The Prevalence and Incidence of Intimate Partner Violence in Older Women in Primary Care Practices

Abstract

Objective

Identify the incidence and prevalence of intimate partner violence (IPV) in women over 55 years of age in primary care offices.

Design

Telephone survey conducted between March and June 2003 by trained female interviewers who gathered self-report information about health and abuse.

Patients

A total of 3,636 women over 55 years of age had at least 1 visit in the past 12 months to primary care offices affiliated with an academic center in Southwestern Ohio were contacted by phone; 995 were deemed competent and completed the interview.

Intervention/Instrument

Thirty-eight page instrument that explored health, history of psychological (controlling behavior and threat of physical harm), physical, and sexual abuse since age 55 years. Interviews lasted 20 to 45 min.

Main Results

The mean age was 69 years (SD 8.35). Physical abuse in intimate relationships was reported by 1.52% since age 55 years (prevalence) and 0.41% in the past year (incidence). Prevalence and incidence rates for sexual abuse were 2.14% and 1.12%, threat of physical harm 2.63% and 1.62%, respectively. Less than half of the victims told someone else about the abuse. The mean number of health conditions was 3.84 for victims and 3.21 for nonvictims (P<.055) with significantly larger percentages of IPV victims reporting problems with chronic pain and depression.

Conclusions

Physical and sexual abuse by an intimate partner does occur in women over 55 years, but rates are lower than those of younger women. Health care providers are reminded to think about IPV in older women and to ask about abuse as disclosure is rare.

Until recently much of what we understand about intimate partner violence (IPV) committed against older women has been gleaned from the research on elder abuse.¹ Intimate partner violence is psychological, physical, or sexual abuse by 1 intimate partner against the other.² Elder abuse is generally a vulnerable adult or physically or mentally incapacitated adult who is unable to care for him/herself (OH Statute 5101.61). In addition to IPV, it also includes neglect, self-neglect, and forms of financial abuse.³

Older women experience higher rates of elder abuse than older men even after accounting for their larger proportion of the aging population. In 2001, females made up 59% of the total national population over 65 years of age. Females were victims in 56% of the substantiated adult protective services reports and spouses were the perpetrators in 30% of the cases.⁴ The most widely cited study on elder abuse, a 1986 phone survey of Boston residents, reported that 59% of the perpetrators were spouses.⁵

There are few sources with rates of IPV among older women. The 1993 to 2001 National Crime Victimization Survey estimated that 0.044% of women over 55 years experienced nonlethal victimizations by an intimate partner annually (both reported and not reported to the police) compared with 1.23% of women aged 12 to 24 years and 0.087% of women 25 to 54 years of age.⁶ A study of women 50 to 79 years old living independently who were enrolled in the Women's Health Initiative demonstrated 2% experiencing physical abuse and 10% reporting verbal abuse in the past year by an intimate partner.⁷

A sample of IPV victims (ages 18 to 64 years) utilized more health care than nonvictims, with annual health care costs almost 50% (48.8%) more than age-matched women who are not abused.⁸ Intimate partner violence victims present with a variety of chronic health and mental health diagnoses such as headache, irritable bowel syndrome, chronic pain, depression, and posttraumatic stress disorder.⁹ As a result primary care physicians typically have frequent contact with IPV victims. Physicians' rates for asking about IPV are less than 10%.^10–13 Focus groups with health care providers about IPV in older women demonstrated that most do not consider that it might be an issue.¹⁴ Hence, physicians are missing opportunities to assist victims or may provide inappropriate referrals. For example, marriage counseling is not recommended for couples living with IPV.

To date studies examining the prevalence of IPV in primary care offices are presented for women of all ages; 5.5% to 22% of women report physical abuse in the past year. Lifetime prevalence of IPV is 21.4% to 39%.^15–18 The primary purpose of this study was to determine the incidence and prevalence of IPV among women over 55 years in ambulatory internal medicine and family medicine offices. Secondly, we examined the relationship between IPV and self-reported health conditions.

