Nexus over Comparators and FMLA Interference by Omission: Sixth Circuit Partially Revives Pregnancy Discrimination and FMLA Claims in Bunnell v. Beaumont Hospital

Court: U.S. Court of Appeals for the Sixth Circuit (unpublished, not recommended for publication)
Date: September 4, 2025
Case: Jackilyn Bunnell v. Wm. Beaumont Hospital, No. 24-1889
Panel: Judges Gibbons, White, and Murphy (White, J., for the Court; Murphy, J., concurring in part and dissenting in part)

Introduction

This appeal arises from a COVID-era layoff of an ultrasonographer, Jackilyn Bunnell, at William Beaumont Hospital’s Royal Oak, Michigan campus. After announcing her pregnancy, Bunnell sought reduced exposure to infectious disease and x-rays. Her supervisor, Tracy Zeiter, initially reacted poorly to a doctor’s note but later apologized and accommodated several requests. As the pandemic unfolded, the hospital redeployed staff, and later instituted a system-wide workforce reduction. Relying on 2019 productivity metrics, Zeiter recommended four employees for layoff, including Bunnell and another ultrasonographer who was pregnant.

Bunnell’s lawsuit asserted multiple federal and state claims: pregnancy discrimination and retaliation (Title VII and Michigan’s ELCRA), disability discrimination and retaliation (ADA and PWDCRA), and FMLA interference and retaliation. The district court granted summary judgment to Beaumont across the board. The Sixth Circuit affirmed most of that ruling, but revived two claims: pregnancy discrimination and FMLA interference.

The decision is significant for two reasons. First, it clarifies that a non-accommodation pregnancy discrimination claim need not present comparator evidence under Young v. UPS to survive summary judgment; showing a nexus between pregnancy and the adverse action suffices at the prima facie and pretext stages, particularly in a workforce-reduction case. Second, it recognizes that an employer may “interfere” with FMLA rights by withholding material job information when the employee returns from leave—even if hiring is coordinated by a different unit and HR controls the formal process.

Summary of the Judgment

Pregnancy discrimination (Title VII/ELCRA): Reversed and remanded. The panel held that a reasonable jury could find pretext in the supervisor’s selection of Bunnell for layoff, given remarks tying Bunnell’s pregnancy-related accommodations to departmental “strain,” the selection of the only two pregnant employees for layoffs, and questions about the “objectivity” and use of productivity metrics controlled in part by the supervisor.
Pregnancy-based retaliation (Title VII/ELCRA): Affirmed for the employer. Temporal proximity between Bunnell’s complaint to her supervisor and the layoff was not enough; no additional evidence of a causal link was presented.
Disability discrimination/retaliation (ADA/PWDCRA): Affirmed for the employer. Severe morning sickness can qualify as a disability in some circumstances, but there was no evidence Beaumont knew of this specific complication (the doctor’s note did not mention it), and no adverse action tied to a disability-based accommodation.
FMLA retaliation: Affirmed for the employer. No evidence that Bunnell opposed a practice made unlawful by the FMLA or that she was retaliated against for doing so.
FMLA interference: Reversed and remanded. The panel found a triable fact issue whether the supervisor’s failure to inform Bunnell of a contemporaneous full-time opening at Beaumont’s Dearborn facility—and her characterization of Bunnell as on “Medical Leave” without mentioning the same-day clearance to return—used FMLA leave as a negative factor in employment actions.

Judge Murphy dissented as to the pregnancy discrimination claim, arguing that the employer relied on neutral, objective productivity data during a bona fide workforce reduction and that the evidence did not permit a reasonable finding of pretext or animus. He concurred in reviving the FMLA interference claim.

