Abstract
Aims: To assess the prevalence and investigate ethnic differences of emotional, physical and sexual violence among a population of both Sami and non-Sami in Norway. Methods: Our study was based on the SAMINOR 2 study, a population-based survey on health and living conditions in multiethnic areas with both Sami and non-Sami populations in Central and Northern Norway. Our study includes a total of 11,296 participants: 2197 (19.4%) Sami respondents and 9099 (80.6 %) non-Sami respondents. Results: Almost half of the Sami female respondents and one-third of the non-Sami female respondents reported any violence (any lifetime experience of violence). Sami women were more likely to report emotional, physical and sexual violence than non-Sami women. More than one-third of the Sami men compared with less than a quarter of non-Sami men reported having experienced any violence in their life. Sami men were more likely to report emotional and physical violence than non-Sami men. However, ethnicity was not significantly different regarding sexual violence experienced among men. Violence was typically reported to have occurred in childhood. Sami participants were more likely to report having experienced violence in the past 12 months. For all types of violence, the perpetrator was typically known to the victim.
Background
Internationally, interpersonal violence (sexual, physical and emotional) is recognized as a public health issue which adversely affects mental and physical health [1]. Depression and post-traumatic stress disorder are considered the most prevalent mental health conditions associated with violence and abuse [2,3]. Moreover, a large range of somatic symptoms is often associated with violence [4]. International studies have indicated a higher prevalence of interpersonal violence in indigenous populations than in non-indigenous population [5–7]. Canadian studies have found indigenous people to be three times more likely to experience violent victimization [5]. In Greenland, a report on the living conditions of young people revealed that violence, including sexual abuse, was a major problem [8]. A comparative study of reported violence in Greenland and Denmark found the overall prevalence to be higher in Greenland [7]. Proportionally more women in Greenland reported sexual abuse than men. The differences was largest when occurring in childhood [9].
Former national studies on violence in Norway have not included information on Sami ethnicity [10,11]. Hence, little is known about the prevalence of interpersonal violence in Norway’s indigenous population. One paper that compared Greenlandic Inuit and Norwegian Sami adolescents reported no differences of experienced violence. In both populations, violence was reported by approximately one in four participants [12].
Historically, the Sami people has been a nomadic people, living in the Artic areas of Norway, Sweden, Finland and Russia’s Kola Peninsula [13]. In the 19th and 20th centuries, the Sami people were subjected to austere assimilation policies, with a resulting loss of language and identity [14].
Aims
This paper aims to explore the prevalence of various types of violence in a population with both Sami and non-Sami inhabitants in Central and Northern Norway, and to explore whether there are differences based on ethnicity and socioeconomic or demographic factors.
Methods
The SAMINOR study
This study is a sub-project of a larger questionnaire-based population study on health and living conditions in areas with both Sami and Norwegian populations, the SAMINOR 2 questionnaire study. The first population-based study on health and living conditions in areas with both Sami and Norwegian populations (SAMINOR 1) was conducted in 2003/2004. This study has been described in a previous paper [15]. The SAMINOR 2 questionnaire study was designed as a follow-up study of issues addressed in the original SAMINOR study. It was also expanded to include additional health issues such as violence and mental health. The SAMINOR 2 questionnaire study has been described in a recently published paper [16].
Sample
In selected municipalities with a mixed Sami and non-Sami population, all residents aged 18–69 years in the National Registry were sent an invitation to participate in the study. A questionnaire was mailed to 44,669 persons. A total of 1424 questionnaires were returned unopened and hence classified as technically missing, leaving 43,245 persons eligible for the study. Among these, 11,600 persons consented by returning the completed questionnaire, yielding a participation rate of 27%. Further methodological details are described elsewhere [16].
In the present study, 304 respondents were discarded due to missing information on ethnicity and violence, leaving 11,296 informants as the study group: 2197 (19.6%) Sami and 9099 (80.6 %) non-Sami individuals (see flow chart Figure 1).
