Methods

Participants

In the 2018/2019 academic year, 5447 students were enrolled in the 1^st and 3^rd year of higher education and integrated master’s degree. Based on the data of the population, a stratified probabilistic sampling was carried out among univertsity students, in accordance with the year of study and the scientific area of study.

This is a cross-sectional observational study with a representative sample of university students (n = 840) of a univeristy in Portugal. The sample is composed of 464 students attending the 1^st year (55,2%) and 302 students (36,0%) are enrolled in enrgineering courses. The majority of enquired students was female (55.4%), was not currently in a romantic affair (58.3%), had changed their residence after entering Higher Education (64.9%), was a full-time student (88.8%) and had an BMI corresponding to the normal weight (73.1%). The average age of the sample is 20.78 (DP = 4.221), varying between 18 and 54, with only 3% of students being 30 years-old or older.

Instruments

Nowadays there are multiple scientific instruments duely validated in order to efficiently monitor the prevalence of risky sexual behaviors in university students and various scales developed to analyze the knowledge and attitudes about sexuality such as those presented in the Handbook of sexuality-related measures (^{Davis et al., 1998}). Therefore, in order to satisfy the purposes of the present study, we developed a questionnaire, in accordance with the procedures defined by ^{Bowling (1998}): building on the scale (1^st stage); validation of content (2^nd stage); psychometric validation (3^rd stage).

A preliminary version of the instrument was reviewed by five researchers in the fields of health and science education and 12 university students to validate its content. A pre-test was then performed with a sample of 32 students to test for comprehension and difficulty, which was not included in the final sample.

In the questionnaire we included socio-demographic variables (sex, age, scientific area, academic year, weight and height (to calculate the BMI), romantic status, professional situation and current residence) and specific questions related to:

Procedure and statistical analysis

All students belonging to the courses selected for the stratitied sampling were personally invited to take part in the study. At the end of a class, randomly selected, belonging to the respective courses, the objectives of the study were presented and after informed consent the students filled a questionnaire in paper format, in the context of the classroom. The answer rate was of 96.2% (95% CI 94.8-97.6), 33 questionnaires were excluded because they hadn’t been answerered or were incorrectly filled. This way, a total of 873 university students were invited.

This study was approved by the Ethics Committee for Research in Social and Human Sciences, of the University of Minho Ethics Council, under the protocol CEICSH 009/2019.

The data was analyzed by means of statistical software, namely IBM SPSS Statistics for Windows, version 25.0 (IBM Corp., Armonk, NY, USA). The results show the analysis of the frequencies and the contingency tables, as well as the execution of tests of statistical inferece, such as Pearson’s correlation, the Chi-Square test and independence tests such as t-student and single-directional unity (ANOVA). The linear generalized model was developed including only the variables which showed to have a statistically significant effect in the probability of resulting in risky sexual behaviors. A significance level of 0.05 was given.

Risky sexual behaviors were defined as: having had sexual relations at 16 years of age or less; having two or more sexual partners in the last 12 months; failing to always use condom during the sexual relation in the last 12 months; having sexual relations after the ingestion of alcohol in the last 12 months.

For the knowledge scale, the number of correct answers was added in order to provide us with a general figure for knowledge, which means that the higher the number in the scale, the higher the level of knowledge is.

The scale regarding perceptions of STIs made up of 3 items varies between 3 and 15 and the scale of perceptions of condom use is composed of 2 items, varying between 2 and 10, with the highest figure attributed to the sum of the items meaning the more negative the attitudes of university students will be regarding STIs and the use of condom.

Results

Results show that the majority of university students (65.6%) was sexually active, with a higher prevalence of sexually active students frequenting the 3^rd year (73.4%) when compared to students in the 1^st year (59.3%) (χ²(1)=18.393,p=.000). Moreover, we noted that at the moment of data-gathering the students that had a romantic relationship (91.9%) were more sexually active and students who had no romantic partner (46.3%) (χ²(1)=186.081,p=.000). We also found that students who worked and studied (87.1%) were also more sexually active than full-time students (62.8%) (χ²(1)=21-603, p=.000). (Table 1).

The students who initiated their sexual life at an early age corresponded to 39.5% of the sexually active students. No statistically significant differences were found between an early start of sexual relations and the socio-demographic characteristics of the enquired students.

