Three-generation family history: the clinical minimum pedigree
Three generations is the default standard for clinically useful family history in genetics. This page explains why, what to record at each generation, efficient intake approaches that save clinic time, and how modern pedigree software compares with paper-based family-history cards such as the NHS HEE template.
Written for genetic counsellors, clinical geneticists, and primary-care clinicians increasingly asked to collect initial family history as part of cancer-risk and reproductive-risk pathways.
Short version. The clinical minimum family history in genetics is three generations: the proband's own generation, the parents and aunts/uncles, and the grandparents. Three generations supports recognition of Mendelian inheritance patterns and is the conventional input depth for cancer risk models. Record sex, year of birth, year and cause of death, known conditions with age at diagnosis, ancestry, and reproductive history on each individual. Paper aide-memoires such as the NHS HEE family-history card work well for patient-completed intake; digital pedigree software turns the captured data into structured analysis.
Why three generations
Three generations is a threshold rather than a hard rule, but it is the threshold at which the pedigree starts to be informative for genetic analysis.
- Inheritance pattern recognition: single-generation or two-generation pedigrees rarely show vertical transmission clearly. Dominant and recessive patterns are usually only distinguishable from three generations up.
- Cancer risk models: BRCAPRO, MMRpro, PancPRO, and similar tools use first-, second-, and often third-degree relatives. Two generations misses the grandparents and aunts/uncles, losing a substantial proportion of the signal.
- Consanguinity loops: shared ancestors typically sit two or more generations back. Shorter pedigrees miss them.
- Founder effects and ancestry: population-specific carrier frequencies matter only if ancestry is recorded, which usually sits at the grandparent generation.
Extending beyond three generations improves risk-model accuracy further, particularly for sparse or cascade families. Three is the floor.
What to record at each generation
| Generation | Individuals | Minimum data to capture |
|---|---|---|
| I (grandparents) | 4 grandparents, plus their siblings where significant | Sex, year of birth, year and cause of death, known conditions with age at diagnosis, ancestry |
| II (parents, aunts, uncles) | Parents and their siblings (and their partners for the children generation) | Parents in full detail; aunts/uncles with conditions and ages |
| III (proband, siblings, cousins) | Proband, full siblings, half-siblings, first cousins | Full clinical detail for proband; key conditions and ages for siblings; known conditions for cousins |
| IV (children, nieces, nephews) | Children of proband and siblings, where present | Full detail if relevant (for example, affected child as the diagnostic anchor) |
What counts as a "known condition"
The conditions worth recording depend on the clinical context, but a general minimum includes:
- Cancers, with site and age at diagnosis (age at diagnosis is a critical input for hereditary cancer risk models).
- Inherited conditions (any diagnosed genetic disorder).
- Developmental disorders and intellectual disability.
- Early-onset cardiovascular disease and sudden unexplained death (relevant for hypertrophic cardiomyopathy, long QT, familial hypercholesterolaemia).
- Recurrent pregnancy loss and stillbirth.
- Congenital anomalies.
- Childhood deaths of unclear cause.
Age at diagnosis is more informative than simply noting the condition. A single relative with breast cancer at 70 is a different clinical signal from a single relative with breast cancer at 35.
Ancestry and consanguinity
Ancestry is typically recorded at the grandparent level because carrier frequencies and founder variants are population-specific. Specific ancestry matters particularly for Ashkenazi Jewish, Mediterranean, sub-Saharan African, South Asian, and French Canadian populations among others, where specific carrier screening panels have higher yield.
Consanguinity — known unions between related individuals — should be asked about directly and respectfully. Many patients do not volunteer the information without prompting. Record both known consanguinity in the parents' generation and community endogamy in the grandparents' generation.
Efficient intake approaches
Getting a full three-generation family history in a short clinical appointment is challenging. Effective approaches include:
- Pre-visit patient-completed questionnaires that capture the core structure and conditions, letting the clinician focus time on clarifying and annotating.
- Structured intake templates (paper or digital) that remind the patient to include aunts, uncles, and cousins — often missed when the patient reports only their "close" family.
- Live gesture drawing during the consultation for building and correcting the pedigree in real time.
- Follow-up rounds for extended family history, particularly for hereditary cancer pathways where the initial pedigree may need additional affected relatives identified through subsequent discussion with the patient's family.
NHS HEE family-history template and paper tools
Paper-based family-history templates have a place in the workflow. The NHS Health Education England (HEE) family-history template card, used widely in UK primary care and genetics, is a pocket-sized aide-memoire for collecting a three-generation history. Similar cards and forms are used internationally. They are useful because:
- Patients can complete them at home, reducing clinic time.
- They provide a consistent prompt for the clinician and patient.
- They are low-friction and universally accessible.
Paper templates are inherently limited. They capture unstructured free text; they do not enforce standard notation; they do not run inheritance pattern analysis or risk models; and they cannot be shared between systems without re-entry. The practical workflow combines paper intake with digital structured entry: the paper card is the patient's recall tool, and the pedigree software is the analytical and archival record.
Common gaps in three-generation histories
Common omissions worth checking for:
- Aunts and uncles, particularly on the paternal side where patients often know less.
- Miscarriages and terminations — frequently unreported unless directly asked.
- Childhood deaths and stillbirths, sometimes not mentioned because they are "not in the family story."
- Consanguinity, frequently not volunteered.
- Adoptions and donor conceptions, with biological vs social ancestry.
- Cancers in older generations where the primary site was unknown or mis-recalled.
How Evagene supports this
Evagene's pedigree model supports three generations as a default and handles extensions to four and beyond when clinically relevant. Core data — sex assigned at birth, gender identity, year of birth and death, known conditions with age at diagnosis, ancestry, consanguinity — are captured as structured fields against each individual, rather than as free text. The full three-generation structure feeds directly into:
- Mendelian inheritance analysis for AD, AR, XR, and other patterns.
- BRCAPRO, MMRpro, and PancPRO cancer risk models.
- Carrier probability calculation with Bayesian updating.
- Consanguinity detection with Wright's coefficient.
- Batch risk screening across the 200+ catalogued conditions.
Intake is supported through multiple routes: gesture drawing during consultation, GEDCOM and JSON import from other pedigree systems, 23andMe uploads for genotype-level data, and OCR from pedigree images where only a picture is available. Services that use paper intake cards continue to do so, with the resulting data captured into Evagene for structured analysis and long-term storage.
Frequently asked questions
Why three generations?
The minimum for pattern recognition and for standard cancer risk models.
What do I capture per individual?
Sex, birth year, death year and cause, conditions with age at diagnosis, ancestry, and reproductive history.
Is age at diagnosis important?
Yes — it is a key input for age-dependent penetrance and for cancer risk models.
Is paper intake still useful?
Yes, as an aide-memoire for patient-completed intake, feeding into digital structured capture.
What about four generations?
Improves accuracy further; capture where practical, especially for hereditary cancer pathways.
Does Evagene handle three generations?
Yes, as the default model, with extension to deeper pedigrees when needed.