Genetic counselling after a first affected child
A couple, both unaffected, have a two-year-old son who has just been diagnosed with a condition caused by a single autosomal recessive gene. There is no other known case in either family. The mother is ten weeks pregnant, and she has one healthy sister, aged 27, who is planning a pregnancy with a partner from the same town, where screening studies suggest that about 1 in 50 people carry the altered gene. The parents tell the genetic counsellor that they have read that the chance of a second affected child is 'one in four', but they believe their family has already had its share of bad luck, so the next baby should be fine. The sister, who has never been tested, wants to know whether she should be worried and whether her risk is the same as her sister's was before her nephew was diagnosed. The counsellor has the parents' test results confirming that both carry one copy of the altered gene.
1. Explain why both parents must be carriers, state the risk to the unborn child, and evaluate the parents' reasoning that they have 'had their share of bad luck'.
Read the worked answer
Because the condition is autosomal recessive, an affected child must have two altered copies, one inherited from each parent. Both parents are unaffected, so each must also have a working copy - that is, each is heterozygous, which the test results confirm. The son's genotype is aa, the parents' is Aa × Aa.
Each pregnancy from Aa × Aa is an independent draw: 1/4 aa (affected), 1/2 Aa (unaffected carrier), 1/4 AA (unaffected non-carrier). The risk to the unborn child is therefore 1 in 4, or 25%, the same as it was before the son was diagnosed and the same as for any future pregnancy.
The parents' reasoning is the gambler's fallacy: previous conceptions do not change the gametes produced in this pregnancy. Saying the family is 'due' a healthy child is no more valid than expecting a coin that has landed heads to land tails. The 1-in-4 figure is also a probability per pregnancy, not a rule about four children: the chance that a couple like this has four unaffected children is (3/4)⁴ ≈ 32%, and the chance of two affected children in a row is (1/4)² = 6.3%, which is uncommon but entirely possible.
2. Work out the sister's probability of being a carrier and her chance of having an affected child with her partner. Explain why you must give a range rather than a single number.
Read the worked answer
The sister's parents are the mother's parents. Since the mother is a confirmed carrier and is unaffected, at least one of the grandparents carries the altered gene.
Case 1: one grandparent is a carrier, the other is not (Aa × AA). Every child has a 1/2 chance of being a carrier and none is affected, so the sister's unaffected status tells us nothing extra: her carrier probability is 1/2.
Case 2: both grandparents are carriers (Aa × Aa). Among unaffected children, the ratio AA : Aa is 1 : 2, so her carrier probability is 2/3.
The grandparents have not been tested, so the honest answer is a range: the sister's carrier probability is between 1/2 and 2/3.
Her partner's carrier probability is about 1/50 from the population screening figure. Risk of an affected child = P(sister is carrier) × P(partner is carrier) × 1/4:
- lower end: 1/2 × 1/50 × 1/4 = 1/400 ≈ 0.25%
- upper end: 2/3 × 1/50 × 1/4 = 1/300 ≈ 0.33%
So her risk is roughly 1 in 300 to 1 in 400 - far lower than her sister's 25%, but still tens of times the population baseline of 1/50 × 1/50 × 1/4 = 1/10 000 (0.01%). Giving only 2/3 would overstate her risk; giving only 1/2 would understate it.
3. The sister asks what she should actually do with this information. Give a reasoned recommendation, including the limitations of testing and the difference between a probability and a prediction.
Read the worked answer
The most useful next step is carrier testing of the sister for the specific variant found in her nephew. Testing for the family's known variant is far more informative than a general panel, because a general test might not cover that particular change and a negative result would then be falsely reassuring. If she is not a carrier, her risk of an affected child with this partner is essentially the population risk (about 1/10 000), not zero. If she is a carrier, her risk is 1/200 (0.5%) with an untested partner and can be refined further by testing the partner.
Limitations to state plainly: a negative result lowers but does not necessarily abolish the risk if the test does not detect the family's variant or if rare variants in the same gene are not covered; results apply to this condition only, not to other genetic conditions; and testing has implications for relatives, so consent and confidentiality matter.
Finally, the figures are probabilities, not predictions. A 1-in-300 risk means that out of 300 couples in this situation about one would have an affected child - it says nothing about which one. The counsellor should present the range, offer the tests, and make clear that a decision to test or not is the couple's, supported by information rather than pressure. The same reasoning applies to the parents: 25% is high enough that prenatal diagnosis or early newborn monitoring may be worth discussing, but it also means a 75% chance of an unaffected child.