Adolescent victimization and psychosocial well-being in a Finnish population-based sample

ABSTRACT

Potentially traumatizing events are common in adolescence, and these experiences have an impact on psychopathology. Furthermore, these events cumulate for certain individuals. More knowledge about the relationship between well-being and victimization experiences’ amount, type, and timing in adolescents is needed to understand the phenomenon of adolescent victimization. In this cross-sectional study, we investigated in a population-based sample of Finnish adolescents, whether victimization experiences associate with psychosocial well-being. The study sample comprised 5743 adolescents aged 15–16 years who filled in a questionnaire about 12 victimization experiences and psychosocial well-being. General Linear Model regression analyses were used for statistical analysis. A greater number of victimization experiences and recent timing, as well as type of victimization such as cyberbullying, were linked to problems related to psychosocial well-being. It is important to recognize current forms of youth victimization to identify the adolescents at risk for psychosocial distress and investigate on specific types of victimization.

KEYWORDS:

• Polyvictimization
• trauma
• adolescents
• well-being
• population-based sample
• cyberbullying

Introduction

The majority of children and adolescents experience one or more potentially traumatic events during their childhood and adolescence (Finkelhor et al., 2009; McLaughlin et al., 2013; Turner et al., 2010). Studies in the United States investigating childhood and adolescent victimization have reported high numbers of victimization experiences and raised awareness for polyvictimization, i.e. the experience of multiple potentially traumatic events over the life span (Finkelhor et al., 2005; Turner et al., 2010). In Europe, in a large Swiss sample of adolescents, 73.2% of participants reported being victimized in the past year, with the mean number of victimization experiences being 2.78 with a maximum of 31 experiences (Lätsch et al., 2017). Further, in a large Swedish sample of adolescents, lifetime victimization was also common, as 84.1% had experienced at least one victimization experience; the mean number of victimization experiences was 4.1 out of 33 experiences (Aho et al., 2016). In line with these results, two-thirds of Finnish 6^th and 9^th graders in a large national sample reported having at least one victimization experience during the past year (Ellonen & Salmi, 2011). In this study, the terms victimization experiences and potentially traumatic events (PTEs) are used interchangeably, as these refer to the same group of experiences such as physical and sexual abuse or violence, war trauma, or natural disasters. However, individual responses (e.g. psychological distress) to these events vary markedly and not every traumatic event exposure develops into a traumatic stress reaction.

Theories of victimization and polyvictimization in children and adolescents

Attempts to understand the impact of PTEs experienced in childhood and adolescence on well-being seem to follow two main approaches. First, the interest narrowly focuses on single events, such as child sexual abuse, physical abuse, or corporal punishment and neglect, and their consequences on mental health. This approach confines the scope and leaves other events less investigated, ignoring the burden of experiencing several individual PTEs. It has been suggested that focusing on only specific PTEs leads to neglect of one of the potentially most vulnerable groups of children and adolescents. Second, there is growing evidence for the frequency and detrimental effects of multiple traumatizations (Haahr-Pedersen et al., 2020; Lee et al., 2022). Besides studying these traditional forms of child and adolescent victimization, it is important to acknowledge the current forms of victimization such as cyberbullying. Use of social media as well as cyberbullying are common among adolescents, and cyberbullying has been shown to have a negative impact on adolescent mental health (Fahy et al., 2016). Furthermore, it is important to gain understanding on how the timing of the experiences (lifetime or recent victimization) relates to psychological well-being to be able to identify the adolescents at risk on time. Studying timing of PTEs increases knowledge about sensitive periods of development and periods in which exposures to PTEs may be particularly detrimental in the long term (Dunn et al., 2017). This knowledge is needed in finding ways to prevent PTE exposure and developing appropriate interventions for PTE victims at different developmental stages. Therefore, the knowledge of the impact of timing of PTE is relevant for both clinicians and policy makers.

Unfortunately, PTEs tend to occur in adolescents who have already experienced PTEs before (Boney McCoy & Finkelhor, 1995). This is often referred to as polyvictimization or multiple traumatization, albeit according to a recent scoping review the definition of polyvictims has been inconsistent (Lee et al., 2022). According to one study, 10% of children and adolescents aged 2–17 years had experienced 11 or more different forms of victimization during their lives (Turner et al., 2010). The same group of researchers defined polyvictims as the most extreme 10% in their sample, with 9–15 PTEs depending on the age group in children and adolescents aged 3–18 years (Finkelhor et al., 2009). In another study (Turner et al., 2010), polyvictims seemed to often experience the most serious types of victimization.

Victimization and mental health outcomes

Experiencing PTEs has been linked to many psychiatric and psychological problems in the short- and long-term, including mental health disorders, cognitive problems, substance abuse, psychopathic traits, physiological symptoms, somatic disorders, and even shorter life expectancy (Fontaine et al., 2018; Hughes et al., 2017; Saukkonen et al., 2016; Schonkoff & Garner, 2012; Waehrer et al., 2020). Some of the PTEs have been shown to have a stronger predictability for the onset of psychiatric disorders than others, sexual abuse and rape having the highest predictability for posttraumatic stress disorder (PTSD) (McLaughlin et al., 2012, 2013). In a recent meta-analysis, the prevalence of PTSD ranged between 0.5% and 67.3% in children and adolescents exposed to a PTE, with an overall rate of 16% (Alisic et al., 2014).