METHODS

Sample

Adult primary care offices affiliated with an academic center in Southwestern Ohio provided patient lists of females 55 years and older. A total of 4,261 names and phone numbers were available for women who had at least 1 visit to the office in the last 12 months. Sample size was calculated using an IPV rate of 1.2% for women aged 55 years or older. A simple random sample of 600 patients from a pool of 3,200 was calculated to estimate the IPV rate in the population with an error bound of 0.9%; 200 from each age group: 55 to 64, 65 to 74, and over 75 years old. To have the same error bound assuming an IPV rate of 2.5%, 1,125 subjects were required. As a result, we attempted to enroll 1,000 subjects.

The patient lists were stratified into the 3 above-mentioned age groups. Trained female interviewers called each woman on the list between March and June 2003. A phone number on the list was called at least 3 times at different times during the day and on differing days of the week before we considered the patient unavailable and took her out of the sample. Participants were called until we had completed 1,000 surveys with at least 200 in each age group.

Women gave verbal consent by agreeing to participate in the Women's Health and Relationship Survey. Respondents were asked to provide their age, birth date, and the current year to assess their mental status. Failing to answer any of these 3 questions correctly resulted in termination from the study. Less than 1% (0.71%, n=26) did not pass this mental competency test.

Of the available phone numbers, 297 of the numbers were disconnected, 261 were wrong numbers, and 67 of the women were deceased, leaving 3,636 working phone numbers. We were unable to reach 50.93% (n=1,852) (i.e., called at 3 different times on different days and repeatedly got the answering machine, no answer, or the interviewer was repeatedly asked to call back). Of the remaining 1,784 available respondents, 19.11% (n=695) refused to participate, 0.41% (n=15) were intercepted by family members refusing the woman's participation, 1.24% (n=45) were too sick, 0.71% (n=26) did not pass the mental status questions, and 0.22% (n=8) were dropped because they refused to answer the abuse questions. There were 995 usable surveys. The unadjusted response rate was 27.4% (995/3,636). Adjusting the denominator for the 1,852 that could not be reached by phone resulted in a 55.8% (995/1,784) response rate.

Instrument

The survey was adapted from validated instruments and included questions about mental status,¹⁹ health conditions,²⁰ IPV,^20,21 and selected demographic information. Items identifying behaviors consistent with the Center for Disease Control and Prevention's definition of IPV were included.² Close-ended questions asked about psychological/emotional abuse (criticized, shouted, jealous), controlling behaviors, and threats of physical harm; physical and sexual abuse since the woman turned 55 years old (prevalence) and in the last year (incidence). For example, “Since you turned 55 has someone close to you pushed, grabbed or shoved you?” Answer options included: never, rarely, occasionally, frequently, or very frequently. If the participant's response was affirmative she was asked, “Who did this to you” and given the options of spouse, child, grandchild, other relative, friend, or other. The respondent was then asked, “In the last year since MONTH, 2002 has someone close to you pushed, grabbed or shoved you?” and given the response options described above. The respondent was then asked, “Can you tell me about the most recent time or the one you remember the most?” This was followed by, “Have you ever told anyone about this such as doctor, clergy, police, friend, relative, other?” Because of question burden we excluded questions about the perpetrator for some psychological/emotional abuse questions and limited questions about whom the subject told to physical and sexual abuse.

Demographic and health questions were asked first, then abuse questions in the following order: psychological/emotional abuse, including controlling behaviors and threats of physical harm; physical abuse; and sexual abuse. Survey administration took 20 to 45 minutes depending on whether or not the woman had experienced any abuse. Local aging and domestic violence resources were offered to each individual upon completion of the interview; 13% of the participants requested this information. The University of Cincinnati IRB approved the study protocol.

Analysis

Descriptive statistics and tests of significance (χ² test of independence and t-tests for independent groups) for bivariate relationships were calculated using SPSS 11.5 software.

RESULTS

The mean age was 69.1 years (SD 8.35). The sample size for each age group who completed the survey was: age 55 to 64 years, n=337 (34%); 65 to 74 years, n=366 (37%); over 75 years, n=285 (29%). Table 1 presents the demographic characteristics of the sample broken down by IPV status. The only significant relationship is between current marital status and IPV status, 8.6% of the currently partnered women experienced IPV since turning 55 years old.

Table 1.

