A vial of blood, and a question that used to take years
A mother mails in a blood sample. About two weeks later, a healthcare provider reads back a result. That result does not describe her child's behavior. It describes her own antibodies - and whether they carry a pattern tied to a biological form of autism. This is MARAbio's whole proposition, and it is stranger and simpler than it sounds.
MARAbio Systems is a roughly 20-person company in Salt Lake City that does one unusual thing well: it turns maternal immunology into a number. In December 2025 it began selling the MAR-Autism test, which it calls the first-ever blood test to identify Maternal Autoantibody Related Autism, or MARA. The company would like you to be skeptical, because skepticism is exactly what a measurement is for.
Autism is usually diagnosed by waiting
For most families, the path to an autism diagnosis runs through observation - milestones missed, checklists filled in, specialists booked months out. It works, eventually. Eventually is the problem. By the time behavior is legible, the developmental window where intervention helps most has often narrowed.
MARAbio's founding scientist, Dr. Judy Van de Water, spent more than two decades at the UC Davis MIND Institute chasing a different lead. In a subset of cases, a mother's own IgG autoantibodies cross the placenta, slip past the fetal blood-brain barrier, and bind proteins that the developing brain is busy using to wire itself. The result is not a behavior that emerges. It is a mechanism that already happened.
The uncomfortable part: this is not a rare footnote. MARAbio cites estimates that MARA may account for up to about one in five autism cases - more common, by that math, than single genetic causes like Fragile X.
Above: the quiet heresy of the company - that some autism has a cause you can see in a tube, not just a pattern you notice over time.
Take 20 years of academic research and ship it
Academic discoveries rarely walk out of the lab on their own. MARAbio's bet was that Van de Water's research was specific enough, and validated enough, to become a regulated product rather than a citation. The company secured exclusive rights to her body of MARA work from UC Davis - the scientific spine of everything it sells.
Then it brought in people who had shipped diagnostics before. Michael S. Paul, a former Lineagen chief, joined as a consultant in 2021 and became CEO in 2023. Dr. E. Robert Wassman - four decades in genetic diagnostics, prior chief medical roles at Genzyme Genetics and Ambry - runs clinical affairs as Chief Medical Officer. The board chair is Vincent Forlenza, who used to run the medical-device giant Becton Dickinson. It is, notably, not a roster assembled to write papers.
Eight proteins, one assay, a fortnight
The MAR-Autism test is an indirect multi-ELISA assay. In plain terms, it checks a mother's blood for specific combinations of IgG autoantibodies against eight proteins that the fetal brain leans on heavily during development: CRMP1, CRMP2, GDA, NSE, LDHA, LDHB, STIP1 and YBX1. Several are ordinary cellular workhorses - two of them are just lactate dehydrogenases - which happen to be heavily expressed exactly where and when a brain is being built.
It is not the presence of any single antibody that matters. It is the pattern. Certain combinations - the CRMP1 plus CRMP2 pairing is the headline example - carry a specificity reported at 99% or higher for the MARA subtype. A provider orders the test, the sample is collected at home or in office, and results come back in roughly 14 days.
One honest limitation, which the company states plainly: the test is given to the mother, and it is currently validated for mothers who already have a child with autism or showing signs of it. General-population validation is still underway. The test does not diagnose a specific child in utero; it reads risk from the parent who carries it.
The road from lab to launch
A decade is a long time to spend turning one good idea into one shippable tube. MARAbio took it anyway.
The numbers that earned the round
Investors do not fund mechanisms; they fund evidence. MARAbio's pitch leaned on specificity. In the underlying research, certain autoantibody combinations were associated with dramatically higher odds of an autism diagnosis - on the order of 22 to 31 times versus controls - and were present in roughly 10 to 18% of autism cases while appearing in essentially none of the controls. That last figure is the one that makes a diagnostic possible: a marker that almost never shows up where it shouldn't.
Why specificity is the whole game
Four bars, one argument: it is the gap between the top bars and the bottom one that makes a test worth selling.
The money followed. In October 2024, MARAbio announced a $19M initial Series A led by MAK Capital One, with the Autism Impact Fund among the named backers and the round staying open into year-end. The use of funds was unglamorous and exactly right: launch the test and push the science toward therapy.
Detection is the first half of the sentence
MARAbio writes its vision as a single line: a world in which the most severe effects of autism do not occur. A test, by itself, does not deliver that world. It only tells you where you stand. The company's longer game is the second half of the sentence - a therapeutics pipeline aimed at neutralizing the offending maternal antibodies before they reach a developing brain at all.
That is the logic that makes the diagnostic more than a diagnostic. If you can reliably detect the antibody, the argument goes, you have a target you can eventually block. Detection becomes the on-ramp to prevention. It is an ambitious bet, and the company is careful not to oversell a therapy it has not yet delivered.
The wait, reconsidered
Go back to the vial of blood. For families with one autistic child weighing another pregnancy, the question has always been answered with a shrug and a statistic. MARAbio replaces the shrug with a measurement - imperfect, still validating, honest about its limits, but a measurement. That is the change. Not certainty. Specificity.
If the general-population validation holds, and if the therapeutics follow, the mother who mails in a sample in 2030 may get back something better than a risk score. She may get a plan. MARAbio has not earned that sentence yet. It has earned the right to try.
The company that decided a hunch was not good enough - and that a fortnight beats a five-year wait.