In adolescents, childhood PTEs have been associated with many psychiatric conditions besides PTSD, most consistently with behaviour and substance abuse disorders (McLaughlin et al., 2012). Evidence suggests that adolescent boys and girls experience the same rates of PTEs, but girls experience more anxiety sensitivity, trait anxiety, and depression than boys (Martin et al., 2014). In their study, Martin et al. (2014) proposed that these differences might be due to the gender role socialization and orientation and genetic factors affecting anxiety sensitivity in girls. In another study with Finnish and Danish participants (Peltonen et al., 2010), boys generally reported more internalizing and externalizing symptoms than girls, but higher parental violence was linked to internalizing symptoms in girls. In summary, there appear to be gender differences in responses to PTEs, warranting further investigation. Repeated exposure to multiple PTEs is linked to more severe PTSD symptoms and depression in adolescents (Fairbank & Fairbank, 2009; Suliman et al., 2009). Polyvictimization has been found to be associated with heightened trauma symptoms in children and adolescents (Finkelhor et al., 2005, 2007b). Accordingly, recent reviews have shown it to be linked to adverse mental health outcomes (Haahr-Pedersen et al., 2020; Lee et al., 2022).

Crime victimization experiences are potentially traumatic events and there is extensive evidence of their association with weaker psychosocial well-being (Mahuteau & Zhu, 2016; Tan & Haining, 2016). Negative associations with well-being have also been reported in crime occurring online (Kaakinen et al., 2018). Besides the risk for psychiatric symptoms, prior victimization experiences have been linked to heightened risk for future sexual abuse victimization in children and adolescents aged 10–16 years (Boney McCoy & Finkelhor, 1995).

Importance of assessing amount, type and timing of victimization

In a seminal study, Finkelhor et al. (2009) stressed the importance of identifying multiple victimization experiences (polyvictims) to be able to target help to the group with the most psychological torment and vulnerability to future victimization. When polyvictimization is not taken into account, it may lead to an underestimation of the patient’s full agony while also overestimating the effect of individual PTEs. Furthermore, some effects of individual victimization experiences may no longer be significant when polyvictimization is considered (Finkelhor et al., 2007a; Lätsch et al., 2017; Turner et al., 2010). In a Swedish study, the number of victimizations was related to psychological consequences, most strongly to posttraumatic stress and dissociation, and affected psychological well-being more in females than males (Aho et al., 2016). Furthermore, Turner et al. (2010) found that experiencing several different types of victimization experiences is more strongly associated with trauma symptoms than repeated victimization of a single victimization type or even chronic exposure to a single victimization type, even if it is of a serious nature. In addition to the amount and type of victimization experiences, the onset of the victimization seems to matter; physical abuse during middle childhood and sexual abuse during middle childhood or adolescence were associated with psychological distress such as depression, anxiety, and PTSD (Adams et al., 2018).

The previous studies investigating adolescent victimization in the Finnish setting have mostly utilized the data from the Child Victim Survey carried out four times since the 1980’s (e.g. Ellonen & Salmi, 2011; Peltonen et al., 2010). The study has covered the effects of physical and sexual abuse, parental violence, and bullying in over 13,000 6^th and 9^th graders. In these studies, polyvictimization was linked to increased psychosocial problems measured with the Strengths and Difficulties Questionnaire (SDQ). Electronical bullying was assessed but was not the focus of the study. It is important to collect new information on victimization experiences regularly to follow up the changes in the frequencies and types of these experiences, for example cyberbullying has emerged as a major victimization experience among adolescents in recent years (Chudal et al., 2022).

Aims and hypotheses

The present study aimed to investigate and report the amount, type and timing of victimization experiences and how these associate with the psychological wellbeing in a population based sample of 15–16-year-old Finnish adolescents. Information on specific forms of modern interpersonal victimization such as cyberbullying and hate violence were included. Based on earlier research, our hypotheses were as follows:

H1:

A higher amount of victimization experiences over the life span as well as during the last 12 months would be linked to lower psychosocial well-being and heightened internalizing and externalizing symptoms in adolescents.

H2:

Some types of interpersonal victimization (parental corporal punishment, physical and sexual abuse) would be most strongly linked to lower psychosocial well-being in both time frames (over the life span and during the last 12 months).

Methods

Participants and procedure

Data collection in the Finnish Self-Report Delinquency Study. The current study utilizes data from the Finnish Self-Report Delinquency Study (FSRD) that was collected in 2020. The FSRD surveys are collected every four years by the Institute of Criminology and Legal Policy at the University of Helsinki, and they cover a wide range of adolescents’ criminal behaviour, victimization experiences, and other related phenomena (Kaakinen et al., 2022). The survey data are based on a random sample of schools, and they represent Finnish 9^th grade students (aged 15–16 years).

