Abuse and risk factors among community-dwelling elderly in South Korea during COVID-19

ABSTRACT

Elder abuse is a serious social problem in South Korea. This study aimed to estimate the prevalence and identify risk factors of elder abuse during the coronavirus disease 2019 (COVID-19) pandemic. By using the 2020 Living Profiles of Older People Survey in Seoul Metropolitan City and considering sociodemographic, health-related, social support, and isolation factors as independent variables, a logistic regression model was established. Among 3,106 samples, the overall rate of elder abuse was 8.5%, with emotional abuse being the most frequent. Social isolation (OR = 1.47, p = .018) and recognition of abuse (OR = 0.90, p = .003) were associated with having experienced abuse. Seniors with a greater number of diseases, higher levels of dependency, depressive symptoms, and/or cognitive impairments were more likely to suffer abuse, as were those who smoked or drank alcohol. To prevent elder abuse by social isolation during and after COVID-19, more active health promotion programs and healthy aging policies should be implemented.

KEYWORDS:

• COVID-19
• elder abuse
• social isolation
• depressive symptoms
• south Korean elders

Introduction

Elder abuse is a growing trend worldwide, posing a serious threat to public health and increasing risk factors for older adults’ health (Koga et al., 2020). According to a 2021 World Health Organization (WHO) report, one in six older adults experiences abuse. The increasing older population and social changes in many countries have produced various types of elder abuse (World Health Organization, 2021). This global phenomenon differs from one nation to another (Yon et al., 2017). The WHO and United Nations (UN) insist that policies related to health and social welfare, as well as laws or other systems, should be established to prevent further risks to elders (Pillemer et al., 2016; World Health Organization, 2014).

In South Korea, the population of individuals aged 65 and above is expected to rise to 11.59 million by 2027 (from 7.07 million in 2017), with the proportion of the elderly surging to 25.0% by 2030 (from 15.7% in 2020), which will make it a super-aged society (Statistics Korea, 2020). According to the annual report of the Korean Elder Protection Agency (2020), the number of elder abuse cases increased from 5,188 in 2018 to 5,243 in 2019, rising significantly to 6,259 cases in 2020. It is noteworthy that 5,505 or 88% of elder abuse cases occurred at home, which was the highest proportion of all cases; moreover, the greatest proportion of elder abuse cases – 2,060 or 32.9% – occurred in households where the elderly lived with their children (Korean Elder Protection Agency, 2020). For this reason, social changes such as smaller family structures and increasing social participation among family members have placed a care burden on children with dependent parents, making it increasingly difficult to support them and, thereby, inducing conflicts (Kim et al., 2016). The analysis of the various causes and effects of elder abuse has resulted in the topic becoming a critical research area, especially in the fields of gerontology and public health.

To prioritize elderly care in the community, long-term care insurance policies have been introduced and the number of medical or nursing facilities has been increasing dramatically. However, new forms of abuse are also emerging (H. M. Lee & Lee, 2020). Therefore, in order to prevent elder abuse, the Korean government established an institution specializing in the protection of older adults in the year 2020, while also encouraging victims of abuse to seek counseling and urging those who know of abuse victims to report the abuse. However, the effect of this institution has not sufficiently reduced the abuse rates (Korean Elder Protection Agency, 2020). Older adults are more likely to face further abuse as a result of the new social situations as well as economic and social exclusion. As elder abuse in South Korea can incur enormous social costs in the future, it is necessary to prioritize this problem.

Elder abuse, which poses a serious threat to older adults’ physical and mental health, is prevalent in modern society and places an additional cost on individuals, communities, and societies (Dong, 2015). It is a burden on the already weakened state of mental health among older adults due to sadness, depressive symptoms, and helplessness, as a result of being alienated from society, work, or home due to the deterioration of physical function (Mawar et al., 2018; Park, 2015). It not only worsens their functional impairment but also leads to early death. Moreover, older people often experience depressive symptoms (Koga et al., 2020; Park & Chung, 2017), loneliness (Zhang et al., 2018), and suicidal ideations (Seo & Ko, 2011; S. Y. Lee & Atteraya, 2019), which can result in extended periods of poor health. Those who suffer severe physical violence also tend to become more violent, obsessed, or socially maladapted, while emotional abuse can cause psychological disorders such as fear or anxiety in social relationships (Park, 2015). Older people who are unable to pursue their individual needs and retain their individuality may be inhibited from engaging in health-promoting behaviors, thereby worsening their potential health problems. From a public health perspective, this is of critical importance (Kim, 2020).

