- Design
- Population-based cohort study with Cox proportional hazards modelling and mutually adjusted component analysis
- Population
- 52,789 adults aged 65+ attending municipal health checkups in Okayama City, Japan, 2006-2007; 49,665 in the fully adjusted model; mortality follow-up to December 2016
- Primary outcome
- All-cause mortality by Kihon Checklist oral-function score (0-3)
- Effect
- aHR 1.09 (95% CI 1.04-1.13) for score 1, 1.26 (1.20-1.32) for 2, 1.49 (1.38-1.60) for 3. Chewing difficulty aHR 1.19 (1.14-1.23), dysphagia 1.18 (1.14-1.23), xerostomia 1.02 (0.98-1.06)
Oral function is assessed, when it is assessed at all, by someone looking in the mouth. This cohort asked whether three yes-or-no questions do the same work.
The Kihon Checklist oral-function score sums positive answers to three self-reported items: chewing difficulty, dysphagia and xerostomia, giving a score from 0 to 3. The study followed 52,789 adults aged 65 and over who attended municipal health checkups in Okayama City, Japan in 2006-2007, with mortality follow-up to December 2016, using Cox proportional hazards models.
In the fully adjusted model (n=49,665), adjusted hazard ratios for all-cause mortality were 1.09 (95% CI 1.04-1.13) for a score of 1, 1.26 (1.20-1.32) for 2 and 1.49 (1.38-1.60) for 3. The gradient is clean and monotonic, which is what makes a score usable.
The component analysis is what changes practice. Chewing difficulty (aHR 1.19, 95% CI 1.14-1.23) and dysphagia (aHR 1.18, 1.14-1.23) were each independently associated with mortality. Xerostomia was not (aHR 1.02, 95% CI 0.98-1.06). So the useful screen is two questions, not three — and dry mouth, which is the symptom most often volunteered and most often drug-related, is the one carrying no independent signal.
Two questions asked in an ordinary review, with no examination and no test, stratify ten-year mortality. That is a cheap addition to a consultation, and in Indian primary care, where dentition and nutrition in older adults are systematically under-assessed, it is cheaper than anything that would replace it.
- Ask about chewing difficulty and about choking — drop xerostomia from the screen
- A positive answer is a prompt to assess weight, dentition, diet and medication
- The gradient from score 1 to 3 is monotonic, which makes it usable as a stratifier
- Self-reported, so this captures what the patient notices, not what an examination would find
- Association, not causation — no trial has shown that treating these symptoms lowers mortality
Why it matters
It reduces an oral health assessment to two questions that any clinician can ask in any consultation.
Don't overread it
Observational and from one Japanese city; these symptoms mark risk rather than cause it, and treating them has not been shown to change mortality.
The statistics, in plain English
Hazard ratios of 1.09 to 1.49 are modest for an individual — a score of 3 is associated with about half again the mortality rate, not a doubling. What makes it useful is the clean gradient and the enormous sample, which makes the estimates precise: the intervals are narrow. The xerostomia result is a genuine null, with an interval (0.98-1.06) tight enough to say that any independent effect is very small, not merely undetected.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for geriatrics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free