The computer-based FSRD survey was filled in a class during a regular school day and supervised by an educated teacher. The study protocol complied with the informed consent procedure and the other principles of the Declaration of Helsinki. The study was pre-reviewed by the Research Ethics Committee in the Humanities and Social and Behavioral Sciences of University of Helsinki and deemed compliant with the research ethics guidelines (statement 33/2019). In 2020, altogether 5674 adolescents from 74 schools took part in the survey; the overall response rate was 78%. The schools were selected using stratified sampling regarding the school size (probability proportional to size sampling) and location (city or rural) to ensure the representativeness of the study sample. From those schools, all 9^th grade students who had given their informed consent took part in the study: no further inclusion or exclusion criteria applied. In the current study, we used the data from 5607 participants who had answered the questions of interest in this study. Due to major differences in gender group sizes, male and female participants were included in further analysis.

Measures

Victimization. Victimization as an independent variable in the FSRD survey was measured by asking whether the respondent had experienced any of 12 types of victimization, including property crimes, violent crimes, and sexual violence. If yes, they were asked to recall whether they had experienced these events in the past 12 months and how often. Sum variables of 0–12 victimization experiences were calculated for each participant. The statistical analyses were conducted using both lifetime victimization and recent victimization variables. Both scales showed acceptable to excellent internal consistencies (Cronbach’s ⍺ = 0.78 for lifetime victimization scale and Cronbach’s ⍺ = 0.94 for recent victimization scale). Both lifetime and recent victimization sum scores were further divided into four groups to form groups of similar size (0 victimization experiences, 1–2 victimization experiences, 3–4 victimization experiences, and 5 or more victimization experiences), coded as dummy variables (1 = belonging to the group; 0 = not belonging to the group).

Psychosocial well-being. The dependent variable adolescent’s psychosocial well-being was measured using the Strengths and Difficulties Questionnaire (SDQ; Goodman, 2001), a widely used 25-item self-report questionnaire about the feelings and behaviour of 3–16-year-old children and adolescents. SDQ targets the adolescent’s emotional symptoms, conduct problems, hyperactivity and inattention, peer relationship problems, and prosocial behaviour. The sum variable of all items, excluding prosocial behaviour, with a maximum score of 40 was used in this study, reflecting the total difficulties and overall adjustment. In addition, subscales of externalizing (including hyperactivity, inattention, and conduct problems) and internalizing symptoms (including emotional symptoms and peer relationship problems) with both 10 items and maximum score of 20 were formed. The sum variable (Cronbach’s ⍺ = 0.79) and the externalizing (Cronbach’s ⍺ = 0.73) and internalizing symptoms subscales (Cronbach’s ⍺ = 0.75) showed acceptable internal consistencies. A factor analysis in a Finnish population-based sample, conducted elsewhere (Peltonen et al., 2010), supported a three-factor solution, including externalizing symptoms, internalizing symptoms, and prosocial behaviour.

Control variables

Family income. In the FSRD survey, the participant was asked to estimate their family’s financial situation compared with other families familiar to them using a 7-point Likert scale (1 = a lot worse to 7 = a lot better). This variable was used to represent the family’s socioeconomic status.

Family structure. Family structure was measured on a 7-point scale (1 = together with my mother and father, 2 = partly with my mother, partly with my father, 3 = with my mother, 4 = with my father, 5 = with my mother and her spouse or my stepfather, 6 = with my father and his spouse or my stepmother, 7 = with other relatives, 8 = in a foster family, 9 = in other circumstances). A dichotomous variable was formed to reflect a nuclear family (1 = living with both mother and father, 0 = not living with both mother and father).

Statistical analyses

Statistical analyses were performed using the IBM SPSS version 28 software package. Gender differences were assessed through suitable mean comparisons (two independent sample t-test or χ²-test). A p-value of under 0.05 was considered statistically significant.

The linear regression analyses for both lifetime and recent victimization experience groups as independent variables were conducted using general linear models (GLMs). First, the models were run for the total sample using SDQ total score reflecting the psychosocial well-being as the dependent variable and the victimization group as an independent variable, adjusted for age and gender. The group with 0 victimization experiences was used as the reference to which the means of other groups’ SDQ scores were compared. The models were then run separately for boys and girls, adjusted for age. Second, the models were run initially for the total sample and then separately for boys and girls, adjusted for age, gender in the model for all, as well as family structure, and financial status as control variables. The differences between boys and girls were measured by adding the interaction term gender*victimization.

GLM was used to assess individual effects of single victimization experiences on psychosocial well-being. First, the models were run using single experiences as independent variables, adjusted for age and gender (univariate model). The models were then adjusted by adding other victimization experiences to the model (multivariate model).

Results

Sample characteristics

The study comprised 5509 participants, 5473 of whom had answered the SDQ questions. Of the sample, 48.5% were male and 49.8% female, and 1.7% of the participants reported their sex as ‘other’. Participants’ age range was 15–16 years (mean age 15.2 years). Most participants (96.1%) were born in Finland. Two-thirds (67%) of the participants (both girls and boys) lived in a nuclear family. On average, the adolescents estimated their family income to be better than in other families (M = 4.43, SD = 1.15) on a 7-point Likert scale, but boys (M = 4.57, SD = 1.16) had estimated their families’ financial situation to be slightly better than girls (M = 4.30, SD = 1.12) (t(5484.37) = −8.64, p < 0.001).

