Our aim in this web-based study was to identify subgroups of students based on risky behaviors, and determine the prevalence of these subgroups among a representative sample of students in Tabriz, Iran. Our results showed that the frequency of smoking, hookah use, alcohol use, substance abuse, and having unsafe sex, as the risky behaviors, were 18.5%, 9.1%, 9.2%, 8.3% and 14.5%, respectively. The results of a previously published meta-analysis also showed the prevalence rate of high-risk sexual behaviors among Ethiopian students to be 42%, approximately . In a study among American students, the prevalence rates of alcohol consumption, cigarette and hookah smoking were 44%, 31% and 22%, respectively . Another study on Canadian students revealed that 55% of students smoked cigarette, 62% consumed alcohol, 36% had drug abuse, and 28% had high-risk sexual intercourse . The reason for the low prevalence of high-risk behaviors among Iranian students compared to other countries may be due to cultural-religious beliefs in Iran, as well as the religious prohibition of alcohol consumption and the legal prohibition of alcohol and drug abuse.
The results of a study among students in Khorramabad, Iran, showed that the prevalence rates of smoking cigarette, drug abuse, and alcohol consumption were 3.7%, 2.4%, and 5.5%, respectively . Another study conducted on Tehran University students showed that the prevalence rates of alcohol, drug abuse, and high-risk sexual behaviors were 4.6%, 2.3%, and 5.6%, respectively . Another study on the prevalence of high-risk behaviors among students in Rudan, Iran, demonstrated that the prevalence rates of drug use, alcohol consumption, and high-risk sexual behaviors were 5.5%, 4.9%, and 6.6%, respectively . Another study conducted in 2011 in Tabriz, Iran, showed that the prevalence rates of smoking cigarette, smoking hookah, alcohol consumption, drug abuse, and high-risk sexual intercourse were 15.8%, 8.5%, 8%, 7.6%, and 10.8%, respectively . As evident, all these studies reported lower levels of risk-taking behaviors compared to our findings in the present study, which may be attributed to the use of online questionnaire in the present study, as evidences suggest that the quality of responses in sensitive issues to online questionnaires is better than that of paper questionnaires [15–20].
Overall, the results of this study showed that smoking cigarette (18.5%, boys: 36.4% and girls: 7.6%) and drug abuse (8.3%, boys: 15.3% and girls: 4.1%) were the most and the least common risky behaviors among Iranian college students, respectively. Our results also showed that boys had more risky behaviors than girls. These findings were in line with those of other studies conducted on Iranian students [29–31]. The low prevalence of risky behaviors among Iranian female students compared to male students can be due to different cultural and social expectations and greater freedom of boys which may facilitate the inclination of boys towards risky behaviors [32, 33].
A useful preventive approach is to pay attention to the concurrency of high-risk behaviors, Various studies have shown that engagement in one high-risk behavior is associated with engagement in other high-risk behaviors . Numerous studies have shown the concurrency of cigarette smoking and hookah smoking , smoking and alcohol consumption , cigarette smoking and drug abuse , and high-risk sexual behaviors and alcohol and drug abuse . The present study showed that the students of both genders had three subgroups in terms of risk taking behaviors including low-risk, smoking cigarette and high-risk, with the prevalence rates of 58.9%, 23.2% and 17.9% in boys and 87.9%, 10.6% and 1.5% in girls, respectively.
In the previous studies that used the clustering method to investigate high-risk behaviors among students, different subgroups of students in terms of high-risk behaviors were identified. A study in Tabriz showed three subgroups or classes of risky behaviors among students: 1. low-risk 2. smoking cigarette and hookah 3. high-risk . According to this study, 3.7% of boys and 0.4% of girls were in the high-risk class. As mentioned above, another reason could be attributed to the quality of responses due to online questionnaires in the present study. In the study conducted by Safiri et al (12), high-risk behaviors were less prevalent among girls than boys, which is in line with the findings of our study. Another study conducted in Tabriz in 2011, four subgroups of high-risk behaviors among students were identified: 1. low-risk 2. smoking cigarette and hookah 3. risky sexual behavior in girls and risky sexual behavior and alcohol consumption in boys, and 4. high-risk. According to this study, 13.3% of boys and 4.3% of girls had high-risk behaviors .
Studies applying the LCA method to investigate the concurrency of high-risk behaviors among students in different societies have different patterns of behavior among university students. For instance, a study on the US students involved in tobacco use, drug abuse, and alcohol consumption showed five latent classes for the behaviors . According to this study, being a boy increased the chance of placement in high-risk class. Additionally,, 61.8% of the students were in non/low user class and 5.6% were in poly-substance user class. Another similar study in Canada identified three latent classes of the behavior: 1. normal, 2. relatively healthy, and 3. high-risk. The prevalence rates of these classes were 65.7%, 14.5% and 19.8%, respectively . Similarly, Chiauzzi et al., in the US found the following four classes for the behaviors: 1. low-risk alcohol consumption/low prevalence of drug abuse, 2. lower alcohol consumption/moderate prevalence of drug abuse, 3. moderate-risk alcohol consumption/moderate prevalence of drug abuse, and 4. high-risk alcohol consumption/high prevalence of drug abuse. The prevalence rates of these classes were 46.0%, 20.2%, 13.6% and 20.2%, respectively . The high prevalence rates of high-risk classes in these studies compared to our study may be due to geographical and cultural differences and the type of norms.
There were several limitations in the present study. First, although we believe that the students provided us with a more honest answers, compared to paper-based surveys, our results were still based on the self-report of the participants. According to previous research, the studies in which online questionnaires are used for data collection have usually a problem with low response rate [40, 41], and this may come true for our study, yet we do not know how much was the number of students who viewed the questionnaire link but did not answer to the questions. Thus, we cannot determine the response rate. Another limitation of our web-based study is participation bias, which may distort the results , because certain people with internet access may have a social network account and wish to participate in the study.