Factors associated with advance directive completion among older adults with and without cognitive impairment: a nationwide cross-sectional study.
Frontiers in public health · 2026
Abstract originale
OBJECTIVE: This study aimed to identify and compare the factors associated with advance directive (AD) completion between older adults with and without cognitive impairments, using a nationwide cross-sectional dataset. METHODS: This study analyzed cross-sectional nationwide data obtained from the 2023 National Survey of Older Koreans, which is conducted every 3 years by the Korean Ministry of Health and Welfare. The study measured cognitive function in a total of 9,951 participants, using the Korean version of the Mini-Mental State Examination for Dementia Screening (K-MMSE-DS). The participants were classified into cognitive impairment (n = 2,448) and normal cognition (n = 7,503) groups, and the differences in AD completion based on demographic characteristics, health-related characteristics, and attitudes toward death of the participants in each group were analyzed using the chi-square test. A multivariable logistic regression model was used to identify the factors associated with AD completion. RESULTS: Multivariable logistic regression analysis showed that religious affiliation, poorer nutritional status, higher exercise frequency, participation in death preparation education, and higher awareness of dying well were commonly associated with AD completion in both the cognitive impairment and normal cognition groups. Group-specific associations were also identified: in the cognitive impairment group, AD completion was additionally associated with younger age, female sex, higher education, increased prescription medication use, higher independence in activities of daily living, and higher depressive symptom scores, whereas in the normal cognition group, it was additionally associated with a higher number of chronic diseases and history of falls within the past year. CONCLUSION: Different patterns factors associated with AD completion were observed between older adults with and without cognitive impairment, suggesting that strategies to promote AD completion may need to be tailored according to cognitive status.
Parole chiave
Leggi il testo integrale originale
Testo estratto dalla pubblicazione: sono esclusi immagini, media, supplementi e parti soggette a diritti distinti.
Factors associated with advance directive completion among older adults with and without cognitive impairment: a nationwide cross-sectional study
Abstract
Objective
This study aimed to identify and compare the factors associated with advance directive (AD) completion between older adults with and without cognitive impairments, using a nationwide cross-sectional dataset.
Methods
This study analyzed cross-sectional nationwide data obtained from the 2023 National Survey of Older Koreans, which is conducted every 3 years by the Korean Ministry of Health and Welfare. The study measured cognitive function in a total of 9,951 participants, using the Korean version of the Mini-Mental State Examination for Dementia Screening (K-MMSE-DS). The participants were classified into cognitive impairment (n = 2,448) and normal cognition (n = 7,503) groups, and the differences in AD completion based on demographic characteristics, health-related characteristics, and attitudes toward death of the participants in each group were analyzed using the chi-square test. A multivariable logistic regression model was used to identify the factors associated with AD completion.
Results
Multivariable logistic regression analysis showed that religious affiliation, poorer nutritional status, higher exercise frequency, participation in death preparation education, and higher awareness of dying well were commonly associated with AD completion in both the cognitive impairment and normal cognition groups. Group-specific associations were also identified: in the cognitive impairment group, AD completion was additionally associated with younger age, female sex, higher education, increased prescription medication use, higher independence in activities of daily living, and higher depressive symptom scores, whereas in the normal cognition group, it was additionally associated with a higher number of chronic diseases and history of falls within the past year.
Conclusion
Different patterns factors associated with AD completion were observed between older adults with and without cognitive impairment, suggesting that strategies to promote AD completion may need to be tailored according to cognitive status.
Keywords: advance care planning, advance directives, aged, cognitive dysfunction, dying well
1.Introduction
With advancements in medical science and rapid population aging, South Korea has transitioned into a super-aged society, intensifying social demands for a “dignified death” and high-quality end-of-life care (1). As of 2025, approximately 10.5 million individuals in South Korea are aged 65 years or older, with an estimated 955,584 (9.09%) living with dementia (2). Furthermore, mild cognitive impairment affects approximately 28.09% of older adults, representing a substantial population at an elevated risk of developing dementia (2). Given the aging population and increasing prevalence of dementia, a growing proportion of older adults are at risk of losing their decision-making capacity (3). In this context, Advance Care Planning (ACP)—a process through which individuals, together with their families and healthcare providers, discuss and document preferences for future care, including for situations in which decision-making capacity may be lost—has become increasingly important (4). As one of the key components of ACP, Advance Directives (ADs) allow individuals to specify their preferences regarding future medical treatment, and have emerged as an important mechanism for protecting patient autonomy and improving the quality of end-of-life care (5). In South Korea, AD completion has been legally institutionalized under the Act on Hospice-Palliative Care and Decisions on Life-Sustaining Treatment for Patients at the End of Life, which took effect in 2018 and enables individuals to document their preferences regarding life-sustaining treatment in advance (6).
Dementia is characterized by a progressive, longitudinal disease trajectory in which cognitive function gradually deteriorates; therefore, ACP should be approached as an ongoing, dynamic process (7). Ideally, ACP should be initiated before individuals lose their communication and decision-making capacities, enabling them to actively participate in discussions and decisions regarding future end-of-life care (7).
When individuals discuss end-of-life care preferences with their families and thereafter complete ADs, they are more likely to receive end-of-life care aligned with their preferences (5) and less likely to undergo nonbeneficial life-sustaining treatments (8). Among patients with dementia, AD completion is associated with reduced use of burdensome interventions, such as feeding tubes, mechanical ventilation, and cardiopulmonary resuscitation, as well as increased utilization of hospice and palliative care focused on symptom relief and comfort (9). Furthermore, AD completion has been shown to significantly reduce the emotional burden experienced by family members during surrogate decision-making processes (10).
Given these benefits, establishing effective institutional frameworks that support ACP and AD completion among older adults with dementia is essential to protect patient self-determination and improve the quality of end-of-life care. Accordingly, there is an urgent need to investigate the current status of AD completion and identify the associated multidimensional factors across different levels of cognitive function.
Previous research on AD completion has predominantly focused on the general older adult population or patients with terminal cancer; therefore, a significant gap remains regarding the specific needs of older adults with cognitive impairment (11). Older adults with cognitive decline exhibit distinct differences from their cognitively normal peers across various dimensions, including health status, chronic disease burden, medication patterns, healthcare utilization, attitudes toward death, and overall life satisfaction (12). Crucially, this cognitive and clinical heterogeneity extends into the realm of end-of-life decision making. The predictive factors and underlying mechanisms influencing AD completion among cognitively impaired older adults diverge significantly from those observed in their cognitively intact peers (13).
Furthermore, factors influencing AD completion are multi-layered, encompassing not only sociodemographic characteristics, but also clinical conditions, health behaviors, and psychological factors (such as suicidal ideation) (14, 15). Despite these complexities, there is a paucity of nationwide, large-scale comparative studies evaluating the distinct predictors of AD completion between cognitively impaired and cognitively normal subgroups. Notably, our group previously examined factors associated with AD completion among community-dwelling older Korean adults using data from the 2020 National Survey of Older Koreans (15); the present study extends this earlier work by employing the more recent 2023 survey wave and, in particular, by stratifying analyses according to cognitive status to examine whether the factors associated with AD completion differ between cognitively impaired and cognitively normal older adults, a distinction not addressed in our previous study.