Descriptive statistics of victimization experiences and psychiatric symptoms

The frequencies of individual lifetime and recent victimization experiences are presented in Table 1. Girls had experienced on average 2.79 (SD = 2.64) lifetime victimizations and 1.04 (SD = 1.70) recent victimizations; boys had experienced 2.70 (SD = 2.57) lifetime victimizations and 1.01 (SD = 1.75) recent victimizations. Boys and girls did not significantly differ from each other in the total number of lifetime (t(5507) = 1.27, p = 0.21) or recent (t(5507) = 0.77, p = 0.21) victimization experiences.

Table 1. Frequencies of victimization experiences (lifetime and recent).

	Lifetime victimization						Recent victimization
	All (n = 5509)		Boys (n = 2717)		Girls (n = 2792)		All (n = 5509)		Boys (n = 2717)		Girls (n = 2792)
Range, Mean, SD	0–12	2.75 (2.61)	0–12	2.70 (2.58)	0–12	2.79 (2.64)	0–12	1.02 (1.73)	0–12	1.01 (1.75)	0–12	1.04 (1.70)
Type of victimization (% yes)
Destruction of personal property	33.9		37.8		30.1		10.8		13.8		7.8
Violent robbery	8.7		12.8		4.7		3.2		4.8		1.5
Theft of personal property	40.1		43.0		37.2		12.6		14.6		10.5
Bullying	40.3		35.1		45.4		10.0		8.6		11.4
Threats of violence	31.5		37.3		25.9		13.5		16.7		10.3
Physical abuse	24.9		32.4		17.6		9.0		11.8		6.2
Hate crime	10.3		11.8		8.9		5.2		5.6		4.8
Parental corporal punishment	22.2		19.2		25.1		3.5		2.4		4.6
Cyberbullying	27.5		22.3		32.5		13.4		12.1		14.7
Sexual abuse (adult)	7.4		1.9		12.8		4.4		1.1		7.6
Sexual abuse (peer)	13.6		5.3		21.7		9.0		3.2		14.8
Internet crime	14.4		11.3		17.4		7.8		5.7		9.8

The descriptive statistics of the study variables (victimization experiences and SDQ scores) are presented in Table 2. Over the life span, more than half of the participants had 0–2 victimization experiences and 23.2% had 5 or more experiences. In the last 12 months, most participants had 0–2 victimization experiences, whereas 5.5% had 5 or more experiences. According to a χ²-test, the number of boys and girls in the groups differed only in the recent victimization groups of 0 and 1–2 experiences, and the differences were negligible.

Table 2. Descriptive statistics of victimization experiences (lifetime and recent) and SDQ scores.

	All (n = 5509)		Boys (n = 2717)		Girls (n = 2792)
	n	%	n	%	n	%
Number of victimization experiences (lifetime)
0	1252	22.7	641	23.6	611	21.9
1–2	1845	33.5	900	33.1	945	33.8
3–4	1135	20.6	555	20.4	580	20.8
5 or more	1277	23.2	621	22.9	656	23.5
Number of victimization experiences (recent)
0	3215	58.4	1630	60.0	1585	56.8
1–2	1493	27.1	702	25.8	791	28.3
3–4	500	9.1	226	8.3	274	9.8
5 or more	301	5.5	159	5.9	142	5.1
	All (n = 5473)		Boys (n = 2696)		Girls (n = 2777)
SDQ total score (Range, Mean, SD)	0–33	10.95 (5.54)	0–31	9.69 (5.21)	0–33	12.17 (5.58)
SDQ externalizing symptoms (Range, Mean, SD)	0–20	5.67 (3.53)	0–18	4.51 (3.24)	0–19	6.80 (3.44)
SDQ internalizing symptoms (Range, Mean, SD)	0–20	5.28 (3.24)	0–18	5.17 (3.11)	0–20	5.38 (3.36)

Note. SDQ = Strengths and Difficulties Questionnaire.

There was a significant difference in SDQ scores between boys and girls in SDQ total score (t(5463.13) = 17.03, p < 0.001), externalizing subscale score (t(5466.29) = 25.31, p < 0.001), and internalizing subscale score (t(5457.41) = 2.31, p < 0.05), with girls reporting higher SDQ scores on all domains (Table 2).

Associations between number of victimization experiences and psychosocial well-being

As seen in Figure 1, the SDQ total scores increased with an increasing number of victimization experiences. Among all victimization groups in both time frames (lifetime or recent), SDQ scores were significantly higher for girls than for boys. In both boys and girls and in both time frames, the SDQ scores were significantly higher in each successively victimized group. The same increasing pattern was observed in externalizing and internalizing symptoms subscale score (see Supplementary table A).

Figure 1. SDQ total score by number of lifetime and recent victimization experiences.

Note. This table illustrates the SDQ total score by number of lifetime and recent victimization experiences in boys and girls. SDQ = Strengths and Difficulties Questionnaire (boys n = 2696, girls n = 2777; participants completing the questionnaire).

The results of GLM regression models for the relation between psychosocial well-being and lifetime and recent victimizations are presented in Table 3. The models were adjusted by age and gender, age being non-significant and gender significant, with girls scoring higher than boys on all SDQ domains in all victimization groups. Lifetime victimization experiences increased the psychosocial distress measured with SDQ total score significantly more in girls than boys. In recent victimization models, the difference between boys and girls was significant in the group with 3–4 victimization experiences, where girls had higher scores on SDQ total score than boys.

