears that nearly all ofthe abnormal though nonschizophrenicco-twins were schizoid. Overall, onlyabout 13 percent of the monozygotictwins of schizophrenics have been re­garded as normal or nearly norlnal, and,because most of the errors inherent inthis sort of research tend to increasethe proportion of apparent norlnals,this is surely an overestimate. But,while a critic could easily quibble aboutany of the proportions in l"'able 3, acrude but critical conclusion is in­escapable: monozygotic twins of schiz-

252

ophrenics are about as likely to beschizoid as schizophrenic. What then isinherited? These considerations ledEssen-Moller (19) to regard schizoidiaas the basic inherited trait, and I(rin­glen, in a careful and sensitive anal­ysis of twin research, including hisown nlajor study, seelns to have reacheda similar conclusion, although he re­garded the predisposition as less spe­cific (12). At the very least a primafacie case has been lnade for con­sidering the whole group of schizoidand schizophrenic disorders as alterna­tive expressions of a single genotype.Moreover, because monozygotic twinsare identical genetically, there is pre­sUlnptive evidence that the range ofvariability within pairs can in principlebe accounted for by environmentalfactors. The genes allow a range ofoutcolnes.

A critical point to be established isthe proportion of schizoids or schizo­phrenics anlong the first-degree relatives(parents, sibs, children) of schizo­phrenics. Table 4 gives Kallnlann's re­sults. No one else has investigated sonlany relatives of schizophrenics, andfew others have' conducted field studiesintensive enough to identify schizoids.l"'he more intensive lnodern studieshave tended to show sOlnewhat largerproportions of afflicted relatives (3, 10,22). So did Slater among dizygotictwins of schizophrenics (13). The pro­portions found by Gottesman andShields (23) and by Odegard (24) weresonlewhat snlaller. Kallmann's valuesmay be taken as fair average estimatesof the proportion of schizoids or schiz­ophrenics anl0ng first-degree relativesof schizophrenics.

Table 4 also shows the results offour studies of the children of twoschizophrenics. An estiInated 66 per­cent of the children of these matingswere schizoid or schizophrenic, again,this is surely an underestimate becausethe subjects were still quite young. The

results of one such study, that ofLewis (25), was not included. Lewisdid not give ages, and he stated thathis follow-up was incomplete. Rosenthalhas recently reviewed these studies (26).

An inlportant uilknown lnust nowbe considered. There is no adequateestinlate of the proportion of schizoidsin the general population. Then, isthe clusteriIig of schizoids anlong rela­tives of schizophrenics greater thannlight occur by chance? Although theproportion of schizoids. found in fanl­ilies of schizophrenics is surely greaterthan that expected by even the mostpessimistic observer of the general pop­ulation, a better answer is that neitherthe relatives of other kinds of psychi­atric patients nor the controls used inpsychiatric studies have been found tobe afflicted in' significant numbers withdisorders of a schizoid character orwith any kind of behavioral disorderto the extent seen in relatives of schiz­ophrenics. Furth~r evidence-the smallproportion of schizoids found amongdescendants of normal relatives ofschizophrenics-is discussed below.

While the lack of data for the gen­eral population and the related lack ofdata for the families of schizoid pro­bands preclude estimates of gene fre­quency, it should be noted that schizoiddisorders surely afflict a large propor­tion of the population. With only iso­lated exceptions, schizophrenia afflictsabout 1 percent of any population. /Ifeach schizophrenic has five living first­degree relatives (about the number inKallmann's study), a simple extrapola­tion yields an estimate of 4 percentfor the proportion of schizoids plusschizophrenics in the general popula­tion. This crude estimate can onlymake the point that any population, and

-especially any psychiatric population(persons identified because they came topsychiatric clinics or hospitals), is like­ly to contain large numbers of schiz­oids. One of the most neutral iInpli­cations of this conclusion has an ob­vious application to the choosing ofcontrol groups for research in schiz­ophrenia. '

Genetic Hypothesis

The nlost parsinlonious explanationof the data is given by the hypothesisthat a defect in a single autosonlalgene accounts for the genetic contri­bution to both schizoid and schizo­phrenic disease (the "donlinance hy­pothesisH

). By including schizoid dis-

SCIENCE, VOL. 167

Table 4. Percentages of first-degree relatives found to be schizophrenic or schizoid.

