Chapter 10:  Developmental Disabilities

The International Classification of Functioning

In 1980 the World Health Organization (WHO) published the International Classification of Impairments, Disabilities and Handicaps (WHO 1980) as a companion classification to the International Classification of Diseases to document the consequences of congenital or acquired illnesses with nonfatal outcomes (figure 10.1). These consequences were evident in impairments, defined as abnormalities of body or organ structures and functions; disabilities, defined as reduction of a person's ability to perform basic and everyday tasks; and handicaps, defined as a person's disadvantage in fulfilling social roles. This framework suggested a hierarchical and linear relation between these three dimensions that entailed a problem-solving sequence in which interventions at one level have the potential to modify succeeding elements (Fryers 1992).

Two of the most important criticisms of this classification that are particularly relevant for this review are its failure to adequately cover disabilities affecting children and its limited utility for public health applications. A closely related criticism has been its failure to take into account the pervasive role of the environment in exacerbating or reducing the nature and extent of disablement.

In its revised International Classification of Functioning, Disability and Health the WHO (2001) has sought to address some of these criticisms and, in particular, to unify the apparent polarities in the medical and social models of disability (figure 10.2). In the International Classification of Functioning a person's functioning and disability is conceived as a dynamic interaction between health conditions (diseases, disorders, injuries, and trauma) and the effect of contextual and environmental factors in limiting activity and restricting participation.

The Effect of Poverty

In developing countries serious developmental disabilities represent only a proportion of the poor developmental outcomes of children and young adults. The neurological and developmental deficits that have their primary origin in adverse social and environmental conditions, such as poverty, poor nutrition, and social deprivation, during the critical years of early brain growth and development usually present later in childhood as cognitive impairments and poor performance at school. This group of at-risk infants and children is likely to exceed by many times the number of children with readily identifiable developmental disabilities. Because the contribution of this group to the burden of developmental disabilities is difficult to measure and the interventions are more closely linked to broad community development and poverty alleviation strategies, they will not be considered, except in passing, in this chapter. This in no way implies that they have any less a priority in policies and programs that address developmental disabilities in low-income countries.

The Health Transition

Infant and child mortality has steadily declined in Sub-Saharan Africa countries during the second half of the twentieth century (Ahmad, Lopez, and Inoue 2000; Hill et al. 1999) along with an accompanying fall in fertility (Cleland, Onuoha, and Timaeus 1994; Cohen 1993). This is a result of a complex process of social and economic change in these countries coupled with sustained implementation of family planning programs (United Nations Population Division 1998) and primary health care interventions (Fauveau et al. 1990; Grant 1992; Velema et al. 1991) that specifically target preventable diseases in infancy and childhood.

This "demographic and epidemiological transition" has occurred at varying rates in different African countries. Infant mortality trends during the health transition in developed countries have shown large absolute decreases in postneonatal mortality accompanied by increases in the proportional contribution of neonatal mortality to overall infant mortality (MacFarlane and Mugford 1984). These shifts in the ratio of neonatal to postneonatal mortality have also occurred in countries in Sub-Saharan Africa with the lowest infant mortality rates. Neonatal mortality now accounts for 45 percent of all infant deaths in South Africa (SADHS 1998, 7). These demographic changes suggest that a higher priority should be given to programs that address perinatal conditions. They also signify a need to move away from an exclusive preoccupation with child survival strategies to those that aim to improve the quality of survival in the 92 to 95 percent of children now surviving beyond five years in many African countries. Pointing to pregnancy and the neonatal period targets high-risk groups not only for residual infant mortality but also for developmental disabilities, because many of the most important causes of these disabilities are addressed by interventions during this period.

The Impact of HIV

Arguments for the assignment of a higher priority to research and programs that address developmental disabilities in low-income countries are regularly based on evidence for the health transition in these countries. These transitions, however, are generally uneven both between and within countries, and this is particularly true of the African subcontinent. The "unfinished agenda" of infectious diseases and malnutrition continues to make a variable contribution to postneonatal mortality and morbidity in the different countries and their subpopulations, depending on social, economic, and political conditions and the coverage and quality of primary health care services. Many African countries are now reporting rising postneonatal mortality with patterns of disease often indistinguishable from those in the previous decade (Ahmad et al. 2000), and in many cases, the increase is attributed to the human immunodefi-ciency virus (HIV) epidemic (Nicoll et al. 1994; United Nations Population Division 1999). Recent evidence suggests that the proportion of under-five mortality attributable to HIV and the acquired immune deficiency syndrome (AIDS) in Sub-Saharan Africa is quite variable (0.1 to 42.4 percent) and that it is highest in some of the countries with the steepest reductions in postneonatal mortality in the past two decades (Walker, Schwartlander, and Bryce 2002).

These data suggest that programs to tackle developmental disabilities may have a greater claim on national resources in some Sub-Saharan Africa countries than in others during this complex health transition. But even in countries with high under-five mortality and HIV seroprevalence rates, 8 or 9 out of every 10 children will survive beyond five years. Many of them will continue to be at risk for developmental disabilities because of preventable biological factors and the lack of services and programs to identify, treat, and ameliorate the impact of these disabilities. Additionally, as will be discussed later, children with HIV infection may be at special risk for developmental disabilities.

Measuring the Burden of Developmental Disabilities

Developmental disabilities, because of their early onset and lifelong requirement for support and care, impose enormous social and economic burdens on affected individuals, their families, and their communities. Calculation of the loss of disability-adjusted life years (DALYs) is now the method widely used to capture and compare the combined effect on the global burden of disease (GBD) of premature mortality and decreased functional capacity resulting from designated health conditions (Murray 1994). An important obstacle to the measurement of the burden of developmental disabilities in low-income countries is the lack of good quality prevalence data. An additional concern stems from the characteristics and applications of the measurement itself.

