Journal of Intellectual Disability Research

179
    pp –  

Children with intellectual disability in rural South Africa:

prevalence and associated disability
A. L. Christianson,1 M. E. Zwane,1 P. Manga,1 E. Rosen,2 A.Venter,3 D. Downs4 &
J. G. R. Kromberg1
1 National Health Laboratory Service and Division of Human Genetics, School of Pathology, Faculty of Health Sciences,
University of the Witwatersrand, Johannesburg, South Africa
2 Department of Paediatrics, University of the Witwatersrand, Johannesburg, South Africa
3 Department of Paediatrics, University of the Free State, South Africa.
4 Department of Speech Pathology and Audiology, University of the Witwatersrand, Johannesburg, South Africa

Abstract children in this population. The prevalence of

severe and mild ID was . per  and . per
Background The objective of the present study was
 children, respectively. The male:female ratio of
to determine the prevalence of intellectual disability
children with ID was :. In the affected children, a
(ID) and its associated disabilities in rural South
congenital aetiology for the ID was determined in
African children aged – years. It was undertaken
 subjects (.%), an acquired aetiology in 
in eight villages in the district of Bushbuckridge,
(.%) and the aetiology was undetermined in 
Northern Province, South Africa.
children (.%). Epilepsy (.%) and cerebral
Methods A two-phase design was utilized.
palsy (.%) were the commonest associated
The first phase involved screening children on a
disabilities.
house-to-house basis by interviewing mothers or
Conclusion The present study represents the first
caregivers using an internationally validated ques-
data on the prevalence of ID and associated disabil-
tionnaire for detecting childhood disability in devel-
ities in rural South African children. The prevalence
oping countries. The second phase consisted of a
of ID was comparable with results from a study
paediatric/neurodevelopmental assessment of the
performed in one other African country (Zambia)
children who screened positive.
as well as those from other developing countries.
Results A total of  children were screened;
The data provide an initial factual insight into ID
 (.%) had a paediatric evaluation and
and its associated disabilities for healthcare, social
 children were diagnosed with ID, giving a
service and educational policy planners. This study
minimum observed prevalence of . per 
provides a basis for the initiation and development
of appropriate and integrated services for the best
possible care of individuals affected with these dis-
Correspondence: A. L. Christianson, National Health Laboratory abilities, and for their possible prevention.
Service and Division of Human Genetics, School of Pathology,
Faculty of Health Sciences, University of the Witwatersrand, PO Keywords childhood intellectual disability, preva-
Box , Johannesburg , South Africa. lence, rural, South Africa
©  Blackwell Science Ltd
 Journal of Intellectual Disability Research
180
A. L. Christianson et al. • Intellectual disability in rural South Africa
Introduction
Intellectual disability (ID) is a serious and
lifelong disability that places a heavy burden on
affected individuals, their families, society and
the healthcare system. The prevalence of ID and
associated epidemiological factors has been well
documented in industrialized countries. Con-
versely, only limited information is currently avail-
able from developing nations, where it is estimated
that between  and  million children with ID
may presently reside (Shah ; Roeleveld et al.
).
Until recently, infant and childhood mortality
and morbidity in developing nations has been
caused predominantly by infectious diseases and
malnutrition. However, in many of these countries,
the implementation of programmes like GOBI-FFF
and the integrated management of childhood ill-
nesses (IMCI) have been successful in addressing
these problems, resulting in declining infant and
childhood mortality (Cash et al. ; WHO ;
UNICEF ). Consequently, problems like ID,
which previously received limited attention from
healthcare planners, have begun to emerge and
demand due consideration. South Africa is one
country in which this epidemiological transforma-
tion is, in the present authors’ opinion, beginning to
occur. The rights of people with disabilities have
also recently been enfranchised in the country’s
new Constitution, enabling the disabled, and those
representing them, to compete more equitably for
healthcare funding (RSA ). The collection of
epidemiological data on ID is essential so that
appropriate healthcare, social welfare and educa-
tional strategies can be planned and implemented
for the benefit of affected people. In addition, such
statistics will be of value as indicators of general
health status as the significance of indicators like
infant and childhood mortality declines (Durkin
et al. ).
The aim of the present study was to document
the prevalence of ID and the disabilities associated
with it in –-year-old rural South African children.
It forms part of a broader study in which the pre-
valence and nature of five specific childhood
disabilities (i.e. ID, motor disability, auditory
disability, visual disability and epilepsy) were
investigated.
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
Subjects and methods
The present study was completed over a .-year
period from  to  in the Mhala district of
Bushbuckridge, Northern Province, South Africa.
The area is adjacent to the Kruger National Park
in the north-east corner of the country. The demo-
graphic features of the Mhala district, as described
by Tollman et al. (), include a population of