In many abuse cases, the caregivers’ experiences of difficulties with older people are associated with the latter’s lack of independence to perform physical functions, being disabled, and psychological characteristics such as dementia and cognitive impairment (Dong, 2015; Dong et al., 2011). Family disharmony, conflicts with people around them, low-income level, and environmental aspects such as living alone may be significant factors for elder abuse (Day et al., 2017). Although there may not be any evidence of direct abuse, elder abuse can be indirectly perpetrated by avoiding – and thereby “neglecting” – older people; this issue requires careful consideration as it lowers the quality of life and life expectancy of older adults by depriving them of fundamental support and medical treatment (Jeong & Park, 2016). Various kinds of personal or social factors such as substance abuse, abuse dependence, geographic location, negative prejudice against older adults, and cultural community norms can also impact elder abuse (Pillemer et al., 2016). In particular, depending on the times and social circumstances, problems such as elder abuse or neglect may be overlooked.

To combat the global spread of the coronavirus disease 2019 (COVID-19), many countries implemented regulations to restrict social activities through lockdowns, stay-at-home orders, and social distancing. Insufficient social ties and networks with family or friends caused by these restrictions may have resulted in older adults experiencing emotional abuse and neglect (Van Orden et al., 2021). They cannot request help because they are isolated and physically dependent on family members, which exerts further personal, social, and economic hardships (Makaroun et al., 2020). Moreover, older people with multiple diseases are exposed to abuse because of their insecure economic conditions and physical weakness, which prevent them from living on their own (Makaroun et al., 2020). Because families caring for older adults with physical and social problems are isolated due to changes in their external environment, they tend to experience an internal socioeconomic impact, which inevitably leads to abuse of the weak (Du & Chen, 2021). Thus, it is important to not only understand the experience of elder abuse but also discuss countermeasures to prevent further social problems.

Aims and hypotheses

As the risk factors in elder abuse have not been sufficiently studied previously, research on elder abuse during COVID-19 has not consistently identified individual sociodemographic factors. The purpose of this study is, thus, to identify the prevalence of abuse experience and abuse-related factors among the community-living elderly during the COVID-19 pandemic. Specifically, the objectives are as follows:

(1) To identify the general and disease-related characteristics of the subject.

(2) To determine whether the subject has experienced abuse.

(3) To identify the type of perpetrator.

Moreover, we hypothesize that:

1. After controlling for the subject’s demographic characteristics, social and protective factors, such as the number of family members and friends, will lower the odds of abuse.

(2) After controlling for the subject’s demographic characteristics, social isolation due to COVID-19 will increase the odds of abuse.

Methods

Study design

The ecological model’s multiple levels of influence on elder abuse – which is complex, multi-layered, and dynamic in terms of personal, family, and social interactions – has been recognized (Blundell et al., 2017). Applying the ecological framework to elder abuse is useful in determining the individual, family, community, and societal factors for the analysis and understanding of complex social issues. In addition to considering the individual characteristics of the elderly as the microsystem; the relationship with their caregivers including family, friends, and neighbors was considered as the mesosystem; and social isolation due to COVID-19 was considered as the macrosystem (Schiamberg & Gans, 1999).

We selected variables at the micro level (sociodemographic factors, individual health status and behavior, and recognition of abuse), meso level (number of family members and friends), and macro level (social isolation due to COVID-19; Kwon, 2007).

Study subjects

This study involved secondary data analysis of the 2020 Living Profiles of Older People Survey in Seoul (LPOPS_SEOUL), conducted biannually since 2012 by the Seoul Metropolitan City (Seoul Welfare Foundation, 2020). It provides national data on the eight categories of life and living, health status, living environment, working activity, leisure activity, respect protection, care, and sociodemographic characteristics based on the recommendations of the Global Network for Age-friendly Cities and Communities (Pfeiffer, 1975; Plouffe et al., 2016). The survey was conducted from July 21 to October 23, 2020, and, while the target sample was 3,000, the final number of participants included 3,106 senior citizens over the age of 65 living in Seoul Metropolitan City who agreed to participate in the 2020 LPOPS_SEOUL. Data were collected using the square root proportional distribution method by the number of categorized Seoul citizens based on gender, age, and region. The limitation of such a sampling method is that it excludes those living in facilities or institutions. However, it suggests directions for future policies related to the elderly by adding questions related to the latest national issues at the time. The questionnaire was completed using computer-assisted face-to-face interviews. Participants’ informed consent was obtained by trained interviewers. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of the Korea National Institute for Bioethics Policy (2012–1622-003).

