The National Elder Mistreatment Study: Race and Ethnicity Findings

Abstract

The prevalence of elder mistreatment with respect to race and ethnicity was examined in an unweighted sample of 5,777 participants (5,776 participants in weighted sample). Random Digit Dialing methodology was used to select a representative sample of community-dwelling older adults, and the survey was available in English and Spanish. Mistreatment types included emotional, physical, and sexual abuse. Race and ethnicity based differences were largely absent, and the only observed increase was for physical mistreatment among Non-White older adults; however this association was not sustained in multivariate analyses controlling for income, health status, and social support. Findings are in contrast to prior reports of increased risk of mistreatment in minority populations, and point to correlated and modifiable factors of social support and poor health as targets for preventive intervention.

INTRODUCTION

Early studies of elder mistreatment indicate that prevalence varies by gender, socioeconomic status, prior victimization, psychiatric history, low social support, and substance use (Acierno, et al., 2010; Bachman, 1994; Biggs, Manthorpe, Tinker, Doyle, & Erens, 2009; Breslau, Davis, Andreski, & Petersen, 1991; Hanson, Kilpatrick, Freedy, & Saunders, 1995; Laumann, Leitsch, & Waite, 2008; Norris, 1992). Demographic factors such as race and ethnicity also have been the foci of study. Considering race, results have been mixed, with some studies showing that Caucasians at increased risk relative to African Americans (Cottler, Compton, Mager, Spitznagel, & Janca, 1992; Norris, 1992) and Hispanics (Burnam, et al., 1988), and other studies placing African Americans at greatest vulnerability (Federal Bureau of Investigation, 2008; Kilpatrick, Edmunds, & Seymour, 1992; Laumann, et al., 2008). However, this variance in race-based risk estimates across studies may be due to confounding effects of income, education, age, gender, geographic location, and respondent willingness to report the crime (Acierno, et al., 2010; Kilpatrick, et al., 2000; Reiss & Roth, 1993). Considering ethnicity, Laumann (2008) reported reduced risk of verbal or financial mistreatment in Latinos relative to African-Americans and Whites. Nevertheless, very little research has focused on ethnicity, and reasons for differential prevalence are not yet clear. Frequently, explanations of low or non-existent elder mistreatment in the Latino population cite family solidarity and its association with higher levels of social support as potentially protective (Acierno, et al., 2010). This seems logical when considering that Latino culture is deeply rooted in the concepts of familia (family) and respeto (respect), especially with reference to older adults. This is typically a positive aspect of family care of the elderly, but as Lowenthal (2007) states, may lead to the protection of intra-family perpetrators and a rejection of involvement of authorities. La vergüenza (shame), which goes against the “family institution”, also may lead to inhibited reporting. Thus, these often beneficial values do not assure that elder mistreatment is not occurring. Instead, while often protective, they may also contribute to underreporting when the event does occur relative to other ethnic/race groups (Sánchez, 1998).

Some investigators posit that very specific experiences of Latinos and minority group members might actually increase the probability of abuse, such as the process of immigration with all its associated stressors (e.g., discrimination, illegal or tenuous or questionable status, low socio-economic status, trouble assimilating into a new culture) (Vazquez & Rosa, 1999). Of course, when considering ethnicity or race as a risk or protective factor, it is also important to consider other risk factors for elder mistreatment, both independently and insofar as they relate to ethnicity per se. For example, older women seem to be at increased risk relative to older men (Biggs, et al., 2009; Klein, Tobin, Salomon, & Dubois, 2008; Pillemer & Finkelhor, 1988; Yaffe, Weiss, Wolfson, & Lithwick, 2007). Age also plays a role, with “younger old” reporting more verbal abuse and financial mistreatment than “older old” (Acierno, et al., 2010; Laumann, et al., 2008). Acierno et al. (2010) recently reported that low social support is a very strong risk factor for mistreatment, and may be more relevant than either gender or age in predicting these events. This finding was supported by other studies (Racic, Kusmuk, Kozomara, Debelnogic, & Tepic, 2006; Yaffe, et al., 2007) where authors noted that reduced participation in social activities may be a risk factor for abuse, particularly among men.

Prior research has not focused on the associations of race and ethnicity, per se, on elder mistreatment insofar as these factors have been considered along with other demographic factors, but have not themselves been the target of higher level analyses or statistical control. Therefore, the present investigation sought to isolate the relevance of race and ethnicity to mistreatment potential by controlling for specific factors identified as highly important in previous studies and as central in the culture of Latinos and African Americans, including history of prior traumatic event exposure, low income, poor perceived health, and low social support (see below for comment regarding correlated variables of gender and age) (Acierno, et al., 2010; Acierno, Ruggiero, Kilpatrick, Resnick, & Galea, 2006; Kilpatrick, et al., 2007; Ruggiero, et al., 2009).

