In this long-term prospective study, young women exposed to violence in late adolescence had increased odds of heavy illness burden and bad self reported health in adulthood compared to non-exposed women, controlling for social demographics, health and smoking and adult violence exposure. The men did not show the same relationship between violence exposure in adolescence and increased odds of heavy illness burden or bad self-reported health.
Research has shown in several important respects that there is a cross-sectional and retrospective relationship between violence exposure and negative health outcomes [1
]. Few, if any, prospective studies showing long-term relationships between exposure to violence and adverse health have been published. The objective of most studies is to prove a casual relationship between two variables; that is, a change in one variable "causes" a change in the other, rather than an associative relationship. An associative relationship is not necessarily causal, but can be explained by the presence of other 'un-seen' variables to which the two variables being studied are themselves separately linked. Prospective studies are often regarded as strong as they deal methodologically with difficulties such as confounding and other biases. This study suggests that exposure to violence in young women may have a longitudinal relationship to negative health outcomes. The men did not show a similar distinct relationship. Instead, poor health status in earlier life was more strongly related to negative health outcomes in the long-term.
Lately, two emerging understandings of how early experiences of violence may affect adult health have been established; (1) latent effects of adversities during critical periods and (2) accumulated exposure of stressful experiences [37
]. The first theory is explained by the existing evidence that suggests that early childhood trauma (including violence exposure, abuse and neglect) activates stress associated hormonal and neurochemical systems in the body that under normal circumstances are protective but become toxic with severe exposure, with resulting negative physical effects on the body [37
The second theory is built on the strong relationship between retrospective adult reports of traumatic childhood or adolescent incidents and/or amount of reported violent episodes and increased prevalence of health impairments in adulthood [21
]. In relation to accumulating traumatic childhood or adolescent events, family characteristics (such as parental psychopathology, parental loss or absence or parental divorce) during the upbringing contribute to the development of subsequent future health-related well-being or problems in adulthood [40
]. Also, persons who have experienced adversities during their upbringing, are more likely to participate in high-risk behaviors [41
], which are related to both negative health and violence [43
]. Continual psychological pressure and/or persistent wear and tear of the body due to repeated stressful or traumatic experiences over the life course might dysregulate the normal physiological adaptations to stress and threats, and later sensitivity to stress [37
], or influence immune functioning which may in turn contribute to increased adult health problems [44
Any of these theories may explain the long-term effects on health seen in our study. It is reasonable to assume that the violence exposure in the life stage of adolescence, as well as in childhood, exercise negative long term effects on health [20
], while several crucial developmental psychological transitions are negotiated, in relation to other stages in life [20
]. Also, we do not know the amount of possible accumulating adverse events, but it is well-known that previous exposure to violence is a strong risk factor for further exposure [47
In our study the results also express distinct gender differences concerning the prospective effects of reported exposure to violence in adolescence and health status in adulthood. This has been demonstrated in one earlier short-term prospective study [50
] and in some cross-sectional studies [50
], Several reasons for this have been put forward. The magnitude, nature and health impact of violence differ greatly for young men and women. In a study by Sundaram et.al. 2004, young men were significantly more likely to experience violence than women [50
] but the associations between physical violence, poor self rated health and self reported morbidity were significant for women, but not men. Danielsson et. al. (2009) showed in their study pronounced gender differences in adolescent and young adults, both in type, prevalence and outcomes of exposure to violence [51
]. The young women reported more severe adverse effects from all types of abuse than the men. It is probable that gender specific experiences of violence and gender differences in health perceptions interact and contribute to a gender specific process of victimization [47
Gender differences in health outcomes could also be understood as having possible biological explanations [53
]. Research has shown sex differences in brain maturation during childhood and adolescence indicating possible diverse developmental pathways due to different or similar adverse experiences such as violence exposure [50
]. One potential mechanism is sex differences in the development of brain structures that process experiences (HPA axis; hypothalamic-pituitary-adrenal axis). In females, there is an increased response of the HPA axis to stress with advancing puberty, while in males the response is decreased, possibly associated with increased testosterone levels [39
]. This, in connection to the stress associated with violence exposure, might differentiate males and females with respect to the rates of onsets, courses and symptomatology of common psychiatric disorders and psychological symptoms [56
]. It is well known that women in the general population in all ages, have higher rates of post traumatic stress disorder (PTSD), which could indicate a psychological fragility were violence exposure could be more harmful to young women than young men [57
]. Furthermore; adolescence has been described as the identity formation life stage [59
]. Many factors may play a role in this period of life, including different stresses associated with social gender expectations related to men and women gender roles, the higher incidence of exposure to trauma experiences in young females, and differences in social cognitive function such as rejection sensitivity, or even a gendered difference in reporting symptoms [56
Life course remarks
In order to isolate a potential link between adolescent exposure to violence and adult health, societal changes have to be looked upon and possibly controlled for [61
]. In our samples, general patterns are apparent, but these patterns were generally not statistically significant; the educational level rose, and the percentage of social class III families decreased from the 1974 to the 2000 cohorts, and so did smoking. On the other hand, the percentages of reported childhood economical problems were almost steady throughout the entire research cohort era. Also, there was a tendency for two parent families to decrease during the same period.
As three different survey year cohorts (1974, 1981, and 1991) are studied in relation to the 2000 health outcome, four different historical contexts are possible to reflect on. During the studied period 1974–2000, a number of significant changes did take place in the Swedish society. In welfare terms, the study period begins when the Swedish welfare state was still expanding and ends during a period of retrenchment. The labor market situation deteriorated from almost full employment to high unemployment [61
]. Although there have been a changing historical context during the whole study period, there has not been an economical collapse with depression and familial deprivation as a consequence. But understanding and controlling the historical context within the performed studies make it more possible to rule out potential societal effects over time and allow the focus to be on the primary exposures and outcomes.
This study has certain limitations. First, the relative low sample size affects the power of the results. The sample size together with some low frequencies could generate numerical problems and introducing wide confidence intervals [64
]. Despite this, significant results were found. Still, an increased chance of false negatives remains, for example with regard to the low number of statistical significant differences found among the social demographic descriptives (Table
). Second, the measures used in this study to capture exposure to violence are crude and possibly underestimate the prevalence of exposure violence. Third, it is possible that respondents’ conceptual understandings of some survey questions might have changed throughout the study period. For example, the importance of different health related assessments varies between adolescence and adulthood since health problems differ between adolescence and adulthood (the study is framed in a broad age range), and probably between the earlier survey cohort and the later [65
]. However, in our study even after controlling for family upbringing related factors, behavioral factors, and adolescent illness burden, there remained a significant relationship between adolescent exposure to violence and adult health status of the women.