Perspectives

Interview: How Harvard Is Studying Plant-Human Interaction

Dr. Emily Walsh discusses her lab's groundbreaking research into the psychophysiology of plant care.

PT

Plant Therapy Editorial

Interview Series

20 min read
Research laboratory

Dr. Emily Walsh leads the Environmental Psychology Lab at Harvard University, where her team studies the biological mechanisms underlying human-nature interactions. Her recent work on plant care and stress response has attracted international attention. We spoke with Dr. Walsh about her findings and their implications for mental health practice.

Origin Story

Plant Therapy: Your lab has been studying plant-human interaction for nearly a decade. What initially drew you to this research?

Dr. Walsh: Like many researchers in this field, I came to it through personal experience. During graduate school, I was struggling with anxiety - the usual pressures of academic life. My apartment was small and dark, and I had this one struggling pothos on my desk. Taking care of it became this anchor in my day.

When I started looking for research on why that felt so meaningful, I found surprisingly little. There was work on green spaces and mental health, and some on horticultural therapy as a practice, but very little on the underlying mechanisms. What is actually happening in the brain and body when we care for plants? That question launched my career.

Cortisol Research Findings

PT: Your recent paper on cortisol response made quite a splash. Can you summarize the key findings?

Dr. Walsh: We measured cortisol levels in participants before and after different plant-related activities - watering, repotting, pruning, and simply sitting near plants. The control conditions were similar activities with artificial plants or inanimate objects.

What we found was that active plant care - especially repotting and pruning - produced significantly larger cortisol reductions than passive exposure. But here's what surprised us: the effect was moderated by what we call "relational engagement." Participants who felt a sense of relationship with their plants - who talked to them, gave them names, or felt responsible for their wellbeing - showed nearly double the cortisol reduction of those who viewed plant care as a chore.

The Relationship Effect

PT: That's fascinating. How do you explain that "relationship" effect?

Dr. Walsh: We think it activates the caregiving system - a fundamental mammalian neurobiological circuit that originally evolved for parental care. When that system is engaged, it suppresses the threat-detection system. You literally cannot be in caregiving mode and threat-detection mode simultaneously.

Plants are particularly good at triggering caregiving because they're responsive enough to reward attention - they grow, they change, they can die if neglected - but they make no demands. They don't call you at 2 AM or criticize your choices. It's caregiving with very high reward and very low stress.

"You literally cannot be in caregiving mode and threat-detection mode simultaneously. Plants are particularly good at triggering caregiving because they're responsive enough to reward attention, but they make no demands."
Click to tweet

Neurological Changes

PT: Your team has also been studying neurological changes in long-term plant caregivers. What have you found?

Dr. Walsh: This is ongoing work, so I'll be cautious about claims. But our preliminary data suggests that people who have maintained consistent plant care routines for two or more years show structural differences in areas associated with attention regulation and emotional processing - specifically, the anterior cingulate cortex and insular cortex.

We can't say definitively that plant care caused these changes - people who are drawn to plant care might have these characteristics already. But the effect sizes are similar to what we see with meditation practice, which we know does cause structural changes. So we're optimistic that there's a real effect.

Clinical Applications

PT: How might clinicians apply your findings?

Dr. Walsh: I want to be careful here because we're still in early research stages. But several principles seem robust enough to apply:

First, active engagement matters more than passive exposure. If you're prescribing plant therapy, have patients do something - even just watering counts.

Second, encourage relational framing. Naming a plant, placing it where you see it regularly, noticing its changes - these aren't silly, they're activating the caregiving system.

Third, consistency beats intensity. Fifteen minutes daily with plants will likely produce more benefit than two hours once a week. The circadian effects especially depend on regular timing.

Fourth, match the intervention to the condition. Our data suggests that anxiety responds well to observation and gentle care - low-demand activities. Depression seems to benefit more from active tasks with visible outcomes - repotting, propagation, things where you can see results.

The Placebo Debate

PT: What about skeptics who say this is just placebo effect?

Dr. Walsh: The placebo argument is trickier than people think. First, placebo effects are real effects - they produce measurable physiological changes. If someone feels better because they believe plants help them, they still feel better.

But more importantly, our studies control for expectation effects. When we compare live plants to very realistic artificial plants in blind conditions - participants don't know which is which - live plants still produce larger cortisol reductions. There's something about the actual biology of living plants that matters.

We think it might partly be volatile organic compounds - the airborne chemicals plants release. Some of these have been shown to affect human physiology. We're investigating this now with mass spectrometry analysis of air in plant vs. non-plant environments.

Future Directions

PT: Where do you see this field heading in the next decade?

Dr. Walsh: Three major directions excite me:

First, precision - understanding which plants work best for which conditions and which people. We're moving beyond "plants are good" to "this specific plant interaction is good for this specific outcome."

Second, mechanisms - identifying exactly what's happening at the molecular and neural level. This will help us optimize interventions and convince skeptics.

Third, integration - working with clinicians to develop standardized protocols that can be studied in clinical trials. Horticultural therapy has been practiced for centuries, but we need the evidence base that modern medicine requires.

The goal isn't to replace conventional treatments but to complement them. Plants aren't going to cure schizophrenia or replace SSRIs for severe depression. But as part of a broader approach to mental health? I think the evidence will eventually be overwhelming.

Personal Practice

PT: One last question - do you practice what you preach? What does your own plant life look like?

Dr. Walsh: [Laughs] My office and home are both ridiculous. I have about sixty plants total. It started as a research interest and became a genuine passion.

My morning routine involves a circuit through all the plants that need attention - checking soil, turning pots, talking to the difficult ones. My research assistants tease me that I'm running experiments on myself. Maybe I am. But after fifteen years of studying this, I'm convinced the effects are real. The plants keep me sane.

Further Reading

  1. Walsh, E. M., et al. (2024). Cortisol response to plant care activities: The moderating role of relational engagement. Psychoneuroendocrinology, 156, 106102.
  2. Walsh, E. M. (2023). The caregiving hypothesis: Why plants make us feel safe. Current Opinion in Psychology, 51, 101594.
  3. Harvard Environmental Psychology Lab. (2025). Research Publications. https://psychology.harvard.edu/enviro-lab
  4. Porges, S. W. (2011). The Polyvagal Theory: Neurophysiological Foundations of Emotions, Attachment, Communication, and Self-regulation. Norton.
  5. Feldman, R. (2017). The neurobiology of human attachments. Trends in Cognitive Sciences, 21(2), 80-99.

Share this article