Flow chart of inclusion in the study population of emotional, physical and sexual violence in the SAMINOR 2 study, 2011–2012.
Questionnaire
Questions regarding ethnicity
Ethnicity was assessed using responses to the following groups of questions: ‘What language does/did you use at home?’, ‘What language did your parents use at home?’, ‘What language did your grandparents use at home?’ and ‘What do you consider yourself to be?’ Response options were ‘Norwegian’, ‘Sami’, ‘Kven’, ‘Other’. Multiple answers were permitted. Participants choosing ‘Sami’ in response to any of the three first groups of questions, in addition to self-identification as Sami, were classified as belonging to the Sami ethnic group. Norwegians, Kvens (descendants of Finnish immigrants) [17] and Others were defined as non-Sami.
Questions on violence
Three sections in the questionnaire addressed experience of emotional, physical and/or sexual violence. Participants who responded positively to the question ‘Have you experienced that someone systematically and over time has tried to repress or humiliate you?’ were classified as ‘having experienced emotional violence’. Participants who responded positively to the question ‘Have you been exposed to physical assault/abuse?’ were classified as ‘having experienced physical violence’. Participants who responded positively to the question ‘Have you been exposed to sexual assault?’ were classified as ‘having experienced sexual violence’. Participants who responded positively to having experienced any type of violence (sexual, physical and emotional) were defined as ‘having experienced any violence’. Moreover, participants were asked to indicate whether the violence had occurred in childhood and/or in adulthood.
For each type of violence, participants were asked to indicate their relationship with the perpetrator, with the following alternatives: ‘Stranger’, ‘Spouse’, ‘Family/relatives’ or ‘Other known’. Multiple responses were permitted.
Background variables
Background information, such as education, income, religion, use of substances (alcohol and tobacco), was collected from the questionnaire. Information regarding age, gender and home municipality of participants were provided by Statistics Norway. Participants were placed in one of the three age groups: 18–34 years, 35–49 years and 50–69 years. A participant’s level of education was measured according to the number of years of education: primary school (0–9 years), secondary school (10–12 years), higher education of 3 years’ duration or less (13–15 years) and higher education of more than 3 years’ duration (16 years and above). Household annual income was categorized into three groups: low (<150,000–300,000 NOK), medium (301,000–600,000 NOK) and high (601,000 to >900,000 NOK).
The 25 municipalities included in the SAMINOR 2 study have been described elsewhere [16]. These were recoded into ‘Sami minority area’ and ‘Sami majority area’. The reasons behind such descriptions are described in more detail in another paper [18]. The question regarding smoking reads as follows: ‘Do you smoke, or have you previously smoked? The available responses categories were: ‘Yes, daily’, ‘Yes, previously’, ‘Yes, sometimes’ and ‘No, never’. The categories were narrowed down to three: No, never (‘No never’), Yes, daily (‘Yes, daily’) and Yes, previously (‘Yes, previously’ and ‘Yes, sometimes’). Respondents were asked to indicate how often they had consumed alcohol in the past year: ‘Never consumed alcohol’, ‘Have not been drinking alcohol during the last year’, ‘A few times during the last year’, ‘About once a month’, ‘Two or three times per month’, ‘About once a week’, ‘Two or three times a week’ and ‘Four to seven times a week’. The three categories that were created were ‘Never/rarely’ (‘Never consumed alcohol’, ‘Not during the last year’ and ‘A couple of times in the past year’), ‘Monthly’ (‘About once a month’ and ‘two or three times a month’), ‘Weekly’ (‘About once a week’, and ‘Four to seven times a week’).
Respondents were asked to indicate their affiliation with the Laestadian Church (a Lutheran denomination particularly widespread in the Northern regions of Norway, Sweden and Finland): ‘Are your grandparents affiliated with the Laestadian church?’, ‘Is your father affiliated with the Laestadian church?’, ‘Is your mother affiliated with the Laestadian church?’ and ‘Are you affiliated with the Laestadian church? Multiple responses were permitted. Participants responding positively to one or more of these options were classified as ‘Laestadianist’. Respondents with no positive response on Laestadian church were classified as ‘non-Laestadianist’.
