Hereditary diseases pass from one generation to another when a pathogenic variant in DNA is copied into a gamete — an egg or sperm — and carried into the zygote at fertilization, then read by every cell that develops from it. Whether disease appears in the child depends on the variant's chromosome location, its dominance behavior, parental origin effects, and random chance.
That single sentence hides most of the biology. Inheritance is not one rule but at least five: autosomal dominant, autosomal recessive, X-linked, mitochondrial, and polygenic. Each follows a different arithmetic, and each fails in different ways. This explainer works through the mechanism, the numbers, the exceptions clinicians actually encounter, and what a family can do with the information.
The molecular chain from parent to child
A hereditary disease begins as a change in the DNA sequence of a germ cell. The change can be a single base substitution, a small insertion or deletion, a repeat expansion, or a larger structural rearrangement. Before meiosis, each chromosome replicates, so the cell holds two identical copies of every chromosome. Meiosis then runs two divisions and deliberately breaks and rejoins chromosomes during crossing over.
Crossing over is the key step most explanations skip. During prophase I, homologous chromosomes physically pair and swap segments at defined breakpoints. This reshuffling is why two siblings share, on average, 50 percent of their segregating alleles but can differ sharply at any single locus. Recombination also separates variants that sit far apart on the same chromosome while keeping close neighbors linked. A pathogenic variant embedded in a large inversion or a region of suppressed recombination can therefore travel with neighboring markers for generations — which is exactly how geneticists map disease genes and why some families show disease tracking with an apparently unrelated trait.
After recombination, the two meiotic divisions segregate chromosomes into haploid gametes. Errors here matter. Nondisjunction leaves a gamete with an extra or missing chromosome, producing aneuploidy such as trisomy 21. The risk rises steeply with maternal age: for Down syndrome, roughly 1 in 1,500 births at maternal age 20, about 1 in 400 at 35, and about 1 in 100 at 40 (CDC). Aneuploidy is a chromosomal, not a single-gene, form of heredity, but it follows the same core logic: the child's cells inherit a dosage error.
Once fertilization restores the diploid set, the zygote's genotype is fixed. Every subsequent cell inherits that genotype by mitosis, with occasional somatic mutations layered on top. A disease that depends on a variant present in every cell is constitutional; one that arises later in a subset of cells is somatic and typically not transmitted. This distinction — germline versus somatic — is the first thing a genetics clinic establishes.
What the odds actually are, condition by condition
Inheritance arithmetic is simple once the mode is known, but the modes produce very different family risks.
| Mode | Variant location | Risk to child if one parent affected | Risk if both parents carriers | Key example |
|---|---|---|---|---|
| Autosomal dominant | One of 22 autosome pairs | 50% per pregnancy | Up to 75% if both affected | Huntington disease |
| Autosomal recessive | Both copies of an autosome | Usually near 0% (parent has two variants) | 25% per pregnancy | Cystic fibrosis, sickle cell disease |
| X-linked recessive | X chromosome | Sons 50%; daughters usually carriers | Not applicable to carrier pair | Duchenne muscular dystrophy |
| X-linked dominant | X chromosome | 50% for either sex, often male-lethal | Variable | Fragile X (premutation effects differ) |
| Mitochondrial | Mitochondrial DNA | Variable, often near 100% from mother | Not applicable | MELAS, Leigh syndrome |
| Polygenic/multifactorial | Many loci plus environment | Recurrence ~2–5% for first-degree relatives | — | Type 2 diabetes, most congenital heart defects |
For autosomal dominant disease, each pregnancy is an independent 50 percent coin flip. A family with four unaffected children after an affected parent is unusual (probability 1 in 16) but not evidence against the diagnosis. For autosomal recessive disease, carrier parents have a 25 percent risk per pregnancy regardless of the outcomes of prior pregnancies. Roughly 1 in 25 people of European ancestry carries one cystic fibrosis variant, and about 1 in 2,500–3,000 newborns has the disease (CDC newborn screening data).
The counterintuitive detail is that most affected children of carrier parents have no family history at all. Recessive variants persist in populations because carriers are healthy, and two carriers usually have no idea they share a gene. Roughly 80 percent of children with a recessive disease are the first known case in their family. That statistic alone overturns the common assumption that heredity means visible family patterns.
Edge cases that break the textbook rules
Three mechanisms routinely surprise clinicians and families.