METHODS

The present study used Random Digit Dialing (RDD) methodology to derive a random, nationally representative sample (based on age, race, and gender) of 5,777 community-residing older adults. RDD was selected over in person interviews in light of successful epidemiological research on interpersonal violence in which rates of physical and sexual assault, including assault of older adults, was documented (see Kilpatrick, et al., 1992). Moreover, this telephone based method was contrasted directly to in-person interview in a sample of older adults (Acierno, et al., 2003) in terms of reporting rates, comfort with the topic, and hearing or technical difficulties. No differences were found on either prevalence or feasibility variables. In fact, responses to open-ended questions at the end of the interview indicated that participants felt more comfortable talking about interpersonal violence over the telephone than directly to another person physically present in the room with them (Acierno, et al., 2003). In addition to demographic factors, participants were interviewed via telephone in English or Spanish (according to the preference of the person) about emotional, physical, sexual, and financial mistreatment as well as neglect occurring since they were age 60.

Variables

EMOTIONAL MISTREATMENT

Emotional mistreatment was defined as an affirmative answer to any one of the following four questions. After it was determined that such an event had occurred, the timing (i.e., when it most recently occurred) was defined.

1. “Now we want to ask you about some things that people in your life might do that make you feel bad, such as saying very mean things to you, or being rude to you. A lot of people say this happens to them, and we really need to find out how often it happens. Sometimes, we call these things emotional mistreatment. The person who might do these things could be a romantic partner, spouse, family member, friend, or someone who helps take care of you. Has anyone ever verbally attacked, scolded, or yelled at you so that you felt afraid for your safety, threatened or intimidated?”

2. “Has anyone ever made you feel humiliated or embarrassed by calling you names such as stupid, or telling you that you or your opinion was worthless?”

3. “Has anyone ever forcefully or repeatedly asked you to do something so much that you felt harassed or coerced into doing something against your will?”

4. “Has anyone close to you ever completely refused to talk to you or ignored you for days at a time, even when you wanted to talk to them?”

PHYSICAL MISTREATMENT

Physical mistreatment was defined as an affirmative answer to any one of the following three questions. As was the case with emotional mistreatment, descriptive parameters of the event were collected after respondents indicated that such an event had occurred.

1. “Another type of stressful event that people sometimes experience is being physically hurt by another person. The person doing these things could be a romantic partner, spouse, family member, friend, or someone who helps take care of you. Has anyone ever hit you with their hand or object, slapped you, or threatened you with a weapon?”

2. “Has anyone ever tried to restrain you by holding you down, tying you up, or locking you in your room or house?”

3. “Has anyone ever physically hurt you so that you suffered some degree of injury, including cuts, bruises, or other marks?”

SEXUAL MISTREATMENT

Sexual mistreatment was defined as an affirmative answer to any one of the following three questions. After it was determined that such an event had occurred, the timing was defined.

• 1“I am going to ask you questions about unwanted sexual advances that you may have experienced over your lifetime. People do not always report such experiences to the police or discuss them with family or friends. The person making the unwanted advances isn't always a stranger, but can be a friend, romantic partner, or even a family member or someone you trust to help you or help take care of you. Such experiences can occur anytime in a person's life. Regardless of how long ago it happened or who made the advances, has anyone ever made you have sex or oral sex by using force or threatening to harm you or someone close to you?”
• 2a(for females) “Has anyone ever touched your breasts or pubic area or made you touch his penis by using force or threat of force?”
• 2b(for males) “Has anyone ever touched your pubic area or made you touch their pubic area by using force or threat of force?”
• 3a(for females) “Has anyone ever forced you to undress or expose your breasts or pubic area when you didn’t want to?”
• 3b(for males) “Has anyone ever forced you to undress or expose your pubic area when you didn’t want to?”

ETHNICITY

Ethnicity was defined as an affirmative response to the question “Are you of Hispanic origin or decent.”

RACE

Race was first defined specifically in response to the question “In which of the following categories do you feel you belong: white; Black, Pacific Islander, American Indian or Alaskan native, Asian or something else?” However, due to small cell sizes for most race subtypes, categories were collapsed into White vs. Non-White.

INCOME (NORMAL VS. LOW)

Low income was defined as cases where the entire household income was less than $35,000 the previous year.