To address this gap, this study utilized nationwide cross-sectional data from the 2023 National Survey of Older Koreans—a representative, population-based survey conducted every 3 years by the Korean Ministry of Health and Welfare. This study aimed to identify and compare the factors associated with AD completion—including demographic profiles, health-related characteristics, and attitudes toward death—between older adults with and without cognitive impairment, using multivariable logistic regression analysis. The findings may provide a foundation for developing cognitive status-tailored strategies to promote AD completion and for informing policies that support older adults’ right to a dignified death.
2.Materials and methods
2.1.Study design
This study was a secondary analysis of nationwide cross-sectional survey data, conducted to compare the factors associated with AD completion between older adults with cognitive impairment and those with normal cognition.
2.2.Data collection and participants
Data for this secondary analysis were drawn from the 2023 wave of the National Survey of Older Koreans, a nationally representative survey administered every 3 years by the Korean Ministry of Health and Welfare (16). Trained interviewers (n = 176) conducted face-to-face interviews using the Tablet-PC Assisted Personal Interview method between September 4 and November 12, 2023; all interviewers had prior experience with similar surveys and completed an additional 8-h training session before data collection. Eligible participants were community-dwelling adults aged 65 years or older residing in one of 17 Korean cities or provinces, identified through a proportional two-stage stratified sampling design in which the population was first stratified by the 17 metropolitan cities/provinces and then, within the nine non-metropolitan provinces, re-stratified by neighborhood (16). Sampling weights accounting for household- and individual-level nonresponse and poststratified to 2023 Statistics Korea population estimates (region, sex, age) were computed by the Ministry of Health and Welfare research team (16); however, as described in the Statistical Analysis section, these weights and the associated strata/clustering variables were not incorporated into the present analysis, which instead relied on conventional (non-survey-adjusted) multivariable logistic regression. Of the 10,078 individuals who responded to the 2023 survey, 9,951 were retained for this study after excluding 127 participants with missing cognitive-function test results; the research team received de-identified, coded data for secondary analysis.
2.3.Ethical considerations
The 2023 National Survey of Older Koreans was approved by Statistics Korea as a statutory survey (Approval No. 117071), and all procedures related to its data collection—including the survey instruments and informed consent process—were reviewed and approved by the Institutional Review Board of the Korea Institute for Health and Social Affairs (IRB No. 2023-078). During the original survey, IRB-approved consent forms were administered by trained investigators; participants were provided with sufficient information about the survey, their autonomy was fully respected, and written informed consent was voluntarily obtained from all participants prior to participation, in accordance with ethical standards. The present secondary-analysis protocol, which used only de-identified data, was separately reviewed and approved by the Institutional Review Board of G University (IRB No. 1044396-202506-HR-099-01), the corresponding author’s affiliated institution. This study was conducted in accordance with all applicable ethical guidelines and regulations.
2.4.Measurements
2.4.1.Cognitive impairment
Cognitive function was assessed using the Korean version of the Mini-Mental State Examination for Dementia Screening (K-MMSE-DS) (17). Participants were classified as having cognitive impairment if their K-MMSE-DS score fell 1.5 standard deviations below the age- and education- adjusted normative mean, consistent with the cut-off criteria established for this screening instrument.
2.4.2.Sociodemographic characteristics
Sociodemographic information—age (continuous, years), sex (male/female), spouse (living with a spouse: yes/no), living arrangement (living alone, married couple, household with children, or other), perceived health status (5-point scale, very healthy to very unhealthy), religion (yes/no), and educational attainment (years of formal education, continuous)—was obtained by self-report.
2.4.3.Physical health-related characteristics
A broad set of physical health indicators was collected, comprising the number of diagnosed chronic diseases (continuous), number of prescribed medications used for more than 3 months (continuous), number of medical visits in the past month (continuous), alcohol consumption (yes/no, based on past-year drinking frequency), smoking status (current, former, never), nutritional status (continuous score, detailed below), body mass index (BMI, kg/m2, continuous), need for assistance with daily activities (a single self-reported yes/no item, distinct from the ADL/IADL classifications below), activities of daily living (ADL) and instrumental activities of daily living (IADL) (each dichotomized as fully independent vs. not fully independent, defined below), exercise frequency (days/week, continuous), history of falls in the past year (yes/no), sleep quality (5-point scale, very satisfied to not at all satisfied), registered disability status (yes/no), and functional limitation (yes/no).
Nutritional status was evaluated with the Nutritional Screening Initiative checklist (18), a 10-item, differentially weighted instrument yielding scores from 0 to 21. Following the instrument’s established cut-offs, participants were grouped into good (0–2), moderate-risk (3–5), or high-risk (≥6) nutritional categories; the continuous score was entered into the regression models.
Participants were classified as fully independent in ADL, if they reported being “fully independent” on all seven ADL items, and as fully independent in IADL if they reported being “fully independent” on all 10 IADL items; participants who did not meet this criterion on any item were classified as not fully independent.
2.4.4.Emotional characteristics
Three self-reported indicators captured participants’ emotional status: depression (continuous score), satisfaction with quality of life (5-point scale, very satisfied to very dissatisfied), and past-year suicidal ideation (yes/no). Depressive symptoms specifically were measured with the 15-item Korean version of the Short Form Geriatric Depression Scale (SGDS-K) (19); items are scored dichotomously (0 or 1) and summed (range 0–15), with higher totals denoting more severe depressive symptoms. The continuous score was entered into the regression models.
2.4.5.Primary outcome and attitude toward death
All items in this section were drawn from the 2023 National Survey of Older Koreans conducted by the Korean Ministry of Health and Welfare (16). AD completion was the primary outcome variable, assessed using the item, “Have you completed an advance directive in preparation for death at the end of life?” with responses coded as 1 = yes and 0 = no for the regression analyses.
Three additional indicators captured participants’ attitudes toward death: agreement with medical care for life prolongation (5-point scale, strongly agree to strongly disagree), prior participation in death preparation education (yes/no), and awareness of the significance of dying well.
The latter construct was derived from five survey items asking respondents to rate the perceived importance of specific end-of-life circumstances—namely, being able to prepare for one’s own dying process, dying free of pain, dying surrounded by family or close friends, not becoming a burden to family or close friends, and dying at home rather than in a hospital or care facility—each scored on a 5-point scale (1 = very important, 5 = not at all important) and summed into a composite ranging from 5 to 25. Because a lower rating on each item denoted greater perceived importance, a lower composite score corresponded to a higher overall level of awareness regarding the significance of dying well. This construct reflects general attitudes toward end-of-life priorities and is conceptually distinct from participants’ specific knowledge of ADs or the AD completion process, which was not separately assessed in the survey.
