From 1 January 1997 to 31 December 2001 we performed a retrospective observational study on the incidence, accuracy of clinical diagnosis, and prevalence estimation of Down’s syndrome in a well defined population of 1.7 million in Northern Ireland.¹

A total of 208 postnatal cases of Down’s syndrome were diagnosed: 197 trisomy (94.7%), 3 translocation (1.45%), and 8 mosaic cases (3.85%) (expected ratios 94% trisomy, 5% translocation, 1% mosaic²). In a population of 114 307 live births, a minimum prevalence of 167.9 per 100 000 (or 1 in 595 births) was calculated.

The detection rate of mosaic variants is higher than quoted rates of 1–3%. This may be accounted for by inclusion of newly diagnosed adult cases in our study, but mosaic variants often do not have dysmorphic features and may not be identified in studies.

Ninety per cent of trisomy and 100% of translocation cases were diagnosed on clinical features alone, with karyotyping carried out for diagnostic confirmation. This figure fell to 37.5% for mosaic cases (p < 0.001), confirming the difficulties with the clinical diagnosis of mosaic Down’s syndrome, where few classical dysmorphic features are present.

The two mosaic cases diagnosed within seven days of life presented with simian creases, hypotonia, and characteristic facial features including epicanthic folds, up-slanting palpebral fissures, and protruding tongue. One patient had a sandal gap.

Three mosaic children were diagnosed after day 7. One clinically felt to be Noonan syndrome was diagnosed at 6 months. Another (diagnosed at 19 months) presented with developmental delay, without dysmorphic features, and the third had a sample sent at 7.5 years, as a check sample and not time of first diagnosis.

Three mosaic patients were diagnosed as adults. One was an inpatient at a regional specialist assessment centre for learning disabilities, and was previously known to have Down’s syndrome. A second presented at 18 years of age and was educationally subnormal with no dysmorphic features. The third was aged 54 years at diagnosis, presenting with short stature, mental retardation, and low white cell count with poor myeloid activity in the bone marrow, but with no dysmorphic features.

Our study concluded that the overall minimum prevalence is around 1 in 595 births. This is a slightly higher prevalence than previously documented, and may be a result of a higher incidence of mosaic cases, which are often without dysmorphic features and therefore more difficult to diagnose. Mosaic Down’s syndrome should always be considered in those who are educationally subnormal, without a specific diagnosis.