Measurements

Dependent variable

The Welfare of Senior Citizens Act defines elder abuse as any act of physical or verbal abuse, financial exploitation, lack of caregiving, or financial neglect of senior citizens (Ministry of Health and Welfare, 2020). We considered elder abuse to include any such behavior toward individuals aged 65 years or older in the last 12 months. If the respondents answered “yes” to one of the following seven items, they were considered to have experienced elder abuse:

• (3) Physical abuse

  • “Has anyone tried to hurt or harm (e.g., push, hit, etc.) you?”

  • “Has anyone tried to lock you in a room or physically restrain you?”

(4) Psychological abuse

• “Has anyone yelled at you, called you names, or insulted you?”

• “Has anyone disregarded your opinions in making decisions related to you?”

(5) Financial exploitation

• “Has anyone taken your income (basic pension, etc.) without your consent?”

(6) Neglect

• “Does someone in your family refuse to care for you (e.g., by refusing to provide food, drink, shelter, etc.) when in need?”

(7) Sexual abuse

• “Has anyone tried to sexually assault, harass, or engage in any sort of sexual misconduct with you?”

Control variables

Based on the literature review, control variables such as sociodemographic factors (gender, age, type of household, job, and socioeconomic status (SES)) were considered (Cho, 2019; Dong, 2015). Participants were divided by age into two groups – young-old (65–74 years) and old-old (≥75 years) – and their current employment status (yes or no) was also assessed. They were also categorized into two groups based on household type – living alone and living with someone.

Independent variables

Psychological factors (depressive symptoms), self-rated health, and health risk behaviors (cigarette smoking, frequency of alcohol consumption, frequency of exercise) were also considered based on previous studies (Cho, 2019; Dong, 2015). Physical health factors (determined based on the Korean version of the Instrumental Activities of Daily Living or K-IADL (number of comorbidities), cognitive function (cognitive impairment), and social and protective factors (social isolation due to COVID-19 and number of family members and friends with whom participants were in contact) were also measured (Eslami et al., 2017; Makaroun et al., 2020), as was recognition of abuse.

Self-rated health and SES were measured through the following questions, respectively:

• (8) “How would you rate your health in general?”

(9) “How would you rate your economic status in general?”

We categorized the five responses into three groups: good (“very good” or “good,”), moderate (“fair”), and poor (“bad” or “very bad”).

Depressive symptoms were assessed using the Korean version of the Geriatric Depression Scale-Short Form (SGDS-K), originally developed and translated into Korean by Bae and Cho (2004). The SGDS-K comprises 15 items with the response options of “yes = 1” or “no = 0.” In this study, a score ≥8 was taken as indicative of significant depressive symptoms. The SGDS-K has satisfactory validity and reliability (Cronbach’s alpha of 0.90; Bae & Cho, 2004).

Cigarette smoking, frequency of alcohol consumption, and frequency of exercise were assessed based on participants’ responses to the following three questions, respectively:

• (10) “Have you smoked a cigarette in the past 30 days?”

(11) “How often have you had a drink in the past 12 months?”

(12) “How often have you exercised in the past 7 days?”

The responses to the question about smoking were categorized as “no” or “yes.” The responses to the question about frequency of alcohol consumption were categorized in terms of times per year, month, or week, if at all (none, 1–11 times/year, 1–3 times/month, and more than 1 time/week). The responses to the question about exercising were categorized as no exercise, 1–2 days/week, 3–4 days/week, and more than 5 days/week.

The K-IADL assesses physical dependency. Respondents were asked whether they needed assistance when performing 10 different instrumental activities (personal hygiene, grooming, housekeeping, preparing meals, making and receiving phone calls, managing money, taking medications as prescribed, use of transportation, shopping, and doing laundry; Won et al., 2002). The scores were categorized as having no limitations in daily activities (K-IADL = 0) and having limitations in daily activities (K-IADL ≥ 1). They ranged from 10 ~ 33 points, with higher scores indicating greater physical dependency. The K-IADL has satisfactory validity and reliability (Cronbach’s alpha of 0.94; Won et al., 2002); in this study, the Cronbach’s alpha was 0.87.

Cognitive impairment was assessed using the 10-item Short Portable Mental Status Questionnaire (SPMSQ; Pfeiffer, 1975). The items included:

• (13) “What is today’s date, month, and year?”

(14) “What is the name of this place?”

(15) “How old are you?”

We considered SPMSQ scores of 7–10 to reflect normal cognitive function and scores of 0–6 to reflect cognitive impairment. The SPMSQ has satisfactory validity and reliability (Cronbach’s alpha of 0.90).

Participants’ comorbidities were self-reported. The number of comorbidities was assessed with the question: “Have you ever been diagnosed for more than three months with any of the chronic diseases listed below such as hypertension, cerebrovascular disease, chronic pulmonary disease, congestive heart failure, dementia, diabetes, etc.?” (Katz et al., 1996).

