International Journal of Epidemiology Vol. 26, No.

1
© International Epidemiological Association 1997 Printed in Great Britain

Prevalence and Time Trends of

Disabilities in School-Age Children
C RUMEAU-ROUQUETTE,* H GRANDJEAN,** C CANS,† C DU MAZAUBRUN* AND A VERRIER*

Rumeau-Rouquette C (Unité 149, INSERM, 123 Boulevard de Port-Royal, 75014 Paris, France), Grandjean H, Cans C,
du Mazaubrun C and Verrier A. Prevalence and time trends of disabilities in school-age children. International Journal of
Epidemiology 1997; 26: 137–145.
Background. Although the evolution of the prevalence of cerebral palsy is now well documented, much less is known
about the evolution of the prevalence of other disabilities such as mental retardation, sensorial defects, autism and
psychosis. The aim of this paper is to determine those trends.
Methods. A population-based survey was carried out in 1992–1993 in three French ‘départements’. All disabled children
born between 1976 and 1985 and receiving a special education and/or financial assistance were systematically
registered.
Results. The comparison of three cohorts of children born in 1976–1978, 1979–1981 and 1982–1984 using the test for
trend in proportion showed a significant decrease (P = 0.03) in the prevalence of severe mental retardation, after exclu-
sion of Down syndrome. This decrease was significant for severe mental retardation associated with psychosis. The time
trend prevalence for cerebral palsy increased (P = 0.03) but was irregular. The time trend prevalence of other disabilities
(other motor defects, severe sensorial disabilities, autism and psychosis) did not change significantly. A detailed analysis
of severe mental retardation and cerebral palsy was performed by geographical area, age at first registration and type of
disability.
Conclusion. The increase in prevalence of cerebral palsy is possibly due to earlier registration of disabled children. The
decrease in prevalence of severe mental retardation does not seem to be due to recruitment bias, but there is a possibility
of classification bias.
Keywords: prevalence, time trends, schoolchildren, disabilities, France