In burden-of-disease calculations the years of life lost for each death are estimated and assigned a relative value based on the age at death (age weighting). Age weighting is intended to reflect differential productivity of an individual at different stages of his or her life cycle and, thereby, assigns a different social value to lives lived at different ages. A year lived at age 2 counts for only 20 percent of a year lived at 25, when the age-weighting function is at a maximum. The effect of this is to reduce the DALYs lost by premature death of children with developmental disabilities and to potentially lower the relative importance of these conditions in decisions about resource allocations (Anand and Hanson 1997).

To measure the additional contribution of disability, the number of years of healthy life lost is estimated by multiplying the expected duration of the condition (to remission or death) by a disability weighting (0–1). Disability weighting does not take into account the way that individual and social resources can compensate for the level of disability experienced. The failure to factor in these compensatory mechanisms distorts burden-of-death estimates and undervalues the therapeutic and rehabilitation benefits offered by rehabilitation professionals and programs (Jelsma, De Weerdt, and De Kock 2002).

Developmental disabilities incorporate many different disease entities with their different causes; time of onset; natural histories; amenability to primary, secondary, and tertiary prevention; and the relative risk of premature mortality and functional limitation. Since the burden-of-disease methodology is based on calculations for individual disease entities, each of these would need to be addressed separately in order to arrive at a burden-of-disease estimate for developmental disabilities as a whole. Without this information, more pressing and immediate issues, such as HIV/AIDS and malaria, will dominate the health agenda and displace developmental disabilities from consideration for resource allocation.

Two-Phase Childhood Disability Surveys

The best available standard for the measurement of the prevalence of developmental disabilities in population-based surveys is the two-phase survey method, which has been validated in many different population settings in the developing world (Durkin et al. 1994; Thorburn, Desai, and Davidson 1992; Zaman et al. 1990). The first phase consists of a survey of all households in a drawn sample or target population in which all children are screened using the "ten questions" questionnaire. In the second phase all children screening positive and a random 10 percent sample of those screening negative undergo a detailed medical and psychological examination from which an eti-ological diagnosis is made and disability rated (none, mild, moderate, and severe) in each of the following areas: gross motor, fine motor, vision, hearing, cognition, speech, and seizures.

Two-phase surveys of this kind have been shown to have a high sensitivity but a low positive predictive value. They are, therefore, ideally suited to serve as screening instruments but completely inappropriate as case-finding tools for epidemiological studies or as the basis for referring children with disabilities for rehabilitation services. Using the ten-question screen as the only basis for determining prevalence of disabilities has been shown to overestimate prevalence by up to 300 percent (Durkin, Hasan, and Hasan 1995).

Many clinical researchers have set the ethical requirement that all children screening positive in a target population should be included in the second phase so that they can be referred to medical and rehabilitation services. Although no similar imperative applies to those screening negative, the inclusion of negatives is necessary (thus the minimal sample of 10 percent) because, using this method, the prevalence estimate for the total population is obtained as a weighted average of the rates in those screening positive versus those screening negative (Shrout and Newman 1989).

Disability Prevalence Data in Sub-Saharan Africa

Reliable data based on the criteria set above on the prevalence of developmental disabilities is scarce in the African subcontinent. An initial Medline search was combined with a search of two major specialist disability journals, Disability and Rehabilitation and The International Journal of Rehabilitation Research, for the years 1995 to 2004 and a detailed review of all peer-reviewed publications within which the two-phase methodology is employed to measure the prevalence of developmental disabilities in developing countries. In the relatively few studies identified in Sub-Saharan Africa, not a single one fulfilled the methodological criteria laid out above for two-phase disability studies.

Most of the identified studies were conducted in South Africa. The prevalence rates for all categories of developmental disability in these studies varied from 11 to 60 per 1,000 children (Cornielje et al. 1993; Couper 2002; Irlam 1996; Katzenellenbogen et al. 1995; Schneider 1997). Although two-phase surveys were used in most studies, the phase 1 screening questionnaires were not standardized in all cases and were not validated for use in African populations in any of the studies (table 10.1).

A more serious limitation is the variation in the methods or content of the evaluation in the second phase. These clinical assessments were not standardized, and they varied from repeat questionnaires by fieldworkers to assessments by rehabilitation assistants, rehabilitation professionals, and neurodevelopmental pediatricians. Additionally, for valid calculations of prevalence in two-phase studies a 10 percent sample of individuals screening negative should be included in the second phase. This did not occur in any of the cited studies, nor was it always clear from the description of the methods how prevalence rates were derived. A final concern is the lack of a standardized approach in most studies to the grading of severity of disability. Since it is clear from carefully conducted studies elsewhere that the ratio between severe and mild disabilities may be very high, comparisons between studies that include children of differing severity may lead to erroneous conclusions about the true size and nature of the problem in different areas.

A tabulation of some of the two-phase studies conducted elsewhere in the developing world is provided as table 10.2 for comparison and similarly shows wide variations in prevalence rates for all categories of disability in childhood from different studies (Durkin et al. 1994; Durkin, Hasan, and Hasan 1995; Islam, Durkin, and Zaman 1993; Milaat et al. 2001; Natale et al. 1992; Tamrat et al. 2001; Zaman et al. 1990). The studies in Bangladesh, Jamaica, and Pakistan that provided the range in prevalence of from 8.1 to 31.0 per 1,000 children were based on identical and established methodologies, used the same set of instruments, and relied on identical rating systems for severity. These rates in all likelihood reflect true differences in the prevalence of developmental disabilities in these three countries and provide a strong platform on which to begin to look at types, determinants, and causes of disability and on which to build a rational approach to intervention.

