Introduction

Falls, one of the major adverse events affecting older adults, lead to serious injuries. Hospitalizations because of fall-related injuries have been increasing (Moreland et al., 2021), and deaths because of falls have risen 30% between 2007 and 2016 (Burns & Kakara, 2018). The older adult population in South Korea is projected to roughly double from 16.5% in 2021 to 32.3% in 2038 (Statistics Korea, 2019). Moreover, falls are expected to become a more prominent issue, as they are considered a significant cause of mortality in older adults (Burns & Kakara, 2018). Research results point to a high risk of fall events during acute episodes of disease and accidental falls coupled with comorbidities leading to life-threatening consequences in older adults (K. Zhang et al., 2022). These authors also noted that older adults may become increasingly frail as they age, increasing the risk of falls and reducing their ability to survive the resultant injuries.

Falls may be prevented by timely screening for fall risk factors and implementing related interventions. Although several multifactorial interventions and tailored fall-prevention programs have been tested, a recent meta-analysis revealed that some of these interventions do not produce sufficient changes and that challenges because of barriers such as attrition and adherence remain (Oluwaseyi et al., 2018). Another meta-analysis suggested that local community characteristics must be considered to develop truly effective fall interventions (Hill et al., 2018).

The occurrence of falls among older adults differs by age and gender. Older adults in the oldest age group (over 75 years) and older women have been reported to face the highest risk of falls (Wu & Ouyang, 2017) and fall-related mortality (Alamgir et al., 2012). The causes of falls may differ in these groups from those in other groups and may reflect age- and gender-related changes in physical and mental functions. Specifically, decreasing limb-muscle strength and physical performance has been identified as a major risk factor for falls (Kawabata et al., 2021) and may play a crucial role in the fall risk faced by women and by individuals in the oldest age group, as these groups are most vulnerable to these factors (Nakano et al., 2014; Winger et al., 2021). Although regular exercise is currently encouraged for community-dwelling older adults to increase their motor skills and physical performance, it is unclear how much exercise is required to effectively prevent falls. Nevertheless, few studies have examined the physical condition and exercise habits of older adults, and even fewer have analyzed the effects of age and gender on fall risk factors in large older adult study samples.

In addition to physical status, other complex factors are also known to contribute to falls in older adults (Chen et al., 2021). Factors previously suggested include impaired physical condition, neuropsychological deficits, visual acuity, multimorbidity, number of medications, and dependency in activities of daily living (ADLs; Bloch et al., 2010). However, few researchers have analyzed fall risk factors considering multiple dimensions of health based on the biopsychosocial model targeting Korean older adults. The biopsychosocial model posits that health issues result from interactions among biological, psychological, and social factors (Wade & Halligan, 2017). As this model adds the social, psychological, and behavioral dimensions of health to the classic biomedical model (Engel, 1981), it may offer a favorable framework for conducting a systemic analysis of complex and multifaceted fall risk factors. In applying the biopsychosocial model, studies involving multifactorial analyses are needed to address the independent influence of each factor on falls stratified by age and gender to help researchers develop effective strategies tailored to the needs of different age and gender groups.

The aim of this study was to use a biopsychosocial model to investigate the prevalence of falls among older adults living in the community; to assess their exercise habits, lower-limb muscle strength, and physical performance; and to elucidate the significant fall-contributing factors by age and gender.

Methods Study Population

This cross-sectional study was based on data from the 2017 National Survey of Older Koreans. This nationwide, 3-year interval survey is regularly conducted by the Ministry of Health and Welfare in Korea. The 2017 survey covered 17 metropolitan areas and provinces in Korea using a proportional two-stage stratified cluster sampling method. First, the data were stratified by Korea's seven constituent metropolitan areas, nine constituent provinces, and Sejong (a self-governing city), with the nine provinces and Sejong city further stratified by urban (neighborhood) and rural (town and township) areas (Ministry of Health and Welfare [MOHW] & Korea Institute of Health and Social Affairs [KIHASA], 2017). The data were obtained via face-to-face interviews conducted by well-trained interviewers from June 12 to August 28, 2017. The 2017 National Survey of Older Koreans included 10,299 individuals aged ≥ 65 years living in standard housing. The raw data used in this study were obtained on December 5, 2020, from the Health and Welfare Data Portal (https://data.kihasa.re.kr/). Of the 10,299 responses, 226 proxy responses were excluded to increase data accuracy. Overall, the data from 10,073 older adults (response rate: 97.8%) were included in the final analysis. The study was approved by the institutional review board with which all of the authors were affiliated (200825-2A).