2.5.Statistical analysis
All statistical analyses were performed in IBM SPSS Statistics for Windows, version 30.0 (IBM Corp., Armonk, NY, USA); statistical significance was set at a two-tailed α of 0.05. Responses recorded as “not applicable” or “no response” were treated as missing data and handled via listwise deletion. Group comparisons between the cognitive impairment and normal cognition groups were conducted with independent-samples t-tests (continuous variables) and chi-square tests (categorical variables).
To identify factors associated with AD completion (coded as 1 = completed, 0 = not completed), multivariable binary logistic regression analyses were performed separately for the cognitive impairment group and the normal cognition group. Independent variables that showed statistically significant associations with AD completion in the bivariate analyses were entered simultaneously into each model using the forced-entry method; reference categories for categorical predictors are specified in the corresponding regression tables. Prior to fitting the regression models, multicollinearity among the independent variables was evaluated using variance inflation factors (VIF), with no evidence of severe multicollinearity detected (VIF < 10). The assumption of linearity of the logit for continuous variables was also verified. Model fit was evaluated using the Hosmer–Lemeshow goodness-of-fit test and Nagelkerke R2. Because separate regression models were fitted for each cognitive group rather than a single pooled model with interaction terms, these analyses identify within-group associations; this methodological boundary is further discussed in the Limitations section. Additionally, as noted in Section 2.2, sampling weights, stratification, and clustering from the complex survey design were not incorporated into these analyses. Because our analytic goal was to estimate within-group associations between covariates and AD completion rather than to generate population-level prevalence estimates, a conventional (non-survey-adjusted) multivariable logistic regression was considered statistically appropriate: established survey-methodology literature indicates that sampling weights are essential for unbiased descriptive or prevalence estimation, but are not necessarily required—and can reduce estimation efficiency—when the analytic objective is to estimate associations within a defined subpopulation under a correctly specified model (20–23).
3.Results
3.1.Differences in participant characteristics by cognitive impairment status
The cognitive impairment group (n = 2,448) and the normal cognition group (n = 7,503) differed significantly across sociodemographic, health-related, emotional, and attitudinal domains (Table 1). The cognitive impairment group was significantly older (74.67 ± 6.96 vs. 73.81 ± 6.67 years, p < 0.001), had a higher proportion of female participants (66.8% vs. 59.9%, p < 0.001), a higher rate of living alone (38.8% vs. 33.0%, p < 0.001), and lower educational attainment (7.75 ± 4.40 vs. 8.40 ± 4.17 years, p < 0.001) than the normal cognition group. Regarding physical health, the cognitive impairment group had more chronic conditions (2.40 ± 1.64 vs. 2.12 ± 1.58, p < 0.001), greater prescription medication use (2.15 ± 1.52 vs. 1.99 ± 1.61, p < 0.001), and more frequent healthcare visits (p < 0.001). The normal cognition group had higher rates of alcohol consumption (34.9% vs. 30.0%, p < 0.001) and current smoking (8.4% vs. 7.4%, p < 0.001). The cognitive impairment group was more often at moderate-to-high nutritional risk (40.3% vs. 32.1%, p < 0.001), had a slightly lower BMI (23.49 ± 2.72 vs. 23.69 ± 2.60, p < 0.001), and showed lower independence in ADL (87.8% vs. 94.3%, p < 0.001) and IADL (75.5% vs. 86.0%, p < 0.001). Conversely, the cognitive impairment group had a higher rate of needing assistance with daily activities (13.2% vs. 6.1%, p < 0.001) and functional limitation (24.5% vs. 14.0%, p < 0.001). Compared with the normal cognition group, the cognitive impairment group had higher depression scores (3.98 ± 3.50 vs. 2.80 ± 3.09, p < 0.001), a higher rate of suicidal ideation (1.8% vs. 0.9%, p = 0.001), and a higher rate of AD completion (12.5% vs. 10.7%, p = 0.017). The two groups also differed significantly on three of the five items assessing awareness of dying well—preparing for one’s own dying process, dying in the presence of family or close friends, and dying at home rather than in a hospital or care facility (p = 0.011 to p < 0.001)—whereas no significant differences were found for dying without pain (p = 0.769) or not burdening family or close friends (p = 0.206); item-level details are presented in Table 1.
Table 1Differences in participant characteristics by cognitive impairment status (n = 9,951).
Categories Variable Cognitive impairment group (n = 2,448) Normal cognition group (n = 7,503) t/χ2 p
General characteristics Age 74.67 ± 6.96 73.81 ± 6.67 5.362 <0.001
Sex Male 813 (33.2) 3,011 (40.1) 37.354 <0.001
Spouse Yes 1,321 (54.0) 4,562 (60.8) 37.73 <0.001
Living arrangements Living alone 950 (38.8) 2,473 (33.0) 37.658 <0.001
Married couple 1,229 (50.2) 4,190 (55.8)
Household with children 243 (9.9) 697 (9.3)
Others 26 (1.1) 143 (1.9)
Perceived health status Very healthy 42 (1.7) 198 (2.6) 220.66 <0.001
Healthy 702 (28.7) 3,171 (42.3)
Moderate 885 (36.2) 2,560 (34.1)
Unhealthy 706 (28.8) 1,401 (18.7)
Very unhealthy 113 (4.6) 173 (2.3)
Religion Yes 980 (40.0) 3,091 (41.2) 1.035 0.16
Education level 7.75 ± 4.40 8.40 ± 4.17 −6.361 <0.001
Physical health-related characteristics Number of chronic diseases 2.40 ± 1.64 2.12 ± 1.57 7.64 <0.001
Number of prescription medications 2.15 ± 1.52 1.99 ± 1.61 4.465 <0.001
Use of medical institutions Fx during 1 month 1.72 ± 2.33 1.50 ± 2.18 4.127 <0.001
Drinking alcohol Yes 734 (30.0) 2,616 (34.9) 19.703 <0.001
Smoking Current 180 (7.4) 627 (8.4) 22.488 <0.001
Past 519 (21.2) 1901 (25.3)
Never 1749 (71.4) 4,975 (66.3)
Nutritional status 2.68 ± 2.76 2.07 ± 2.39 9.748 <0.001
Good 1,462 (59.7) 5,095 (67.9) 63.875 <0.001
