A Comparative Assessment of Intimate Partner Violence Perpetration among Male Military Personnel and Civil Servants in Ibadan, Nigeria

Abstract

The mode of training and work of the military personnel is different from that of civil servants and may affect their relationships with their intimate partners. This study assessed and compared the prevalence and correlates of self-reported Intimate Partner Violence (IPV) perpetration against female partners by male military personnel and civil servants in Ibadan, south-western Nigeria. A cross sectional survey of 1240 respondents, comprising 631 military personnel and 609 civilians, was conducted using a multistage sampling technique. A self-administered questionnaire was used to collect data on the respondents’ characteristics and the different types of IPV perpetrated by them. Chi square test was used to compare the proportions of the different types of IPV perpetration and logistic regression analysis was used to determine the predictors of IPV perpetration in the two study groups. Military personnel reported significantly more IPV such as physical abuse, psychological abuse and controlling behaviors than the civil servants. Childhood exposure to inter-parental IPV and history of physical fight with another woman significantly increased the odds of perpetration of each type and any form of IPV in the two populations after controlling for other variables. Military personnel were significantly more likely to perpetrate any form of IPV than the civil servants. IPV was prevalent in both groups but was more among the military personnel. There is a need for multidisciplinary interventions such as psycho-education and conflict management skills to address violence against female intimate partners especially among the military population in Nigeria.

Introduction

Intimate Partner Violence (IPV) is a global public health problem. It is an act, behaviour or attitude which result in, or is likely to result in physical, sexual or psychological harm or suffering and it can affect social, economic, religious and cultural groups (Krug et al. 2002). The World Health Organization (WHO) multi-country study described IPV experienced by women as physical, emotional, or sexual abuse and controlling behaviors by a current or former intimate male partner (Garcia-Moreno et al. 2005).

These actions can cause substantial burden on the victims and their children as well as the communities and health care systems (Kimberg 2008). IPV is an important cause of morbidity and mortality which can result in negative physical, mental, sexual and reproductive health outcomes for women (Garcia-Moreno et al. 2005; Krug et al. 2002). Evidence has shown that exposure to IPV is associated with increased likelihood of traumatic physical consequences which vary from minor to life threatening injuries for women (Antai 2011b; Ellsberg et al. 2008). Fatal health consequences include maternal mortality, homicide and suicide (WHO 2009). Children exposed to IPV are more likely to display behavioural, emotional and academic problems (Kernic et al. 2002, 2003) and to have psychological and physical health problems in adulthood (Dube et al. 2002) as well as the likelihood to perpetrate IPV in adulthood (Wood and Sommers 2011). In addition to its human costs, violence places an enormous economic burden on societies in terms of lost productivity and increased use of social services (WHO 2012).

Military population are not less likely than the civilian population to perpetrate IPV. Studies comparing the military population with the civilian population in the United States of America showed higher rates of IPV perpetration among the military (Cronin 1995; Griffin and Morgan 1988; Heyman and Neidig 1999; Smith-Marek et al. 2016). Studies indicated that IPV occurs in military families (Hoyt et al. 2014; Jones 2012; Rentz et al. 2006; Rodrigues et al. 2015). The rates of IPV across US military population range from 13.5% to 58% (Jones 2012) and 36.9% of Korean Air Force personnel perpetrated IPV (Seo et al. 2014). The military environment possesses a unique set of circumstances and stressors that may promote family violence including IPV. These include periods of separation from family, balancing the demands of family life and work, stress related to the military mission, frequent relocation, as well as financial stressors (Bradley 2007; Brannen et al. 1999; Newby et al. 2005; Rentz et al. 2006).

There is a gap in literature on self-reported perpetration of IPV and the associated risk factors by the military personnel in developing countries including Nigeria. A related study on IPV among the military population in Nigeria focused on the experience of female victims (Chimah et al. 2015). However, studies among civilian men in Africa also showed that IPV is a major public health problem. About 40% of married men reported IPV perpetration in Uganda (Speizer 2010) and 27.5% of men reported using physical violence against their current or most recent female partners in South Africa (Gupta et al. 2008). In Ile-Ife, Nigeria, 50.5% of the men reported carrying out at least one episode of psychological abuse, 13.1% of them reported physical violence while 6.8% reported sexual abuse against their wives (Fatusi and Alatise 2007). In a study among civil servants in Ibadan, Nigeria, 42.5% of men perpetrated physical IPV (Fawole et al. 2005) while in another cross-sectional survey of married men in Ibadan, 44.1% reported perpetration of IPV in their lifetime (Fawole et al. 2010).

One of the motives behind domestic abuse including IPV is to gain control over the victims. Military culture, notably its male dominance (Stander et al. 2011) and its use of violence as a legitimate tool to settle conflicts, might play a role in the perpetration of IPV against their female partners (Adelman 2003; Jones 2012; Zamorski and Wiens-Kinkaid 2013). The power theory suggests that power imbalances between partners may increase tension within the family unit and consequently increases the risk of IPV. According to the power theory, individuals employ violence to settle conflicts within the family and between intimate partners because this way of addressing conflicts has been learned in childhood by either witnessing or experiencing physical abuse. (Burelomova et al. 2018). Abusers use different tactics to exert power over their partners. These include humiliation, threats, dominance and isolation (Garcia-Moreno et al. 2005). Other risk factors for the perpetration of IPV against female partners include intrapersonal vulnerability such as misuse of alcohol, exposure to abuse by witnessing parental IPV as children and history of perpetration of maltreatment (Bowen et al. 2017; Dube et al. 2002; McCarroll et al. 2009; Smith-Marek et al. 2016).

