Intimate partner violence exacts an enormous toll each year in the United States, affecting 3% to 5% of adult intimate relationships.1–3 Compared with urban settings, much less is known about the prevalence and correlates of partner violence in rural areas.4–6 Rural women are more isolated, have access to fewer services, and face different attitudes and norms than urban women.5–7
This study examined the prevalence of severe physical abuse, measured by the Conflict Tactics Scale,8 and controlling emotional abuse, measured by Yllo’s Controlling Behavior Questions,9 as reported by a cohort of cohabiting couples in a rural Iowa county. The objectives were to estimate the prevalence of abuse victimization reported by men and women on each scale and to identify factors associated with violence against men and women.
The study population included members of a prospective cohort study in a rural county. The county, cohort, and overall methods are described elsewhere.10 Respondents aged 18 and older were asked whether they were currently living with a spouse or partner and, if so, were asked to respond to questions about partner violence. Participants were interviewed in person and without their partner being present. Of the 1633 adult cohort members, 1310 (80.2 %) lived with their partners.
We used 5 items from the Conflict Tactics Scale8 to measure severe physically abusive acts perpetrated by the respondent’s partner during the last 12 months. As recommended by Straus,11 these severe abuse items are kicking, hitting with a fist, hitting with some other object, threatening with a knife or gun, and using a knife or gun to harm.
We used Yllo’s Controlling Behavior Questions to assess controlling emotional abuse in the previous 12 months. We used the following items:
You felt intimidated and frightened (e.g., by your husband’s/wife’s/partner’s shouting, looks, smashing things). You felt isolated by your husband/wife/partner controlling whom you could see or call or where you could go. You felt you were treated like a subordinate, like a servant by your husband/wife/partner, making you wait on him or her or making important decisions alone. You felt you were made to do sexual things against your will.1. 2. 3. 4.
The Conflict Tactics Scale and the Controlling Behavior Questions were independently coded into dichotomous variables, so that a positive response to any one item indicated the presence of that type of abuse.
Prevalence of abuse measured by each scale was compared by gender, and χ2 tests of independence were used to determine whether reported abuse varied by age, marital status, education, residence, and farm work. Multivariate logistic regression analysis was conducted to identify factors associated with abuse as measured on each scale independently and to determine whether these differed among men and women. The odds ratios (ORs) should not be interpreted as risk ratios on the Controlling Behavior Questions because emotional abuse was not a rare outcome.
Based on the Conflict Tactics Scale, 2.9% of the women and 4.7% of the men reported at least 1 incidence of severe physical violence directed at them by their partners (Table 1) (P = .11). On the Yllo Controlling Behavior Questions, however, 46.7% of the women and 30.2% of the men reported experiencing emotional abuse perpetrated by their partners (P < .001).
Responses of the individuals on the 2 original scales were significantly associated (Spearman rank correlation coefficient = 0.123 for the overall sample, 0.120 for men, and 0.149 for women; all correlations, P < .001).
For the overall sample, women reported experiencing more controlling emotional abuse from their partners (OR = 1.84; 95% confidence interval [CI] = 1.30, 2.61) (Table 2). Increasing age and being married were protective against both types of victimization for both men and women. Education was unrelated to prevalence of any type of abuse.
Women living in a rural nonfarm residence reported the highest victimization of severe physical abuse (OR = 5.30; 95% CI = 1.77, 15.88). When controlling for farm residence, men who engaged in farm work experienced more severe physical abuse (OR = 4.00; 95% CI = 1.46, 10.86) and controlling abuse (OR = 1.51; 95% CI = 0.91, 2.49) than men who did not. A similar result was not seen among women.
Partner violence is an important public health problem in rural and urban areas. Indicators of abuse in rural populations differ from those in urban areas. Gender appears to be a strong effect modifier for both severe physical and emotional abuse. Women who live outside of towns but not on farms reported more physical abuse than other women. Increased isolation might explain this finding. Women living on farms may have more interaction with others (e.g., farm workers) than women who do not live on farms. In addition, some individuals might choose rural nonfarm areas because of the isolation, and one reason to prefer isolation is to hide abusive behavior.
Men, but not women, who engaged in farm work reported experiencing more abuse than did nonfarmers. This finding has not been reported previously. The strains of farm work on relationships, which include exhausting physical labor, long hours, and financial uncertainty, may lead to increased risk of abuse by women, but this finding needs to be studied further.
Measurements of abuse should include both physical and emotional components. Further study of factors associated with different types of abuse in rural environments will help direct the development and implementation of programs that offer assistance to victims and prevent violent behavior.
