Frye Requires Objective Scientific Support for Medical Causation: Association Between Prematurity/Low Birth Weight and Autism Is Insufficient

1. Introduction

Terehoff v Frenkel (2026 NY Slip Op 02688) is a medical malpractice appeal arising from obstetrical care. The plaintiff (an infant, by his mother and natural guardian) sued the mother’s obstetrician and the obstetrician’s practice, alleging a failure to diagnose preterm labor. The infant was born at 25 weeks, weighing 770 grams, and was later diagnosed with autism.

The central appellate issue was evidentiary: whether the plaintiff could present expert testimony that the defendants’ alleged failure to diagnose and treat preterm labor—specifically, the claimed failure to prolong the pregnancy by about two weeks— was a substantial factor in causing the child’s autism. The defendants sought to preclude that causation theory under Frye v United States (293 F 1013 [DC Cir]).

2. Summary of the Opinion

The Appellate Division, Second Department reversed the judgment insofar as appealed from and held that the Supreme Court erred in denying the defendants’ motion to preclude the plaintiff’s neurological expert’s causation testimony. The court concluded that the expert’s opinion rested on association/correlation drawn from observational studies, not on a generally accepted scientific basis establishing that prematurity (or being born at 25 weeks rather than 27 weeks) causes autism. It therefore granted preclusion of that causation theory and remitted for a new trial. The plaintiff’s cross-appeal regarding damages (including the lack of an award for future pain and suffering) was dismissed as academic.

3. Analysis

3.1 Precedents Cited

  • Frye v United States (293 F 1013 [DC Cir])
    The foundational New York rule: scientific expert testimony is admissible only if the underlying principle/procedure has gained general acceptance in the relevant field. The court applied Frye not to a novel lab test, but to a causation theory.
  • Lugo v New York City Health & Hosps. Corp. (89 AD3d 42)
    The opinion draws heavily from Lugo for three propositions: (i) Frye governs scientific-principle evidence; (ii) Frye’s focus is on whether the opinion rests on sufficiently established, generally accepted principles—not whether a particular expert is persuasive; and (iii) New York has expanded Frye to psychological/physiological theories and theories of causation. Critically, Lugo supplies the refined causation-focused formulation: the inquiry asks for a “substantive, demonstratable, objective” basis and a “reasonable quantum of legitimate support” in the literature.
  • People v Wesley (83 NY2d 417)
    Cited for New York’s adherence to Frye as the admissibility framework for scientific evidence.
  • Zito v Zabarsky (28 AD3d 42)
    Used for two core points: (i) the proponent bears the burden to show general acceptance; and (ii) a causation theory must be more than “theoretical speculation” or a “scientific hunch.”
  • Del Maestro v Grecco (16 AD3d 364)
    Reinforces that the burden of proving general acceptance rests on the party offering the expert testimony.
  • Marsh v Smyth (12 AD3d 307)
    Quoted (via Lugo) for the causation-specific Frye lens: the focus is not mere popularity of a view, but whether the field has an objective, demonstrable basis supporting the expert’s conclusion and legitimate support in the literature.
  • Ratner v McNeil-PPC, Inc. (91 AD3d 63)
    This decision supplies the key doctrinal bridge between association and causation. Ratner is cited to explain that observational studies can show association, but expert testimony may be precluded where it extrapolates from such studies to assert causation, particularly where there is an “analytical gap” between data and conclusion.
  • General Elec. Co. v Joiner (522 US 136)
    Quoted (via Ratner) for the “analytical gap” concept: courts may reject expert opinions where the leap from data to conclusion is too great.
  • Heckstall v Pincus (19 AD3d 203)
    Quoted (via Ratner) for the proposition that observational studies or case reports are not generally accepted in the scientific community on questions of causation.