The prevalence of disabilities in children has been MATERIAL AND METHODS

evaluated by a number of surveys or registers, but the
data most often concern cerebral palsy and severe
mental retardation. Several of these studies, notably
in Sweden,1–4 the UK,5,6 and Australia,7–10 have shown
that the prevalence of cerebral palsy has remained un-
changed or increased. In France, a previous survey
allowed us to compare the children living in 16 départe-
ments11–13 who were born in 1972, 1976, and 1981 and
to show that the rate of the main disabilities has re-
mained relatively stable. Since then, there has been
progress, especially in resuscitation of new-borns, which
may influence the long term outcome of children. We
conducted a new survey to study the evolution of dis-
abilities among more recent cohorts of children. The aim
of this paper is to describe the prevalence of severe
motor, mental and sensorial disabilities and to analyse
the trends in prevalence among children born from
1976 to 1984.
* Unité 149, INSERM, 123 Boulevard de Port-Royal, 75014 Paris,
France.
** CJF 94-04, INSERM, Hôpital de la Grave, 31052 Toulouse, France.
†
RHEOP, 23 rue Albert 1er de Belgique, 38000 Grenoble, France.
137
Registration of Disabled Children
This study was carried out in 1992–1993 in three French
‘départements’ (Haute-Garonne, Isère, and Saône-et-
Loire) and concerned all disabled children born between
1976 and 1985 whose parents lived in one of these
départements at the time of the study. At that time the
children were aged 8–17 years.
These geographical areas were chosen because local
epidemiological teams had experience in research into
disabilities and perinatal problems. The Haute-Garonne,
located in the region called Midi-Pyrénées in the South-
west of France, includes the region’s main city, Tou-
louse, which contains university hospitals and public
and private institutions. On the other hand, Isère is a
part of the large Rhône-Alpes region, whose main city,
Lyon, is not in the département. The Saône-et-Loire is
a rural area in Burgundy, whose main city is Dijon, also
in a different département.
The survey was carried out in co-operation with the
CDES (Commission Départementale de l’Education
Spéciale) of each département. They are the only au-
thority authorized under French law to refer disabled
children for special education and care in public or
138 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
private institutions and to provide financial assistance
to their families. Families may request referral or assist-
ance for children aged ,20 years. The request for chil-
dren with severe disabilities is generally made, however,
before school age or during the first school year. To
verify the completeness of the records, severe disab-
ilities were also registered through institutions that can,
in special circumstances, accept children without CDES
authorization, such as day hospitals, residential centres
and CAMSP (Centre d’Action Médico-Social Précoce).
The latter, which are centres for early medical-social
treatment, were included only in Isère.
The study was carried out by one of INSERM’s
public health networks. A team of researchers in each
département gathered and verified data and was re-
sponsible for a part of the analysis. The study as a
whole was co-ordinated by a team from INSERM
Research Unit 149. At the local level, the procedures
were the following:
– the CDES sent a request for authorization to parents,
in accordance with the procedures established by the
CNIL (Commission Nationale de l’Informatique et des
Libertés);
– unless parents refused, a physician-investigator,
approved by the CDES, completed a computer-
readable file containing some demographic informa-
tion (but excluding name and address), a description
of the disabilities and their probable cause, and a
summary of placements; this information came from
the CDES files, and children were not examined
especially for the study;
– the investigator also obtained information from day
hospitals, residential centres and CAMSP;
– teams in each département verified and coded the
files thus compiled;
– the data were then entered without any mention of
the children’s identity; additional verification pro-
cedures aimed at corroborating the homogeneity of
both the data gathering and the coding were carried out.
The diagnoses were coded using the nomenclature
of the World Health Organization (ICD-9), except for
mental disorders, which were coded according to the
French classification of childhood and adolescent men-
tal illnesses.14
Definition of Disabilities
The analysis considered the following disabilities.
Mental retardation (MR) was divided into two main
groups:
– Down syndrome (trisomy 21), regardless of IQ level;
– other severe mental retardation, which was defined,
following Kiely,15 as IQ ø 50; when IQ was not
available, mental retardation specified as severe or
profound was included in this group.
Motor disabilities were considered, whatever their
severity, and classified in three groups:
– cerebral palsy (CP) without any mention of post-
neonatal origin,
– other motor disabilities without any mention of post-
neonatal origin,
– motor disabilities with mention of post-neonatal origin.
The following CP defects were included: diplegia,
tetraplegia, hemiplegia and non-paralytic forms.
Only severe sensorial defects were considered, using
the following definitions:
– blindness or amblyopia in both eyes (visual acuity
less than 3/10 with the best possible correction);
– bilateral deafness or severe hearing loss (levels
.70 dB).
Only two types of severe psychiatric disorders were
studied:
– infantile autism,
– other psychoses.
In the part of the analysis concerning each kind of
disability, a child with multiple disabilities was counted
in two or more groups. We then analysed associations
between different disabilities.
The cases above do not include all of the children un-
der CDES jurisdiction, for we excluded from the study
such diseases as cancer, chronic lung diseases, malforma-
tions not entailing the disabilities described above, and
mental disorders except for autism and other psychoses.
Method of Analysis
Prevalence rates were calculated by dividing the
number of disabled children living in the département
at the time of the survey (1992–1993) by the number of
children born during the same year and living in the
département during the 1990 census. To take into ac-
count any possible effect of migration, the prevalence
rates were also calculated by dividing the number of
disabled children born and living in the surveyed area
by the number of all children born and living there.
After comparing these two rates, we performed a more
detailed study of the disabilities and their associations.
 DISABILITIES IN SCHOOL-AGE CHILDREN 139

TABLE 1 Prevalence rates of disabilities among residents and among native residents

Residents Native residents P

n per 1000 n per 1000

Base population 325 347 231 571

Down syndrome 320 0.98 222 0.96 0.77
Other severe mental 841 2.58 581 2.51 0.58
retardation
Cerebral palsy 600 1.84 395 1.71 0.23
Other motor defects
post-neonatal 282 0.87 174 0.75 0.14
other 427 1.31 293 1.27 0.63
Blindness and amblyopia 220 0.68 131 0.57 0.10
Deafness and severe 226 0.69 130 0.56 0.05
hearing loss
Autism 173 0.53 107 0.46 0.25
Other psychoses 463 1.42 307 1.33 0.33

