Ageing of the world's population is accelerating. Despite the health challenges they face, older people should not be stereotyped as frail and dependent. Healthy ageing is possible if adequate care is provided and effective intervention is delivered for healthy living. Common oral diseases (dental caries and periodontal diseases) are determined to a great extent by health behaviours (diet and oral hygiene) and are controllable through effective interventions. Hence, this study aims to test the theoretical models for health promotion, develop theory-derived intervention, and evaluate its effectiveness in improving oral health of older adults. The study hypothesis is that theory-derived intervention is more effective than conventional health education in eliciting positive behaviours and controlling oral diseases in older adults.

This study targets older adults and aims to (i) test the explanatory power of the dominant theoretical models/ frameworks, (ii) develop theory-derived intervention, and (iii) evaluate the effectiveness of the theory-derived intervention. The hypothesis is that theoretical-derived intervention is more effective than conventional health education in improving oral health behaviours and preventing oral diseases among older adults.

The sample size needed for this study was calculated by using the software G*Power version 3.1.9.2 (Franz Faul, University of Kiel, Germany). The sample size calculation was based on the primary outcomes, namely increment of coronal caries and increment of root caries, and independent t-tests for comparing means of two independent groups. A caries incidence study among the elderly reported that the mean (SD) 5-year caries incidence were 2.65 (3.14) for coronal caries and 2.21 (2.83) for root caries 17. The 2-year incidence of coronal caries and root caries were therefore estimated to be 1.06 (1.26) and 0.88 (1.13), respectively, in the control group. The effect of an intervention is considered clinically significant if it reduces the caries incidence by 40%. Based on a significance level of 5% and a power of 80%, 140 subjects per group are needed for coronal caries and 165 subjects are needed for root caries outcome measures. To take both outcome measures into account, the higher number (i.e., n=165) will be adopted. Allowing for a 25% attrition rate, 220 elderly persons will need to be recruited into each group, giving a total sample size of 440.

Each participant will be required to complete a detailed questionnaire at the initial visit through face-to-face interviews. Information on socio-demographic background (gender, age, education level, past occupation, and type of housing) and oral health behaviours (diet, toothbrushing and dental flossing) will be collected. The questionnaire will also include the scales for three dominant theoretical models, namely, Health Belief Model (HBM), Theory of planned behaviour (TPB) and Social Cognitive Theory (SCT).

The theoretical model that best explains the health behaviours will be selected for designing oral health intervention. Intervention materials and activities will be carefully designed to address all the constructs/ domains of the selected theoretical model(s). Relevant medical literatures will be referred to. The needs, interest, and health literary of the target group (i.e. older adults of varied social and educational backgrounds) will be taken into full consideration. The intervention will be tested in a small group of 15-20 older people to ensure its relevance, practicality, and acceptability.

Participants will be randomly assigned to two groups, stratified by gender and education level. Allocation concealment will be ensured by using sealed opaque envelopes. The two groups will receive conventional health education and theory-derived intervention, respectively.

At baseline and 6-, 12-, and 24-months post-intervention, each participant will a) undergo an oral examination (dental checkup) by a trained dentist, who will be calibrated against an experienced oral epidemiologist and blinded to the group allocation, and b) complete a questionnaire (detailed questionnaire at baseline and short questionnaire for follow-ups). Chi-square tests will be used for comparing proportions. Parametric or non-parametric methods will be adopted for comparing means. ANCOVA (analysis of covariance) and multiple regressions will be performed to evaluate the effects of various factors on the outcomes.

Inclusion Criteria:

• 55-70 years old
• Living in the community (alone or with family members), instead of residing in a nursing home
• Having at least 8 natural teeth
• No plan to immigrate in 3 years

Exclusion Criteria:

• Life-threatening disease; impaired cognitive function (e.g. Alzheimer's syndrome); radiotherapy in the head and neck region