Cognitive Disabilities

Population-based prevalence data on cognitive disabilities are sparse in developing countries, and what little is available comes mainly from outside Sub-Saharan Africa. Severe mental retardation (MR), defined as a decreased general intelligence quotient of less than or equal to 55, accompanied by significant limitations in adaptive capability, is consistently found to be in the range of 3 to 5 per 1,000 persons in developed countries. In the few available studies from low-income countries, rates are significantly higher and range from 6 per 1,000 and 22 per 1,000 for severe MR and 14.5 per 1,000 and 65.3 per 1,000 for mild MR in Bangladesh (Islam, Durkin, and Zaman 1993) and Pakistan (Durkin, Hasan, and Hasan 1998), respectively. The only equivalent figures from well-designed population-based disability surveys come from Zambia (Stein, Belmont, and Durkin 1987), where rates of 5 per 1,000 were recorded for severe MR and 30 per 1,000 for mild MR, and in a recent study from rural South Africa (Christianson et al. 2002), where similar rates of 6.4 per 1,000 were noted for severe MR and 29.1 per 1,000 for mild MR. These figures are a little lower than suggested averages for severe MR in developing countries (9.3 per 1,000) but similar to suggested averages for mild MR (29.8 per 1,000; Roeleveld, Zielhuis, and Gabreels 1997). The failure to ascertain a specific biological cause in many of these children suggests that many cognitive disabilities may have their origins in maternal and infant malnutrition and impoverished environments, which have pervasive adverse effects on growth and psychological development.

Motor Disabilities

No methodologically sound studies could be found that reported population-based prevalence estimates for motor disabilities or its subtypes in Sub-Saharan Africa. One report, which lacked a detailed second-phase medical evaluation and did not include children screening negative, set the prevalence at 28 per 1,000 in a rural South African district (Couper 2002). A comparative rate of 19.5 per 1,000 for severe motor disabilities and 52.5 per 1,000 for mild motor disabilities was reported from Pakistan. There is a clear and immediate need for studies to document independent prevalence rates for motor disabilities, given its common association with severe cognitive disabilities and the observation from studies elsewhere that it is an important and common functional limitation in children with developmental disabilities (Durkin, Hasan, and Hasan 1995).

Vision Disabilities

There are estimated to be 1.5 million blind children worldwide (WHO 1992). The prevalence of blindness in children in European countries varies from 0.2 to 0.4 per 1,000 children and in African countries from 0.5 to 1.1 per 1,000 (Gilbert et al. 1999). Several studies have estimated that as much as 47 percent of blindness or severe visual impairment in developing countries is preventable or curable (Adeoye 1996; Nwosu 1998; Silver et al. 1995). Data from studies done in schools for the blind in East, central, and West African countries show that the most common causes of blindness in children are acquired diseases, such as vitamin A deficiency (29 percent) and measles (27 percent) (Gilbert et al. 1993; Gilbert et al. 1995).

Quite variable rates might be expected in countries at different stages in the health transition. A recent study on the prevalence of blindness in South African schools for the blind supports this view (O'Sullivan, Gilbert, and Foster 1997). An overall estimate of blindness prevalence was 0.35 to 0.6 per 1,000 children. Although 39 percent of causes of blindness or severe visual impairment were found to be preventable, only 5 percent of the affected children had conditions amenable to primary preventive measures, such as vitamin A deficiency or measles. Almost a quarter of the children (23 percent) had inherited conditions, intrauterine infections, or retinopathy of prematurity; the majority of these problems were potentially preventable through genetic counseling and improved antenatal and neonatal care. The remaining 11 percent needed sophisticated surgery for such conditions as cataracts of unknown origin and glaucoma.

Hearing Disabilities

Estimates of hearing loss and profound deafness in developed countries are on the order of 1 per 1,000, compared with 1.4 to 4.0 in developing countries. The WHO estimates that there are 120 million people worldwide with hearing impairment, and 78 million of those affected are in developing countries (WHO 1995b). As in other domains, there are few prevalence data on hearing disabilities in Sub-Saharan Africa. Where data are available, it is often difficult to separate methodological limitations from true differences in prevalence between these studies (WHO 1995b). Reported rates of profound hearing loss range from 2.1 per 1,000 (sensorineural loss) in Swaziland (Swart et al. 1995) to 4.0 per 1,000 in Sierra Leone (Seely et al. 1995; see table 10.3). Chronic otitis media has been demonstrated to be the most frequent cause of hearing impairment in many developing countries (Smith et al. 1996). Estimates suggest that as much as 50 to 66 percent of all hearing impairment in the developing world is preventable (Smith and Hatcher 1992).

Learning Disabilities

A learning disability is traditionally defined as a disorder in one or more of the basic psychological processes involved in understanding or in using language, spoken or written, resulting in an imperfect ability to listen, think, speak, read, write, spell, or do mathematical calculations. The definition specifically excludes learning problems that are primarily the result of visual, hearing, or motor disabilities and those resulting from mental retardation or emotional disturbance. In Sub-Saharan Africa it may be difficult to distinguish between traditional learning disabilities and the consequences of adverse social and environmental conditions.

No studies could be found that described the prevalence of learning disabilities in Sub-Saharan Africa. A Cape Town study in a population not typical of most of the rest of the subcontinent found that the origin of 45 percent of learning disabilities was prenatal, 17 percent was perinatal, 9 percent was postnatal, and about 25 percent was unknown (Molteno and Lachman 1996). The distribution between these etio-logical categories varied by ethnic group; children from white and mixed race families had relatively high prenatal contributions (55 percent) compared with children from black African families (23 percent). These relative contributions were reversed in the perinatal category; black African children contributed 37 percent, and children from mixed race groups, 8 percent. Similar distributions were found in an earlier Zimbabwean study but with a much higher proportion being of unknown cause (Axton and Levy 1974). Forty percent of all cases were considered to be preventable. Many of the children in these studies had other major disabilities, including motor, cognitive, and sensory deficits, raising doubts as to whether they fit the traditional definition for this disorder.

As developing countries go through the health transition, it seems likely that learning disabilities will become an increasingly important concern as the countries begin to make more qualitative investments in the future human potential and productivity of their populations. In a country like South Africa, where economic development and rapid political change have coincided, the great demand for access to educational opportunities has highlighted the learning deficits and, in some cases, the disabilities of many children.

Congenital Disorders

Congenital disorders are defined here as any potentially disabling condition arising before birth and including those caused by environmental, genetic, and unknown factors, whether they are evident at birth or become manifest later in life. It does not include congenital infections or nutritional factors influencing intrauterine growth, which will be discussed later.