I~OI

MZtwins

49.045.844.0

66.1

Total: schizoidplus schizophrenic

(%)

32.631.534.8

32.2

Schizoid(%)

§ From Kallmann (11), Kahn (45),

0.61

in schizoidia-schizophrenia. First of all,there remain small deviations fromthe theoretical expectations under thedominance hypothesis, deviations whichhave been cited by Shields (2). Thesemainly take the forn1 of a greater re­semblance between relatives than canbe explained by sim.ple dominance. Forexample, the monozygotic twin of aseverely afflicted individual is morelikely to be schizophrenic than thetwin of a mildly affected individual.If only a single gene \vere involvedone would expect the risk of schizo­phrenia for a monozygotic twin of anyschizophrenic to be equal to that ofany other. Likewise, the larger theproportion of schizophrenic relatives is,

16.414.3

9.2

33.9

jo.a1st degree Children of tworelatives schizophrenics

Schizophrenia:::(%)

:(: From (43),

0.4

100011912741

171

NORMAL

Numberof individuals

0.2

Degree of genetic relationship

1.0---------------------~

Fig. 1. Observed and expected proportion of schizoids and schizophrenics.

253

0.4

0.2

0.6

0.8-0Q)+-'UQ.)

'+­'+-menC1>>....,COQ)'-

"'t­O

enc::o

........'-oC-O'--a..

But research 111ust proceed fronl hy­potheses based on present understand~

ing. From that viewpoint, and forpractical purposes, it is not at all un­reasonable to proceed on the workingassumption that IUOSt schizoidia-schizo­phrenia is associated with defects in asingle basic biochemical or physiolog­ical pathway, translnitted by a singlemode of inheritance. It matters littlethat new research will no doubt turnup complexities that cannot even beimagined today.

Apart from insights gained frolnanal~gies to other genetic diseases, thereare factual reasons for expecting thatmany elements in addition to a singleDlain gene go into the mix that results

Relationship

* Age-corrected rates. t From (11).Schulz (46), and Elsasser (47).

Childreni'Siblings~

Parents~:

Children of twoschizophrenics§

ease (schizoidia), this hypothesis ex­tends that of Slater (27). The view thatschizoidia and schizophrenia are asingle disease genetically is supportedby their clinical similarity and is virtual­ly required by the finding that the dis­orders occur with equal probability inlTIonozygotic twins of schizophrenics.Further 8upport for the hypothesis ispresented in Fig. 1. The proportions ofaffected first-degree relatives fit reason­ably well with the theoretical propor­tions expected under the donlinancehypothesis.

Kallmann presented SOUle data 011second-degree relatives (11). Among822 grandchildren of his schizophrenicprobands he found 4.3 percent to beschizophrenic and 22.8 percent to beschizoid. The corresponding rates fornephews and nieces were considerablylower (3.9 and 6.2 percent). However,Kallmann pointed out that the normalsibs of his schizophrenia probands con­tributed many more nephews and niecesthan the schizoid or schizophrenic sibsdid. While the total of 27.1 percent foraffected grandchildren is certainly closeto the 25 percent expected under thedOlTIinance hypothesis, the proportionsof affected nephews and nieces nlay ornlay not be compatible \vith that hy­pothesis.

The segregation of schizophrenia andschizoidia within families fits well withthe dominance hypothesis. In I(all­nlann's study, which included threegenerations, the normal children of hisschizophrenic proballds, produced fewschizophrenic or schizoid children (1.8and 2.6 percent, respectively), no nlorethan might be expected in a generalpopulation. This is in contrast to thecorresponding values of 13.7 and 33.4percent for the ,children of the schizoidor schizophrenic children of Kallnlann'sschizophrenic probands (11).

The matter cannot be so sinlple, ofcourse. The mechanisms involved ina disease like schizoidia-schizophreniawill surely be found to be extrenlelyconlplex. Even phenylketonuria, whichonly a few years ago provided a pro­totype of rigorous simplicity for be­havioral genetics, has. turned out tobe enormously complicated by second­ary biochemical effects and by other,1110Stly unknown, factors (28). Hetero­genity is also likely. Probably the mostcOll1pletely known genetic disease inhU111ans, glucose-6-phosphate dehydro­genase deficiency, occurs in at least 18variants, each one presumably due toan amino acid substitution at a differ­ent place in the sanle enzyme (29).