In order to plan for the development of effective interventions, it is important to understand the type of IPV and context in which they occur. This comparative study assessed the prevalence and identified the risk factors associated with the perpetration of different types of IPV among male military personnel and civil servants in Ibadan, Nigeria. This study also compared and assessed the health complaints as a result of the physical IPV perpetrated on the female partners. Information obtained from this study would be useful for the public health providers in developing targeted interventions to reduce the occurrence of IPV in military and civilian populations.

Method

The study area was Ibadan, the capital city of Oyo State in south-western Nigeria.

Participants

Study Population

The study population comprised of male civil servants working in the selected ministries in the Oyo State Secretariat, Ibadan and military personnel working in 2 Division Nigerian Army, Ibadan. This consisted of both junior and senior staff in the civil service as well as commissioned and non-commissioned officers in the Army. All consenting men in the selected ministries or units who had been working for at least two years and had been in intimate relationships with female partners for at least one year were eligible to participate in the study.

Study Design

A comparative cross sectional study design was used.

Procedures

Sample Size Determination

Sample size was estimated using the sample size formula to compare independent proportions (Fielding et al. 1993) based on the prevalence of intimate partner violence perpetration among married men (44.4%) in the general population in Ibadan, Nigeria (Fawole et al. 2010). An expected difference of 15% was used at 95% confidence interval and a power of 90%. After adjusting for clustering effect (2.0) by doubling the calculated sample size to increase precision (Fielding et al. 1993) and after considering non response rate by increasing the calculated sample size by 20%, a minimum sample size of 570 respondents per group was obtained.

Sampling Technique

A multi-stage sampling technique was used in this study.

Selection of civil servants: In stage one, eight ministries were selected by simple random sampling (balloting) from a sampling frame of all the fifteen ministries in the Oyo State Secretariat, Ibadan. In stage two, from each selected ministry, four departments were selected by balloting without replacement from the list of the departments. In stage three, each department was taken as a cluster and all consenting male civil servants, present in each department, in the selected ministries were interviewed.

Selection of military personnel: In stage one, seven branches were selected by simple random sampling balloting) from a sampling frame of all the twenty-four branches in 2 Division Nigerian Army, Ibadan. In stage two, from each of the selected branches, three units were selected by balloting without replacement from the list of the units. In stage three, each unit was taken as a cluster and all consenting male army personnel, present in each unit, in the selected branches were interviewed.

Data Collection

A semi-structured and self-administered questionnaire was used for data collection. The questionnaire was developed from the Revised Conflict Tactics Scale (Straus et al. 1996; Thompson et al. 2006) and the WHO standardized questionnaire for the multi-country study on women’s health and domestic violence (Garcia-Moreno et al. 2005). The questions were modified to address the objectives of this study. The questionnaire elicited information on respondent’s sociodemographic characteristics, perpetration of controlling behaviours, psychological abuse, sexual and physical violence. Pre-testing of the questionnaire was carried out among staff of the Federal Civil Service and Military personnel in the Nigerian Air Force both located in Ibadan. Ambiguous questions were revised to ensure clarity. To ensure that questionnaires were completed, trained research assistants were available to explain the questions to respondents who required assistance.

Ethical approval, with reference number AD 13/479/128, was obtained from the Oyo State Ministry of Health Ethical Review Committee. Permission to conduct this study was also given by the Head of Service, Oyo State Secretariat, Ibadan and the General Officer Commanding, 2 Division Nigerian Army, Ibadan. Written informed consent was obtained from each participant. Confidentiality and anonymity was maintained. Participants in need of help such as counseling were referred to appropriate institutions.

Measures

Study Variables

The dependent variable was perpetration of IPV (physical violence, sexual violence, psychological/emotional abuse and controlling behavior).

Physical violence was assessed with six questions. Each respondent was asked if he slapped or threw something that could hurt at his partner; pushed, shoved or pulled his partner’s hair; hit his partner with his fist or some object that could hurt; kicked, dragged or beat his partner up; tried to choke or burn his partner on purpose; or threatened to use or actually used a gun, knife, cutlass or other weapon against his partner. Respondents who gave a positive answer to any question had a score of 1 and negative answers to all questions had a score of 0. A score of 1 or more in any category of physical intimate partner violence was taken as perpetration of physical violence.

Sexual violence was assessed with four questions. Each respondent was asked if he physically forced his partner to have sexual intercourse with him when she did not want it; if he threatened her to have sexual intercourse with her; if he forced her to do something sexually that she found humiliating or degrading; or if he made his partner have sexual intercourse with him without a condom when she indeed wanted to use it. Respondents who gave a positive answer to any question had a score of 1 and negative answers to all questions had a score of 0. A score of 1 or more in any category of sexual intimate partner violence was taken as perpetration of sexual violence.

Psychological/emotional abuse was assessed with five questions. Each respondent was asked if he insulted or made his partner feel bad about herself; humiliated or disgraced her in front of other people; threatened to hurt his partner or someone she cared about; destroyed something belonging to his partner intentionally; or if he did some things to scare or intimidate her on purpose. Respondents who gave a positive answer to any question had a score of 1 and negative answers to all questions had a score of 0. A score of 1 or more in any category of psychological/emotional abuse was taken as perpetration of psychological/emotional abuse.

Controlling behavior was assessed using six questions. Each respondent was asked if he tried to keep his partner from seeing her female friends; restricted his partner’s contact with her family of birth; insisted on knowing where his partner was at all times; got angry/jealous if his partner talked with other men; frequently accused his partner of being unfaithful; or if the respondent expected his partner to ask for his permission before seeking health care for herself. Respondents who gave a positive answer to any question had a score of 1 and negative answers to all questions had a score of 0. A score of 1 or more in any category of controlling behaviour was taken as perpetration of controlling behavior.

