Patterns of abuse of elderly people in Brazil: analysis of notifications

This study aimed to describe the characteristics of elderly people abuse notifications by gender and to assess notification patterns according to gender. We analyzed data from the Brazilian Information System for Notificable Diseases (SINAN) in 2017. We carried out a descriptive analysis of victim characteristics, violence, and the probable perpetrator according to gender. Pearson’s χ2 test was used to assess the significance between groups. Then, we verified the main relationships between the studied characteristics and the victim’s gender by simple correspondence analysis (SCA). Thus, 17,311 cases/suspicions of elderly people abuse were notified, corresponding to 7.2% of the total number of violence notifications. Of these victims, 50.4% were white, 42.3% were married, and 17.2% had a disability/disorder; 76.9% occurred at home, 62.8% included physical violence, and 49.5% were cases of repeated violence. Most perpetrators were men (62%), and violence by two or more perpetrators was observed in 62.8% of the cases. SCA evidenced inequalities in older adults’ gender, which proved to be higher among women. Physical violence was the most common among younger and old individuals, whereas neglect/abandonment tended to occur more frequently among the oldest individuals, and was most often committed by daughters. In sum, this study demonstrated evidence of gender-based violence, especially among older adults. Disability proved to be an essential characteristic for neglect/abandonment in older adults. In this context, policies are needed to reduce gender inequalities and implement a care network for older adults who are victims of violence.


Introduction
Abuse of elderly people is widely recognized for its numerous social and public health repercussions.It is defined as a "single or repeated act, or lack of appropriate action, occurring in any relationship in which an expectation of trust causes harm or distress to an elderly person" 1 (p. 126-7).In older adults, its effects can be especially severe due to greater vulnerability and health conditions 1,2 .The main consequences include a high risk of morbidity and mortality 1,2,3 , an increase in chronic diseases, physical injuries, as well as mental and behavioral disorders 1,2 .
Population aging is one of the most significant challenges today 4 , and it is believed that the problems arising from this process and the occurrence of violence will increase 1,2,3 .Geriatric syndromes such as frailty, common in older adults, can favor events of violence 3 .Furthermore, studies have shown a relationship between violence, social inequalities 5,6 , and factors related to perpetrators, such as financial dependence, mental disorders, and substance abuse 7 .
The occurrence of elderly people abuse changes between countries.A systematic review and metanalysis carried out in 2016, data from 28 countries worldwide, obtained a combined prevalence of 15.7% of elderly people abuse 8 .In Norway, a high-income country, the prevalence was 9.2%, and proved to be higher among women 9 .In Colombia, a country with an upper middle-class income, 15.1% of older adults, most aged 60-69 years, female, with a low schooling level, poor socioeconomic status, and living alone or with their children, reported some form of violence 10 .In Brazil, elderly people abuse accounted for 5.7% of all reported cases in the period from 2009 to 2013, and was higher among women 11 .
Although several studies have been found in the literature, authors claim that the visibility of the issue is still insufficient 11,12 .Information regarding the nature and extent of the violence is scarce, especially in low-income countries 7,11 .Furthermore, the accelerated aging of the population, the impact of violence, and the inequalities of the country make it essential for different studies and approaches to monitor this issue.This information can expand knowledge, support the development of public policies, provide greater visibility to violence, enable comparisons with other countries, and guarantee the rights of older adults.
Therefore, this study aims: (1) to describe the characteristics of elderly people abuse notifications carried out in the Brazilian Information System for Notificable Diseases (SINAN) according to the perpetrator's gender and (2) to assess notification patterns made to SINAN according to the victim's gender.

Study design and data source
This is a cross-sectional, exploratory study carried out by analyzing data from the SINAN component of the Violence and Accident Surveillance System (VIVA) for 2017.

Population
All notifications of elderly people abuse in SINAN for Brazil were included in the study, except selfharm.An older adult was considered to be any person aged 60 and older, according to the Brazilian Statute of the Elderly 13 .
The characteristics of violence included in this study were: (a) type of violence: physical, financial, psychological, neglect/abandonment, other types of violence (sexual, torture, human trafficking, legal intervention); (b) means of agression: force/beating, sharp object, blunt object, threat, others (firearm, substance/hot object, poisoning/intoxication, others); (c) location: household, public road, other places (school, sports venue, industries/construction, collective housing, bar, or similar commerce/ services); and (d) repeat violence: yes, no.
The following perpetrator characteristics were also studied: (a) probable perpetrator: family member (father, mother, stepfather, stepmother, sibling), partner (spouse, ex-spouse, companion, ex-companion), older adults' children, friend/acquaintance, unknown, caregiver, others (boss, person with institutional relationship, police, others); (b) probable perpetrator's gender: male, female, both; (c) if the perpetrator is an elderly person: yes, no; (d) number of people involved: one, two, or more; and (e) suspected alcohol use by the perpetrator: yes, no.

