Study design and participants
This was a cross-sectional study based on a survey on 34,564 community-dwelling older adults aged ≥ 60 years in Iwamizawa, a regional city in Hokkaido, northern Japan. This study included older adults who participated in wellness checkups in CHEER Iwamizawa (a research project titled “Checkup to discover HEalth with Energy for senior Residents” in Iwamizawa) in October 2020. The participants of CHEER Iwamizawa were recruited through a public relations magazine published by the city council and flyers posted at major public facilities in the city. In addition, local government employees visited older adults’ social clubs to describe the study’s purpose and recruit participants. The survey content was explained verbally and in writing to the participants, and written informed consent was obtained prior to the survey. This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Faculty of Dental Medicine, Hokkaido University (approval number: 2020–9).
Evaluation of outcome
OHRQoL was evaluated using the Japanese version of the GOHAI46. The GOHAI consists of 12 questions rated on a five-point Likert scale assessing the frequency of oral health-related problems over the past three months. The total score ranges from 12 to 60; the higher the score, the higher the OHRQoL. The median scores, as per the GOHAI national norms for Japanese individuals, are 56.0 and 52.5 for men aged 60 − 69 years and 70 − 79 years, respectively, and 54.0 and 53.0 for women aged 60–69 years and 70 − 79 years, respectively. Therefore, in this study, participants scoring below these cutoff values were defined as the poor OHRQoL group, while the remaining participants were assigned to the good OHRQoL group.
Defining the risk of poor OHRQoL
CS and LEH were defined as risk factors for poor OHRQoL, and their presence or absence was examined using self-administered questionnaires. CS was assessed by the question, “Do you feel stressed by the lifestyle changes and restrictions on going out due to the COVID-19 pandemic?” (Yes/To some extent/Not really/No). Participants who answered either yes or to some extent were considered to have CS. LEH was assessed by the question, “Do you perform exercises, such as walking, at least once a week?” (Yes/No). Participants who answered no were considered to lack exercise habits. Then, four mutually exclusive groups (risk of poor OHRQoL) were created based on the presence or absence of both CS and LEH (Group 1: no CS and no LEH; Group 2: no CS and LEH; Group 3: CS and no LEH; Group 4: both CS and LEH).
Data on participants’ age, sex, smoking status, body mass index, and medical history (malignant neoplasm, stroke, myocardial infarction, depression, and osteoarthritis) were collected using a self-administered questionnaire.
Instrumental activities of daily living
IADL was assessed using the JST-IC38. The JST-IC consists of 16 questions with a “Yes/No” response format; the total score ranges from 0 to 16. No specific cutoff value was set; higher scores indicated higher IADL. The internal consistency obtained in the study population and example items of questionnaires are shown in Supplemental Table S4 (available online).
Assessment of depressive tendencies
The Japanese version of the Geriatric Depression Scale short form (GDS-15) was used to assess depressive tendencies. The GDS-15 comprises 15 questions in a “Yes/No” response format, with the total score ranging from 0 to 15. In the present study, a total score of ≥ 6 indicated the presence of depressive tendencies47. The internal consistency obtained in the study population and example items of questionnaires are shown in Supplemental Table S4 (available online).
Number of teeth
The number of teeth erupted in the oral cavity was recorded, excluding stump teeth and teeth with severe mobility.
Diagnosis of oral hypofunction
Qualified examiners who received two hours of instruction and training from the authors (K.M. and Y.W.) regarding the appropriate data collection methods for measurements performed oral function measurements. However, interexaminer reliability measures were not obtained. There was no strict calibration of examiners for oral function measurement. However, this potential nondifferential misclassification would probably bias results toward the null hypothesis and not lead to overestimation of the observed associations. Significant robust associations were observed under these circumstances. The oral function was objectively assessed using six of the seven parameters previously described to diagnose oral hypofunction48: oral hygiene, oral moisture, occlusal force, tongue and lip movement, tongue pressure, and masticatory function (excluding swallowing function). Oral hygiene was assessed using the tongue coating index49 to determine the degree of tongue coating by visual inspection. Oral moisture was measured at the center of the tongue dorsum, approximately 10 mm from the apex of the tongue, using an oral moisture checker (Mucus, Life Co., Ltd., Saitama, Japan). Each measurement was taken thrice, and the median value was used. The occlusal force of the entire dentition was measured using a pressure-indicating film (Dental Prescale II, GC Corp, Tokyo, Japan) during three seconds of clenching in the intercuspal position. For denture users, the occlusal force was measured with their dentures in place. Tongue and lip movements were assessed using oral diadochokinesis. Participants were asked to repeat the syllables /pa/, /ta/, and /ka/ for five seconds, and the number of each syllable pronounced per second was counted using an automatic counter (Kenkoukun Handy, Takei Scientific Instruments Co., Ltd., Niigata, Japan). Tongue pressure was measured thrice using a tongue pressure measuring instrument (JMS tongue pressure measuring instrument, JMS Co., Ltd., Hiroshima, Japan), and the maximum value was used. The masticatory function was measured using a masticatory ability testing system (Glucosensor GS-II, GC Corp., Tokyo, Japan).
Using previously established guidelines48, the cutoff values for the above six items were as follows: tongue coating index ≥ 50%; oral moisture < 27.0; occlusal force < 500 N; any of the /pa/, /ta/, or /ka/ syllables repeated < 6 times/second for diadochokinesis; tongue pressure < 30 kPa; and masticatory function < 100 mg/dL. Oral hypofunction was defined when at least three of the six measurements met these criteria.
Sample size calculation
Sample size calculation was performed using G*Power 188.8.131.520. Assuming a two-tailed Mann–Whitney U test with α = 0.05 and d = 0.50, an estimated 134 participants were required to achieve a power of 0.80.
Descriptive statistics, comparisons between the two groups with and without CS, and comparisons between the two groups with and without LEH were conducted by performing Mann–Whitney U tests for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. The scoring of categorical valuables is shown in Supplemental Table S5 (available online). Comparisons among the four groups of risk factors for poor OHRQoL were made by performing the Kruskal–Wallis test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables. Thereafter, Poisson regression with robust standard errors51 was used to calculate the prevalence ratio for poor OHRQoL of CS alone and LEH alone, respectively. Adjustment factors included sociodemographic characteristics (i.e., age, sex, body mass index, JST-IC), depressive tendencies27, and the number of teeth24. Subsequently, the prevalence ratio for poor OHRQoL regarding the risk factors of poor OHRQoL was calculated. When the study design is similar to the present study, the odds ratio is often obtained by logistic regression analysis. However, in logistic regression analysis, the odds ratio is known to deviate from the true relative risk as the frequency of the outcome increases, overestimating the odds ratio when the risk ratio is greater than 1.0 and underestimating the odds ratio when the risk ratio is less than 1.0. 52. Therefore, in this study, Poisson regression with robust standard errors was used to calculate estimates that approximate the true relative risk. All analyses were performed using SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA), and the significance level (two-tailed) was set at 5%. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.