The independent variables (coded as categorical predictors) included socio-demographic data such as age (coded as ≥37 years or < 37 years using mean age of 37 years), marital status (coded as married or unmarried), level of education (coded as tertiary or secondary), grade level/rank (coded as senior/commissioned officer or junior/non-commissioned officer) as well as other factors such as childhood exposure to parental IPV (coded as Yes or No), history of physical fight with a woman (coded as Yes or No), use of alcohol (coded as Yes or No) and length of years in the current intimate relationship (coded as ≥11 years or < 11 years).

Data Analysis

Data analysis was done using SPSS version 18. Frequency distributions were presented with appropriate tables. Chi-square statistic was used to compare the proportions of the different types of IPV perpetration and independent variables were entered into multivariate logistic regression model to determine the predictors of perpetration of each type and any form of IPV in the two study groups. Coefficient alpha was set at 0.05 for all analyses.

Results

A total of 1240 men, comprising 631 military personnel and609 civil servants, responded to this survey and provided complete data.

Characteristics of the Respondents

The respondents in the civil service were older than those in the military with the mean age of 38.85 ± 9.95 years compared to 35.13 ± 9.1 years respectively. The civil servants in this study were better educated than the military personnel. About half (50.7%) of the civil servants were senior officers (Grade level 7 and above), while 18.7% of the military personnel were commissioned officers. Majority (75.3%) of the respondents were married; 74.5% of the men in the civil service and 76.1% of the men in the military were married.

Clustering of Violent Behavior

An overview of violent behavior in the present study showed that violence tends to occur in clusters; 5.6% of participants reported engaging in all forms of IPV. Approximately one third (33.6%) of respondents perpetrated at least one form of IPV, 19.9% perpetrated two forms of IPV, 12.8% perpetrated three forms of IPV. Military personnel endorsed perpetration of one or more combination of different forms of IPV than civil servants (p < 0.0001).

Self- Reported Perpetration of IPV

Men in the military perpetrated more acts of physical violence than men in civil service, which included slapping (13.3% vs. 6.7%, p < 0.0001), pushing (7.9% vs. 4.4%, p = 0.011), threatening to use or actually used guns or other weapons (3.5% vs. 1.3%, p = 0.013) and perpetrating any form of IPV against the partner while pregnant (3.3% vs. 1.3%, p = 0.019) Overall, military personnel endorsed higher rates of perpetration of physical IPV than civil servants in the year preceding the study (18.5% vs. 11.7%, p = 0.001). Civil servants were significantly more likely to force their partners to have sexual intercourse without a condom (14.4% vs. 9.5%, p = 0.007). There were no other significant differences between military and civil servant groups with regard to perpetration of sexual violence. (see Table 1).

Table 1.

Comparison of self-reported perpetration of IPV by male military personnel and civil servants

Types of intimate partner violence	Military personnel n = 631 n (%)	Civil servants n = 609 n (%)	Total n = 1240 n (%)	χ2 (p value)
Physical violence: In the last one year, I
Slapped or threw something that could hurt at my partner	84 (13.3)	41 (6.7)	125 (10.1)	14.80(<0.0001)*
Pushed, shoved or pulled my partner’s hair	50 (7.9)	27 (4.4)	77 (6.2)	6.48 (0.011)*
Hit my partner with a fist or some object that could hurt	34 (5.4)	21 (3.4)	55 (4.4)	2.75 (0.097)
Kicked, dragged or beat my partner up	22 (3.5)	16 (2.6)	38 (3.1)	0.77 (0.380)
Tried to choke or burn my partner on purpose	14 (2.2)	11 (1.8)	25 (2.0)	0.27 (0.605)
Threatened to use or used a gun, knife, cutlass or other weapon against my partner	22 (3.5)	8 (1.3)	30 (2.4)	6.20 (0.013)*
Did any of the above to my partner in pregnancy	21 (3.3)	8 (1.3)	29 (2.3)	5.51 (0.019)*
Perpetrated any physical violence	117 (18.5)	71 (11.7)	188 (15.2)	11.42 (0.001)*
Sexual violence: In the last one year, I
Forced my partner to have sexual intercourse	85 (13.5)	72 (11.8)	157 (12.7)	0.76 (0.383)
Used threat to have sexual intercourse with my partner	63. (10.0)	46 (7.6)	109 (8.8)	2.28 (0.131)
Forced my partner into degrading sexual act	32 (5.1)	18.30	50 (4.0)	3.59 (0.058)
Made my partner to have sexual intercourse without a condom even when she wanted a condom	60.(9.5)	88 (14.4)	148 (11.9)	7.20 (0.007)*
Perpetrated any sexual violence	144 (22.8)	140 (23.0)	284 (22.9)	0.01 (0.944)
Psychological abuse: In the last one year, I
Insulted or made my partner feel bad about herself	218 (34.5)	150 (24.6)	368 (29.7)	14.61 (0.001)*
Did something to scare or intimidate her on purpose	61.97	39 (6.4)	100 (8.1)	4.45 (0.035)*
Humiliated or disgraced her in front of other people	70 (11.1)	26 (4.3)	96 (7.7)	20.21(<0.0001)
Threatened to hurt my partner or someone she loves	50 (7.9)	35 (5.7)	85 (6.9)	2.30 (0.129)
Destroyed something belonging to her intentionally	35 (5.5)	28 (4.6)	63 (5.1)	0.58 (0.447)
Perpetrated any psychological abuse	252 (39.9)	190 (31.2)	442 (35.6)	10.32 (0.001) *
Controlling behavior: In the last one year, I
Insisted on knowing where my partner was at all times	355 (56.3)	205 (33.7)	560 (45.2)	63.90(<0.0001)*
Got angry/jealous when she talked with other men	245 (38.8)	137 (22.5)	382 (30.8)	38.78(<0.0001)*
Tried to keep her from seeing her female friends	195 (30.9)	98 (16.1)	293 (23.6)	37.67(<0.0001)*
Expected her to ask for permission before seeking health care for herself	125 (19.8)	109 (17.9)	234 (18.9)	0.74 (0.390)
Frequently accused her of being unfaithful	100 (15.8)	80 (13.1)	180 (14.5)	1.84 (0.175)
Restricted her contact with her family of birth	77 (12.2)	60 (9.9)	137 (11.0)	1.75 (0.187)
Perpetrated any controlling behavior	438(69.1)	318 (52.2)	754 (60.8	37.05(<0.0001)*