16 JANUARY 1970

the greater is the risk of schizophreniafor any given individual. Another sortof problem is that of accounting for thevariability seen among schizophrenics;this becomes more difficult when schiz­oids are included. Although there areno grounds for expecting 'any partic­ular degree of resemblance between af­fected persons, it has often been arguedthat, if only one gene were involved,the range of obs~rvable phenotypesshould be slnaller than is the case. Andthe persistance of schizophrenia pre­sents 'a problem. Before the introduc­tion of antipsychotic drugs, schizo­phrenics reproduced at a rate 30 per­cent lower (16), and slchizoids at arate 22 percent lower (11), than therate for the general population. Suchreproductive deficits should have low­ered the rates of occurrence of a dis­order due to a main gene of largeeffect far below the presently observedrates for schizophrenia.

Attempts to account for such findingshave led to widespread espousal ofpolygenic theories of schizophrenia (12,24, 30). As Gottesman 'and Shieldshave pointed out (31), the facts areexplained adequately by polygenic the­ory. Most polygenic theorists have re­garded schizophrenia as a thresholdtrait. But clinically schizoidia and schiz­ophrenia seem to form a continuumof psychopathology, much as first de­scribed by Kretschmer (32). If thereis a threshold it probably falls betweenthe schizoid and the normal condition,but it seenlS that any such "threshold"is as likely to be a function of lacko.f diagnostic precision as a function ofthe disease. It is not necessary to con­sider other aspects of the polygenicargument here. Known modifiers of thephenotypic expression of the diseasepoint toward plausible solutions of theproblems encountered by the dominancehypothesis and toward resolution of theapparent differences between main-geneand polygenic theories.

Modifying Factors

One class of modifiers must be en­vironmental events in the broadest sense-events occurring from conception on­ward that produce some change in theorganism. The nature-nurture dilemmais unreal. It is change in the environ­ment of the cell that induces changein the genetically mediated .metabolicsystems of the cell. The functional stateof the cell is a result of the interplay

254

of these deternlinants. But realizationthat phenotypic traits depend on inter­action between gene and environmentimposes conditions on research aimedat assessing the environment contribu­tion. Genes function within cells. Theyinteract with chemical, thermal, or otherphysical events and not with the ab­stractions ("stress," for example) thattoo often have passed for environmentaldata. The ultimate questions implicitin the concept of ge~e-environmentin­teraction are, for example: How doesa noxious learning experience alter theenvironment of the ,cell? What responseis elicited from the genetic program ofthe cell? How is the later operation ofthe cell modified? Of course, such ques­tions cannot be approached directlytoday. But unless the environmentalcontribution is too variable from caseto case to allow generalization, it shouldbe possible to build up a series of as­sociations between environment and be­havior that would point toward theenvironmental events that enter intothe gene-environment interaction. Thecritical requirement is that such asso­ciations be potentially translatable intoevents that occur at the level of thegene. Despite all the research that hasbeen done on the effects of environmenton the development of schizophrenia,and despite the scope for environmentalfiactors demonstrated by the differencesbetween members of monozygotic twinpairs, practically no associations thatmeet this requirement have been estab­lished. Clinicians have learned to pre­dict the effects of environmental fea­tures on their patients, but it is difficulttp see any etiological clues in this bodyof experience. On general clinicalgrounds it makes sense to continue tostudy the effects of environmentallystinlulated autonomic and endocrineresponses. An association between lowerbirth weight and the development ofschizophrenia in one member of amonozygotic twin pair has been re­ported (21), but it must be quite im­perfect in view of the failure of otherinvestigators to confirm it (12, 23).Perhaps differences in autonomic re­sponses among children of schizo­phrenics that were described in a pre­liminary report from a wide-rangingprospective study (33) are the mostpromising associations so far defined.Alnlost everything remains to· be done.