Then the analysis focused on the time trends of
prevalence rates. Since the children’s ages varied at
data collection, and because the eldest among them had
the highest probability of having been referred to CDES,
only children registered by the CDES before age
9 years have been included. Children born in 1985 were
excluded because many of them were ,9 years at the
time of the survey. Those born between 1976 and 1984
were subdivided into three cohorts by regrouping three
birth years (1976–1978, 1979–1981, and 1982–1984)
with the aim of having a reasonable number of disabled
children in each cohort when the prevalence rate was
low. Comparing those cohorts it was possible to detect
a prevalence rate difference of 0.3 to 0.5 per 1000, with
an α risk of 0.05 and a β risk of 0.20. The prevalence
rates observed for those three cohorts were compared
by using the test of linear trend of proportions.16 When
a difference (P , 0.05) was observed, prevalence rates
were calculated for each birth year to verify if the above
subdivision could introduce an artefact. Then a more
detailed analysis was performed, taking into account
the geographical area, the age at first registration by the
CDES and the associations of disabilities.
RESULTS
Only 6% of parents refused to allow their children to be
included in the study: the refusal rate was between 5%
and 7% according to the département. In the three
départements as a whole, 6174 disabled children were
reported to our study. Of this number, 6103 had been
reported to the CDES, and 71 came from the records of
the day hospitals and CAMSP. There were 6100 whose
parents lived in these three départements at the date of
data collection.
The base population comprised 325 347 children
born between 1976 and 1985 whose parents lived in the
three départements during the 1990 population census:
71% of those children were born in the surveyed départe-
ments; 3% in a foreign country; and the others in dif-
ferent areas of France.
Table 1 shows the overall prevalence rates, calcu-
lated for current (i.e. 1990) residents (R) and for native
residents (NR) of the same département. They did not
differ significantly. For deafness and severe hearing
loss, however, the P level was 0.054. This figure results
from a significant difference in Haute-Garonne: 0.80 per
1000 (R) versus 0.49 per 1000 (NR) (P = 0.01), where-
as the rates in Isère were 0.65 per 1000 (R) versus
0.59 per 1000 (NR), and in Saône-et-Loire, 0.62 per
1000 (R) versus 0.60 per 1000 (NR). The subsequent
analysis was thus performed considering prevalence
rates only for current residents. For deafness and severe
hearing loss, however, both rates were calculated.
The prevalence of MR as a whole was 3.56 per
1000. This group was subdivided in two groups: cases
of Down syndrome (0.98 per 1000), and other severe
MR (2.58 per 1000). Associations between severe MR
and other disabilities were frequent (Table 2). Severe
MR was most often associated, in order of descending
prevalence, with cerebral palsy (CP) and other motor
disabilities, then psychosis, autism and severe visual
defects. Association with severe hearing loss and deaf-
ness was infrequent. The prevalence of severe MR not
associated with the disabilities listed here was only
0.77 per 1000.
140 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

The prevalence of CP without any mention of a post- TABLE 2 Prevalence ratesa of severe mental retardation (MR)
neonatal origin was 1.84 per 1000. The CP cases included associated with other disabilities and prevalence rate of dis-
the following defects: tetraplegia (40%), diplegia (17%), abilities not associated with severe MR
hemiplegia (21%); the main symptoms of non-paralytic
CP were major inco-ordination and tonus anomalies; 43% No. Prevalence
of CP children could not walk, while 17% were able to per 1000
walk only with help. Cerebral palsy was principally associ-
ated with severe MR (Table 2) and severe visual defects. Severe MR associated with:b
cerebral palsy 253 0.78
The prevalence of the latter association was 0.10 per 1000
other motor defects 164 0.50
and that of CP not associated with any deficiency under blindness or amblyopia 65 0.20
consideration in this paper was 0.84 per 1000. deafness or severe hearing loss 9 0.03
Of other motor disabilities not specified as post- autism 71 0.22
neonatal (1.31 per 1000), 19% involved a malformation psychoses 156 0.48
of the nervous system (NS), and 30% a malformation Severe MR without above disabilities 251 0.77
of the osteo-muscular system. The prevalence of spina Disabilities not associated with severe MR
bifida among the NS malformations was 0.22 per 1000. cerebral palsy 347 1.07
The prevalence of motor defects specified as post- other motor defects 545 1.68
neonatal was 0.87 per 1000. blindness or amblyopia 155 0.48
deafness or severe hearing loss 217 0.67
The prevalence of blindness and amblyopia was autism 102 0.31
0.68 per 1000, and the rate of blindness in both eyes, psychoses 307 0.94
0.27 per 1000. Deafness and severe hearing loss to-
gether were as prevalent (0.69 per 1000) as the serious a
Prevalence rates among residents.
visual disabilities, but the rate of bilateral deafness b
Children with three or more disabilities are counted several times.
(0.41 per 1000) was slightly higher than that of blind-
ness in both eyes. Visual defects were often associated
with severe MR and CP, as described above.
The prevalence of autism was 0.53 per 1000, and The prevalence of CP increased (P = 0.03). The
41% of the cases were associated with severe MR trends for other motor disabilities, specified as post-
(Table 2). The prevalence of psychoses was higher neonatal or not, showed a slight but non-significant
(1.42 per 1000), but the association with severe MR, decrease. Spina bifida decreased significantly in the
slightly lower (34%). third cohort: its prevalence in cohorts 1, 2 and 3 was,
respectively, 0.26, 0.28 and 0.09 (P = 0.01).
Trends of Prevalence by Birth Period (Table 3) A detailed analysis by birth year was performed for
The prevalence of Down syndrome was very stable in severe MR and CP (Figures 1 and 2). The time trend of
the three cohorts. For other severe MR, however, the severe MR was less irregular than that of CP. The pro-
prevalence rate decreased (P = 0.03). portion of severe MR decreased and was significant