Table 3. Results of GLM regression. Victimization experiences (lifetime and recent) regressed on SDQ total score and externalizing and internalizing symptoms subscale scores. All adjusted for gender and age; separate analyses for boys and girls, adjusted for age.

	SDQ total score^a			SDQ externalizing symptoms subscale score^b			SDQ internalizing symptoms subscale score^c
	All (n = 5473)	Boys (n = 2696)	Girls (n = 2777)	All (n = 5473)	Boys (n = 2696)	Girls (n = 2777)	All (n = 5473)	Boys (n = 2696)	Girls (n = 2777)
	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)
Lifetime victimization
0 experiences	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)
1–2 experiences	1.23 (0.97–1.68)***	0.84 (0.36–1.33)***	1.82 (1.32–2.33)***	0.82 (0.60–1.05)***	0.51 (0.20–0.83)**	1.12 (0.81–1.46)***	0.51 (0.29–0.72)***	0.33 (0.03–0.63)*	0.69 (0.37–1.01)***
3–4 experiences	3.44 (3.05–3.84)***	2.76 (2.22–3.31)***	4.13 (3.56–4.70)***	1.82 (1.56–2.07)***	1.43 (1.08–1.78)***	2.20 (1.84–2.57)***	1.63 (1.38–1.87)***	1.33 (1.00–1.67)***	1.93 (1.58–2.29)***
5 or more experiences	6.12 (5.75–6.52)***	5.43 (4.89–5.96)***	6.85 (6.30–7.40)***	3.45 (2.89–3.38)***	2.77 (2.43–3.11)***	3.51 (3.16–3.87)***	3.00 (2.76–3.24)***	2.66 (2.33–2.98)***	3.34 (2.99–3.68)***
Linear trend per event	0.93 (0.88–0.98)***	0.84 (0.77–0.91)***	1.01 (0.94–1.08)***	0.46 (0.43–0.50)***	0.42 (0.38–0.47)***	0.50 (0.46–0.55)***	0.47 (0.44–0.50)***	0.42 (0.38–0.47)***	0.52 (0.47–0.55)***
Recent victimization
0 experiences	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)	(reference)
1–2 experiences	2.43 (2.12–2.74)***	2.10 (1.67–2.54)***	2.74 (2.30–3.18)***	1.07 (0.89–1.27)***	1.01 (0.73–1.29)***	1.14 (0.86–1.42)***	1.35 (1.17–1.54)***	1.09 (0.83–1.35)***	1.60 (1.33–1.87)***
3–4 experiences	4.51 (4.03–4.98)***	3.74 (3.06–4.42)***	5.16 (4.50–5.82)***	1.86 (1.55–2.16)***	1.65 (1.21–2.10)***	2.03 (1.61–2.46)***	2.65 (2.36–2.94)***	2.09 (1.68–2.50)***	3.13 (2.72–3.53)***
5 or more experiences	7.50 (6.91–8.10)***	7.12 (6.32–7.92)***	7.93 (7.04–8.82)***	3.43 (3.05–3.82)***	3.17 (2.66–3.68)***	3.73 (3.16–4.31)***	4.07 (3.71–4.43)***	3.95 (3.47–4.43)***	4.20 (3.65–4.74)***
Linear trend per event	1.21 (1.13–1.28)***	1.10 (0.99–1.20)***	1.32 (1.21–1.43)***	0.53 (0.48–0.58)***	0.48 (0.42–0.55)***	0.58 (0.51–0.65)***	0.67 (0.63–0.72)***	0.61 (0.55–0.67)***	0.74 (0.67–0.81)***

Note. SDQ = Strengths and Difficulties Questionnaire. *** = p < 0.001, ** = p < 0.01, * = p < 0.05.

Differences between boys and girls were assessed by including the interaction term gender*victimization in the model.

^aThe interaction term was significant in linear trend (lifetime p < 0.001 and recent p < 0.01). In lifetime victimization, the difference was significant in all three groups (p < 0.01, p < 0.001, p < 0.001). In recent victimization, the difference was significant in the group with 3–4 experiences (p < 0.01).

^bThe interaction term was significant in linear trend (lifetime p < 0.01 and recent p = 0.05). In lifetime victimization, the difference was significant in all three groups (p < 0.01, p < 0.01, p < 0.01). In recent victimization, the difference was not significant.

^cThe interaction term was significant in linear trend (lifetime p < 0.01 and recent p = 0.01). In lifetime victimization, the difference was significant in the groups with 3–4 experiences (p < 0.05) and 5 or more experiences (p < 0.01). In recent victimization, the difference was significant in the groups with 1–2 experiences (p < 0.01) and 3–4 experiences (p < 0.001).

In both time frames and for all victimization groups, all SDQ scores were significantly higher than in the reference group; the highest SDQ scores were measured for the group with 5 or more victimization experiences in SDQ total score and internalizing and externalizing symptoms subscales. A linear trend per victimization event was calculated to present the change in SDQ score by one victimization experience at a time; in lifetime victimization SDQ total score increased by 0.84 points for boys and 1.01 points for girls and in recent victimization by 1.10 points for boys and 1.32 points for girls by one victimization experience.