Sexual relations with multiple partners was the risky sexual behavior with the lesser prevalence among sexually active students, even though 18.7% of university students had 2 or more sexual partners in the last 12 months. The students who were not in a romantic relationship (31.1%) and those who had changed residence after entering higher education (25.7%) were the ones who presented a higher proportion of multiple partners in comparison with the students who were in a romantic relationship (χ²(1)=38.265, p=.000) and were physically dislocated (χ²(1)=10.623, p=.001), respetively.

Table 1 Prevalence and Chi-Square test of risky sexual behaviors depending on socio-demographic characteristics 

Note: ^* p < .05; ^** p < .01; ^*** p < .001; ns - not significant

The unprotected sexual relations were carried out by the majority of sexually active students (60.1%). And about a third of the sexually active students revealed having already had sexual relations after the consumption of alcohol in the last 12 months (32.3%). We can conclude through the observation of table 1 that 3^rd year students (37.8%) acknowledged this type of practice more frequently when compared with the 1^st year students (χ²(1)=7.689, p=.006). Similarly, statistically significant differences were identified between sexual relations under the effect of alcohol and the variables sex, romantic partnership and residence.

Correlations between the different risky sexual behaviors demonstrated the following: early sexual relations with multiple partners (r _sp =.111, p<.01), unprotected sexual relations (r _sp =.103, p<.05) and sexual relations under the effect of alcohol (r _sp =.124, p<.01). Moreover, sexual relatons with multiple partners are positively correlated with sexual relations under the effect of alcohol (r _sp =.307, p<.01). Generally, university students displayed, on average, 1.50±1.071 risky sexual behaviors, with 57.1% of sexually active students referring 2 or more behaviors of this type and 17.8% of students referring engaging in no risky sexual behavior.

Table 2 Average, pattern deviation and independence tests (ANOVA and t-student) for the socio-demographic characteristiscs and the risky sexual behaviors depending on the knowledge about STIs and the perception of STIs and condom use 

Note: ^* p < .05; ^** p < .01; ^*** p < .001; ns - not significant

The level of knowledge about sexuaility was of 3.99±1.194 correct answers, varying between 0 and 6 correct answers, allowing us to conclude that university students display a moderate level of knowledge.

An analysis of table 2 indicated the existence of statistically significant difference between the level of knowledge and the scientific area of study, the professional situation and the practice of sexual relations under the effect of alcohol. In this sense, students attending courses in the area of exact and natural sciences, working-students and students who had sexual relations after ingesting alcohol posess a level of knowledge higher than that of students belonging to engineering sciences courses (Z(3,824)=3.590, p=.013), full-time students (t(818)=-2.046, p=.041) and those who reported not having engaged in that type of behavior (t(539)=-2.159, p=.031), respectively.

Table 3 Adjusted odds ratio (OR) and confidence intervals of 95 (CI) of the linear generalized model for predicting risky sexual behaviors 

Note: ª Reference category; OR: odds ratio; 95 CI: 95 confidence interval

The perceptions of university students regarding STIs and the use of condom showed themselves to be moderate, with the average of scales being 6.54±2.162 and 5.44±1.800, respetively.

Regarding the perceptions of STIs, male students displayed a more negative level of perception regarding STIs when compared to female students (t(823)=2.156, p=.031) (Table 2). In what concerns the scale of perceptions towards use of condom, students who were attending the 3^rd academic year displayed a higher average when compared to students in the 1^st year (t(829)=-2.844, p=.005). This means that the first possess more unfavorable perceptions towards the use of condom. In the same sense, male students and students who were in a romantic partnerhsip, a higher average is shown in the scale of perceptions of when compared to female students (t(829)=4.676, p=.000) and to the students who were not in a romantic partnership (t(822)=3.850, p=.000).

Regarding sexual behaviors, sexually active students who, in the last 12 months, had unprotected sexual relations and engaged in sex after consuming alcoholic beverages displayed more unfavorable perceptions towards the use of condom when compared to students who never had sexual relations (t(829)=7.171, p=.000), those who did not have unprotected sexual relations (t(539)=-6.767, p=.000) nor under the effect of alcohol (t(540)=-2.203, p=.028).

In the model, the variables which had a statistically significant effect in the model were: Perceptions regarding the use of condom (G ²(1)=30.541, p=.000); being in a romantic partnership (G ²(1)=6.135, p=.013) and current residence (G (1)=7.102, p=.008) (Table 3).