TABLE 3 Trends of prevalence rates among children registered by CDESa before the age of 9 years

1976–1978 1979–1981 1982–1984 P trend

n per 1000 n per 1000 n per 1000

Down syndrome 88 0.93 96 0.95 94 0.97 0.76

Other severe mental retardation 225 2.37 231 2.28 186 1.92 0.03
Cerebral palsy 144 1.52 173 1.71 186 1.92 0.03
Other motor defects
post-neonatal 67 0.71 71 0.70 63 0.65 0.64
other 108 1.14 120 1.18 96 0.99 0.34
Blindness and amblyopia 59 0.62 58 0.57 56 0.58 0.69
Deafness and severe hearing loss 70 0.74 63 0.62 66 0.68 0.64
Autism 43 0.45 69 0.68 41 0.42 0.76
Other psychoses 114 1.20 87 0.86 94 0.97 0.11

a
Commission Départementale de l’Education Spéciale.
 DISABILITIES IN SCHOOL-AGE CHILDREN 141

FIGURE 1 Trend of prevalence of severe mental retardation by

FIGURE 2 Trend of prevalence of cerebral palsy by birth year
birth year (rates and 95% confidence intervals)
(rates and 95% confidence intervals)

TABLE 4 Prevalence rates by birth period and geographical areas and age at first registration by CDESa among children registered by
CDES before the age of 9 years

1976–1978 1979–1981 1982–1984 P trend

n per 1000 n per 1000 n per 1000

Other severe mental retardation

Haute-Garonne 74 2.34 82 2.44 63 1.96 0.31
Isère 94 2.27 93 2.05 80 1.84 0.16
Saône-et-Loire 57 2.62 56 2.49 43 2.03 0.20
Cerebral palsy
Haute-Garonne 67 2.12 62 1.85 84 2.61 0.18
Isère 56 1.35 85 1.88 77 1.77 0.96
Saône-et-Loire 21 0.97 26 1.15 25 1.18 0.61
Other severe mental retardation
,6 156 1.65 174 1.72 145 1.50 0.42
6–8 69 0.73 57 0.56 41 0.42 0.05
Cerebral palsy
,6 121 1.28 150 1.48 165 1.70 0.01
6–8 23 0.24 23 0.23 21 0.22 0.71

a
Commission Départementale de l’Education Spéciale.

(P = 0.024), and the increase in CP was also signific-
ant (P = 0.047).
The prevalence rates of sensorial defects dropped
slightly between the first and the second cohort, but the
difference was not significant. As explained above, the
prevalence rates of deafness and severe hearing loss
were also calculated among the residents born in the
same département. These rates were 0.58, 0.51 and
0.58 per 1000 for cohorts 1, 2 and 3, respectively.
The prevalence rate of autism had a rather irregular
time trend. Although it decreased significantly between
the second and third cohort, the trend was not sig-
nificant. An irregular but not significant trend was also
observed for the group of other psychoses.
The analysis was also performed by taking into ac-
count the geographical area and age at the first registra-
tion by the CDES (Table 4). The time trends of severe
MR showed a regular decrease in Isère and in Saône-et-
Loire, but the drop was observed only in the third co-
hort in Haute-Garonne. The prevalences of CP observed
in each area, increased slightly from cohort 1 to 3 in the
three areas. However, no time trend was significant
when each département was considered.
Table 4 shows the trends for children registered by
the CDES before and from the age of 6 years. Severe
MR prevalence rates decreased significantly among
children registered between 6 and 8 years but not for those
registered before the age of 6 years. On the contrary, for
142 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