Demographic Characteristics of Women 55 Years of Age and Over in Adult Primary Care Offices by Intimate Partner Violence (IPV) Victim Status (n=995)

Demographic Characteristics	IPV Victim Status Since Turning 55 Years Old
	Victim	Nonvictim
Age
Mean (SD)	67.54 (7.51)	69.19 (8.4)
Race, n (%)
Caucasian	64.8 (35)	54.4 (499)
African American	35.2 (19)	45.6 (418)
Current marital status
Currently partnered†	66.7 (36)*	41.4 (381)
Not currently partnered	33.3 (18)	58.9 (547)
Current household composition
Lives alone	21.8 (12)	39.1 (365)
Lives only with spouse	34.5 (19)	25.9 (242)
Lives w/spouse+at least 1 other family member	23.6 (13)	7.5 (70)
Lives w/children or grandchildren	12.7 (7)	21.9 (204)
Lives w/one other person‡	7.3 (4)	5.6 (52)
Income
<$20,000	44.4 (24)	46.3 (428)
≥$20,000 and ≤$40,000	24.1 (13)	15.0 (139)
>$40,000	14.8 (8)	12.0 (111)
Refused or did not know	16.7 (9)	26.6 (246)
Current employment status
Never worked outside the home	13.0 (7)	7.0 (65)
Currently employed	20.4 (11)	15.7 (146)
Unemployed and looking for work	13.0 (7)	12.1 (112)
Retired	53.7 (29)	65.2 (605)
Education level
Less than high school	35.2 (19)	33.7 (313)
High school graduate	27.8 (15)	30.4 (282)
Some college/college graduate	29.6 (6)	29.6 (275)
Some graduate school/graduate school
Graduate	7.4 (4)	6.3 (58)
Independence
Dresses self	90.9 (50)	95.7 (892)
Administers own medications	98.2 (54)	94.6 (882)
Drives self	67.3 (7)	60.7 (566)

A t-test for independent samples was performed on the age variable. χ² test of independence was calculated for the other relationships. The only significant relationship is between current marital status and IPV status.

Note all the percentages are based on n=995. The percentages are based on the number of women who gave valid responses to a specific demographic question. “Don't know” and “refused” responses were excluded for the calculation of the responses.

^*P<.001.

^†Current partner includes spouse, common law husband, boyfriend, or gentleman caller.

^‡Other includes parents, in-laws, nephew, friend, godson, roommate, and unspecified family member.

Psychological/emotional abuse was reported most frequently, with 45.2% occurring since age 55 years (prevalence) and 31.7% in the past year (incidence). Because of question burden the perpetrator was not identified for questions about name calling, criticizing, shouting, swearing and jealous, or possessive behaviors. Table 2 describes the IPV rates in the sample for other forms of psychological/emotional abuse, controlling behaviors and threats; physical abuse; and sexual abuse. Controlling behaviors and threats of physical harm were the highest with 2.43% and 2.63%, respectively, since turning 55 years. Physical abuse was reported with rates of 1.52% (prevalence) and 0.41% (incidence). Sexual abuse rates were 2.14% (prevalence) and 1.12% (incidence). Overall abuse rates were 5.36% (prevalence) and 3.14% (incidence).

Table 2.

Types of Controlling Behavior, Threats of Physical Violence, Physical and Sexual Abuse by an Intimate Partner Reported by Women Aged 55 Years and Over