Recognition of abuse was assessed with the question “Do you consider the following to be abuse?” Nine items related to physical and psychological abuse, financial exploitation, neglect, and sexual assault were listed with response options of “yes = 1” or “no = 0.”

Relationship with children, relationship with friends and neighbors, and social isolation were evaluated to assess social and protective factors. Number of family members and number of friends were calculated based on the following questions, respectively:

• (16) “How many family members/relatives do you see or hear from regularly?”

(17) “How many family friends do you see or hear from regularly?”

If the respondents answered “yes” to one of the following two items, they were considered to experience social isolation due to COVID-19:

• (18) “During the COVID-19 pandemic do you find it difficult to contact family or friends?”

(19) “During the COVID-19 pandemic do you find it difficult to participate in social activities?”

Statistical analysis

T-tests and chi-square tests assessed the difference between abused and non-abused elderly based on the independent variables. Logistic regressions and hierarchical modeling were used to examine and determine the factors affecting abuse experience within the past 12 months. Hierarchical regressions were performed in order to evaluate whether one or more variables statistically predict(s) a criterion variable while statistically “controlling for” sociodemographic variables. For example, after statistically controlling for socioeconomic status and level of educational attainment, does age and gender predict Korean elders’ abuse? Statistical significance in this study was defined as a p-value of p < .05, and adjusted p-values with odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using SPSS Statistics 27.0 (IBM Corp, Armonk, NY). To ensure unbiased national estimates, sampling weights were computed for the participants to reflect a representative sample of Seoul Metropolitan City’s adults aged 65 years and older. The variance inflation factor (VIF) test is used to assess multicollinearity in our regression model. VIFs between 1.1 and 2.0 suggest that the correlation is not severe enough to warrant corrective measures. Hierarchical regressions were used to evaluate whether one (or more) variables statistically predict(s) a criterion variable while controlling for sociodemographic variables. The analysis was carried out in two stages through hierarchical regression: In step 1, demographic factors were used as control variables and, in step 2, independent variables were added.

Results

Sociodemographic factors

Of the 3,106 community-dwelling elderly participants, 1,375 were male and 1,731 were female; the majority had poor SES (N = 2,181, 70.2%). One-third had depressive symptoms (N = 1,126, 36.3%) and did not drink alcohol (N = 1,135, 36.5%). The abused were more likely to report cognitive impairment (p < .001), depressive symptoms (p < .001), and physical dependency (p < .001), while those who were not abused were more likely to have closer ties with family (p < .001) and a higher recognition of abuse score (p = .031; Table 1).

Table 1. Sociodemographic and health-related characteristics of older adults who did and did not suffer abuse in Korea (N = 3,106).

Factors		N (%)/M± SD										p-value*
		Total (N = 3,106)			Not abused (N = 2,839)			Abused (N = 267)			χ^2 or t
		n		%	n		%	n		%
Gender	Male	1,375	(44.3)		1,244	(90.5)		131	(9.5)		2.72	0.099
	Female	1,731	(55.7)		1,595	(92.1)		136	(7.9)
Age (tears)	65–74	1,902	(61.2)		1,759	(92.5)		143	(7.5)		7.25	0.007*
	≥75	1,204	(38.8)		1,080	(89.7)		124	(10.3)
Type of household	Living together	2,063	(66.4)		1,910	(92.6)		153	(7.4)		10.89	0.001*
	Living alone	1,043	(33.6)		929	(89.1)		114	(10.9)
Job	No	2,159	(69.5)		1,968	(91.2)		191	(8.8)		0.57	0.452
	Yes	947	(30.5)		871	(92.0)		76	(8.0)
SES	Poor	2,181	(70.2)		1,961	(89.9)		220	(10.1)		32.86	<0.001*
	Moderate	829	(26.7)		796	(96.0)		33	(4.0)
	Good	97	(3.1)		83	(85.6)		14	(14.4)
Self- rated health	Poor	491	(15.8)		407	(82.9)		84	(17.1)		62.77	<0.001*
	Moderate	1,467	(47.2)		1,343	(91.5)		124	(8.5)
	Good	1,148	(37.0)		1,089	(94.9)		59	(5.1)
Cognitive impairment	No	3,023	(97.3)		2,789	(92.3)		234	(7.7)		105.40	<0.001*
	Yes	83	(2.7)		50	(60.2)		33	(39.8)
Smoking	No	2,546	(82.0)		2,351	(92.3)		195	(7.7)		15.79	<0.001*
	Yes	560	(18.0)		488	(87.1)		72	(12.9)
Depressive symptoms	No	1,980	(63.7)		1,873	(94.6)		107	(5.4)		70.83	<0.001*
	Yes	1,126	(36.3)		966	(85.8)		160	(14.2)
Social isolation due to COVID-19	No	2,173	(70.0)		1,974	(90.8)		199	(9.2)		2.904	0.088
	Yes	933	(30.0)		865	(92.7)		68	(7.3)
Frequency of alcohol consumption	None	1,135	(36.5)		1,032	(90.9)		103	(9.1)		37.11	<0.001*
	1–11 times/year	497	(16.0)		451	(90.7)		46	(9.3)
	1–3 times/month	847	(27.3)		811	(95.7)		36	(4.3)
	More than 1 time/week	626	(20.2)		544	(86.9)		82	(13.1)
Frequency of exercise	No exercise	579	(18.6)		485	(83.8)		94	(16.2)		74.77	<0.001*
	1–2 days/week	1,161	(37.4)		1,050	(90.4)		111	(9.6)
	3–4 days/week	389	(27.0)		794	(94.6)		45	(5.4)
	More than 5 days/week	526	(16.9)		509	(96.8)		17	(3.2)
K-IADL	10–29	11.1 ± 2.44			11 ± 2.32			12.16 ± 3.33			5.55	<0.001*
Number of comorbidities	0–11	2.41 ± 1.66			2.33 ± 1.63			3.23 ± 1.77			7.91	<0.001*
Number of family members	0–39	2.02 ± 5.88			2.14 ± 6.04			0.69 ± 3.49			−6.03	<0.001*
Number of friends	0–24	0.36 ± 2.2			0.35 ± 2.18			0.48 ± 2.46			0.95	0.342
Recognition of abuse	0–9	8.17 ± 1.68			8.19 ± 1.69			7.96 ± 1.47			−2.16	0.031*