Ethics
Written informed consent was obtained from all participants. The study was approved by the Regional Committee for Medical and Health Research Ethics of Northern Norway and the Norwegian Data Protection Authority.
Statistical analysis
Descriptive statistics are presented according to ethnicity and gender. The chi-square test was used in the comparison of groups. Level of significance was set to 5%. Logistic regression analyses were performed to adjust for age, education, living region, Laestadianism and alcohol intake. The results of the logistic models are presented with confidence intervals of 95% (95% CI). The software package for statistical analysis program, the SPSS version 22 was used for the analyses. We also investigated potential interaction between ethnicity and living area.
Results
Prevalence of different types of violence
Population characteristics across groups are presented in Table I. Almost half of the Sami population, 45% (n=989) reported having been subjected to any violence (Table II). For the non-Sami population, the figure was 32.6% (n=3682) (Table II).
Population characteristics in the Sami and non-Sami population of emotional, physical and sexual violence by gender in the SAMINOR 2 study, 2011–2012.
Prevalence of emotional, physical, sexual and any violence as a child only, adult only, both as a child and as an adult and past 12 months among the Sami and non-Sami by gender in the SAMINOR 2 study, 2011–2012.
Among both women and men, ethnicity was associated with all types of violence, except for sexual violence among men. Women were more likely to report being subjected to sexual violence than men, irrespective of ethnicity. Emotional violence was the most common type of violence reported to have occurred both in child- and adulthood, independent of ethnicity and gender.
There was a significant age variation for any violence (Tables III and IV): violence was less reported by respondents in age group 50–69. When assessing the three types of violence separately, age variation remained the same for non-Sami women. For Sami women the pattern remained the same, but was only significant for age group 35–49. Among men, the age variation for non-Sami was only significant for emotional violence. For Sami men the pattern was different: emotional, physical and sexual violence increased by age. However, it was only significant for physical violence between age group 35–49 and 50–69 and sexual violence between age group 18–34 and 50–69.
Crude and adjusted odds ratio (OR) for any lifetime violence among women. The results are adjusted for ethnicity, age, education, living area, affiliation to Leastadianism and alcohol intake, the SAMINOR 2 study 2011–2012.
Crude and adjusted odds ratio (OR) for any lifetime violence among men. The results are adjusted for ethnicity, age, education, living area, affiliation to Leastadianism and alcohol intake, the SAMINOR 2 study 2011–2012.
To have experienced any type of violence was positively associated with living in Sami majority areas, also when adjusted to ethnicity (Tables III and IV). When assessing the three types of violence separately, the pattern remained the same. However, it was only significant for physical violence among women.
Participants who reported experience of violence in childhood: 724 (33.7%) had been exposed to two or more types of violence (data not shown). Among men, this was found to be associated with ethnicity: highest among non-Sami, 32.7%, compared with 28.8% among Sami. No effect on ethnicity was observed among women. Of those who reported any violence in adulthood, 27.4% had been exposed to two or more types of violence (data not shown). This proportion did not differ between Sami women (35.0%) and non-Sami women (34.2%). However, a larger proportion of Sami men reported being exposed to two or more types of violence compared to non-Sami men (14.7% vs. 7.9%, p⩽.001).
Overall 2.9% of the study population reported that they had been exposed to some type of violence in the past 12 months (Table II).
Sami respondents were nearly twice as likely to report being subjected to violence in the past 12 months (Sami respondents: 4.1%; non-Sami respondents: 2.6%, Table II). Estimates of associations between any violence, ethnicity and participants’ background and selected lifestyle characteristics are shown for in Table III (women) and Table IV (men).
Ethnicity was strongly associated with being subjected to any violence both among women and men, and remained after adjusting for age, education, living area, religion and alcohol intake (Tables III and IV).