Type of Abuse	Prevalence of Women Reporting Abuse,* % (n) [95% CI]	Incidence of Women Reporting Abuse†, % (n) [95% CI]
Psychological/emotional abuse
Controlling behaviors‡	2.43 (24) [1.47 to 3.39]	1.21 (12) [0.53 to 1.90]
Routinely checked up on you in a way that made you afraid	1.32 (13) [1.0 to 2.0]	0.71 (7) [0.0 to 1.0]
Put you on an allowance	1.01 (10) [0.0 to 2.0]	0.41 (4) [0.0 to 1.0]
Not letting you go to work or social activities, talk or see friends	1.11 (11) [0.0 to 2.0]	0.51 (5) [0.0 to 1.0]
Threats of physical violence‡	2.63 (26) [1.63 to 3.63]	1.62 (16) [0.83 to 2.41]
Said things to scare you	1.22 (12) [1.0 to 3.0]	0.81 (8) [0.0 to 1.0]
Thrown, hit, kicked or smashed something	2.04 (20) [1.0 to 3.0]	1.12 (11) [0.0 to 2.0]
Physical abuse‡	1.52 (16) [0.75 to 2.28]	0.41 (4) [0.01 to 0.80]
Pushed, grabbed, or shoved you	0.81 (8) [0.0 to 1.0]	0.20 (2) [0.0 to 0.0]
Slapped, hit, or punched you	0.71 (7) [0.0 to 1.0]	0.70 (2) [0.0 to 0.0]
Hit you with an object	0.41 (4) [0.0 to 1.0]	0.0 (0) [0.0 to 0.0]
Choked or attempted to drown you	0.61 (6) [0.0 to 1.0]	0.20 (2) [0.0 to 0.0]
Sexual abuse‡	2.14 (21) [1.22 to 3.03]	1.12 (11) [0.0 to 2.0]
Pressured you to have sex in a way you did not like or want	2.04 (20) [1.00 to 3.0]	1.12 (11) [0.0 to 2.0]
Physically forced you to have sex	0.41 (4) [0.0 to 1.0]	0.20 (2) [0.0 to 0.0]
Attacked the sexual parts of your body	0.31 (3) [0.0 to 1.0]	0.10 (1) [0.0 to 0.0]
Total	5.36 (53) [3.96 to 6.77]	3.14 (31) [2.05 to 4.23]

^*Prevalence is defined as women who reported abuse since the age of 55 years old.

^†Incidence is defined as women who reported abuse in the last 12 months.

^‡Respondents who indicated “do not know” or refused to answer the question are not included in the denominator. The greatest number of respondents not included in any 1 question was 18.

CI, confidence interval.

We asked women who had experienced physical and sexual abuse since 55 years if they had told anyone about it. The results are in Table 3. More than half (11/16, 68.8%) of the victims of physical abuse had told someone, usually “another person” (relative, friend, or unspecified person). Sexual abuse was disclosed less frequently (7/21, 33.3%) and always to another person. No victim of sexual abuse reported it to a doctor and only 2 of the 16 victims of physical abuse told a doctor.

Table 3.

The Prevalence of Types of Intimate Partner Violence Reported by Women Over 55 Years and Whom They Told

Type of Abuse	Percent of, Abused Women, Who, Told Someone (n)	Who Women Told*
		Doctor (n)	Clergy (n)	Police (n)	Other†, Person (n)
Physical abuse (n=16)	11	2	0	6	8
Sexual abuse (n=21)	7	0	0	0	7

^*Respondents could give more than 1 response.

^†The “other” category includes relative, friend, or unspecified person.

Intimate partner violence victims who had experienced abuse in the past year reported a mean number of 3.84 health conditions compared with 3.21 for non-IPV victims (t-test, 2 tailed, P=.055). Mean number of conditions for IPV victims and nonvictims since age 55 years were 3.49 and 3.22, respectively (t-test, 2 tailed, P=.280). Table 4 shows health conditions associated with victim and nonvictim status. Both chronic pain and depression were significantly higher in IPV victims than nonvictims since age 55 years and in the last year (P<.05). When asked if their health condition interfered with normal daily activities such as “seeing friends or relatives, going out, doing the things you need or like to do,” there was no significant difference between IPV victims and nonvictims. (Results not shown.)

Table 4.

Self-Reported Health Conditions Associated with Women 55 Years and over Who Were Victims of Intimate Partner Violence (IPV) Compared with Nonvictims

Health Condition	Prevalence			Incidence
	Victim, % (n)	Nonvictim, % (n)	P-Value	Victim, % (n)	Nonvictim, % (n)	P-Value
Chronic pain (e.g., back pain, migraines)	50.9 (27)	36.8 (343)	.039	58.1 (18)	36.9 (352)	.017
Depression or anxiety	49.1 (26)	30.4 (287)	.005	54.8 (17)	30.7 (293)	.004
Digestive problems (e.g., irritable bowel, ulcer, heartburn)	34.0 (18)	31.1 (290)	.664	45.2 (14)	30.8 (294)	.090
High blood pressure or heart problems	59.6 (31)	74.4 (697)	.160	48.4 (15)	74.7 (713)	.001
Lung problems (e.g., asthma, emphysema, or COPD)	23.1 (12)	23.5 (219)	.948	30 (9)	23.2 (222)	.390
Diabetes or thyroid problems	37.3 (19)	37.4 (349)	.978	44.8 (13)	37.2 (355)	.404
Bone or joint problems (e.g., arthritis or osteoporosis)	66.0 (35)	63.5 (591)	.707	67.7 (21)	63.5 (605)	.628
Stroke, nerve problems (e.g., MS or Parkinson's)	17.0 (9)	12.7 (119)	.370	19.4 (6)	12.7 (122)	.281

χ² tests of independence were performed.