Moderate risk 697 (28.5) 1815 (24.2)
High risk 289 (11.8) 593 (7.9)
BMI 23.50 ± 2.68 23.69 ± 2.60 −3.129 0.002
Need for assistance in daily activities Yes 323 (13.2) 461 (6.1) 66.954 <0.001
ADL Fully independent 2,149 (87.8) 7,079 (94.3) 117.997 <0.001
IADL Fully independent 1848 (75.5) 6,455 (86.0) 148.447 <0.001
Exercise Days per week 2.15 ± 2.61 2.70 ± 2.71 −9.018 <0.001
Fall experience Yes 193 (7.9) 419 (5.6) 16.910 <0.001
Quality of sleep Very satisfied 80 (3.3) 239 (3.2) 103.94 <0.001
Satisfied 932 (38.1) 3,675 (49.0)
Somewhat 901 (36.8) 2,438 (32.5)
dissatisfied 495 (20.2) 1,065 (14.2)
Not at all 40 (1.6) 86 (1.1)
Disability status Yes 133 (5.4) 271 (3.6) 15.716 <0.001
Functional limitation status Yes 600 (24.5) 1,048 (14.0) 148.45 <0.001
Emotional characteristics Depression Yes, cutoff ≥8 3.98 ± 3.50 2.80 ± 3.09 14.893 <0.001
412 (16.8) 690 (9.2) 109.22 <0.001
Subjective satisfaction (QoL) Very satisfied 25 (1.0) 123 (1.6) 228.79 <0.001
Satisfied 679 (27.7) 3,156 (42.1)
Neutral 1,406 (57.4) 3,704 (49.4)
Dissatisfied 319 (13.0) 498 (6.6)
Not at all 19 (0.8) 22 (0.3)
Suicide intention Yes 43 (1.8) 70 (0.9) 11.152 0.001
Attitude toward death AD completion Yes 305 (12.5) 804 (10.7) 5.666 0.017
Attitude toward medical care for life prolongation Very agree 34 (1.4) 94 (1.3) 30.552 <0.001
Agree 101 (4.1) 333 (4.4)
Disagree 834 (34.1) 2,882 (38.4)
Very disagree 1,178 (48.1) 3,516 (46.9)
I do not know 301 (12.3) 678 (9.0)
Take death preparation education Yes 89 (3.6) 208 (2.8) 4.752 0.029
Awareness of dying well Composite 9.83 ± 2.93 9.78 ± 2.90 0.779 0.436
(1) Prepare for dying by oneself Very important 860 (35.1) 2,490 (33.2) 13.005 0.011
Important 1,245 (50.9) 3,895 (51.9)
Neutral 273 (11.2) 892 (11.9)
Not important 56 (2.3) 210 (2.8)
Not at all 14 (0.6) 16 (0.2)
(2) Dying without pain Very important 1,252 (51.1) 3,740 (49.8) 1.817 0.769
Important 846 (34.6) 2,690 (35.9)
neutral 283 (11.6) 881 (11.7)
Not important 61 (2.5) 173 (2.3)
Not at all 6 (0.2) 19 (0.3)
(3) Dying in the presence of family or close friends Very important 658 (26.9) 2,225 (29.7) 15.266 0.004
Important 1,194 (48.8) 3,492 (46.5)
Neutral 476 (19.4) 1,408 (18.8)
Not important 105 (4.3) 359 (4.8)
Not at all 15 (0.6) 19 (0.3)
(4) Not burdening family or close friends Very important 1,017 (41.5) 3,236 (43.1) 5.913 0.206
Important 1,058 (43.2) 3,094 (41.2)
Neutral 301 (12.3) 990 (13.2)
Not important 66 (2.7) 170 (2.3)
Not at all 6 (0.2) 13 (0.2)
(5) Dying at home rather than in a hospital or care facility Very important 270 (11.0) 955 (12.7) 19.608 <0.001
Important 962 (39.3) 3,074 (41.0)
Neutral 866 (35.4) 2,370 (31.6)
Not important 296 (12.1) 869 (11.6)
Not at all 54 (2.2) 235 (3.1)
fx, frequency; BMI, body mass index; ADL, activities of daily living; IADL, instrumental activities of daily living; ADs, advance directives; QoL, quality of life.
3.2.Differences in participant characteristics by AD completion among older adults with cognitive impairments
Within the cognitive impairment group (n = 2,448), participants who completed an AD differed significantly from those who had not across several characteristics (Table 2). Participants who completed an AD were significantly younger (p < 0.001), more likely to have a spouse (p = 0.032), more likely to have a religious affiliation (p < 0.001), and had higher educational attainment (p < 0.001). Those who completed an AD also had more chronic conditions and greater prescription medication use (p < 0.001), and more frequent healthcare visits (p = 0.037), as well as higher rates of past or current smoking (p < 0.001) and more frequent exercise (p < 0.001). They were more likely to have depression (p < 0.001) and suicidal ideation (p = 0.002), had higher participation in death preparation education (p < 0.001), and had lower scores on the dying-well awareness scale (p < 0.001).
Table 2Differences in participant characteristics by AD completion in older adults with cognitive impairment (n = 2,448).
Categories Variable Older adults who completed ADs (n = 305) Older adults who did not completed ADs (n = 2,143) t/χ2 p
General characteristics Age 72.91 ± 5.99 74.92 ± 7.05 −5.367 <0.001
Sex Male 113 (37.0) 700 (32.7) 2.314 0.135
Spouse Yes 182 (59.7) 1,139 (53.1) 4.572 0.032
Living arrangements Living alone 112 (36.7) 838 (39.1) 19.081 <0.001
Married couple 179 (58.7) 1,050 (49.0)
Household with children 14 (4.6) 229 (10.7)
Others 0(0.0) 26(1.2)
Perceived health status Very healthy 7 (2.3) 35 (1.6) 1.759 0.78
Healthy 83 (27.2) 619 (28.9)
Moderate 106 (34.8) 779 (36.4)
Somewhat unhealthy 93 (30.5) 613 (28.6)
Unhealthy 16 (5.2) 97 (4.5)
Religion (Yes) Yes 168 (55.1) 812 (37.9) 32.869 <0.001
Education level 9.47 ± 4.35 7.51 ± 4.35 7.378 <0.001
Physical health-related characteristics Number of chronic diseases 2.79 ± 1.41 2.35 ± 1.67 5.025 <0.001
Number of prescription medications 2.57 ± 1.41 2.10 ± 1.53 5.084 <0.001
Use of medical institutions Fx during 1 month 1.98 ± 2.08 1.68 ± 2.36 2.089 0.037
Drinking alcohol Yes 103 (33.8) 631 (29.4) 27.418 <0.001
Smoking Current 21 (6.9) 159 (7.4) 19.392 <0.001
Past 94 (30.8) 425 (19.8)
Never 190 (62.3) 1,559 (72.7)
Nutritional status 3.62 ± 2.73 2.54 ± 2.74 6.42 <0.001
BMI 23.54 ± 2.67 23.48 ± 2.73 0.37 0.356
Need for assistance in daily activities Yes 32 (10.5) 291 (13.6) 2.477 0.29
Exercise Days per week 3.11 ± 2.69 2.01 ± 2.57 6.708 <0.001
Fall experience Yes 33 (10.8) 160 (7.5) 4.135 0.042
ADL Fully independent 272 (89.2) 1877 (87.8) 0.632 0.427
IADL Fully independent 240 (78.7) 1,608 (75.0) 1.926 0.165
Disability status Yes 18 (5.9) 115 (5.4) 0.149 0.7
Functional limitation status Yes 65 (21.3) 535 (25.0) 1.926 0.165
Emotional characteristics Depression Yes 76 (24.9) 336 (15.7) 16.282 <0.001
Suicide intention Yes 12 (3.9) 31 (1.4) 9.576 0.002
Attitude toward death Take death preparation education Yes 30 (9.8) 59 (2.8) 38.233 <0.001
Awareness of dying well Composite 8.94 ± 2.48 9.96 ± 2.97 −6.533 <0.001
fx, frequency; BMI, body mass index; ADL, activities of daily living; IADL, instrumental activities of daily living.