Measures

The biopsychosocial model proposed by Engel (1981) was used in this study to highlight the contributions of biological, psychological, and social factors to determining an individual's health concerns. After the use of meta-analysis and the application of the International Classification of Functioning on fall predictors in previous studies (Bloch et al., 2010; Soh et al., 2020), the selected fall-related factors were integrated into the biopsychosocial model. In this study, the fall-related biological factors were chronic diseases, number of medications, visual difficulties, ADL dependence, lower-limb muscle strength, and physical performance; the fall-related psychological factors were depression, cognitive ability, regular smoking, alcohol consumption in the last year, nutritional status, and exercise; and the fall-related social factors were educational level, annual income, living conditions, and instrumental ADL (IADL) dependence.

Demographic and health-related characteristics

The demographic characteristics considered in this study included age, gender, years of education, marital status, living conditions, and household income. The participants were divided into two age categories: young–old (< 75 years) and old–old (≥ 75 years; Yoshimura et al., 2013).

The health-related characteristics considered in this study included number of chronic diseases, number of medications currently taken, nutritional status, ADL and IADL dependence, level of depression, cognitive status, smoking status, and alcohol consumption in the past year. Nutritional status was evaluated using the Nutritional Screening Initiative checklist (Posner et al., 1993), which includes 10 items and a total checklist score range of 0–21. Each participant was asked to circle the item corresponding to their situation, and each item was assigned a weighted score ranging from 1 to 4. Circling no items resulted in a total checklist score of 0, whereas circling all 10 items resulted in a total checklist score of 21 (Posner et al., 1993), with 0–2 indicating adequate nutritional status, 3–5 indicating at-risk nutritional status, and ≥ 6 indicating malnourishment. ADL and IADL dependence statuses were assessed using the Korean version of the ADL scale (Won, Rho, Kim, et al., 2002) and the Korean version of the IADL scale (Won, Rho, SunWoo, et al., 2002). The Korean version of the ADL scale comprises seven questions on basic self-care abilities such as hygiene, bathing, dressing, eating, toileting, control of urination and defecation, and indoor activities. The Korean version of the IADL scale contains 10 questions on abilities related to grooming, going out, shopping, food preparation, housekeeping, laundry, transportation, using the telephone, self-medication, and handling finances. Each item is scored from 1 (completely independent) to 3 or 4 (completely dependent). Depression level was measured using the Geriatric Depression Scale-Short Form (GDS-SF), Korean Version (Kee, 1996). The GDS-SF Korean Version consists of 15 items in a binary response format (yes/no) with a total score ranging from 0 to 15 and higher scores indicating a more severe level of depression. The Cronbach's alpha of the GDS-SF has been calculated as .80 (Park et al., 2017) and was calculated as .89 in this study. Cognitive status was assessed using the Mini-Mental Status Examination for Dementia Screening (MMSE-DS; Seoul National University Bundang Hospital, 2009). The MMSE-DS consists of 19 items. Each item is weighted from 1 to 5 for a maximum score of 30, with higher scores indicating better cognitive status.

Falls

The participants were asked whether they had fallen (or slipped) within the past 12 months for any reason and regardless of whether the incident had resulted in injury. The respondents were considered to have experienced a fall if they answered “yes” to the question. Number of falls, reasons for falls, and related medical treatments were also recorded.

Lower-limb muscle strength

A five-times sit-to-stand test (FSTST) was performed to assess lower-limb muscle strength (F. Zhang et al., 2013). Older adults were instructed to stand up and sit down 5 times from a 45-cm-high chair or bed, with a score assigned between 1 and 4 (1 representing “unable to perform,” 2 representing “very difficult to perform,” 3 representing “slightly difficult to perform,” and 4 representing “able to perform without difficulties”) and higher scores indicating stronger lower-limb muscle strength.

Physical performance

Physical performance was evaluated using five items related to mobility from the Physical Functioning Scale developed by Lee et al. (2002). These items respectively assess ability to run 400 meters, climb 10 steps without a break, kneel or squat, reach out to an object overhead, and lift an object ≥ 8 kg. Each item is scored from 1 to 4 (1 representing “unable to perform,” 2 representing “very difficult to perform,” 3 representing “slightly difficult to perform,” 4 representing “able to perform without difficulties”), with higher score indicating better physical performance. The Cronbach's alpha of the physical functioning scale was previously calculated as .81 (H. Kim & Park, 2014).