Note: χ² – Chi square test

^*- Statistically significant

More men in the military perpetrated psychological/emotional abuse against their intimate partners than those in the civil service, which included insulting or making their intimate partners to feel bad about themselves (34.5% vs. 24.6%, p < 0.0001) and humiliating or disgracing their intimate partners in front of other people (9.7% vs. 4.3%, p < 0.0001). 317 Overall, military personnel endorsed higher rates of perpetration of any form of psychological/emotional abuse against intimate partner in the last year preceding the study (39.9% vs. 31.2%, p = 0.001). (see Table 1).

Men in the military perpetrated all forms of controlling behaviors more than those in civil service in the year prior to the study, which included insisting on knowing where their partners were at all times (56.3% vs. 33.7%, p < 0.0001), getting angry or jealous when their intimate partners talked with other men (38.8% vs. 22.5%, p < 0.0001) and trying to keep their intimate partners from seeing their friends (30.9% vs.16.1%, p < 0.0001). Overall, military personnel endorsed higher rates of perpetration of any form of controlling behavior against intimate partners in the year prior to the study (69.1% vs. 52.2%, p < 0.0001). (see Table 1).

Self-Reported Major Causes of Perpetrating the Last Episode of IPV

Of the respondents who reported perpetration of at least one form of IPV, more military personnel than the civil servants reported disobedience (22.6% vs. 17.6%), control issues (20.6% vs. 15.8%), difficulty at work (12.1% vs. 9.6%) and drunkenness (6.2% vs. 4.9%) as the major causes of perpetrating the last episode of IPV. However, more respondents in civil service than in the military reported financial problems (19.6% vs. 13.0%), refusal of sex (13.6% vs. 8.5%), non-availability of food at home (9.8% vs. 6.4%) and infidelity (15.8% vs. 11.9%) as the major causes of perpetrating the last episode of IPV.

Predictors of Perpetration of Different Types and any Form of IPV

After adjusting for the other independent variables among the military personnel and civil servants; previous history of a physical fight with another woman and having had childhood exposure to inter-parental IPV were the significant predictors with the highest odds of perpetrating physical violence, sexual violence, psychological abuse and controlling behaviour. In both groups, respondents with previous histories of physical fights with other woman and childhood exposures to inter-parental IPV were more likely to perpetrate each type and any form of IPV. (see Tables 2, 3, 4, 5 and 6).

Table 2.

Comparison of chi-square and logistic regression analysis of predictors for perpetrating physical violence

	Military personnel n = 631				Civil servants n = 609
	Chi-square analysis (χ2)		Logistic regression		Chi-square Analysis (χ2)		Logistic regression
Predictors	Yes	χ2	OR	95% CI	Yes	χ2	OR	95% CI
Age
≥ 37 years	46(20.4)	0.84	1.204	0.676–2.144	38(11.7)	0.003	1.384	0.633–3.023
< 37 years (Ref)	71(17.5)		1		33(11.6)		1
Marital status
Not married	26(17.2)	0.23	0.711	0.396–1.277	14(9.0)	1.39	0.517	0.245–1.094
Married (Ref)	91(19.0)		1		57(12.6)		1
Level of education
Tertiary	38(15.7)	2.10	0.456**	0.253–0.821	56(11.5)	0.04	0.945	0.477–1.875
Secondary (Ref)	79(20.3)		1		15(12.2)		1
Grade level/ Rank
Senior Officers	25(21.2)	0.67	1.387	0.693–2.775	30(9.7)	2.32	0.531*	0.284–0.992
Junior Officers (Ref)	92(17.9)		1		41(13.7)		1
Childhood exposure to inter-parental IPV
Yes	36(30.0)	12.88***	2.049**	1.199–3.501	19(29.2)	21.82***	2.460**	1.232–4.911
No (Ref)	81(15.9)		1		52(9.6)		1
Ever had a physical fight with another woman
Yes	52(53.6)	93.31***	7.914***	4.809–13.024	25(43.9)	63.31***	6.801***	3.529–13.105
No (Ref)	65(12.2)		1		46(63.9)		1
Current use of alcohol
Yes	61(25.0)	10.99***	1.456	0.915–2.317	30(18.4)	9.84**	1.788*	1.014–3.154
No (Ref)	56(14.5)		1		41(9.2)		1
Length of years in relationship
≥ 11 years	30(19.5)	0.12	1.053	0.563–1.969	26(10.5)	0.52	0.514	0.240–1.102
< 11 years (Ref)	87(18.2)		1		45(12.4)		1

Ref - Reference value; OR - Odds ratio; CI - Confidence interval

^*- p value < 0.05

^**- p value < 0.01

^***- p value < 0.001

Table 3.

