Newborn Screening, Carrier Panels, and an Affected Child of Two Negative Parents
A provincial lab screens 250,000 newborns for an autosomal recessive metabolic disorder and confirms 4 affected infants. A carrier panel covering the five most common pathogenic variants, with an estimated detection rate of 90% for this disorder, was run on 1,000 randomly chosen adults from the same province; 12 were carriers. Hardy-Weinberg calculations from the newborn data predict a carrier frequency near 0.8%. The province also contains a religiously isolated community of 6,000 people with a long history of marriage within the community; a pilot screen there found 3 affected infants among 900 births. Separately, a couple from the general population each tested negative on the carrier panel and then had an affected child.
1. Calculate the expected carrier frequency from the newborn data, compare it with the observed 12/1000, and decide whether the two are in genuine conflict.
Read the worked answer
Calculation
- Affected frequency = q^2 = 4/250,000 = 1.6 x 10^-5.
- q = sqrt(1.6 x 10^-5) = 0.004, i.e., about 1 in 250 alleles.
- Carrier frequency = 2pq = 2(0.996)(0.004) = 0.0079, about 0.8% or 1 in 127.
Comparison
- Observed: 12/1000 = 1.2%.
- With n = 1,000 the 95% confidence interval around 12 carriers runs roughly from 0.6% to 2.1% (Wilson interval). The predicted 0.8% sits comfortably inside it.
- Expected number of carriers in 1,000 adults is about 8; observing 12 is a difference of 4 individuals, well within Poisson-scale fluctuation.
Verdict: no real conflict yet.
Do not invoke population structure to explain a difference that sampling error already explains. Also note the Hardy-Weinberg assumptions (random mating, no selection, no migration, large population, no mutation) and that newborn screening can undercount cases (infant deaths before diagnosis, missed specimens, variant forms), which would bias q downward - a reason for caution in either direction. A larger sample, or direct variant genotyping, would be needed before claiming the province has an elevated carrier rate.
2. Analyze the isolated community's data. Distinguish the roles of founder effect, genetic drift, and consanguinity, and state what should not be concluded.
Read the worked answer
Numbers
- q^2 = 3/900 = 0.00333, so q = 0.058 (about 1 in 17 alleles).
- Carrier frequency 2pq = about 0.109, roughly 11% - more than ten times the provincial estimate. Affected births are about 1 in 300 versus 1 in 62,500.
Separating the mechanisms
- Founder effect and drift: a small, long-isolated population starts from a limited set of alleles, and random sampling of alleles across generations can raise a rare variant's frequency well above the provincial value. This changes allele frequencies - the explanation for the high carrier rate.
- Consanguinity: mating among relatives does not change allele frequencies. It increases the chance that two carriers of the same variant marry and that a child inherits two identical copies, raising the proportion of affected births relative to what allele frequencies alone predict. Both effects are probably operating here.
What should not be concluded
- That the community is genetically inferior, or that the disorder is caused by inbreeding in any simple sense; the variant was likely present at low frequency before isolation.
- That a high carrier rate applies to individuals who married in from outside.
- Anything clinical from n = 900 births without confirming genotypes: the 3 cases could be a different disorder phenocopy.
What to do: genotype the affected infants to identify the actual variants, estimate kinship from pedigrees, offer community-designed counseling and carrier testing, and keep results confidential to avoid stigmatization.
3. Explain how two parents who both tested negative could have an affected child, and give a defensible residual risk estimate.
Read the worked answer
How it happens
- An affected child of two carriers must inherit two pathogenic variants. If both parents test negative on a five-variant panel, then at least one pathogenic allele in this family is a variant the panel does not interrogate - a rare known variant, a novel missense or nonsense change, or a small deletion not detected by the assay.
- Other, less likely routes: a de novo variant on one parental allele; uniparental disomy or a large deletion on one side; gonadal mosaicism; or the child's diagnosis is wrong (a phenocopy caused by a different gene or an environmental cause).
Residual risk arithmetic
- Prior carrier probability in the general population is 0.008.
- With a 90% detection rate, a negative result leaves about 0.008 x 0.10 = 0.0008, i.e., roughly 1 in 1,250 per parent.
- Risk that both parents are carriers and the child is affected is about (0.0008)^2 x 0.25 - very small, but not zero, and clearly not zero in a family that already has an affected child. Once an affected child exists, the family's prior is no longer the population prior.
Correct response: sequence the affected child first (or the parents for the gene), identify the two variants, then offer targeted testing. Report the negative panel as "reduced risk," never as "not a carrier."