Reduced penetrance and variable expressivity. A dominant variant does not guarantee disease. Penetrance is the fraction of carriers who show any sign; expressivity is how severe it is. A variant in a BRCA gene may confer a markedly elevated lifetime cancer risk without causing cancer in every carrier, and the same variant can produce early-onset disease in one relative and late-onset disease in another. Age-dependent penetrance means a 30-year-old carrier may simply not have reached the time window yet.
Anticipation and repeat expansion. Some diseases worsen across generations because a short tandem repeat grows longer when transmitted. In Huntington disease, the CAG repeat in HTT expands, particularly when inherited from the father, and larger repeats produce earlier onset. A parent with 40 repeats may have a child with 50 and a much earlier age of onset. This is genuine biological change across generations, not just earlier detection.
Parent-of-origin imprinting and mitochondrial bottleneck. Some genes are expressed only from the maternal or only from the paternal copy. Prader-Willi and Angelman syndromes arise from the same chromosomal region but differ by which parent contributed the deleted or silenced segment. Mitochondrial DNA is inherited almost exclusively from the mother, but the fraction of mutant mitochondria passed on varies because of a bottleneck during oocyte development — so a mildly affected mother can have a severely affected child, or the reverse.
A professional caveat belongs here: genotype does not equal prognosis. Two people with identical pathogenic variants can have different outcomes because of modifier genes, environment, and stochastic developmental events. Genetic testing identifies risk architecture, not destiny, and counseling must reflect that uncertainty rather than promise a course.
Misconceptions worth correcting
Several false beliefs drive unnecessary fear or false reassurance.
"It skipped a generation, so the gene is gone." Dominant variants with reduced penetrance can appear absent in one generation and reappear in the next; the variant was present throughout. "No family history means no genetic risk." New mutations and recessive carrier matings produce affected children with no prior cases, as noted above. "A carrier is mildly affected." For recessive conditions, carriers are typically unaffected; being a carrier is not a mild form of the disease. "Boys only get X-linked disease from their mother." True for the X chromosome itself, but a de novo variant can arise in the mother's egg or early in the embryo without any maternal carrier status. "50 percent risk means half the children will be affected." Each pregnancy is independent; the 50 percent figure describes probability per conception, not a quota.
A related misconception is that genetic risk is fixed at birth. Epigenetic regulation, somatic mutation accumulation, and environmental exposures modulate expression over a lifetime, which is why identical twins can diverge clinically.
Practical actions a family can take
The steps below are general and should be tailored by a genetics professional to the specific condition.
Gather a three-generation family history: ages at diagnosis, causes of death, consanguinity, and any early-onset cancers or congenital anomalies. Ask a clinician whether the suspected condition has a validated genetic test and whether testing is indicated before or after symptoms appear. For carrier screening, know the panel's limits: expanded panels cover hundreds of recessive conditions, but a negative result does not exclude every possible variant. If a pathogenic variant is found, discuss reproductive options — prenatal diagnosis, preimplantation genetic testing, or gamete donation — with a counselor rather than acting on internet summaries. For dominant conditions with adult onset, understand the distinction between diagnostic and predictive testing and the insurance and psychological implications before testing.
One explicit threshold matters for interpretation: a variant is classified on a five-tier scale from pathogenic to benign, and a "variant of uncertain significance" is not a diagnosis. Roughly a substantial share of variants initially reported as uncertain are later reclassified, so recontact with the testing laboratory every few years is reasonable practice.
What inheritance models cannot tell you
Inheritance models describe probabilities, not individual outcomes. They cannot predict which child inherits a variant, how severe disease will be, or when symptoms begin. Polygenic conditions — most cases of type 2 diabetes, hypertension, and common congenital heart defects — do not follow single-gene arithmetic at all; their familial recurrence of roughly 2–5 percent reflects many small-effect variants plus environment. Epigenetic effects, somatic mosaicism, and structural variants below the resolution of standard testing add further uncertainty.
Testing technology also sets limits. Standard sequencing misses large deletions, repeat expansions, and deep intronic variants unless specific assays are ordered. A negative result therefore means "not detected by this test," not "no genetic cause." This is why genetic counseling pairs test interpretation with family history rather than replacing one with the other. Understanding the mechanism tells you the rules of the game; it does not tell you the score.
This article is general educational information about inheritance, not a diagnosis, treatment plan, or genetic counseling opinion. It does not promise any outcome for any individual or family. For personal risk assessment, testing decisions, or reproductive planning, consult a qualified clinical geneticist or genetic counselor in your area; for a medical emergency, contact local emergency services.