Tobacco use remains a serious public health problem as it is a major cause of preventable diseases and death,1,2 or in other words, the most avoidable cause of death.3 Furthermore, secondhand smoke (SHS) exposure increases morbidity and mortality from coronary heart disease, lung cancer, respiratory infections, and other illnesses. Despite (inter)national tobacco prevention policies, such as a ban on smoking in public places, increased taxes on tobacco products and social education about the harmful effects of smoking, the recognition of the severe health consequences of smoking has been a slow process.4 While these types of policy interventions are only indirectly aimed at young adults, they demonstrate the positive effects of smoking bans in public spaces, being associated with reduced smoking progression, reduced consolidation of experimentation with regular smoking and increased smoking cessation among young adults.5 This is particularly relevant in areas where alcohol is served, given the strong association between tobacco and alcohol consumption in this age group.6

Starting smoking early increases the risk of regular smoking, and early adulthood is often associated with increased cigarette smoking and the establishment of regular smoking habits.7 The academic environment may constitute a context that favors tobacco use,8 as well as initiation/experimentation9 through the way it is socially accepted in this context (Nolen-Hoeksema, 2004). Several studies have shown an increase in the number of smokers on academic courses, both in the number of students who began to smoke regularly and in the number of cigarettes smoked daily.10 For example, Tercyak, Rodriguez and Audrain-McGovern (2007)11 found that 25% of those who reported never having smoked in high school started smoking a year later.

Given the impact at various levels of tobacco use and the initiation and increase of smoking among young adults, it is important to understand how knowledge and attitudes toward tobacco use relate to smoking behavior, as this understanding will enable the development of health education programs in higher education from a preventive and educational perspective. In scientific literature, several investigations have addressed the relationship between knowledge, attitudes and smoking habits in higher education.12–14 These studies take into account the fact that university students are in the process of being educated and that in future, they will be agents of change, responsible for promoting healthy lifestyles.15

The purpose of this study was to assess the prevalence of active smoking and secondhand smoke exposure among college students in the north of Portugal and analyze the relationship between knowledge about tobacco use and attitudes toward smoking.

The results showed that 68.8% of the students surveyed were non-smokers, 11.2% were former smokers and 20.1% were current smokers (7.3% occasional smokers, 2.9% regular smokers and 9.9% daily smokers). Regular and daily smokers consume on average 8.43 (SD=6.462) cigarettes per week and 8.33 (SD=4.870) cigarettes per day, respectively. Most current smokers started smoking aged under 17 (61.4%) and reported never having tried to quit smoking (59.7%). Only 34.4% of students reported (almost) never having been in enclosed spaces with smokers in the previous 7 days.

Tobacco use is positively correlated with smoker peers (rsp=.329, p<0.01) and with SHS exposure (rsp=.295, p<0.01), so, smokers tend to have more friends who smoke (χ2(8)=114,594, p=.000) and are more exposed to SHS than former smokers and non-smokers (χ2(8)=86,536, p=.000) (Table 1).

Smoking was significantly associated with sociodemographic characteristics such as which academic year, scientific area, age, being in an affective relationship, current residence, and professional situation. This means that there is a higher prevalence of smoking in: graduating students compared to the year first students (χ2(2)=8.260, p=.016); students of the economics and law compared to the other scientific areas (χ2(6)=20.103, p=.003); older students compared to the younger ones (χ2(2)=11.370, p=.003); students who are in a relationship compared to those who are not (χ2(2)=23,206, p=.000); students who moved away from home at the time of entering higher education compared to non-displaced students(χ2(2)=12,942, p=.002); student-workers compared to full-time students (χ2(2)=23,984, p=.000) (Table 1). Note that no statistically significant differences were found in tobacco use according to sex or BMI of respondents.

In general, students showed favorable attitudes toward smoking, because the mean score 2.01 (SD=.924) showed that most respondents disagreed or strongly disagreed with the items of AtS. But, there were significant differences based on smoking status, smoker peers, SHS exposure, scientific area, sex and current residence, as shown in Table 2. Not surprisingly, smokers had more negative attitudes than those who had never smoked or stopped smoking (F(2, 824)=85.886, p=.000). And students who report that most or almost all friends are smokers and are mostly or almost always exposed to SHS exhibit more negative attitudes than students with no, few or some smoker friends (F(4, 811)=15.581, p=.000) and are never or rarely exposed to SHS (F(4, 814)=7.298, p=.000), respectively.

Engineering sciences respondents showed more negative attitudes, while those in social and human sciences showed the most favorable attitudes (F(3, 823)=6.791, p=.000). Girl respondents and students who lived at home had more favorable attitudes than boy respondents (t(825)=4.478, p=.000) and students who had moved away from home (t(813)=2.516, p=.012), respectively.

There were no significant differences by year of frequency, age, being in a relationship, IBM and professional situation.

Table 3 shows respondents’ knowledge of tobacco use. The mean score was 3.11±1.26 (out of 6) correct answers. There were no significant differences in knowledge scores because of smoker peers, SHS exposure, year of frequency, scientific area, age, being in a relationship, current residence or IBM. However, one-way ANOVA showed significant differences between TUK and smoking status. Respondents who had ceased smoking had a significantly higher mean knowledge score than those who still smoked and those who had never smoked (F(2, 829)=3.559, p=.029). Furthermore, t-test showed significant differences between TUK and sex of respondents and current residence. That is means girls and displaced students had a lower level of knowledge than boys (t(830)=2.336, p=.020) and students living at home (t(822)=−2.334, p=.020), respectively.