Exercise

The question “Do you exercise regularly?” was asked to determine whether the participants exercised regularly. Exercise frequency and duration were respectively determined using the questions “How often do you exercise?” and “How long do you exercise in a single workout?” The answer to the exercise frequency question was scored from 0 to 2 (0 = none, 1 = 1–2 times per week, and 2 = ≥ 3 times per week). The answer to the exercise duration question was scored from 0 to 3 (0 = none, 1 = < 30 minutes, 2 = 30–59 minutes, and 3 = ≥ 60 minutes).

Procedures

On the basis of the 2014 National Survey of Older Koreans questionnaire, a panel of experts (25 professors and researchers in the fields of public health and gerontology) reviewed and provided comments for further refinement of the study instruments. The instruments were revised by the research team members based on recommendations from the panel of experts. The revised instruments were pretested with 45 older adults in Korea. After the pretest, some items were deleted, the order of some items was adjusted, and answer choices for the multiple-choice questions were added (MOHW & KIHASA, 2017).

Sixty trained interviewers and 15 supervisors were involved in collecting data. The interviewers were divided into 15 teams, with four in each group, and each team was supervised by one supervisor. The survey was conducted in a presampled survey area. All of the households in the survey area were visited by trained interviewers, and all residents aged ≥ 65 years living in those households were interviewed for data collection purposes. Data collection was continued until the requisite number of older adults in the survey area had completed the questionnaire (i.e., 10 older adults in urban areas and 20 older adults in rural areas; MOHW & KIHASA, 2017).

Data Analysis

IBM SPSS Statistics Version 25 for Windows (IBM Inc., Armonk, NY, USA) was used for the analysis. Descriptive statistics (means, standard deviations, frequencies, and percentages) were used to describe general and health-related characteristics. Chi-square tests were performed to analyze age- and gender-related differences in the participants' physical performance. Finally, logistic regression analyses were performed to identify the significant age- and gender-related factors associated with falls among the participants. Our logistic regression analysis included the fac­tors significant for fall in univariate analysis.

Results Demographic and Health-Related Characteristics of the Participants

The general characteristics of the study population are summarized in Table 1. The mean age was 73.9 years (SD = 6.54, range: 65–106 years), and 4,224 (41.9%) participants were in the old–old group (≥ 75 years old). Of the 10,073 participants, 5,787 (57.5%) were women. The average educational level was 7.18 years (SD = 4.59). Most participants were married, and 3,138 (31.2%) were widowed. Twenty-four percent (n = 2,416) lived alone. The mean annual income per household was ₩25,710,000 (equivalent to 24,000 USD). The average number of chronic diseases diagnosed by physicians and the total medications currently taken were 2.7 (SD = 1.84) and 4.1 (SD = 3.41), respectively. Thirty-four percent (n = 3,409) of the participants had visual difficulties. Of the sample, 9,369 (93.0%) and 7,766 (77.1%) participants were completely independent in terms of ADLs and IADLs, respectively. The average depression score was 4.1 (SD = 4.08), and the average MMSE-DS score was 25.2 (SD = 3.84). One thousand twenty-nine (10.2%) older adults were regular smokers, and the average number of times they consumed alcohol during the last year was 1.1 (SD = 2.09). The nutritional status was “adequate” for 4,102 (40.7%) of the participants.

Table 1 - Participant Demographic and Health-Related Characteristics (N = 10,073) Variable n % Median IQR Age (years; M and SD) 73.9 6.54 73.00 10  Young–old adults < 75 years 5,849 58.1  Old–old adults ≥ 75 years 4,224 41.9 Gender  Male 4,286 42.5  Female 5,787 57.5 Education (years; M and SD) 7.18 4.59 6.00 6 Marital status  Married 6,416 63.7  Widowed 3,138 31.2  Divorced 365 3.6  Other 154 1.5 Living alone  Yes 2,416 24.0  No 7,657 76.0 Household income (10k, KRW; M and SD) 2,571.9 2,244.32 1,888.62 2,202 Number of chronic diseases (M and SD) 2.7 1.84 3.00 3 Diseases diagnosed  Hypertension 5,945 59.0  Anemia 215 2.1  Benign prostrate  hypertrophy 893 8.9  Thyroid disease 328 3.3 Total number of medications (M and SD) 4.1 3.41 3.00 5 Visual difficulty  Yes 3,409 33.8  No 6,664 66.2 ADL  Completely independent 9,369 93.0  Help needed 704 7.0 IADL  Completely independent 7,766 77.1  Help needed ≤ 4 domains 1,549 15.4  Help needed ≥ 5 domains 758 7.5 Depression (M and SD) 4.1 4.08 3.00 15 MMSE† (M and SD) 25.2 3.84 26.00 5 Regular smoking  Yes 1,029 10.2  No 9,044 89.8 Number of drinks during the last 1 year (M and SD) 1.1 2.09 0.00 1 Nutritional status  Adequate 4,102 40.7  At risk 3,646 39.2  Malnutrition 2,025 20.1 Fall experience during the last 1 year  Yes 1,585 15.7  No 8,488 84.3 Number of falls (n = 1,585)  1 time 1,035 65.3  ≥ 2 times 550 34.7 Treatment because of fall injury (n = 1,585)  Yes 1,032 65.1  No 553 34.9 Fall reason (n = 1,585)  Slippery surface 427 26.9  Wobbly leg 306 19.3  Tripping over something 266 16.8  Losing one's footing 224 14.2  Sudden dizziness 184 11.6  Steep slope 59 3.7  Others 119 7.5

