On this page
In India, cousin marriage is a real family question, not just a textbook one. An analysis of the National Family Health Survey 2015–16 by Sharma and colleagues in the Journal of Biosocial Science found that 9.9% of ever-married women were in a consanguineous marriage, rising to 23% in the South region. First-cousin marriage made up most of it, at 8.7%.
You can work out the genetic risk yourself with two numbers and one cross.
Why does cousin marriage raise the risk of recessive disorders?
A recessive condition appears only when a child inherits two faulty copies of the same gene. If the parents are unrelated, the two copies arrive independently, so the chance is q × q = q², where q is the frequency of the faulty allele.
Cousins share a pair of grandparents. That means one grandparent's allele can travel down both sides of the family and meet itself in the child.
The inbreeding coefficient F is the probability that a child's two copies of a gene are both copies of the same ancestral allele, as defined in MIT's 7.03 Genetics lecture on inbreeding.
For the child of first cousins, F = 1/16, according to the US National Institute of Justice's population genetics course.
How do you run the cross for first cousins?
Suppose the shared grandfather carries one faulty allele, a. Follow it down the pedigree:
It reaches his son with chance 1/2, and his daughter with chance 1/2.
The son passes it to his child (cousin 1) with chance 1/2; the daughter passes it to her child (cousin 2) with chance 1/2.
So both cousins carry that same allele with chance 1/2 × 1/2 × 1/2 × 1/2 = 1/16.
Their child gets it from both parents with chance 1/2 × 1/2 = 1/4, which gives 1/16 × 1/4 = 1/64.
The two shared grandparents carry four alleles at that gene between them, and each can do this, so 4 × 1/64 = 1/16. That is F. The MIT lecture works through the same pedigree and gets the same answer.
How often does a recessive disorder appear in children of cousins versus unrelated parents?
With inbreeding, the chance that a child has two copies of the recessive allele is q² + Fpq, where p = 1 − q, as set out in Steven Carr's notes on inbreeding and health at Memorial University. Put in F = 1/16:
q = 1/100 (roughly 1 in 50 people is a carrier): unrelated parents, 1 in 10,000; first cousins, about 1 in 1,390. About 7 times as often.
q = 1/1,000: unrelated parents, 1 in a million; first cousins, about 1 in 15,800. About 63 times as often.
q = 1/10,000: unrelated parents, 1 in 100 million; first cousins, about 1 in 160,000. About 630 times as often.
The rarer the allele, the bigger the multiplier, because for rare alleles the ratio is roughly F/q. The MIT lecture gets about 600 times for q = 1/10,000. But the absolute chance stays small: even at 630 times, it is about 1 in 160,000.
Does this mean cousin marriage is dangerous?
For one specific rare condition, the multiplier is large. Across all conditions, the extra risk is modest. In 2002 a working group of the National Society of Genetic Counselors, led by genetic counsellor Robin Bennett, estimated the additional risk for first cousins at 1.7 to 2.8 percent on top of the background risk every couple has, as reported by the University of Washington. A 2011 review by Hanan Hamamy notes that first-cousin couples are generally given a risk of birth defects about double the population risk.
Why so much smaller? Many conditions are not caused by a single rare recessive allele, and inbreeding makes no difference to dominant traits, as the MIT lecture notes.
The NSGC recommendations (Bennett et al., 2002) advise a thorough family history and the same genetic screening offered to any couple from the same ethnic group. If a condition already runs in the family, the numbers change completely; that is a pedigree probability problem, not an F × q one.
How do you solve these questions in an exam?
Use this four-step check:
Find q. If you are given how common the disorder is, q is its square root. Our Hardy-Weinberg numerical guide shows the steps.
Unrelated parents: q².
First cousins: q² + (1/16) × q × (1 − q).
Sanity check: the ratio should be close to 1 + 1/(16q). If your cousin answer is not bigger than q², check the Fpq term.
Real families are small, so they rarely show textbook ratios; see why small families often miss 3:1.
If the idea of one allele travelling down both sides of a family is hard to picture, the Fermi Simulation Lab has interactive simulations for ideas that are easier to see than to read.