  mainly Tsonga-speaking people, who reside
in  rural villages between Acornhoek in the
North and the Sabie River in the South. Three
rural hospitals, supported by satellite clinics in sur-
rounding villages, serve the health needs of the
population.
The subjects for the present study were children
aged – years, who were recruited in eight selected
villages in the Mhala district. The villages were
spread geographically across the district and a
community-based rehabilitation worker was present
in each village.
A standard two-phase design was used to under-
take the research, using the internationally validated
Ten Question Questionnaire (TQQ), a research tool
that was developed and has been successfully used
in developing countries, to screen for five disabil-
ities (i.e. ID, motor disability, visual disability, audi-
tory disability and seizures) in –-year-old children
(Stein et al. ; Zaman et al. ; Durkin et al.
; Thorburn et al. b, c). In the first phase,
a woman resident in each village was selected
and trained to administer the TQQ to mothers
and caregivers on a house-to-house basis. Quality
control interviews were carried out at random to
ensure the accuracy of data collection. Data
obtained from the TQQ were collected and com-
puterized. In the second phase, children who
screened positive on the TQQ were referred for
paediatric evaluation at clinics held in each village.
These paediatric assessments included a presenting
history, family history, birth history, developmental
history and clinical examination, and were under-
taken by two neurodevelopmental paediatricians
and a clinical geneticist with neurodevelopmental
expertise. The results of the paediatric evaluations
were recorded on a standard medical assessment
form (MAF; Durkin et al. ). After completing
the examination of the children, a neurodevelop-
mental assessment using the Griffiths Scale of
 Journal of Intellectual Disability Research
181
A. L. Christianson et al. • Intellectual disability in rural South Africa
Mental Development (GSMD) was performed on
those children whom the clinicians considered were
at risk of being developmentally delayed (Griffiths
). For these children, a general intelligence
quotient (GIQ) was thus obtained. Visual testing,
using a Snellen notation of visual acuity chart, was
undertaken on those children capable of undergo-
ing the test. Children considered to be at risk for a
hearing problem after clinical assessment by the
clinicians were referred to a subsequent auditory
assessment clinic that was undertaken in the eight
villages as part of the present study. Auditory evalu-
ation included otoscopy, followed by imittance
testing in both ears using tympanometry and ipsi-
lateral acoustic reflex threshold testing. Air and
bone conduction thresholds were measured with
pure tone audiometry, using play audiometric tech-
niques when required.
Utilizing the information obtained from the
TQQ, MAFs, and neurodevelopmental, visual and
auditory assessments, a final diagnosis was derived.
In those children with a developmental disability,
the disability was classified as severe ID (SID)
where the GIQ was £ and mild ID (MID) where
it was –. Classification of the hearing disabil-
ities into mild, moderate and severe was done
according to the criteria of Thorburn et al. (a).
All of these data were then collated and analysed.
Ethical approval to undertake this research had
been granted by the Committee for Research
on Human Subjects of the University of the
Witwatersrand, Johannesburg, South Africa.
Results
In phase , a total of , –-year-old children
were screened using the TQQ by fieldworkers in the
Males Females
Age (years) n % n %
2–5 9 20.9 12 27.9
6–9 17 39.5 5 11.6
Total 26 60.5 17 39.5
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
eight villages. The subjects came from  house-
holds and  (.%) were aged – years, while
 (.%) were – years old. Most of these
children (n = , .%) had been born in hospi-
tal, a further  (.%) were born in a clinic or
health centre and  (.%) were born at home.
Out of the  children screened by the TQQ, 
(.%) had a paediatric evaluation in phase  of
the study.
Intellectual disability was diagnosed in 
(.%) children, giving a ID prevalence of . per
 (one per ) children in the present sample.
Severe ID was present in  (.%) children,
giving a prevalence of . per  children. There
were  (.%) children with MID, giving a
prevalence of . per , and the MID: SID
ratio was . : . The age and sex characteristics of
the children with SID and MID are detailed in
Tables  and . For both groups, the ratio of
affected males to females closely approximated  : .
Aetiology for ID was determined in only 
(.%) cases. The cause was found to be congeni-
tal in  (.%) of these patients, acquired in 
(.%) and undetermined in  (.%). For the
 children with SID, the aetiology could be deter-
mined in  (.%) children and was congenital in
origin in  (.%) of these cases. By comparison,
in children with MID, the aetiology could only be
determined in  (.%) out of the  patients
(Table ).
Epilepsy was associated with the ID in 
(.%) children (Table ). Generalized seizures
were recorded in  (.%) subjects and the
remaining two (.%) suffered from simple partial
seizures.
Among the  patients with SID and epilepsy,
eight (.%) were receiving antiepileptic drugs
(AEDs), while only seven out of the  (.%)
Table 1 Age and sex of children with
Total severe intellectual disability (male : female
ratio :)
n %
21 48.8
22 51.2
43 100
 Journal of Intellectual Disability Research      
182
A. L. Christianson et al. • Intellectual disability in rural South Africa