HEALTH STATUS (GOOD VS. POOR)

Health status over the prior month was assessed using the general health question number 1 from the World Health Organization Short-Form 36 Health Questionnaire (Ware & Gandek, 1998). Participants were asked to rate the following question: “In general, would you say your health is “Excellent, Very Good, Good, Fair, or Poor.” These responses were dichotomized into Poor Health (self rating of fair or poor) and Good Health (self-rating of excellent, very good, or good). This assessment is consistent with previously validated single item measures of general subjective health, which have shown both good reliability and validity (Sibthorpe, Anderson, & Cunningham, 2001), and has been found related to morbidity and mortality (Grant, Piotrowski, & Chappell, 1995; Idler & Benyamini, 1997). Poor health was considered as a possible risk factor due to the associated increase in caregiver burden resulting from increased medical visits, healthcare regimens, and caretaking responsibilities.

EXPERIENCE OF PRIOR TRAUMATIC EVENTS (YES VS. NO)

Participants were asked to report if they had been exposed to the following events and indicated fear that they would be killed or seriously injured during this exposure: natural disasters such as earthquake, hurricane, flood, or tornado; serious accident at work, in a car, or somewhere else; or being in any other situation where you thought you would be killed. Past research indicates that exposure to one form of trauma is associated with increased risk of interpersonal violence (likely due to common risk factors in the environment) (Acierno, et al., 2006).

SOCIAL SUPPORT (HIGH VS. LOW)

Perceived social support during the past month was assessed via a modified five-item version of the Medical Outcomes Study module for social support (Sherbourne & Stewart, 1991). Participants were asked about emotional (e.g., “someone available to love you and make you feel wanted”); instrumental (e.g., “someone available to help you if you were confined to bed”); and appraisal (e.g., “someone available to give you good advice in a crisis”) social support and responded to items using a four-point scale from “none of the time” to “all of the time” (sample range=0–20; M=15.9 [SD=4.8]). Low social support was operationalized as a score in the lower quartile of the sample ratings, and the comparison high social support was operationalized as a score in the upper quartile of sample ratings. Social support has been identified as a protective factor for mental and health outcomes in prior research (Acierno, et al., 2006), and was predicted to be a potentially modifiable protective factor for elder mistreatment insofar as social connections would lessen the likelihood that abuse would occur due to increased activities with others in the social network, leaving less time in the isolated company of a perpetrator.

Participants

Data were collected from 5,776 older adults (weighed sample; 5,777 older adults in unweighed sample) with an average age of 71.5 years (SD = 8.1), range of 60 to 97 years; 60.2% (3,477) of the older adults were women and 39.8% (2,300) were men. Of the total, about 56.8% (3,281) were married or cohabitating, 11.8% (677) were separated or divorced, 25.1% (1,450) were widowed, and 5.2% (303) were never married and the remainder did not report. Considering race in order of magnitude, 84.4% (4,876) indicated that they were White, 6.7% (386) Black, 2.3% (132) American Indian or Alaskan Native, 0.9% (49) Asian, 0.2% (13) Pacific Islander, and the remainder chose not to identify. Considering ethnicity, 4.3% (245) indicated that they were of Hispanic or Latino origin.

The sample also was characterized in terms of contextual factors that might serve to increase or decrease risk of elder mistreatment, yielding the following risk factor information: low household income (i.e., less than $35,000 per year combined for all members of the household): 46.1% (2,002); poor reported health: 22.3% (1,279); a prior traumatic event: 62.0% (3,566); perceived low social support 26.6% (1,379).

Statistical Plan

Prevalence of each type of mistreatment was compared across race and ethnicity via univariate chi square analyses. Chi square analyses also were used to determine ethnicity and race based differences in contextual variables that might influence risk, including income, social support, perceived health status, and exposure to prior traumatic events. Contextual variables that were significantly different across ethnicity and race were included as covariates in multivariate logistic regression analyses for each mistreatment type to determine if control of these factors affected ethnicity and race based risk of mistreatment. Note that logistic regressions did not include gender and age in order to limit the number of non-modifiable variables in the logistic regression (e.g., variables that could not be targeted in preventive interventions) and maximize power. Thus, all logistic regressions were re-run with the addition of gender and age to verify that controlling for these variables would not change observed patterns of risk relationships. No risk relationships were altered by this addition, and logistic regression results reported below omitted these two variables.