Note: *p < .05; M = Mean, SD = Standard Deviation; K-IADL = Korean version of the Instrumental Activities of Daily Living; SES = Socioeconomic status

Types and perpetrators of elder abuse

In total, 267 elderly reported 358 total cases of abuse, averaging 1.35 cases per person. Perpetrators included spouses; offspring or offspring’s spouses; neighbors, friends, roommates, relatives, and others; and institutional workers. Offspring or offspring’s spouse (N = 128) constituted the highest proportion of perpetrators, at 35.8%, followed by neighbors, friends, roommates, relatives, and others (N = 97, 31.0%). There were 97 spouse-related cases (27.1%) and 22 institutional worker-related cases (6.1%; Table 2).

Table 2. Types of abuse and perpetrators (N = 358).

Types of abuse	Types of perpetrators
	Spouse		Offspring or offspring’s spouse		Neighbors, friends, roommates, relatives, others		Institutional workers		Total
	n	%	n	%	n	%	n	%	n	%
Physical abuse	1	(0.3)	3	(0.8)	17	(4.7)	0	(0.0)	21	(5.9)
Psychological abuse	94	(26.3)	111	(31.0)	81	(22.6)	13	(3.6)	299	(83.5)
Financial exploitation	0	(0.0)	6	(1.7)	9	(2.5)	0	(0.0)	15	(4.2)
Neglect	2	(0.6)	8	(2.2)	4	(1.1)	9	(2.5)	23	(6.4)
Total	97	(27.1)	128	(35.8)	111	(31.0)	22	(6.1)	358	(100.0)

Note: Institutional workers = medical personnel, senior welfare facility workers, and workers in other institutions.

Factors influencing elder abuse

Compared to the analysis of the presented model 1, the explanatory power (Nagelkerke’s R²) increased from 4.3% to 19% in the regression analysis including the independent variables added in step 2. Specifically, Model 1 revealed that gender was significant (OR = 0.76 (CI 0.58–0.99), p = .048) and SES was partially significant (OR = 0.41 (CI 0.27–0.60), p < .001). Model 2 revealed that gender was insignificant (OR = 0.92 (CI 0.65–1.30), p = .641) and the old-old group was more likely to experience elder abuse than was the young-old group (OR = 0.64 (CI 0.45–0.90), p = .011). Moderate SES was associated with elder abuse (p < .001). Those with depressive symptoms (OR = 1.65 (CI 1.20–2.27), p = .002), greater physical dependency (OR = 1.10 (CI 1.05–1.16), p < .001), cognitive impairment (OR = 5.79 (CI 3.34–10.05), p < .001), a greater number of comorbidities (OR = 1.22 (CI 1.12–1.33), p < .001), and those who smoked (OR = 1.70 (CI 1.13–2.56), p = .011) or consumed alcohol 1–3 times per month (OR = 0.54 (CI 0.34–0.84), p. = .006), were more likely to experience elder abuse than were those who did not.