A supplementary table (Supplementary table B) is provided with further adjustments (family income and family structure) that resulted in only minor changes in the model.

Impact of the type of victimization experiences

In addition to polyvictimization experiences, individual effects of all 12 victimization experiences were analysed (Table 4). First, the analyses were run separately for each experience, adjusted for gender and age. As the cumulative nature of the effect of PTEs is known, all victimization experiences were then included in the multivariate model. When controlled for all other victimization experiences, threats of violence in recent victimization experiences as well as parental corporal punishment, hate crime, bullying, and cyberbullying in both time frames had the greatest impact on elevating the SDQ total score. Notably, the differences were not particularly large based on confidence intervals.

Table 4. Linear regression results of effects of single lifetime and recent victimization experienceson SDQ total score (n = 5473). First run as univariate analysis separately for each experience, adjusted for gender and age; then run as multivariate analysis, adjusted for gender, age, and other victimization experiences.

	Lifetime victimization		Recent victimization
	Univariate	Multivariate	Univariate	Multivariate
	B (95% CI)	B (95% CI)	B (95% CI)	B (95% CI)
Destruction of personal property	2.56 (2.27–2.86)***	0.20 (−0.12–0.51)	2.95 (2.49–3.40)***	0.39 (−0.09–0.86)
Violent robbery	4.31 (3.81–4.82)***	1.00 (0.49–1.51)***	4.93 (4.11–5.74)***	0.03 (−0.83–0.88)
Theft of personal property	2.13 (1.84–2.41)***	0.45 (0.16–0.74)**	2.45 (2.03–2.88)***	0.44 (0.00–0.87)*
Bullying	2.98 (2.70–3.27)***	1.21 (0.92–1.49)***	4.52 (4.05–4.98)***	2.49 (2.02–2.96)***
Threats of violence	3.64 (3.35–3.94)***	0.93 (0.58–1.29)***	4.19 (3.78–4.59)***	1.31 (0.82–1.81)***
Physical abuse	3.48 (3.16–3.81)***	0.63 (0.26–0.99)***	4.23 (3.74–4.73)***	0.92 (0.35–1.49)**
Hate crime	4.44 (3.98–4.90)***	1.40 (0.93–1.87)***	4.86 (4.23–5.50)***	1.30 (0.65–1.96)***
Parental corporal punishment	3.17 (2.84–3.51)***	1.67 (1.35–1.99)***	4.74 (3.96–5.51)***	2.30 (1.56–3.04)***
Cyberbullying	3.70 (3.40–4.01)***	1.33 (1.00–1.67)***	4.28 (3.88–4.69)***	1.96 (1.52–2.41)***
Sexual abuse (adult)	3.69 (3.13–4.25)***	0.95 (0.41–1.49)***	3.82 (3.11–4.52)***	1.17 (0.47–1.86)***
Sexual abuse (peer)	3.81 (3.39–4.23)***	1.14 (0.72–1.57)***	3.78 (3.28–4.28)***	1.24 (0.73–1.76)***
Internet crime	3.56 (3.16–3.96)***	0.85 (0.45–1.26)***	4.21 (3.69–4.74)***	1.07 (0.52–1.63)***

Note. SDQ = Strengths and Difficulties Questionnaire. *** = p < 0.001, ** = p < 0.01, * = p = 0.05.

Discussion

The purpose of this population-based study was to evaluate the frequency of both individual PTEs and polyvictimization over the life span and during the recent year in a population-based sample of Finnish adolescents and to investigate the effects of these experiences on adolescents’ psychosocial well-being. In line with earlier research (Ellonen & Salmi, 2011), victimization experiences were common among Finnish adolescents, as the majority (77.3%) reported having at least one victimization during their lives, with the mean number of victimizations being 2.75. In the study of Ellonen and Salmi (2011), 9% of participants had experienced five or more victimizations during the past year and were identified as polyvictims, whereas in our study the corresponding proportion was 5.5%. This study contributes to the adolescent victimization research with a large sample, amount, type and timing of the victimization experiences as well as with current victimization types such as cyberbullying and hate crime.

Amount of victimization experiences and psychosocial well-being

As expected, an association between polyvictimization and psychosocial well-being was found in both time frames. The more victimization experiences the adolescents had, the more overall psychosocial distress as well as internalizing and externalizing psychological symptoms they reported. These findings are consistent with previous research in which polyvictimization has been linked to elevated risk for decreased social and emotional functioning and mental health issues, such as PTSD and depression, in adolescents (Fairbank & Fairbank, 2009; Lätsch et al., 2017; Suliman et al., 2009).