3.3.Differences in participant characteristics by AD completion among older adults without cognitive impairments
In the normal cognition group (n = 7,503), participants who completed an AD also differed significantly from those who had not (Table 3). They were significantly older (p = 0.005), more likely to live alone (p = 0.032), more likely to have a religious affiliation (p < 0.001), and had more chronic diseases and prescription medications (p < 0.001) than those who had not. Rates of alcohol consumption did not differ significantly between the two groups (p = 0.465), although those who completed an AD had lower rates of smoking (p = 0.002). As in the cognitive impairment group, those who completed an AD reported poorer nutritional status (p < 0.001), lower BMI (p = 0.026), more frequent exercise (p < 0.001), more falls (p < 0.001), and higher depression (p = 0.007). They also had higher participation in death preparation education (p < 0.001) and lower dying-well awareness scores (p < 0.001).
Table 3Differences in participant characteristics by AD completion in older adults without cognitive impairment (n = 7,503).
Categories Variable Older adults who completed ADs (n = 804) Older adults who did not completed ADs (n = 6,699) t/χ2 p
General characteristics Age 74.44 ± 6.41 73.74 ± 6.70 2.838 0.005
Sex Male 313 (38.9) 2,698 (40.3) 0.54 0.47
Spouse Yes 496 (61.7) 4,066 (60.7) 0.299 0.593
Living arrangements Living alone 271 (33.7) 2,202 (32.9) 8.839 0.032
Married couple 468 (58.2) 3,722 (55.6)
Household with children 54 (6.7) 643 (9.6)
Others 11 (1.4) 132 (2.0)
Perceived health status Very healthy 23 (2.9) 175 (2.6) 8.042 0.09
Healthy 331 (41.2) 2,840 (42.4)
Moderate 266 (33.1) 2,294 (34.2)
Somewhat unhealthy 173 (21.5) 1,228 (18.3)
Unhealthy 11 (1.4) 162 (2.4)
Religion Yes 408 (50.7) 2,683 (40.1) 33.898 <0.001
Education level Years 8.62 ± 4.31 8.37 ± 4.16 1.562 0.119
Physical health-related characteristics Number of chronic diseases 2.40 ± 1.58 2.09 ± 1.57 5.393 <0.001
Number of prescription medications 2.18 ± 1.52 1.97 ± 1.62 3.51 <0.001
Use of medical institutions Fx during 1 month 1.62 ± 2.26 1.49 ± 2.17 1.676 0.094
Drinking alcohol Yes 271 (33.7) 2,345 (35.0) 0.533 0.465
Smoking Current 42 (5.2) 585 (8.7) 12.695 0.002
Past 223 (27.7) 1,678 (25.0)
Never 539 (67.0) 4,436 (66.2)
Nutritional status 2.40 ± 2.51 2.03 ± 2.37 3.967 <0.001
BMI 23.50 ± 2.59 23.72 ± 2.60 −2.234 0.026
Need for assistance in ADL Yes 54 (6.7) 407 (6.1) 0.511 0.475
Exercise Days per week 3.45 ± 2.75 2.61 ± 2.69 8.397 <0.001
Fall experience Yes 74 (9.2) 345 (5.2) 22.375 <0.001
ADL Fully independent 745 (92.7) 6,334 (94.6) 4.808 0.02
IADL Fully independent 679 (84.5) 5,776 (86.2) 1.87 0.178
Disability status Yes 39 (4.9) 232 (3.5) 3.97 0.057
Functional limitation status Yes 125 (15.5) 923 (13.8) 1.87 0.178
Emotional characteristics Depression Yes 95 (11.8) 595 (8.9) 7.4 0.007
Suicide intention Yes 8 (1.0) 62 (0.9) 0.038 0.845
Attitude toward death Takes death preparation education Yes 85 (10.6) 123 (1.8) 203.26 <0.001
Awareness of dying well 9.10 ± 2.54 9.86 ± 2.93 7.86 <0.001
fx, frequency; BMI, body mass index; ADL, activities of daily living; IADL, instrumental activities of daily living.
3.4.Factors associated with AD completion in those with cognitive impairments
Multivariable logistic regression analysis was conducted to identify factors associated with AD completion in the cognitive impairment group (Table 4). AD completion was significantly associated with younger age (adjusted odds ratio [aOR] = 0.965, 95% CI = 0.941–0.989), female sex (aOR for male sex = 0.399, 95% CI = 0.229–0.695), religious affiliation (aOR = 1.748, 95% CI = 1.336–2.286), higher education (aOR = 1.099, 95% CI = 1.057–1.142), greater prescription medication use (aOR = 1.202, 95% CI = 1.053–1.371), never having smoked relative to current smoking (aOR = 0.457, 95% CI = 0.231–0.904), poorer nutritional status (aOR = 1.146, 95% CI = 1.089–1.206), higher exercise frequency (aOR = 1.162, 95% CI = 1.106–1.220), higher ADL independence (aOR = 1.715, 95% CI = 1.082–2.719), higher depressive symptom scores (aOR = 1.057, 95% CI = 1.010–1.105), participation in death preparation education (aOR = 3.300, 95% CI = 1.991–5.470), and higher awareness of dying well, reflected by a lower total score (aOR = 0.866, 95% CI = 0.820–0.915).
Table 4Factors associated with AD completion in older individuals with cognitive impairments (n = 2,448).
Independent variable B SE p Adjusted OR 95% CI
Lower Upper
Age −0.039 0.012 0.002 0.962 0.939 0.985
Sex Male −0.919 0.283 0.001 0.399 0.229 0.695
Spouse Yes 0.205 0.144 0.155 1.228 0.926 1.628
Religion Yes 0.558 0.137 <0.001 1.748 1.336 2.286
Education level 0.094 0.020 <0.001 1.099 1.057 1.142
Number of chronic diseases −0.022 0.067 0.746 0.978 0.858 1.116
Number of prescribed medications 0.184 0.067 0.006 1.202 1.053 1.371
Number of medical visits in the past month 0.032 0.028 0.262 1.032 0.976 1.092
Alcohol use Yes 0.098 0.159 0.539 1.103 0.807 1.506
Smoking (current) Past 0.463 0.281 0.100 1.589 0.915 2.758
Never −0.782 0.348 0.024 0.457 0.231 0.904
Nutritional status 0.136 0.026 <0.001 1.146 1.089 1.206
Exercise (fx) During 1 week 0.150 0.025 <0.001 1.162 1.106 1.220
Fully independent ADL Yes 0.540 0.235 0.022 1.715 1.082 2.719
Depression Yes 0.055 0.023 0.016 1.057 1.010 1.105
Death preparation education Yes 1.194 0.258 <0.001 3.300 1.991 5.470
Suicide intention Yes 0.541 0.416 0.193 1.718 0.760 3.882
Awareness of dying well −0.143 0.028 <0.001 0.866 0.820 0.915
Nagelkerke R2 = 0.211
SE, standard error; OR, odds ratio; CI, confidence interval; fx, frequency; ADL, activities of daily living. Awareness of dying well was scored such that lower total scores indicate greater perceived importance of dying well; thus, an aOR < 1 indicates that higher awareness of dying well was associated with a greater likelihood of AD completion.
