119
reference to Professor Leck’s intended was to alert the reader to Coffey and Jessop’s finding of an increase in anencephaly after the 1957 influenza pandemic,4 and to put these results in the wider context of Leck’s report of a similar increase in the number of congenital malfonnations,5 though not in anencephaly. Several case-control and cohort studies support an increase in neural tube defects following maternal influenza in early pregnancy.6 Intranasal administration of influenza to pregnant mice reduces fetal growth and increases the incidence of congenital malformations.7 Leck did not fmd an increase in anencephaly in Scotland after the influenza epidemics of the 1950s,5 but Doll, Hill, and Sakula reached the opposite conclusion, that the 1957 and 1958 epidemics were followed by a z10% increase in anencephaly in Scotland.8 Part of the continuing controversy stems from the fact that anencephaly is rare and subject to appreciable random fluctuations. More detailed statistical analysis may help to resolve the issue. Dr Jongbloet and colleagues suggest ovopathy as an alternative explanation for the season-of-birth effect in schizophrenia, but this hypothesis does not predict the observed increase in schizophrenic births after the 1957 influenza pandemic. With regard to the low monozygotic (MZ) twin concordance in schizophrenia, there are reports, from as long ago as 1937, that viral infections in pregnant mothers with MZ twins sometimes affect only one twinWe do not propose that all cases of schizophrenia are related to influenza; many may be genetic in origin, and some of these are associated with chromosomal anomalies. Indeed, it is possible that part of the genetic predisposition to schizophrenia is related to abnormal host responses to viral infections. We are flattered to receive a comment from Mr LaFosse and Dr Mednick, for it was the pioneering work of Mednick’s group that first drew attention to the relation between the 1957 influenza pandemic and schizophenia. It is difficult to pinpoint the time of maximum effect of the epidemic from recorded data on laboratory reports, sickness absences, and influenza-related deaths because these measures are subject to variable degrees of time lag beween the occurrence and report. Furthermore, the timing of the peak of the epidemic varied in different parts of England and Wales. The calculation of the time of maximum effect is therefore more difficult in our study than in Mednick’s in Helsinki,9 and it is possible that the critical month is the 6th rather than the 5th. We have data on the effects of influenza epidemics over a period of 22 years in England and Wales that is compatible with this, but at present we would confine ourselves to saying that the window of vulnerabilty is between the 5th and 7th month of fetal development. We
apologise for our ambiguous
paper. What
Autistic dimension
we
Department of Psychological Medicine, Institute of Psychiatry and King’s College Hospital, London SE5 8AF, UK
EADBHARD O’CALLAGHAN PAK SHAM NORIYOSHI TAKEI GYLES GLOVER ROBIN MURRAY
HA, Sempos CT. Statistical methods in epidemiology. New York: Oxford University Press, 1989: 45-71. 2. Woodall J, Rowson KEK, McDonald JC. Age and Asian influenza. Br Med J 1958; ii: 1. Kahn
1316. 3. Penrose LS. Congenital syphilis in monovular twins. Lancet 1937; i: 322. 4. Coffey VP, Jessop WJE. Maternal influenza and congenital deformities: a prospective study. Lancet 1959; ii: 935-38. 5. Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963; 17: 70-80. 6. Leck I. Epidemiological clues to the causation of neural tube defects. In: Dobbing J, ed. Prevention of spina bifida and other neural tube defects. London: Academic Press, 1983: 155-82. 7. Adams JM, Heath HD, Imagawa DT, Jones MH, Shear HH Viral infections of the embryo. Am J Dis Child 1956; 93: 109-14. 8 Doll R, Hill AB, Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 1960; 14: 167-72. 9. Mednick SA, Machon RA, Huttenen MO, Bonet D. Adult schizophrenia following prenatal exposure to an influenza epidemic Arch Gen Psychiatry 1988; 45: 189-92.
