Materials and methods

Study design and participants

This study employed a cross-sectional design to investigate the associations between OHRQoL, emotional loneliness, sleep experiences, and dream recall among older adults.

Participants were 609 community-dwelling older adults in Greece, aged 60 years and older. They were recruited through community centers, local health units, and social associations using a non-probability convenience sampling strategy. Approximately 700 community-dwelling older adults were approached for participation, and 609 eligible participants with complete data for the variables examined in the present study were included in the analyses.

Before recruitment, the minimum sample size was estimated using a standard single-population-proportion calculation, assuming a 95% confidence level (two-sided α=0.05), a 5% margin of error, and an expected proportion of 0.50 to represent maximum variability. This calculation yielded a target sample of approximately 384 participants. It was a precision-based recruitment estimate rather than an effect-size-based power analysis for the correlational or regression models. The final analytic sample of 609 participants exceeded this target. No additional exclusion criteria were applied beyond the predefined eligibility criteria, and participants with incomplete data for the variables examined in the present study were not included in the analyses.

Eligibility criteria included the following: (a) age ≥60 years and (b) the ability to understand the study information, provide informed consent, follow the questionnaire instructions, and complete the questionnaire independently. No standardized cognitive screening instrument was administered; therefore, cognitive status was not formally established.

Ethical approval for the study was obtained from the Ethics Committee of the University of Ioannina (approval number: 27934; date: July 24, 2024). All procedures were conducted in accordance with the principles of the Declaration of Helsinki and the General Data Protection Regulation (GDPR). Participants provided electronic informed consent prior to participation, were assured of anonymity and confidentiality, and retained the right to withdraw from the study at any time without consequence. No sensitive personal data was collected.

Measures

Oral Health-Related Quality of Life (OHRQoL)

OHRQoL was assessed using the Geriatric Oral Health Assessment Index (GOHAI) [26]. The GOHAI is a widely used self-report instrument designed to evaluate OHRQoL in older adults across three domains: physical function (e.g., chewing and speaking), psychosocial function, and pain or discomfort. The scale consists of 12 items assessing functional, psychosocial, and pain-related aspects of oral health. Higher scores indicate better perceived OHRQoL. The GOHAI has demonstrated satisfactory reliability and validity across diverse populations. Cronbach’s α for the present sample was 0.732.

Emotional Loneliness

Loneliness was assessed using the six-item version of the De Jong Gierveld Loneliness Scale [27], a widely used instrument in aging research with well-established psychometric properties. The scale measures two distinct dimensions of loneliness: emotional loneliness, reflecting the absence of close emotional attachment and intimate relationships, and social loneliness, reflecting deficiencies in the broader social network. Subscale scores were calculated according to the recommended scoring procedures, with higher scores indicating greater levels of loneliness. In the present study, emotional loneliness was operationalized using the three-item Emotional Loneliness subscale, with higher scores indicating stronger feelings of emotional isolation and a lack of meaningful emotional connectedness. The three-item Emotional Loneliness subscale demonstrated satisfactory internal consistency in the present sample (Cronbach’s α=0.80). This coefficient was calculated specifically for the three emotional loneliness items included in the analyses, rather than for the full six-item scale.

Sleep Quality and Dream Variables

Subjective sleep and dream-related indicators were assessed using five brief study-specific self-report items. The exact wording was: “Do you think you get enough sleep?” for sleep sufficiency; “Do you think your sleep is good?” for subjective sleep quality; “Do you tend to take naps during the day?” for daytime napping; “Do you remember your dreams?” for dream recall frequency; and “Do you like remembering your dreams?” for attitudes toward dream recall.

The items assessing sleep sufficiency, subjective sleep quality, daytime napping, and dream recall frequency were rated on five-point ordinal scales, with higher scores indicating greater sleep sufficiency, better perceived sleep quality, more frequent daytime napping, and more frequent dream recall, respectively. Attitudes toward remembering dreams were assessed on a three-category scale (no, indifferent, yes).

These items were developed for the purposes of the present study to provide a brief and feasible assessment of subjective sleep and dream experiences in community-based older adults. As they were not part of a previously validated standalone questionnaire, they should be interpreted as brief subjective indicators rather than comprehensive or psychometrically validated measures. Consequently, they may be more susceptible to measurement error and limited construct validity compared to standardized multi-item instruments.

Despite these limitations, similar single-item or brief measures have been widely used in population-based aging research as pragmatic indicators of subjective sleep experience [12,13].

Sociodemographic Characteristics

Information regarding age, gender, educational level, marital status, living arrangements, and chronic disease status was also collected.

Statistical analysis

Data were analyzed using IBM SPSS Statistics for Windows, Version 25 (Released 2017; IBM Corp., Armonk, New York, United States).

Descriptive statistics, including means, standard deviations, frequencies, and ranges, were calculated to characterize the sample and the distributions of the study variables. Pearson product-moment correlation coefficients were computed to examine bivariate associations among oral health-related quality of life, emotional loneliness, sleep sufficiency, subjective sleep quality, daytime napping, dream recall frequency, and age.

Group differences in the primary study variables were examined using independent-samples t-tests and one-way analyses of variance, as appropriate. Effect sizes were considered alongside statistical significance.

Hierarchical multiple regression analyses were conducted to examine whether OHRQoL remained independently associated with subjective sleep quality and dream recall frequency. The two outcomes were examined in separate models. At Step 1, age and gender were entered as demographic covariates. At Step 2, OHRQoL and emotional loneliness were entered simultaneously to evaluate their incremental contribution beyond age and gender. Changes in explained variance (ΔR²), standardized regression coefficients (β), 95% confidence intervals, and corresponding p-values were examined.

Sleep sufficiency was not examined as an additional regression outcome because it was almost perfectly correlated with subjective sleep quality (r=0.989). Separate regression analyses of these two indicators would therefore have been substantively redundant. Subjective sleep quality was retained as the broader global appraisal of sleep, while the bivariate findings for sleep sufficiency were reported descriptively and interpreted cautiously.

Regression analyses were conducted using complete-case data (n=607), following listwise exclusion of participants with missing values in the variables included in the models. Analyses concerning emotional loneliness and subjective sleep quality were conducted in relation to the theory-guided hypotheses stated in the Introduction, whereas analyses involving dream recall frequency were exploratory. Because the hypotheses and analytical distinctions were not preregistered, the findings should not be regarded as strictly confirmatory.

No formal adjustment for multiple comparisons was applied. Accordingly, the reported p-values are unadjusted, and the possibility of Type I error cannot be excluded. Weaker findings, particularly those involving dream recall frequency, were therefore interpreted as exploratory and not as statistically robust evidence of an association. All statistical tests were two-sided, and statistical significance was set at p<0.05.