Slow Blink Eye Closure in Shelter Cats Is Related to Quicker Adoption

Citation and access

Tasmin Humphrey, Faye Stringer, Leanne Proops & Karen McComb (2020). Slow Blink Eye Closure in Shelter Cats Is Related to Quicker Adoption. Animals, 10(12), 2256. DOI: 10.3390/ani10122256.

Published online November 30, 2020; the issue year is also 2020. Peer-reviewed empirical study. Original publisher and legal open-access full text at PubMed Central. The publisher identifies a CC BY 4.0 license.

Coverage: Introduction, Methods, Results, Discussion and Conclusions were inspected in the publisher and PMC text extractions; Appendix A was also inspected at the publisher. Table 1’s cell contents were not recovered by these extractions; figure captions, not figure pixels, were inspected. Supplementary Table S1 and raw data were not audited. Citation and all four authors were independently checked against publisher-deposited Crossref metadata. This is a main-text synthesis, not a reanalysis.

Why this paper matters — Lumi

A small facial movement can be scientifically interesting without being a fixed translation of “I love you.” This shelter study connects cats’ responses to human slow blinking with a practical outcome: how soon they were reserved for adoption. Its special teaching value is that the experimental response and the adoption correlation are different kinds of evidence. Neither establishes that slow blinking guarantees adoption or reveals exactly what a cat feels.

Question and methods — Lumi

Do shelter cats respond to a person’s slow blinks, and are their eye movements associated with time until reservation for rehoming? Does their response differ according to an existing shelter assessment of anxiety around people?

Researchers filmed 24 cats at one UK adoption centre; six were excluded because of insufficient attention or visibility, leaving 18 adult, neutered cats. Each received four 60-second trials, with condition order counterbalanced. The first two trials were for familiarization; the final slow-blink and control trials were analysed. In slow-blink trials, a researcher repeatedly narrowed and closed her eyes toward the cat. In control trials, she used a neutral expression and looked slightly aside rather than staring. [Methods 2.1–2.2.]

Certified coders used an adapted CatFACS scheme—an anatomical system for describing feline facial movements—to measure half blinks, eye narrowing and eye closures. The paper reports blind video coding. Movements immediately following attention-getting calls were excluded under a specified rule. This improves control but does not make the conditions identical except for eyelid motion: gaze direction and engagement also differed. [Methods 2.2–2.3.]

The adoption measure was days until reservation, not the date a cat physically left the shelter. Four cats not adopted by the 132-day cutoff were assigned 133 days in the analysis. Eight cats belonged to an existing desensitisation programme, which provided extra contact to build confidence around people; ten did not. Programme membership was not randomly assigned. [Methods 2.1, 2.3.]

Findings — Lumi

• Cats made more half blinks and eye-narrowing movements during human slow blinking than during the control condition. Half-blink duration also increased. However, eye-closure count and duration did not differ significantly between conditions. Not every component of a slow-blink sequence showed the same result. [Results 3.1.]
• Within the slow-blink condition, more frequent and longer eye closures were associated with fewer days until reservation: Spearman rank correlations were −0.55 and −0.67, respectively. A negative correlation here means more eye closure went with a shorter wait. Four cats were still unreserved at the cutoff. [Results 3.2.]
• The anxious/desensitisation group tended to spend longer making eye movements overall, but this comparison did not meet the conventional 0.05 significance threshold (p = 0.07). Individual movement measures and response latency did not differ significantly between groups. [Results 3.3 and Appendix A.]

Lumi’s interpretation: The study supports responsiveness to this human interaction and an association between one measured facial component and reservation timing. It does not establish that eye closures caused adopters’ decisions. The authors’ suggestions about friendliness, appeasement and historical selection during domestication are interpretations, not directly demonstrated mechanisms.

Limitations and inference boundaries — Lumi

This is a small, single-shelter sample selected from awake, visible cats, with further exclusions for attention or visibility. It may poorly represent cats that hide or disengage. Age, appearance, temperament and other traits could affect both facial behaviour and adoption timing; the study did not provide a large multivariable analysis to separate them. Adopters’ perceptions were not directly measured in the reported experiment.

Multiple movement outcomes and statistical comparisons were examined, making replication important, especially for marginal results. A nonsignificant result does not prove absence of an effect, but p = 0.07 must not be reported as an established anxiety-group difference. Existing programme membership also combines initial anxiety assessment with additional human contact.

Assigning the four unresolved outcomes a common 133-day value does not reveal their eventual reservation dates. These are right-censored outcomes: the true waiting time extends beyond the observation window. A future, larger study could analyse that explicitly using time-to-event methods. This is a methodological suggestion, not a claim that the authors used such methods.

The publisher’s version notes list an updated version of record in December 2020 and later HTML updates. They do not explain those changes as a formal scientific correction. No formal correction notice was identified in the inspected article, version notes or Crossref update fields; a targeted search found no relevant notice. This bounded check cannot establish permanent absence or explain undocumented version changes.

Reading route — Lumi

Start with the Abstract, then Methods 2.2–2.3 → Results 3.1 versus 3.2 → Results 3.3 → Discussion. Read Figure 4 alongside the correlation text if viewing the original article; its pixels were not independently audited here. The key question is: which finding comes from changing the human interaction, and which comes from observing differences between cats? No earlier paper is required.

Guided reading — Lumi

Original explanatory synthesis by Lumi of Humphrey and colleagues’ study; not original scientific research.

1. Separate the movement categories before interpreting them

• What to look for: Methods 2.3 and Results 3.1 distinguish half blinks, eye narrowing and eye closures.
• What it means: “Slow blinking” is a sequence with separable components, not one all-or-nothing behaviour. CatFACS describes visible facial movement rather than assigning an emotion to it.
• Why it matters: The human slow-blink condition increased some components, but not eye closures significantly. Describing every movement as an increased “loving blink” would erase the actual pattern.

2. Split the experiment from the correlation

• What to look for: Results 3.1 compares each cat’s two conditions; Results 3.2 relates eye closures during slow-blink trials to reservation timing across cats.
• What it means: A within-cat condition comparison asks about responses to different interactions. A correlation asks whether two measurements vary together; it does not manipulate adoption decisions.
• Why it matters: Eye closures can correlate with shorter waits even though their average count did not significantly increase under slow blinking. Those findings are not contradictory, because the analyses ask different questions.

3. Read the endpoint, not just the headline

• What to look for: Methods 2.3 defines days until reservation and assigns 133 days to four cats beyond the observation cutoff.
• What it means: Reservation is a useful practical indicator, but not completed adoption. A cutoff tells us a wait is longer than the observed period, not when it eventually ends.
• Why it matters: This limits precision and causal interpretation. It also shows why recording actual reservation dates and handling unresolved outcomes explicitly would strengthen a follow-up.

4. Keep the anxiety explanation provisional

• What to look for: The pre-existing desensitisation groups, the p = 0.07 total-duration comparison, and the authors’ discussion of both friendly and submissive contexts.
• What it means: The observed tendency is a hypothesis-generating signal, not a confirmed group difference. Programme membership includes extra exposure to people as well as an initial anxiety assessment.
• Why it matters: A movement may have different social functions in different contexts. This paper does not license a universal emotion dictionary, nor does it show that a cat must be relaxed whenever its eyes narrow.

Two reflection questions — Lumi

1. What experiment could test whether eye closures themselves influence adopters, rather than another cat characteristic influencing both eye closure and selection?
2. Why can eye closure correlate with faster reservation even when its frequency does not differ significantly between slow-blink and control trials?