Blood Group Typing in Paternity and Kinship Testing

Every blood group antigen is encoded by one or more genes located on specific chromosomes. Because humans are diploid, each individual carries two alleles at each locus, one inherited from each parent. The phenotype observed serologically reflects which alleles are expressed, depending on whether they are codominant (both expressed, as in ABO and MNS), dominant-recessive (as in some Rh variants), or amorphic (a silent allele that contributes no detectable antigen, as in the ABO O allele).

The logic of paternity testing by serology follows directly from these rules. A child receives exactly one allele at each blood group locus from its biological mother and exactly one from its biological father. If the mother's genotype and phenotype are known, the allele she contributed to the child can often be deduced. The remaining allele in the child must then have come from the biological father. If the alleged father cannot carry that allele based on his own phenotype, he is excluded. Where the mother's phenotype is ambiguous (because she could be homozygous or heterozygous), multiple interpretations are possible, and the exclusion may be uncertain rather than definitive until further testing resolves her genotype.

A concrete example in ABO: a child is blood group B, mother is blood group O (genotype ii). The mother could only contribute allele i. The child's B antigen therefore came from the alleged father. An alleged father of blood group O (genotype ii) cannot carry a B allele, so he is excluded. An alleged father of group AB (genotype I^A I^B) carries both A and B alleles and is not excluded on ABO grounds alone. The power to exclude depends on the father's phenotype, the mother's phenotype, and the child's phenotype considered together, not each in isolation.

A standard pre-DNA paternity panel tested multiple systems to maximise the cumulative exclusion probability. The systems varied by laboratory and era, but by the 1980s most accredited paternity laboratories in North America and Europe were testing at least six to eight systems.

The combined exclusion probability from all six systems listed above reaches approximately 90 to 95 percent in European reference populations. This means that testing a random unrelated man against a European mother-child pair, about 1 in 20 unrelated men would pass all six panels without exclusion. Later panels that added HLA typing (human leucocyte antigen, a highly polymorphic system) raised the figure closer to 99 percent, but HLA typing was technically demanding and expensive. By the time HLA panels were widely available, DNA STR typing was already proving more cost-effective and more powerful.

Secretor status as an adjunct: individuals who secrete ABO antigens into saliva and other body fluids carry a functional FUT1 gene. Approximately 80 percent of people are secretors. Secretor status is itself genetically determined and can be tested independently, adding a small increment of discrimination to the overall panel. Its primary forensic use, however, is in body-fluid identification rather than paternity testing, and it is rarely determinative in a paternity context.

Two statistical measures dominate serological paternity interpretation: the exclusion probability and the paternity index. They address different questions and should not be confused.

The exclusion probability (PE) for a single system is calculated from population allele frequencies. It represents the fraction of all men in the reference population who would be excluded as the father of a child with the observed maternal and child phenotypes. For the ABO system in a European population with a child of group B and a mother of group O, the PE depends on the frequency of the B allele and the frequency of all phenotypes that cannot carry a B allele. PEs from independent systems are combined using the formula: cumulative PE = 1 minus the product of (1 minus PE_i) for each system i. This assumes independence between systems, which is generally true for loci on different chromosomes.

The paternity index (PI) is a likelihood ratio. It answers a different question: given that the alleged father has not been excluded, how much more probable is the observed result if he is the biological father compared to a random man from the same population? A PI of 10 means the results are 10 times more likely if the alleged father is the biological father than if a random man is. The combined paternity index across all systems is the product of the individual system PIs, and it is then combined with a prior probability (usually 0.5 by convention, reflecting no prior assumption either way) using Bayes' theorem to produce a posterior probability of paternity.

Serological paternity testing has several recognised failure modes. Understanding them is essential both for the laboratory scientist and for the expert witness explaining results to a court.

• Chimerism: A chimeric individual carries two genetically distinct cell populations, typically arising from twin zygosity or bone marrow transplantation. Blood typing may yield discordant results between different sample types, or may show antigens inconsistent with the individual's apparent parentage. Chimerism can cause both false exclusions and confusing non-exclusions.
• Rare antigen variants and silencing mutations: Rare alleles that abolish antigen expression (null phenotypes, such as the Bombay Oh phenotype in ABO) can make a person appear to be of a group inconsistent with their true genotype. An Oh individual, who lacks all H antigen and appears to be group O regardless of their ABO genotype, can produce a child of group A or B, which would be falsely interpreted as an exclusion if the examiner is unaware of the Bombay phenotype.
• Recent blood transfusion: Transfused red cells from a different donor can persist in the recipient's circulation for several weeks. If the donor's blood group differs from the recipient's, the blood type result will be a mixture, potentially containing antigens the recipient does not genetically carry. All paternity samples should be obtained from individuals not recently transfused, and the history must be documented.
• Technical error: Sample mislabelling, reagent deterioration, and weak agglutination read as negative are sources of laboratory error. Kidd system antibodies (anti-Jk^a, anti-Jk^b) are notoriously labile and require fresh samples and careful technique. All results should be performed in duplicate with appropriate controls.
• Mutation at the antigen locus: Extremely rare spontaneous mutation in the child could produce an antigen not inherited from either parent. This is theoretically possible but documented cases are vanishingly rare and are not a routine explanation for an exclusion.

