Abstract
Objectives
Medication non-adherence among older adults with non-communicable diseases (NCDs) remains prevalent worldwide, which causes hospitalization and mortality. Our study aimed to examine the association of medication non-adherence with level of overall intrinsic capacity (IC), pattern of IC, and specific IC component among older adults with NCDs.
Methods
A cross-sectional questionnaire-based survey of 1268 older adults aged 60 years and above was conducted in 2022 in southern Taiwan. Among them, 894 suffered from 1 more NCD were included in this study. The Integrated Care for Older People Screening Tool for Taiwanese and the Adherence to Refills and Medication Scale were used to assess IC and medication non-adherence, respectively. Latent class analysis (LCA) was used to identify patterns of IC impairment, and binary logistic regression was used to assess the association between medication non-adherence and IC.
Results
Older adults in the moderate (score: 1–2) or low (score≧3) overall IC groups were more likely to experience medication non-adherence (moderate: adjusted odds ratio (aOR) 1.57 [95% CI: 1.05–2.36]; low: 2.26 [1.40–3.67]). The “physical and nutritional impairments accompanied by depressive symptoms” group was associated with statistically higher odds of medication non-adherence (aOR 1.66 [1.01–2.73]). Older adults with cognitive impairment, hearing loss, or depressive symptoms showed greater likelihood of medication non-adherence (cognitive impairment: aOR 1.53 [1.03–2.27]; hearing loss: aOR 1.57 [1.03–2.37]; depressive symptoms: aOR 1.81 [1.17–2.80]).
Conclusions
Intervention for improving medication non-adherence among older adults with NCDs should consider IC.
Keywords: Older adults, Intrinsic capacity, Medication adherence
1. Introduction
Medication non-adherence among older adults with non-communicable diseases (NCDs) remains prevalent, ranging from 22% to 61% worldwide [[1], [2], [3], [4], [5], [6]]. Medication non-adherence could lead to all-cause hospitalization and mortality in older people [7] and a substantial cost burden on healthcare systems [8]. Prior research underscored the importance of medication self-management support and patient-centered care in improving medication adherence among older adults with NCDs [[9], [10], [11]].
Medication self-management and adherence are related to functional ability. Prior research found that cognitive impairment was associated with poor medication self-management [12], difficulty taking medicines [13], and poor medication adherence [14]. Visual impairment was related to difficulty taking medicines [13]. Older adults with depressive symptoms were 4.16 times more likely to have medication non-adherence [6]. The development and maintenance of functional ability have been identified as the first priority for most older people to achieve healthy ageing [15].
Functional ability is influenced by intrinsic capacity (IC), the environments where people live, and the interactions between these elements [15]. IC is defined by the World Health Organization (WHO) as the composite of all the physical and mental capacities an individual can draw on [15]. WHO further proposed Integrated Care for Older People (ICOPE) and highlighted the importance of comprehensive assessments of six intrinsic capacities, including cognitive ability, locomotion, vitality, vision, hearing, and psychosocial, before care planning [16]. A prior study based on ICOPE in Taiwan revealed that compared to the robust group, the “visual impairment” group, the “physio-cognitive decline with sensory impairment” group, the “depression with cognitive impairment” group, and the “impairments in all domains” group were more likely to occur excess polypharmacy, potentially inappropriate medications, and adverse drug reactions [17]. However, little is known about the potential influences of IC levels, cluster of IC impairment, and specific IC impairments on medication non-adherence among older adults with chronic diseases in light of comprehensive assessments of the six ICs globally. Previous studies on the determinants of medication non-adherence regarding functional capacity focused only on one or two ICs [6,[12], [13], [14]].
IC shows potential for improving medication non-adherence among older adults. However, determinants for medication non-adherence based on comprehensive assessments of the six ICs remain unknown. Therefore, this study aimed to investigate the associations of IC levels and patterns as well as specific IC component on medication non-adherence among older adults with NCDs. This study enriches the knowledge of how the IC levels and patterns, as well as specific IC components, were associated with medication non-adherence among older adults with NCDs. This study might further contribute to developing IC based strategies that can improve medication non-adherence and extend the clinical and policy implications of comprehensive IC assessments.
2. Methods
2.1. Data collection and study population
A face-to-face cross-sectional questionnaire survey was conducted by well-trained interviewers from April to November 2022 among older people in southern Taiwan. We applied convenient sampling to recruit older people from communities in Tainan City, outpatient clinics, or inpatient wards at National Cheng Kung University Hospital (NCKUH). People aged 60 years or older who could communicate using Mandarin Chinese or Taiwanese and provided consent for participation were eligible to participate in this study, and 1268 participated in this survey. Those who did not have NCDs and did not complete measurements of six IC domains and scales of the Adherence to Refills and Medications Scale (ARMS) were excluded. The above-mentioned non-communicable diseases include hypertension, diabetes mellitus, cardiovascular disease, cerebrovascular disease, lung disease, kidney disease, liver disease, bone and joint problems, emotional problems, and cancers. Finally, 894 older people were included in this study. This study was approved by the Institutional Review Board of NCKUH (IRB No.: A-ER-110-249).