TABLE 5 Prevalence rates of severe mental retardation (MR) associated with other disabilities (children registered by CDES before the
age of 9 years)

1976–1978 1979–1981 1982–1984 P trend

n per 1000 n per 1000 n per 1000

Severe MR associated with

cerebral palsy 78 0.82 72 0.71 77 0.80 0.83
other motor defects 50 0.53 49 0.48 38 0.39 0.17
blindness or amblyopia 23 0.24 18 0.18 16 0.17 0.22
autism 20 0.21 27 0.27 14 0.14 0.31
other psychoses 52 0.55 42 0.41 26 0.27 0.002
Severe MR without other
disability 46 0.49 58 0.57 42 0.43 0.68

TABLE 6 Prevalence rates of cerebral palsy by association with mental retardation (MR) and by type of defects (children registered before
the age of 9 years)

1976–1978 1979–1981 1982–1984 P trend

n per 1000 n per 1000 n per 1000

Associations:
cerebral palsy plus severe 78 0.82 72 0.71 77 0.80 0.83
MR
cerebral palsy without severe
MR 66 0.70 101 1.00 109 1.13 0.02
Type of defects:
hemiplegia 27 0.28 29 0.29 41 0.42 0.09
diplegia 25 0.26 30 0.30 25 0.26 0.94
quadriplegia 57 0.60 66 0.65 67 0.69 0.44
non-paralytic form 32 0.34 45 0.44 51 0.53 0.04