For the final model, we kept the variables that had a statistically significant effect on the logit probability of the smoking status: AtS (G2(2)=101.993, p=.000); Smoker peers (G2(8)=20.492, p=.009); SHS exposure (G2(8)=21.693, p=.006); In relationship (G2(2)=16.247, p=.000); Professional situation (G2(2)=14.410, p=.001). The adjusted model was statistically significant (G2(22)=257.946, p=.000) and correctly predicted the status smoking 72.5% of the time (93.9% for non smokers).

Table 4 summarizes the results of the multinomial logistic regression for smoking status. The model made it possible to predict that: students who were less exposed to SHS were less likely to be smokers compared to non-smokers; students who had fewer smoking friends were less likely to be smokers and ex-smokers compared to non-smokers; being in a relationship compared to not being in a relationship increases the chances of being a former smoker in relation to a non-smoker; being a full-time student compared to student workers reduces the chances of being a former smoker by 67.7% and of being a smoker by 58.4%; students scoring higher on AtS were more likely to be a former smoker or current smoker than non-smoker.

This study examined current smoking status, smoking history, cessation attempts and knowledge of tobacco use and attitudes toward smoking among students at one university in Portugal. Our data showed that 1 in 5 university students are smokers but that the prevalence of smoking in this university was lower than identified in other national23–26 and international studies (Spain10,27,28; Greece29; Iran30; Australia31; Jordan32; Palestine33; Liban34; United Kingdom35; Italy36; Czech Republic37; Serbia38; Brazil39; Saudi Arabian40; Belgium41; Chile42; Poland43; Turkey44). Similarly, we found a high proportion of students who were former smokers (11.2%), considering that they are young adults and despite the fact that most current smokers (61.4%) started smoking before entering higher education.10,45 This could be explained because before leaving home, there had been an aspect of parental control which was lost when entering higher education.46 The prevalence of smoking increased with permanence in higher education, which corroborates scientific literature.3,32,33,47,48 Consequently, older students included more smokers than younger students (13.6% versus 6.4%) and there were also statistically significant differences depending on the number of cigarettes consumed per day or per week. That is, older smokers smoke, on average, 12.10 (±7.340) cigarettes per week or 8.84 (±4.969) cigarettes per day, while younger smokers smoke, on average, 5.09 (±3.048) cigarettes per week or 5.09 (±2.468) cigarettes per day (t=−2.909, p=.009 and t=−2.445, p=0.17, respectively).

A national study49 showed that 37.2% of young people aged between 15 and 34 years old had used tobacco in the previous 30 days, meaning this prevalence was higher than verified in this study.

Unlike other national24,26 and international studies30,33,42,47,48,50–53 in which male students smoked more than female students, our study did not find any differences in smoking between the sex of respondents. This is consistent with a study conducted in 2004 at the same university.54

Most smokers reported never having attempted to quit smoking (59.7%), a very high percentage compared to other studies.10 This indicates, efforts should be made to understand why students are reluctant to quit smoking.

The prevalence of SHS in closed public spaces is high, due to the fact that more than half of the students (65.6%) had been exposed to SHS in the previous week. Despite prohibitive smoking policies in enclosed public places, college students had a high level of exposure to SHS.48,55,56 Furthermore, we found that smokers are more exposed to SHS than non-smokers.57

Most respondents (52.2%) reported that at least some of their friends were smokers, consistent with the literature, not forgetting that having smoking friends seems to influence tobacco use among university students.30,38,40,45,50,51,53,58–60

Regarding attitudes toward tobacco, we found that most students showed favorable attitudes toward smoking31 but, as in other studies,31,33 non-smokers demonstrated higher scores on positive attitudes than smokers and former smokers. The influence of smoking friends, exposure to SHS and being enrolled in engineering science courses seems to reduce positive attitudes toward tobacco use.

In terms of knowledge about tobacco use, several studies have shown that smokers have low levels of knowledge,61 suggesting that increasing knowledge about the effects of smoking would decrease smoking rates during academic courses.47 In our study, there was a moderate level of knowledge about tobacco,30,33,48 higher in students who had already quit smoking.31,32

Unlike other research findings, where girls were better informed than boys31,32,62 or where there were no differences,30 we found that boys were better informed than girls.33

It is also important to highlight the sociodemographic variable “current residence”, because students who had left their familial home smoked more, knew less about the harmful consequences of smoking and showed more negative attitudes compared to students who had not changed their residence after entering higher education.

Finally, the limitation of this study should be noted. Restricting the study to a single university limits generalizability to the total population of university students in Portugal.

No data were collected concerning the smoking habits of parents/households and, although scientific studies state that children of smoking parents have a greater tendency to be smokers, this type of question did not fit within the scope of our study. One final limitation should be considered. Since the data were collected in 2018–2019, the use of e-cigarettes or “heat-not-burn” tobacco products are at least worth mentioning, especially looking at young people.

This study suggests an urgent need for higher education institutions to implement socio-educational programs to discourage tobacco use among university students. In addition, it is also highly recommended that during academic training students develop personal and social skills to deal with the tobacco epidemic.

The authors have no conflicts of interest to declare.