Our study revealed that 22.4% of studied high school students were abnormal as total difficulties’ scores by using the Strength and Difficulty questionnaire. This prevalence rate was near to similar study conducted in Tehran which 20 percent of high schools students get abnormal scores . Although another study in Tehran showed 13.7 percent abnormality as total problem scoring among teenagers . Similar studies in other countries have presented somewhat similar findings, such that 25 percent prevalence among 11- 17 years old adolescents in a study in Nepal  and 18.7 percent abnormality in total problem scores among Brazilian teenagers 
In our study most common sub scale was peer relation problem (23%) followed by conduct disorder (18.1%) that differs from other studies. Rimal et al reported most prevalence abnormal scores in emotional subscale followed by Peer Relationship Problem . Also the results of a survey that was conducted in the Gaza Strip revealed that 34 % of participant adolescents gained abnormal scores in emotional problems, 22.7% of them had abnormal scores in conduct disorder subscale and 8.5 % were abnormal as hyperactivity scores .
Cury et al demonstrated 30.8% prevalence of emotional problem abnormal score, 17.7% abnormality in conduct disorder scores and 6.8 % abnormal scores in hyperactivity subscales in Brazilians youths .
Our study showed statistically significant difference in some subscales scores between girls and boy, such as emotional problem, peer relation problem subscales and total difficlties. The results of a similar survey that was conducted in Nepal also revealed significant statically difference in emotional problem subscale scores between two genders .similar study which conducted in Finland Indicated that girl students in 7th and 9th grade in high school gained higher scores in emotional problem subscale in comparison to boy students . Koskelainen and colleagues showed that girl students significantly have had more abnormality in emotional problem subscale scores . A similar study in southern European countries revealed statically significant difference between Italian girls and boys as prosocial behavior. Portuguese students demonstrated significant difference as hyperactivity subscale scores between two genders. 
in this study, variables can be classified as demographic, familial and economical factor .Table 3 in the crude analysis and subsequent along with table 4 in multivariate analysis revealed that females (OR: 2.52, 95% CI: 1.68-3.66), grade9 in school (OR: 0.52, 95% CI : 0.32 to 0.83) and the number of siblings (OR: 0.88, 95% CI: 0.78-0.99) in the crude analysis and in the multivariate analysis are important variables that associated with mental health problems. Another study showed that gender was associated with mental health problems, But Prevalence of mental health problems more men than women were reported . In this study, the number of siblings and grade 9 in school are protective factors and female gender is risk factor . Considering the observational nature of the study, the authors do not report the cause and effect relation between the variables.
Table 3 and 4 showed the separation of the parents was near to significant in univariable analysis (P=0.07), but not significant in multivariate analysis(P=0.24) . Another study showed that parental separation was associated with children’s behavioral/emotional problems 
In the present study were not observed the relation between education and job of parents with mental health problems .Another study indicated that father’s education is significant associated with total difficulties.  The results of the study by seyf hashemi et al. showed that abnormal SDQ total scores in adolescents were assoÂ¬ciated with lower level of education in both mothers and fathers. The children ‘s mother are housewives compared to children who are employing mothers , and children of unemployed parents were more odds to have a total difficulties .
Grade 9 in school is significantly associated with total difficulties, while in grade 10, 11 this association was not observed. Considering that grade in school may indicate the child’s age. So we can say there is a significant correlation between age and total difficulties. In another study has shown that the chances of total difficulties in young children more .
Univailable and multivariable analysis showed that there are significant association between number of siblings and total difficulties, more siblings protective role in the prevention of mental health problems. The results of another study showed that ADHD is more in children who have poor sibling’s relationship  The findings of this study showed that individual and family factors can increase the risk of mental health problem in adolescents.
Limited study population and lack of comparison among urban and rural students or other subcultures are our study limitations. Yet our study demonstrated high prevalence of relation problem and conduct disorder among high school students that it could bring a lot of problems in them developmental and educational Process .