2.2. Medication non-adherence assessment
A dependent variable in this study is medication non-adherence, measured by the ARMS, which was developed to evaluate medication adherence suitable for use across literacy levels among patients with chronic diseases [18]. The 12-item ARMS scale encompasses two subscales: eight items for assessing medication adherence and the other four for refilling prescriptions, respectively. Responses were assessed using a 4-point Likert scale, with “none”, “some of the time”, “most of the time”, and “all of the time” corresponding to values ranging from 1 to 4, respectively. The total scores ranged from 12 to 48; the lower the score, the better adherence. Participants in the adherence group were identified with scores <16 points, and those in the non-adherence group with scores ≥16 points [18,19]. The internal consistency in the present study was evaluated with Cronbach’s α coefficient, suggesting an adequate reliability (α = 0.718) [20].
2.3. Intrinsic capacity assessment
The IC assessment, which was developed in accordance with the World Health Organization’s Integrated Care for Older People (ICOPE) guidelines [21], facilitates a rapid and feasible evaluation of six IC domains. The cognition subscale included three-item recall memory, time orientation, and location orientation, and impairment in the cognition domain was defined as a failure in either the item recall test or time or location orientation. The locomotion subscale included the Five Times Sit to Stand test, and impairment in the locomotion domain was defined as a failure to complete the test within 12 seconds [22]. Within the vitality subscale, participants who underwent an unintentional weight reduction exceeding 3 kg or reported appetite loss during the previous three months were defined as malnourished. Participants were classified as having a visual impairment in the vision subscale if they self-reported difficulty in watching. Within the hearing subscale, the hearing ability was evaluated using the whisper test to repeat a series of three numbers, 6, 1, and 9, conducted once on each ear, with participants experiencing difficulty hearing in either ear classified as having impaired hearing. In the psychological subscale, participants were deemed to have a psychosocial impairment if they responded affirmatively to either of the two evaluative questions: “Have you been bothered by feeling down, depressed, or hopeless in the past two weeks?” or “Have you lost interest in your usual activities in the past two weeks?”. All the subscales were then converted into 0 (no problems) or 1 (having problems), with higher scores ranging from 0 to 6, suggesting poorer IC.
This study used three types of measures of IC. First, each IC impairment in the six domains was dichotomously converted into 0 (no problem) or 1 (having the problem). Second, overall IC was categorized into three levels: a total score of 3 and above representing low IC, 1-2 indicating moderate IC, and 0 designating high IC [23]. Lastly, IC patterns were identified by the latent class analysis (LCA) which classifies heterogeneous subgroups of older adults with various patterns of IC impairments.
2.4. Covariables
The covariates analyzed in this study included sociodemographic variables, lifestyle, and settings for receiving IC assessments. Sociodemographic variables comprised gender (male and female), age (60–69, 70–79, 80+), educational level (uneducated, primary school, junior school, or senior high school or above), and marital status (currently married or others). Health status is represented by numbers (1, 2, 3, ≧4) and types of NCDs, including hypertension, diabetes mellitus, hyperlipidemia, cardiovascular disease, cerebrovascular disease, lung disease, kidney disease, liver disease, bone and joint problems, emotional problems, and cancers. Among these NCDs, cardiovascular disease was specifically included in the adjusted model due to its significant association with a higher likelihood of medication non-adherence in the bivariate analysis. Lifestyle factors occurring in the past month comprised smoking, alcohol consumption, chewing betel quid, mild exercise, and vigorous exercise, as self-reported by the participants. Furthermore, settings for receiving IC assessments of older people encompassed community, outpatient, and inpatient.
2.5. Statistical analysis
LCA was conducted to classify IC patterns. The optimal number of IC impairment patterns was determined by comparing the goodness-of-fit of various models using the Akaike Information Criteria (AIC), Bayesian Information Criteria (BIC), and sample-size adjusted BIC (SABIC), with lower values of these criteria indicating a superior model fit [24], and the Bayes factor (BF), with values of 3 or higher being desirable [25]. In the process of final class selection, four statistical methods were further employed: the Bootstrap Log-Likelihood Ratio (BLR) test, Entropy, and average latent class posterior probability (ALCPP). The BLR, in particular, generates p-values that statistically determine if one model is superior to another [26]. Entropy, a measure of the model’s class definition accuracy, ideally approaches 1, with values above 0.8 considered acceptable [27]. ALCPP represents the average likelihood of a class model correctly identifying the class membership of individuals, with values above 0.8 considered acceptable [28]. The fit statistics from the 1- to 5-class models to identify appropriate IC impairment patterns are shown in Supplementary Table S1. AIC and SABIC values favored the 3- or 4- class model; BIC, BLRT, and ALCPP values favored the 2- or 3- class model; BF values favored 3- to 5-class models; Entropy favored the 3-class model. We selected the 3-class model to define IC impairment patterns in this study. Qualitative labels were assigned to the three impairment patterns based on the conditional probabilities: “Low IC impairments (LIC)”, “Physical and Nutritional impairments accompanied by Depressive symptoms (PND)” and “Cognitive and Physical impairments accompanied by Hearing loss (CPH)” (Fig. 1).
Fig. 1.
Latent Profiles of IC impairment for 3-class model.