Hosmer & Lemeshow test 15.204 (p = 0.055).
3.5.Factors associated with AD completion in those without cognitive impairments
Multivariable logistic regression analysis was conducted to identify factors associated with AD completion in the normal cognition group (Table 5). Unlike in the cognitive impairment group, sex was not significantly associated with AD completion in this group (aOR = 1.010, 95% CI = 0.847–1.204, p = 0.913). AD completion was significantly associated with religious affiliation (aOR = 1.448, 95% CI = 1.242–1.689), a higher number of chronic diseases (aOR = 1.151, 95% CI = 1.063–1.246), and poorer nutritional status (aOR = 1.050, 95% CI = 1.014–1.087). Higher exercise frequency (aOR = 1.118, 95% CI = 1.087–1.150) and a history of falls within the past year (aOR = 1.647, 95% CI = 1.245–2.178) were also significant physical health–related correlates, whereas BMI showed only a borderline, non-significant association (aOR = 0.971, 95% CI = 0.942–1.001, p = 0.054). Participation in death preparation education (aOR = 6.198, 95% CI = 4.599–8.354) and higher awareness of dying well, reflected by a lower total score (aOR = 0.904, 95% CI = 0.878–0.931) were also significantly associated with AD completion in this group.
Table 5Factors associated with AD completion in older individuals without cognitive impairments (n = 7,503).
Independent variable B SE p Adjusted OR 95% CI
Lower Upper
Age 0.012 0.006 0.051 1.012 1.000 1.025
Sex Male 0.010 0.090 0.913 1.010 0.847 1.204
Living alone Yes −0.068 0.088 0.439 0.934 0.787 1.110
Religion Yes 0.370 0.078 <0.001 1.448 1.242 1.689
Number of chronic diseases 0.141 0.040 <0.001 1.151 1.063 1.246
Number of prescribed medications −0.070 0.040 0.081 0.932 0.861 1.009
Alcohol use Yes 0.072 0.090 0.428 1.074 0.900 1.283
Nutritional status 0.049 0.018 0.006 1.050 1.014 1.087
BMI −0.030 0.015 0.054 0.971 0.942 1.001
Exercise (fx) During 1 week 0.112 0.014 <0.001 1.118 1.087 1.150
Fall experience during the past 1 year Yes 0.499 0.143 <0.001 1.647 1.245 2.178
Death preparation education (yes) Yes 1.824 0.152 <0.001 6.198 4.599 8.354
Independent ADL Yes 0.086 0.167 0.607 1.090 0.785 1.512
Awareness of dying well −0.101 0.015 <0.001 0.904 0.878 0.931
Nagelkerke R2 = 0.087
SE, standard error; OR, odds ratio; CI, confidence interval; BMI, body mass index; fx, frequency; ADL, activities of daily living. Awareness of dying well was scored such that lower total scores indicate greater perceived importance of dying well; thus, an aOR < 1 indicates that higher awareness of dying well was associated with a greater likelihood of AD completion.
Hosmer & Lemeshow test 7.091 (p = 0.527).
4.Discussion
In this study, cognitive impairment was defined as a score on the K-MMSE-DS that was 1.5 standard deviations or more below the mean for the participant’s respective age and education group (16). As a result, 24.6% of the entire sample was determined to have cognitive impairment. With medical advancements improving the possibility of having a prolonged life, people increasingly seek the right to self-determination regarding both life and death (24). When caring for end-of-life patients, ACP helps clinicians and families honor and implement the patient’s preferences (24). As South Korea has transitioned into a super-aged society due to the rapidly increasing older adult population, AD completion is becoming increasingly important (25). However, according to this study, only 12.5% of older adults in the cognitively impaired group and 10.7% in the cognitively normal group had completed an AD.
The results of this study demonstrate that end-of-life education is strongly associated with higher odds of AD completion, regardless of cognitive function levels. Notably, in the normal cognition group, the aOR associated with death preparation education was approximately twice as high as that observed in the cognitive impairment group. Furthermore, a staggering 90.3% of individuals have never independently explored life-sustaining treatment systems or AD completion (21). In 2025, the registration rate for AD completion among South Korean adults aged ≥19 years was 6.8% (26). Although 82.2% of participants in the cognitive impairment group and 85.3% in the normal cognition group in this study did not agree with medical care for life prolongation, the overall rate of AD completion remained low. This discrepancy suggests that while negative attitudes toward non-beneficial life extension are widespread, personal willingness does not automatically translate into formal documentation. This low completion rate may reflect the fact that factors associated with AD completion are fundamentally multi-layered—encompassing not only sociodemographic characteristics, but also clinical conditions, health behaviors, psychological states, and structural accessibility (14, 15). Consequently, addressing this low completion rate requires going beyond generalized educational programs to consider these complex, multi-layered determinants, which may operate differently depending on cognitive status. Furthermore, religious affiliation, poorer nutritional status, higher exercise frequency, and higher awareness of the significance of dying well were identified as significantly associated with AD completion in both the groups. However, in a previous study, 75.1% of community-dwelling older adults have never been introduced to the formal concept of “dying well” (27). While greater perceived importance of dying well was associated with AD completion, the exact underlying mechanism remains unclear. One possible explanation is that prioritizing end-of-life values may foster general receptivity toward advance care planning; however, systematic exposure to these concepts remains low, highlighting an area that warrants further empirical investigation.
Several additional factors were significantly associated with AD completion within the cognitive impairment group, including a complex interplay of sociodemographic factors (such as age and educational attainment), number of prescription medications used, independence in ADL, and depressive symptoms. A particularly notable group-specific finding was sex: female sex was significantly associated with a higher likelihood of AD completion in the cognitive impairment group (aOR for male sex = 0.399, 95% CI = 0.229–0.695), whereas sex showed no significant association with AD completion in the normal cognition group (aOR = 1.010, 95% CI = 0.847–1.204). Prior research on sex differences in AD completion has yielded mixed and context-dependent findings: a study of community-dwelling Korean older adults reported higher AD completion among men than women, a pattern the authors themselves noted was inconsistent with earlier Korean studies reporting higher completion among women (15). Similarly, a dyadic study of older American couples found that the association between cognitive function and AD completion differed according to spousal sex, suggesting that gender and cognitive status may jointly shape advance care planning behavior (28). Although the present study did not formally test for a statistical interaction between sex and cognitive status, our finding that sex was significantly associated with AD completion only in the cognitively impaired group—but not in the cognitively normal group—adds to this body of evidence suggesting that the role of sex in AD completion may not be uniform across populations and may depend on cognitive status, warranting further investigation through studies formally designed to test sex-by-cognition interactions. Regarding depression, the association with AD completion in the cognitive impairment group was statistically significant but modest in magnitude (aOR = 1.057, 95% CI = 1.010–1.105). This finding is broadly consistent with prior literature indicating that depression is more prevalent among individuals with cognitive impairment, reaching up to 32% (29), and is recognized as a factor associated with perceptions of mortality and death anxiety in older adults (29, 30). However, given the cross-sectional design and modest effect size observed here, this association should be interpreted as preliminary, and the underlying mechanism cannot be established from the present data.