RESULTS

The cooperation rate was 69% for the sample. 60.2% of the older adults were women and 39.8% were men. Table 1 presents the prevalence for each mistreatment type in terms of ethnicity and race. No significant differences were observed for any mistreatment type with respect to ethnicity. Considering race, Non-Whites were at twice the risk of Whites (3.0% vs. 1.4%) for physical mistreatment (OR 2.19), but there were no observed differences for emotional or sexual mistreatment.

TABLE 1.

X² Analyses for Past Year Mistreatment in Terms of Ethnicity and Race.

Predictive Factor	%	N	χ2	OR	CI	p
Emotional Mistreatment
Hispanic			0.09	1.14	0.58–2.17	.461
No	4.7	242
Yes	4.2	10
Race			1.84	1.30	0.89–1.91	.108
Non-White	5.8	32
White	4.5	208
Physical Mistreatment
Hispanic			0.01	0.96	0.35–2.63	.538
No	1.6	82
Yes	1.7	4
Race			8.01	2.19	1.26–3.83	.007
Non-White	3.0	16
White	1.4	62
Sexual Mistreatment
Hispanic			1.51	1.01	1.00–1.01	.229
No	0.6	33
Yes	0.0	1
Race			0.10	0.83	0.25–2.72	.520
Non-White	0.5	3
White	0.7	31

Note: The level of the variable given first represents the reference value of the variable, which is also the level the variable hypothesized to be associated with increased risk. OR below 1 indicates reduced risk for the reference variable. Confidence Intervals that do not cross the value 1.00 indicate increased (if CI ranges above 1.00) or reduced (if CI ranges below 1.00) risk for the reference value of the variable.

Considering differences in terms of contextual variables presented in Table 2, greater proportions of Hispanics reported low income (61% vs. 46%), poor social support (42% vs. 26%), and poor health (37% vs. 21%). There were no differences in terms of exposure to prior traumatic events. Similar patterns were observed with race, with a greater proportion of Non-Whites reporting low income (65% vs. 43%), poor social support (38% vs. 24%), and poor health (32% vs. 20%).

TABLE 2.

X² Analyses of Contextual Variables in Terms of Ethnicity and Race

Predictive Factor	%	N	χ2	OR	CI	p
Low Income
Hispanic			16.17	1.89	1.39 – 2.63	.000
Yes	61.3	103
No	45.5	1895
Poor Health
Hispanic			32.79	2.16	1.65 – 2.83	.000
Yes	37.1	89
No	21.4	1161
Prior Traumatic Event
Hispanic			0.16	1.05	0.81 – 1.38	.372
Yes	63.3	155
No	62.0	3365
Low Social Support
Hispanic			27.0	2.05	1.56 – 2.71	.000
Yes	41.7	91
No	25.9	1265
Low Income
Race			72.24	2.39	1.94 – 2.93	.000
Non-White	64.6	286
White	43.3	1612
Poor Health
Race			42.53	1.86	1.54 – 2.24	.000
Non-White	32.1	185
White	20.3	982
Prior Traumatic Event
Race			2.54	1.15	0.96 – 1.38	.072
Non-White	64.9	373
White	61.7	2994
Low Social Support
Race			49.0	1.94	1.61 – 2.34	.000
Non-White	38.4	205
White	24.3	1064

Note: See Note Table 1.

Thus, logistic regression analyses were conducted to examine risk of mistreatment in terms of ethnicity and race, controlling for the significantly different contextual variables of low income, social support, and poor health. Ethnicity based logistic regressions continued to reveal no predictive risk of being Hispanic for any form of elder mistreatment. However, low social support increased risk of all forms of mistreatment, even after effects of ethnicity and other factors were controlled. Similarly, poor health was associated with both emotional and physical mistreatment. Race based logistic regressions also indicated that Non-White status was not, in itself, associated with any increased prevalence of any form of mistreatment, including physical mistreatment once income, social support, and health status were controlled. As with ethnicity based multivariate analyses, low social support was independently predictive of all forms of mistreatment, and poor self-rated health was associated with increased emotional mistreatment. Thus, overall, no differences for ethnicity or race were observed for any mistreatment type in multivariate analyses.

