INTRODUCTION
1. Background
Life expectancy has increased globally, accelerating population aging [1]. South Korea, one of the most rapidly aging countries worldwide, is projected to have adults aged ≥65 comprise 34.3% of its population by 2040 [2]. This demographic shift presents major social and economic challenges, especially the growing burden of chronic diseases [3]. Therefore, the development of systematic and sustainable strategies for managing chronic diseases among older adults is an urgent priority.
In older adults, the management of chronic diseases involves pharmacotherapy, lifestyle modifications, and regular medical check-ups, and pharmacotherapy is a core component. This reliance on pharmacotherapy is particularly evident in South Korea, where 86.1% of older adults have at least one chronic disease and 63.9% have multimorbidity (≥2 diseases) [3]. This multimorbidity frequently necessitates multiple medications, leading to polypharmacy [4]. In older adults, age-related physiological declines in absorption, distribution, metabolism, and excretion, the concurrent use of multiple drugs substantially increases the risk of adverse drug reactions (ADRs) [5].
Existing evidence indicates that the burden of ADRs is consistently shown to be higher in older adults than in the general adult population. In adult outpatients seen in primary care, a systematic review and meta-analysis estimated a pooled ADR prevalence of 8.32% (95% confidence interval [95% CI]=7.82~8.83) [6]. In community-dwelling older adults, a 6-year prospective cohort reported that 26.9% experienced at least one ADR [7]. Regarding hospital admissions, a large-scale study in Japan found that 5.0% (95% CI=4.5%~5.6%) of admissions in the general adult population were attributable to ADRs [8], whereas in older adults, ADR-related admissions accounted for 11.0% (95% CI=5.1%~16.8%) [9]. Collectively, these findings highlight that ADRs in older adults are not only more frequent but also associated with more severe outcomes, such as hospitalization, functional decline, and increased healthcare costs.
Despite its importance, most studies on ADRs in polypharmacy have focused primarily on medication counts, showing that risk increases with more prescriptions. Doherty et al. [7] found that, relative to those taking fewer than five medications, older adults using 5–9 drugs had a significantly higher risk of ADRs (odds ratio [OR]=1.81, 95% CI=1.17~2.82), with the risk rising further among those taking 10 or more medications (OR=3.33, 95% CI=1.62~6.85) [7]. Similarly, a prospective inpatient analysis showed that patients who experienced ADRs were taking more medications than those without ADRs (10.5 vs. 7.8, p<.01) [10].
Beyond medication counts, multimorbidity poses additional challenges, as each condition requires individual treatment and prescriptions are often fragmented across providers [11]. Such inadequate prescription management can further elevate the risk of adverse outcomes due to drug–drug interactions or duplicate prescriptions [12]. Therefore, it is important to focus not only on the number of prescribed medications but also on the associations between the types and number of chronic conditions and the occurrence of ADRs. Nevertheless, previous research has not sufficiently examined these differential associations, and studies reflecting the complex health status of older patients remain limited.
A prior study showed that patients who experienced ADRs had a greater number of comorbidities compared with those without ADRs (6.1 vs. 5.2, p<.01) [10]. However, this study relied on simple group comparisons and did not provide a quantitative estimate of risk according to the number of chronic conditions. In addition, disease-specific studies have also reported elevated ADR risks, such as in cardiovascular disease [13] and chronic kidney disease [14], yet these findings remain limited to selected conditions.
Despite valuable insights from prior work, existing studies have been limited in scope and have not fully captured the complexity of multimorbidity in older adults. To address this gap, the present study stratifies the number of chronic conditions (1, 2, 3, 4, ≥5) and classifies their types into 11 categories, evaluating their associations with ADRs simultaneously. This more granular approach, beyond simple summation or coarse grouping, clarifies risk differences attributable to both disease count and composition.
Accordingly, this study aims to quantify ADR risk across the defined disease counts and 11 disease types in older adults, and to identify high-risk constellations that are most relevant in real-world care. By reflecting the clinical profiles of older patients, the findings are intended to inform targeted risk stratification and monitoring, support safer medication management under polypharmacy, and ultimately contribute to improved healthcare quality for older adults.
