Prevalence of elder abuse and victim-related risk factors during the COVID-19 pandemic in China

From April to May 2020, we conducted a questionnaire survey among elderly people aged 65 and above. People with language/communication impairments were excluded. We chose Changsha County, Ningxiang County, and Pingjiang County in Hunan Province as the survey locations. A total of 15 communities and 15 villages were randomly selected from each country. After obtaining the demographic information of the local elderly individuals over 65 years of age, we conducted a household questionnaire survey.

This research is based on a cross-sectional design. The sample size was calculated using the formula: N = (Z_α/2)²PQ / E² (Z_α/2, standard normal deviate at 95% confidence level; E, 1% relative error; P, prevalence = 50%; Q, 100% – prevalence). So the sample size should be at least 9604. Due to the entry conditions and the validity of the questionnaire, we needed to increase the number of survey subjects by 50%, so our sample size was 15,000 [21, 22]. A total of 15,000 questionnaires were distributed, and 10,362 valid questionnaires were recovered, with a recovery rate of 69.1%. A total of 5689 questionnaires were recovered from rural areas, and 4673 were recovered from urban areas. The survey takers were undergraduates majoring in epidemiology and health statistics who had undergone uniform and strict training. We communicated with each elderly individual alone in a room and told him/her that we would keep the conversation confidential. To prevent the spread of the epidemic, we wore masks. We also provided masks to the elderly individuals. A distance of 1 m was maintained during the conversation. To increase the likelihood of elderly individuals participating in the survey, daily necessities were distributed, and health consultations were performed. The questionnaire survey was conducted after obtaining verbal “informed consent” from the elderly individual.

The dependent variables of this study included physical abuse, emotional abuse, financial abuse, and neglect. The elderly abuse questionnaire in the “Third Survey on Chinese Women’s Social Status” is widely used to measure the prevalence of elder abuse in China (Cronbach’s alpha 0.81) [21, 23, 24]. To ensure the consistency of the measurement standards, we used this questionnaire. The investigator asked the elderly individual face-to-face, “Did your family exhibit any of the following behaviors towards you in the past year?” Eight behaviors that constitute domestic abuse were included in the survey: 1 Long-term refusal to visit, to greet, or speak to you; 2 Does not provide you with basic living expenses or embezzles your money privately; 3 Does not take care of you when needed; 4 Insults or threatens you; 5 Hits you; 6 Does not provide you with a fixed residence; 7 Does not feed you enough or feeds you poorly; 8 Does not allow you to leave the house. According to the aforementioned classifications, behaviors 5 and 8 were classified as physical abuse, behaviors 1 and 4 were classified as emotional abuse, behavior 2 was classified as financial abuse, and behaviors 3, 6, and 7 were classified as neglect. As long as one of the 8 behaviors was reported, the individual was considered to be abused, and a score of 1 was recorded; otherwise, a score of 0 was recorded.

According to related research, we selected the following 10 risk factors as independent variables. (1) Age: the actual age of the elderly individual [11, 22]. (2) Gender: 0 = male; 1 = female [22, 23]. (3) Marital status: 0 = married; 1 = divorced/widowed/unmarried [15]. (4) Educational level: 0 = illiterate; 1 = primary school to high school; 3 = university or above [21, 25]. (5) Income: 1 = under 5000; 2 = 5000 ~ 30,000; 3 = over 30,000 (RMB). If he or she is affected by COVID-19, the income of an elderly individual may decline [8, 9]. (6) Number of children: the number of living children. During the epidemic, the dependence of elderly individuals on their children increased [8, 22]. (7) Health condition (5-point system): 5 = very healthy; 1 = very unhealthy [18]. (8) Cognitive ability, or the basic cognition and calculation ability of the elderly individuals. This section mainly included questions such as “what day is today” and “how much is 30 minus 5.” There were a total of 30 questions in the survey. When an individual provided a correct answer, he or she was given 1 point. When he or she provided a wrong answers, he or she was not given any points. The lowest possible score was 0, and the highest possible score was 30 [26, 27]. (9) Social frequency: the frequency of participation in social activities over a week. If the elderly individual participated in social activities on a particular day, he or she was given a score of 1; otherwise he or she was given a score of 0. The lowest possible score was 0, and the highest possible score was 7. Less social participation frequency increases the risk of abuse, especially among men [28, 29]. During the epidemic, social distancing measures reduced the social participation frequency of the elderly [8, 9]. (10) Religious belief: 0 = yes; 1 = no [22, 23].

This study revealed that the number of cases of abuse was 1596 and that the prevalence rate of elder abuse was 15.4%. Physical abuse (163) and emotional abuse (412) were less common than financial abuse (647) and neglect (698). A total of 31.2% of participants reported two or more types of abuse (Fig. 1).

The study found that the prevalence of elder abuse increased with age, with the prevalence being 12.8% in the 65- to 69-year-old group, 16.2% on the 70- to 74-year-old group, and 21.1% in the 75-plus-year-old group. In terms of gender, women (16.7%) were more likely to be abused than men (13.7%). The prevalence of elder abuse among divorced, widowed, and unmarried individuals (20.0%) was higher than that among married individuals (13.3%). As the education level of the elderly individuals increased, the prevalence of abuse gradually decreased. The lower the income of an elderly person was, the higher the likelihood of abuse. The prevalence of abuse in the group earning more than RMB 30,000 was only 8.2%. The more children an elderly individual had, the higher the risk of abuse. The better the health and self-care ability of an elderly individual was, the lower the likelihood of abuse. The lower the cognitive ability of an elderly individual was, the higher the likelihood of abuse. The higher the social participation frequency of an elderly individual was, the lower the risk of abuse. Among elderly individuals with a weekly social frequency of 5–7, the prevalence rate of elder abuse was only 3.6%. People without religious beliefs (15.5%) were more likely to be abused than those with religious beliefs (11.0%). The chi-square test showed that the difference in the prevalence of abuse in each group was significant (p < 0.05) (Table 1).

Age has a significant negative impact on the prevalence neglect, which means that the greater an individual’s age is, the greater the likelihood of being neglected. With increasing age, elderly individuals’ daily and psychological dependence on family members increases [30]. This imposes a very large burden or financial pressure on caregivers; thus the risk of elder neglect increases with age. Gender is significantly related to emotional abuse and neglect, as women are more likely to be emotionally abused and neglected [31]. The high prevalence of abuse among females may be related to the personality and psychological characteristics of females. In China, women are mostly housewives, and they depend financially on other family members. These women are in a disadvantaged position in the traditional sense and are relatively more vulnerable to abuse.

There is a significant correlation between income and the four types of abuse. The higher an individual’s income is, the less likely he or she is to be abused. Studies have found that the more financially well-of an elderly individual is, the better the support he or she can give to his or her children, and the less likely he or she is to be abused [32]. The more children there are, the greater the likelihood of physical abuse, financial abuse, and neglect. This may be because when they face support problems, the children tend to shirk their responsibilities and regard their elderly parent as a burden to the family, even beating and scolding him or her [19].

Moreover, we found that elderly people with a common-law spouse, those with a higher education level, those that frequently participate in social activities, and those with religious beliefs are less likely to suffer abuse (Table 2).