Comparison of chi-square and logistic regression analysis of predictors for perpetrating sexual violence

	Military personnel n = 631				Civil servants n = 609
	Chi-square analysis (χ2)		Logistic regression		Chi-square Analysis (χ2)		Logistic regression
Predictors	Yes	χ2	OR	95% CI	Yes	χ2	OR	95% CI
Age
≥ 37 years	50(22.2)	0.07	1.018	0.618 – 1.677	58(17.9)	10.12***	0.913	0.505–1.649
< 37 years (Ref)	94(23.2)		1		82(28.8)		1
Marital status
Not married	28(18.5)	2.06	0.533*	0.317–0.898	45(29.0)	4.29*	0.917	0.548–1.536
Married (Ref)	116(24.2)		1		95(20.9)		1
Level of education
Tertiary	67(27.7)	5.28*	1.059	0.655–1.712	118(24.3)	2.27	1.364	0.795–2.341
Secondary (Ref)	77(19.8)		1		22(17.9)		1
Grade level/ Rank
Senior Officers	43(36.4)	15.29***	1.854*	1.062–3.238	55(17.8)	9.54**	0.631	0.394–1.011
Junior Officers (Ref)	101(19.7)		1		85(28.3)		1
Childhood exposure to inter-parental IPV
Yes	42(35.0)	12.48***	1.736*	1.077 – 2.796	26(40.0)	11.90***	1.968*	1.100–3.523
No (Ref)	102(20.0)		1		114(21.0)		1
Ever had a physical fight with another woman
Yes	41(42.3)	24.61***	2.718***	1.671–4.419	27(47.4)	21.11***	3.214***	1.756–5.882
No (Ref)	103(19.3)		1		113(20.5)		1
Current use of alcohol
Yes	75(30.7)	14.16***	1.530*	1.016–2.303	49(30.1)	6.29*	1.426	0.923–2.203
No (Ref)	69(17.8)		1		91(20.4)		1
Length of years in relationship
≥ 11 years	26(16.9)	4.08*	0.499*	0.279–0.893	39(15.8)	12.16***	0.545*	0.302–0.983
< 11 years (Ref)	118(24.7)		1		101(27.9)		1

Ref - Reference value; OR - Odds ratio; CI - Confidence interval

^*- p value < 0.05

^**- p value < 0.01

^***- p value < 0.001

Table 4.

Comparison of chi-square and logistic regression analysis of predictors for perpetrating psychological abuse

	Military personnel n = 631				Civil servants n = 609
	Chi-square analysis (χ2)		Logistic regression		Chi-square Analysis (χ2)		Logistic regression
Predictors	Yes	χ2	OR	95% CI	Yes	χ2	OR	95% CI
Age
≥ 37 years	95(42.2)	0.76	1.019	0.662 – 1.569	105(32.4)	0.47	1.708	0.985–2.963
< 37 years (Ref)	157(38.7)		1		85(29.8)		1
Marital status
Not married	52(34.4)	2.50	0.627*	0.405–0.973	50(32.3)	0.11	1.074	0.648–1.780
Married (Ref)	200(41.7)		1		140(30.8)		1
Level of education
Tertiary	112(46.3)	6.59**	1.365	0.919 – 2.027	160(32.9)	3.33	1.697*	1.035–2.780
Secondary (Ref)	140(36.0)		1		30(24.4)		1
Grade level/ Rank
Senior Officers	57(48.3)	4.24*	0.964	0.586–1.587	92(29.8)	0.59	0.707	0.457–1.092
Junior Officers (Ref)	195(38.0)		1		98(32.7)		1
Childhood exposure to inter-parental IPV
Yes	66(55.0)	14.02***	1.747**	1.135 – 2.689	37(56.9)	22.43***	2.757***	1.583 – 4.802
No (Ref)	186(36.4)		1		153(28.1)		1
Ever had a physical fight with another woman
Yes	60(61.9)	22.96***	2.615***	1.642 – 4.165	35(61.4)	26.73***	3.302***	1.821 – 5.988
No (Ref)	192(36.0)		1		155(28.1)		1
Current use of alcohol
Yes	122(50.0)	16.80***	1.660**	1.168–2.358	61(37.4)	4.02*	1.195	1.797–1.790
No (Ref)	130(33.6)		1		129(28.9)		1
Length of years in relationship
≥ 11 years	62(40.3)	0.01	0.894	0.558–1.433	75(30.4)	0.14	0.757	0.455–1.262
< 11 years (Ref)	190(39.8)		1		115(31.8)		1

Ref - Reference value; OR - Odds ratio; CI - Confidence interval

^*- p value < 0.05

^**- p value < 0.01

^***- p value < 0.001

Table 5.

Comparison of chi-square and logistic regression analysis of predictors for perpetrating controlling behavior

	Military personnel n = 631				Civil servants n = 609
	Chi-square analysis (χ2)		Logistic regression		Chi-square Analysis (χ2)		Logistic regression
Predictors	Yes	χ2	OR	95% CI	Yes	χ2	OR	95% CI
Age
≥ 37 years	144(64.0)	4.25*	0.818	0.517–1.292	159(49.1)	2.74	1.545	0.930–2.565
< 37 years (Ref)	292(71.9)		1		159(55.8)		1
Marital status
Not married	102(67.5)	0.22	0.630*	0.402 – 0.988	103(66.5)	16.89***	2.061**	1.285–3.305
Married (Ref)	334(69.6)		1		215(47.4)		1
Level of education
Tertiary	169(69.8)	0.10	0.771	0.513–1.159	261(53.7)	2.13	1.342	0.872–2.054
Secondary (Ref)	267(68.6)		1		57(46.3)		1
Grade level/ Rank
Senior Officers	93(78.8)	6.42**	2.040*	1.168–3.563	143(46.3)	8.87**	0.663*	0.448–0.982
Junior Officers (Ref)	343(66.9)		1		175(58.3)		1
Childhood exposure to inter-parental IPV
Yes	96(80.0)	8.25**	1.977**	1.188–3.290	45(69.2)	8.44**	1.851*	1.035–3.308
No (Ref)	340(66.5)		1		273(50.2)		1
Ever had a physical fight with another woman
Yes	75(77.3)	3.63	1.455	0.858 – 2.465	40(70.2)	8.13**	2.076*	1.115 – 3.868
No (Ref)	361(67.6)		1		278(50.4)		1
Current use of alcohol
Yes	178(73.0)	2.77	1.128	0.775–1.644	100(61.3)	7.44**	1.447	0.986–2.124
No (Ref)	258(66.7)		1		218(48.9)		1
Length of years in relationship
≥ 11 years	89(57.8)	12.19***	0.498**	0.308–0.807	115(46.6)	5.33*	0.781	0.487–1.253
< 11 years (Ref)	347(72.7)		1		203(56.1)		1

