Does poor dental health predict becoming homebound among older Japanese?

This prospective cohort study used data from the Japan Gerontological Evaluation Study (JAGES) Project [22, 23]. The JAGES Project investigated social, behavioral, and health factors in people aged 65 years or older. The JAGES sample was restricted to people who did not already have physical or cognitive disabilities, which were defined as those without eligibility for receiving long-term public care insurance benefits. Thus, the goal of the JAGES Project was to assess the health status and social determinants of able-bodied people aged 65 years or older. The present longitudinal study used panel data from surveys conducted in 2006 and 2010.

We mailed self-reported questionnaires to people who were randomly selected from the list of older residents not receiving long-term public care insurance benefits in each municipality. However, as the study fields were not randomly selected from whole Japanese municipalities, we did not apply weight analysis. In 3 municipalities where the dental health questionnaires were distributed for both study periods, 7270 individuals completed the questionnaire in 2006, and 5589 responded to our questionnaire in 2010 (follow-up rate: 72.39 %). The present study determined which respondents had become homebound between 2006 and 2010, and we excluded 640 individuals who were homebound at baseline (2006). Of these respondents, 211 were excluded because of a lack of information regarding being homebound in 2006, 348 were excluded because of a lack of information regarding being homebound in 2010. Finally, 92.66 % of the participants were included in our analyses. Data collected from 4390 respondents were included in the analysis (Fig. 1).

The JAGES protocol was reviewed and approved by the Ethics Committee on Research of Human Subjects of Nihon Fukushi University. The questionnaire was sent via mail, with a written explanation of the study aim. The people who returned the questionnaire were regarded as having provided consent to participate in the survey. The authors obtained the permission to use the data from the JAGES data management committee.

Homeboundness has several similar definitions. The criteria of homeboundness include going out of one’s home once weekly or less often [4], less often than once weekly [24] and once monthly or less often [25]. Since being homebound was considered to be affected by social and cultural background, we applied the definition used in Japan, which is going out of one’s home less often than once weekly [24]. This reflects not only physical reasons for being confined to one’s home, such as being bedridden, but also psychosocial or geographical reasons. Frequency of going outdoors was measured by using the question, “How often do you usually go outside of the house (this includes shopping, meeting with people, walking, going to the hospital, and other activities)?”

The number of remaining teeth and the use of dentures at baseline (2006) were used as variables representing dental health status. They were assessed by responses to the self-administered questionnaire. Previous studies have indicated that older people with 20 or more teeth ingested more food relative to those with fewer than 20 teeth [26]. Therefore, we categorized the numbers of remaining teeth as follows: 20 or more teeth, fewer than 20 teeth with dentures, and fewer than 20 teeth without dentures.

From previous studies, we selected physical, psychological, and geographical variables as the covariates [1, 7, 27, 28]. As in the previous studies, the following questions regarding sociodemographic characteristics, baseline health status, and risk factors of homeboundness were included in the analyses as covariates: age, sex, marital status, educational attainment, annual household income, comorbidity, depression, walking time (min/day), living alone, and area of residence at baseline (2006). To minimize the effect of age as a strong confounding factor, we conducted a stratified analysis according to age and included age as a continuous variable in the models. Marital status was categorized as married, widowed, separated, never married, and other. Educational attainment was categorized as follows: < 6 years, 6–9 years, 10–12 years, and ≥ 13 years. Annual household income was categorized as follows: < $20,000 (< ¥2,000,000), $20,000–$29,999 (¥2,000,000–¥2,999,999), $30,000–$39,999 (¥3,000,000–¥3,999,999), and ≥ $40,000 (≥¥4,000,000) (US$1 = ¥100). Comorbidity was measured by using the question, “Do you receive treatment now?” (to which respondents answered yes or no). Depressive symptoms were measured via the short version of the Geriatric Depression Scale (GDS-15) by using a simple yes/no format suitable for self-administration. Previous studies have concluded that a cutoff point of 5 is appropriate for use with the GDS as a screening tool for depression in community-dwelling older adults in the United States and Japan [29, 30]. Therefore, a cutoff point of 5 was used as an indication of depression in this study. Walking time was categorized as follows: < 30, 30–59, 60–89, and ≥ 90 min/day. The categories representing the numbers of people living in respondents’ households were “living alone” or “living with others.” Because the risk of being homebound was affected by residential area, residential municipality was also used as a categorical covariate as follows: living area A, living area B, or living area C. The 2005 population densities [31] for areas A, B, and C were 1,026.6, 566.9, and 572.9 persons/km², respectively.

We calculated prevalence and 95 % confidence interval (CI) for respondents who were homebound due to dental health. We then calculated the odds ratios (ORs) of the baseline dental health status for homeboundness four years later. Univariate and multivariate ORs and 95 % CIs for dental health were calculated. As age was strongly associated with both dental health and homebound status, we applied a stratified analysis according to age. The participants were stratified into two groups as follows: 65–74 years old or 75 years or older. In the multivariate model, age, sex, marital status, educational attainment, annual household income, comorbidity, depression (GDS-15), walking time (min/day), living alone, and area of residence at baseline were included. We put emphasis on the theoretical importance of the covariates, and included all covariates in the multivariate model. If data were missing for explanatory variables, the corresponding observations were assigned to “missing” categories. In the present analysis, we selected the study population from three municipalities. The sampling rates and populations of the municipalities were similar. As a result, the sampling weight was not applied in the analyses. All analyses were performed by using SPSS version 22.0 statistical software.

In addition, the results of the logistic regression were used to calculate the population-attributable fraction (PAF) of dental health for being homebound. The PAF is defined as “the proportional reduction in population disease or mortality that would occur if exposure to a risk factor were reduced to an alternative ideal exposure scenario (e.g., no tobacco use)” by the World Health Organization [32]. The PAF has been used to examine the impact of risk factors on health outcomes in various public health studies [33, 34]. The impact of dental health on the population’s risk of being homebound was estimated according to the PAF concept.

Data are from the JAGES study. All enquiries are to be addressed at the data management committee via e-mail: dataadmin@jages.net. All JAGES datasets have ethical or legal restrictions for public deposition due to inclusion of sensitive information from the human participants.