Sip,—M:K: was launched on Oct 4,1957. Space and space travel figure increasingly in schizophrenic delusional systems. More research is needed. Institute of Psychiatry, London SE5 8AF, UK
KEITH LLOYD GLYN LEWIS
SIR,-Your May 18 editorial states that disorders of communication are uncommon, occurring in 20-40 per 100 000 children, and are referred to by the diagnostic manual of the American Psychiatric Association (DSM-IIIR) as pervasive developmental disorders. Communication disorders are much commoner, having a prevalence of 3-15% of children,’ and DSM-IIIR has a separate category of "developmental language disorders". The much rarer communication disorder which accompanies autism can also occur on its own as semanticpragmatic disorder. 2,3 The grounds given for linking together the proposed conditions in an autistic dimension are suspect. Rett’s syndrome cannot be seen as a forme fruste of autism when it has a quite different sex distribution, clinical features, and course. The assertion that Asperger’s syndrome and autism have the same underlying pathology is unwarranted, being based on a single case-report of Asperger’s syndrome, and the magnetic resonance study cited found diverse and differing abnormalities in autism: the brain lesions in these conditions remain unestablished. However, new genetically based research of twins and firstdegree relatives of autistic children shows that they have an increased incidence of cognitive disorders (mainly developmental disorders of speech, language, or reading) and impairments in reciprocal social interaction.4 This provides empirical evidence for what might be included in an "autistic dimension". Department of Child and Adolescent Psychiatry, Institute of Psychiatry, London SE5 8AF, UK
STEPHEN SCOTT
1. Silva PA.
Epidemiology, longitudinal course, and some associated factors: an update. W, Rutter M, eds. Language development and disorders. Oxford: Blackwell, 1987: 1-15. 2. Rapin I, Allen DA. Developmental language disorders: nosologic considerations. In: Kirk U, ed. Neuropsychology of language, reading and spelling. New York: Academic Press, 1983: 155-84. 3. Adams C, Bishops DVM. Conversational characteristics of children with semanticpragmatic disorder. Br J Disorders Commun 1989; 24: 211-39. 4. Rutter M. Autism as a genetic disorder. In: McGuffin P, Murray R, eds. Mental illness and the new genetics. Oxford: Heinemann Medical, 1991. In Yule
Severe immune haemolytic anaemia associated with newer cephalosporins SIR,-Up to 1987 very few cases of immune haemolytic anaemia due to cephalosporins had been reported. Only five welldocumented cases due to first-generation cephalosporins were recorded in 25 years;l-4 all five cases were associated with cephalothin (one case was also associated with cephazolin). Several others have been reported but the serological evidence was incomplete. Second-generation cephalosporins began to be used in the 1970s, but up to 1987 there was only one report of a patient having immune haemolytic anaemia associated with their use; this was attributable to cephamandole.s In 1987 Salama et al6 reported severe intravascular haemolysis due to a third-generation cephalosporin, cefotaxime. Since this report nine other cases of immune haemolytic anaemia associated with cephalosporins have been recorded. Doctors may not be aware of the number of cases since only three have been published."’ Six of the nine have, so far, been recorded only as abstracts (two in Transfusion 1989, vol 29 (suppl): 51 S, and four in XXI Congress of International Society of Blood Transfusion and American Association of Blood Banks, 1990: 32-33): these six were associated with second-generation cephalosporins (five were due to cefotetan and one to cefoxitin). Three others were associated with third generation cephalosporins (cefotaximeceftriaxone,8ceftazidime9). Most cases were associated with severe haemolysis and three had fatal haemolytic anaemia (associated with ceftriaxone, cefoxitin, and cefotetan). Doctors need to be aware that the second and third-generation cephalosporins have caused twice as many cases of immune haemolytic anaemia in the past 4 years than the first-generation drugs did in 25. The haemolysis seems much more severe and has