Table 2 Age and sex of children with

Males Females Total mild intellectual disability (male : female
ratio  : )
Age (years) n % n % n %

2–5 38 19.5 21 10.8 59 30.3

6–9 82 42 54 27.7 136 69.7
Total 120 61.5 75 38.5 195 100

Table 3 Aetiology of intellectual

SID (n = 43) MID (n = 195) Total (n = 238) disability (ID): (SID) severe ID; and
(MID) mild ID (SID : MID ratio . : )
Aetiology n % n % n %

Congenital 22 51.2 27 13.9 49 20.6

Acquired 4 9.3 11 5.6 15 6.3
Undetermined 17 39.5 157 80.5 174 73.1
Total 43 100 195 100 238 100

Table 4 Disorders associated with

SID MID Total intellectual disability (ID): (SID) severe
(n = 43) (n = 195) (n = 238) ID; and (MID) mild ID

Disorder n % n % n %

Epilepsy 11 25.6 26 13.3 37 15.5

Cerebral palsy 11 25.6 9 4.6 20 8.4
Auditory disability 2 4.6 15 7.7 17 7.1
Microcephaly 9 20.9 2 1.0 11 4.6
Down’s syndrome 5 11.6 0 0 5 2.1
Hydrocephaly 2 4.6 2 1.0 4 1.7
Undiagnosed syndromes 2 4.6 2 1.0 4 1.7
Macrocephaly 0 0 3 1.5 3 1.3
Spina bifida 0 0 2 1.0 2 0.8
Tuberous sclerosis 1 2.3 1 0.5 2 0.8

children with MID and epilepsy were receiving
AEDs.
Cerebral palsy was documented in  (.%)
children with ID. It was recorded in  (.%)
children with SID and was less frequent (n = ,
.%) in children with MID (Table ). Micro-
cephaly was documented in  (.%) children with
ID, and in six of these subjects, the condition was
considered to be congenital in origin. Thus, congen-
ital microcephaly (six cases) had a minimum preva-
lence of . per  children. Down’s syndrome
(DS) was diagnosed in five children, giving a
prevalence of . per  children (Table ).
The diagnosis in these DS children was confirmed
by chromosome analysis. All were assessed to have
GIQs <  (mean = ) and only one had a con-
genital cardiac lesion. Spina bifida was recorded in
only two children in the present study, giving a
prevalence of . per  children. These children
had had surgical closure of their spina bifida, one
had required a ventricular-peritoneal shunt for
hydrocephalus and both were MID (Table ).

