Mental retardation (MR), defined by the World Health Organization (WHO) as an IQ of < 70, is a common problem that affects approximately 2% of the world's population. Studies to determine the underlying causes of MR indicate that, in the majority of patients, the etiology remains unknown. A diagnosis (either genetic or environmental) is reached in about 64% of moderate-profound MR (IQ < 50) and 24% of the mildly mentally handicapped group (IQ 50-70). Recurrence risk for MR is variable, ranging from 2 - 21% (dependent on the degree of MR and the study), yet demonstrates a clear familial component to MR. While improved cytogenetic and DNA diagnostic techniques have improved our ability to provide a specific diagnosis and familial recurrence risk for individuals with syndromic MR, identification of genes associated with intelligence and MR by quantitative trait loci (QTL) mapping has been unsuccessful.

Definitions and Prevalence

Mental retardation consists of a clinically and etiologically heterogeneous group of conditions. It is defined by the DSM-IV (American Psychiatric Association 1994) as

1. tested IQ at or under approximately 70,
2. concurrent adaptive functioning has to be deficient or impaired and
3. onset of problems before 18 years of age.

Although there are a number of methods of subclassification, the most diagnostically relevant is the subdivision of MR into mild (IQ 50-70) and moderate-severe (IQ < 50) categories. The prevalence of mild MR has been reported between 0.5% - 2%, while moderate-severe MR affects 0.3 - 0.5%.(1) Though the groups differ from each other in etiology and recurrence risk, both subsets of MR have clear genetic components.

Etiology

Studies to determine the causes of MR vary in results. Most of the studies were published in the 1980s and their data is imperfect, as all were performed prior to the availability of molecular diagnostic testing for Fragile X, and at a time when higher resolution cytogenetic techniques were not available, possibly missing subtle chromosomal abnormalities and microdeletions.

In cases of mild MR, a specific risk factor has been identified in 42 - 43%. However, a convincing diagnosis is reached in only 24%(2),(3) with 14% classified as genetic (chromosomal disorder, Fragile X, known genetic or metabolic diagnosis) and 10% environmental.

While moderate-severe MR is less prevalent, a cause is determined in up to 64% (45% genetic and 19% environmental) with only 36% being idiopathic.(4) It is postulated that this higher rate of diagnosis should be expected in moderate-severe MR based on the normalization of IQ to a bell-shaped curve. IQ is presumed to be a polygenic trait with environmental and genetic influences.

While some individuals who fall into the mild MR category may have a specific etiology for their MR, others simply represent the lowest end of the normal distribution imposed by the nature of measuring intelligence. In contrast, those in the moderate-severe MR category are not part of the normal distribution but are at the lowest end of the spectrum, as a result of an abnormal environmental or genetic process. One study, however, that looked at the outcome of routine assessment of 411 patients with MR referred to the genetics clinic at the University of Ottowa(5) found that, while patients with more severe MR were more likely to have a chromosome analysis, the rate of abnormal findings was unaffected by the severity of MR.

With respect to specific genetic disorders, Trisomy 21 accounts for 5% of mild MR and 30% of severe MR,(6),(7) while Fragile X is seen in 5% of both mild and moderate-severe MR.(8) A recent study of individuals with moderate-severe MR found 7.5% to have a submicroscopic deletion identified using telomeric FISH probes. Those with mild MR were found to have such deletions only 0.5% of the time.(9) Over 500 other genetic diseases, mostly very rare, have also been associated with MR (10) and it is reasonable to suppose that a considerable proportion of cases of unknown etiology have a genetic origin.

Information on the incidence of various environmental factors that cause MR varies widely, depending on the country of origin of the study, how patients were ascertained and what decade patients were born. To complicate matters further, many studies use non-specific categories, such as "multiple congenital anomalies," that include both genetic, infectious and drug/alcohol etiologies. Therefore, we will not attempt to detail the frequency of each cause but rather will simply cite the major causes using our own categorization:

Pregnancy problems - placental insufficiency, toxemia, multiple pregnancy, infections (rubella, toxoplasmosis, herpes, syphilis), trauma, drug/alcohol abuse

Perinatal problems - prematurity, neonatal asphyxia, hyperbilirubinemia, hypoglycemia, central nervous system hemorrhage, congenital hypothyroidism

Childhood diseases - complications of infections (meningitis, encephalitis, pertussis, varicella), lead poisoning, cranial trauma, cerebral tumors, cardiac arrest, asphyxia

Psychosocial factors - low socioeconomic status, psychosocial deprivation, parental neurosis/psychosis/character disorder.(11),(12)

Co-morbidity

Various studies have looked at associations between medical conditions and MR. Most studies are compromised by their having ascertained patients from a genetics clinic or tertiary care hospital(13) which would likely bias toward increased comorbidity (since it is unlikely that individuals with a mild, uncomplicated seizure disorder or deafness would need services of a genetics clinic or tertiary care hospital). While these studies report an increased association with epilepsy, cerebral palsy, blindness and deafness (Table 1), they fail to address whether these associations represent a pleiotropic disease (a collection of symptoms or infirmities that affect seemingly unrelated organs, but have a single common genetic cause), or whether the MR is a consequence of the medical diagnosis.

* Data pooled from references 6 and 13.

In addition, neither the severity (of seizures or cerebral palsy) nor the possibility of an environmental cause (such as trauma) is taken into account as a possible basis for the diagnosis.

A more recent study utilized the British Columbia Health Surveillance Registry to ascertain cases of nonspecific MR in individuals born between 1952 and 1970. Identified cases were linked by birth registration number to family sibships from computer-linked groupings of birth and marriage records in British Columbia. Family information was retrieved for 97% of the cases. Cases with a documented genetic or environmental cause or those with a cerebral malformation were excluded from the study. Though the risk for associated conditions was somewhat lower (Table 2), it was similar to that of other studies.(14)

+ Cases were not separated into mild and moderate-severe groups for this analysis and, therefore, not pooled with data in Table 1.

However, this study was performed prior to currently available cytogenetic banding techniques, FISH and DNA diagnostic testing for Fragile X. Therefore, it is impossible to know how these diagnoses would modify the risk of a comorbid condition.

It is clear that conditions such as seizures and cerebral palsy are associated with MR. How they relate to MR is difficult to determine. It could be that there are unrecognized, non-genetic prenatal or perinatal events (such as trauma or infections) that caused the condition and associated MR. Another possibility is that complications from seizures produced the MR, rather than representing true co-occurrence. Lastly, a single genetic cause (such as a single gene mutation, DNA deletion or duplication) may be responsible for both MR and the co-occurring condition.

We believe that all of these possibilities for the co-morbidity are likely. However, some portion of conditions, such as microcephaly and seizures, represents a variety of medical genetic diagnoses. This assertion is supported by new diagnostic techniques, such as telomere FISH, Uniparental disomy (UPD) studies and high resolution chromosome banding, that have provided an etiology for individuals previously thought to have non-syndromic MR.(15),(16) Similarly, a recent whole genome microsatellite scan of children with multiple congenital anomalies revealed submicroscopic deletions and duplications to be a cause of multiple congenital anomalies.(17) Based on this information, it is possible that submicroscopic deletions and duplications may be a cause of MR.

Recurrence

All studies of sibling recurrence risk in MR have found a significant increase above the general population rate. The studies vary in population studied, exclusion criteria and methods of ascertainment, counting and analysis. A summary of the studies is in Table 3.