Ref - Reference value; OR - Odds ratio; CI - Confidence interval

^*- p value < 0.05

^**- p value < 0.01

^***- p value < 0.001

Table 6.

Comparison of chi-square and logistic regression analysis of predictors for perpetrating any form of IPV

	Military personnel n = 631				Civil servants n = 609
	Chi-square analysis (χ2)		Logistic regression		Chi-square Analysis (χ2)		Logistic regression
Predictors	Yes	χ2	OR	95% CI	Yes	χ2	OR	95%CI
Age
≥ 37 years	169(75.1)	1.48	0.887	0.529 – 1.485	202(62.3)	4.14*	1.627	0.946 – 2.799
< 37 years (Ref)	322(79.2)		1		200(70.2)		1
Marital status
Not married	117(77.5)	0.12	0.711	0.428 – 1.181	120(77.4)	12.06***	1.563	0.927 – 2.635
Married (Ref)	374(77.9)		1		282(62.1)		1
Level of education
Tertiary	200(82.6)	5.31*	1.099	0.694 – 1.740	335(68.9)	9.15**	2.029**	1.306 – 3.152
Secondary (Ref)	291(74.8)		1		67(54.5)		1
Grade level/ Rank
Senior Officers	106(89.8)	12.14***	2.561**	1.264–5.190	182(58.9)	14.13***	0.485***	0.318–0.739
Junior Officers (Ref)	385(75.0)		1		220(73.3)		1
Childhood exposure to inter-parental IPV
Yes	109(90.8)	14.55***	2.905**	1.480–5.700	53(81.5)	0.01**	1.997*	1.005–3.967
No (Ref)	382(74.8)		1		349(64.2)		1
Ever had a physical fight with another woman
Yes	87(89.7)	9.36**	2.458**	1.217–4.965	49(86.0)	11.16***	3.037**	1.36–6.779
No (Ref)	404(75.5)		1		353(63.9)		1
Current use of alcohol
Yes	202(82.8)		1.236	0.806–1.897	125(76.7)		1.733*	1.127–2.667
No (Ref)	287(74.7)		1		227(62.1)		1
		5.70*				11.31***
Length of years in relationship
≥ 11 years	108(70.1)	6.97**	0.539*	0.315 – 0.921	147(59.5)	7.81**	0.683	0.413 – 1.130
< 11 years (Ref)	383(80.3)		1		255(70.4)		1

Ref - Reference value; OR - Odds ratio; CI - Confidence interval

^*- p value < 0.05

^**- p value < 0.01

^***- p value < 0.001

Among the military personnel, being a commissioned officer was a significant predictor with increased odds of perpetrating sexual violence, controlling behavior and any form of IPV. Use of alcohol was a significant predictor with increased odds of perpetrating sexual violence and psychological abuse. Those with tertiary level of education had reduced odds of perpetrating physical violence. However, while the unmarried military personnel had reduced odds of perpetrating sexual violence and controlling behavior, being in intimate relationship for more than 11 years showed significant reduced odds of perpetrating sexual violence, controlling behavior and any form of IPV. (see Tables 2, 3, 4, 5 and 6).

Among the men in civil service, use of alcohol was a significant predictor with increased odds of perpetrating physical violence and any form of IPV. Civil servants with tertiary level of education also had increased odds of perpetrating psychological abuse and any form of IPV while the unmarried men in civil service had increased odds of perpetrating controlling behavior. Being a senior officer showed significant reduced odds of perpetrating physical violence, controlling behavior and any form of IPV. Being in intimate relationship for more than 11 years showed significant reduced odds of perpetrating sexual violence. (see Tables 2, 3, 4, 5 and 6) Overall, male military personnel were more likely to perpetrate any form of IPV than the civil servants [OR: 1.806 (95% CI: 1.404–2.323)]. (see Table 7).

Table 7.

Predictor for perpetration of any form of IPV by work group

	Chi-square analysis (χ2)		Logistic regression
Work group	Yes	χ2	Odds ratio	95% Confidence interval
Military personnel	491 (77.8)	21.4 ***	1.806	1.404–2.323 ***
Civil servants (Ref)	402 (66.0)		1

^***- p value < 0.001

Health Complaints by the Partner as a Result of Physical IPV Perpetrated by the Respondents

The various health complaints the partners made as a result of physical IPV perpetrated against them were reported by the respondents. There was no significant difference in the health complaints by their partners as a result of physical IPV perpetrated by the military personnel and civil servants. (see Table 8).

Table 8.

