The Rabbit Effect

Briefing for the Psychology board — compiled 06/10/2026

1. Two things share the name

1. The original experiment — Nerem, Levesque & Cornhill, Science 208(4451):1475–1476, 27 June 1980. DOI: 10.1126/science.7384790.
2. The popular book — Kelli Harding, The Rabbit Effect: Live Longer, Happier, and Healthier with the Groundbreaking Science of Kindness (2019, ~272 pp, ISBN 9781501184260). Harding is a Columbia University physician, described in reviews as a psychiatrist who worked in the psychiatric emergency room at New York-Presbyterian.

The book opens with the experiment and uses it as the hook for a much wider argument about social ties and health.

2. The experiment, as published

Aim. The authors say earlier animal work had looked at negative stressors and that, to their knowledge, no study had tested a positive social factor.

Animals. Young male New Zealand White rabbits, fed a regular rabbit diet supplemented with 2% cholesterol.

Intervention. A one-to-one relationship between each animal and a single experimenter (M.J. Levesque):

• an early-morning half-hour visit: handled, stroked, talked to, played with
• an hour-long feeding period, also with touching and talking
• several 5-minute visits during the day
• left alone for about 10 hours overnight
• a 2-week adaptation period beforehand so the animals got to know her

Control animals received normal laboratory animal care. All experiments used the same experimenter and the same protocol.

Three studies, not one.

Group	Role	Timing
A	Handled, no concurrent control (earlier control studies existed)	late 1977
B vs C	Handled vs control	early 1978
D vs E	Handled vs control	late 1978

The authors say group A's result was first considered anomalous and is what led them to run the two further studies. Groups A–C were killed after 5 weeks, D and E after 6.

Outcome measures. Percentage of aortic surface stained by Sudan IV (sudanophilic lesions), scored from photographs by independent investigators; serum cholesterol (weekly, fasted); blood pressure; heart rate.

Results.

• Handled groups showed more than 60% less aortic surface covered by lesions than controls.
• B vs C was significant at P = .015. D vs E was also reported significant; the scan I read is garbled there (it appears to say about P = .034), so check the original.
• Serum cholesterol was extremely high and not different between groups (roughly 1,400–1,900 mg/dl).
• Blood pressure and heart rate were also not significantly different.

So the effect did not run through the usual measured risk factors.

What the authors concluded. The social environment had a marked effect on diet-induced atherosclerosis in this model. They said the mechanism was unknown and suggested looking at blood hormone levels, arterial wall permeability, endothelial regeneration and other factors. They acknowledged social interventions are less quantifiable than physical or biochemical ones. Their sharpest methodological point: if social environment alters outcomes this much, it could explain why different labs get contradictory results from nominally identical experiments, so animal protocols should take socio-psychological factors into account.

3. How the story is told in the popular version

Popular accounts and the book describe a lab technician or postdoc who happened to be unusually kind, who talked to and petted the rabbits, and an unexpected group with about 60% fewer deposits. Nerem is quoted in the book as saying they were not social behavioural scientists but could not ignore the finding.

Points where popular tellings and the paper differ:

• Many tellings say 1978; the paper shows the first (anomalous) group was late 1977, and publication was 1980.
• The paper reports three studies with a total of five groups, not a single accidental observation. Only the first was a surprise; the other two were deliberate follow-ups.
• Some retellings call the original "controlled" and say "same hutches". I did not find anything in the paper's text about housing or cage position, so any claim that the handled rabbits were all on a particular shelf should be treated as unverified until the full paper or the book's endnote is checked.

4. Strengths and limits

Strengths

• Deliberate replication of an anomalous first result, with concurrent controls, in two further cohorts.
• Same direction and similar size across cohorts, in an objective-ish outcome (stained area of the aorta).
• Cholesterol, blood pressure and heart rate were comparable, so the obvious explanations were measured and did not account for it.

Limits

• One experimenter did all the handling. Handling and the individual person are inseparable here. A different handler could differ.
• Same lab, same team. No independent replication of this exact design that I found.
• Small groups and a short paper (two journal pages).
• Blinding is not described in the text I read ("independent investigators" scored the stain; whether they knew group membership is not stated).
• Mechanism unknown; the authors themselves list it as open.
• Model. Rabbits fed a cholesterol-heavy diet are a standard but artificial atherosclerosis model.
• Group A had no concurrent control.

5. Related later work (converging, not identical)

None of these repeat the petting design. They support the broader idea that social environment affects atherosclerosis.

Study	Species	Finding
Kaplan et al., Science 220:733–735 (1983)	Cynomolgus monkeys	Socially stressed males on a low-fat, low-cholesterol diet developed more coronary atherosclerosis than controls, with similar serum lipids and blood pressure
Nation et al., Circulation (2002), DOI 10.1161/hc0302.102144	Watanabe heritable hyperlipidemic (WHHL) rabbits	Stable social pairings slowed atherosclerosis progression relative to unstable pairings or individual caging
Stress (2012), DOI 10.3109/10253890.2012.676696	Rabbits	Chronic unpredictable stress increased vascular inflammation and oxidative stress markers, without significant effects on serum lipids

Note the direction: later work leans on stress and unstable social conditions causing harm, which is a different claim from "kindness protects".

6. The book's wider argument

Harding uses the rabbits as a metaphor and then moves to human evidence on social ties, community and environment. Reviews and chapter summaries mention:

• Roseto, Pennsylvania — a close-knit town with lower heart disease than expected (the "Roseto effect").
• Okinawa — longevity linked to community and social ties.
• Chapter 3, "Social Ties: Your Community".
• Loneliness and isolation as health risks; kindness, connection and purpose as protective.

I did not read the whole book, only the opening excerpt and descriptions of it, so treat the list above as a guide to themes, not a complete summary.

7. Caveats for discussion

• The human evidence is mostly observational. Healthier people may simply have more social contact (reverse causation), and social ties correlate with income, education and neighbourhood.
• The rabbit result is good support for "social environment can change disease outcomes in an animal model". It does not show that kindness specifically is the active ingredient, since touch, attention, handler identity and stress reduction are all bundled together.
• Popular science books favour narrative over methodological detail, and "the rabbit study" in retelling is tidier than the paper.

8. Discussion prompts

1. Is a single handler a fatal confound, or does the replication across cohorts handle it?
2. What would an ideal replication look like (multiple handlers, blinded scoring, randomised assignment, independent lab)?
3. If cholesterol, blood pressure and heart rate did not differ, what plausible pathway is left (stress hormones, inflammation, endothelial function)?
4. How far can an animal model of atherosclerosis stretch toward claims about human kindness and community?

9. Sources

• Nerem RM, Levesque MJ, Cornhill JF. Social environment as a factor in diet-induced atherosclerosis. Science 1980;208:1475–1476. (Read via a hosted copy of the paper.)
• Kaplan JR et al. Social stress and atherosclerosis in normocholesterolemic monkeys. Science 1983;220:733–735.
• Nation DA et al. Social environment influences the progression of atherosclerosis in the Watanabe heritable hyperlipidemic rabbit. Circulation 2002.
• Chronic psychological stress induces vascular inflammation in rabbits. Stress 2012.
• Harding K. The Rabbit Effect (2019); publisher excerpt, retailer descriptions and review articles (Psychiatric Times; Kathy Pike PhD).