CP the trend increased significantly among the children
registered before the age of 6 years.
All associations between severe MR and other dis-
abilities (Table 5) showed decreasing time trends, ex-
cept for those with CP, but only the decrease for severe
MR associated with psychosis was significant. The pre-
valence rate of severe MR not associated with other
disabilities had an irregular trend.
Except for CP with severe MR and for diplegia,
which both showed an irregular trend, the prevalence
rates of every form of CP increased slightly (Table 6).
These increasing trends reached a significant level for
CP without severe MR and non-paralytic forms.
DISCUSSION
This survey was based upon information about disab-
ilities furnished to the CDES by the children’s physicians.
The notification of disabled children is generally made
by the parents and is not compulsory, but the CDES are
the only authorities that can provide financial assist-
ance and/or authorize a child’s admission into a public
or private institution or school. Other sources of in-
formation, such as psychiatric day treatment hospitals
and general hospitals, were also included in the survey,
because the parents may be directly reimbursed by the
‘Sécurité Sociale’ for care in those hospitals. Only a
small number of disabled children were treated in those
hospitals without notice to the CDES (1%). The parent
refusal rate was rather low. The registration of severe dis-
abilities may thus be considered as reasonably complete.
As a continuous stream of immigrants enters départe-
ments under study (71% of children were born in
the département in which they lived in 1990), and a
high number of specialized institutions exist in Isère
and Haute-Garonne, we compared the prevalence rates
 DISABILITIES IN SCHOOL-AGE CHILDREN
calculated for all residents and for residents born in the
same département. The rates for all residents were
slightly but not significantly higher, and when con-
sidering each area the only significant difference was
an excess of severe hearing loss in Haute-Garonne. So
the analysis was performed using the prevalence rates
calculated for all residents.
This study required a more detailed investigation
than the previous survey of Unit 149;11–13 because the
co-operation of local epidemiological teams was needed,
this survey was performed in only three départements.
The results from these three départements cannot be
considered as representative of CDES registration
nationwide, in contrast to the earlier work, in which a
concern for representativity guided the selection of the
16 studied départements.11 The comparison of the res-
ults of the two studies should therefore be made with
caution. Certain rates are higher in the current study:
Down syndrome (0.98 per 1000, versus 0.82 per 1000);
other severe mental retardation (2.58 per 1000, versus
1.69 per 1000), although some of this difference may be
explained by increased precision in inclusion criteria;
motor disabilities (4.02 per 1000 versus 3.24 per 1000).
For sensory disabilities and autism, the results of the
two surveys are rather similar. The difference in pre-
valence of disabilities between the two surveys may be
explained by the regional differences which were also
observed in our previous survey.11
The comparison of our results with those of other
studies indicates that the prevalence of Down syndrome
is quite similar to that noted at birth in the main mal-
formation registers.17,18 Foreign studies do not separate
Down syndrome from severe MR, as we have done, so
comparisons are difficult. Looking at school-age chil-
dren, the prevalence of severe MR (IQ , 50) varies
between 2.4119 and 4.99 per 1000,20 but most of the
rates fall between 3.0 per 1000 and 3.5 per 1000.15 The
total prevalence of 3.56 per 1000 (Down syndrome plus
severe MR) that we observed in the current study is
similar to the foreign data. The prevalence of motor
disabilities seen here is higher than that observed else-
where, especially in Sweden, where it is approximately
3 per 1000.21 That of CP is just ,2 per 1000; the rate
observed in most studies. Some variation exists among
the data on sensory disabilities. Blindness varies from
0.15 to 0.41 per 1000.22,23 The rate we observed (0.27 per
1000) falls within these limits. For deafness, our results
are close to those described by the Oxford Handicap
Register (0.51 per 1000 in 5-year old children)24 and
the Finnish study of a cohort of children born in 1966
and followed through the age of 14 years.22 The pub-
lished prevalence of autism fluctuates substantially (from
0.07 and 1.39 per 1000). Studies of French children
143
whose age is comparable to those of the children in our
study report a rate between 0.4 and 0.5 per 1000.25
The time trend of the prevalence of disabilities in this
study is different from that seen in our earlier study.11
In the present survey a tendency towards the diminution
of severe MR was observed especially in the last co-
hort; this was not observed previously. There was also
a significant increase in the CP rates, which was not
previously observed. The decrease in the rates of deaf-
ness and autism that we observed between the 1976 and
1981 cohorts11 does not appear in this study.
The stability in the prevalence of Down syndrome