Comparison of the health complaints by the partners as a result of physical IPV perpetrated by the military personnel and civil servants (As reported by the respondents)

Health complaints of the partner	Military personnel n = 117 n (%)	Civil servants n = 71 n (%)	Total n =188 n (%)	χ2 (p value)
As a result of the physical violence I perpetrated against my female intimate partner in the last one year, she compalined of:
Physical pain that still hurt the next day	17 (14.5)	14 (19.7)	31 (16.5)	0.87 (0.353)
Penetrating eye or internal injuries	15 (12.8)	4 (5.6)	19 (10.1)	0.14# (0.088)
Broken teeth or broken bones	13 (11.1)	5 (7.0)	18 (9.6)	0.85 (0.358)
Loss of consciousness from being hit on the head	12 (10.3)	3 (4.2)	15 (8.0)	0.17# (0.112)
Having to spend nights in the hospital	11 (9.4)	4 (5.6)	15 (8.0)	0.42# (0.263)
Sprain, bruise or small cut	10 (8.5)	7 (9.9)	17 (9.0)	0.09 (0.761)

Note: χ² – Chi square test

^#- Fishers exact test

Discussion

This comparative cross sectional study aimed to bridge the existing knowledge gap on the prevalence of different forms of IPV and the associated risk factors among a representative sample of men in the military and civil service in Ibadan, Nigeria. Self-reported perpetration of different forms of IPV was significantly higher among the military personnel compared to the civil servants in this study. This is consistent with findings by some other researches from developed countries where the prevalence of IPV was higher among the military population when compared with the civilian population (Cronin 1995; Heyman and Neidig 1999; Smith-Marek et al. 2016). This finding is also consistent with report of female partners who are victims of IPV in military and civilian population in Abuja, Nigeria (Chimah et al. 2015).

Men in the military perpetrated more acts of physical violence than men in civil service; this is consistent with the findings from studies that showed higher rate of physical partner abuse among the military compared to the civilian population (Chimah et al. 2015; Heyman and Neidig 1999; Rentz et al. 2006). A study showed that self- and partner-report of physically assaultive and injurious acts were perpetrated by respondents with military background (Hoyt et al. 2014). About 16% of Korean Air Force personnel perpetrated physical abuse against their intimate partners (Seo et al. 2014). This could be as result of training the military personnel are exposed to which entails giving and obeying orders. They may expect their partners to obey their instructions and violence could ensue if this is not so. Stressors associated with military lifestyle and military training may also be accountable for this difference (McCarroll et al. 2003; Newby et al. 2005; Rentz et al. 2006; Stamm 2009). In this study, significantly more respondents in the military compared to those in civil service threatened to use or actually used gun or other weapon against the partners. Access to weapon among the military personnel could account for this difference. Weapon use of any type by an intimate partner is associated with a wide range of violent offender behavior (Sorenson 2017). A systematic review on the use of firearms in IPV suggested that IPV increases in severity and deadliness when violent intimate partners have access to firearms (Zeoli et al. 2016). This is important when planning for interventions for the victims because a study reported that there may be reduced likelihood that IPV victimization in the military will be reported to the police compared to their civilian counterparts (Becker and Bachman 2020).

The prevalence of perpetration of sexual violence of 22.9% was similar in both groups of men but was higher when compared with similar studies in Ile-Ife (6.8%) and Ibadan (16%) (Fatusi and Alatise 2007; Fawole et al. 2010). Higher educational and economic status has been found to be associated with sexual violence in intimate relationships (Krug et al. 2002; WHO 2012). Majority of the respondents in this study were in the working class with high educational and economic status. This could have accounted for the higher prevalence of sexual violence among this study population. Self-reported prevalence of perpetration of sexual violence in this study was also higher when compared to the reports of female intimate partners in military and civilan population in Nigeria (Chimah et al. 2015). A significantly higher proportion of the civil servants made their partners have sexual intercourse without a condom when the partner wanted to use it compared to the military personnel. This may be as a result of increased awareness of the need for consistent condom use among military personnel as advocated through Armed Forces Programme on AIDS Control (AFPAC) because of higher prevalence of Human Immunodeficiency Virus (HIV) infection among the military than that of the general population (Nwokoji and Ajuwon 2004). Different studies showed that there is increased awareness on condom use among the military personnel (Essien et al. 2006, 2007; Holmes et al. 2008; Miles 2003) and this can lead to increased condom use among them. An interventional study on HIV and sexually transmitted diseases risk reduction in a cohort of Nigerian military personnel showed significant increase in reported condom use with causal partners and positive condom beliefs at six and twelve months follow up (Essien et al. 2006).

The prevalence of psychological or emotional abuse was significantly higher among the military personnel compared to the civil servants; studies among the military personnel reported similar findings (Heavey et al. 2017; Seo et al. 2014). This prevalence is lower when compared with prevalence of perpetration of psychological abuse (50.5%) among married men in Ile-Ife, Nigeria (Fatusi and Alatise 2007; Fawole et al. 2010). In addition, some African traditional gender norms encourage men to control their wives which include insulting them (Fawole et al. 2010). Research findings indicated that emotional violence is associated with low self-worth and low self-esteem (Nojomi et al. 2007). Psychological abuse can lead to negative mental health outcomes, and mental health problems can render a person more vulnerable to experiencing IPV (Trevillion et al. 2015).

Controlling behaviors in intimate relationships were perpetrated by about two thirds of the respondents. This prevalence was significantly higher among the military personnel than the civil servants. Studies among African families revealed that the control of women and their sexuality by men is similar to the control of property and might be accompanied with violence (Antai and Antai 2008; Balogun and John-Akinola 2015; Ilika 2005). The patriarchal nature of some African cultures, including cultural values in Nigeria, supports male superiority while women have limited decision-making power (Olayanju et al. 2013). Law enforcement officers including men in military and the police exercise more control in their intimate relationships because they may experience authoritarian spillover and sometimes transfer the use of force and control from the job to their homes (Adelman 2003; Anderson and Lo 2011; Sgambelluri 2000). Even though the difference is not statistically significant, a higher proportion of military personnel than civil servants expected their intimate partners to ask for permission before seeking health care for themselves. This behavior can have a negative effect on health seeking behavior and reproductive health outcome of women (Babu and Kusuma 2016; García-Morenoa and Amina 2016).