This pattern was found for both time frames of victimization experiences and for both genders. Girls’ psychosocial well-being seemed to be more negatively affected by their experiences, although boys and girls had experienced similar numbers of victimizations in both time frames. Parallelly, earlier research has shown that girls suffer more often from PTSD than boys (Alisic et al., 2014; McLaughlin et al., 2013). This same pattern considering PTSD symptoms also been shown in adults, even though female participants had experienced less PTEs than boys (Olff et al., 2007; Tolin & Foa, 2006). These differences in PTSD symptoms may be related to differences in psychobiology, for example, variations in the impact on various regions of the brain (e.g. Klabunde et al., 2017), but they remain understudied (Olff, 2017). According to Olff (2017), there are gender differences in handling with stressful situations, problem-focused coping being more prevalent in men and emotion-focused and defensive coping in women. Furthermore, there are indications that the adoption of socially accepted masculine traits can serve as a protective element for male mental well-being, whereas the conventional societal norms promote conformity to gender roles, such as the expectation that females should exhibit emotional passivity, appear to have less beneficial effects on women’s mental health (Yuan et al., 2021). Additionally, gender differences in different types of victimization experiences were not in the centre of this study, but they might contribute to the differences in psychosocial well-being. Contrary to findings in the present study, some research has not found different effects for boys and girls (McLaughlin et al., 2012). Even though the assessment instrument used in our study did not assess PTSD or other mental health disorders, general psychological distress likely represents some of the aspects of mental health symptoms related to victimization experiences.

Research to gender differences in other symptoms than PTSD after PTE exposure is scarce. In a study with Ghanaian adolescents, girls reported more mental health symptoms (depression, anxiety and suicidal ideation) and poorer well-being than boys, but the prevalence of adverse life experiences was similar across the genders and the association between adverse life experiences and mental health was not largely influenced by gender (Adjorlolo et al., 2022). Girls have been reported to be more likely to show non-suicidal self-injury and suicidal behaviours after adverse childhood experiences in comparison to boys (Xiao et al., 2023). The results of the current study are novel and gender differences in the general psychosocial well-being (i.e. other than PTSD) after PTE exposure need to be further investigated.

Timing of victimization experiences

The linear trend for recent victimization predicting the decrease in psychosocial well-being was higher than for lifetime victimization. This result might have several explanations. The self-report nature of the study may have had an impact on the reported victimization experiences; adolescents might remember the most recent experiences better and report them more precisely. Also, trauma-related symptoms have been observed to decrease over time (Meiser-Stedman et al., 2005). Notably, the effects of prior victimization experiences might interact with the effects of the most recent victimization experiences, and these effects were not controlled in this study when investigating recent victimization experiences.

Effects of the type of victimization experiences

A further objective of our study was to investigate the effects of specific types of individual victimization experiences (parental corporal punishment, physical and sexual abuse). First, as expected, parental corporal punishment was among the victimization experiences with the greatest impact on overall psychosocial well-being, both that having happened earlier and more recently. Evidence exists for detrimental effects of interpersonal victimization, such as war, terrorism, and violence, especially in girls (Alisic et al., 2014). Interestingly, the strong predictive effect of sexual abuse (McLaughlin et al., 2012) or physical abuse on psychosocial well-being was not replicated in this sample after controlling for other victimization experiences. Second, having experienced hate crime, bullying, or cyberbullying in both time frames and threats of violence more recently were shown to be associated with decreased psychosocial well-being.

In fact, these types of victimization reflect interpersonal violence in an adolescent’s everyday life. Cyberbullying and bullying were common victimization types in this sample, and hate crimes and hate-motivated violence might have a growing tendency among youth, as the number of refugees, immigrants, and foreign residents in Finland is on the rise (Official Statistics of Finland, 2022, March, 31), with threats of violence being distinctive to bullying. Unfortunately, since the sample of the current study consisted mostly of adolescents born in Finland, the group size differences of the ethnic groups did not allow for further comparisons in this study to investigate this speculation. However, there is evidence for polyvictimization accounting for differences in mental health symptoms in adolescents with different ethnic backgrounds (Andrews et al., 2015), reflecting the need to investigate the effect of hate crimes and hate-motivated violence as a single trauma type. The participants reporting their sex as ‘other’ pose another interesting minority that warrants further investigations. Recent research has shown that transgender youth, especially with non-binary identity, appear to be involved in bullying both as the victim and as the perpetrator (Heino et al., 2021). Therefore, these results highlight an important future direction for polyvictimization research, as new forms of violence among adolescents must be considered.

Moreover, a current review on cyberbullying has described similar consequences to traditional bullying; targets of cyberbullying suffer from depression, anxiety, suicidal behaviour, and somatic symptoms (Nixon, 2014). Interestingly, one study found a connection between some intrafamilial adverse events (violence, alcohol or drug problems, divorce) and cyberbullying victimization in adolescents in 7^th to 9^th grades (Gualco et al., 2022). A large study among adolescents in different countries in Europe and Asia has reported prevalence rates of both traditional bullying and cyberbullying and suggested that future bullying prevention programmes include both forms of bullying (Chudal et al., 2022). As most adolescents use the internet for various purposes (schoolwork, social life, browsing), it is important to also focus on victimization that happens online and how to identify the adolescent victims of cyberbullying.

Future directions

The benefits of knowledge about the prevalence of adolescent victimization on a national level are undisputable, allowing for better recognition of trauma symptoms and more efficient treatment planning for children, adolescents, and their families. The results of the study showed that vast majority of Finnish adolescents have experienced at least one PTE and underline the importance of trauma-informed treatment. To our knowledge, this is the first attempt to include a detailed examination of the more recently acknowledged effects of interpersonal victimization – hate crime, cyberbullying, and internet crimes – to the adolescent polyvictimization research, raising important directions for future studies. These results encourage to longitudinal study designs concerning those PTE types to find out their impact on adolescent mental health.