Standard practice requires that any exclusion be confirmed by repeat testing of a fresh independently drawn sample. Most accreditation bodies, including those following International Society for Forensic Genetics (ISFG) guidelines, require that an exclusion be observed in at least two independent systems, or confirmed in the same system on a fresh sample, before it is reported as definitive.

Serological methods were not limited to mother-child-alleged-father trios. They were also applied to broader kinship questions: full versus half siblings, grandparentage, and immigration cases where a claimed family relationship had to be assessed without the direct parent available for testing. The analytical logic extends from the same principles but involves additional steps and carries greater uncertainty.

Sibling testing by serology: two full siblings share, on average, half their genes. At any given blood group locus, the probability that both siblings inherited the same allele from one specific parent is 0.5. When both parental phenotypes are known, the probability of observing a particular sibling pair profile under the hypothesis of full sibship versus unrelated can be calculated. The discrimination is poor compared to DNA STR profiling, however, because most blood group systems have only two or three common alleles, and the chance of identical profiles by coincidence is relatively high. Serological sibling testing was most useful as a subsidiary corroboration, not as a primary determination.

Immigration kinship cases arose frequently in countries with family reunification visa programmes. UK courts handled a number of cases in the 1970s and 1980s where blood typing was used to assess claimed parent-child relationships in immigration disputes. The limitations of serology were acknowledged in these proceedings, and the introduction of DNA testing in the late 1980s rapidly transformed this practice. The UK Home Office began funding DNA testing for immigration kinship cases in 1989, effectively ending the era of serological immigration testing within a few years.

The legal treatment of serological paternity evidence varies across jurisdictions, largely reflecting when each jurisdiction updated its evidentiary rules to incorporate DNA. In most active paternity litigation today, DNA STR profiling is the method of choice, but serological evidence retains legal relevance in specific circumstances: archived historical cases, jurisdictions with limited laboratory infrastructure, and cases where samples exist only in serological form.

India: the Bharatiya Sakshya Adhiniyam 2023 (BSA 2023, which replaced the Indian Evidence Act 1872) governs the admissibility of scientific evidence including blood group reports. Indian courts have treated a definitive serological exclusion as sufficient to rebut a presumption of legitimacy, but a non-exclusion result alone has not been held to establish paternity. Courts have increasingly ordered DNA testing where blood group results are inconclusive or contested, following the Supreme Court's guidance in cases such as Banarsi Dass v. Teeku Dutta (2005).

United Kingdom: the Family Law Reform Act 1969 gave courts power to direct blood tests in paternity disputes and set out the use of test results. Amendments through the 1990s extended the power to direct DNA tests. The Human Fertilisation and Embryology Act 1990 and subsequent Family Procedure Rules allow courts to consider scientific evidence including DNA, and serological evidence is now rarely adduced as a standalone finding. A blood group exclusion remains legally valid if it meets the reliability standards applicable to any scientific evidence, but courts and practitioners will generally insist on DNA confirmation if any doubt exists.

United States: the Uniform Parentage Act (most recently revised in 2017 and adopted in various forms by US states) allows courts to order genetic testing and sets standards for the weight given to results. Under Daubert v. Merrell Dow Pharmaceuticals (1993), courts apply a reliability test to scientific evidence. Blood group serology satisfies Daubert for exclusion evidence, but courts have consistently required DNA as the primary method in contested cases. Historical blood group results in archived cases are sometimes re-evaluated by DNA typing from stored samples where available.

European Union: member states vary in their procedural treatment of paternity evidence, but all major forensic laboratories follow ISFG guidelines, which now mandate DNA STR profiling as the primary method and set minimum quality thresholds for the reporting of kinship statistics. Serological evidence from older cases may still be encountered in legal contexts, particularly in civil proceedings involving inheritance or social security entitlement.