Chi-squared tests and Mann-Whitney U tests were conducted to comparing categorical variables and continuous variables, respectively, in relation to medication non-adherence. Chi-squared tests were further conducted to compare characteristics between different subgroups based on LCA. To adjust for the potential confounders, we conducted binary logistic regressions, which showed independent association of medication non-adherence with various IC measures. LCA was conducted using Mplus 8.4 (Muthen & Muthen, Los Angeles), and SPSS 29.0 (IBM Corp, Armonk) was used to perform descriptive, bivariate, and multivariate statistics. A p-value of <0.05 was considered significant.
3. Results
Table 1 summarizes the descriptive statistics and bivariate analyses of older people based on their medication adherence status. A total of 894 older adults were recruited as the study population, with an average age of 72.76 ± 7.10 years. Approximately 37.1% of the study participants had more than 1 NCD, with hypertension being the most prevalent (62.0%), followed by diabetes mellitus (37.2%), bone and joint problems (33.1%), cardiovascular diseases (24.8%), and kidney disease (22.1%). The prevalence of medication non-adherence in the study sample was 19.60%. Some 47.1% of participants had moderate IC with 1 or 2 impairments, and 17.7% had low IC with 3 or more impairments. The class prevalence of the LIC group was 75.20%, indicating that our study participants were relatively healthy. Approximately 13.10% of the study participants had co-occurrent cognitive impairment, limited mobility, and hearing loss and were categorized into the CPH group. The PND group included those suffering from limited mobility, malnutrition, depressive symptoms, and 11.70% of the study population displayed this impairment pattern. Among all study participants, impairments in locomotion were most prevalent (29.8%), followed by cognitive ability (23.9%), hearing (19.6%), vision (18.7%), psychosocial (18.7%), and vitality (18.6%). Older adults with moderate or low IC had a significantly higher prevalence of being medication non-adherent. Regarding IC patterns, those with malnutrition, depression, and limited mobility were more likely to experience medication non-adherence. Concerning IC impairments, those with cognitive impairments, malnutrition, hearing loss, or depressive symptoms were significantly more likely to be medication non-adherent.
Table 1.
Bivariate analysis of factors related to medication non-adherence of the study subjects (N = 894).
| Variables | N (%) | Adherent | Non-Adherent | p-Value |
|---|---|---|---|---|
| N (%) | N (%) | |||
| Total | 894 (100.00) | 719 (80.4) | 175 (19.6) | |
| Overall IC | 0.003 | |||
| High IC (score = 0) | 315 (35.2) | 270 (85.7) | 45 (14.3) | |
| Moderate IC (score = 1–2) | 421 (47.1) | 334 (79.3) | 87 (20.7) | |
| Low IC (score≧3) | 158 (17.7) | 115 (72.8) | 43 (27.2) | |
| IC patterns | 0.049 | |||
| LIC | 672 (75.2) | 553 (82.3) | 119 (17.7) | |
| PND | 105 (11.7) | 78 (74.3) | 27 (25.7) | |
| CPH | 117 (13.1) | 88 (75.2) | 29 (24.8) | |
| Specific IC component | ||||
| Cognitive impairment | 214 (23.9) | 158 (73.8) | 56 (26.2) | 0.005 |
| Limited mobility | 266 (29.8) | 210 (78.9) | 56 (21.1) | 0.469 |
| Malnutrition | 166 (18.6) | 124 (74.7) | 42 (25.3) | 0.039 |
| Visual impairment | 167 (18.7) | 127 (76.0) | 40 (24.0) | 0.114 |
| Hearing loss | 175 (19.6) | 128 (73.1) | 47 (26.9) | 0.007 |
| Depressive symptoms | 167 (18.7) | 117 (70.1) | 50 (29.9) | <0.001 |
| Female | 451 (50.4) | 363 (80.5) | 88 (19.5) | 0.962 |
| Age (years), median (IQR) # | 72 (68–77) | 72 (68–77) | 72 (67–77) | 0.565 |
| Age categories (years) | 0.496 | |||
| 60–69 | 319 (35.7) | 250 (78.4) | 69 (21.6) | |
| 70–79 | 417 (46.6) | 339 (81.3) | 78 (18.7) | |
| 80+ | 158 (17.7) | 130 (82.3) | 28 (17.7) | |
| Education | 0.253 | |||
| Uneducated | 76 (8.5) | 60 (78.9) | 16 (21.1) | |
| Primary School | 299 (33.4) | 230 (76.9) | 69 (23.1) | |
| Junior School | 126 (14.1) | 104 (82.5) | 22 (17.5) | |
| Senior High School and above | 393 (44.0) | 325 (82.7) | 68 (17.3) | |
| Currently Married | 635 (71.0) | 509 (80.2) | 126 (19.8) | 0.752 |
| Number of non-communicable diseases | 0.076 | |||
| 1 | 294 (32.9) | 247 (84.0) | 47 (16.0) | |
| 2 | 275 (30.8) | 217 (78.9) | 58 (21.1) | |
| 3 | 183 (20.5) | 150 (82.0) | 33 (18.0) | |
| ≧4 | 142 (15.9) | 105 (73.9) | 37 (26.1) | |
| Hypertension | 554 (62.0) | 443 (80.0) | 111 (20.0) | 0.657 |
| Diabetes mellitus | 333 (37.2) | 260 (78.1) | 73 (21.9) | 0.173 |
