Background
Trust is a cornerstone of social and economic cooperation, yet its neurobiological basis had been poorly characterized. Animal research had implicated oxytocin in pair bonding and maternal behavior, but direct evidence that oxytocin modulates trust decisions in humans was lacking prior to this study.
Kosfeld and colleagues at the University of Zurich designed an economic trust paradigm — the classic investor-trustee game — where “investors” could transfer money to anonymous “trustees” knowing the trustee could return any amount. This operationalized trust as a quantifiable, incentivized behavioral measure rather than a self-report outcome.
Methods
Double-blind, placebo-controlled, between-subjects design. 58 healthy adult male participants randomized to:
- Intranasal oxytocin: 24 IU (3 puffs per nostril of 4 IU/puff) administered 45 min before task
- Placebo: intranasal vehicle
Trust game: Investors chose to transfer 0–12 monetary units (MU) to a trustee. Transferred amounts were tripled; trustees then decided how much to return. Higher transfers reflect greater investor trust.
Control condition (separate cohort): Transfers to a computer (random return) rather than a human trustee, to isolate trust from risk-taking.
Key Findings
| Condition | Oxytocin Group | Placebo Group | p-value |
|---|---|---|---|
| Mean MU transferred (human trustee) | 17 | 13 | 0.019 |
| % transferring max (12 MU) | 45% | 21% | < 0.05 |
| MU transferred (computer/random) | No difference | No difference | ns |
- Oxytocin selectively increased trust toward human trustees — not toward a computer performing random returns
- This rules out a general risk-taking or reward-processing explanation; the effect is specific to social trust
- Participants in the oxytocin group were more likely to exhibit full trust (maximum transfer) even knowing it created maximum vulnerability
Clinical Significance
This landmark paper provided the first direct causal evidence that a neuropeptide — administered intranasally in humans — could substantially alter a core prosocial behavior. Key implications:
- Oxytocin as a prosocial neuromodulator: The study operationally validated the “social engagement” hypothesis for oxytocin beyond animal models
- Therapeutic potential: Social deficits in autism spectrum disorder, social anxiety, PTSD, and personality disorders could theoretically be addressed via oxytocin modulation
- On-demand dosing window: Effects emerged within 45 minutes of administration — consistent with intranasal kinetics and CNS delivery
The study generated enormous interest and spawned hundreds of replication and extension studies across clinical and healthy populations.
Limitations
- Male participants only; oxytocin’s effects on trust in women are more complex and context-dependent (estrogen interactions)
- Intranasal delivery to CNS is pharmacologically uncertain — extent of direct brain penetration vs. peripheral effects remains debated
- Effect was observed in a specific economic paradigm; generalization to naturalistic trust behavior is assumed but not proven
- Some replication attempts have produced smaller or null effects, underscoring context-sensitivity of oxytocin’s prosocial actions