©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
 Journal of Intellectual Disability Research
183
A. L. Christianson et al. • Intellectual disability in rural South Africa
Auditory disability was diagnosed in two (.%)
out of the  children with SID. In one, the hearing
loss was mild, and in the other, it was moderate.
Because of difficulties completing valid hearing
tests on all  children with SID, this is considered
to be an underestimate of the number of affected
children. Fifteen (.%) out of the  children
with MID were assessed to have a hearing loss. In
five (.%) of these children, the auditory disabil-
ity was severe, and it was moderate in one (.%)
child and mild in nine (%) others.
Discussion
Intellectual disability can be defined as a signifi-
cantly below average level of general intellectual
functioning, accompanied by significant limitations
in adaptive capability (APA ). General intellec-
tual functioning is defined by the intelligence quo-
tient (IQ), the GSMD equivalent of which is the
GIQ (Griffiths ). The GSMD was preferred for
the present study because the three clinicians were
trained in its usage and its use is accepted in South
Africa (Luiz ). Severe ID was defined as a GIQ
£  to allow for comparison between the present
findings and those of other studies (Stein et al.
). A GIQ of – was used to define MID. In
the GSMD, a GIQ of – is considered to be
borderline ID (Griffiths ). However, given the
socio-economic and educational circumstances in
rural Bushbuckridge, and by extension, a large part
of rural and urban South Africa, borderline intellec-
tual functioning will usually predispose a child to
school failure which, in these circumstances, repre-
sents a significant limitation in adaptive capability.
Thus, the decision was made to extend the upper
limit of GIQ to  for MID.
Given these definitions, an ascertained prevalence
rate of . per  children with ID was docu-
mented in the present study. There is a paucity of
data in the literature with which to compare this
figure. Stein et al. () presented data from nine
developing nations using the same methodology
applied in the present study. However, the sample
sizes for the studies in each of these countries were
significantly smaller. The prevalence of children
with ID in the nine countries ranged from nine
per  in the Philippines to  per  in
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
Bangladesh. Zambia ( per ) and Pakistan
( per ) had prevalences of childhood ID
which were similar to that found in the present
study, while the prevalence was lower in Sri Lanka
( per ) and Malaysia ( per ). Stein
et al. () noted that prevalence rates for SID of
about five per  and MID of about  per 
had been found in several of the ‘more’ developed
countries in their study, giving these countries an
overall ID rate of about  per  and an
MID : SID ratio of  : . These observations are in
keeping with those of the present study in South
Africa which, in the authors’ opinion, could be con-
sidered a ‘more’ developed country.
In a study on the prevalence of childhood disabil-
ity in Jamaica, ID was recorded in . per 
children between the ages of  and  years (Paul
et al. ). The above study utilized the TQQ, and
in addition, a statistical methodology was developed
and used to compensate for false-negatives cases
missed by the TQQ, as well as for children screened
positive but who did not present for paediatric eval-
uation (Shrout & Newman ). Consequently,
the estimated prevalence of ID in Jamaican children
was significantly greater than that recorded in the
present study. The authors were cognisant of the
fact that methodological difficulties had limited
the ascertainment of children with ID, particularly
MID, in the present study. The TQQ screening
method may not have been sensitive enough to
detect some children with MID who were otherwise
outwardly normal. This possibility was recognized
while visiting schools in the selected villages to
discuss the research project with the teaching staff.
Some children in the –-year-old age group in the
schools were reported to be experiencing repeated
school failure, but had not been screened positive
by the TQQ. For the TQQ, these children repre-
sented potential false negatives in that, despite
being considered normal by their parents/caregivers,
they were obviously at risk of having MID. Thus,
the prevalence rates in the present study for MID
and, by extension, total ID should be considered as
‘minimum’ observed frequencies.
The SID (GIQ £ ) prevalence in the present
study was . per  children, which approaches
twice the rate of between three and four per 
children with SID estimated to be present in indus-
trialized nations. However, it is lower than the cal-
 Journal of Intellectual Disability Research
184
A. L. Christianson et al. • Intellectual disability in rural South Africa
culated average . per  children with SID in
developing countries (Roeleveld et al. ). In
eight out of the nine developing countries docu-
mented by Stein et al. (), the prevalence of
SID ranged from five per  to  per 
children, and was unusually high in India ( per
). Numerous methodological problems com-
plicate the calculation of accurate figures for MID
in both industrialized and developing countries.
Roeleveld et al. () calculated an average preva-
lence rate of MID of . per  children in
developing countries. This is almost the same as the
prevalence rate for MID (. per ) obtained
in the present study, but as noted above, this figure