| Cardiovascular disease | 222 (24.8) | 168 (75.7) | 54 (24.3) | 0.040 |
| Cerebrovascular disease | 72 (8.1) | 58 (80.60) | 14 (19.4) | 0.977 |
| Lung disease | 61 (6.8) | 52 (85.2) | 9 (14.8) | 0.326 |
| Kidney disease | 198 (22.1) | 157 (79.3) | 41 (20.7) | 0.649 |
| Liver disease | 119 (13.3) | 92 (77.3) | 27 (22.7) | 0.358 |
| Bone and joint problems | 296 (33.1) | 239 (80.7) | 57 (19.3) | 0.866 |
| Emotional problems | 73 (8.2) | 60 (82.2) | 13 (17.8) | 0.691 |
| Cancers | 105 (11.7) | 81 (77.1) | 24 (22.9) | 0.367 |
| Smoking | 194 (21.7) | 157 (80.9) | 37 (19.1) | 0.842 |
| Alcohol consumption | 96 (10.7) | 74 (77.1) | 22 (22.9) | 0.382 |
| Chewing betel quid | 62 (6.9) | 45 (72.6) | 17 (27.4) | 0.107 |
| Mild exercise | 705 (78.9) | 569 (80.7) | 136 (19.3) | 0.679 |
| Vigorous exercise | 163 (18.2) | 124 (76.1) | 39 (23.9) | 0.122 |
| Settings receiving IC assessments | 0.092 | |||
| Community | 296 (33.1) | 235 (79.4) | 61 (20.6) | |
| Outpatient | 519 (58.1) | 427 (82.3) | 92 (17.7) | |
| Inpatient | 79 (8.8) | 57 (72.2) | 22 (27.8) |
#, Mann-Whitney test; IC, intrinsic capacity; LIC, low IC impairments; PND, physical and nutritional impairments accompanied by depressive symptoms; CPH, cognitive and physical impairments accompanied by hearing loss.
Table 2 compares the characteristics between different IC impairment patterns based on LCA, namely, LIC, PND, and CPH. IC impairment patterns were not statistically associated with gender, age, and settings receiving IC assessment. Those in the PND group and CPH group had more impairments in various domains of IC (p < 0.001).
Table 2.
Comparisons of demographic and intrinsic capacity between the three groups of participants with different IC patterns based on latent class analysis.
| Variables | Low IC impairments | Physical and nutritional impairments accompanied by depressive symptoms | Cognitive and physical impairments accompanied by hearing loss | p-Value | |||
|---|---|---|---|---|---|---|---|
| N | % | N | % | N | % | ||
| Numbers | 672 | 100.0 | 105 | 100.0 | 117 | 100.0 | |
| Female | 337 | 50.1 | 52 | 49.5 | 62 | 53.0 | 0.834 |
| Age (years) | |||||||
| 60–69 | 246 | 36.6 | 30 | 28.6 | 43 | 36.8 | 0.448 |
| 70–79 | 313 | 46.6 | 51 | 48.6 | 53 | 45.3 | |
| 80+ | 113 | 16.8 | 24 | 22.9 | 21 | 17.9 | |
| Settings receiving IC assessments | |||||||
| Community | 222 | 33.0 | 43 | 41.0 | 31 | 26.5 | 0.262 |
| Outpatient | 390 | 58.0 | 54 | 51.4 | 75 | 64.1 | |
| Inpatient | 60 | 8.9 | 8 | 7.6 | 11 | 9.4 | |
| Specific IC component | |||||||
| Cognitive impairment | 96 | 14.3 | 11 | 10.5 | 107 | 91.5 | <0.001 |
| Limited mobility | 81 | 12.1 | 73 | 69.5 | 112 | 95.7 | <0.001 |
| Malnutrition | 47 | 7.0 | 91 | 86.7 | 28 | 23.9 | <0.001 |
| Visual impairment | 105 | 15.6 | 19 | 18.1 | 43 | 36.8 | <0.001 |
| Hearing loss | 97 | 14.4 | 15 | 14.3 | 63 | 53.8 | <0.001 |
| Depressive symptoms | 48 | 7.1 | 81 | 77.1 | 38 | 32.5 | <0.001 |
IC, intrinsic capacity; low IC impairments, LIC; physical and nutritional impairments accompanied by depressive symptoms, PND; cognitive and physical impairments accompanied by hearing loss, CPH.
Table 3 shows the association between medication non-adherence and various measures of IC impairment. Older adults in the moderate and low IC group were 57% and 126% more likely to experience medication non-adherence compared with those with the high IC (moderate: adjusted odds ratio (aOR) 1.57 [95% confidence interval (CI) 1.05–2.36], P = 0.028; low: aOR 2.26 [1.40–3.67], P = 0.001) after adjustment for age, sex, education, marital status, numbers of non-communicable diseases, cardiovascular disease, smoking, alcohol consumption, chewing betel quid, mild exercise, vigorous exercise, and settings receiving IC assessments, respectively. Older adults in the PND group exhibited a 66% increase in the odds of medication non-adherence compared with the LIC group (aOR 1.66 [1.01–2.73], P = 0.044) after adjustment for the covariates. Regarding individual IC impairment, older adults with cognitive impairment had a 53% higher likelihood of medication non-adherence compared to those without cognitive impairment after considering covariates (aOR 1.53 [1.03–2.27], P = 0.037). Additionally, individuals with hearing loss showed a 57% increase in the odds of medication non-adherence, while those with depression symptoms were 81% more likely to experience medication non-adherence (hearing loss: aOR 1.57 [1.03–2.37], P = 0.035; depressive symptoms: aOR 1.81 [1.17–2.80], P = 0.008).