Analysis

Precedents Cited and Their Role

Cline v. Catholic Diocese of Toledo, 206 F.3d 651 (6th Cir. 2000): Recited the prima facie elements of pregnancy discrimination: pregnancy, qualification, adverse action, and nexus. The majority applied Cline’s framework and focused on the “nexus” element, particularly at the pretext stage.
Barnes v. GenCorp Inc., 896 F.2d 1457 (6th Cir. 1990): In workforce reduction cases, a plaintiff must present “additional” evidence to indicate improper singling out. The majority reiterated Asmo’s teaching that showing a nexus between pregnancy and termination satisfies Barnes’s “additional evidence” requirement.
Asmo v. Keane, Inc., 471 F.3d 588 (6th Cir. 2006): Key comparator. The court drew parallels: a decisionmaker’s remarks and timing can support an inference of pregnancy-related animus and pretext even when employers cite neutral criteria. Although not identical factually, Asmo anchored the majority’s conclusion that Bunnell’s evidence surpasses mere disagreement with business judgments.
Young v. United Parcel Serv., 575 U.S. 206 (2015): The Supreme Court recognized disparate-treatment claims based on failure to accommodate pregnant workers relative to nonpregnant comparators. The majority distinguished Young, emphasizing that Bunnell’s claim was not accommodation-based. Therefore, she was not required to marshal comparator evidence of similarly situated nonpregnant employees—she could proceed by proving a nexus and pretext.
Chen v. Dow Chemical Co., 580 F.3d 394 (6th Cir. 2009): If an employer “reasonably and honestly” relies on particularized facts, it is entitled to summary judgment even if the decision later appears mistaken. The majority held Bunnell’s record evidence—especially supervisor remarks and the selection of both pregnant employees—created a fact dispute as to whether the reliance was honest or pretextual.
Marshall v. The Rawlings Co. LLC, 854 F.3d 368 (6th Cir. 2017): Cat’s-paw liability permits employer liability where a biased lower-level supervisor influences a neutral decisionmaker. The majority applied this framework: even though a higher-level manager approved the layoff, a jury could find the recommendation was tainted by the supervisor’s bias.
Risch v. Royal Oak Police Dep’t, 581 F.3d 383 (6th Cir. 2009): Courts assess the probative value of discriminatory remarks by considering position, content, timing, and connection to the challenged action. The majority found Zeiter’s remarks probative of animus.
Fuhr v. Hazel Park Sch. Dist., 710 F.3d 668 (6th Cir. 2013); Tuttle v. Metro. Gov’t of Nashville, 474 F.3d 307 (6th Cir. 2007); Mickey v. Zeidler Tool & Die Co., 516 F.3d 516 (6th Cir. 2008): Retaliation standards. Temporal proximity alone generally cannot establish causation, except in narrow circumstances involving immediate firing upon protected activity. These cases supported affirmance of the retaliation dismissals.
Seeger v. Cincinnati Bell Tel. Co., LLC, 681 F.3d 274 (6th Cir. 2012); Edgar v. JAC Prods., Inc., 443 F.3d 501 (6th Cir. 2006); Wysong v. Dow Chem. Co., 503 F.3d 441 (6th Cir. 2007); 29 C.F.R. § 825.220(c): FMLA interference doctrine and “negative factor” rule. The majority relied on these to hold that withholding information about a job opening and mischaracterizing a return-to-work status could be viewed as using FMLA leave as a negative factor.

Dissent’s authorities (illustrating the other side of the ledger):

Hedrick v. W. Reserve Care Sys., 355 F.3d 444 (6th Cir. 2004): Courts should not second-guess employers’ business judgments. The dissent saw the hospital’s productivity-based selection as a classic business judgment.
Brewer v. New Era, Inc., 564 F. App’x 834 (6th Cir. 2014); Rowe v. Cleveland Pneumatic Co., 690 F.2d 88 (6th Cir. 1982): Suspicion increases when employers rely on subjective criteria; here, by contrast, the employer relied on facially objective metrics.
McClellan v. Midwest Machining, Inc., 2022 WL 203205 (6th Cir. Jan. 24, 2022): Distinguishing Asmo; the dissent concluded Bunnell’s facts were more akin to McClellan, where neutral business reasons prevailed.