A second class of modifiers consistsof complex traits that have been linkedto schizophrenia by decades of empir­ical research. Somatotype has been

found by several investigators to beassociated with major modification ofschizophrenia. Mesomorphs are un­derrepresented among schizophrenics,and especially underrepresented amongschizophrenics younger than 25. Ecto­morphs are correspondingly overrepre­sented. Schizophrenic mesomorphs arepredominantly paranoid and have ashorter mean period of hospitalizationthat other schizophrenics. Parnell (34),who has reviewed the subject and con­tributed his own data, found all theseassociations to be statistically significant.A relation between intelligence and theprognosis in schizophrenia is wellknown: the higher the intelligence thebetter the prognosis. But higher intelli­gence may also affect the expression ofschizophrenia. Lane and Albee (35)found that the I.Q. of children who laterbecame schizophrenic was seven pointslower than that of their si~lings whoremained nonschizophrenic. There area host of other established associationsbetween complex traits and schizo­phrenia-for example, patterns of auto­nomic nervous system reactivity, im­munological phenomena, resistance tocertain chronic diseases, and toleranceof wound shock. Some such traits ap­pear to be only oddities, given our pres­ent knowledge; others are known to belinked to favorable or unfavorable prog­nosis in schizophrenia, and still othersare known only to be more frequent orinfrequent among schizophrenics. Sev­eral reviews of thes,e findings are avail­able (36).

The large number of such complextraits and the magnitude of the modi­fication of schizophrenia associated withsome of them nlust mean that they havea significant role in the ecology of thedisease. F or one thing, they suggesta plausible solution to the puzzle posedby the persistence of high rates of schiz­ophrenia. Sir Julian Huxley et ale (37)postulated that the gene responsible forschizophrenia conferred sufficient phys­iological or reproductive advantages tonlaintain a balanced polymorphism.They listed several physiological tr~its

found in schizophrenics that could bedue to pleiotropism. Although the num­ber of traits listed seems large, wide­spread pleiotropism might result fronla mutation at a regulatory locus (38).But many modifying traits are clearlynot due to pleiotropism, and some ofthose-particularly differences in soma­totype and intelligence-which demon­strably affect the outcome in schizo­phrenia must have conferred general

SCIENCE, VOL. 167

biological advantages through nluch ofman's history as well. In either event,schizophrenics possessing advantageoustraits would be expected to reproduceat relatively higher rates than those notpossessing such traits. Over tinle, theevolutionary process would, theoretical­ly, act to establish sets of favorabletraits that, on the average, would tendto accompany schizophrenia. Theoryaside, the popular association betweengenius and insanity, thought to be er­roneous by Kallmann, was given SOOle

substance by Karlsson's finding thatcreative achievements and schizophreniaoccurred in the same faolily lines (39).I reported a similar impression; how­ever, the evidence was not gatheredsystematically (40). Althoughthe prob­lenl posed by the persistence of schiz­ophrenia relnains theoretical and un­solved, further exploration of modifyingtraits provides as likely a path as anyother now in view toward solution ofthe puzzle.

Modifying traits also suggest an ap­proach to the problenl of deviationsfrom strict expectations under the dOIni­nance hypothesis. As pointed out above,polygenic theory can account for suchdeviations. But traits like S0111atotypeand intelligence are themselves alnlostcertainly polygenic. Polygenic n10difiersof a single n1ain gene explain the samefacts, and indeed would yield the sanlC111athematical results as sin1ple ·additivepolygenic theory per see A n1ultitude ofgenes summating to produce schizo­phrenia directly or a single main geneplus groups of genes summating toproduce n10difying traits account equal­ly well for findings such as the tendencyof nlonozygotic twins to be concordantwith respect to severity of illness.

