Doll therapy — offering a lifelike baby doll to a person with dementia to hold, cradle, and care for — works in one-on-one care when the caregiver watches for specific signs and lets the person's response decide each session. Engagement looks like reaching for the doll, cradling it, talking to it in a warm tone, and settling; distress looks like agitation, refusal, anxious possessiveness, or upset when the doll is nearby, and it means the doll comes away. The approach has real evidence behind it for agitation and behavioral symptoms in moderate-to-severe dementia, but the benefit is not guaranteed for any individual. That makes observation the core skill: the doll is an offer, and what the person does with it in the next few minutes is the answer.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Table of Contents
- What the evidence says doll therapy can do
- How to introduce the doll without forcing it
- Signs the session is working
- Signs of distress, and what to do about them
- What to document after each session
- Talking with family before you start
- Frequently Asked Questions
What the evidence says doll therapy can do
A 2023 systematic review and meta-analysis in Geriatric Nursing pooled nine studies of nursing home residents with dementia (845 people) and found a significant reduction in agitation — the only outcome the reviewers rated high-quality — with no clear cognitive improvement. Doll therapy calms behavior; it does not restore memory or thinking. A 2022 randomized single-blind trial of 134 women with moderate-to-severe dementia, published in the Journal of Clinical Medicine, compared a doll against a sham object (a cube) over 30 sessions. The doll group showed significant drops in Neuropsychiatric Inventory total and distress scores.
A smaller Turkish randomized study in the Journal of Geriatric Psychiatry and Neurology found eight weeks of doll therapy reduced agitation measured with the Cohen-Mansfield Agitation Inventory. The limits matter. In the 2022 trial, stress biomarkers did not differ between groups, and a pilot cluster trial reported in Frontiers in Public Health found no significant reduction in anxiety, agitation, or aggression versus usual care. Benefits also appeared across repeated sessions, not from a single introduction. Treat the doll as an option to offer and re-offer, judged each time by what you observe.
How to introduce the doll without forcing it
Practice guidance from dementia-care sources such as DementiaHub.SG is consistent: introduce the doll non-intrusively. Place it within the person's reach — on a nearby chair or table — rather than putting it into their hands. Let them notice it, pick it up, or ignore it. Avoid framing that decides the doll's meaning for them.
Some people relate to it as a baby, others as a doll or a companion; follow their lead in how you refer to it. Never insist, never use it to distract from a request they are making, and never present it as a test. If the person ignores the doll, that is a neutral result, not a failure. Remove it quietly and try again another day, perhaps at a different time or in a calmer setting. The trial benefits accumulated over many sessions, so one flat response tells you little.
Signs the session is working
In the 2022 trial, engagement showed up as greater interest in the doll than in the control object, easier acceptance of the nurse stepping away, and spontaneous caregiving and exploratory behaviors — rocking, patting, examining the doll's face and clothes. Those are the behaviors to watch for in your own sessions.
Concrete positive signs, drawn from the trial and from practice guidance: Engagement can be quiet. A person who simply holds the doll and watches it calmly is engaged; you do not need conversation or visible delight to count the session a success.
- Reaching for or accepting the doll without prompting
- Cradling, rocking, patting, or adjusting its clothing
- Talking or singing to it in a warm, loving tone
- Smiling, softened facial expression, relaxed posture
- Staying settled when you step back or leave briefly
Signs of distress, and what to do about them
Distress signs mean the doll is not comforting this person right now, and guidance from DementiaHub.SG and Home Instead is direct about the response: withdraw it. Watch for: Remove the doll calmly and without scolding, and redirect to another activity. Possessiveness is the subtle one: it can look like deep engagement, but if the doll starts displacing eating, sleep, or personal care, it has become a source of strain.
Shorten sessions and build in a natural handover — "shall I hold the baby while you have lunch?" — rather than taking the doll abruptly. A distressing session does not permanently rule the doll out, but repeated distress across several attempts does. The evidence supports doll therapy for people who respond to it; it says nothing in favor of persisting against a clear negative reaction.
- Agitation, anger, or anxiety that starts or worsens with the doll present
- Pushing the doll away, throwing it, or repeated refusal
- Distressed conviction that a real baby is in danger, hungry, or lost
- Anxious possessiveness — inability to put the doll down, panic if it is moved, refusing meals or care to guard it
- Tearfulness or grief reactions the person cannot settle from
What to document after each session
A 2024 scoping review in Nursing Reports distilled what staff should record: physical and verbal interaction with the doll, caregiving behavior, duration of use, emotional response, refusal, distress, and possessiveness. That list doubles as the observation checklist during the session itself. DementiaHub.SG adds structure for the care plan: note each session's goal, time, and duration, the person's responses when the doll is introduced, played with, and withdrawn, and their mood afterward.
Mood afterward is easy to skip and highly informative — a person who is calmer for an hour after the doll goes away is benefiting even if the session itself looked unremarkable. Keep records comparable between sessions and hand them over between caregivers or shifts. Research protocols vary widely — the 2023 meta-analysis found session lengths from 15 minutes to 3 hours across studies — so there is no standard dose to copy. Your own notes are how you find this person's right length and frequency, and how you spot fatigue with the doll over weeks.
Talking with family before you start
Mitchell and Templeton's ethics analysis in Nursing Ethics concluded doll therapy supports the person's wellbeing and autonomy but raises real dignity concerns — including distress in family members who find it infantilizing to see a parent cradle a doll. Their advice: train caregivers and talk with families before implementation, with the person with dementia treated as the rights holder at the center of the decision.
In practice, explain to relatives what the evidence shows about agitation, what you will watch for, and that the doll stops the moment it causes distress. Families who see the observation-and-documentation approach — rather than a doll simply left in a lap — usually understand it as care, not condescension. If the person visibly benefits, that observed comfort is the deciding fact, and it belongs to them.
Frequently Asked Questions
How long should a doll therapy session last?
There is no evidence-based standard — the 2023 Geriatric Nursing meta-analysis found study sessions ranging from 15 minutes to 3 hours. Start short, watch the person's response, and let your session notes set the length.
Does doll therapy improve memory or cognition?
No. The 2023 meta-analysis found a high-quality-rated reduction in agitation but no clear cognitive improvement. It is a comfort and behavior intervention, not a cognitive one.
What if the person thinks the doll is a real baby?
Follow their lead rather than correcting or confirming. It only becomes a problem if the belief causes distress — fear the baby is in danger, or refusing food or care to guard it — in which case shorten sessions and use gentle handovers, or withdraw the doll.