Table 3.
Odds ratio of medication non-adherence in association with various measures of intrinsic capacity.
| Various measures of IC | Unadjusted model |
Adjusted model 1 |
Adjusted model 2 |
Adjusted model 3 |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OR | 95% CI | p-Value | aOR | 95% CI | p-Value | aOR | 95% CI | p-Value | aOR | 95% CI | p-Value | |
| Overall IC | ||||||||||||
| High IC (score = 0) | ref | ref | ||||||||||
| Moderate IC (score = 1–2) | 1.56 | (1.05, 2.32) | 0.026 | 1.57 | (1.05, 2.36) | 0.028 | ||||||
| Low IC (score≧3) | 2.24 | (1.41, 3.60) | 0.001 | 2.26 | (1.40, 3.67) | 0.001 | ||||||
| IC patterns | ||||||||||||
| LIC | Ref | ref | ||||||||||
| PND | 1.61 | (1.00, 2.60) | 0.052 | 1.66 | (1.01, 2.73) | 0.044 | ||||||
| CPH | 1.53 | (0.96, 2.44) | 0.072 | 1.50 | (0.93, 2.42) | 0.094 | ||||||
| Specific IC component | ||||||||||||
| Cognitive impairment | 1.54 | (1.04, 2.26) | 0.030 | 1.53 | (1.03, 2.27) | 0.037 | ||||||
| Limited mobility | 0.79 | (0.53, 1.17) | 0.234 | 0.75 | (0.50, 1.13) | 0.172 | ||||||
| Malnutrition | 1.26 | (0.81, 1.95) | 0.312 | 1.31 | (0.83, 2.06) | 0.240 | ||||||
| Visual impairment | 0.83 | (0.55, 1.26) | 0.380 | 1.27 | (0.83, 1.96) | 0.268 | ||||||
| Hearing loss | 1.56 | (1.04, 2.34) | 0.030 | 1.57 | (1.03, 2.37) | 0.035 | ||||||
| Depressive symptoms | 1.83 | (1.20, 2.79) | 0.005 | 1.81 | (1.17, 2.80) | 0.008 | ||||||
aOR = adjusted odds ratio from the multivariate logistic regression with adjustment for age, sex, education, marital status, numbers of non-communicable diseases, cardiovascular disease, smoking, alcohol consumption, chewing betel quid, mild exercise, vigorous exercise, and settings receiving IC assessments.
IC, intrinsic capacity; LIC, low IC impairments; PND, physical and nutritional impairments accompanied by depressive symptoms; CPH, cognitive and physical impairments accompanied by hearing loss.
4. Discussions
To the best of our knowledge, this research is the first to examine the associations of medication non-adherence with IC levels and patterns as well as specific domain of IC impairments among older adults with NCDs, in which 64.8% of participants had impairments in at least one specific domain of IC, and those in the moderate or low level of IC group were more likely to experience medication non-adherence. This study identified 3 distinct IC impairment patterns (i.e., LIC, PND, and CPH), and the PND group was associated with statistically higher odds of medication non-adherence. Among specific domains of IC impairments, those with cognitive impairment, hearing loss, or depressive symptoms showed a greater likelihood of medication non-adherence.
Our study revealed that lower IC levels were related to higher odds of medication non-adherence. It has brought insight about adverse outcomes of poor IC in older adults, which has never been discovered yet [29]. In our study, 47.1% of participants who had moderate IC with 1 or 2 impairments and 17.7% who had low IC with 3 or more impairments were identified through comprehensive IC assessments and then had opportunities to receive further detailed assessment and corresponding interventions. Our findings responded to the WHO’s recommendation on ICOPE through comprehensive assessments of the six ICs [16]. The comprehensive IC assessment could contribute to patient-centered care and medication self-management support, which can help older adults develop functional abilities to improve medication non-adherence [[9], [10], [11]].
We also observed that the PND group was associated with a statistically higher likelihood of medication non-adherence. Malnutrition, depression, and limited mobility among older people were closely interconnected. A study in France revealed that people with appetite and weight loss showed 4.33 times and 2.38 times higher likelihood of screening positive for psychological and locomotion impairments, respectively [30]. Research in South Korea found that malnutrition was significantly related to depression in community-dwelling older adults [31]. Another study in Finland showed that poorer life-space mobility was related to a greater prevalence of depressive symptoms [32]. A prior study found that older adults with substantial declines in locomotion, psychological, cognition, and vitality domains were at twice the risk of hospitalization compared with relatively healthy older adults [33]. Medication non-adherence could lead to all-cause hospitalization in older people [7]. Medication non-adherence might be a mediator of the association between IC and hospitalization. Medication adherence may be increased by improving IC in older adults with NCDs, through multidomain intervention programs which include the revision and optimization of pharmacological treatments [34]. Future attention should be focused on older adults who have malnutrition, depression, and limited mobility, and early multidomain intervention which aims to improve IC, which may in turn lead to improved medication non-adherence and reduction of hospitalization.