The frequency of congenital disorders is best described in terms of birth prevalence, that is, affected births per 1,000 in the absence of a prevention program. Since there are no prevalence data from developing countries, projections are based on data from developed countries; adjustments are made for regional differences in the prevalence of hemoglobin disorders (WHO 1994), glucose-6-phosphate dehydrogenase deficiency (Luzzato and Mehta 1989), and the effect of customary consanguineous marriage (Alwan and Modell 1997), all of which have their greatest burden in less-developed parts of the world (table 10.4). The estimated rate of 61 per 1,000 in Sub-Saharan Africa, which approximates or exceeds infant mortality in some Sub-Saharan African countries, is inflated by the inclusion of fetal losses, but it does provide some indication of the increasing contribution of congenital disorders to burden of disease and disability as infant mortality rates fall (WHO 1999).

In West Africa 2 to 3 percent of all children have a serious hemoglobinopathy (sickle-cell anemia, thalassemia) (Adeoye 1973; Obama et al. 1994). These children are at risk for nervous system complications, the frequency of which may be as high as 12.8 percent. Complications of sickle-cell disease in children include mental changes, cerebrovascular accidents, cranial nerve palsies, dural sinus thrombosis, and increased susceptibility to meningitis, especially salmonella and pneumococcal meningitis. A recent study in the United States revealed that 33 percent of children observed with sickle-cell disease had mild mental retardation (Steen et al. 1999).

Down syndrome has until quite recently been regarded as rare in black African populations. Reports of a birth prevalence of 1.16 (Adeyokunnu 1982) and 2.09 (Venter et al. 1995) per 1,000 from Nigeria and South Africa, respectively, suggest that it may be more common than previously thought. These figures may also underrepresent the true prevalence, because evidence suggests that the syndrome in some of these children may go unrecognized or undeclared in rural black communities (Christianson and Kromberg 1996). The prevalence is likely to be higher in Sub-Saharan Africa countries that have high fertility rates, where effective family planning programs and prenatal screening programs are often lacking and where 11 to 15 percent of births occur to mothers over the age of 35 years (Drugan et al. 1999).

The reported incidence of neural tube defects varies from country to country, from region to region within the same country, and from time to time (Windham and Edmonds 1982). A reported prevalence for neural tube defects of 7 per 1,000 children in Nigeria may have overestimated the incidence, because the study was based at a tertiary referral hospital (Windham and Edmonds 1982). The only available population-based estimate of 3.35 per 1,000 comes from a previously cited study in rural South Africa.

Although consanguineous marriages are considered to be an important cause of congenital abnormalities in many parts of the world, having particularly high prevalence in parts of South Asia, the Middle East, and North Africa, the contribution of such marriages to the burden of developmental disabilities appears to be relatively small in most parts of Sub-Saharan Africa. It has been extrapolated from the birth incidence of single gene disorders in developed countries that the equivalent birth incidence in Sub-Saharan Africa is likely to be on the order of 25 per 1,000 (table 10.4). The collective impact is significant, but none of the individual disorders is a public health problem with sufficiently high population-attributable risk to currently merit targeted intervention.

Perinatal and Neonatal Conditions

Perinatal events, such as preterm delivery, low birthweight, intrauterine growth retardation (IUGR), and birth asphyxia or injury, are commonly associated with an elevated risk of early neonatal death and, in those who survive, of impaired physical, sensory, or mental development in infancy and childhood. Many factors, acting singly or in combination, contribute to the elevated frequency of these events in low-income countries. These include the effect on the growing and developing fetus of maternal macro- and micronutrient deficiencies before and during pregnancy; the direct and indirect effects of maternal systemic and genital tract infections, such as syphilis, rubella, cytomegalovirus, and malaria; and the neurological effects of low blood glucose, hypox-ia, bilirubin toxicity, and acquired infections in the first few days of life. Many of these risks may be aggravated or ameliorated, depending on the availability and quality of antenatal, delivery, and postnatal services.

In areas in which maternal and neonatal services are poor and birth asphyxia is an important cause of developmental disabilities, operational research to develop and evaluate alternative approaches to the delivery of these services should be a first priority. In settings in which adequate maternal and neonatal services are available, research is urgently needed on the etiology and prevention of adverse perinatal outcomes, such as low birthweight, preterm birth, and IUGR; on the causal pathways between these factors and developmental disabilities; and on the differential impact of their prevention on the prevalence of neurodevelopmental disabilities in low-income countries. In undertaking this research it will be important to define IUGR and its subtypes (Goldenberg et al. 1989); distinguish between IUGR and its antecedents, many of which are independent risk factors for poor neurodevelopmental outcome (Breart and Poisson-Salomon 1988); control for poor social and environmental conditions that operate postnatally and that may modify neurodevelopmental outcomes (Breart and Poisson-Salomon 1988); select a set of outcomes that are sufficiently prevalent, well defined, and stable over time so that they can be measured with precision at defined time points; and use reasonable sample sizes.

Congenital Rubella

Congenital rubella is a major global cause of preventable hearing impairment, blindness, and intellectual disability. Mathematical modeling has yielded a global disease burden estimate for congenital rubella syndrome (CRS) of 110,000 to 300,000 new cases per year (Cutts and Vynnycky 1999). The incidence of CRS has been variably set at 0.5 to 2.2 per 1,000 live births in developing countries during epidemics, which occur every four to seven years (Cutts et al. 1997). Although many developed countries have set elimination goals, only 28 percent of developing countries routinely vaccinate against rubella (Robertson et al. 1997). No countries in Sub-Saharan Africa include rubella in their national immunization program, and rubella serology, which is essential for surveillance, is unavailable in much of the subcontinent (Robertson 2000).