Those who reported having a greater number of family members (OR = 0.96 (CI 0.92–0.99), p = .027) and experiencing social isolation due to COVID-19 (OR = 1.47 (CI 1.07–2.01), p = .018) were more likely to experience elder abuse than were those who had a smaller number of family members and had not experienced social isolation. Those with a higher score of recognition of abuse (OR = 0.90 (CI 0.83–0.97), p = .003) were less likely to experience elder abuse than were those with a lower score (Table 3).

Table 3. Logistic regression analysis of factors associated with elder abuse (N = 267).

Factor	Model 1				Model 2
		OR	95% CI	p-value	OR		95% CI	p-value*
Gender	Male	1			1
	Female	0.76	(0.58–0.99)	0.048	0.92	(0.65–1.3)		0.641
Age (years)	65–74	1			1
	>75	1.18	(0.88–1.59)	0.277	0.64	(0.45–0.9)		0.011*
Type of household	Living together	1			1
	Living alone	0.77	(0.57–1.02)	0.069	1.28	(0.91–1.8)		0.153
Job	No	1			1
	Yes	1.10	(0.79–1.53)	0.559	1.22	(0.84–1.77)		0.287
SES	Poor	1			1
	Moderate	0.41	(0.27–0.60)	<0.001*	0.38	(0.25–0.57)		<0.001*
	Good	1.60	(0.88–2.91)	0.122	2.52	(1.31–4.85)		0.006*
Self-rated health	Poor				1
	Moderate				0.79	(0.54–1.14)		0.201
	Good				0.67	(0.41–1.09)		0.103
K-IADL	10–29				1.1	(1.05–1.16)		<0.001*
Cognitive impairment	No				1
	Yes				5.79	(3.34–10.05)		<0.001*
Number of comorbidities	0–11				1.22	(1.12–1.33)		<0.001*
Smoking	No				1
	Yes				1.7	(1.13–2.56)		0.011*
Frequency of	None				1
alcohol consumption	1–11 times/year				1.1	(0.73–1.67)		0.649
	1–3 times/month				0.54	(0.34–0.84)		0.006*
	More than 1 time/week				1.53	(0.99–2.37)		0.056
Frequency of exercise	No exercise				1
	1–2 days/week				0.7	(0.5–0.97)		0.035*
	3–4 days/week				0.39	(0.26–0.58)		<0.001*
	More than 5 days/week				0.19	(0.1–0.34)		<0.001*
Depressive symptoms	No				1
	Yes				1.65	(1.2–2.27)		0.002*
Number of family members	0–39				0.96	(0.92–0.99)		0.027*
Number of friends	0–24				1.1	(1.03–1.17)		0.005*
Recognition of abuse	0–9				0.9	(0.83–0.97)		0.003*
Social isolation due to	No				1
COVID-19	Yes				1.47	(1.07–2.01)		0.018*

Note: * p < 0.05; OR = Odds ratio; CI = Confidence interval; K-IADL = Korean version of the Instrumental Activities of Daily Living; SES = Socioeconomic status

Discussion

This study reflected the current status and risk factors of elder abuse as a result of the social isolation during the COVID-19 pandemic in Seoul. The main findings including general characteristics and factors are discussed below along with findings from previous studies.

Sociodemographic factors of elder abuse

First of all, there was no significant difference between males and females in terms of the general characteristics of elder abuse experience. This result differs from that of a previous study (Jeon et al., 2019) conducted in South Korea before the COVID-19 pandemic, which showed that female elders were more vulnerable than were male elders. In Korea, because there are more women living alone, the conclusion that they are more vulnerable should be interpreted with caution. Despite findings from previous scoping reviews conducted in other countries also drawing the same conclusions (Du & Chen, 2021; Pillemer et al., 2016), the COVID-19 situation may not have a significant effect on gender in terms of abuse or may vary depending on the understanding of the sociocultural status of each elderly male and female. Thus, empirical conclusions should be drawn through further studies in the future.

The fact that elder abuse differs depending on individual participants’ characteristics or where they live are findings that are reflected in some previous studies. Unlike other studies, this study reclassified and analyzed the age groups of the elderly and found a significant difference. By categorizing age groups into young-older adults (below 75 years) and old-older adults (over 75 years), we discovered that the younger group experienced more abuse than did the older group. Young-old adults who are still physically active could have a poor relationship with family and neighbors due to their interfering or high maintenance nature. With regard to economic status, most related studies consistently report that the lower the economic level of the victim, the higher the risk of elder abuse, confirming the economic condition of the victim as a major influencing factor for elder abuse (Kim et al., 2016). Contrary to previous studies, the results of this study showed that the high-income group experienced a lot of abuse. It can be inferred that some abuse occurred to the elderly who experienced job loss and reduced income due to the COVID-19 pandemic. However, further research is needed because a comparative study has not yet been conducted to determine the factors that lead to the abuse of these young-older adult groups. This needs to be cautiously studied because victims’ age groups and effects of isolation may differ among countries.