COPD, chronic obstructive pulmonary disease

DISCUSSION

Cross-sectional surveys of U.S. households demonstrate that physical abuse becomes less frequent with age.²² Consistent with these findings, our rates of physical abuse were less than those of other IPV primary care studies that sampled predominantly younger women. However, rates of psychological/emotional abuse in our sample of older women (not including controlling behaviors and threats of physical violence) approached those of younger women.^15–18 What this study adds to IPV research is an assessment of the rates of IPV (controlling behaviors, threats, physical, and sexual abuse) among women over 55 years of age who are receiving care in family medicine and internal medicine practices.

We were unable to find any published study that examined IPV perpetrators in older women. Therefore, only the elder abuse literature is available for comparison. Pillemer identified spouses as the perpetrators of elder mistreatment (physical violence, chronic verbal aggression, and neglect) 59% of the time (37/63) and children 24% (15/63).⁵ The perpetrator of physical violence was most commonly the spouse. Hwalek et al.²³ also found that caregivers perpetrated 52% of elder abuse, with 39% being adult children and 14% spouses. Our data serves as a reminder that IPV does occur in older women.

Older women infrequently told someone about their abuse and when they did, it was either to a friend or relative. This is consistent with other studies; abuse is most commonly reported to nonauthority figures.²⁴

The lack of disclosure to physicians is also noteworthy. Other studies suggest that women want physicians to ask about IPV, but may chose not to disclose for a myriad of reasons, such as shame and embarrassment, worried about repercussions from the abuser and wanting to protect the perpetrator.^25–30 Our own work with older women demonstrated that only 20/38 had spoken with their physicians about the abuse at some point. Their reasons were similar to those of younger women, but they also talked about traditional mores of the times, i.e. do n'ot air your dirty laundry in public, and dependency on the abuser or visa versa because of failing health.^31,32

In younger women, both Campbell and Coker found that victims reported poorer health status than nonvictims.^33,34 Mouton³⁵ evaluated self-reported health in older women using a validated instrument that measures physical and mental functional status (SF36). Intimate partner violence victims had lower physical and mental health scores than nonvictims. In line with these findings, our data confirm other studies that report depression^34,36 and chronic pain^33,34,37 as common among IPV victims. Although digestive problems only approached significance in our sample, it is a common condition in younger IPV victims.^8,33,38 Thinking of these as “red flag” diagnoses, conditions frequently associated with IPV, should trigger health care providers to inquire about IPV when an older patient with 1 of these conditions is seen.^8,39

There are limitations to this study. First, we sampled a predominantly urban population, which included significantly more African Americans and lower socio-economic status than the statistics for the county or state. Therefore, these results may not be generalized to nonurban settings. Second, some women may have failed to disclose abuse to the interviewers because of the generational mindset that IPV is a private issue.³¹ Third, all data are self-report and not corroborated. Our methods did not allow for obtaining demographic information for nonrespondents. Finally there are limitations to phone interviewing as a data collection method. Forty-four percent were not available to talk with an interviewer. Patients who are more transient, have limited English language skills or do not have access to a telephone may have been missed during our field period.

Two strengths of our work are that we used questions from validated national-level instruments and employed and trained female interviewers. We were also mindful of the ordering of questions, with the more sensitive abuse questions asked during the later part of the survey. As a result only 8 women did not answer the abuse questions.

In conclusion, this study represents a first effort to establish prevalence and incidence rates of IPV among older female patients in primary care. Although rates of physical abuse are less than those for younger women, psychological/emotional, both physical and sexual abuse occur. Our findings serve to remind health care providers to think about IPV in older women, especially those with depression or chronic pain. Directly asking older women about abusive experiences may be beneficial, as our results suggest that they may not initiate the disclosure.