It is recommended that countries wishing to undertake prevention programs for CRS should either mount vaccination programs for adolescent girls or women of reproductive age or offer universal vaccination in infancy as part of routine childhood immunizations, accompanied by serological surveillance of women of reproductive age. These programs should be undertaken only if the current expanded programs of immunization are already achieving coverage of 80 percent or more. Coverage of less than 80 percent may result in reduced transmission in childhood but leave a large number of women susceptible when they reach reproductive age. A recent cost-benefit analysis of universal rubella vaccination indicates economic benefits comparable to Haemophilus influenzae type B (Hib) and hepatitis B virus (HBV) vaccines (Hinman et al. 2002).

Congenital Syphilis

Congenital syphilis is a common and important cause of diverse clinical manifestations in the newborn infant that include deafness, interstitial keratitis, and mental retardation. It is largely preventable through screening and adequate treatment in pregnancy of the 4 to 15 percent of women known to be affected in Sub-Saharan Africa (Schulz, Cates, and O'Mara 1987).

In spite of the availability of an established and highly cost-effective intervention, it was found, in a recent survey of 22 countries in Sub-Saharan Africa, that only 38 percent of women attending antenatal clinics were being screened and treated for syphilis (Gloyd, Chai, and Mercer 2001). It has been roughly estimated that every year up to 600,000 opportunities are missed to reduce adverse fetal and infant outcomes in Sub-Saharan Africa. Although reductions in fetal wastage and neonatal mortality may be the main benefits, more effective antenatal treatment will also reduce defined risks for developmental disabilities in these children.

Other congenital infections such as cytomegalovirus, tox-oplasmosis, and herpes infections are also responsible for important neurological sequelae, but because they occur less frequently and are less amenable to primary or secondary prevention, they are given less weight in this review.

HIV Infection

HIV infection is known to have an adverse effect on the developing central nervous system and could potentially make a substantial contribution to the burden of developmental disabilities in populations with high HIV seroprevalence. HIV infection causes damage to the central nervous system through direct cytopathic effects such as occurs in HIV-associated encephalopathy (Brustle et al. 1992), or as a result of vasculopathy or immune-mediated factors. Secondary complications of immunodeficiency, such as opportunistic infections, malignancy, and intracranial hemorrhage, may lead to brain damage. Thrombocytopenia, from direct damage to the bone marrow or as an indirect consequence of opportunistic infections, predisposes HIV-infected children to intracranial hemorrhages and strokes (Mueller 1994). These children are also at higher risk for opportunistic infections of the central nervous system, such as toxoplasmosis or cryptococcal meningitis (Aylward et al. 1992; Mueller 1994), although such infections occur less frequently in children than in adults. Neurological problems in HIV-infected children have been described with varying frequency from different parts of the world. In a natural history study from South Africa, neurological abnormalities were found in approximately 50 percent of children followed to 18 months of age (Bobat et al. 1998). In Rwanda, 15 to 40 percent of HIV-infected infants were found to have abnormalities by 6 months of age (Msellati et al. 1993).

The developmental trajectory of such infected children is confounded by maternal, social, and biological risk factors during pregnancy and early childhood. Maternal substance and drug abuse, more common in HIV-infected women, is known to have an independent adverse effect on brain growth and neurodevelopmental outcome. Low birthweight and prematurity, poverty, protein calorie malnutrition, and micronutrient deficiencies, more frequently seen in HIV-infected children, particularly in developing countries, may similarly compromise early child development (Brouwers et al. 1996). Children who are persistently ill lose a sense of mastery motivation and hence fail to practice new development skills, especially during the first two years of life (Trad et al. 1994). Maternal-child interaction is affected as HIV disease progresses and as maternal emotional availability decreases, resulting in irregular attachment. The observation has been made that HIV-uninfected children of HIV-positive mothers are also at higher risk for cognitive academic and language delays than the general population (Condini et al. 1991). This may be similarly mediated through the social, economic, and environmental consequences of the infection on other household members (Faithfull 1997; Kotchik 1997; Miles et al. 1997).

Malaria

Malaria is the leading cause of childhood mortality and morbidity in large tracts of the subcontinent. Because cerebral malaria is a well-known and not infrequent complication and may result in neurological sequelae in survivors, malaria has the potential to make a significant contribution to the burden of developmental disabilities in Sub-Saharan Africa.

The neuropathology of cerebral malaria stems from a series of complex mechanisms, which may operate independently or in combination to adversely affect the developing brain. These include hypoglycemia, multiple seizures, reduced cerebral perfusion associated with raised intracranial pressure, hypoxia associated with microvascu-lar obstruction, and tissue damage following induction of cytokine cascades (Marsh 1995). The sequelae reported in order of frequency were hemiplegia or hemiparesis, speech disorders, blindness, hearing impairment, cerebral palsy, and epilepsy. Because children often have multiple neurological sequelae it is not possible to disaggre-gate these data to provide absolute rates for each type of sequelae.

In a pooled analysis from five recent studies with similar definitions of cerebral malaria and comparable methodologies and diagnostic criteria, a neurological sequelae rate in survivors of 16 percent was reported (Snow et al. 1999). A range in reported incidence of neurological sequelae from 9 percent (Molyneux et al. 1989) to 23 percent (van Hensbroek et al. 1997) reflects differences in what are considered to be significant deficits on discharge and those that resolve over subsequent periods of observation. The overall rate of persisting neurological sequelae in studies that lasted at least six months was 5.6 percent and provides a better estimate of true incidence. Assuming that only children with cerebral malaria who reach hospitals are likely to survive, that even these children have a case-fatality rate of 16.7 percent, and that only 5.6 percent of survivors have persistent neurological sequelae after six months, it has been estimated that the annual risk of neurological sequelae for cerebral malaria is 0.03 per 1,000 in children zero to four years of age and 0.006 per 1,000 in children five to nine years of age (van Hensbroek et al. 1997). This amounts to 2,443 and 402 annual neurological sequelae events in these two age groups, respectively. Because the underlying assumptions for these estimates are quite conservative, it is likely that the burden of sequelae is higher than the figure presented. These assumptions, if correct, also suggest that primary prevention of malaria may have differential impact on malaria-specific mortality and disability.