Statistical analysis
The frequencies of characteristics of the victim, violence, and probable perpetrator were stratified by gender.Pearson's χ 2 test was performed to verify the differences among the groups, with a 5% significance level.
Then, the relationship between the studied characteristics was assessed by the simple correspondence analysis (SCA), which is a statistical technique of multivariate analysis, commonly indicated to analyze categorical data 14 , enabling visualization through graphs, in which each category is presented as a point.These points indicate the relationship levels between the categories 14 ; smaller distances indicate more significant association and larger distances indicate dissociations 15 .The categories furthest from the point of origin in the graph present the most significant influence 16 .
All variables studied in SCA were considered, except for variables with a p-value ≥ 0.20 in the analyses obtained by the chi-square test.The caregiver as the probable perpetrator was not excluded, as they are greatly relevant to studies with older adults 17,18 .
SCA was stratified by the victim's gender and considered the structure of a contingency table.The top row included violence and perpetrator's characteristics (location, type of violence, repeated violence, perpetrator, older adult, probable perpetrator, and means of aggression), whereas the columns included victim's characteristics (age group, marital status, and presence of disability/disorder).After processing, all categories with a mass ≤ 0.025 and those with similar contributions in the dimensions were excluded.
The data are presented in table and graph forms.In the graph, the characteristics with smaller distances from each other were grouped in each quadrant of the graph and were called a "pattern".Data were compiled in the TabWin software (http://siab.datasus.gov.br/DATASUS/index.php?area=060805&item=3), and processing was performed in Stata, version 14 (https://www.stata.com).

Ethical aspects
This study used secondary data in the public domain, not requiring approval by the Research Ethics Committee.However, all ethical aspects involved in scientific research were respected.

Descriptive
In 2017, 17,311 (7.2%) cases of suspected/confirmed domestic/interpersonal elderly people abuse were reported in SINAN.Among these, 55.5% were women, 30.7% were 70-79 years old, 50.4% were white, 42.3% were married, and 17.2% had a disability/disorder.Most occurred at home (76.9%); the main type was physical violence (68.4%), followed by neglect/abandonment (31.4%).The main form of violence was force/beating, and 49.5% reported being victims of repeat violence.Regarding perpetrators' gender, 62% were men, and 62.8% of the occurrences were committed by two or more perpetrators (Table 1).Elderly women of white skin color and widows reported the most events of violence.Household occurrences (86%), through physical violence (57.3%), by means of aggression, force/beating, threats, and repeat violence (58%) were more frequent.Regarding perpetrators, 58.8% were men; and 65.8% of elderly women have suffered violence by two or more perpetrators.The main perpetrator were their children (44.3%), with no significant difference between genders, p-value > 0.05.Among elderly men, violence occurred mainly among those of black/mixed-race skin color (50.6%) and married (50.6%).Household was the main location (64.7%), followed by public roads (21.1%), whereas physical violence was present in 72% of cases.A total of 34.9% of the perpetrators were suspicious of alcohol consumption, with no difference between genders (p-value > 0.05).The main perpetrators were older adults' children (31.1%) and unknown (24.3%) (Table 1).

Simple correspondence analysis
For presenting p-value < 0.20 in the descriptive analysis, the following characteristics were excluded: elderly women, public roads and other location; financial violence and other types of violence; perpetrator's gender; caregiver, friend/acquaintance, and unknown as probable perpetrators; and the use of blunt and sharp objects as a means of aggression.Two dimensions were needed to explain the joint variance of column and row profiles.The contribution of Dimension 1 to inertia was 87.1%, whereas Dimension 2 was 9.5%, totaling 96.6% of the explanation.We verified a major association between variables, χ 2 test = 6,025.50and p-value < 0.01.Among the relative contributions of each dimension to total inertia, age groups ≥ 80 (28.4%) and 60-64 years (21.1%), and married individuals (21.7%) stood out.For the other characteristics, neglect/abandonment (32.5%), as well as partner (22.4%) and children as probable perpetrators (13.2%) stood out.
For elderly men, the following categories (p-value < 0.20) were excluded: other places of occurrence; financial violence and different types of violence; perpetrator's gender; caregiver, friend/ acquaintance, and partner as probable perpetrators; use of blunt and sharps objects; and threats as a means of aggression.Two dimensions explained 96.8% of the profiles' joint variance.The association between variables was high, χ 2 test = 3,343.47and p-value = 0.01.Among the relative contributions of each dimension and the sociodemographic variables, individuals aged ≥ 80 (25.9%) and 60-64 years (25.3%), and the presence of disability/disorder (25.2%) stood out.

Table 2
Coordinates and contributions of sociodemographic characteristics of victim, type of violence, and perpetrator for elderly women.Brazil, 2017.

