Simulated Presence Therapy: What the Research Shows

In 1999, a research team led by Camberg and colleagues published a paper that would quietly change how dementia care professionals thought about managing anxiety in people with Alzheimer's disease. Their intervention was simple, almost surprising in its simplicity: they recorded family members speaking warmly, sharing memories, offering reassurance, and played those recordings to patients who were distressed. The results were positive. The familiar voice worked.

That was the beginning of simulated presence therapy. More than two decades of clinical research followed. KindredMind is, in one sense, where that research has arrived.

What Simulated Presence Therapy Is

Simulated presence therapy (SPT) is a non-pharmacological dementia care approach designed to reduce behavioral and psychological symptoms of dementia (BPSD), the anxiety, agitation, repetitive behavior, and distress that accompany cognitive decline. It works by providing a person with dementia with access to a familiar, emotionally significant voice, typically a family member's, in a way that feels like genuine contact.

The underlying clinical rationale is rooted in what dementia does and does not damage. In Alzheimer's disease and most other dementias, emotional memory is preserved longer than episodic memory. A person who cannot remember that their daughter called five minutes ago will still feel the emotional warmth and relief of hearing her voice. The emotional register of an interaction reaches the person even when the factual content does not.

SPT works because it speaks to the part of the brain that dementia has not yet touched.

Origins: Camberg et al., 1999

Camberg and colleagues developed SPT as a practical intervention for nursing home residents with Alzheimer's disease who were experiencing agitation and distress. Their approach involved recording personalized audio messages from family members, not generic soothing phrases, but specific references to shared memories, familiar expressions, and names that mattered to the individual patient.

The reasoning was clear: a generic voice offering generic reassurance does not reach a person with dementia in the same way that a specific, familiar voice does. Personalization was not a detail, it was the mechanism. The research found positive effects on agitation and distress, and the approach began to be adopted in memory care settings.

Research Limitations and the Next Step

As SPT became more widely studied, researchers also noted its inherent limitation: a static recording cannot respond, adapt, or pick up on what the person actually says in the moment. It is a monologue, not a conversation. The clinical benefit is real, but it is constrained by the medium.

The 2024 RCT (PubMed 38646703)

More recent research has continued to support the role of familiar voice interventions in dementia anxiety. A 2024 randomized controlled trial (PubMed 38646703) examined the effects of familiar voice exposure on anxiety and agitation in people with dementia, finding measurable reductions in distress. The study adds to a growing body of evidence that the neurological mechanism identified by Camberg and colleagues, emotional memory preservation and its responsiveness to familiar voice, is robust and clinically meaningful.

From Static Recording to Responsive Conversation: Where KindredMind Fits

The limitation of original SPT was always the static nature of the recording. A cassette tape plays the same message regardless of what the listener says. It cannot respond to "where are you?" with warmth. It cannot pick up on a shift in tone and adapt. It cannot remember that the person loves gardenias, and ask how the garden is looking.

KindredMind represents the natural evolution of the SPT framework: from static recording to responsive, personalized conversation. When a person with dementia calls, they are not met with a pre-recorded message. They are met with a real-time, adaptive conversation in the caregiver's voice, a voice that knows their name, their family members, their daily routines, their favorite memories, and their emotional triggers. KindredMind supports 15 companion languages, including Spanish, French, Portuguese, Italian, Polish, Russian, Mandarin, German, Korean, and Hindi (with Tagalog, Ukrainian, and Greek in beta), so the familiar voice speaks in the language the person has always felt most at home in.

This distinction matters clinically. The therapeutic mechanism of SPT is not the recording, it is the familiar voice and the emotional comfort it provides. KindredMind preserves that mechanism and removes the limitations imposed by the static medium.

The Alzheimer Society of Canada's Framework

The Alzheimer Society of Canada publishes specific guidance on dementia-friendly communication, principles for how family members and professionals should approach conversations with people with dementia. The framework emphasises warmth, patience, short and clear sentences, and meeting the person in their emotional reality rather than correcting factual errors.

This is, in essence, the operating principle of simulated presence therapy, and it is the operating principle of KindredMind. The AI is not optimised for efficiency or brevity. It is trained to respond with warmth, to listen before it redirects, to use specific language and references that are meaningful to this particular person. It does what the Alzheimer Society of Canada says a caring family member should do, which is exactly what it is: a system through which a caring family member is present.

On the Deception Question

The most common concern about familiar voice systems in dementia care is whether using them is deceptive, whether it is wrong for a person with dementia to believe they are talking to their daughter when they are, in some sense, talking to an AI.

This is a serious question, and it deserves a serious answer. The clinical consensus in dementia care, from the Alzheimer Society of Canada, from the Alzheimer's Association, from every professional memory care training framework, is that literal reality is not always the most compassionate truth. Simulated presence therapy does not claim that a recording is a person. It claims that a familiar voice provides comfort, and that comfort is real and valid and worth providing. The same is true of KindredMind. For a fuller treatment of this question, see our approach page on the ethics of AI in dementia phone care.