For both genders, the association between ethnicity and any violence was stronger in minority areas (data nor shown).
Perpetrators
Among those reporting any violence, most reported it to be a known perpetrator, and one in five reported the perpetrator to be a stranger.
Discussion
Sami respondents of both genders were more likely to report having been subjected to any violence. Sami women reported highest prevalence of violence exposure. Across ethnic groups and genders, emotional violence was the most frequently reported type of violence. Women were more likely to report sexual violence than men, and the proportion of Sami women who reported sexual violence was higher than non-Sami women. The impacts of ethnicity and gender were distinct, particularly in the case of sexual violence.
Comparison of results to other indigenous studies
Our findings are consistent with other studies which show that indigenous people are more likely to be exposed to violence than non-indigenous people [5–7]. Findings for Sami women in our study (49.1%) are congruent with a study of the Inuit population in Greenland [9] that reported that 47% of Inuit women were exposed to violence. However, the reported prevalence for Inuit men (48%) was higher than for Sami men in our study (39.7%).
In the study by Curtis et al. [9], sexual violence was reported by one in four Inuit women (25%) and 6% of Inuit men. In our study, one in five Sami women reported sexual violence (21.8%). The corresponding figure for Sami men was 5% in our study. This might suggest that the prevalence of sexual violence in the Inuit and Sami people is rather similar. Further, Curtis et al. reported that 8% of Inuit women and 3% of Inuit men had been subjected to childhood sexual violence. In our study, sexual violence in childhood was reported by 16.7% for Sami women and 4.9% of Sami men. Discrepancies may be explained by differences in phrasing the questions: in the Curtis study the question regarding sexual assault was phrased ‘have you ever been forced to sex’, while in our study the question regarding sexual violence was phrased more generally: ‘Have you been exposed to sexual assault’. The age cut-off was also lower in the study by Curtis et al.: less than 13 years; the cut-off in our study was 18 years. Moreover, regarding the potential impact of the period under study, Curtis et al. conducted their study in Greenland in 1993–1994. An increased openness in society in general and the establishment of various health facilities addressing sexual violence may also have resulted in a higher prevalence of reported sexual violence in childhood in our study.
Comparison of results to other Nordic studies
Emotional violence
Our results for emotional violence in childhood in the total sample (14.1% for women and 13.8% for men) is similar to a recent Norwegian National study which reported a prevalence of exposure to violence in childhood of 15.4% for women and 11.2% for men [10]. Moreover, the prevalence of emotional violence for women in our study (28.2%) is within the estimate given in an earlier, Nordic cross-sectional study of women seeking reproductive health services that reported a prevalence of 19–37% for emotional violence [19].
Physical violence
As for any experience of physical violence we found a prevalence of 19% for women and 11.3% for men. The prevalence for women was slightly lower than the prevalence presented in a national study among pregnant women in Norway, which presented a prevalence of 21.5%. As to physical violence in childhood by family/relatives, we found a prevalence of 6% for women and 4.3% for men (5.4% total). Other Norwegian studies have reported similar figures. In a population-based national study of students graduating from upper secondary school, Mossinge and Stefansen [20] found that 8% had been victim of severe physical violence from their parents. A Norwegian National study published in 2014 [10] found that 5.0% of men and 4.9% of women had experienced severe physical violence from parents/caretaker (5% total). The figures from the latter study are in accordance with our estimate and might suggest that the question regarding physical violence in our study may have been interpreted as severe.
Sexual violence
The prevalence estimate found in our study for sexual violence was 17.3% for women and 4.4% for men. The prevalence for women corresponds well with a national study among pregnant women in Norway that presented a prevalence of 17%. In a study, Wijma et al. [19] presented a prevalence of 22.9% for sexual violence for women in Norway. This finding among abused women in shelter in Norway corresponds well with our estimate for Sami women (21.8%). For sexual violence in childhood (before age 18), our estimates for women are 12.5% and 3.9% for men. Our findings are in line with a recent Norwegian national study which found a prevalence of sexual violence of 12.3% for women and 5% for men before the age of 16 [21].