Correspondence graph
Pattern A consisted of married women and partners as probable perpetrators of violence.Pattern B corresponds to elderly women aged ≥ 80 years with disabilities/disorders, who were assaulted at home, with the presence of a perpetrator, and who suffered negligence/abandonment and repeat violence.Pattern C includes older adults, aged 60-69 years, single, who suffered psychological and physical violence by threat and force/beating.In Pattern D, widowhood, children as probable perpetrators, and female perpetrators stood out (Figure 1).Pattern A consists of married men, attacked by male perpetrators, with two or more individuals involved in the elderly person abuse.Pattern B consists of widowers and children as the probable perpetrator; it also consists of women who have suffered psychological violence and repeat violence, at home, and with only one person involved.Pattern C includes older adults, aged 60-69 years, single, who most commonly suffered violence on public roads, with the means of aggression being force/ beating, performed by an unknown probable perpetrator.Finally, Pattern D included older adults, aged ≥ 80 years and victims of neglect/abandonment (Figure 2).

Table 3
Coordinates and contributions of sociodemographic characteristics of victim, violence and perpetrator for elderly men.Brazil, 2017.

Discussion
This study assessed the main characteristics and patterns of elderly people abuse registered in SINAN.Most victims were women, aged 70-79 years, white, married, and approximately one fifth with a disability/disorder.The main type of violence was physical, followed by neglect/abandonment, emphasizing the use of force/beating.The most common location was the household, and repeat violence occurred in half of the analyzed population.Men were the most common perpetrator of elderly people abuse and, when known, the main perpetrator were the older adults' children.Alcohol consumption was reported for about one third of the perpetrators.We noted essential differences when considering the victim's gender in the studied notifications.Women were the primary victims, mostly white and widows, and the place of occurrence was mainly at home with repeated reports.Male victims are mostly black/mixed-race and married, and the violence occurred primarily on public roads.Regarding age, violence committed by the children and repeat violence were more common among older adults -an unknown perpetrator with suspected alcohol use, was the most common among young-old individuals.
The differences in this study characteristics regarding gender refer to the attributions constructed historically and culturally for men and women.Discrepancy and hierarchy mark these attributions, which are present in the social imagination and tend to be daily repeated 19 .In this study, the significant number of notifications for elderly women stood out.However, this number may be even higher, as they may find it more challenging to seek help due to health conditions and dependence on the perpetrator 20 .Moreover, the consequences in the most dependent group can be even more severe, Cad.Saúde Pública 2023; 39(1):e00075722 with psychological and health impacts, which is often accentuated depending on the duration and exposure to violence 20 .
The findings of this study are similar to those obtained in Europe, which found more violence against women.However, it was concluded that abuse in elderly men often goes unrecognized, even if it is frequent 21 .Elderly men can be just as exposed to violence as women, demystifying the belief that elderly people abuse only occurs against women 21 .One Brazilian study, with notification data from 2009 to 2013, found 45.7% of violence among elderly men 11 , which proved to be similar to our study (44.5%).
Different patterns were observed in the occurrence of elderly women abuse.Pattern A shows married women and partners as probable perpetrators.This profile is consistent with the literature, which shows a high prevalence of intimate partner violence (IPV) against women, ranging from 16.5% to 54.5% 20 .In a study conducted in Spain 22 , approximately 16% of the women surveyed suffered IPV; however, when compared to younger women, the frequency among elderly women was lower.In Brazil, few studies describe IPV in elderly women, most covering the younger age group, as it is the most commonly affected population 23,24,25 .
Pattern B shows elderly women, aged 80 years and older, with disability/disorder and repeat negligence/abandonment violence in their own homes.The disability results from a complex and dynamic process, mainly defined by difficulties in executing activities of daily living (ADL) 26 , and is responsible for a reduced quality of life and loss of autonomy.In a situation of dependency, older adults need care, which, in most cases, is provided by family members exposed to conditions of stress, tension, and conflict.Generally, only one person is responsible for the care 10 , favoring violence 1,2,27 .Furthermore, studies show that dependence on ADLs or instrumental ADLs increases the likelihood of violence, especially among women 10,27 .
This pattern is consistent with other studies that show a predominance of neglect/abandonment in women, aged 70 years and over, which occurred at home and repeatedly 28,29 .One study showed Cad.Saúde Pública 2023; 39(1):e00075722 that neglect/abandonment was more frequent among older adults aged 80 years and older when compared to those aged 60-69 years, and among those with disabilities/disorders 30 .Another survey showed that the victim's home was considered a place of risk for older adults, especially dependent individuals 10 , as found in our study.
Pattern C included elderly women, aged 60-69 years, single, and who suffered psychological and physical violence through threats and force/beating.This pattern is also consistent with the literature.Another study 29 , carried out with notifications from the state of Minas Gerais (Brazil), highlighted the association of physical and psychological violence among elderly women.That same study showed that only physical violence is more significant in elderly men when compared to elderly women 29 , which differs from the pattern observed in our study.In another study, physical violence figured prominently, followed by psychological violence, use of force/beating, and threats.Notably, psychological violence was more frequent among elderly women 28 .