may appear surprising, but recent results in Paris showed
that although the number of therapeutic abortions rose
between 1981 and 1989, the prevalence of Down syn-
drome at birth did not decrease: it was 1.08 per 1000 in
1981 and 1.23 per 1000 in 1985, probably due to the in-
crease in births to older mothers.26 According to other
data from the International Clearinghouse for Birth De-
fects Monitoring System about births from 1974 through
1988, the prevalence of this disability increased in
Finland, in Strasbourg, and in Canada, and showed a
tendency to decrease in east central France and in two
regions of Italy.18 Even where Down syndrome birth
rates decrease, however, survival of Down syndrome
infants shows a tendency to increase.13
The decrease in the prevalence of severe MR was
significant, dropping 19% between the first and the
third cohort, and this decrease was significant whatever
the method of analysis of the time trends. It was only
significant for children registered by the CDES be-
tween the ages of 6–8 years. The tendency towards a
diminution was also observed in the three départements
and in various associations between severe MR and
other disabilities. This evolution must none the less be
confirmed by other studies because of the possibility
that a change occurred in the classification of severe
MR. This hypothesis must be considered for the fol-
lowing reasons: i) the drop was only significant when
severe MR was associated with psychosis; ii) for many
of the psychoses no IQ was indicated and severe MR
was defined by physicians using qualitative criteria. In
other studies time trends of the prevalence of severe
MR have been studied less than those of CP, and the
results are variable.11,15 According to Fryer and Mac-
kay,27 the prevalence rates of MR (IQ , 50) increased
from 1961 to 1977 and then began to decrease.
We observed a significant increase in CP prevalence
rates, more specifically in non-paralytic forms and for
CP not associated with severe MR. This increase was
only significant among children registered before the
age of 6 years and may be related to an increase of
early registration by the CDES. This means that the
144 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY
possibility of a recruitment bias has to be taken into
consideration. However, classification bias can be
excluded, for the classification of motor disabilities in
the group of CP were made a posteriori without any
difference from one birth year to another.
Among studies of cerebral palsy, the longest chrono-
logical series are those of Hagberg in Sweden, 1,4
Pharoah in the UK,5,6 and Stanley in Australia.7,9 Their
results indicate a stable tendency in the UK and in
Australia, and an increase in Sweden. When pooling
data from western Sweden and South-West Germany,
the prevalence of bilateral spastic CP showed no sig-
nificant change for children of normal birthweight born
between 1975 and 1986, while for children ,1500 g this
prevalence was irregular.28 In England5 the prevalence
has increased for children with a birthweight ,1500 g.
Contrary to our previous results, the time trends of
sensorial disability prevalence rates did not show any
significant change. These results underline how cau-
tiously any change in time trends must be interpreted
and the importance of comparisons with other results.
Comparisons are difficult, however, for sensorial dis-
abilities, except for those associated with congenital
malformations whose evolution is better known. The
prevalence of severe congenital anomalies of the eyes
(anophthalmos, microphthalmos and congenital catar-
acts) among live births is 0.25 per 100017 and its evolu-
tion may influence that of visual defects. However,
the prevalence of anotia and microtia is only 0.07
per 1000:17 these defects account for only a small pro-
portion of children with severe hearing loss. According
to the Registry of Congenital Malformations in Paris,26
the prevalence of anophthalmos and of microphthalmos
increased significantly from 1981 to 1990, but the rate
of induced abortions for these malformations has also
been increasing since 1988. The prevalence of congen-
ital cataract has shown annual fluctuations. The preval-
ence of anotia and microtia increased significantly from
1981 to 1992, and that of induced abortion for these
defects increased.
The change in the prevalence rate of autism is dif-
ficult to interpret: the test of trend was not significant,
the trend was irregular and there was a significant
decrease between the second and third cohorts. For
psychoses the time trend prevalence was also irregular,
with no significant change. The prevalence of psy-
choses decreased significantly only for those associated
with severe MR. Because distinguishing between aut-
ism and psychosis on the basis of the data collected by
the CDES was difficult, all cases were classified by a
single psychiatrist without taking either geographical
area or birth period into consideration, so that classi-
fication bias cannot explain those fluctuations. In our
previous survey, the prevalence rate of autism de-
creased significantly (0.56, 0.47 and 0.31 per 1000 for
children born in 1972, 1976, 1981).11 The high rate
observed now among children born in 1979–1981 is