Legal Reasoning

1) Pregnancy discrimination: Nexus and pretext in a workforce reduction

The majority framed Bunnell’s claim as a standard disparate-treatment case—not a Young-style accommodation claim. Accordingly, comparator proof was not required. Instead, the court applied Cline and Asmo to ask whether Bunnell presented evidence from which a jury could infer a nexus between pregnancy and the layoff and find Beaumont’s proffered productivity rationale pretextual.

Several facts raised triable issues:

Zeiter’s remarks explicitly connected Bunnell’s pregnancy-related limitations to departmental productivity “strain,” and urged greater “flexibility” by entering contact rooms. The majority found these statements probative of animus under Risch.
Of the employees Zeiter selected for layoff, the only two pregnant (or recently pregnant) ultrasonographers were chosen.
Although Zeiter relied on 2019 productivity measures, she controlled (or strongly influenced) some metrics (e.g., portable assignments), did not account for factors potentially depressing Bunnell’s counts (e.g., higher no-shows in the Women’s Heart Center, where Bunnell worked hundreds of hours), and nevertheless laid off an employee who worked among the highest hours in the department.
Cat’s-paw theory was apt: even if the ultimate approval came from a higher-level manager (Burgett), a jury could find that Zeiter’s biased recommendation tainted the decision.

The majority concluded these facts could persuade a reasonable juror that the productivity rationale was not honestly relied upon, satisfying pretext under Chen.

Dissent’s response: Judge Murphy would have affirmed. He emphasized:

The undisputed, hospital-wide COVID-induced reduction for valid business reasons.
That Bunnell initially appeared on a layoff list supplied to Zeiter, and Zeiter first questioned why Bunnell was on it, suggesting the absence of bias.
The hospital’s reliance on seemingly objective productivity metrics, and the lack of concrete counter-data from Bunnell showing those metrics were manipulated or inaccurate.
The need to avoid substituting judicial judgments for business judgments about how to measure productivity.
Unlike Asmo, here Zeiter congratulated Bunnell and accommodated requests; there were no overtly pregnancy-hostile statements.

2) Disability discrimination and retaliation (ADA and PWDCRA)

The court reaffirmed that pregnancy itself is not a disability, though pregnancy-related complications can be. Bunnell asserted “severe morning sickness,” but the record lacked evidence that Beaumont knew of this specific condition (her doctor’s note did not mention it). The temporary removal from a scrub-distribution assignment was based on a documented complaint naming Bunnell for unprofessional behavior, and there was no evidence similarly situated comparators were treated better. Summary judgment for Beaumont was affirmed.

3) Retaliation (Title VII/ELCRA and ADA/PWDCRA)

Bunnell engaged in protected activity when she objected to Zeiter’s “strain” comments. Still, temporal proximity alone does not establish causation absent exceptionally close timing and circumstances. Here, the record lacked additional evidence linking the complaint to the layoff, and Zeiter’s immediate apology undermined animus. All retaliation claims were properly dismissed.

4) FMLA interference (but not retaliation)

On the same day Bunnell’s doctor cleared her to return from FMLA leave, a Dearborn department inquired about her availability for a full-time opening. Zeiter responded that Bunnell had been furloughed and moved to “Medical Leave,” without mentioning the same-day clearance, and did not inform Bunnell about the opening. She later testified she withheld the lead to avoid the perception of “pawning [Bunnell] off,” and preferred HR to run the process.

Drawing on 29 C.F.R. § 825.220(c), Edgar, and Wysong, the court held a jury could find that Zeiter’s omission used FMLA leave as a “negative factor” in employment actions—classic interference. The fact that HR had formal oversight and the posting was at a different campus did not foreclose liability; the key was that Zeiter was asked specifically about Bunnell and failed to convey material information to either side accurately. Deposition testimony suggested the higher-level manager (Burgett) was surprised to learn Bunnell had not been told, supporting a credibility issue for trial. The separate FMLA retaliation theory failed because there was no evidence Bunnell opposed an unlawful practice or suffered retaliation for such opposition.