Furthermore, our study found that the CPH group was not related to a statistically higher likelihood of medication non-adherence (aOR = 1.50, p = 0.094); however, those with cognitive impairment (aOR = 1.53, p = 0.037) or hearing loss (aOR = 1.57, p = 0.035) showed a greater likelihood of medication non-adherence. The adjusted odd ratios are similar in magnitude, but the sample size of the non-adherent group in those with CPH (n = 29) is much smaller than those with cognitive impairment (n = 56) or hearing loss (n = 47). The smaller sample size in the CPH non-adherent group might be a reason for not being statistically significant.
Regarding individual IC impairments, our study revealed that older adults with cognitive impairment, hearing loss, or depressive symptoms were also at greater odds of medication non-adherence. In this study, the prevalence rates of cognitive impairment, depression symptoms, and hearing loss were 23.9%, 18.7%, and 19.6%, respectively, all of which are higher than the rates of 20.3%, 7.6%, and 10.8% found in a previous study [17]. That might be because our study sample was limited to those with NCDs, and NCDs were related to cognitive impairment [35], depression symptoms [36], and hearing loss [37]. Our findings corroborated with those of previous research, which suggested that cognitive impairment [[12], [13], [14]] and depression symptoms [6] were associated with medication non-adherence. Furthermore, our study highlighted the potential impact of hearing loss on medication non-adherence. Previous research revealed that medication non-adherence could lead to all-cause hospitalization and mortality in older people [7], and hearing loss was related to a higher risk of hospitalizations, readmission, and mortality [38]. Early intervention to improving or managing hearing loss could enhance medication adherence and decrease hospitalizations, readmissions, and mortality rates. Our findings responded to the current advocate of monitoring for hearing for person-centered interventions in older adults by WHO [21,39].
Some limitations should be acknowledged for this study. First, this study was based on a cross-sectional questionnaire survey, which could not confirm the causal relationship between IC and medication non-adherence. Additional longitudinal and interventional studies are necessary to validate our findings further and to explore their clinical implications. Second, in our analyses, we did not perform a comprehensive adjustment for all potential confounders, such as living arrangement [40]. Lastly, the study sample was drawn from a single city in Taiwan, which may restrict the generalizability of the findings.
5. Conclusion
Older adults with lower levels of IC tended to experience higher odds of medication non-adherence. The PND group was more likely than other groups to experience medication non-adherence. Older adults with cognitive impairment, hearing loss, or depressive symptoms exhibited a greater likelihood of medication non-adherence. Our results point to the importance of comprehensive IC assessments among older adults as well as the development of interventions towards cognitive impairment, hearing loss, depressive symptoms, and a combination of malnutrition with depression and limited mobility in reducing medication non-adherence.
Author contributions
CB Lee and CY Li: Conceptualization, Methodology and Funding acquisition; CY Li: Data curation, Supervision, Validation and Project administration; CB Lee: Formal analysis, Visualization and Writing - original draft; CY Li, YC Yang, LJE Ku, YT Chou, HY Chen, HC Su, YL Wu, YT Lo: Investigation and Writing - review & editing. All authors approved the final version of the manuscript.
Ethics declaration
The study was conducted in accordance with the guidelines of the Declaration of Helsinki. All participants signed a written informed consent before their participation in the study. This study was approved by the Institutional Review Board of NCKUH (IRB No.: A-ER-110-249).
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgment
The authors are grateful for grants from the National Science and Technology Council (NSTC 112-2314-B-006-067-MY3, NSTC 112-2410-H-039-004), China Medical University (CMU111-S-10), and National Health Research Institutes (NHRI-13A1-CG-CO-04-2225-1). The funder has no role in the conduct and submission of this work. The guarantor is CY Li, who takes full responsibility for the work as a whole, including the study design, data access, and decision to submit and publish the manuscript.
Footnotes
Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jnha.2024.100303.