Bacterial Meningitis

A recent survey of mostly hospital-based epidemiological studies throughout Sub-Saharan Africa provides an initial basis for calculating incidence rates for bacterial meningitis and its most common pathogens in the subcontinent (Peltola 2001). These studies also provide important information on overall and pathogen-specific case-fatality and neurological sequelae rates in African children.

In the absence of prospective population-based studies, the few incidence data available for all causes of bacterial meningitis permit a tentative estimate of annual incidence at about 25 per 100,000, or 180,000 cases for the subcontinent as a whole. On the basis of published epidemiological and laboratory data, 50 percent of cases can be assumed to be caused by Hib, giving an estimate of about 90,000 cases of Hib meningitis per year. Because the vast majority of cases occurred in children under the age of five years, the incidence in this age group is estimated to be 74 cases per 100,000. This calculation tallies quite well with data in the same age group from five African countries. Annual incidence rates of Hib meningitis (per 100,000 children under the age of five years) was 72 in Senegal (Cadoz, Denis, and Mar 1981), 62 in Burkina Faso (Tall et al. 1994), 60 in The Gambia (Bijlmer and van Alphen 1992), 53 in Niger (Campagne et al. 1999), and 51 among black children in South Africa (Hussey et al. 1994). Although meningococcal meningitis predominates in the meningitis belt, it has its greatest impact on adults, and overall, Hib and Streptococcus pneumoniae predominate as causes of meningitis in young children in the subcontinent as a whole.

On the basis of data from several studies evaluating neurological sequelae of bacterial meningitis according to etiology, it is estimated that about 40 percent of those surviving Hib meningitis, 50 percent of those surviving pneumococcal meningitis, and 10 percent of those surviving meningococcal meningitis had long-term sequelae.

Measles

For the vast majority of children with measles infection the main risk is death rather than nonintact survival. A notable exception, as has been mentioned earlier, is the occurrence of blindness in measles survivors, which has been cited as one of the most important preventable causes of blindness in the populations of Sub-Saharan Africa. Measles immunization has long been accepted as one of the most cost-effective interventions in child health. The main challenge will be to replicate more widely the high vaccination coverage and virtual elimination of measles that has been achieved in many countries on the subcontinent.

Tetanus

Neonatal tetanus (NNT) remains an important cause of infant mortality in Sub-Saharan Africa, where it has been estimated that 150,000 neonates suffer from NNT each year (Galazka and Gasse 1995). The full extent of the problem may be much larger than this, and the WHO has suggested that in some areas as many as 95 percent of cases may go unrecognized (WHO 1997). Since high case-fatality rates of up to 90 percent have been reported for NNT, the main end point of interest in most studies has been neonatal death rather than neurological sequelae. In early studies it was thought that no permanent neurological damage occurred in survivors of NNT (Gadoth et al. 1981; Sharma et al. 1976). However, frequent uncontrolled spasms associated with prolonged apnea and drops in oxygen saturation may lead to hypoxic brain damage. Later long-term studies from India and Turkey have suggested that 13 to 37.5 percent of survivors had significant impairments (Anlar, Yalaz, and Dizmen 1989; Gadoth et al. 1981). In a recent study in Kenya significant reductions in head circumference were noted in NNT survivors (Barlow et al. 2001). These children also had more problems with hand-eye coordination, lower developmental scores, more mild neurological abnormalities, and more frequently reported behavioral problems.

The main intervention to prevent NNT is the provision of three doses of tetanus toxoid (TT) to women in their first pregnancy and a single booster dose in each of the subsequent pregnancies, in addition to following safe hygienic practices at birth and in the postnatal period. In a global end-of-decade assessment of TT coverage, 71.3 percent of pregnant women were reported to have received at least two doses of TT in their previous pregnancy (UNICEF 1999). It has been shown in tetanus seroprevalence surveys in Sub-Saharan Africa that estimates of TT coverage significantly under measure protection against NNT (Deming et al. 2002) and that many countries may be approaching World Summit of Children year 2000 goals of 90 percent TT coverage. This is reflected in a dramatic decline in the number of clinical cases in some parts of the subcontinent (Jeena, Wesley, and Coovadia 1999).

Given that vaccination coverage is reasonably high and serological protection even higher, and that high mortality rates in unvaccinated populations result in high neonatal mortality, we speculate that NNT makes a very modest contribution to burdens of developmental disability in Sub-Saharan Africa as a whole. It may represent a more significant problem in selected countries, where the main thrust must be to improve TT coverage.

Iodine Deficiency

Iodine deficiency has multiple and serious adverse effects, including impaired cognitive function, and it is considered to be the leading cause of preventable mental retardation and brain damage worldwide. A 1999 review of data on the status of iodine deficiency disorders (IDD) demonstrated that IDD is a public health problem in 44 out of 46 countries in Sub-Saharan Africa and that an estimated 295 million people living in these countries are potentially at risk for iodine deficiency (WHO, UNICEF, and ICCIDD 1999). Its well-known effect on mental development has played an important role in mobilizing political, public health, and nutritional activists in support of national and international prevention programs. Efforts at prevention and control focus mainly on the iodization of salt, and as a result of concerted public health prevention efforts, salt iodization was reported, by 1999, to have reached 63 percent of households in Sub-Saharan Africa. In a recent study of seven African countries one year following the introduction of salt iodiza-tion, increases in median urinary iodine to above levels considered to constitute iodine deficiency was observed in all countries (Delange, de Benoist, and Alnwick 1999). Progress toward the elimination of IDD through universal salt iodization, with its anticipated impact on cognitive impairment, appears to be one of the most significant successes in the field of noncommunicable diseases. However, the challenge still remains to ensure that salt iodization reaches all populations. Recent studies in South Africa and Lesotho showed that, in spite of compulsory iodization of salt, iodine deficiency remains a significant problem in primary school children in more remote areas of the country (Sebotsa et al. 2003; van Stuijvenberg et al. 1999).