METHODS
Ethic statement: As this study used publicly available data, it was reviewed and approved as exempt from review by the Institutional Review Board (IRB) of the Incheon Catholic University (IRB No. 2024-ICCU-IRB-02).
1. Study Design
This cross-sectional study investigated the associations of chronic disease type and number with ADR risk in older adults. It was reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist (https://www.strobe-statement.org).
2. Data Source and Study Population
Data were obtained from the 2021 Korea Health Panel Survey (KHPS), jointly conducted by the Korea Institute for Health and Social Affairs and the National Health Insurance Service. The KHPS sampling frame was derived from the 2016 Registered Census compiled by Statistics Korea. Using this frame, a two-stage stratified cluster sampling method with probability proportional to size was employed to select approximately 8,500 households from 708 survey districts. Data access was granted through a formal data use agreement with the KHPS. Of the 14,847 individuals who participated in the 2021 KHPS household survey, 4,478 respondents aged ≥65 years and with complete data on both types of chronic diseases and the prevalence of ADRs were included in the final analysis.
3. Measures
1) Types and Number of Chronic Diseases
The KHPS included 31 physician-diagnosed chronic conditions, originally organized into eight clinically related groups in the participant questionnaire. For this study, these groupings were largely maintained, except that hypertension and diabetes mellitus were treated as independent categories due to their high prevalence and clinical significance. Conditions listed as ‘others’ in the questionnaire were retained as a separate category because of their heterogeneous nature, comprising three disease types: (1) depression or bipolar disorder, (2) dementia (Alzheimer’s, vascular, or alcohol-related), and (3) chronic renal failure. Accordingly, the 31 conditions were classified into 11 categories: hypertension, diabetes mellitus, cardio-cerebrovascular diseases, neoplasm, arthritis and spinal diseases, liver diseases, respiratory diseases, depression, dementia, chronic renal failure, and thyroid problems.
The total number of chronic diseases per participant was calculated as the sum of all currently present, physician-diagnosed conditions, and categorized as 0, 1, 2, 3, 4, or ≥5. The cutoff of five or more diseases was based on longitudinal evidence showing that multimorbidity (≥5 conditions) is associated with a higher risk of institutionalization among older adults [15].
2) Adverse Drug Reactions
ADRs were operationally defined using a single KHPS item asking, “In the past year, have you experienced any side effects from medications, including over-the-counter drugs?” This self-report captured side effects from both prescribed and over-the-counter medicines. Because detailed medication data and clinical adjudication were unavailable, this self-reported item was used as a practical indicator in the large community-based survey, with the analysis focusing on comparing relative differences in ADR occurrence according to the types and number of chronic diseases rather than estimating a precise prevalence. Responses were coded as ‘yes’ or ‘no.’
3) Covariates
Based on a review of the literature, potential confounders of ADRs were identified. The covariates comprised sociodemographic characteristics such as age, sex, educational level, employment status, marital status, and health-related characteristics, including regular physical activity and body mass index (BMI) [16,17].
Age groups were classified as ‘young-old’ (65~74 years) and ‘old-old’ (≥75 years). Educational level was classified into four categories based on participants’ responses: elementary school or less, middle school, high school, and college or higher. Employment status was determined according to whether the participant was engaged in work for income, with responses categorized as ‘yes’ or ‘no’. The presence of a spouse included a de facto marital relationship; participants living with a spouse were coded as ‘yes’, while those who were separated, widowed, divorced, or never married were coded as ‘no’. Regular physical activity was assessed based on participants’ responses to whether they had engaged in sports or exercise, including walking, regularly during the past year, and was coded as ‘yes’ or ‘no’. BMI was calculated as weight (kg) divided by height (m2) and classified according to World Health Organization categories [18].
4. Data Analysis
The collected data were analyzed using SPSS version 23.0 (IBM Corp.), and cross-sectional weights from the 2021 KHPS were applied to percentage estimates and bivariate analyses to reflect the national representativeness of the study population. The level of statistical significance was set at p<.05. Descriptive statistics, including frequencies and percentages for categorical variables and means and standard deviations for continuous variables, were used to summarize the participants’ sociodemographic and health-related characteristics. Differences in ADR prevalence according to these characteristics were assessed using the chi-square test. Multiple logistic regression analyses were then conducted to examine the associations between chronic disease types and number and ADR occurrence, adjusting for significant covariates identified in the chi-square tests. Participants with zero or one chronic disease were used as the reference group, reflecting a broad range of medication exposure typical among community-dwelling older adults, including the use of over-the-counter for minor health issues. Results were presented as ORs with 95% CIs, and the goodness of fit of the logistic regression models was assessed using the Nagelkerke R2 and the Hosmer–Lemeshow test.
