What Word Games Help Caregivers Support Memory Loss Patients?

The crossword evidence, the twice-weekly dose that worked, and the UC San Diego technique that stops a puzzle feeling like a memory test.

Crossword puzzles are the best-supported word game for caregivers to use with someone who has memory loss, with word searches, anagrams, word-association games and category-naming games as easier alternatives when crosswords are too hard. The evidence behind them is real but modest: word games are part of cognitive stimulation programmes that produce small cognitive gains and clearer benefits in communication, mood and the caregiver relationship. Cognitive stimulation means structured activities — word games, puzzles, music and discussion of past and present events — designed to engage thinking and conversation rather than to teach or test. The 2023 Cochrane review of cognitive stimulation for dementia pooled 37 randomized trials with 2,766 participants with mild or moderate dementia, average age 79, and lists word games among the core activities of these programmes.

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 actually shows about word games

Across 25 studies and 1,893 participants, Cochrane found a 1.99-point advantage on the Mini-Mental State Examination (a 30-point cognitive screening test) for cognitive stimulation over control conditions. The reviewers rated this moderate-certainty evidence and called the difference clinically important, and they also found improvements in communication and social interaction plus slight gains in participants' own mood ratings. Crosswords specifically have a dedicated trial. The COG-IT study randomized 107 people with mild cognitive impairment, average age 71, to web-based crossword puzzles or computerized cognitive games for 12 weeks with booster sessions out to 78 weeks.

As reported in NEJM Evidence in October 2022, crosswords beat the games on the ADAS-Cog cognitive scale at 78 weeks (difference −1.44, 95% CI −2.83 to −0.06, p=0.04), on daily function, and on MRI measures of brain shrinkage. The size of that win matters. The effect sizes were modest — Cohen's d of 0.43 at 12 weeks and 0.34 at 78 weeks — and crosswords outperformed computerized games mainly in participants at a later disease stage, with the two about equal earlier on. Crosswords are a sensible default, not a large effect.

How often, and when in the illness

Cochrane found the benefit was larger when sessions ran twice weekly or more rather than once weekly, and larger where the person's dementia was mild at the start. Frequency and timing matter more than which puzzle you pick. That points to a practical priority order for caregivers.

Start earlier rather than waiting for a decline that seems to justify it, aim for at least two short sessions a week, and choose the game the person will actually sit down for instead of the one that looks most therapeutic. Short and repeated beats long and occasional. A ten-minute crossword twice a week is closer to the tested pattern than an hour-long session every other Sunday.

What caregiver-delivered word games realistically deliver

The iCST trial tested exactly this arrangement: 356 carer–patient pairs at eight UK sites assigned to 75 thirty-minute family-delivered cognitive stimulation sessions over 25 weeks. Published in PLoS Medicine in 2017, it found people with dementia rated their relationship with the carer more favourably (MD 1.77, 95% CI 0.26–3.28, p=0.02), and carers' own quality of life on the EQ-5D improved (MD 0.06, p=0.01). It found no difference in cognition, and no difference in patients' self-reported quality of life.

Adherence was poor: 22% of iCST pairs (39 of 180) completed no sessions at all, and only 40% managed two sessions per week. Read together, those two results reset expectations. The dependable payoff from word games you run at home is a better relationship and a better week for you — not a change in memory test scores. That is still worth half an hour, and it is a different reason to do it than the one most caregivers are given.

Running a crossword with someone who has memory loss

UC San Diego's Alzheimer's Disease Research Center, in its caregiver activities guide, sets out an approach that removes the test-taking feel: The Alzheimer's Association adds the troubleshooting rule. In its guidance on daily activities, it notes that activities the person loves can be modified to their current ability, and that when something isn't working the problem is often the wrong time of day or too much complexity — try again later or simplify, rather than dropping the activity for good.

  • Start the crossword yourself, then ask the person for assistance — you are the one doing the puzzle, and they are helping.
  • Expect that the puzzle may simply be too difficult for some people, and have an easier one ready.
  • Accept that the person may not remember the rules or whose turn it is; supply both without comment.
  • Use the puzzle's theme to start a conversation, which is often where the real value of the session sits.

Limits and risks worth knowing

Cochrane reported high inconsistency between studies on cognition and quality of life, and an evidence gap on longer-term programmes. It also reported no negative effects.

That combination defines what word games are: low-risk, worth doing, and not a treatment that halts decline. If a session consistently produces frustration or distress, the activity is the wrong fit for that person on that day — simplify it, move it to a better hour, or switch to a word-association or category-naming game with no right answer to miss.

Frequently Asked Questions

Are word searches as good as crosswords?

No trial has compared them directly. Crosswords are the game with trial evidence behind them, but Cochrane's finding that session frequency and early timing matter more than the specific puzzle suggests a word search the person will do twice a week beats a crossword they refuse.

Will word games improve my relative's memory scores?

Probably not when delivered at home. The iCST trial of 356 carer–patient pairs found no cognition difference, while the relationship rating and carer quality of life both improved.

Should we use an app or paper?

COG-IT's crossword arm was web-based, so screens are not a problem in themselves. Use whichever format the person handles without extra help, since the barrier in trials was doing the sessions at all, not the medium.


You Might Also Like

HelpDementia.com

Dementia, Alzheimer's, Caregiving & Healthy Aging Guidance

© 2026 HelpDementia.com. All rights reserved.

Educational information only. It is not medical advice and does not replace care from a qualified clinician.