reference to Professor Leck’s intended was to alert the reader to Coffey and Jessop’s finding of an increase in anencephaly after the 1957 influenza pandemic,4 and to put these results in the wider context of Leck’s report of a similar increase in the number of congenital malfonnations,5 though not in anencephaly. Several case-control and cohort studies support an increase in neural tube defects following maternal influenza in early pregnancy.6 Intranasal administration of influenza to pregnant mice reduces fetal growth and increases the incidence of congenital malformations.7 Leck did not fmd an increase in anencephaly in Scotland after the influenza epidemics of the 1950s,5 but Doll, Hill, and Sakula reached the opposite conclusion, that the 1957 and 1958 epidemics were followed by a z10% increase in anencephaly in Scotland.8 Part of the continuing controversy stems from the fact that anencephaly is rare and subject to appreciable random fluctuations. More detailed statistical analysis may help to resolve the issue. Dr Jongbloet and colleagues suggest ovopathy as an alternative explanation for the season-of-birth effect in schizophrenia, but this hypothesis does not predict the observed increase in schizophrenic births after the 1957 influenza pandemic. With regard to the low monozygotic (MZ) twin concordance in schizophrenia, there are reports, from as long ago as 1937, that viral infections in pregnant mothers with MZ twins sometimes affect only one twinWe do not propose that all cases of schizophrenia are related to influenza; many may be genetic in origin, and some of these are associated with chromosomal anomalies. Indeed, it is possible that part of the genetic predisposition to schizophrenia is related to abnormal host responses to viral infections. We are flattered to receive a comment from Mr LaFosse and Dr Mednick, for it was the pioneering work of Mednick’s group that first drew attention to the relation between the 1957 influenza pandemic and schizophenia. It is difficult to pinpoint the time of maximum effect of the epidemic from recorded data on laboratory reports, sickness absences, and influenza-related deaths because these measures are subject to variable degrees of time lag beween the occurrence and report. Furthermore, the timing of the peak of the epidemic varied in different parts of England and Wales. The calculation of the time of maximum effect is therefore more difficult in our study than in Mednick’s in Helsinki,9 and it is possible that the critical month is the 6th rather than the 5th. We have data on the effects of influenza epidemics over a period of 22 years in England and Wales that is compatible with this, but at present we would confine ourselves to saying that the window of vulnerabilty is between the 5th and 7th month of fetal development. We
apologise for our ambiguous
paper. What
Autistic dimension
we
Department of Psychological Medicine, Institute of Psychiatry and King’s College Hospital, London SE5 8AF, UK
EADBHARD O’CALLAGHAN PAK SHAM NORIYOSHI TAKEI GYLES GLOVER ROBIN MURRAY
HA, Sempos CT. Statistical methods in epidemiology. New York: Oxford University Press, 1989: 45-71. 2. Woodall J, Rowson KEK, McDonald JC. Age and Asian influenza. Br Med J 1958; ii: 1. Kahn
1316. 3. Penrose LS. Congenital syphilis in monovular twins. Lancet 1937; i: 322. 4. Coffey VP, Jessop WJE. Maternal influenza and congenital deformities: a prospective study. Lancet 1959; ii: 935-38. 5. Leck I. Incidence of malformations following influenza epidemics. Br J Prev Soc Med 1963; 17: 70-80. 6. Leck I. Epidemiological clues to the causation of neural tube defects. In: Dobbing J, ed. Prevention of spina bifida and other neural tube defects. London: Academic Press, 1983: 155-82. 7. Adams JM, Heath HD, Imagawa DT, Jones MH, Shear HH Viral infections of the embryo. Am J Dis Child 1956; 93: 109-14. 8 Doll R, Hill AB, Sakula J. Asian influenza in pregnancy and congenital defects. Br J Prev Soc Med 1960; 14: 167-72. 9. Mednick SA, Machon RA, Huttenen MO, Bonet D. Adult schizophrenia following prenatal exposure to an influenza epidemic Arch Gen Psychiatry 1988; 45: 189-92.