Conclusion

A main gene of large effect nlodifiedby multiple factors, including polygenictraits, suggests a number _of testablehypotheses. Biochemical or other effectsof a main gene should be present inschizoids as well as in schizophrenics.In family studies, the critical test of theplace of the schizoid would be his re­productive performance in matings withnorn1al individuals; 50 percent of theoffspring of such matings should beschizoid or schizophrenic. However,po'ygenic modifiers should, on the aver­age, maintain lesser degrees of dis­ability in particular families. Thus,'schizoid parents should have fewer

16 JANUARY 1970

schizophrenic but n10re schizoid chil­dren than-schizophrenic parents. Thereis' incomplete support for this conten­tion in Kallmann's study (11) of thegrandchildren of his schizophrenic sub­jects: the schizoid children of his schiz­ophrenic probands had nlore schizoidand fewer schizophrenic children thantheir schizophrenic siblings, but nlen1­bers of the third generation, the grand­children of the probands, \vere tooyoung to yield decisive evidence. Alongthe sanle Jines, it would be expectedthat nearly all schizophrenics shouldhave at least one schizoid or schizo­phrenic parent. -Although the work ofKalln1ann and the intensive familystudies of Alanen (10) and IJdz (22)support this expectation, luore rigorousevidence is needed. The traits, that fa­vorably nl0dify schizophrenia shouldbe nlore apparent alnong schizoid thanamong schizophrenic relatives of schiz­ophrenics. One would hypothesize, forexalnplc, that the nlore mesomorphicor m.ore intelligent an10ng the childrenof schizophrenics would tend to haveless severe illnesses and to have morechildren than the less mesomorphic orless intelligent. These hypotheses, andmany n10re that are ilnplicit in the pre­ceding discussion, constitute a signifi­cant refinement of the genetic hypoth­eses so far explored in schizophrenia.

l'he inlportance of genetic factorsin the development of schizophreniahas by now been established beyondreasonable dispute, although it is clearthat environment too plays its etiologicrole. The results of recent researchhave refocused attention on schizoiddisorders, a term applied to psychiatricdisorders reselnbling schizophreniawhich afflict relatives of schizophrenics.The many conceptual and researchproblems presented by the schizoid areconsidered.

Schizoids and schizophrenics" occurwith about the saIne frequency amongmonozygotic twins of schizophrenics.About 45 percent of the _sibs, parents,and children of a schizophrenic areschizoid or schizophrenics, as are about66 percent of the children of two schiz­ophrenics. Fron1 the known risk ofschizophrenia for the population as awhole, it is estimated that at least 4percent of the general population willbe afflicted with schizoid-schizophrenicdisease.

Since lnonozygotic twins are identicalgenetically, it appears that the samegenotype is compatible with either schiz­ophrenic or schizoid disease. The pro­portions of affected first-degree rela­tives and the segregation of affectedindividuals -within falnilies closely ap­proximate theoretical expectations basedon the hypothesis of a defect in asingle autosomal dominant gene. How­ever, modifying traits play a significantrole; this is 'discussed and integratedinto the main genetic hYP9thesis. Em­phasis is placed on hypotheses testableby future research.

References and Notes

L E. Slater, in The Transmission of Schizophre­nia, D. Rosenthal and S. Kety. Eds. (Perga­mon, Oxford, 1968); D. Rosenthal, in TheOrigins 0/ Schizophrenia. J. Romano, Ed.(Excerpta Medica Foundation, New York,1967).

·2. J. Shields, in The TranS1'11ission of Schizophre­nia, D. Rosenthal and S. Kety, Eds. (Perga­mon, Oxford, 1968).

3. L. L. Heston t Brit. J. Psychiat. 112, 819(1966).

4. D. Rosenthal, P. Wender, S. Kety, F. Schul-singer, L. Ostergard, J. 'VeIner, in The Trans­mission of Schizophrenia, D. Rosenthal andS. Kety, Eds. (Pergamon,' Oxford, 1968).

5. J. Karlsson, The Biological Basis of Schizo­phrenia (Thomas, Springfield, Ill.. 1966).

6. P. Wender, D. Rosenthal, S. Kety, in TheTransmission of Schizophrenia, D. Rosenthaland S. Kety, Eds. (Pergamon, Oxford, 1968).

7. S. Kety, ·n.Rosenthal, F. Schulsinger, P.Wender, ibid.

8. I. Ray, Mental llygiene (Hafner, New York,new ed., 1968).

9.K. Planansky, J•.Nerv. Ment. Dis. 142, 318(1966); E. Essen-l\.1011er, Alschl'. Psychiat.Neuro1. 112, 258 (1946).

10. Y. 0.· Alanen, Acta Psychiat. Scand., SuppleNo. 189 (1966).

1L F. J. Kallmann, The Genetics of Schizophre­nia (Augustin, New York, 1938).