difficult to explain. The prevalence rates of autism and
psychosis are too irregular to allow us to conclude that
the apparent decrease constitutes a real trend.
In conclusion, the prevalence of severe deficiencies
had not changed for children born between 1976 and
1984 except for CP and severe MR. The annual trend
for those deficiencies was irregular, particularly for the
annual rate of CP. A classification bias may explain
the decreasing prevalence of severe MR. For CP rate
the rising prevalence may be due to a recruitment bias
with earlier registration at CDES for CP children. Fur-
ther studies with follow-up of children from 0 to 15 years
need to be set up in order to control this recruitment
bias.
ACKNOWLEDGEMENTS
This study was carried out within the framework of a
public health network of INSERM, thanks to financing
by INSERM, the Ministry of Health, and the Caisse
Nationale d’Assurance Maladie. It benefited from
the co-operation of the CDES, the day hospitals, and
the CAMSP, all of which we warmly thank for their
co-operation.
REFERENCES
1
Hagberg B, Hagberg G, Olow I. The changing panorama of
cerebral palsy in Sweden 1954–1970. II. Analysis of the vari-
ous syndromes. Acta Paediatr Scand 1975; 64: 193–200.
2
Hagberg B, Hagberg G, Olow I. The changing panorama of
cerebral palsy in Sweden 1954–1970. III. The importance
of foetal deprivation of supply. Acta Paediatr Scand 1976;
65: 403–08.
3
Hagberg B, Hagberg G, Olow I. The changing panorama of
cerebral palsy in Sweden IV. Epidemiological trends
1959–78. Acta Paediatr Scand 1984; 73: 433–40.
4
Hagberg B, Hagberg G, Olow I. The changing panorama of
cerebral palsy in Sweden VI. Prevalence and origin during
the birth year period 1983–1986. Acta Paediatr Scand
1993; 82: 387–93.
5
Pharoah P O D, Cooke T, Cooke R W I, Rosenbloom L. Birth
weight specific trends in cerebral palsy. Arch Dis Child
1990; 65: 602–06.
6
Pharoah P O D, Cooke T, Rosenbloom L, Cooke R W I. Trends
in birth prevalence of cerebral palsy. Arch Dis Child 1987;
62: 379–84.
7
Stanley F J, Blair E. Cerebral palsy. In: Pless I B (ed.). The Epi-
demiology of Childhood Disorders. Oxford: Oxford Uni-
versity Press, 1994, pp. 473–97.
8
Stanley F J, Watson L. Trends in perinatal mortality and cerebral
palsy in Western Australia, 1967 to 1985. BMJ 1992; 304:
1658–63.
 DISABILITIES IN SCHOOL-AGE CHILDREN
9 20
Stanley F J, Watson L. The cerebral palsy in Western Australia:
trends, 1968 to 1981. Am J Obstet Gynecol 1988; 158:
89–93.
10
Stanley F J. Spastic cerebral palsy: changes in birthweight and
gestational age. Early Hum Dev 1981; 5: 167–78.
11
Rumeau-Rouquette C, du Mazaubrun C, Verrier A, Mlika A.
Prévalence des Handicaps Evolution dans Trois Généra-
tions d’Enfants 1972, 1976, 1981. Paris: INSERM, 1994.
12
Rumeau-Rouquette C, du Mazaubrun C, Mlika A, Dequae L.
Motor disability in children in three birth cohorts. Int J
Epidemiol 1992; 21: 359–66.
13
Mlika A, du Mazaubrun C, Rumeau-Rouquette C. Prévalence
des retards intellectuels sévères et des trisomies 21 dans
3 générations: 1972, 1976 et 1981. Rev Epidémiol Santé
Publique 1993; 41: 44–52.
14
Mises R. Classification Française des Troubles Mentaux de
l’Enfant et de l’Adolescent. Présentation générale et mode
d’utilisation. CTNERHI, Ministère des Affaires Sociales et
de la Solidarité. Paris, 1990.
15
Kiely M. The prevalence of mental retardation. Epidemiol Rev
1987; 9: 194–218.
16
Armitage P, Berry G. Statistical Methods in Medical Research.
London: Blackwell Scientific Publications, 1994, pp. 402–11.
17
EUROCAT working group. Surveillance of Congenital Anomal-
ies in Europe 1980–1992. Brussels: EUROCAT report,
1995.
18
International Clearinghouse for Birth Defects Monitoring Sys-
tems. Congenital Malformations Worldwide. Amsterdam:
Elsevier, 1991.
19
Kushlick A, Cox G R. The epidemiology of mental handicap.
Dev Med Child Neurol 1973; 15: 748–59.
145
Drillien C M, Jameson S, Wilkinson E M. Studies in mental
handicaps. Part 1: Prevalence and distribution by clin-
ical type and severity of defect. Arch Dis Child 1966; 41:
528–38.
21
Lagergren J. Children with motor handicaps, epidemiological,
medical and socio-paediatric aspects of motor handicapped
children in a Swedish County. Acta Paediatr Scand 1981;
(Suppl.): 289.
22
Rantakallio P. The longitudinal study of the Northern Finland birth
cohort of 1966. Paediatr Perinat Epidemiol 1988; 2: 59–88.
23
Riise R, Flage T, Hansen E, Rosenberg T. Visual impairment in
Nordic children. I. Nordic registers and prevalence data.
Acta Ophthalmol 1992; 70: 145–54.
24
Oxford Register of Early Childhood Impairments. Annual
report. Oxford: Level 3, Women’s Centre, Oxford Radcliffe
Hospital, 1994.
25
Fombonne E, du Mazaubrun C. Prevalence of infantile autism in
four French regions. Soc Psychiatry Psychiatr Epidemiol
1992; 27: 203–10.
26
Goujard J, de Vigan C, Vodovar V. Le Registre des Malforma-
tions Congénitales de Paris. Evolution 1981–1992. Rap-
port. Paris, 1995.
27
Fryers T, Mackay R I. The epidemiology of severe mental hand-
icap. Early Hum Dev 1979; 3: 277–94.
28
Krägeloh-Mann I, Hagberg G, Meisner C et al. Bilateral spastic
cerebral palsy—a comparative study between south-west
Germany and western Sweden. II: Epidemiology. Dev Med
Child Neurol 1994; 34: 473–83.
(Revised version received June 1996)