Sip,—M:K: was launched on Oct 4,1957. Space and space travel figure increasingly in schizophrenic delusional systems. More research is needed. Institute of Psychiatry, London SE5 8AF, UK
KEITH LLOYD GLYN LEWIS
SIR,-Your May 18 editorial states that disorders of communication are uncommon, occurring in 20-40 per 100 000 children, and are referred to by the diagnostic manual of the American Psychiatric Association (DSM-IIIR) as pervasive developmental disorders. Communication disorders are much commoner, having a prevalence of 3-15% of children,’ and DSM-IIIR has a separate category of "developmental language disorders". The much rarer communication disorder which accompanies autism can also occur on its own as semanticpragmatic disorder. 2,3 The grounds given for linking together the proposed conditions in an autistic dimension are suspect. Rett’s syndrome cannot be seen as a forme fruste of autism when it has a quite different sex distribution, clinical features, and course. The assertion that Asperger’s syndrome and autism have the same underlying pathology is unwarranted, being based on a single case-report of Asperger’s syndrome, and the magnetic resonance study cited found diverse and differing abnormalities in autism: the brain lesions in these conditions remain unestablished. However, new genetically based research of twins and firstdegree relatives of autistic children shows that they have an increased incidence of cognitive disorders (mainly developmental disorders of speech, language, or reading) and impairments in reciprocal social interaction.4 This provides empirical evidence for what might be included in an "autistic dimension". Department of Child and Adolescent Psychiatry, Institute of Psychiatry, London SE5 8AF, UK
STEPHEN SCOTT
1. Silva PA.
Epidemiology, longitudinal course, and some associated factors: an update. W, Rutter M, eds. Language development and disorders. Oxford: Blackwell, 1987: 1-15. 2. Rapin I, Allen DA. Developmental language disorders: nosologic considerations. In: Kirk U, ed. Neuropsychology of language, reading and spelling. New York: Academic Press, 1983: 155-84. 3. Adams C, Bishops DVM. Conversational characteristics of children with semanticpragmatic disorder. Br J Disorders Commun 1989; 24: 211-39. 4. Rutter M. Autism as a genetic disorder. In: McGuffin P, Murray R, eds. Mental illness and the new genetics. Oxford: Heinemann Medical, 1991. In Yule
Severe immune haemolytic anaemia associated with newer cephalosporins SIR,-Up to 1987 very few cases of immune haemolytic anaemia due to cephalosporins had been reported. Only five welldocumented cases due to first-generation cephalosporins were recorded in 25 years;l-4 all five cases were associated with cephalothin (one case was also associated with cephazolin). Several others have been reported but the serological evidence was incomplete. Second-generation cephalosporins began to be used in the 1970s, but up to 1987 there was only one report of a patient having immune haemolytic anaemia associated with their use; this was attributable to cephamandole.s In 1987 Salama et al6 reported severe intravascular haemolysis due to a third-generation cephalosporin, cefotaxime. Since this report nine other cases of immune haemolytic anaemia associated with cephalosporins have been recorded. Doctors may not be aware of the number of cases since only three have been published."’ Six of the nine have, so far, been recorded only as abstracts (two in Transfusion 1989, vol 29 (suppl): 51 S, and four in XXI Congress of International Society of Blood Transfusion and American Association of Blood Banks, 1990: 32-33): these six were associated with second-generation cephalosporins (five were due to cefotetan and one to cefoxitin). Three others were associated with third generation cephalosporins (cefotaximeceftriaxone,8ceftazidime9). Most cases were associated with severe haemolysis and three had fatal haemolytic anaemia (associated with ceftriaxone, cefoxitin, and cefotetan). Doctors need to be aware that the second and third-generation cephalosporins have caused twice as many cases of immune haemolytic anaemia in the past 4 years than the first-generation drugs did in 25. The haemolysis seems much more severe and has