Impact

For employees and plaintiffs’ counsel:

Comparator evidence is not always necessary. In non-accommodation pregnancy discrimination claims, survival at summary judgment can hinge on a nexus and pretext showing without comparator proof under Young.
Supervisor remarks matter. Statements linking pregnancy-related limitations to productivity or “strain,” coupled with selection patterns and control over metrics, can create a jury issue even in workforce reductions.
Cat’s-paw remains potent. Bias held by a recommending supervisor can taint an otherwise neutral approval structure.
FMLA interference by omission. Employers who fail to share material job information or who mischaracterize leave status risk an interference finding—no formal adverse action is required if leave is used as a negative factor.
Disability claims tied to pregnancy complications need employer knowledge. If a condition like hyperemesis gravidarum is central, ensure documentation informs the employer of the specific limitation.

For employers and defense counsel:

Document “honest reliance” carefully. In reductions, ensure metrics are consistently applied, account for supervisor-controlled variables (e.g., assignments), and document why chosen measures fairly reflect productivity.
Train supervisors on language and implications. Remarks about “strain” or “flexibility” in response to pregnancy-related accommodations can supply powerful pretext evidence.
Coordinate FMLA return-to-work communications. When an employee is cleared to return, ensure managers accurately convey status internally, including to other sites, and promptly inform the employee of relevant openings.
Centralized HR processes are not a shield. Decentralized omissions by line managers can support FMLA interference. Implement protocols requiring timely, documented notice to returning employees about suitable vacancies.
Retaliation prevention. Apologizing and correcting course—as Zeiter did—can help, but contemporaneous documentation and follow-through remain essential to defeat causation in retaliation claims.

Complex Concepts Simplified

Workforce reduction (Barnes) “additional evidence” requirement: In mass layoffs, plaintiffs must show something beyond the adverse action itself—direct, circumstantial, or statistical evidence—that suggests impermissible singling out. In pregnancy cases, showing a nexus between pregnancy and termination can satisfy this requirement.
Young comparators vs. non-accommodation claims: Young addresses accommodation disparities (pregnant vs. nonpregnant employees “similar in their ability or inability to work”). When the claim is not about accommodations but about discriminatory motive, comparator evidence is not strictly required; other evidence of pretext can suffice.
Cat’s-paw liability: An employer may be liable if a biased supervisor’s recommendation influences the ultimate decision, even if the final decisionmaker lacks bias.
Pretext: The employer’s explanation is a pretext if a jury could find it false or not honestly relied upon. Evidence can include inconsistent rationales, suspicious timing, selective metrics, or biased remarks.
FMLA interference vs. retaliation: Interference focuses on whether the employer impeded or penalized leave use (e.g., using leave as a negative factor), while retaliation requires proof the employee opposed an unlawful practice and suffered reprisal for it.
Temporal proximity: Close timing between protected activity and adverse action can support causation but usually needs additional evidence unless the timing is immediate and striking.

Conclusion

Although unpublished, Bunnell offers persuasive guidance within the Sixth Circuit. It underscores that pregnancy discrimination claims not grounded in accommodation disparity do not require comparator evidence under Young; plaintiffs can reach a jury by demonstrating a nexus between pregnancy and an adverse action with evidence of pretext, including supervisor remarks, selection patterns, and control over productivity metrics. On the FMLA front, the decision reinforces the “negative factor” rule: interference can arise from omissions—such as failing to inform a returning employee about an open position—and from mischaracterizing leave status, even when HR owns the formal process or the opening is at a different facility.

At the same time, the court reaffirmed the high bar for retaliation claims premised on temporal proximity and for pregnancy-related disability claims where the employer lacks knowledge of the specific disabling condition. The split between the majority and the dissent on the pregnancy discrimination claim highlights the continuing tension between deference to employer business judgments in reductions-in-force and the obligation to scrutinize seemingly objective criteria for potential bias, especially when decisionmakers’ remarks and selection outcomes cast doubt on the purported neutrality of the process.

Key takeaway: Words and omissions by line supervisors can be outcome-determinative. In reductions-in-force and leave-return scenarios alike, employers should ensure both their data and their communications are fair, accurate, and complete; plaintiffs should preserve and present concrete evidence connecting decisionmaker attitudes and actions to adverse outcomes.