Strengths and limitations
The large number of participants (n=11,296) is a strength of our study. Data were collected in multiethnic municipalities making it possible to assess differences based on ethnicity. The main aim of the study was to assess the differences in prevalence of violence. We used a modified version of the NorVold Abuse Questionnaire (NorAQ). The validation study conducted in Sweden has showed that the abuse variables in the NorAQ have shown good reliability and validity [22]. Our modified version has previously been used in The Survey on Health and Living conditions in Oslo in 2000–2001 (the HUBRO study) [23]. The abuse questions in NorAQ have not been validated, either in the Sami population or among non-Sami in Norway. Differences in cultural and lingual interpretations may have influenced the observed differences among the two groups. It is a strength of the study that the questions give information of various types of violence and information about when it had occurred. Hence, this gives us a broader picture of the exposure of violence in this population. When classifying ethnicity, linguistic affiliation and self-identity were used as criteria. Both criteria are used by the Sami Parliament for register of voters. Hence, misclassification of responders as to ethnicity is unlikely.
The participation rate was 27%. This low rate must be regarded as a limitation of the study. We have limited information about the non-respondents. Since ethnicity is not recorded in any official register in Norway, we are not able to assess whether the non-respondents in the two ethnic populations differed. However, a comparison between participations in SAMINOR 2 and those participating in the first SAMINOR has been conducted [16]. The proportion of participants classified as Sami did not differ between SAMINOR 1 and SAMINOR 2. Since the participation rate in SAMINOR 1 was considerable higher, 60.9%, this population may have been representative for the background population. Further, compared with participants in SAMINOR 1, participants in SAMINOR 2 tended to have higher education. Our results show that the proportion reported violence was slightly higher in the group with high education. This may have influenced our overall estimate by making our estimate slightly higher than would be the case if there were no differences in education between responders and non-responders.
The information letter in SAMINOR 2 did not specifically address violence, which makes it unlikely that participation was influenced by selection based on status as to having experienced violence in general. Also, our observed ethnic difference in reporting of violence is unlikely to be influenced by differential misclassification, since it is unlikely that non-Sami participants were less likely to report violence than Sami. Compared with the total invited population, participants in our study were older. Since the proportion reported violence in our sample was higher among the youngest, our estimates on violence are probably conservative. On the other hand, there is a growing openness in reporting violence in society: this may overestimate both the prevalence and the differences in reporting violence between age groups. The low response rates require that generalizing of the result must be considered with care. However, the differences between respondents and non-respondents are often important but rarely enough to undermine studies. We believe that our results can be generalized to the Sami and non-Sami living in rural areas in Central and Northern Norway.
We had information on several assumed confounding factors and hence were able to control for these, but unknown confounders may exist. Recall bias is also a possible challenge in retrospective reporting such as this, but difficult to avoid in these kinds of surveys [24]. It is generally believed that experiences of violence in childhood tend to be under-reported in adulthood [25]. This may lead to a lower prevalence estimate. On the other hand, age may have a positive impact on reporting exposition to violence, as negative exposures denied earlier in life could come to awareness later on.
Conclusion
For all types of violence, Sami respondents more frequently reported having been subjected to violence, and women in both ethnic groups reported more frequently having been subjected to sexual violence. Our findings have implications for both health and policy in terms of improving violence prevention strategies. Health professionals should be aware of ethnic differences in violence exposure.
Footnotes
Acknowledgements
We are grateful to all the participants who took part in this study. We would like to thank Marita Melhus, Department of Community Medicine, The Artic University of Norway, and Hein Stigum, Department of Community Medicine, University of Oslo for statistical support.
Conflict of interest
None declared.
Funding
Funding was provided by the Nothern Norway Regional Health Authority (Helse Nord RHF) and Sami Norwegian Advisory for Mental Health and Substance use (SANKS).