Vitamin A Deficiency

Vitamin A is an essential micronutrient for normal growth, for normal functioning of the visual system, for the maintenance of normal epithelial integrity and immune function, and for normal reproduction. Consequently, vitamin A deficiency (VAD) results in increased severity of certain infections and an increased risk of disease and death in young children. More severe vitamin A depletion leads to night blindness, which can evolve to irreversible partial or total blindness if the depletion continues (Sommer and West 1996; WHO 1996). Evidence for the contribution of VAD to the burden of visual impairment in Sub-Saharan Africa is the main interest in this review and has been presented earlier. There is evidence to suggest that there has been a decrease in clinical VAD (ACC/SCN 1997), manifest as eye lesions, largely as a result of more effective vitamin A supplementation programs in many developing countries, including Sub-Saharan Africa. However, recent estimates for clinical VAD and subclinical VAD still range between 2.8 million and 3.3 million and 140 million to 251 million preschool children, respectively (UNICEF and Tulane University 1998; WHO 1995a). It, therefore, continues to be a major public health problem for which there are relatively simple and cost-effective interventions.

Iron Deficiency

The WHO has estimated that about 40 percent of the world's population suffer from anemia and that a substantial proportion of that burden is attributable to iron deficiency. There is a particularly high prevalence of anemia in pregnant women (50 percent), children in the first two years of life (40 percent), and schoolchildren (40 percent) (WHO 2000). With regard to preschool children, anemia prevalence is the highest in Africa (42 to 53 percent) and Asia of all WHO regions (WHO 1998). An association between hemoglobin concentration and psychomotor performance has been demonstrated at all stages of life. Although there is a good biological basis for claiming that a deficiency of iron might impair mental and motor development, an Expert Group, in weighing up the available evidence (Idjradinata and Pollitt 1993), concluded that only anemia, and not iron deficiency without anemia, impairs the behavior and development of infants (Draper 1997). There now appears to be sufficient evidence to show that iron supplementation of anemic children over two years of age improves development but that this effect is less conclusive in children under the age of two years (Grantham-McGregor and Ani 2001).

Poverty and Protein-Energy Malnutrition

Although there is evidence that the levels of protein-energy malnutrition (PEM) in Sub-Saharan populations have improved and the spectrum of these deficiencies has changed over the past 30 to 40 years, there is still a substantial burden of PEM in many populations in the subregion. For many reasons it has been difficult to establish a causal relationship between undernutrition and behavioral and cognitive development in these children. The main classification of malnutrition used in these studies defines a mixture of clinical signs that are a product of coexistent infections and many other deficiencies, such as zinc, magnesium, copper, and iron, as well as protein and energy, each of which may have an independent and different effect on developmental outcomes (Grantham-McGregor 1995). Malnourished children usually come from families who suffer from numerous disadvantages. These include poor social, economic, and environmental living conditions and unstable family units with large numbers of closely spaced children. The parents are often unwell, poorly nourished, and depressed; young with low intelligence and levels of education; either unemployed or in low-skilled occupations; and likely to have low social and media contacts. Few toys or books can be found in the homes of these families, and parents participate little in play activities; thus there is little stimulation.

The best evidence for a causal effect of undernutrition on cognition and behavior is likely to come from randomized controlled trials. It has been considered unethical to conduct randomized controlled trials to study the effects of malnutrition on behavior and development. It has therefore been necessary to rely on less satisfactory epidemiological study designs, such as case-control studies, which have been unable to control for the wide-ranging nature of these children's disadvantages.

There is also some uncertainty about whether cognitive impairment is an inevitable consequence of severe early malnutrition and what form it is likely to take. Animal experimentation initially suggested that early malnutrition significantly reduced brain growth and left it permanently smaller in size (Winick and Noble 1966). More recent work has shown that many of these dramatic anatomic alterations are reversible. It also shows that parallel alterations in neu-rotransmitter and receptor characteristics are evident, resulting in subtler neurodevelopmental deficits in motivation, emotional reactivity, and cognitive flexibility rather than in frank reductions in intelligence (Levitsky and Strupp 1995).

In spite of all these limitations and uncertainties evidence suggests that previously malnourished children show a deficit in tests of cognitive function or intelligence. This effect is particularly strong if the undernutrition is chronic and if they return to poor environments. The effect is less clear in children exposed to acute episodes of malnutrition who do not return to impoverished social conditions (Grantham-McGregor 1995). In unraveling the complex relationship between malnutrition and cognition, several questions remain unanswered—these include questions about the duration of the cognitive deficits, the specific types of cognitive functions that are affected, the relative effect of malnutrition at different points in early childhood, whether complete recovery can occur following placement in enriched environments, and the relative contribution of individual nutrients to cognitive deficits.

Alcohol

Fetal alcohol syndrome is the most common single preventable cause of mental retardation worldwide (Viljoen 1999), and the overall rate for the developed world, where the vast majority of prevalence studies have been based, has been placed at 0.97 per 1,000 children (Abel 1995). It has its highest rates in subgroups of the population characterized by low socioeconomic status and confounded by race. For example, in American Indians and African Americans, rates of 8 per 1,000 children (May et al. 1983) and 2.29 per 1,000 have been found, considerably in excess of rates in the U.S. population as a whole.

The only published studies or reports on fetal alcohol syndrome in Sub-Saharan Africa are confined to a subgroup of the "coloured," or mixed race, community in the Western Cape of South Africa, where a prevalence rate of 40.5 to 46.4 per 1,000 children age five to nine years has recently been measured (May et al. 2000). This is the highest rate for fetal alcohol syndrome ever recorded anywhere in the world and appears to be attributable to very high alcohol intake in this South African subpopulation as a consequence of a special set of historical and social conditions.