The final pattern observed for elderly women was being a widow and having children who were a probable perpetrator.One study showed that single women suffer more violations than married women; however, the study grouped single, widowed, and separated women together, totaling 65.3% of the notifications 28 .Analysis with data from notifications from the city of Recife (state of Pernambuco, Brazil) also showed that widows suffered more violence than married/single women 30 .Furthermore, daughters stood out as the main perpetrators, which contrasts with findings in the literature, where the occurrence of violence is more common in sons 28,31 .
Violence presents itself with different configurations among older adults.Pattern A shows married male perpetrators with two or more perpetrators involved.The literature shows that the most common is that married men depend on their partners for care; therefore, marriage is considered a protective factor against illnesses and injuries 32 .However, in this study, married older adults suffered more violence than the others.Note that this study does not allow for the identification of risk or protective factors for violence.A study carried out in Brazil with notification data showed a similar frequency of violence between single and married people, and most aggressors were also male (72.1%) 28 .
Pattern B includes widows, with a daughter as the probable perpetrator, psychological violence, repeat violence at home, and only one perpetrator involved.One study shows that the chance of an elderly men suffering psychological violence is lower when compared to elderly women, and, in general, this violence occurs mainly among elderly men 28 .In another international study, men suffered more psychological violence (20%) than women (18.9%) 21.Thus, as for elderly women, daughters appear as the main perpetrators.We emphasize that women, in general, are the main caregivers of the elderly people in a situation of dependency 33 ; therefore, actions and strategies aimed at supporting elderly women and ensuring gender equity are needed.
Pattern C consists of older adults, aged 60-69 years, single, who suffered violence on public roads, through force/beating, and by unknown probable perpetrator.A Brazilian study 29 , also with notification data, showed that 87.5% of older adults suffered physical violence by an unknown perpetrator, whereas about 18% occurred on public roads, which corroborates the pattern observed in our present study.Another study conducted with elderly Mexicans 34 shows that young-old individuals suffered physical violence mainly on public roads.Results from the Brazilian National Health Survey 35 show that young-old adults, aged 60-69 years, suffered twice as much violence committed by unknown people compared to oldest adults, which is in accordance with our study.
The final pattern includes older adults, aged ≥ 80 years, with violence by negligence/abandonment.As for elderly women, this type of violence mainly occurs due to the fragile conditions presented by these people 36 .Although studies show that neglect/abandonment among elderly women is more prevalent 28,30 , it is essential to pay attention to this type of violence, as disability is considered a risk factor, since it occurs in both genders 37 .
Among our study limitations, the use of notification data carried out by healthcare services stands out, which prevents the attribution of causality in events of violence.Another important aspect is the high incompleteness of the data, especially for the schooling level and alcohol consumption variables, thus limiting the overall analysis.Furthermore, it is believed that underreporting exists and that available data do not reveal the real frequency of violence 11,12 .Another important aspect is the lack of knowledge about how to fill out the Individual Notification Form (FNI).As observed in this study, the sum of the variable type of violence is greater than 100%, which occurs because it is possible to Cad.Saúde Pública 2023; 39(1):e00075722 mark more than one option on the FNI; in the same case, more than one type of violence could be present.However, for the purposes of reporting in SINAN, only the main types of violence should be marked.If any secondary violence is committed by the same perpetrator, the information should be reported in the additional comments field 38 .In this sense, it is essential that the professionals are trained to properly fill out the FNI.This study is relevant since it deals with notifications of elderly people abuse in Brazil and it contributes to better strategies, adequacy of actions, and healthcare services for victims of violence.
In summary, the results showed differences when considering older adults' gender.Elderly women were the main victims, mostly white and widowed, and the place of occurrence was mainly at home with repeated reports.By contrast, elderly men were mostly black/mixed-race and married, and the violence occurred mainly on public roads.IPV in elderly women was also shown, and disability is a characteristic of violence among older adults, mainly through negligence/abandonment.In this sense, the differences in patterns of elderly people abuse indicate that health must consider characteristics such as gender, age, and the presence of disabilities in events of violence.
Actions to combat violence against elderly individuals can be carried out through policies that ensure better material and social conditions and structural changes in order to reduce inequalities and end discrimination, especially gender inequalities.Therefore, we consider imperative to advance public policies, promote a culture of peace, prevent and control violence at all ages, and structure and implement an effective care network for elderly people that are victims of violence.

Figure 1
Figure 1Graphic of correspondence analysis biplot for elderly women.

Figure 2
Figure 2Graphic of correspondence analysis biplot for elderly men.