Note. IQR = interquartile range; ADL = activity of daily living; IADL = instrumental activity of daily living; MMSE = Mini-Mental Status Examination.

Of the 10,073 participants, 1,585 (15.7%) had experienced falls during the last year, 550 (34.7%) of whom had fallen more than twice. The major causes of falls were slippery surfaces and wobbly legs. One thousand thirty-two (65.1%) of those who fell received treatment because of their fall-related injury during the previous year.

Lower-Limb Muscle Strength, Physical Performance, and Exercise by Age and Gender

In terms of lower-limb muscle strength, 8,055 (80.0%) of the participants were able to sit and stand 5 times without difficulty (Table 2). Significantly fewer older adults in the old–old group (p < .001 for both men and women) and female group (p < .001 for both young and old age groups) were able to finish the test. In terms of physical performance, 1,589 (15.8%) and 5,741 (57.0%) participants were able to run 400 m and climb 10 stairs without difficulty, respectively. Furthermore, 8,743 (86.8%) participants were able to reach out and touch objects overhead, and 7,226 (71.7%) were able to lift or move objects weighing ≥ 8 kg. Significantly fewer older adults in the old–old group (p < .001 in both men and women) and female participants (p < .001 in both young–old and old–old age groups) were able to perform either activity. Of the entire sample, 6,854 (68.0%) exercised regularly, 6,096 (60.5%) exercised more than 3 times a week, and 3,352 (33.3%) exercised more than 60 minutes per session. The young–old group reported significantly more regular exercise (p = .005 in men, p < .001 in women), more frequent exercise sessions (p < .001 in both men and women), and longer exercise periods (p < .001 in both men and women) than the old–old group. Older women exercised significantly less than their male counterparts in the old–old group in terms of regularity (p < .001), frequency (p < .001), and duration (p < .001).

Table 2 - Descriptive Statistics of Lower Limb Muscle Strength, Physical Performance, and Exercise, by Age and Gender (N = 10,073) Variable Total Young–Old Group
(< 75 Years Old) Old–Old Group
(≥ 75 Years) n % Male Female Male Female n % n % n % n % Five-times sit-to-stand test  Unable to stand 80 0.8 12 0.5*** 12 0.4***,††† 18 1.0 38 1.5†††  Having slight difficulties 1,938 19.2 142 5.6 468 14.1 344 19.6 984 39.8  Able to stand 8,055 80.0 2,382 93.9 2,833 85.5 1,389 79.3 1,451 58.7 Physical performance  Run 400 m (n = 10,057)   Unable to perform 1,888 18.8 196 7.7*** 568 17.2***,††† 328 18.8 796 32.2†††   Very difficult to perform 3,425 34.1 524 20.7 1,112 33.6 643 36.8 1,146 46.4   Having slight difficulties 3,155 31.4 938 37.0 1,161 35.1 577 33.0 479 19.4   Able to perform 1,589 15.8 877 34.6 465 14.1 199 11.4 48 1.9  Climb 10 stairs without rest   Unable to perform 334 3.3 29 1.1*** 69 70.4***,††† 63 3.6 173 7.0†††   Very difficult to perform 1,212 12.0 108 4.3 332 10.0 192 11.0 580 23.5   Having slight difficulties 2,786 27.7 315 12.5 913 27.5 508 29.0 1,050 42.4   Able to perform 5,741 57.0 2,083 82.1 2,000 60.3 988 56.4 670 27.1  Kneeling or squatting   Unable to perform 423 4.2 29 1.1*** 159 4.8***,††† 58 3.3 177 7.2†††   Very difficult to perform 1,141 11.3 114 4.5 375 11.3 159 9.1