The model accurately predicted that university students who had a romantic partnership and hadn’t changed their residence after entering higher education had a lesser propensity towards engaging in risky sexual behaviors. Moreover, the more negative the perceptions regarding condom use, the greater the probability will be of students engaging in risky sexual behaviors.

Discussion

This study evaluated the knowledge, perceptions and risky sexual behaviors of students attending a university in Northern Portugal, demonstrating, based on the results and the discussion presented later on, the need for higher education institutions to develop socio-educational programs with education in sexuality content.

The results of this study corroborate the results of the national study done about the same topic (^{Santos et al., 2016}) and, in the same way as in international studies, the majority of university students was sexually active (^{Fernández-Silva & Sánchez-Martínez, 2018}; ^{Kebede et al., 2018}; ^{Leon-Larios & Macías-Seda, 2017}; ^{Mavhandu-Mudzusi & Asgedom, 2016}; ^{Spindola et al., 2019}; ^{Villafañe-Ferrer & González-Navarro, 2017})

Generally, potentially risky sexual behaviors were reported by 82.2% of sexually active university students. This high level of prevalence of risky sexual behaviors is corroborated by the findings of recent studies carried out with students of Higher Education (^{Sherab et al., 2019}; ^{Yang et al., 2019}), although a sistematic review of the literature with Ethiopian university students estimates a prevalence of 41.62% (^{Amare et al., 2019}).

The majority of enquired individuals had started their sexual lives at 17 years of age or less, as in other investigations (^{Spindola et al., 2019}; ^{Villafañe-Ferrer & González-Navarro, 2017}), showing that entering higher education tends to co-occur with the age of the first sexual experience or occurs during the academic path (^{Mavhandu-Mudzusi & Asgedom, 2016}). On another note, it is necessary to point out that a minimum of about 40% university students reported already having had sexual relations before enrolling in university (^{Magalhães et al., 2016}).

Although some studies (^{Mukherjee et al., 2019}) demonstrate a tendency for boys to become sexually active at an earlier age than girls, the results obtained did not produce evidence of these differences (^{Mavhandu-Mudzusi & Asgedom, 2016}; ^{SaraçoĞlu et al., 2014}).

The inconsistent use of condom during sexual relations was the analyzed risky behavior with the highest prevalence among university students. The ones enquired in this study, just like the univeristy students recently enquired in the context and scope of studies carried out by others, reported an inconsistent use of condom in the majority of sexual relations (^{Castro, 2016}; ^{Fairfortune et al., 2020}; ^{Folasayo et al., 2017}; ^{Golbasi & Kelleci, 2011}; ^{Hoffman et al., 2017}; ^{Kebede et al., 2018}; ^{Magalhães et al., 2016}; ^{Makgale & Plattner, 2017}; ^{Mavhandu-Mudzusi & Asgedom, 2016}; ^{Pastor & Rojas-Murcia, 2019}; ^{Perera & Abeysena, 2018}; ^{Santos et al., 2016}; ^{SaraçoĞlu et al., 2014}; ^{Xu et al., 2019}).

Sexual relations under the effect of alcohol in the last year were reported by 32.3% of university students, which, although comprising what can easily be considered a high prevalence, is lower than the majority of analyzed studies (18.9% - ^{Makgale & Plattner, 2017}; 25.20% - ^{Mavhandu-Mudzusi & Asgedom, 2016}; 27.1% - ^{Gebresllasie et al., 2017}; 39.9% - ^{Magalhães et al., 2016}; 33.0% - ^{Santos et al., 2016}; 55% - ^{Villafañe-Ferrer & González-Navarro, 2017}; 56.3% - Á. ^{Castro, 2016}; 56.8% - ^{Leon-Larios & Macías-Seda, 2017}; 64.2% - ^{Golbasi & Kelleci, 2011}).