12. Kringlen, "IIeredity and Environment inFunctional Psychoses," Norwegian

Monogr. Med. Sci. Univ. Oslo (1967).13. E. Slater, with the assistance of J. Shields,

"Psychotic and Neurotic Illness in· Twins,"Med. Res. COU1lC~ (Great Brit.) Spec. Rep.Sere No. 278 (1953).

14. B. Jlallgren and T. Sjogren, Acta Psychiat.Neurol. Scand., Suppl. No. 140 (1959).

15. L. S. Penrose, A Clinical and Genetic Studyof 1280 Cases o/Mental Defect (Her Maj­esty's Stationery Office, London, 1938).

16. J. A. Book, Acta Genet. 4, 1 (1953).17. E. W. Reed and S. C. Reed, Mental Retarda­

tion: A F amity Study (Saunders, Philadelphia,1965).

18. P. E. Meehl, Amer. Psychol. 17, 827 (1962).19. E. Essen-Moller, Acta Psychiat., Supple No.

23 (1941.).20. ·P. Tienari, ActaPsychiat., Supple No. 171

(1963).21. W. Pollin, J. R. Stabenau, J. Tupin, Psychia­

try 28, 60 (1965).22. T. Lidz, S. Fleck, A. Cornelison, Schizo­

phrenia and the Family (International Uni­versities Press, New York, 1966).

23. I. I. Gotteslnan and J. Shields, Brit. J.Psychiat. 112, 809 (1966).

24. O. Odegard, Acta Psychiat., Supple No. 169(1963), p. 94.

25. A. Lewis, Acta Genet. 7, 349 (1957).26. D. Rosenthal, J. Psychiat. Res. 4, 169 (1966).27. E. Slater, Acta Genet. 8, 50 (1958).28. C. Johnson, J. Iowa Med. Soc. 59, 27 (1968).29. H. Harris, Brit. Med. J. 2, 135 (1968).30. D. Rosenthal, The Genain Quadruplets (Basic

Books, New York, 1963).31. 1. Gottesman and J. Shields, Proc. Nat. Acad.

Sci. U.S. 58, 199 (1967).32. E. Kretschmer, Physique and Character (W.

Sprott, Trans.) (Paul, Trench and Trubner,London, 1925).

255

Casualties of OUf Time

Social and technological changes produce new sources

of death and disability which raise public issues.

33. S. Mednick and F. Schulsinger,· in TheTransmission of Schizophrenia, D. Rosenthaland S. Kety, Eds. (Pergamon, Oxford, 1968).

34. R. Parnell, Behavior and Physique (Arnold,London, 1958).

35. E. A. Lane and G. W. Albee, Amer. J.Orthopsychiat. 35, 747 (1965).

36. E. Gellhom and G. Loofbourrow, Emotionsand Emotional Disorders: A Neurophysiologi­cal Study (Harper & Row, New York,1963); W. Ross, J. Hay,M. McDowal,Psychosom. Med. 12, 170 (1950); P. Hustonand M. Pepemik, in Schizophrenia, A Review0/ the SJ'ndrome, L. Bellak, Ed. (Logos, NewYork t 1958); C. Rosenbaum t J. N'erv. Ment.

"It is changes that are chiefly respon­sible for diseases, especially the greatestchanges, the violent alterations both inthe seasons and in other things."

-HIP:POCRATES

Violent alterations in the humanenvironment have occurred at an in­creasing rate since the beginning ofthe Industrial Revolution. From thelate 18th century, dislocation fromthe land, turbulent crowding in grow­ing cities, and the econonlic depriva­tions of factory life aff.ected thehealth of the people. Old diseases liketuberculosis flared up, and newsources of death and disability devel­oped, such as the industrial injurieswhich were incurred 'by inexperi­enced hands attempting to masternew machinery. Great changes, vis­ible in a man's lifetime, gave motivesfor new laws and institutions. Socialhygiene, with tardy assistance fromtherapeutic medicine, broughteffec­tive lneasures to bear on the newhealth problems, while hospitals, asy­lunls, and other institutions were es­tablished to take the place of the nowobsolete welfare systems of farm andvillage. Economic and technologicalchanges thus produced specific newkinds of casualties, along with newr.esources for coping with them.