The principal effect of alcohol teratogenicity is brain damage with consequent lowering of intelligence (mean intelligence quotient 65), behavioral abnormalities (attention deficit, hyperactivity, aggressiveness), and poor language assimilation. The syndrome also presents with a recognizable cluster of facial features and a maternal history of heavy alcohol intake in pregnancy, often imbibed in a pattern of binge drinking. There is no reason to believe that the overall prevalence in Sub-Saharan Africa is particularly high, but it may occur in other subpopulations with low socioeconomic status and histories of high alcohol intake. In these subpopulations, interventions to prevent or reduce alcohol ingestion should assume high priority.

Other Toxins

Other toxins, such as drugs, nicotine, and heavy metals, may result in cognitive and developmental deficits in young children. No information could be found on the prevalence of disabilities attributable to these toxins in Sub-Saharan Africa, although there is clear evidence of exposure to lead, for example, and every reason to expect that some contribution, albeit small, is made by these and other toxins to the burden of disabilities in this subregion.

Primary Prevention

Population-based interventions aimed at primary prevention of diseases the risk of which is attributable to high population offer the most attractive approach to tackling developmental disabilities in Sub-Saharan Africa. A general conceptual framework that has been proposed to guide the development and evaluation of all public health interventions (de Zoysa et al. 1998) has potential applications for interventions in the field of developmental disabilities (figure 10.4).

The starting point in this framework is a detailed description of the prevalence, types, severity, and distribution of developmental disabilities in methodologically sound two-phase population-based surveys. The case for good descriptive epidemiological data on developmental disabilities in Sub-Saharan Africa has already been made. Given the expense and effort in mounting these surveys, these descriptions have often been combined in previous studies with the next steps in the research framework, namely, the determination of biological risk factors and social determinants for developmental disabilities. Together these data provide the necessary information to formulate potentially viable interventions and to set up efficacy trials to test them. Subsequent steps, of less immediate relevance to this chapter, involve the refinement of these interventions based on trial findings and their application and evaluation in "real-world" settings.

Secondary Prevention

Neonatal screening programs for congenital disorders, the identification of high-risk newborns for early multidiscipli-nary intervention programs, or screening in infancy for developmental delays at primary health care clinics have all been proposed as secondary prevention strategies to minimize the effect of neurological disorders once they have already occurred. In some Sub-Saharan countries, such as South Africa, that have high attendance for antenatal care, supervised deliveries, postnatal care, and "well-baby" services in the first year of life, these strategies provide the possibility of early intervention within established health facility networks at strategic points during the neonatal and post-neonatal periods. Once screening has occurred, they permit definitive assessments and the institution of appropriate programs of management for identified children. The feasibility and relative benefit of these access strategies and early intervention programs need to be carefully weighed against primary and tertiary prevention in different population settings.

A full consideration of the available primary and secondary interventions and their relative cost-effectiveness is beyond the scope of this chapter. An exercise of this kind is under way and will be the subject of a chapter in a separate monograph (Disease Control Priorities in Developing Countries, 2nd ed.) Because tertiary prevention is concerned with ameliorating the impact of established developmental disabilities and has been an important, if not the main, thrust in Sub-Saharan Africa, it merits special attention.

Tertiary Prevention

Tertiary prevention programs are needed for all children with established disabilities and this is particularly true for Sub-Saharan Africa, where a large burden of established disability already exists as a result of past failures in primary and secondary prevention of the many conditions discussed earlier in this chapter. As the focus turns from mortality to disability as a public health end point, a better understanding is needed of the dynamic interplay between existing child survival strategies and interventions to specifically reduce or ameliorate developmental disabilities in developing countries, in general, and in Sub-Saharan Africa, in particular.

For the majority of countries in Sub-Saharan Africa, service provision to people with established disabilities will be the starting point and, for some time, the dominant approach in addressing the needs of people with developmental disabilities. The concern in these programs is less with underlying causes than with improvement of functional capacity of affected individuals and the enhancement of their participation in all aspects of community life. Their aim will be to combine direct and indirect therapeutic inputs by health and rehabilitation workers with support and training for families and caregivers and societal interventions to limit stigma and equalize opportunities in all walks of life. In all these processes people with disabilities and their immediate family members will be expected to take a leading role with support and guidance, but not direction, from health and rehabilitation professionals.

For health and rehabilitation professionals in developing countries faced with the large number of children and adults with established disability, tertiary prevention continues to remain an important priority. The failure to meet the rehabilitation needs of the majority of people with disabilities through existing services and growing concerns about the inherent limitations of institutional rehabilitation gave rise to an alternative approach to care for the disabled in developing countries, called community-based rehabilitation (CBR).

Community-Based Rehabilitation

The concept of CBR has been defined by the WHO as a community development strategy focusing on equalization of opportunities and social integration of people with disabilities (WHO, UNESCO, and ILO 1994). CBR activities are holistic in nature and, in attempting to integrate social and medical models of disability, include all or some of the following: awareness raising in communities; advocacy for the rights of people with disabilities; development of parents' and caregivers' groups; income generation; networking with educational, social, and employment authorities; and provision of rehabilitation and health services. The focal person in the CBR program is the person with disabilities, and the main goals, in addition to providing access to rehabilitation services, are to ensure that people with disabilities have rights to self-determination; enjoy the full benefits of family membership; are active and responsible community members; have equal access to education, skills training, work, and recreation; play a significant role in the CBR program itself; participate in organizations that cater to people with disabilities; and act as lobbyists for the disabled and their families.

CBR programs vary in their content, but most include some of the following services:

• a decentralized approach to service delivery with the primary focus and integration of these services at district level within established models of primary health care

• training and support for general-purpose mid-level rehabilitation workers who operate at household and community level

• screening and early identification of children with developmental disabilities

• home-based support and training for activities of daily living

• promotion of inclusive education in pre- and primary schools

• provision of vocational training at secondary schools

• referral to specialist rehabilitation services

• reorientation of the roles of rehabilitation professionals, where available, to support, train, and manage rehabilitation teams and CBR programs in addition to directly providing rehabilitation services.

Defining the sequence and level of input at each of these levels may vary from country to country and from program to program but will be facilitated in all cases by the existence of national health systems based on primary health care and national rehabilitation strategies at all tiers of the health system.