In this study, controlling behavior was the most prevalent IPV perpetrated by men in both groups, followed by psychological/emotional abuse, sexual violence and physical violence. This is at variance with the findings of a review of literature of studies comparing the military and civilian population in the United States of America which showed that physical violence accounted for over 90% of cases of partner violence in the military families, followed by emotional violence, neglect and sexual abuse (Rentz et al. 2006). More than one third of the respondents in this study perpetrated more than one type of IPV in the preceding year and this was higher among the military personnel than the civil servants. This is consistent with other studies where physical violence in intimate relationship is often accompanied by other forms of abuse such as psychological and sexual abuse (Antai 2011a; Nwabunike and Tenkorang 2017; Zamorski and Wiens-Kinkaid 2013). Controlling behavior by husband/partner significantly increases the likelihood of physical and sexual IPV, thus acting as a precursor to violence (Antai 2011a). Women with domineering husbands are significantly more likely to experience physical, sexual, and emotional violence (Nwabunike and Tenkorang 2017). In the WHO multi-country study on women’s health, men who were physically violent towards their wives also exhibited higher rates of controlling behaviour than those who were not (Garcia-Moreno et al. 2005).

Concerning the major causes of perpetrating the last episode of IPV by the respondents, more respondents in the military reported disobedience, control issues, difficulty at work and drunkenness as the major causes of perpetrating the last episode of IPV than those in the civil service. This is consistent with prior literature in which domineering and authoritarian spillover was seen among law enforcement officers and their intimate partners (Adelman 2003; Anderson and Lo 2011) and alcohol problem is related to spousal abuse among U.S. Army personnel (McCarroll et al. 2009). However, more respondents in the civil service reported financial problems, refusal of sex, non-availability of food at home and infidelity as the major causes of perpetrating the last episode of IPV than the respondents in the military. This is similar to the findings of studies conducted among women about the causes of IPV where disobedience, refusal to have sex when the husband wants it, infidelity and financial problems were indicated as causes of IPV against women (Antai and Antai 2008; Balogun and John-Akinola 2015). However, women also justified some of these reasons as being acceptable causes of IPV against women (Balogun and John-Akinola 2015; Fawole et al. 2005).

Despite the variations in the prevalence of perpetration of IPV among men in military and civil service in this study, similar factors appear to predict the perpetration of different types and any form of IPV in the two groups. Childhood exposure to inter-parental IPV, history of previous fight with another woman, alcohol consumption and grade level increase the risk of the different forms of IPV. These risk factors were reported by other studies (Dube et al. 2002; McCarroll et al. 2009; Smith-Marek et al. 2016). Interventions for preventing IPV in the two groups must address childhood exposure to inter-parental IPV as well as control of physical aggression by men. Alcohol consumption should also be controlled.

The health complaints from the partners as a result of physical IPV perpetrated by the respondents in this study is consistent with the findings of a previous study (Antai 2011b). Even though there were no statistically significant differences, more respondents in the military reported that their partners complained of body pain that still hurt the next day, penetrating eye and internal injuries, broken bones and teeth and loss of consciousness than the respondents in the civil service. IPV is related to increased health problems and health care utilization (Campbell et al. 2002). Similarly, more respondents in the military reported that their partners were admitted in the hospitals because of the injuries sustained from physical IPV perpetrated by them than the respondents in the civil service. However, these findings were not consistent with the study report where abused women reported more of gynecological health problems (Campbell et al. 2002). There may be underreporting about the gynecological problems because the perpetrators in this study reported the complaint on behalf of their partners. However research findings showed that between 19% and 55% of women who had ever been physically abused by their partners were injured (Ellsberg et al. 2008).

Strengths and Limitations

Certain strengths and limitations of this study should be recognized. The self-reported nature of this study could have made the respondents underestimate the extent to which violence was used in intimate relationships especially since partners were not interviewed. Although selection effect was minimized by ensuring that the groups were similar in terms of key characteristics, generalization of the findings to other groups or population must be done with caution. Similarly, the respondents represented different socio-economic classes with at least primary level of education but the findings may not be generalized to those at the extremes of socio-economic class with no formal education.

In this study, recall bias was minimized by assessing IPV one year prior to the study. Social desirability bias was also minimized by ensuring confidentiality. During the period of data collection, participants in need of help such as counseling were referred to appropriate institutions but the anonymous nature of the survey made it impossible to identify specific individuals who may require additional help and support after the study. Despite these limitations, this study provides useful information about IPV among men at the population level in these work groups which is useful in planning community health services and interventions.

Conclusion

The prevalence of perpetration of IPV is higher among the military personnel than the civil servants in Ibadan, Nigeria. Men in the military perpetrated all forms of violence against their intimate partners compared to men in the civil service. The findings from this study emphasize the important roles the public health as well as welfare workers should play in the prevention, identification and management of IPV and the consequences at the community level. Addressing this public health problem will involve different sectors such as the military and civil organisations, health, social services, counselors, religious organisations, media, judiciary and the police. There should be regular public health education, individual psycho-education, conflict management skills development and surveillance in the military and civil organisation for the prevention of IPV. This will require different strategies with immediate and long-term commitment. Individuals, families and the community also have to play crucial roles in ending partner violence in intimate relationships. There is a need to conduct research on the barriers to reporting IPV in the military population. Additional research on military-specific aspects of IPV such as the effect of military operations and posttraumatic stress disorders on family relationships, is also needed.