is almost certainly an underestimation of the actual
situation.
For ID, SID and MID the male: female ratio of
affected children in the present study was : ,
giving an overall % excess of affected males. A
preponderance of ID males secondary to X-linked
genetic factors is to be expected in any community.
In industrialized countries, the excess of males with
SID approximates %. A possible reason for the
discrepancy in the excess of SID males to females
observed between the present study and that
recorded in developed countries may be that male
children with SID received better care than affected
female children in the community studied. This
may have resulted in greater longevity of males with
SID. In addition, it is possible that caregivers were
more likely to bring their affected male children to
the attention of the research team. The excess of
males with MID in industrialized nations varies
between % and %, similar to that found in the
present study (Roeleveld et al. ).
Aetiology for the ID was documented for just
in excess of one-fourth (.%) of cases, of
which .% were congenital and .% acquired.
However, as might be expected, there were large
and obvious differences in the aetiology of SID and
MID in affected children. Aetiology was determined
in .% of children with SID, in half of whom
(n = , .%) it was considered to be congenital.
An obviously acquired aetiology was present in only
four (.%) SID patients and the aetiology could
not be determined in  (.%) patients. By com-
parison, no definite aetiology was found in more
than twice as many children with MID (n = ,
.%), while  (.%) cases had a congenital
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
aetiology and  (.%) cases had an acquired
one. The acquisition of a defined aetiology for ID
was limited in the present study by the lack of
investigative facilities. Curry et al. () reviewed
the causes of ID reported in literature surveys
and found that the proportion of cases with
unknown aetiology was –% in several
different populations.
Previous studies have documented a clear associ-
ation between social class, ethnicity and/or parental
occupation, and MID, with poor rural areas and
urban ghettos being singled out as areas in which
these problems are potentially greatest (Drews et al.
; Roeleveld et al. ). The Bushbuckridge
district represents a socio-economically deprived
rural area. Factors observed in Bushbuckridge by
the present authors which are thought to be associ-
ated with a high rate of MID include poor living
conditions, malnutrition, limited intellectual stimu-
lation of infants and children, and unattended
home births (.%). Thus, the present authors
would propose that, for many of the  (.%)
MID children with an undefined aetiology, their
problem could have been the result of, or aggra-
vated by, their environment. Many of these factors
could potentially be avoided with the improvement
of social circumstances, maternal and child diet,
maternal education, and early infant and childhood
intellectual stimulation. Having noted this, the
present authors are also cognisant of the fact that
other specific aetiological factors could have been
present in this group, including the sequelae of
endemic diseases (e.g. HIV/AIDS), specific nutri-
tional deficiencies (e.g. iodine), maternal intoxica-
tions (e.g. alcohol) and sub-optimal obstetric care.
These factors also need to be investigated and
addressed.
Epilepsy, the most frequent disorder associated
with ID, was recorded in  (.%) patients. It was
twice as common in SID children (n = , .%)
than it was in those with MID (n = , .%).
However, it was notable that the majority of the
children with epilepsy and SID (.%) were on
AEDs. In comparison, only .% of the children
with epilepsy and MID were receiving such medica-
tion. The children with SID and epilepsy had
multiple, obvious and severe problems which
probably led to their parents seeking care for them
from the local Western medical system, and thus,
 Journal of Intellectual Disability Research
185
A. L. Christianson et al. • Intellectual disability in rural South Africa
obtaining AEDs. The fact that so few children with
MID and epilepsy obtained AEDs is cause for
concern. The reasons why management from the
local Western medical system was not sought and/or
obtained by many children in this population suf-
fering from epilepsy requires further investigation
and amelioration. It also remains a point of conjec-
ture as to what role the untreated seizures in the
children with MID and epilepsy had in the genesis
of their MID (Christianson et al. ).
Cerebral palsy (n = , .%) was the second
most common disorder associated with ID, and as
expected, was more frequent in children with SID
(.%) than in those with MID (.%). It was fol-
lowed in rank order of associated disorders by audi-
tory disability (.%) and microcephaly (.%).
Out of the  children with microcephaly, at least
six were considered to have a prenatal cause. Using
the latter figure, the minimum prevalence of con-
genital microcephaly was . per  children,
which is higher than the . and . per 
birth incidences recently recorded in African infants
in an urban academic hospital and in a rural
Northern Province hospital, respectively (Delport
et al. ; Venter et al. ). Possible reasons for
this difference between the birth incidence and
prevalence of congenital microcephaly is under-
ascertainment at birth, and children with micro-
cephaly, despite their disabilities and the harsh
circumstances of their environment, seem to
