On November 20, 2025, the New Jersey Supreme Court issued one of the most significant evidentiary rulings in the state’s criminal law in years. In State v. Nieves, the Court held that expert testimony diagnosing Shaken Baby Syndrome/Abusive Head Trauma (SBS/AHT) — specifically, the theory that shaking an infant without any impact to the head can cause a recognized cluster of injuries — is not reliable enough to be presented to a jury. The decision affirmed rulings by two trial courts and the Appellate Division, and it represents a rare and important judicial reckoning with what has long been treated as settled medical fact in the courtroom.
This post explains what the Court decided, why it matters, and what it says about the broader problem of unreliable science being presented as certainty in the courtroom.
The Two Cases
The decision consolidated two prosecutions involving similar facts. In the first, Darryl Nieves was the primary caregiver of an infant, D.J., who had a complicated medical history including premature birth and cardiac surgery. At eleven months old, D.J. experienced three episodes of seizure-like symptoms over two weeks, each occurring while Nieves was caring for him. D.J. was found to have bilateral retinal hemorrhages and bilateral subdural bleeding. A child abuse pediatrician reviewed his history, found no other explanation, and diagnosed SBS/AHT “within a reasonable degree of medical certainty.” Nieves was indicted on aggravated assault and child endangerment charges. A second defendant, Michael Cifelli, faced a similar diagnosis from the same physician under similar circumstances.
Both men challenged the admissibility of the State’s SBS/AHT testimony before trial, arguing the underlying science was not reliable enough to meet New Jersey’s evidentiary standard.
How New Jersey Tests Reliability
Under New Jersey Rule of Evidence 702, expert testimony is admissible only if the field of inquiry is developed enough that an expert’s opinion can be considered reliable. New Jersey applied the “Frye standard” for these cases, which asks whether the underlying science has gained general acceptance in the relevant scientific field — and courts can identify more than one relevant field, with general acceptance required in each. This matters enormously here: the SBS/AHT diagnosis is delivered by pediatricians, but its scientific foundation traces back to biomechanics, the engineering discipline that studies how physical forces affect the human body — the same field behind crash-test research used in vehicle safety design.
Tracing SBS/AHT Back to Its Scientific Roots
The Court’s opinion traces the diagnosis back to its scientific roots, and that history matters. In 1968, a neurosurgeon studying whiplash injuries from car accidents concluded that brain injury could result from rotational forces on the head and neck alone, without direct impact — research that had nothing to do with infants or shaking. Other researchers later relied on that study to hypothesize that a shaken infant could suffer similar injuries, and through a series of papers in the early 1970s, the theory of infant “whiplash-shaking” began to spread through the medical community.
In 1987, the first biomechanical study actually designed to test the shaking hypothesis using infant models was published — and it concluded that shaking alone does not produce the injuries associated with SBS/AHT. In 2002, the original whiplash researcher published a paper criticizing the way his own work had been used to support the theory. A decade later, the neurosurgeon who first proposed the shaking hypothesis questioned it as well. In other words, the diagnosis used for decades to support convictions was built on biomechanical assumptions that the biomechanical community itself never fully validated, and that some of the theory’s own originators eventually disavowed.
Why the State Lost
At the Frye hearing in Nieves, the State called a single expert, a child abuse pediatrician, who testified that shaking alone could produce the recognized triad of symptoms. The defense countered with biomechanical engineers who testified that no scientific testing has ever established that a human being can generate the physical force necessary to cause those injuries through shaking alone. The trial court found this decisive, noting that the few studies attempting to establish a force threshold relied on monkeys, wooden dolls, or other mechanical stand-ins for an infant’s body — models that cannot reliably substitute for human anatomy. No study has ever measured what shaking a real infant actually produces in force, because no such study could ethically be performed.
The Supreme Court agreed, and its reasoning turned on a critical concession by the State itself. In supplemental briefing, the State acknowledged that disagreement exists among biomechanical engineers generally regarding whether shaking alone can cause abusive head trauma. The Court treated that admission as significant evidence, on its own, that the biomechanical community has not reached general acceptance on the question — and emphasized that biomechanics is not peripheral, but the very field from which the SBS/AHT theory originated. The Court acknowledged general acceptance within much of the medical and pediatric community, but held that was not enough. The State did not meet its burden because biomechanics is also a relevant scientific community and general acceptance was lacking there.
What the Decision Does Not Do
The ruling is narrower than it may sound. The Court did not hold that abuse involving head trauma is unprovable, and did not bar physical evidence of abuse from reaching a jury. Where there is independent evidence — bruising, fractures, an admission, or any other physical indicator of impact or injury — the State remains free to present it. What the Court excluded is expert testimony asserting, as a matter of medical certainty, that the symptoms alone prove the child was shaken without any impact, when the underlying biomechanical science does not support that conclusion. The Court also left the door open for the future: New Jersey adopted a Daubert-based reliability standard for some expert testimony in 2023, and the Court noted that if new, reliable scientific evidence develops, the State could attempt to establish SBS/AHT’s reliability under that standard in a later case. This is not a permanent ban — it is a determination that, on this record, the science was not there yet.
The decision was also not unanimous. One justice dissented, arguing that SBS/AHT is endorsed by every major medical association involved in its diagnosis and treatment, that every other state to consider the question has allowed the testimony, and that the majority let a handful of biomechanical engineers override the broader medical consensus. The majority’s response was that the State’s own concession of disagreement among biomechanical engineers, combined with that field’s direct relevance to the theory’s scientific foundation, was enough on its own to defeat a claim of general acceptance — regardless of how many medical organizations had endorsed the diagnosis.
Why This Case Matters Beyond These Two Defendants
SBS/AHT prosecutions are not rare. For decades, a diagnosis delivered with confidence by a pediatric specialist has been sufficient, in courtrooms across the country, to support a conviction — even when the medical diagnosis itself was the only evidence of abuse. New Jersey’s highest court has now said, in a published and precedential decision, that this practice rests on a scientific foundation that has not been adequately tested, and that the field most directly responsible for that foundation has not reached consensus on its core premise.
This is what happens when a court takes seriously its role as a gatekeeper against unreliable expert evidence. Science adapted specifically to answer a legal question — rather than developed independently and later applied to one — deserves exactly this kind of scrutiny. The history traced in this opinion shows a theory that migrated from automobile whiplash research into infant medicine, gained momentum through repetition and institutional endorsement, and was never tested against the basic question of whether shaking an infant can actually generate the force the theory requires. That is not how reliable science works, and it is not how evidence that can take away someone’s liberty should reach a jury.
For families and defendants facing an SBS/AHT allegation in New Jersey, this decision is significant. It does not mean such a charge can never be proven — independent physical evidence of abuse remains fully admissible. But the diagnosis alone, offered as medical certainty that shaking without impact occurred, can no longer reach a New Jersey jury without first surviving the scrutiny this opinion demands.