Healthopia Institute

Evidence

What the evidence supports, and what it does not.

We publish what the evidence shows, and what it does not yet show, in the same place.

An open medical journal, reading glasses and a notebook on a desk by a window

What is well supported

Modifiable risk factors are associated with dementia risk at population level

Successive Lancet Commission reports identify a set of factors — vascular risk, hearing loss, physical inactivity, smoking, depression, social isolation, diabetes, excess alcohol, air pollution, traumatic brain injury, and others — associated with a substantial population attributable fraction of dementia cases. [1]

Multidomain lifestyle intervention has shown modest cognitive benefit in trials

FINGER demonstrated a small benefit on a composite neuropsychological battery over two years in at-risk older adults; multi-country and U.S. successors have examined delivery in community settings. [2][3]

Most of these domains warrant clinical attention regardless of cognition

Blood pressure control, sleep-disordered breathing, depression, hearing loss, and physical inactivity each carry independent clinical importance. Addressing them is defensible on its own merits.

What is genuinely uncertain

  • Effect sizes in the interventional literature are modest, and comparison arms were not inert.
  • The relationship between a change in composite test score and a change in dementia incidence years later is not established.
  • The optimal dose — intensity, duration, and combination of components — is unknown.
  • Whether trial-grade delivery transfers to routine independent practice is untested.
  • Composite scores have limited individual predictive validation; they describe a domain profile, not a prognosis.
  • Unsupervised digital cognitive testing differs materially from supervised neuropsychological testing in conditions, effort, and interpretability.

Our validation roadmap

Phase 1

Instrument characterisation

In progress

Test–retest reliability, ceiling and floor behavior, and device effects for the unsupervised cognitive tasks.

Phase 2

Concurrent validity

In progress

Comparison of the digital battery and composite against supervised neuropsychological assessment in a sub-sample.

Phase 3

Implementation fidelity

Planned

How faithfully the protocol is delivered in independent practices, override rates, and where the protocol fails clinicians.

Phase 4

Behavioral and intermediate outcomes

Planned

Change in domain-level behavior and vascular risk measures over 12 months, prespecified and published irrespective of result.

Phase 5

Long-term cognitive outcomes

Not started

Longitudinal follow-up designed with the Scientific Advisory Board. Until this phase reports, no outcome claims are permitted.

Publications

Evidence dossiers

Protocols

Position statements

Peer-reviewed literature

References

  1. [1] Livingston G, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. The Lancet. 2024.
  2. [2] Ngandu T, et al. A 2 year multidomain intervention (FINGER): a randomized controlled trial. The Lancet. 2015;385:2255–63.
  3. [3] Baker LD, et al. U.S. POINTER: a multidomain lifestyle intervention in older adults at risk. 2025.
Institute policy

No Institute material, and no material published by a certified clinician, may state or imply that the program prevents, treats, reverses, or reduces the risk of dementia or cognitive decline, or state a percentage risk reduction. Claims are reviewed by Ethics and Compliance.