Limitations and strengths

The study has some limitations that must be addressed. First, this was a cross-sectional study representing the adolescent’s situation at a specific point of time, which makes interpretation of associations more complicated and drawing causalities impossible. Cross-sectional and retrospective studies are also susceptible to response bias, e.g. recall bias, which might be an important factor in assessing adolescent’s victimization experiences over their life span. Second, a consensus on the definition of the polyvictimization construct (Lee et al., 2022) is needed to allow for more concordant and comparable study designs. In our study, polyvictims were not categorically identified. Future studies should control for an individual’s polyvictimization status when assessing the impact of a single trauma type since there is evidence for a reduced impact of single trauma types when polyvictimization is controlled (Finkelhor et al., 2007a; Lätsch et al., 2017; Turner et al., 2010).

This study was based on adolescents’ self-reports about their victimization experiences and their psychosocial well-being, which can be seen as both a strength and a limitation. When providing sensitive information, such as that here, the participants may be unable or unwilling to answer truthfully. However, the study was conducted anonymously, decreasing the possibility of under- or overreporting. In this study, the victimization experiences were assessed in two different time frames, which might strengthen the credibility of answers, particularly concerning recent events; children and adolescents may be able to remember more recent events better than those in the distant past. In our study, using information about the perpetrator, frequency or the severity of the individual victimization types could have provided more detailed insight into effects on psychosocial well-being. Moreover, using multiple respondents, e.g. parents and teachers, would have increased the quality of the study, especially when assessing the psychosocial well-being of adolescents.

In addition, psychosocial well-being was assessed with a general instrument measuring well-being or distress across a broad spectrum. Other studies have utilized all SDQ subscale scores for adolescents (Lätsch et al., 2017) or SDQ total score for children (Gustafsson et al., 2009) in order to assess polyvictimization consequences. In the current study, a more general approach, including total adjustment and internalizing and externalizing symptoms, was chosen to focus on the effects of both time frames and individual events. As known, victimization experiences can affect an individual in many ways, and a more detailed assessment of psychological distress would have provided more information about the possible consequences of victimization experiences such as specific mental health disorders. Furthermore, a clinical interview would be a more sophisticated, yet resource-intensive way to assess adolescents’ mental health problems.

The population-based design and large sample size are major strengths of the study, whereas the limited age range (15–16 years) of participants only allows cautious conclusions to be drawn. Including a wider age range would provide more information about the associations between victimization experiences and well-being in youth. As far as we know, only a few studies investigating polyvictimization in Finnish adolescent samples are based on adolescent self-report, and thus, our study provides preliminary findings on the topic. Future studies should address these issues in more detail.

Conclusion

Findings of the study demonstrate precursory knowledge about the association between victimization experiences and psychosocial well-being in Finnish 15–16-year-old adolescents. A pattern of weakened psychosocial well-being with an increasing number of victimization experiences emerged. Taken together, our results are in line with earlier research, as there was a significant increase in psychosocial distress in individuals who had experienced several different PTEs. In addition, parental corporal punishment, bullying, cyberbullying, hate crimes, and threats of violence were preliminarily identified to have an association with weakened psychosocial well-being, providing important directions for further research. The results reflect firstly the importance of preventing adolescents from experiencing PTEs and secondly the necessity of identifying adolescents at risk of victimization and those potentially traumatized and in need of treatment.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/02673843.2023.2273385

Additional information

Funding

This work was supported by grants from the following non-profit organizations: the Pediatric Research Foundation of Finland and the Helsinki University Hospital Research Funds (grants TYH2021301 and TYH2022203).

Notes on contributors

Katri Lahti

Katri Lahti, MSc., is clinical psychologist and doctoral researcher in Child Psychiatry, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. Her main interests are child and youth victimization and its impact on mental health.

Heidi Backman

Heidi Backman, Ph.D., is clinical psychologist and researcher in Child Psychiatry, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. Her scope of interests includes child and adolescent behavioral disorders, callous-unemotional traits, and parental practices.

Taina Laajasalo

Taina Laajasalo, Ph.D., is Assistant Professor (Docent) in Forensic Psychology at the University of Helsinki, Finland, and Chief Researcher at the Finnish Institute of Health and Welfare. Her main research activities relate to various aspects of child maltreatment, as well as child and adolescent behavioral disorders.

Markus Kaakinen

Markus Kaakinen, Ph.D., is a senior researcher at the Institute of Criminology and Legal Policy at the University of Helsinki. His research interests are related to youth delinquency and crime prevention, with a current emphasis on the importance of groups, technology and identities in criminal behavior.

Eeva T. Aronen

Eeva Aronen, M.D., Ph.D., is Professor in Child Psychiatry at the University of Helsinki and Senior Child Psychiatrist at Helsinki University Hospital, Helsinki, Finland. Her scope of interests includes sleep, executive functions, stress and adverse life events in children and their relevance to child psychiatric disorders, behavioral disorders in children, and parenting.