Osteogenesis imperfecta is caused by genetic mutations that disrupt the production or structure of type I collagen, the protein that gives bones their flexibility and strength. At least 90% of people with the condition have a defect in one of two genes, COL1A1 or COL1A2, which carry the instructions for building this collagen. The remaining cases trace back to mutations in other genes that help process, fold, or transport collagen, or that guide bone cell development.
How Collagen Mutations Lead to Brittle Bones
Type I collagen is the most abundant protein in bone. It forms a rope-like structure called a triple helix: three protein chains wind tightly around each other, creating a fiber that resists stretching and absorbs impact. For this helix to form correctly, every third amino acid in the chain must be glycine, the smallest amino acid. Glycine is small enough to tuck into the interior of the helix where there’s almost no room.
When a mutation swaps glycine for a larger amino acid, the helix can’t fold properly. The misfolded chain slows assembly, causes the cell to over-modify the protein, and ultimately prevents normal collagen from being secreted into the bone matrix. The result is bone tissue that contains less collagen and collagen of poorer quality. Where the glycine substitution sits along the chain, which of the three chains carries it, and what amino acid replaces glycine all influence how severe the disease becomes. Substitutions in certain critical regions near one end of the helix tend to cause the most severe disruption of fiber assembly.
Quantitative vs. Qualitative Defects
Not all mutations work the same way. The distinction between “quantitative” and “qualitative” collagen defects is central to understanding why osteogenesis imperfecta ranges from barely noticeable to fatal.
Quantitative defects reduce the amount of collagen without changing its structure. Typically, one copy of the COL1A1 gene is silenced by a premature stop signal (a frameshift, nonsense, or splice site variant), cutting collagen production roughly in half. The collagen that does get made is normal, so bones form correctly but are thinner and more fragile. This mechanism underlies the mildest form, sometimes called type I, which is characterized by increased fracture risk, blue-gray tinting in the whites of the eyes, and hearing loss that often begins in adolescence or early adulthood. Long bone deformity is uncommon.
Qualitative defects produce collagen that is structurally abnormal. The glycine substitutions described above fall into this category. Because each collagen molecule contains chains from both the normal and mutant gene copies, even one bad copy can poison the final product. This “dominant-negative” effect means that far more than half of the collagen is defective. Qualitative mutations cause the moderate to severe forms of the disease, including progressively deforming types and the perinatally lethal form (type II), in which bones fracture before or during birth.
Inheritance Patterns
Most cases follow an autosomal dominant pattern, meaning a mutation in just one copy of the gene is enough to cause disease. A parent with the condition has a 50% chance of passing it to each child. However, many severe cases arise from new (de novo) mutations that neither parent carries. Type II, the most severe form, was once thought to be autosomal recessive but is now recognized as a dominant-negative disorder frequently caused by spontaneous mutations. Even when a child’s case results from a new mutation, there is an estimated 7% chance that a subsequent pregnancy could be affected, likely because one parent carries the mutation in a fraction of their egg or sperm cells without showing symptoms themselves.
Autosomal recessive inheritance accounts for fewer than 5% of cases. These involve mutations in genes other than COL1A1 and COL1A2, where a child must inherit a defective copy from each parent to develop the disease.
Genes Beyond COL1A1 and COL1A2
Researchers have identified more than a dozen additional genes that cause osteogenesis imperfecta when mutated. These genes don’t encode collagen directly. Instead, they affect the cellular machinery that modifies, folds, and transports collagen, or they influence how bone-forming cells develop. The conditions they cause are rarer but can be just as severe.
Collagen Processing Genes
Three genes, CRTAP, LEPRE1, and PPIB, encode the components of a complex that chemically modifies collagen after it’s built. Without this modification (called prolyl 3-hydroxylation), collagen chains don’t fold or assemble correctly. Mutations in CRTAP cause type VII, LEPRE1 causes type VIII, and PPIB causes type IX. All three are recessive, so both gene copies must be defective. These types tend to produce severe bone fragility and deformity.
Collagen Folding Genes
After collagen chains are assembled, specialized chaperone proteins guide the triple helix into its final shape and escort it out of the cell. Mutations in SERPINH1 (which encodes a chaperone called HSP47) cause type X, while mutations in FKBP10 cause type XI. FKBP10 mutations are particularly interesting because they can produce a spectrum of overlapping conditions previously thought to be unrelated, including osteogenesis imperfecta and Bruck syndrome, which combines brittle bones with joint contractures.
Mineralization and Bone Cell Genes
Some forms have nothing to do with collagen processing at all. Type V is caused by a single specific mutation in IFITM5, a gene active in bone cells during mineralization, the process of depositing mineral crystite onto the collagen scaffold. People with type V develop distinctive calcification of the membrane between their forearm bones, restricting wrist rotation, and are prone to exaggerated callus formation after fractures. Type VI results from recessive mutations in SERPINF1, which encodes a factor involved in regulating how minerals are deposited in bone.
Still other genes, including WNT1 and SP7, affect how bone-forming cells (osteoblasts) develop and mature. Dominant mutations in WNT1 have been linked to early-onset osteoporosis, while recessive mutations cause a more recognizable form of osteogenesis imperfecta. These discoveries have expanded the definition of the disease well beyond its original collagen-centered understanding.
How the Condition Is Detected
Genetic testing panels using next-generation sequencing can now screen all known osteogenesis imperfecta genes simultaneously, with analytical sensitivity above 98%. For families with a known mutation, testing can confirm or rule out the diagnosis quickly. For new cases, broader panels help distinguish osteogenesis imperfecta from other conditions that cause fragile bones.
Severe forms are sometimes detected before birth. On ultrasound, the hallmark findings are shortened, bowed, or angulated femurs (thighbones) that measure below the 5th percentile for gestational age, often visible by the mid-second trimester. Fractures may already be present. Increased nuchal translucency in the first trimester, a measurement of fluid at the back of the fetal neck, can serve as an early clue in some cases. Because several other skeletal conditions can look similar on ultrasound, genetic testing is typically needed to confirm the specific diagnosis.
Why Severity Varies So Much
One of the most striking features of osteogenesis imperfecta is its range. Some people experience only a handful of fractures in childhood and live with few limitations. Others are born with dozens of fractures and bones too fragile to support normal growth. This variation comes down to the type and location of the underlying mutation.
A person whose COL1A1 gene simply produces less collagen (a quantitative defect) will generally have milder disease than someone whose gene produces abnormal collagen that disrupts every fiber it touches (a qualitative, dominant-negative defect). Within qualitative mutations, those closer to certain critical regions of the collagen chain, or those that substitute particularly bulky amino acids for glycine, tend to cause more severe disease. The chain carrying the mutation matters too: substitutions in the COL1A1-encoded chain and the COL1A2-encoded chain can produce different severities even at equivalent positions.
For recessive forms caused by non-collagen genes, severity depends on whether the mutation eliminates the protein entirely or just reduces its function. Complete loss of the prolyl 3-hydroxylation complex, for instance, produces severe deformity, while partial loss may result in a milder presentation. This genetic complexity is why two people with “osteogenesis imperfecta” can have vastly different experiences, and why genetic testing has become essential for accurate prognosis and family planning.