In the cognitively normal group, a higher number of chronic diseases and a history of falls were significantly associated with increased odds of AD completion. A history of falls is associated with fear of falling among older adults (31), which in turn negatively affects quality of life (32). For cognitively intact older adults, these physical health factors may highlight bodily decline and prompt earlier engagement with advance directives; however, this interpretation is a hypothesis that was not directly tested in the present cross-sectional study and warrants further investigation, for example through longitudinal designs. In contrast, these physical factors showed no significant association with AD completion among older adults with cognitive impairment, indicating that key determinants of advance care planning may vary according to cognitive status.
Strategies to encourage AD completion may need to be tailored to individuals’ level of cognitive function. Specifically, advance directive education and communication for older adults with cognitive impairment may benefit from sensitively accounting for sociodemographic factors, functional status, depressive symptoms, and—based on the present findings—sex. Conversely, educational approaches for cognitively intact individuals might leverage current health status and intact functional capacities. Overall, these findings highlight the potential value of tailored death education programs and suggest that communication strategies to encourage AD completion warrant further refinement and empirical testing.
4.1.Strengths and limitations
This study has notable strengths. It used data from a large, nationally representative sample of community-dwelling older Koreans, which allowed factor associated with AD completion to be examined separately among individuals with and without cognitive impairment. To our knowledge, this is among the first studies to explore these factors by cognitive status using a nationwide dataset, and the findings may inform future efforts to tailor advance care planning approaches to individuals’ cognitive status.
However, several limitations warrant consideration. First, because of the cross-sectional nature of the dataset, causal relationships between the identified factors and AD completion could not be established. Longitudinal studies are required to clarify the temporal relationships and potential causal pathways underlying AD completion, particularly in relation to cognitive status. Second, as a secondary data analysis, the findings were restricted to variables available in the original dataset. Other potentially relevant factors—such as cognitive and psychosocial processes and individuals’ knowledge or awareness of ADs—were not captured. Given that awareness of ADs is likely to influence completion, the lack of this measure may have resulted in residual confounding. Future studies incorporating theory-driven variables, including AD-specific knowledge, are warranted. Third, only community-dwelling older adults who were able to provide self-reported data were included. Participants classified as cognitively impaired were therefore likely to have relatively mild impairment, and the findings may not be generalizable to older adults with more severe cognitive impairment or to those in institutional or hospital settings. Future studies should include more diverse populations across different care settings. Fourth, cognitive impairment was determined using a cut-off score on a cognitive screening instrument (K-MMSE-DS) rather than a comprehensive clinical diagnostic assessment. Participants classified as cognitively impaired may not have had a clinically confirmed diagnosis of dementia or mild cognitive impairment, and such misclassification may have attenuated differences between groups. More robust diagnostic assessments should be used in future research. Fifth, unweighted multivariable logistic regression was used rather than complex survey design adjustments. While suitable for assessing within-group covariate associations, omitting sampling weights may affect standard error precision and limit the generalizability of population-level descriptive prevalence estimates. Finally, because factors associated with AD completion were examined using separate regression models for each cognitive group, these within-group findings do not statistically confirm between-group differences. Future research should incorporate formal interaction analyses to directly test for moderation by cognitive status.
5.Conclusion
This study demonstrated that different patterns of factors associated with AD completion were observed between cognitively impaired and cognitively normal older adults, although several factors were common to both the groups. Religious affiliation, nutritional status, exercise frequency, participation in death preparation education, and higher awareness of dying well were significantly associated with AD completion in both cohorts.
However, several factors were uniquely associated with AD completion within each cognitive group. In the cognitive impairment group, younger age, female sex, higher educational attainment, greater prescription medication use, never having smoked, higher ADL independence, and higher depressive symptom scores were significantly associated with AD completion. In the normal cognition group, a greater number of chronic diseases and a recent history of falls were uniquely associated with AD completion; notably, sex was not significantly associated with AD completion in this group, in contrast to the cognitive impairment group.
These findings suggest that the factors associated with AD completion may differ according to cognitive status, indicating that strategies to promote AD completion may need to be tailored accordingly. For older adults with cognitive impairment, approaches that address functional independence, psychosocial factors such as depressive symptoms may warrant particular attention. In addition, men with cognitive impairment may warrant closer attention in advance care planning discussions, although the effectiveness of sex-tailored approaches requires further investigation. For cognitively normal older adults, the association between falls and chronic disease burden with AD completion suggests that routine health encounters—such as those following a fall or a new diagnosis—may represent opportunities to introduce AD-related discussions, although this hypothesis was not directly tested in this study and warrants further investigation.
Acknowledgments
We would like to thank the Korea Institute for Health and Social Affairs for their help in collecting the data, as well as the older adults who participated in this study.
Funding Statement
The author(s) declared that financial support was received for this work and/or its publication. This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (Ministry of Science and ICT, MSIT) (RS-2025-21722979, Project Title: Development and Evaluation of a Dementia Literacy Tailored Advanced Care Planning Intervention for Dignity Care of Older Adults with Mild Cognitive Impairment, 100%).
Footnotes
Edited by: Ping Guo, University of Birmingham, United Kingdom
Reviewed by: Yupawadee Kantabanlang, University of Michigan, United States
Miki Fujimoto, King’s College London, United Kingdom
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Ethics statement
The studies involving humans were approved by the Institutional Review Board of Gachon University. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.
Author contributions
SC: Conceptualization, Methodology, Software, Supervision, Validation, Writing – original draft, Writing – review & editing. HK: Conceptualization, Funding acquisition, Methodology, Project administration, Validation, Visualization, Writing – original draft, Writing – review & editing.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that Generative AI was used in the creation of this manuscript. Generative AI was used only to improve the English language, grammar, and readability of the manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1.Park HY, Kim MS, Yoo SH, Lee J, Song IG, Jeon SY, et al. For the universal right to access quality end-of-life care in Korea: broadening our perspective after the 2018 life-sustaining treatment decisions. J Korean Med Sci. (2024) 39:e123. doi: 10.3346/jkms.2024.39.e123
2.Central Dementia Center . Dementia Today. Seoul: Central Dementia Center; (2023).
3.Alzheimer’s Disease International . World Alzheimer Report 2021: Journey through the Diagnosis of Dementia. London: ADI; (2021).
4.Rietjens JA, Korfage IJ, Dunleavy L, Preston NJ, Jabbarian LJ, Johnstone M, et al. Definition and recommendations for advance care planning: an international consensus supported by the European Association for Palliative Care. Lancet Oncol. (2017) 18:e543–51. doi: 10.1016/S1470-2045(17)30582-X
5.Silveira MJ, Kim SYH, Langa KM. Advance directives and outcomes of surrogate decision making before death. N Engl J Med. (2010) 362:1211–8. doi: 10.1056/NEJMsa0907901,
6.Ministry of Health and Welfare, National Agency for Management of Life-Sustaining Treatment . Act on Hospice and Palliative Care and Decisions on Life-Sustaining Treatment for Patients at the End of Life. Act No. 14013 (Enacted Feb 3, 2016, Effective Feb 4, 2018).