Appendix A. Supplementary data
The following is Supplementary data to this article:
References
- 1.Cao W., Kadir A.A., Wang J., Hu L., Wen L., Yu M., et al. Medication non-adherence and associated factors among older adult stroke survivors in China. Front Pharmacol. 2022;13 doi: 10.3389/fphar.2022.1054603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bouwman L., Eeltink C.M., Visser O., Janssen JJWM, Maaskant J.M. Prevalence and associated factors of medication non-adherence in hematological-oncological patients in their home situation. BMC Cancer. 2017;17(1):739. doi: 10.1186/s12885-017-3735-1. 2017/11/09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Franchi C., Ardoino I., Ludergnani M., Cukay G., Merlino L., Nobili A. Medication adherence in community-dwelling older people exposed to chronic polypharmacy. J Epidemiol Community Health. 2021;75(9):854–859. doi: 10.1136/jech-2020-214238. [DOI] [PubMed] [Google Scholar]
- 4.Félix I.B., Henriques A. Medication adherence and related determinants in older people with multimorbidity: a cross-sectional study. Nurs Forum. 2021;56(4):834–843. doi: 10.1111/nuf.12619. [DOI] [PubMed] [Google Scholar]
- 5.Chew S.M., Lee J.H., Lim S.F., Liew M.J., Xu Y., Towle R.M. Prevalence and predictors of medication non-adherence among older community-dwelling people with chronic disease in Singapore. J Adv Nurs. 2021;77(10):4069–4080. doi: 10.1111/jan.14913. [DOI] [PubMed] [Google Scholar]
- 6.Ge L., Heng B.H., Yap C.W. Understanding reasons and determinants of medication non-adherence in community-dwelling adults: a cross-sectional study comparing young and older age groups. BMC Health Serv Res. 2023;23(1):905. doi: 10.1186/s12913-023-09904-09908. 2023/08/24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Walsh C.A., Cahir C., Tecklenborg S., Byrne C., Culbertson M.A., Bennett K.E. The association between medication non-adherence and adverse health outcomes in ageing populations: a systematic review and meta-analysis. Br J Clin Pharmacol. 2019;85(11):2464–2478. doi: 10.1111/bcp.14075. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cutler R.L., Fernandez-Llimos F., Frommer M., Benrimoj C., Garcia-Cardenas V. Economic impact of medication non-adherence by disease groups: a systematic review. BMJ Open. 2018;8(1) doi: 10.1136/bmjopen-2017-016982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yang C, Lee Diana Tze F, Wang X, Chair Sek Y. Developing a medication self-management program to enhance medication adherence among older adults with multimorbidity using intervention mapping. Gerontologist. 2022;63(4):637–647. doi: 10.1093/geront/gnac069. [DOI] [PubMed] [Google Scholar]
- 10.Yang C., Zhu S., Lee D.T.F., Chair S.Y. Interventions for improving medication adherence in community-dwelling older people with multimorbidity: a systematic review and meta-analysis. Int J Nurs Stud. 2022;126 doi: 10.1016/j.ijnurstu.2021.104154. 2022/02/01. [DOI] [PubMed] [Google Scholar]
- 11.Poitras M.-E., Maltais M.-E., Bestard-Denommé L., Stewart M., Fortin M. What are the effective elements in patient-centered and multimorbidity care? A scoping review. BMC Health Serv Res. 2018;18(1):446. doi: 10.1186/s12913-018-3213-3218. 2018/06/14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Advinha A.M., Nunes C., de Barros C.T., Lopes M.J., de Oliveira-Martins S. Key factors of the functional ability of older people to self-manage medications. Sci Rep. 2021;11(1):22196. doi: 10.1038/s41598-021-01434-01439. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Davies L.E., Todd A., Sinclair D.R., Robinson L., Kingston A. Is polypharmacy associated with difficulty taking medicines in people aged ≥85 living at home? Findings from the Newcastle 85+ Study. Br J Clin Pharmacol. 2023;89(10):3217–3227. doi: 10.1111/bcp.15858. [DOI] [PubMed] [Google Scholar]
- 14.Smaje A., Weston-Clark M., Raj R., Orlu M., Davis D., Rawle M. Factors associated with medication adherence in older patients: a systematic review. Aging Med. 2018;1(3):254–266. doi: 10.1002/agm2.12045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.WHO . World Health Organization; 2015. World report on ageing and health. [Google Scholar]
- 16.WHO . World Health Organization; 2017. Integrated care for older people: guidelines on community-level interventions to manage declines in intrinsic capacity. 9241550104. [PubMed] [Google Scholar]
- 17.Meng L.C., Hsiao F.Y., Huang S.T., Lu W.H., Peng L.-N., Chen L.K. Intrinsic capacity impairment patterns and their associations with unfavorable medication utilization: a nationwide population-based study of 37,993 community-dwelling older adults. J Nutr Health Aging. 2022;26(10):918–925. doi: 10.1007/s12603-022-1847-z. 2022/10/01. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kripalani S., Risser J., Gatti M.E., Jacobson T.A. Development and evaluation of the Adherence to Refills and Medications Scale (ARMS) among low-literacy patients with chronic disease. Value Health. 2009;12(1):118–123. doi: 10.1111/j.1524-4733.2008.00400.x. [DOI] [PubMed] [Google Scholar]
- 19.Fadil H.A., Sammman W.A., Elshafie R.M. Prevalence of nonadherence to medications among geriatric patients in Al-Madinah Al-Munawara City’s Hospitals, Kingdom of Saudi Arabia. Int J Clin Pract. 2023;2023(1) doi: 10.1155/2023/3312310. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Bland JM, Altman DG. Statistics notes: Cronbach’s alpha. BMJ. 1997;314(7080):572. doi: 10.1136/bmj.314.7080.572. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.WHO World Health Organization; 2019. Integrated care for older people (ICOPE): guidance for person-centred assessment and pathways in primary care. https://www.who.int/publications/i/item/WHO-FWC-ALC-19.1 [cited 2024 3 January]; Available from:
