Frequency and associated factors related to sexual addiction in medical students from 16 Latin American cities, 2016: a regional

Examples of addiction problems that have been reported in Background: growing populations are those related to sexual impulses and addictions. However, such studies have not been carried out in Latin America. The aim of this study was to characterize and identify possible associations of sexual addiction in medical students in Latin America. An analytical cross-sectional study was carried out among the Methods: university students of a medical school in 16 cities; students of medical schools were interviewed during the first semester of 2016. To define sexual addiction, the multi-cage cad-4 test was used, categorizing individuals as possibly or not a potential problem. Additionally, associations with several social and educational variables were obtained. In our study, 6% (221) of the 3691 respondents exhibited a Results: possible problem of sexual addiction; men had 95% more problems (95% confidence interval (95%CI): 21-214, p=0.006), for each year of age it increased by 9% (95%CI: 1-18%, p=0.034 ), those who had a partner were 67%  more likely to exhibit sexual addiction (95%CI: 1.34-2.08%, p <0.001) and those who professed a religion present 44% less frequency (95%CI: 20-60%, p: 0.001). When adjusted for marital status, having children, year of studies, and the university where the respondent studied were not associated. Although the percentage of students who had problems with Conclusion: sexual addiction is minimal, screening programs should be created to find students who suffer from these problems, to avoid the possible consequences that may arise. 1 2 3


Introduction
Throughout life, young people experience a series of changes, including biological, cognitive and psychosocial type changes, which are influenced by various factors, such as gender, social, cultural, political and economic environment, as well as the level of physical and psychosocial/cognitive maturity of the person 1 . Some studies have found that by submitting these young people to stressful environments can trigger them to be disinhibited, which leads them to adopt risky behaviors 1,2 . It is essential to know that one of the populations with the most considerable stress and interpersonal relationships is the students 3 . If early sexual initiation and the non-use of protective measures are added to this, it becomes an immediate health problem 4 , especially if these individuals have had an inadequate sexual education 5 .
Some studies have explored the sexual sphere of university students, for example, in Turkey, this topic was characterized, and other topics considered "taboo" for their culture were touched 6 . In Finland, thousands of university students were surveyed to determine continuity/frequency according to age group 7 , another one performed in India measured sexual knowledge and practices 8 . All of these studies highlight the importance of the student having an adequate orientation for the correct expression of their sexuality. In the future, this will help them to advise their patients, allowing them to establish better interactions and communication with them 9 . In Latin America, the studies performed in the sexual sphere try to link the sexual behavior with factors that could influence this behavior, finding only one study, performed in Colombia, that analyzes this relationship through the lifestyle of university students 10 . For all these reasons mentioned, the objective of the study was to characterize and find the associations of sexual addiction that medical students have in Latin America.

Study background
We performed an analytical cross-sectional study; this was based on the data that was generated after the application of surveys to medical students from 16 cities in Latin America. A multicenter study was conducted, because the minimum sample size required was 3506 medical students, this is important to find minimum differences of 1% of sexual impulses according to marital status (the smallest difference detected in the prior study); this was calculated with a potency of 80% and a confidence level of 95%. Calculation was performed using STATA 11.1 with the SAMPSI command for sample size calculation, We calculated the statistical power for the crossing of the variables; in almost all cases, the statistical power obtained was greater than 99%. However, in two of the crosses, the power was not reached (for marital status: 56% and for the type of the university: 25%); Therefore, the results obtained in these two cases must be taken as purely exploratory.

Study participants
Medicine students of both sexes who agreed to participate voluntarily between 18 and 26 years in the study were included, they were recruited by direct contact. Written informed consent was obtained, and 148 surveys that had no answers from the primary survey were excluded, from which the dependent variable was extracted.

Ethics
The study began in April 2016, after the Ethics Committee of "Hospital Nacional Docente Madre Niño San Bartolome" approved the research with the code CIEI-170023, and was completed in April 2017. The study was performed by capturing the necessary amount of medical schools to reach the minimum sample size; those in charge of surveying in each campus made the request for institutional permission.
We explained the purpose of the study and that the participation was voluntary and anonymous; The students were surveyed in environments where students had the necessary privacy conditions for the proper resolution of the survey.

Participant assessment
The primary variable was based on a previously validated questionnaire 11 . The CAD-4 Multi Cage Test used consist in the validation of eight kind of addiction, so we based our questionnaire in four questions related to sexual addiction: 1. Has your sexual activity prevented you from performing usual tasks in your life, such as work or family obligations?
2. Have your partners complained about your excessive sexual activity?
3. Have you ever considered that your sexual activity is excessive?
4. Have you ever tried unsuccessfully to moderate your sexual activity?
For this investigation, only the questions that measured sexual addiction were taken. The questionnaire is self-administered and is answer on a dichotomous scale (yes/no) and it is estimated that none or an affirmative answer indicates not an existence of that problem; two affirmative answers indicate a possible presence of that problem; three affirmative answers suggest a very probable presence of that problem; and answer affirmative at four suggests the sure existence of that problem 11 . The cut-off point was two definite answers, as an answer of zero or one indicated the absence of this problem. On the other hand, two or more positive answers indicated there was the possible existence of addiction. Frequencies and percentages were described. Also, this was disclosed according to the number of affirmative answers that all surveyed students had.