Sejong: Ministry of Health and Welfare; (2018).
7.Ha JH, Lee C. The need for advance care planning for individuals with dementia in South Korea. Health Soc Welf Rev. (2019) 39:41–72. doi: 10.15709/hswr.2019.39.3.41
8.Mack JW, Weeks JC, Wright AA, Block SD, Prigerson HG. End-of-life discussions, goal attainment, and distress at the end of life: predictors and outcomes of receipt of care consistent with preferences. J Clin Oncol. (2010) 28:1203–8. doi: 10.1200/JCO.2009.25.4672,
9.Dixon J, Karagiannidou M, Knapp M. The effectiveness of advance care planning in improving end-of-life outcomes for people with dementia and their carers: a systematic review and critical discussion. J Pain Symptom Manag. (2017) 55:132–150.e1. doi: 10.1016/j.jpainsymman.2017.04.009,
10.Carr D, Luth EA. Advance care planning: contemporary issues and future directions. Innov Aging. (2017) 1:igx012. doi: 10.1093/geroni/igx012
11.Kim B, Choi J, Lee I. Factors associated with advance directives documentation: a nationwide cross-sectional survey of older adults in Korea. Int J Environ Res Public Health. (2022) 19:3771. doi: 10.3390/ijerph19073771,
12.Nicholas LH, Halpern SD, Weir DR, Baum MY, Nolan M, Gallo J, et al. Decision making for patients with severe dementia versus normal cognition near the end of life. Innov Aging. (2023) 7:igad081. doi: 10.1093/geroni/igad081,
13.Owsley KM, Langa KM, Macis M, Nicholas LH. Treatment preferences among adults with normal cognition and cognitive impairment. J Am Geriatr Soc. (2022) 70:3390–401. doi: 10.1111/jgs.18032,
14.Lin J, Yang M, Xu R, Deng Y, Huang T. Influencing factors and profile differences in death attitudes of advance care planning readiness among patients with chronic heart failure: a latent profile analysis. Front Public Health. (2026) 14:1825995. doi: 10.3389/fpubh.2026.1825995,
15.Choi S, Ko H. Factors affecting advance directives completion among older adults in Korea. Front Public Health. (2024) 12:1329916. doi: 10.3389/fpubh.2024.1329916,
16.Korea Institute for Health and Social Affairs (2024) Survey of Korea older adults 2024. Available online at: https://mods.go.kr/board.es?mid=a10301010000&bid=10820&act=view&list_no=432917 (Accessed 5 August, 2025)
17.Kim TH, Jhoo JH, Park JH, Kim JL, Ryu SH, Moon SW, et al. Korean version of mini mental status examination for dementia screening and its' short form. Psychiatry Investig. (2010) 7:102–8. doi: 10.4306/pi.2010.7.2.102,
18.Boyle MA, Morris DH. Community Nutrition in Action: An Entrepreneurial Approach. Belmont, CA: Wadsworth Press; (2003).
19.Jae NB, Maeng JC. Development of the Korean version of the geriatric depression scale and its short form among elderly psychiatric patients. J Psychosom Res. (2004) 57:297–305. doi: 10.1016/j.jpsychores.2004.01.004,
20.DuMouchel WH, Duncan GJ. Using sample survey weights in multiple regression analyses of stratified samples. J Am Stat Assoc. (1983) 78:535–43. doi: 10.1080/01621459.1983.10478006
21.Winship C, Radbill L. Sampling weights and regression analysis. Sociol Methods Res. (1994) 23:230–57. doi: 10.1177/0049124194023002004
22.Korn EL, Graubard BI. Epidemiologic studies utilizing surveys: accounting for the sampling design. Am J Public Health. (1991) 81:1166–73. doi: 10.2105/ajph.81.9.1166,
23.Gelman A. Struggles with survey weighting and regression modeling. Stat Sci. (2007) 22:153–64. doi: 10.1214/088342306000000691
24.Heppner HJ, Hag H. Vorsorgeplanung am lebensende aus geriatrischer sicht [advanced care planning and end of life decisions - the geriatricians’ view]. Dtsch Med Wochenschr. (2023) 148:1360–5. doi: 10.1055/a-1825-5874
25.Kim S-H, Lee J-H. A study on awareness of life-sustaining treatment decision system and advance directives, and comparative analysis by socio-demographic factors. J Korean Soc Sch Community Health Educ. (2025) 26:31–48. doi: 10.35133/kssche.20250228.03
26.Seo JH. (2025) Bank of Korea report gains renewed attention following president lee’s order to review incentives for withdrawing life-sustaining treatment. Bravo My Life. Available online at: https://bravo.etoday.co.kr/view/atc_view/18092 (Accessed July 13, 2026)
27.Ryu EJ, Choi SE. Relationships among perceptions of dying well, attitudes toward advance directives, and preferences for advance directives among elderly living alone. Hanguk Hosupisu Wanhwa Uiryo Hakhoe Chi. (2020) 23:241–51. doi: 10.14475/kjhpc.2020.23.4.241,
28.Lu P, Kong D, Lee J, Shelley M. Advance directive completion among older American couples: a dyadic perspective on the role of cognitive function and other factors. J Gerontol B Psychol Sci Soc Sci. (2022) 77:1144–53. doi: 10.1093/geronb/gbab210,
29.Ismail Z, Elbayoumi H, Fischer CE, Hogan DB, Milikin CP, Schweizer T, et al. Prevalence of depression in patients with mild cognitive impairment: a systematic review and Meta-analysis. JAMA Psychiatry. (2017) 74:58–67. doi: 10.1001/jamapsychiatry.2016.3162,
30.Kulakçı Altıntaş H, Ercan Şahin N. Attitudes to aging, depression, and death anxiety among community-dwelling older adults: a cross-sectional correlational study. BMC Geriatr. (2025) 25:454. doi: 10.1186/s12877-025-05839-3,
31.Tsai C-Y, Chen C, Tung T-H, Li Y-T, Chen W-C. The relationship between falling and fear of falling among community-dwelling elderly. Medicine (Baltimore). (2021) 100:e26492. doi: 10.1097/MD.0000000000026492,
32.Schoene D. A systematic review on the influence of fear of falling on quality of life in older people: Is there a role for falls?
Clin Interv Aging. (2019) 14:701–19. doi: 10.2147/CIA.S197857,
Associated Data
Data Availability Statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Come citare questo studio
Seunghye Choi, Hana Ko. (2026). Factors associated with advance directive completion among older adults with and without cognitive impairment: a nationwide cross-sectional study.. Frontiers in public health. DOI: 10.3389/fpubh.2026.1957140
Consulta la pubblicazione originale ↗