In the present century, the rate of

256

Dis. 146, 103 (1968); H. Freeman, in Schizo­phrenia, A Review of the Syndrome, L. Bel­lak, Ed. (Logos, New York, 1958); L. Rees,in Schizophrenia: Somatic. Aspects, D. Rich­ter, Ed. (Macmillan, New York, 1957).

37. J. Huxley, E. Mayr, H. Osmond, A. Hoffer,Nature 204, 220 (1964).

38. R. J. Britten and E. H. Davidson, Science165, 349 (1969).

39. J. Karlsson, in The Transmission 0/ Schizo­phrenia, D. Rosenthal and S. Kety, Eds.(Pergamon, Oxford, 1968).

40. L. Heston and D. Denney, ibid.41. P. Tienari, ibid.42. E. Inouye, in Proceedings, Third World Con..

Amasa B. Ford

change has accelerated. Many oldproblems have been mastered, butnew ones have arisen. "Poor l'aws"and workhouses have gone the way ofphthisis and chlorosis. Now we mustask whether general hospitals cancope with increasing drug addictionamong alienated youth or how healthdepartments can protect the publicagainst ,cigarettes and overeating. Butbefore we can restructure our healthservices we must assess what we knowabout the particular health needs oftoday. Because established social in­stitutions have great inertia, change isslow and· tends to lag behind need.New problems, therefore, call for spe­cial attention, since they foreshadowfuture needs.

The purpose of this report is toidentify sources of death and majordisability which are new or are ofnew inlportance in. developed coun­tries in the two decades since WorldWar II. Using examples from GreatBritain and the United States, wem1ake some estimates of how peopleare being affected.

Certain casualties result from im­mediate causes, such as the toxic ef­fects of a new drug or the increaseduse of nlotor vehicles. Many, possiblymore, take the fornl of nlajor disabil­ity resulting fronl conditions that have

gress of Psychiatry (Univ. of Toronto Press,Montreal, 1961), vol. 1, p. 524.

43. F. J. Kallmann, Amer. 1. Psychiat. 103, 309(1946).

44. J. Shields, I. Gottesman, E. Slater, ActaPsychiat. Scand. 43, 385 (1967).

45. E. Kahn, Monogr. Gesamtgeb. Neural.Psychiat. 36, 1 (1923).

46. B. Schulz,Z. Gesamte N eurol. Psychiar. 168,322 (1940).

47. G. Elsasser, Die N achkommen geisteskrankerElternpaare (Thieme, Stuttgart, 1952).

48. I thank James Shields, John Price, IrvingGottesman', and Russell Noyes, who com­mented on various phases of this manuscript.

complex origins. Examples are theextended survival of old people ·withchronic disease and the social aliena­tion of young people. A rough classi­fication by cause will serv.e as an out­line, since an understanding of howdisease originates is the most reason­able basis for control and prevention.Effects on health may be produced bychanges in population, by technologi­cal develqplnents, by new factors inmedicine, or by shifting social andcultural patterns. These categories,however, should not obscure the factthat specific casualties may resultfrom multiple causes.

Signs of Change

,Prosperity and life expectancy havereached unprecedented levels in de­veloped countries during the past 20years, but there are indications thatwe may be approaching the limit ofeffectiveness of current methods ofdis.ease control and prevention.

In the early 1950's infant mortalityrates ceased to improve at the ratewhich had prevailed for many. years.

lIn the decade 1946-56, rates had de­creased 46 percent in the UnitedStates and 45 percent in England andWales. In the subsequent 10 years,they decreased' only 16 and 22 p.er­cent, respectively. The estimated aver­age length of life, which in effect isinversely related to infant mortality,has increased to 70.1 years in theUnited States and 68.4 years in Eng­land and Wales, but the rates of in­crease since 1956 have been less thana fifth of what they wer.e in the pre­vious 10.years. The progressive reduc­tion of deaths in the first year of life,which has been one of the finest fruitsof social and medical development

Dr. Ford is assistant professor of medicine andpreventive medicine at. Case Western ReserveSchool of Medicine, Cleveland, Ohio 44106.

SCIENCE, VOL. 167