Statistical analysis
The dependent variable was crossed with the independent variables using the chi-square statistical test for crossing with sex, marital status, having a romantic partner currently, having children, and studying in a private university. The dependent variable was compared to the quantitative variables (age and year of studies) with the sum of Wilcoxon ranks. The software used to perform the analysis was STATA 14.0.
A 95% confidence level was used for bivariate and multivariate analysis, prevalence ratios (RP), 95% confidence intervals (95% CI) and p values (considering values p<0.05 as statistically significant), through the use of generalized linear models, with the Poisson family, the log link function, robust models and using the university as a cluster adjustment group. For the preparation of the multivariate model, all the variables in Table 1 were taken into account (both those that were statistically significant and those that were not, this being important in the context of a student with a possible addiction to sex). The variables that were significant in the bivariate analysis were maintained in the multivariate. Of those that were not significant, age became significant in the multivariate model (the others served as an adjustment).

Results
A total of 3691 students were surveyed in the medical schools, according to the answers of the cad-4 multi cage test, it was obtained that 87.3% (n=3222) and 6.7% (n=248) had no positive answers and had only one positive answer regarding sexual addiction, respectively. These individuals were classified as having no sexual addiction. In total, 3.1% (n=113), 1.9% (n=70), and 1.0% (n=38) had 2, 3, and 4 definite answers, respectively. These individuals were categorized with sexual addiction problems. Table 1 shows the characteristics of the students determined to exhibit potential sexual addiction problems. De-identified responses for each participant are available as Underlying data 12 . Figure 1 shows the percentage of sexual addiction according to the year of study of the surveyed students. Students residing in the capitals of Peru and Colombia had the highest frequency of sexual addiction problems, both with 11%, while those who studied in the cities of Tucumán, Argentina and Ciudad Bolivar, Venezuela both had 0% ( Figure 2).
In the multivariate analysis, an association was found between the highest frequency of sexual addiction and male sex On the other hand, those who described themselves as followers of a religion had a lower frequency of sexual addiction (PR: 0.56; 95% CI: 0.40-0.80; p-value: 0.001); all these variables were adjusted for marital status, having children, the year of studies and using the university where the surveyed studied as a cluster (adjustment variable) ( Table 2). Figure 3 shows the percentages of sex addiction according to the statistically significant categorical variables in the multivariate model. Figure 4 shows the percentages of specific answers to the four questions of the sexual addiction test in medical students in Latin America.

Discussion
Our study found that few students had a sexual addiction. Our findings differ from a study performed at a private university in the US, which showed that 26% of the students had symptoms of risk for sexual addiction 13 ; this could be due to population differences. Broader population studies have obtained similar percentages to ours. A nationwide study conducted in Sweden found that 10% of the population is considered "hypersexual" 14 .
Regional and local surveys have been performed in the US, which suggests that approximately 5% of the general population can meet the criteria of a compulsive sexual disorder 15 . Another factor that led to finding different results could be the test that was applied, because some research is not based on validated instruments. These results should be confirmed, since the objective of the study was not to give prevalence, because a random sampling was not performed out in each campus, so each institution should rate this situation of its students.
Tt was found that men had higher frequencies of sexual addiction; this is similar to what was found in studies conducted     in Mexico, where more male university students than female had a sex addiction and risky sexual behavior, such as practicing sex without condoms 16 . Therefore, counseling activities should focus on them so that they can receive recommendations for safe sex practices.
With regard to age, the frequency of sexual addiction increased at an older age; these results are consistent with studies performed in Sweden and the United States of America; in which it was determined that young adults are more likely to suffer from "hypersexuality" and to show behaviors related to compulsive sexuality 14,17 . However, it is essential to highlight that the relationship between younger age and hypersexuality has not always shown an association, being demonstrated by a study that used a virtual platform 18 . Therefore, it is necessary to characterize the age groups that present more significant problems of addiction in each reality, to focus on them the programs of prevention of sexually transmitted diseases.
It was also found that people who have a sexual partner had a higher frequency of sexual addiction problems; this is similar to what was found in an investigation in Mexico 19 . Those who professed to have a religion had lower frequencies of sexual addiction; this could be due to the less liberal life they present and the religious principles they have 20 .

Limitations
Our study had the limitation that there was no random sampling, which generated a selection bias. However, this is the first report of this problem in the student population, and because of the large number of students that were achieved, this generates results that are pertinent to the subject addressed.

Conclusion
It is concluded that sexual addiction problems occurred in a small proportion of Latin American medical students. These addiction problems were positively associated with male sex, age, and having a romantic partner, being negatively associated with professing a religion. This project contains basic demographic information about each participant, alongside their answers to each of the questions asked.

Data availability
Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).