- 22.Chen L.-K., Woo J., Assantachai P., Auyeung T.-W., Chou M.-Y., Iijima K., et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21(3):300–307.e2. doi: 10.1016/j.jamda.2019.12.012. 2020/03/01. [DOI] [PubMed] [Google Scholar]
- 23.Chang Y.-H., Chen Y.-C., Ku L-JE, Chou Y.-T., Chen H.-Y., Su H.-C., et al. Association between sleep health and intrinsic capacity among older adults in Taiwan. Sleep Med. 2023;109:98–103. doi: 10.1016/j.sleep.2023.06.016. [DOI] [PubMed] [Google Scholar]
- 24.Lanza ST, Rhoades BL. Latent class analysis: an alternative perspective on subgroup analysis in prevention and treatment. PREV SCI. 2013;14(2):157–168. doi: 10.1007/s11121-011-0201-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Masyn KE. Vol 2. Oxford University Press; New York, NY, US: 2013. Latent class analysis and finite mixture modeling; pp. 551–611. (The Oxford Handbook of Quantitative Methods: Statistical Analysis). [Google Scholar]
- 26.Nylund K.L., Asparouhov T., Muthén B.O. Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equ Modeling. 2007;14(4):535–569. doi: 10.1080/10705510701575396. [DOI] [Google Scholar]
- 27.Weller BE, Bowen NK, Faubert SJ. Latent class analysis: a guide to best practice. J Black Psychol. 2020;46(4):287–311. doi: 10.1177/0095798420930932. [DOI] [Google Scholar]
- 28.Weden MM, Zabin LS. Gender and ethnic differences in the co-occurrence of adolescent risk behaviors. Ethn Health. 2005;10(3):213–234. doi: 10.1080/13557850500115744. [DOI] [PubMed] [Google Scholar]
- 29.Yang Y., Ma G., Wei S., Wei X., Yan B., Yuan Y., et al. Adverse outcomes of intrinsic capacity in older adults: a scoping review. Arch Gerontol Geriatr. 2024;120 doi: 10.1016/j.archger.2024.105335. 2024/05/01. [DOI] [PubMed] [Google Scholar]
- 30.Gaussens L., González-Bautista E., Bonnefoy M., Briand M., Tavassoli N., De Souto Barreto P., et al. Associations between vitality/nutrition and the other domains of intrinsic capacity based on data from the INSPIRE ICOPE-care program. Nutrients. 2023;15(7):1567. doi: 10.3390/nu15071567. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Hwang G., Cho Y.H., Hong C.H., Son S.J. Differential effects of sleep disturbance and malnutrition on late-life depression among community-dwelling older adults. Front Psychiatry. 2022;13 doi: 10.3389/fpsyt.2022.820427. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Polku H., Mikkola T.M., Portegijs E., Rantakokko M., Kokko K., Kauppinen M., et al. Life-space mobility and dimensions of depressive symptoms among community-dwelling older adults. Aging Ment Health. 2015;19(9):781–789. doi: 10.1080/13607863.2014.977768. [DOI] [PubMed] [Google Scholar]
- 33.Yu J., Si H., Jin Y., Qiao X., Ji L., Bian Y., et al. Patterns of intrinsic capacity among community-dwelling older adults: identification by latent class analysis and association with one-year adverse outcomes. Geriatr Nurs. 2022;45:223–229. doi: 10.1016/j.gerinurse.2022.04.021. 2022/05/01. [DOI] [PubMed] [Google Scholar]
- 34.Ferrara M.C., Pérez L.M., Sole A.R., Villa-García L., Ars J., Soto-Bagaria L., et al. Sustained improvement of intrinsic capacity in community-dwelling older adults: the +AGIL Barcelona multidomain program. J Intern Med. 2023;294(6):730–742. doi: 10.1111/joim.13710. [DOI] [PubMed] [Google Scholar]
- 35.Jin Y., Liang J., Hong C., Liang R., Luo Y. Cardiometabolic multimorbidity, lifestyle behaviours, and cognitive function: a multicohort study. Lancet Healthy Longev. 2023;4(6):e265–e273. doi: 10.1016/S2666-7568(23)00054-5. [DOI] [PubMed] [Google Scholar]
- 36.Bobo W.V., Grossardt B.R., Virani S., St Sauver J.L., Boyd C.M., Rocca W.A. Association of depression and anxiety with the accumulation of chronic conditions. JAMA Netw Open. 2022;5(5):e229817-e. doi: 10.1001/jamanetworkopen.2022.9817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wu K.-L., Shih C.-P., Chan J.-S., Chung C.-H., Lin H.-C., Tsao C.-H., et al. Investigation of the relationship between sensorineural hearing loss and associated comorbidities in patients with chronic kidney disease: a nationwide, population-based cohort study. PLoS One. 2020;15(9) doi: 10.1371/journal.pone.0238913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Hsu A.K., McKee M., Williams S., Roscigno C., Crandell J., Lewis A., et al. Associations among hearing loss, hospitalization, readmission and mortality in older adults: a systematic review. Geriatr Nurs. 2019;40(4):367–379. doi: 10.1016/j.gerinurse.2018.12.013. 2019/07/01. [DOI] [PubMed] [Google Scholar]
- 39.Chadha S., Dillard L.K., Mariotti S.P., Keel S. Monitoring hearing and vision functions in older adults: rationale and process. Age Ageing. 2023;52(Supplement_4):iv158–iv161. doi: 10.1093/ageing/afad123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lindsay-Rahman R., Lennie T., Moser D., Chung M. A moderated moderation of living arrangements and social support on the relationship between depressive symptoms and medication adherence in patients with heart failure. Eur J Cardiovasc Nurs. 2021;20(Supplement_1) doi: 10.1093/eurjcn/zvab060.108. zvab060. 108. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.