RESULTS
1. Sociodemographic and Health-Related Characteristics of the Participants
The participants had a mean age of 74.02 years, with 57.9% aged 65~74. Females comprised a larger proportion (57.2%) than males. Nearly half had an education level below elementary school (43.4%), and more than half were not engaged in economic activity (57.0%). About 60% lived with a spouse. Regarding health-related characteristics, 54.6% reported engaging in regular exercise, and most were within the normal BMI range (65.9%). Hypertension was the most prevalent chronic condition (56.8%), followed by arthritis and spinal disorders (48.6%), diabetes mellitus (24.0%), and cardio-cerebrovascular diseases (16.5%). On average, participants had 1.91 chronic diseases, with over half having two or more. Notably, about 36% had three or more conditions, and 4.5% had five or more. A total of 3.2% of participants reported experiencing ADRs (Table 1).
2. Differences in Adverse Drug Reactions According to Sociodemographic and Health-Related Characteristics
ADRs differed significantly across all sociodemographic and health-related characteristics (p<.001). ADRs were slightly more prevalent among the old-old (≥75 years) compared with the young-old, and the proportion was markedly higher in females than in males. Individuals with below elementary school education, those without economic activity, and those without a spouse showed higher proportions of ADRs. Regular exercise was somewhat more common in the ADR group, although the difference was modest. With respect to BMI, a statistically significant difference was observed; however, the distribution across categories was largely similar between the two groups, with normal weight being the most prevalent in both (Table 2).
3. Risk of Adverse Drug Reactions According to Chronic Disease Profile
Multivariable logistic regression analyses revealed that specific chronic disease types were significantly associated with ADRs. Arthritis and spinal diseases showed the strongest association with ADRs (OR=2.16, 95% CI=1.44~3.24), followed by depression (OR=2.15, 95% CI=1.22~3.79) and cardio-cerebrovascular diseases (OR=1.86, 95% CI=1.25~2.77) (Table 3).
Regarding the number of chronic diseases, the risk of ADRs increased progressively with the number of conditions. After adjusting for all covariates, compared with participants with one or no chronic disease, the risk of ADRs was higher among those with two chronic diseases (OR=1.87, 95% CI=1.14~3.06), three chronic diseases (OR=2.63, 95% CI=1.58~4.39), four chronic diseases (OR=3.15, 95% CI=1.69~5.88), and five or more chronic diseases (OR=6.22, 95% CI=3.32~11.64) (Table 4).
DISCUSSION
This study investigated the risks of ADRs according to both the types and number of chronic diseases in older adults, aiming to provide evidence for more precise prevention and management strategies. Our findings highlight that, beyond medication counts, disease-specific profiles should be considered as an integral component of medication safety management.
In this study, multimorbidity was highly prevalent, with more than half of the participants experiencing at least two chronic conditions and over one-third reporting three or more. Given the anticipated rise in multimorbidity with population aging, these results underscore the need for comprehensive approaches that address condition–condition interactions and the complexity of treatment regimens. Incorporating disease-specific risk profiles into such strategies may further enhance tailored interventions for medication safety.
The prevalence of ADRs in this study was 3.2%. However, underreporting remains a critical issue. A systematic review of 37 studies found a median under-reporting rate of 94%, meaning that only about 6% of actual ADRs are captured in spontaneous reporting systems [19]. This suggests that the true prevalence of ADRs among older adults may be considerably underestimated. Contributing factors include reliance on self-reports, underrecognition of mild symptoms, and limited awareness of drug-related reactions [20]. Older adults in particular may misinterpret or fail to notice such symptoms due to cognitive or sensory decline. To address this, future community-based surveys should use symptom-specific checklists and medication-oriented guided questions, and link self-reports with clinical or pharmacy data for validation. From a nursing perspective, improving patients’ ability to recognize and report ADR-related changes through education, counseling, and communication with caregivers is essential for early detection and safer medication management.