3.2 Legal Reasoning

The court’s reasoning proceeds in four steps:

  1. Frye applies to causation theories, not only to tests.
    Relying on Lugo v New York City Health & Hosps. Corp., the court reiterates that New York courts use Frye to screen expert causation opinions where the theory itself depends on scientific propositions that must be generally accepted.
  2. The plaintiff bore the burden to show general acceptance/objective support.
    Under Zito v Zabarsky and Del Maestro v Grecco, the plaintiff had to demonstrate that the scientific field (here, pediatric neurology/related disciplines) generally accepts the proposition being offered: that prematurity/low birth weight, and specifically a two-week prolongation from 25 to 27 weeks, is a causal mechanism for autism.
  3. The offered proof showed association, not causation.
    The challenged neurologist expert testified that his theory was one of “association or causal inference” and acknowledged that the causes of autism are unknown, relying on observational studies. The court found that none of the relied-upon studies concluded that premature birth causes autism; at most, they reported an association between prematurity and later autism diagnosis. Under Ratner v McNeil-PPC, Inc. and Heckstall v Pincus, that is an insufficient foundation to offer a causation opinion to the jury.
  4. The “analytical gap” was too large.
    Even accepting statistical findings about prevalence of autism among preterm/low-birth-weight infants, the expert simultaneously maintained that autism’s pathophysiology and causes are unknown. The court held this created too great an “analytical gap” between the data and the proffered causation conclusion (citing Ratner v McNeil-PPC, Inc. and General Elec. Co. v Joiner). Because the plaintiff could not show acceptance that a 25-week vs. 27-week gestational age difference increases autism risk in a way that supports legal causation, the testimony should have been precluded.

The remedy followed from the evidentiary error: because the jury found liability and causation based on a theory the court deemed inadmissible under Frye, the judgment could not stand, and the case was remitted for a new trial.

3.3 Impact

Practical tightening of causation proof in prenatal/neonatal malpractice claims. The decision underscores that plaintiffs cannot bridge the causation gap from prematurity/low birth weight to autism using observational association alone. Where the medical community acknowledges that the etiology of autism is unknown, courts will scrutinize whether the expert can point to an objectively supported and accepted causal pathway (or other robust scientific support), rather than epidemiological correlation.

Stronger role for Frye hearings in “mechanism-unknown” conditions. The court’s focus on the lack of accepted mechanism and the absence of field acceptance (e.g., among pediatric neurologists) signals that litigants should expect intensive Frye gatekeeping when experts attempt to convert statistical associations into case-specific causation for complex neurodevelopmental diagnoses.

Litigation strategy consequences. Defendants may increasingly frame motions around (i) whether cited literature actually states causation, (ii) whether the relevant specialty accepts the causal inference, and (iii) whether there is an “analytical gap” between epidemiology and the expert’s ultimate conclusion. Plaintiffs, correspondingly, must build a record showing not just correlation but scientifically grounded causation support that satisfies Frye as articulated through Lugo v New York City Health & Hosps. Corp. and Ratner v McNeil-PPC, Inc..

4. Complex Concepts Simplified

  • Frye test (general acceptance).
    A rule that limits certain expert scientific testimony unless the underlying scientific principle is generally accepted as reliable in the relevant expert community.
  • Association vs. causation.
    An association means two things occur together more often than expected; it does not prove one causes the other. Causation requires evidence that one factor actually produces the other (not merely that they correlate).
  • Observational studies.
    Studies that observe outcomes without controlling exposures through random assignment. They can suggest correlations but are often limited in proving causation because other factors (“confounders”) may explain the relationship.
  • Analytical gap.
    The logical leap between data (e.g., population statistics) and an expert’s conclusion (e.g., “this caused the plaintiff’s autism”). If the leap is too large, the testimony may be excluded.
  • Substantial factor (legal causation).
    In negligence/medical malpractice, the plaintiff must show the defendant’s departure was a meaningful contributing cause of the injury—not merely present in the background.

5. Conclusion

Terehoff v Frenkel reinforces and operationalizes a stringent Frye gatekeeping rule for medical-causation opinions: where the expert’s theory rests on observational associations and the condition’s etiology/mechanism remains unknown, New York courts may find an impermissible analytical gap and preclude the testimony. The decision’s central significance is its insistence that correlation-based literature—without objective, field-accepted support for causation—cannot be used to present to a jury the claim that prematurity (or a specific two-week gestational difference) caused autism, warranting reversal and a new trial when such testimony drives the verdict.