This study has shown that many adolescents experience peer victimisation in schools in northern Russia and that victimisation is strongly associated with psychological and somatic health problems. These findings are consistent with those of a recent meta-analysis of the consequences of bullying and victimisation for psychosomatic health . Moreover, the relation we observed between experiencing more victimisation and having higher odds of poor health accords with findings from the earlier HBSC study conducted in 28 countries in Europe and North America  and a recent smaller-scale study from Norway  where a graded association was noted between the frequency of having been bullied and the likelihood of reporting different negative health outcomes. However, it was noticeable in the current study that in terms of somatic symptoms, for more than half of the symptoms there were higher odds among those girls and boys scoring 9–10. This was not observed for the psychological symptoms, where with the sole exception of anxiety among girls, those in the highest victimisation category (scoring 11–27) had the highest odds of reporting poor health. This and the fact that even relatively few instances of victimisation (i.e., scores of 1–2) were associated with poorer health outcomes in some cases highlights the necessity of future research using more finely graded categories of victimisation (i.e. relating to both type and intensity of victimisation) to better understand the effects of peer victimisation on adolescent health. Moreover, it seems unlikely that our findings are an artefact of the categorisation system we employed as when the victimisation variable was entered into the regression analysis as a continuous variable it was significantly associated with all of the health problems.
It has been suggested that stress may be the mechanism that links the experience of peer victimisation to negative health outcomes . In relation to this, it is possible that social support, which can act to buffer the effects of stressful environments , may reduce the detrimental effects of peer victimisation on health outcomes . This notion is supported by research that showed how differences in familial warmth protected against subsequent behavioural disorders in identical twins subject to victimisation  and by evidence that support from both parents and teachers may mitigate the effects of victimisation . If support does act to mitigate the detrimental effects of victimisation on well-being this may explain the strong relationship we observed between victimisation and negative health outcomes in the current study. Specifically, some research indicates that the majority of Russian adolescents tend not to report experiencing peer victimisation and they feel that they cannot turn to teachers for help .
This suggests that the better training of teachers to recognise what have been described as the physical, psychosomatic and behavioural ‘warning signs’ of peer victimisation  may be one potentially effective intervention when it comes to addressing this issue. This could perhaps be one element in comprehensive school-based anti-bullying programmes which recent review articles have linked to a reduction in the occurrence of both bullying and victimisation in schools in other settings [25, 26]. However, as other review evidence questions the extent to which school-based interventions reduce actual bullying behaviours , it is also important that possible actions to mitigate bullying and its effects are not restricted solely to schools. For example, other adults who come into contact with children – such as doctors – should also be made aware of the potential signs of bullying and what to do when children present with possible symptoms as a result of being bullied .
There are several possible limitations to this study that should be mentioned. First, as the data were self-reported with no means of verification there is the potential for reporting bias. Second, there is also a possibility of selection bias as we were only able to gather information from those children in school on the day of the survey. This may have been problematic as previous research has linked school absenteeism to victimisation . Third, we equated frequency of victimisation with the intensity of the victimisation experience. However, the effects of being sworn at several times might differ markedly, say, from those of being badly physically beaten on only one occasion. Fourth, the questions on victimisation and health outcomes referred to different time periods i.e. this school year and the previous 30 days. The use of different reference periods may have introduced the possibility of bias into the study. Fifth, although we have followed previous authors in using 2–3 times as a cut-off to determine what constitutes victimisation, in the study we referenced, the precise definition was “‘2 or 3 times a month’ (in the past couple of months)” [15, p. 263]. In the current study however, the victimisation took place ‘During this school year’ i.e. the school year began in September and the survey was undertaken in March to May of the following year (more than 6 months after the beginning of the school year). Over this much longer time period the effects of experiencing 2–3 instances of victimisation might be very different from those suggested in the reference article. This indicates that the prevalence estimates from this study may not be strictly comparable with those from earlier studies using this victimisation cut-off point. Sixth, the somatic symptom ‘problems with eyes’ was not precisely defined and may have been interpreted in different ways by different respondents. Finally, the data we collected were cross-sectional so it is impossible to determine the order of events. A recent review of longitudinal research studies has suggested for example, that the relation between peer victimisation and internalising problems may be bi-directional where peer victimisation both leads to, and is a consequence of such problems .