Sexual relations with multiple partners is one of the most important factors associated to STIs (^{Chen et al., 2016}), however, in the last 12 months, this risky behavior was reported by 18.7% of university students. Analogous results were shown in international studies (^{Mavhandu-Mudzusi & Asgedom, 2016}; ^{Yi et al., 2018}), in which, for example, 12.9% of students reported having had more than one sexual partner in the last six months (^{Magalhães et al., 2016}), 26.2% of the enquired ones reported having had two or more sexual partners in the three months prior to the study (^{Makgale & Plattner, 2017}) and the majority had already had more than one sexual partner in their lives (^{Spindola et al., 2019}). Other international studies showed higher percentages of sexual relations with multiple partners (59.3% - ^{Golbasi & Kelleci, 2011}; 64.4% - ^{Kebede et al., 2018}; 66.7 % - A. T. ^{Folasayo et al., 2017}). Although in the present study we did not specifiy the type of sexual partners, a national study showed that a considerable proportion of students reported having had sexual relations with occasional partners (32.0%) (^{Santos et al., 2016}).

Although globally the gender of the enquired individuals does not predict the risky sexual behaviors (^{Makgale & Plattner, 2017}), when they are analyzed separately, boys reported the practice of sexual relations after the ingestion of alcohol more frequently when compared to girls, and we found existing studies which provide similar data and evidence or results (^{Hoffman et al., 2017}). In the same sense, contrary to what was to be expected, due to the fact that it was verified in various studies (^{Amare et al., 2019}; ^{Gebresllasie et al., 2017}; Hoffman et al., 2017; ^{Leon-Larios & Macías-Seda, 2017}; ^{Magalhães et al., 2016}; ^{Mavhandu-Mudzusi & Asgedom, 2016}; ^{Perera & Abeysena, 2018}; ^{Sherab et al., 2019}; ^{Yi et al., 2018}), no statistically significant differences were identified in the use of condom and the number of sexual partners between girls and boys.

The current scientific investigations revealed other correlations, namely between unprotected sexual relations and sexual relations under the effect of alcohol (^{Castro, 2016}; ^{Magalhães et al., 2016}) and between unprotected sexual relations and having multiple sexual partners (^{Makgale & Plattner, 2017}; ^{Sherab et al., 2019}), although the data arising from the present study does not corroborate these findings (^{Leon-Larios & Macías-Seda, 2017}; ^{Peltzer & Pengpid, 2015}).

Students attending Higher Education displayed a moderate level of knowledge of STIs (^{Folasayo et al., 2017}; ^{Santos et al., 2016}; ^{Sohbet & Geçici, 2014}; ^{Soleymani et al., 2015}; ^{Villafañe-Ferrer & González-Navarro, 2017}), showing a higher level of knowledge than the one registered in other studies (^{Chen et al., 2016}; ^{Dutt & Manjula, 2017}; ^{Elias et al., 2017}; ^{Evcili & Golbasi, 2017}; ^{Fernández Rouco et al., 2018}; ^{Mukherjee et al., 2019}; Soleymani et al., 2015). Even so, the results showed that a lot of that knowledge remains superficial and full of myths (^{Dutt & Manjula, 2017}; ^{Spindola et al., 2019}; ^{Xu et al., 2019}). Mukherjee and colleagues (2019) refer that the fact that the main source of information is the group of peers could in part account for high indexes of wrong information transmitted.

Previous studies had already demonstrated that women had better knowledge (^{Folasayo et al., 2017}; ^{Fonte et al., 2018}; ^{Leon-Larios & Macías-Seda, 2017}; ^{Oppong Asante, 2013}; ^{Santos et al., 2016}; ^{SaraçoĞlu et al., 2014}; ^{Sohbet & Geçici, 2014}), however, in the present study, we did not identify those differences, with the average of correct answers given by boys and girls being identical.

Unlike what was verified in other researches in this area (^{Evcili & Golbasi, 2017}; ^{Folasayo et al., 2017}; ^{Fonte et al., 2018}; ^{Oppong Asante, 2013}; ^{Sohbet & Geçici, 2014}; ^{Soleymani et al., 2015}), in which it was determined that the level of knowledge increases with age, this associated was not identified (^{Leon-Larios & Macías-Seda, 2017}).

Similarly to what was verified in the study by ^{Santos, Ferreira and Santos (2016}), the scientific area of study also influence the knowledge of students. In accordance wity these authors, those who have an education in the areas of life and health sciences displayed a higher level of knowledge. In the presente study, however, the students belonging to exact and natural sciences were the ones who displayed more knowledge about STIs.