Chi-square tests indicate significance at level: *P < .05; **P < .01; ***P < .001. Note. OR = odds ratio; CI = confidence interval. aAll variables were included in logistic models. bControlling behavior is not a rare outcome, so odds ratios do not approximate risk ratios. *P < .05.
Distribution by Gender Conflict Tactics Scale Yllo’s Controlling Behavior Questions Men Women Men Women Men Women No. (%) No. (%) No. Abused Prevalence No. Abused Prevalence No. Abused Prevalence No. Abused Prevalence Total 621 (47.4) 689 (52.6) 29 4.7 20 2.9 187 30.1 322 46.7 Age, y 18–29 18 (2.9) 36 (5.2) 1* 5.6 3* 8.3 13*** 72.2 19* 52.8 30–39 111 (17.9) 161 (23.4) 12 10.8 8 5.0 48 43.2 92 57.1 40–49 144 (23.2) 159 (23.1) 4 2.8 3 1.9 54 37.5 68 42.8 50–65 194 (31.2) 210 (30.5) 8 4.1 2 1.0 43 22.2 96 45.7 ≥ 66 154 (24.8) 123 (17.9) 4 2.6 4 3.3 30 19.5 48 39.0 Marital status Married 593 (95.6) 663 (96.2) 25* 4.2 15** 2.3 176 29.7 308 46.5 Other 27 (4.4) 26 (3.8) 4 14.8 5 19.2 12 44.4 15 57.7 Education ≤ High school 364 (59.7) 293 (44.7) 17 4.7 6 2.0 93* 25.5 134 45.7 > High school 246 (40.3) 362 (55.3) 12 4.9 14 3.9 92 37.4 175 48.3 Residence Farm 238 (38.3) 251 (36.4) 12 5.0 3*** 1.2 55*** 23.1 130 51.8 Rural nonfarm 120 (19.3) 150 (21.8) 4 3.3 12 8.0 38 31.7 64 42.7 Town 263 (42.4) 288 (41.8) 13 4.9 5 1.7 95 36.1 129 44.8 Farm work Yes 320 (51.6) 181 (26.3) 21* 6.6 1*** 0.6 90 28.1 90 49.7 No 299 (48.2) 507 (73.6) 8 2.7 19 3.7 97 32.4 232 45.8 Conflict Tactics Scale Yllo’s Controlling Behavior Questions Total Men Women Total Men Women Independent variablea OR 95% CI OR 95% CI OR 95% CI ORb 95% CI OR 95% CI OR 95% CI Female gender 0.66 0.23, 1.86 1.84 1.30, 2.61 Increasing age category 0.72 0.55, 0.94 0.68 0.48, 0.96 0.77 0.50, 1.18 0.76 0.68, 0.84 0.64 0.54, 0.76 0.84 0.73, 0.97 Married 0.24 0.10, 0.55 0.35 0.11, 1.15 0.16 0.08, 0.54 0.69 0.39, 1.22 0.72 0.31, 1.67 0.67 0.29, 1.50 > High school education 1.04 0.56, 1.94 0.76 0.34, 1.72 1.58 0.56, 4.46 1.15 0.91, 1.47 1.39 0.96, 2.02 0.99 0.72, 1.37 Rural nonfarm residence 0.62 0.18, 2.01 0.52 0.16, 1.68 5.30 1.77, 15.88 1 0.63, 1.61 0.75 0.46, 1.23 0.91 0.61, 1.36 Residence on farm 0.62 0.25, 1.50 0.42 0.16, 1.13 1.60 0.34, 7.60 0.85 0.61, 1.20 0.36 0.20, 0.63 1.40 0.91, 2.16 Farm work 3.17 1.18, 8.54 4.00 1.46, 10.86 0.18 0.02, 1.62 0.89 0.59, 1.35 1.51 0.91, 2.49 0.90 0.58, 1.41 Interaction of gender and farm work 0.86 0.01, 0.79 1.29 0.77, 2.14 Interaction of gender and rural nonfarm residence 6.51 1.50, 28.24 0.73 0.41, 1.33 Model goodness of fit 6.13 9.4 17.9* 12.9 5.7 8.8
Support for this project was provided by the Great Plains Center for Agricultural Health (Centers for Disease Control and Prevention/National Institute for Occupational Safety and Health, U07/CCU706145) and the University of Iowa Injury Prevention Research Center (Centers for Disease Control and Prevention/National Center for Injury Prevention and Control, CCR 703640), both housed in the University of Iowa Department of Occupational and Environmental Health.
Human Participant Protection
The Keokuk County Rural Health Cohort Study was approved by the University of Iowa human subjects review board.