The analysis by chronic disease types revealed significant differences in ADR experience, with the highest risk observed among older adults with arthritis and spinal diseases. This pattern is clinically plausible, as arthritis and spinal diseases in older adults are commonly managed with medications such as nonsteroidal anti-inflammatory drugs, opioids, and muscle relaxants, which are known to increase the risk of ADRs in later life [21]. In older adults, these medications have been associated with gastrointestinal, renal, and central nervous system–related adverse effects, including dizziness, sedation, and impaired balance [21,22]. Importantly, ADRs in older adults often present not as discrete clinical symptoms but as nonspecific functional changes, such as reduced mobility or balance instability, which may complicate early recognition [23].
Older adults with depression also demonstrated an elevated risk of ADRs, identifying this group as another major vulnerable population. This may be attributed to the fact that antidepressants act directly on the central nervous system and belong to a drug class with a high potential for interactions and adverse effects. In practice, psychotropic medications commonly prescribed in late-life depression are associated with drowsiness, dizziness, confusion, falls, and delirium [20]. Tricyclic antidepressants and certain selective serotonin reuptake inhibitors are also classified as potentially inappropriate for older adults [24], and age-related declines in drug metabolism and excretion further heighten vulnerability. Therefore, this group should be prioritized for close monitoring, with medication review supported by emotional care and patient education.
In addition, older adults with cardio-cerebrovascular diseases had nearly twice the risk of ADRs compared to those without this condition. Medications commonly used for cardio-cerebrovascular conditions have been associated with adverse effects such as bleeding and orthostatic hypotension, which are clinically relevant in older adults [25]. Furthermore, cardiovascular pharmacotherapy has been linked to an increased risk of falls and fall-related injuries in older adults [26]. Accordingly, symptom changes such as bleeding tendencies, blood pressure fluctuations, or dizziness should be interpreted with potential ADRs in mind, beyond disease progression [25,26].
Given that arthritis and spinal diseases, depression, and cardio-cerebrovascular diseases were significantly associated with ADRs, tailored nursing interventions are essential for older adults with these conditions. For older adults with arthritis and spinal diseases, nurses should monitor not only patients’ responses to medication but also their everyday mobility and balance. In particular, pain and muscle weakness can obscure early indicators of ADRs [27]. Therefore, nurses should regularly evaluate gait stability, orthostatic dizziness, and recent falls or near-fall episodes to detect emerging risk signals. Furthermore, providing guidance on low-intensity nonpharmacological pain-management approaches—such as heat application, joint-protection techniques, and brief stretching—can support pain relief while potentially lowering analgesic requirements and thereby reducing the risk of ADRs [28]. Collectively, these nursing strategies can enhance pain management and functional capacity, while also mitigating the risk of ADRs in this vulnerable population. Taken together, these findings underscore the importance of a proactive, function-oriented nursing approach to medication safety among older adults with arthritis and spinal diseases.
For individuals with depression, regular medication review and emotional support should be combined with education to improve medication adherence and recognition of drug-related symptoms. In practice, nurse-led medication review for older adults with depression can involve verifying all prescribed and over-the-counter psychotropic medications and recent regimen changes, checking indications and dosing, and asking specifically about new or worsening cognitive, mood, or mobility symptoms during routine contacts.
Older adults with cardio-cerebrovascular diseases often receive multiple medications, which may increase the risk of dizziness, hypotension, and falls [29]. Therefore, nurses should monitor vital signs closely, assess for early symptoms of ADRs, and guide safe medication practices (e.g., adequate hydration, posture changes, and activity tolerance). In addition, incorporating routine assessment of orthostatic blood pressure and other hemodynamic-related functional indicators, along with any recent near-fall episodes, into nursing encounters can help identify early signs of increased vulnerability to medication-related harm. These disease-specific approaches can help prevent ADRs and promote safer medication use among high-risk older adults.
These associations should be interpreted with caution because disease severity, treatment intensity, and detailed medication exposure were not incorporated into the models. Taken together, these findings highlight that older adults with multimorbidity require systematic, individualized approaches to medication safety. Regular medication review, vigilant monitoring, and multidisciplinary collaboration—including active involvement of family caregivers—are key components of effective medication management. Integrating disease-specific considerations into these processes may further enhance the impact of nursing interventions and reduce ADR risk in vulnerable populations.