However, our finding that elderly people with high and moderate SES experience more abuse differs from previous findings that elderly people with lower SES experience more abuse and neglect in countries such as Mexico, Ireland, and India (Pillemer et al., 2016). This can be interpreted as meaning that people with good finances may feel more exploited and usually have more interactions with others that may induce further abuse. A study by Du and Chen (2021) conducted in China revealed a somewhat different conclusion that the higher the economic level, the less abuse the elderly experience. Older adults with a moderate SES may have felt obligated to support their family despite their situation worsening during the pandemic. Therefore, the fact that even the elderly who are financially stable can experience abuse should not be overlooked. However, as with other factors, it is necessary to consider the possibility that the SES of the respondents was interpreted differently depending on their place of residence since this study only focused on one city in South Korea.

Psychological abuse has remained the most common type of abuse in South Korea (Park, 2015; Seo & Ko, 2011). This kind of abuse also includes perpetrators yelling and swearing at the elders. Elderly abuse and mental health affect the quality of life and mortality of older adults, and their association has been extensively and comprehensively discussed by previous researchers (Dong et al., 2013). Therefore, while the pandemic itself can adversely affect the mental well-being of the elderly and their families, the resulting abuse during the pandemic will also inevitably take its toll on their psychological status, which could ultimately lead to illness and death. Therefore, it is necessary to actively assess the mental status of and formulate mental health interventions for the elderly and their families that can be implemented during the COVID-19 pandemic.

Health-related factors of elder abuse

Regarding the factors that influence elder abuse, older adults who are dependent on others for their IADL experience more abuse (Han & Mosqueda, 2020). On the other hand, older adults with cognitive impairment suffer more abuse than do those who do not suffer cognitive impairment, and participants with depressive symptoms tend to experience greater abuse, showing that the more depressed older adults are, the more abuse they experience (Evandrou et al., 2017; Mawar et al., 2018; Pillemer et al., 2016). Accordingly, it is necessary to approach those who are clinically diagnosed as depressed as those who can potentially be abused and consider this to solve or prevent abuse before it gets serious. During the ongoing COVID-19 pandemic, because social isolation worsens depressive symptoms and emotional abuse, as has been reported in other Asian countries such as China (Du & Chen, 2021; Makaroun et al., 2020), it is important to consider formulating relevant policies and programs for those who experience such psychological conditions. Moreover, depressive symptoms and isolation in this crisis is a common experience among not only older adults who are abused but also the people around them. Thus, active strategies need to be considered for a practical resolution.

This study explains health-related behavioral factors such as smoking, alcohol consumption, and exercise and their relationship with elder abuse. Although alcohol consumption is commonplace in social interactions with others, an older person who drinks too much can be a burden on people around them, which can result in elder abuse. Conversely, the more abused the older adults, the higher is the probability that they will drink excessively. Older adults who frequently exercise find it easier to cope with their daily life; however, the lack of exercise among older adults can be a risk factor for abuse (Carmona-Torres et al., 2019). Older adults should be encouraged to engage in more exercise and participate in activities to prevent abuse. Moreover care for abused elders should be provided so that they can live healthy lives and reduce their caregivers’ burden.

Sociocultural factors of elder abuse

Each risk factor may reveal different results from specifically collected data, depending on the type of abuse. However, we recategorized original data similar to the World Health Organization’s (2021) analysis for comparison. In this study, the consequences of abuse by facility workers were considered along with the expansion of long-term care services in South Korea and global concerns about institutional abuse (Beaulieu et al., 2020). Elder abuse in long-term care facilities occurs frequently in various forms and has become a critical problem in many countries. Although countermeasures have long been discussed, the lack of facilities, cautiousness due to research ethics, and inadequate prior research have presented obstacles in approaching the victims (Moon & Lee, 2017). While this study did not focus on these issues, they should be considered because an increasing number of older people will receive care in facilities other than their homes in the future.