manage to survive, and therefore, the mortality as
the child ages is possibly less than would be
expected.
Only five children were diagnosed with DS in the
present study, all with SID. Thus, the prevalence for
DS in –-year-old children in Bushbuckridge was
. per  (one per ). A birth incidence of
. per  live births (one per ) in a rural
hospital approximately  km from Bushbuckridge
has been recorded, suggesting that two out of three
children with DS die in this area prior to  years of
age. In the same rural hospital, the birth incidence
of neural tube defects was . per  and that of
spina bifida was . per  live births (Venter
et al. ), whereas in the present community-
based study only . per  children had spina
bifida and no children were diagnosed with
encephalocoeles. These figures highlight the mor-
bidity of DS and neural tube defects in infancy and
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
early childhood in rural South Africa. It is well
accepted that infants and children with severe con-
genital anomalies living in developing countries
tend to die early, and their deaths are either
unrecorded or are absorbed into the general mor-
tality statistics, thus passing unrecognized for what
they truly represent (WHO ). The above figures
on these two common severe congenital anomalies
in South Africa, i.e. DS and neural tube defects,
represent some of the first data from Africa that
have quantified this reality.
In conclusion, the present authors submit the
first data on the prevalence and disabilities associ-
ated with ID in rural South African children. Since
Bushbuckridge is probably a fairly typical socio-
economically deprived rural area and because a
sizeable sample of children was investigated over a
widespread geographic area, these results could be
considered to be representative of much of rural,
Central and Southern Africa. Thus, these findings
provide an initial factual insight for healthcare,
social service and educational policy planners. In
the immediate future, they have implications for
the provision of primary-healthcare-based medical
services for pregnant woman, infants and children,
pre-school education, community-based rehabilita-
tion services for affected individuals and their fami-
lies, remedial education facilities in schools, and
social welfare services. By improving and integrat-
ing such services, it should be possible to ensure
the best possible care for affected individuals and
services for their prevention, thus reducing the
prevalence of ID, especially MID.
Acknowledgements
The authors would like to acknowledge the coop-
eration and patience of the people in the commu-
nity, and especially, the children and their mothers
in the eight villages in Bushbuckridge, without
whom this work could not have been accomplished.
We are also indebted to the local field workers,
the community-based rehabilitation workers, the
Tintswalo hospital nurses, and the clinic sisters
for their cooperation, hard work and hospitality.
Funding for the research was obtained from the
Iris Ellen Hodges Trust and the University of the
Witwatersrand Medical Faculty Endowment Fund.
 Journal of Intellectual Disability Research
186
A. L. Christianson et al. • Intellectual disability in rural South Africa
We also acknowledge: the Wits Rural Facility, who
assisted with accommodation and infrastructure
during the fieldwork; Dr Steve Tollman and his
staff at the Health Systems Development Unit
of the University of the Witwatersrand; and the
Department of Speech Pathology and Audiology
of the University of the Witwatersrand.
References
American Psychiatric Association (APA) () Diagnostic
and Statistical Manual of Mental Disorders (DSM-IV).
American Psychiatric Association, Washington, DC.
Cash J., Kenesch G. T. & Lamstein J. () Child Health
and Survival. The UNICEF/GOBI-FFF Program. Croom-
Helm, London.
Christianson A. L., Zwane M. E., Manga P., Rosen E.,
Venter A. & Kromberg J. G. R. () Epilepsy in rural
South African children – prevalence, associated disabil-
ity and management. South African Medical Journal ,
–.
Republic of South Africa (RSA) () Constitution of the
Republic of South Africa. Act  of . Government
Printers, Pretoria
Curry C. J., Stevenson R. E., Aughton D., Byrne J., Carey
J. C., Cassidy S., Cunniff C., Graham Jr J. M., Jones
M. C., Kaback M. M., Moeschler J., Schaefer G. B.,
Schwartz S., Tarleton J. & Opitz J. () Evaluation of
mental retardation: recommendations of a consensus
conference. American Journal of Medical Genetics ,
–.
Delport S. D., Christianson A. L., Van den Berg H. J. S.,
Wolmarans L. & Gericke G. S. () Congenital
anomalies in black South African liveborn neonates at
an urban academic hospital. South African Medical
Journal , –.
Drews C. D., Yeargrin-Allsopp M., Decoufle P. & Murphy
C. C. () Variation in the influence of selected
sociodemographic risk factors for mental retardation.
American Journal of Public Health , –.
Durkin M., Zaman S., Thorburn M., Meher H. &
Davidson L. () Population-based studies in child-
hood disability in developing countries. International
Journal of Mental Health , –.
Griffiths R. () The Abilities of Young Children. ARICD,
Amersham.
Luiz D. M. (ed.) () Griffiths Scales of Mental
Development: South African Studies. Research Paper C,
University of Port Elizabeth, Port Elizabeth.
Paul T. J., Desai P. & Thorburn M. J. () The preva-
lence of childhood disability and related medical diag-
©  Blackwell Science Ltd, Journal of Intellectual Disability Research , –
     