DISCUSSION

There is significant interest in identifying and explaining race and ethnicity-based differences in the prevalence of interpersonal violence. However, studies on children and adolescents, as well as data presented here on older adults from the National Elder Mistreatment Study, indicate that these differences are less apparent than one might expect. In this study, race and ethnicity exerted no effect on risk of any form of mistreatment when other contextually related variables were controlled. Indeed, the only significant univariate finding, that of increased risk of physical mistreatment for Non-Whites, was eliminated when effects of income, perceived health, and social support were considered. Of particular interest was the finding that the most important predictive factors for all forms of mistreatment were those that are potentially modifiable, specifically social support and health. This finding provides some specific direction to those developing preventive and secondary interventions for elder mistreatment in that it may well be worthwhile to leverage existing agencies and organizations that could potentially facilitate social support or improve health. For example, church or college-based volunteer groups could initiate services dedicated to providing transportation to potentially social events or health-related opportunities (or both in the form of health fairs, etc.) for the significant proportion of older adults who find it difficult or impossible to drive or use public transportation (U.S. General Accounting Office, 1991).

Findings related to low social support and increased risk of negative outcomes are consistently reported in the literature. For example, older adults reporting low social support who were exposed to extreme stress events such as natural disaster were at increased risk of post-traumatic stress disorder, depression, and generalized anxiety disorder (Acierno, et al., 2006). Of course, this was not a longitudinal study, and the direction of risk cannot be determined; it may be the case that those who are mistreated report lower social support and this factor results from, rather than causes, elder mistreatment. The same could be said for perceived poor health. However, it seems most likely that the relationship is bidirectional, with poor support or health both indicative of, and predictive of, mistreatment. Thus, the potential importance of these factors to negative outcomes in older adults, and by connection, their relevance to developing preventive interventions for both interpersonal violence and psychopathology appears great.

Limitations

This study is characterized by several limitations. First, all estimates of mistreatment and contextual variables were based entirely on self-report. Mistreatment events are often under-reported in this age group, and thus the validity of estimates is less than perfect. A second major limitation was a failure to include some measure of cognitive functioning, either as a covariate or risk factor. Thus, these findings reflect responses of a cognitively intact, community-residing sub-population of older adults. By connection, generalization of these results to what may be the group most at risk of mistreatment, the cognitively impaired elderly, is not appropriate. We also did not include individuals who did not have a home phone (i.e., those with only cell phones), introducing the potential for bias. However, our data is nationally representative and is weighted on Census estimates, increasing its generalizability.

TABLE 3.

Logistic Regression for Mistreatment in terms of Ethnicity and Race

Variable	OR	95% CI	B	W	p
Emotional Mistreatment
Ethnicity (Hispanic)	1.35	0.61 – 2.99	.30	.56	.456
Income (Low)	0.93	0.68 – 1.26	−.08	.24	.625
Health (Poor)	1.61	1.15 – 2.23	.47	7.91	.005
Social Support (Low)	2.28	1.67 – 3.12	.83	26.79	.000
Physical Mistreatment
Ethnicity (Hispanic)	1.37	0.33 – 5.75	.31	.18	.670
Income (Low)	1.18	0.65 – 2.13	.17	.30	.584
Health (Poor)	1.92	1.06 – 3.50	.65	4.59	.032
Social Support (Low)	2.24	1.25 – 4.02	.81	7.32	.007
Sexual Mistreatment
Ethnicity (Hispanic)	1.13	--	16.24	.000	.996
Income (Low)	2.25	0.77 – 6.58	.81	2.18	.140
Health (Poor)	1.22	0.44 – 3.44	.20	0.15	.702
Social Support (Low)	4.32	1.56 – 11.95	1.46	7.96	.005
Emotional Mistreatment
Race (Non-White)	1.01	0.63 – 1.61	.01	0.00	.984
Income (Low)	0.90	0.65 – 1.24	−.11	0.45	.502
Health (Poor)	1.67	1.19 – 2.34	.510	8.68	.003
Social Support (Low)	2.16	1.56 – 2.98	.77	21.67	.000
Physical Mistreatment
Race (Non-White)	1.20	0.52 – 2.79	.18	0.18	.671
Income (Low)	1.40	0.74 – 2.65	.34	1.06	.304
Health (Poor)	1.53	0.79 – 2.95	.42	1.58	.209
Social Support (Low)	1.94	1.94 – 3.64	.66	4.25	.039
Sexual Mistreatment
Race (Non-White)	0.45	0.07 – 2.72	−.81	0.77	.381
Income (Low)	2.34	0.80 – 6.85	.85	2.40	.121
Health (Poor)	1.29	0.46 – 3.63	.26	0.24	.627
Social Support (Low)	4.43	1.60 – 12.25	1.49	8.25	.004

Note: The level of the variable given first represents the reference value of the variable, which is also the level the variable hypothesized to be associated with increased risk. Confidence Intervals that do not cross the value 1.00 indicate increased (if CI ranges above 1.00) or reduced (if CI ranges below 1.00) risk for the reference value of the variable.