In addition, analyzing the impact of COVID-19 on social isolation and social support revealed that older people who felt isolated were more affected by the abuse experience. In the United States, the pandemic placed them closer to their caregivers, which caused further stress, making them easy targets for abuse (Han & Mosqueda, 2020; Makaroun et al., 2020). On the other hand, in India, restrictions on outdoor activities were stressful for the elderly, as it meant remaining indoors for long periods, which facilitated abuse (Kumar, 2021). Non-country-specific data using social network analysis revealed various kinds of abuses due to COVID-19 isolation in many countries (Xue et al., 2020). The isolation mandate implemented by various countries during the pandemic can be considered another form of abuse, because this unfamiliar phenomenon of social disconnection increases the danger to older adults around the world by exposing them to potential abuse (Han & Mosqueda, 2020). An unusual finding revealed by this study was that older adults experience less abuse if they have more family members who visit them; on the other hand, they experience more abuse if they have more friends who visit them. Because further analysis was not possible due to study data limitations, it can be interpreted that most older adults these days participate in many social communities rather than engage with their own family and tend to establish diverse relationships with friends or neighbors living around them, which can make them vulnerable to elder abuse (Van Orden et al., 2021). Thus, it cannot be concluded that having sufficient social interactions reduces the likelihood of elder abuse. Because abuse can occur in any relationship according to changes in society, a diverse approach should be considered to solve this problem.

Elder abuse perpetrators in the COVID-19 era

The findings of the study on the types of perpetrators reveal that offspring and their spouses account for the most common type, followed by neighbors, friends, roommates, and relatives. This is similar to previous studies, which revealed that many older people were abused by those with whom they shared close relationships such as their families and neighbors, especially in situations where other social activities were limited by social isolation (Makaroun et al., 2020; Organization, 2021; Xue et al., 2020). As older adults become isolated due to the COVID-19 pandemic-related restrictions and their circle of friends narrows, the number of people they meet becomes limited, inevitably exposing them to abuse from people who are close to them.

These results reveal that more family members result in a lower burden of interaction, less support, and decreased abuse by the family toward the elderly. However, abuse increases when frequent contact with acquaintances and neighbors other than family members increases. This is an especially stressful situation that has been insufficiently studied. A critical problem is that abuse perpetrated by people close to the victims is difficult to report or expose. Thus, further studies are needed to identify the risk factors in such relationships.

Implications

To appropriately respond to a prolonged COVID-19 situation, healthcare and protection guidelines for community-dwelling older adults should be reformed to ensure abuse prevention. Moreover, systematic improvements integrating untact technology with traditional face-to-face services are required. To alleviate the family’s burden of caring for older adults, social perceptions should be improved and better treatment should be offered to bring respite to families and professional care within communities. Finally, healthcare professionals and policymakers should prevent elder abuse by imparting education and promoting elders’ safety.

More specifically, it is necessary to improve public awareness through the “Age with Rights Campaign” of the Global Alliance for the Rights of Older People (GAROP). We should consider establishing action plans and policies to publicize the human rights of the elderly as a social agenda. As part of these efforts, education to protect and promote the human rights of the elderly related to poverty, alienation, abuse, health, education, and social activities should be collaboratively imparted. Considering this pandemic, untact trainings for professionals and nonprofessionals who work with the elderly should also be improved so that they can better recognize abuse in domestic settings and facilitate interventions and provide public assistance such as intensive case management and financial support services to vulnerable older adults. Finally, since the prolonged situation may lead to blind spots for elder abuse, we propose further studies on strengthening protection through the elder abuse reporting system, counseling, education, follow-up management for abusers, and expanding infrastructure.

Conclusion and limitations

This study described elder abuse and the risk factors among the community-dwelling elderly in South Korea during the COVID-19 pandemic. Physical dependency, depressive symptoms, cognitive status, and other health-related statuses were found to affect the relationship between elder abuse and social isolation. To understand the impact of social isolation on elder abuse, more structured studies are needed, especially regarding the type of abuse and specific social relationships.

A limitation of this study is that participants were only asked about their interactions with family and relatives; thus, it is conceivable that being isolated from meeting family members because of the COVID-19-related social distancing protocol may have caused them to disconnect from other social relationships as well. Because this study aimed to identify the influencing factors using the Korean elder abuse research model based on secondary data from a single area – Seoul – there is a limit to the generalizability of the research findings. However, because Seoul is the largest city and home to more than half of the Korean population, it can be viewed as representative of the cultural and situational characteristics of the social isolation brought about by the COVID-19 pandemic that is experienced by Koreans.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data were made available by the Seoul Welfare Foundation after permission was obtained on June 3, 2021 (https://wish.welfare.seoul.kr/swflmsfront/board/boardr.do?bmno=10015&opno=10005&bno=89640)

Additional information

Funding

This paper was supported by research funds for newly appointed professors of Gangneung-Wonju National University in 2022.