The perceptions of STIs were moderate, as is consistently reported in the literature (^{Santos et al., 2016}; ^{Sohbet & Geçici, 2014}). In this context, international scientific research and investigations reported that youngsters, in a general way, consider it unlikely to contract STIs or HIV (^{Amare et al., 2019}; ^{Haile et al., 2017}), which means that they indicated a low level of perception of the susceptibility to risk. Therefore, the programs for prevention of STIs in the academic context must continue to address STIs and their ways of transmission (^{Haile et al., 2017}).

The average scores attributed to the perception of STIs among smale students was higher than those attributed to female ones (^{Evcili & Golbasi, 2017}; ^{Fernández-Silva & Sánchez-Martínez, 2018}; ^{Leon-Larios & Macías-Seda, 2017}), showing perceptions considered to be more unfavorable towards these infections.

The perceptions regarding the use of condom, although also moderate, with an obtained average which was only slightly higher than the average point of the scale, implies unfavorable perceptions. As can be verified in more recente investigations report that many university students believed that condoms reduced pleasure during the sexual relation(s) (^{Mukherjee et al., 2019}) and that the trust put on sexual partners and the stability of romantic relationships justified the absence of the use of condom (^{Pastor & Rojas-Murcia, 2019}).

In accordance with a systematic review of the existing literature carried out by ^{Kabwama and Berg-Beckhoff (2015}), inconsistent results can be observed regarding the correlation between knowledge about HIV and the perception of risk among university students. And the other studies which explicit a lesser degree of knowledge concluided that this fact was directly associated to a higher prevalence of risky sexual behaviors (^{Li et al., 2017}; ^{Perera & Abeysena, 2018}). This way, although knowledge is obviously essential for young people to make informed and responsible redicions, one cannot ignore their inability to understand the risk even when they possess a good or fair level of knowledge. And for that reason, in a general way, this study demonstrated that the level of knowledge is not directly correlated with the perceptions and risky behaviors (^{Dutt & Manjula, 2017}; ^{Xu et al., 2019}).

As happened in previous studies (^{Santos et al., 2016}; ^{Yang et al., 2019}), the perceptions regarding the use of condom were a prominent predictor of risky sexual behaviors. Moreover, it is necessary to consider and take into serious account the students who changed their residence at the time of enrolling in higher education and the students who are not in a romantic partnership or relationship, due to the fact that these groups of students have a higher probability of engaging in risky sexual behaviors.

Two limitations should be taken into account. Firstly, it should be noted that the data-gethering was carried out in only one university, and for that reason transversal data make it more difficult to generalize the data provided by other Portuguese universities. And, secondly, the data-gathering was carried out by resorting to the appliance of a self-report questionnaire and taking into account the sensible character of the topic, students may have over or under-reported their sexual behaviors in their accounts or answers.

Ther present study evaluated the knowledge and the perceptions of STIs and the risky sexual behaviors in the context of Higher Education. University students displayed a high prevalence of risky behaviors, such as unprotected sexual relations, those engaged upon under the effect of alcohol or with multiple sexual partners. The level of knowledge and the perceptions about STIs and the use of condom was moedrate, although we did not detect any correlation between these variables.

Among the predictors, or predictive factors, variables such as the perceptions regarding the use of condom, having a romantic relationship or the current residence were statistically significant and directly correlated with the inconsistency of condom use, multiple sexual partners, practicing sexual relations after the ingestion of alcohol and early sexual relations.

This study made it possible to further undertstand the factors that constitute an influence towards risky sexual behaviors among higher education students, and to conclude that understanding those factors is indeed of invaluable importance, not only for their predictive character, but also because the development of educational programs on sexuality may significantly improve when these interactions are subject to scientific study and known. In this sense, the data gathered in this investigation reinforces the need for higher education institutions - also due to the fact that the beginning of sexual activity usually occurs during the academic period - to incorporate debates on sexuality, to inluce in their curricular componente actual educational content dealing specifically with the subject of sexuality and the prevention of STIs, in order to enable youngsters and young university students in general to make conscious and responsible decisions.

Author contributions

Regina Alves: Project administration, Acquisition of financing, Data curation, Formal analysis, Investigation, Methodology, Visualization, Writing - original draft

José Precioso: Supervision, Writing - Review and editing, Methodology

Elisardo Becoña: Supervision, Writing - Review and editing

Funding

This work was supported by Foundation for Science and Technology --- FCT (Portuguese Ministry of Education and Science) (SFRH/BD/120758/2016).