Finally, this study found a clear trend of increasing risk of ADRs with a higher number of chronic diseases in older adults. These results suggest that the accumulation of chronic diseases may be an important contributor to the occurrence of ADRs. In particular, the finding that ADR risk more than doubled from three or more chronic conditions illustrates how the burden of multimorbidity can substantially heighten vulnerability. This heightened vulnerability is likely due not only to the cumulative effect of polypharmacy but also to the complex pathophysiological interactions among multiple conditions and the increased potential for drug–drug interactions. Thus, in clinical practice, the number of chronic diseases in older patients should be considered not merely as a diagnostic list but as a practical indicator of potential medication-related risk. This highlights the importance of more rigorous monitoring of polypharmacy and regular medication reviews as comorbidities accumulate. From a nursing perspective, older adults with three or more chronic conditions may warrant particular attention in medication safety interventions, with tailored education on safe medication use and multidisciplinary reviews serving as valuable strategies to prevent ADRs in this high-risk group.
This study has several limitations. First, because it used cross-sectional survey data, causal relationships between chronic diseases and ADRs cannot be established, and the temporal sequence between disease onset, medication exposure, and ADR occurrence cannot be confirmed, limiting causal inference. Second, the outcome was self-reported, which introduces recall and reporting bias; in older adults, cognitive or sensory decline may hinder recognition of mild events and lead to underestimation, and the reliability of the self-reported ADR measure is therefore uncertain. In this study, ADRs were operationally defined from a single yes/no item on side effects without clinical adjudication, and the lack of information on type, frequency, or severity means the measure may not capture the multidimensional nature of ADRs and may lead to misclassification of drug-related symptoms versus underlying disease manifestations; future work should use validated instruments or link surveys to clinical and pharmacy records to enhance the validity and reliability of ADR assessment. Third, we lacked data on disease severity, treatment intensity, medication dose and duration, and potential drug–drug interactions, as well as underlying disease burden and treatment complexity, leaving residual confounding. Finally, the sample comprised community-dwelling older adults in a single year and country, so generalizability to institutionalized populations or other settings is limited.
Despite these limitations, this study is significant because it systematically analyzed ADR risk according to both the types and number of chronic diseases among older adults using large-scale community-based data. Our findings identify high-priority groups for targeted interventions and underscore the critical, unique role of nurses in long-term patient monitoring and proactive symptom detection. Building on these results, future research could make an important contribution by developing and evaluating nurse-led intervention programs that address medication management, patient and family education, and tailored monitoring strategies to prevent ADRs among older adults with multimorbidity. Future research should employ long-term cohort designs and incorporate objective clinical evaluation data to more comprehensively identify factors contributing to ADR occurrence.
CONCLUSION
This study analyzed the risk of ADRs according to the types and number of chronic diseases in older adults, providing population-based evidence to promote medication safety in this population. Older adults with arthritis and spinal diseases, depression, or cardio-cerebrovascular diseases had a significantly higher risk of ADRs, and the likelihood of ADR occurrence increased progressively with the number of chronic conditions.
In interpreting these findings, it is important to recognize that ADR risks may vary depending on disease severity, treatment complexity, and patterns of medication use. Nevertheless, the present results identify specific chronic disease groups that require careful medication management and nursing assessment to reduce drug-related harm.
From a nursing perspective, integrating chronic disease type and burden into nursing assessments can support the development of individualized care plans that include medication education, lifestyle counseling, and early recognition and monitoring of ADRs. Strengthening collaboration among nurses, physicians, pharmacists, families, and caregivers can further enhance medication safety. In particular, nurses caring for older adults with arthritis and spinal diseases, depression, or cardio-cerebrovascular diseases should prioritize comprehensive medication review, monitoring for early signs of ADRs, and patient education to promote safe medication practices in these high-risk groups. A nursing-centered, integrated approach of this nature can help minimize drug-related harm, improve the quality of life of older adults, and provide a foundation for policy and educational initiatives aimed at fostering safe medication practices in aging populations.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation 
Print 