noses in Clarendon. Jamaica West Indian Medical Journal
, –.
Roeleveld N., Veelhuis G. A. & Gabreels F. () The
prevalence of mental retardation: a critical review of
recent literature. Developmental Medicine and Child
Neurology , –.
Shah P. M. () Prevention of mental handicaps in chil-
dren in primary health care. Bulletin of the World Health
Organization , –.
Shrout P. E. & Newman S. C. () Design of two-phase
prevalence surveys of rare disorders. Biometrics ,
–.
Stein Z., Belmont L. & Durkin M. () Mild mental
retardation and severe mental retardation compared:
experiences in eight less developed countries. Upsala
Journal of Medical Science  (Suppl.), –.
Thorburn M. J., Desai P. & Davidson L. L. (a)
Categories, classes and criteria in childhood disability
– experience from a survey in Jamaica. Disability and
Rehabilitation , –.
Thorburn M., Desai P., Paul T. J., Malcolm L., Durkin
M. & Davidson L. (b) Identification of childhood
disability in Jamaica: the ten question screen. Interna-
tional Journal of Rehabilitation Research , –.
Thorburn M., Desai P., Paul T. J., Malcolm L., Durkin
M. & Davidson L. (c) Identification of childhood
disability in Jamaica: evaluation of the ten question
screen. International Journal of Rehabilitation Research ,
–.
Tollman S., Herbst K. & Garenne M. () The
Agincourt Demographic and Health Study. University of
the Witwatersrand, Johannesburg.
United Nations Children’s Fund (UNICEF) () The
State of the World’s Children. United Nations Children’s
Fund, New York, NY.
Venter P. A., Christianson A. L., Hutamo C. M., Makhura
M. P. & Gericke G. S. () Congenital anomalies in
rural Black South African neonates – a silent epidemic?
South African Medical Journal , –.
World Health Organization (WHO) () Improving
Child Health. Integrated Management of Childhood
Illnesses. WHO/CHD/., World Health Organization,
Geneva.
World Health Organization (WHO) Advisory Group
() Community Approaches to the Control of Hereditary
Diseases. World Health Organization, Geneva.
Zaman S. S., Khan N. Z., Islam S., Banu S., Dixit S.,
Shrout P. & Durkin M. () Validity of the Ten
Questions’ for screening serious childhood disability:
results from urban Bangladesh. International Journal of
Epidemiology , –.
Received  September ; revised  May 